Datasets:

medarc

/

clinical_pile_v1_minhash_deduped

like 1

post-synthetic_modification_of_covalent_organic_frameworks_via_in_situ_polymerization_of_aniline_for

## Abstract: Covalent organic frameworks (COFs) with layered architecture with open nanochannels and high specific surface areas are promising candidates for energy storage. However, the low electrical conductivity of two-dimensional COFs often limits their scope in energy storage applications. The conductivity of COFs can be enhanced through post-synthetic modification with conducting polymers. Herein, we developed polyaniline (PANI) modified triazine-based COFs via in situ polymerization of aniline with the porous frameworks. The composite materials showed high conductivity of 1.4-1.9 x 10 -2 S cm -1 at room temperature with 10-fold enhancement in specific capacitance than the pristine frameworks. The fabricated supercapacitor exhibited a high energy density of 24.4 W h kg -1 and a power density of 200 W kg -1 at 0.5 A g -1 current density. Moreover, the device fabricated using the conducting polymer-triazine COF composite can light up a green lightemitting diode for 1 min after being charged for 10 s. ## Introduction: Porous organic materials owing to the low skeleton density, high specific surface area, tunable pore size, and robust hydrothermal stability due to covalent bonds have attracted a great deal of attention for task-specific applications ranging from gas adsorption, catalysis, molecular separation, water purification, sensing to electrical energy storage. In the context of affordable and sustainable energy sources, supercapacitors have emerged as promising energy storage systems in the carbon-neutral economy due to the high-power density, fast charge/discharge capability, high coulombic efficiency, and long cycling life. 5,6 The mechanism for storing charges in a supercapacitor relies on two distinct processes: nonfaradaic double-layer formation via the electrostatic adsorption of ions on the electrolyte/electrode interface (electric double-layer capacitance, EDLC) and faradaic redox reactions at the surface of the electrode (pseudocapacitance, PC). 7 Carbon based amorphous polymeric porous materials have been studied extensively as supercapacitive electrode materials due to their large specific surface area and relatively high packing density. 8 However, the lack of control over the pore-size distribution and pore structure often leads to the sluggish ion-diffusion resulting in undesirable capacitance loss during the fast charging-discharging. 9 Owing to the controlled pore size distribution and special pore functionality, covalent organic frameworks (COFs), an emerging class of crystalline porous organic materials obtained through dynamic covalent chemistry, 10 have been explored for supercapacitor applications. 11 The layered architecture of 2D COFs possesses 1D channels that facilitate ion diffusion and mass transport. 12 However, the strong π-π stacking of COF layers leads to inaccessibility of redox-active sites deeply buried inside the pore channels, limiting the charge storage capacity. 13,14 On the other hand, the electron localization on the heteroatom linkages and defects in the framework arising due to insufficient reversible condensation interrupt the extended π-electron conjugation. 15 Consequently, it results in low electrical conductivity of COFs retarding practical usefulness in energy storage applications. 15 The electrical properties of COFs can be improved by doping conducting polymers into the porous frameworks. Dichtel and co-workers first showed the strategy by introducing polyethylenedioxythiophene (PEDOT) via electropolymerization within a redox-active anthraquinonebased COF (DAAQ-TFP COF) to achieve high volumetric energy storage capacity with fast charging rates. 16 Awaga and coworkers adopted in situ solid-state polymerization of PEDOT inside the nanochannels of anthraquinone-based COF with the enhancement of electrical conductivity and high specific capacitance. 17 Later on, Mai and co-workers grew the COF on the surface of commercial polyaniline. 18 In this strategy, the coating of COF on conducting polyaniline hides the redox-active sites that diminish the pseudocapacitance contribution. The electronic charge density on heteroatoms in the frameworks with a high specific surface area facilitates the ion adsorption to enhance the double-layer capacitance. 19,20 Besides, the electroactive building units, like N-rich triazine cores enhance the relative permittivity of the electrode materials and also contribute to the pseudocapacitance. 21 Herein, we synthesized two triazine-based covalent organic frameworks via solvothermal Schiff-base polycondensation by combining C 2 and C 3 -symmetric (triazine) building units, TCOF-1, and connecting two C 3 -symmetric triazine units, TCOF-2 (Fig. 1a). We modified TCOFs via in situ polymerizations of aniline using a chemical oxidation method to improve the electrical conductivity resulting in polyaniline modified composites, PANI@TCOF-1, and PANI@TCOF-2. The composite materials showed high electrical conductivity, which further improved the specific capacitance with cyclic stability. We fabricated asymmetric supercapacitor devices derived from PANI@TCOF-2, and the lighting up of a green light-emitting diode (LED) was demonstrated. ## Results and discussion: Triazine-based covalent frameworks (TCOFs) are the class of porous organic networks that can be synthesized via ionothermal trimerization of aromatic polynitriles in the presence of molten ZnCl 2 . 22a Most of the CTFs obtained under the harsh ionothermal conditions are amorphous. 22b We employed triazine moiety within the monomeric building unit to get highly ordered crystalline frameworks with a large specific surface area. The triazine-based covalent organic frameworks, TCOF-1, and TCOF-2, were fabricated via Schiffbase polycondensation reactions following reported procedures (Fig. 1a). 23,24 TCOF-1 was synthesized by a solvothermal reaction of 2,4,6-tris(4-formylphenoxy)-1,3,5triazine (TFPOT, C 3 ) and p-phenylenediamine (C 2 ), resulting in a pale-yellow solid with 70% yield. Similarly, TCOF-2 was synthesized using TFPOT (C 3 ) and 2,4,6-tris(4-aminophenyl)-1,3,5-triazine (TAPT, C 3 ′ ) to afford a yellow solid with 75% yield. The post-synthetic modification of TCOFs was carried out through in situ polymerization of aniline via chemical oxidation method within the porous frameworks (Fig. 1d). Fourier transform infrared (FTIR) spectroscopy was used to confirm the formation of TCOFs and PANI@TCOFs. The disappearance of the C=O (1704 cm -1 ) and N-H (3460-3210 cm -1 ) stretching bands indicates the consumption of aldehyde and amino groups of the monomers. Whereas, the appearance of the C=N stretching band at 1622-1626 cm -1 confirmed imine bond formation in TCOFs via condensation reaction. The FTIR spectra of polyaniline-TCOF composites (PANI@TCOFs) showed characteristic benzenoid-quinonoid nitrogen vibration at 1564 and 2,4,6-tris(4-formylphenoxy)-1,3,5triazine (TCOF-1) and 2,4,6-tris(4-aminophenyl)-1,3,5-triazine and 2,4,6-tris(4-formylphenoxy)-1,3,5-triazine (TCOF-2). PXRD patterns of (b) TCOF-1 and (c) TCOF-2: comparison between the experimental pattern (blue), Pawley refined profile (red), the refinement differences (grey), and the simulated pattern (green) for eclipsed AA stacking mode of TCOF-1 and slipped AA stacking mode of TCOF-2; space-filling models representing the stacking along the c-axis with the layer distances are shown. (d) Schematic representation depicting in situ polymerization of aniline with TCOFs. (e) FT-IR spectra of TCOFs and PANI@TCOFs. (f) Pore size distribution plots of TCOFs and PANI@TCOFs obtained through the analysis of respective adsorption isotherms employing the non-local density functional theory (NLDFT) method. cm -1 and aromatic amine peak at 1296 cm -1 (Fig. 1e). The blue shift of characteristic bands of polyaniline (8-12 cm -1 ) in composite materials ascertains the inclusion of polyaniline within the TCOFs. 25,26 The TCOFs were characterized by powder X-ray diffraction (PXRD) analysis. The PXRD pattern of TCOF-1 showed an intense diffraction peak at 2.68° attributed to the (100) plane (Fig. 1a). The other diffraction peaks appearing at 4.69°, 5.43°, 7.21°, and 9.46° correspond to the (110), ( 200), (210), and ( 220) facets, respectively. The PXRD profile of TCOF-2 exhibited a prominent peak at 3.98° and relatively weak peaks at 6.87°, 7.96°, and 10.56°, which were indexed to (100), ( 110), ( 200), and (210) facets, respectively (Fig. 1c). The experimental PXRD pattern accorded well with the simulated pattern based on the AA stacking for TCOF-1 and slipped AA stacking for TCOF-2 (Fig. 1b corresponding to (001) plane were observed at 2θ = 16.6° and 25.8° for TCOF-1 and TCOF-2, respectively (Fig. 1b, 1c). The PXRD data and the structural analysis of TCOF-1 and TCOF-2 are consistent with the earlier reports. 23,24 The suppression of the high intensity peaks at low 2θ in PXRD suggests a lowering of crystallinity in PANI@TCOFs. The surface area and porosity of TCOFs and PANI@TCOFs were estimated by the nitrogen adsorption−desorption isotherms at 77 K. The sorption curves of both TCOFs showed type-IV isotherms. The TCOF-1 sorption curves having the H1 hysteresis loop indicate the cylindrical-like pore channels. 27 Whereas, the H4 hysteresis loop of TCOF-2 suggests the narrow slit-shaped pores. 27a The Brunauer−Emmett−Teller (BET) specific surface area of TCOF-1 and TCOF-2 were found to be 1756 and 1110 m 2 g −1 , respectively. The pore volumes were 1.64 cm 3 g −1 for TCOF-1 and 0.85 cm 3 g −1 for TCOF-2, estimated from the nitrogen adsorption amount at P/P 0 = 0.95. The uniform pore size distribution (2.6 nm for TCOF-1; 1.4 and 3.9 nm for TCOF-2) was obtained based on the non-local density functional theory (NLDFT) method. The lowering of BET specific surface area and the changes in the pore size distribution of PANI@TCOFs infer the successful integration of PANI within the porous framework of TCOFs (Fig. 1f). The field emission scanning electron microscopy (FESEM) images depict the aggregated-particles like morphology for TCOF-1 and fused flake-like morphology for TCOF-2 (Fig. 2a, 2b). A distinctly different morphology was noticeable for PANI@TCOFs compared to the pristine TCOFs (Fig. 2c, 2d). PANI@TCOF-1 and PANI@TCOF-2 showed the well-distributed intricate network structure attributing to polyaniline loading with the porous framework. PANI@TCOF-2 showed whisker-like morphology of polyaniline that grew over the framework of TCOF-2 resulting in faster ion diffusion through the nanochannels. 25 The transmission electron microscopy (TEM) images of TCOFs and PANI@TCOFs further revealed the porous nature of the frameworks. The heteroatom rich porous framework structure of TCOFs and PANI@TCOFs motivated us to explore the electrochemical performance of TCOFs and PANI@TCOFs as active electrode materials for supercapacitors. The electrochemical measurements were carried out using a typical three-electrode electrolytic cell with 1 M H 2 SO 4 as an aqueous electrolyte. TCOFs and PANI@TCOFs were coated on platinum-foil and used as the working electrode. A platinum wire and saturated calomel electrode (SCE) were employed as the counter and reference electrodes, respectively. The electrochemical performances of the electrodes were investigated by cyclic voltammetry (CV), galvanostatic charge-discharge (GCD) measurements and electrochemical impedance spectroscopy (EIS). The cyclic voltammetry was carried out at different scan rates, and specific capacitance from the cyclic voltammograms (CVs) was calculated using the following equation. ## 𝐶 𝑠 = ∫ 𝐼 . 𝑑𝐸 2𝜈𝑚𝐸 (1) C s , I, ν, m, and E denote the specific capacitance, current, scan rate, mass of active material deposited on the electrodes, and the potential applied, respectively. Fig. 3a and 3c depict the cyclic voltammograms of PANI@TCOF-1 and PANI@TCOF-2, respectively, recorded at different scan rates from 10 to 100 mV s −1 in the potential window from 0 to 0.8 V. The CV curves of the composite materials show a rectangular shape featuring redox peaks. As revealed from CVs, the capacitive response is due to the combined effect of both electric double-layer capacitance and pseudocapacitance. Here, the pseudocapacitance contribution comes from the reversible redox activity of polyaniline. The voltammograms show two sets of distinct redox peaks. A redox couple between 0.1 and 0.35 V vs. SCE is associated with the conversion of the fully reduced leucoemeraldine base to the partially oxidized emeraldine (Fig. 3c). The redox peaks occurring between 0.35 and 0.6 V vs. SCE are due to the conversion of emeraldine to the oxidized pernigraniline form of PANI. 28 The highest specific capacitance of PANI@TCOF-1 and PANI@TCOF-2 were found to be 156 and 258 F g -1 at 1 mV s -1 scan rate, respectively. The galvanostatic charge-discharge (GCD) experiments were carried out at different current density for further investigations. The specific capacitance was calculated from the GCD plots using the following equation. ## 𝐶 𝑠 = 𝐼 . 𝑡 2𝐸 (2) I, t, and E represent the applied current density, time taken for the charge/discharge process, and the potential window, respectively. The charge-discharge curves of PANI@TCOF-1 and PANI@TCOF-2 at varying current densities exhibited typical triangular shapes (Fig. 3b and 3d). The asymmetric nature of the GCD curves also suggests the contribution from pseudocapacitance along with EDLC to the total specific capacitance value. 21c The specific capacitance of PANI@TCOF-1 and PANI@TCOF-2 from GCD curves were calculated to be 154 and 275 F g -1 at 0.5 A g -1 current density. PANI@TCOFs exhibited contrastingly higher specific capacitance compared to pristine TCOFs (10-20 F g -1 , Fig. 1e). Despite of high specific surface area and narrow pore size distribution, the low capacitance values of TCOFs could be due to poor electrical conductivity. 15 We measured the electrical conductivity by four-probe current-voltage (I-V) method. As anticipated, the electrical conductivity of TCOFs was very low, below the acceptable range of detection. The electrical conductivities of PANI@TCOF-1 and PANI@TCOF-2 were estimated as 1.4 x 10 -2 and 1.9 x 10 -2 S cm −1 , respectively, at room temperature and increased linearly with temperature (Fig. 1f). The significant enhancement of conductivity of PANI@TCOFs compared to pristine TCOFs is attributed to the integration of conducting polymer with porous frameworks. Besides, the whisker-like morphology of polyaniline reduces the diffusion length improving ion transport within the framework that results in considerably high conductivity and specific capacitance of PANI@TCOF-2. 25 The specific capacitance of PANI@TCOFs decreased with increasing the current density, and 43% retention was observed for PANI@TCOF-2, suggesting substantial rate capability of the composite materials (Fig. 3g). The cyclic stability of PANI@TCOFs coated electrodes at a 100 mV s -1 scan rate up to 1000 cycles was examined. The capacitance retention of 69% and 62% was observed for PANI@TCOF-1 and PANI@TCOF-2, respectively (Fig. 3h). Furthermore, the charge-discharge cycling employing PANI@TCOF-2 was performed for 500 cycles at 10 A/g current density. The high retention of coulombic Fig. 3 The electrochemical performance of composite materials in 1 M H 2 SO 4 . Cyclic voltammograms at different scan rates and galvanostatic charge-discharge curves at different current densities (a, b) for PANI@TCOF-1 and (c, d) PANI@TCOF-2, respectively. (e) Specific capacitances of TCOFs and PANI@TCOFs at 0.5 A g -1 current density. (f) Temperaturedependent conductivity profiles of PANI@TCOF-1 and PANI@TCOF-2 (symbols and solid lines denote the experimental and fitted data, respectively). (g) Specific capacitances of PANI@TCOFs obtained at different current densities. (h) Cyclic stability tests of PANI@TCOFs at a scan rate of 100 mV s -1 . (i) Nyquist plots of PANI@TCOFs at frequency range 100 mHz-100 kHz; symbols and solid lines denote the experimental and fitted data, respectively. The lower inset shows the magnified Nyquist plots in the high-frequency region. The upper inset shows the equivalent circuit model to fit the Nyquist plots; R S is the solution resistance or equivalent series resistance, R ct is charge transfer resistance at the electrode/electrolyte interface, Q dl is a constant phase element for double layer capacitance, W d is Warburg element, and Q ps is the pseudocapacitive charging. (j) Ragone plots of gravimetric energy density against power density for PANI@TCOFs-based supercapacitors. (k) Four individual solid-state asymmetric supercapacitors employing PANI@TCOF-2 are connected in a series; the device can power a green light-emitting diode (LED) glowing for ∼1 min. efficiency of 93% suggests the efficient charge storage and durability of the composite materials. The electrochemical impedance spectroscopy (EIS) was performed further to get a detailed insight into the electrochemical behaviour of the materials. EIS measurements were carried out in a frequency range from 10 5 Hz to 0.1 Hz at open circuit potentials. The Nyquist plots for PANI@TCOF-1 and PANI@TCOF-2 feature a nearly vertical rise of impedance along the imaginary axis, indicating an excellent capacitive performance (Fig. 3i). 21c,29 The electrical equivalent circuit (EEC) based on EIS results was constructed (Fig. 3f, upper inset), and the different components of the electrochemical system were analyzed. R s is the solution resistance or the equivalent series resistance, and R ct is the charge transfer resistance at the electrode/electrolyte interface. The lower value of R ct for PANI@TCOFs compared to pristine TCOFs indicates the facile charge transfer among electrolyte ions and PANI@TCOFs at the electrode interface. 30 The double layer capacitance at the electrode/electrolyte interface because of the surface inhomogeneity of the electrode is represented by a constant phase element (CPE, Q dl ). The Warburg element (W d ) and the pseudocapacitive charging (Q ps ) arise due to the diffusion limitation and electron transport inside the nanoporous framework, respectively. At higher frequencies, the diffusion of ions inside the nanopores was restricted due to a shorter time period. 31 The 'knee frequency', the frequency at which the diffusion of ions begins, was 121 Hz for PANI@TCOF-2 compared to 82 Hz for pristine PANI. The high knee frequency of PANI@TCOF-2 indicates a facile ion transport inside the porous networks. The Bode plots also suggest facile diffusion of electrolyte ions throughout the nanoporous surfaces of PANI@TCOFs. The transition point between the resistive and the capacitive behaviour was obtained by the characteristic frequency, f 0 , at the phase angle of -45°. 32 The dielectric relaxation time constant (τ 0 = 1/f 0 ) represents the time required for a supercapacitor to deliver half of its power. 33 The estimated τ 0 for PANI@TCOF-2 was 87 ms. The Ragone plots for PANI@TCOFs are presented in Fig. 3j. The maximum energy density and power density for PANI@TCOF-2 were 24.4 W h kg -1 and 4000 W kg -1 , respectively, promising for the device fabrication. As discussed above, the encouraging results prompted us to fabricate a solid-state asymmetric capacitor using PANITCOF-2 and acetylene black as the electrode materials. A homogeneous slurry containing PANI@TCOF-2 (70%), carbon black (15%), and polyvinylidene fluoride (PVDF) binder (15%) in Nmethylpyrrolidone (NMP) was coated on a platinum foil (1 cm 2 ). The electrodes were assembled in a two-electrode set-up using a Whatman filter paper as a separator with a gel electrolyte (polyvinyl alcohol and H 2 SO 4 ). The four individual solid-state devices derived from PANI@TCOF-2 were connected in a series to enhance the working potential window. After charging the device for 10 s, a green light-emitting diode (2.2-3.3 V) can be powered for ∼1 min. ## Conclusions In summary, we successfully synthesized conducting polymer modified triazine-based covalent organic frameworks (PANI@TCOF-1 and PANI@TCOF-2) through in situ polymerization of aniline using chemical oxidation method. The characteristic stretching bands and the variation in the pore size distribution of PANI@TCOFs implied the inclusion of PANI within the porous frameworks of TCOFs. The SEM and TEM images showed a well-distributed intricate network structure attributing to the loading of polyaniline with the porous framework that resulted in significantly enhanced electrical conductivity (1.4-1.9 x 10 -2 S cm -1 ). More interestingly, the whisker-like morphology of PANI in PANI@TCOF-2 enhanced the ion diffusion within the framework. The synergistic effect of high electrical conductivity, faster ion diffusion, and reversible redox reactions of PANI led to the 10-fold enhancement in specific capacitance of composite materials compared to pristine TCOFs. Owing to significant energy (24.4 W h kg -1 ) and power (4000 W kg -1 ) density, PANI@TCOF-2 was used as a cathode material to devise a solid-state asymmetric supercapacitor that can light up a green LED. Thus, the present study delineates a cost-effective in situ polymerization strategy for doping the conducting polymer into the porous framework for large-scale production of all-organic electrode materials for the futuristic energy storage devices. ## Conflicts of interest There are no conflicts to declare.

chemsum

{"title": "Post-synthetic modification of covalent organic frameworks via in situ polymerization of aniline for enhanced capacitive energy storage", "journal": "ChemRxiv"}

dissolving_used_rubber_tires

## Abstract: Used automobile tires present an enormous environmental burden. Efficient methods for degradation of the sulfur crosslinks in organic elastomers have proven elusive. We show that the reductive silylation of RS-SR bonds to silyl thio ethers RSSiR'' 3 in up to 90% yield using a variety of hydrosilicones occurs in the presence of <1 mol% B(C 6 F 5 ) 3 for model compounds. Sulfur-cured automotive rubber required 10 wt% catalyst for efficient sulfide cleavage. At temperatures ranging from room temperature to 100 °C recoveries of organic polymers as oils from tires using this one step process ranged from 56% for complex mixtures of rubber crumb from ground tires to 93% for butyl rubber (bicycle inner tubes; 87% yield at 100 °C over 30 minutes). After removal of inorganic materials by simple filtration, the recovered polymeric oils were radically or oxidatively crosslinked to generate new elastomers that can be optionally reinforced with the solids recovered in the initial reduction procedure. This mild process constitutes a facile route to reutilize the organic polymers found in automobile and other sulfur-crosslinked rubbers. † Electronic supplementary information (ESI) available: Tables of reactivity of diand tetrasulfides, rubber components, efficiency of repeated reduction and GPC data of produced oils. Figures showing NMR data of starting rubbers, product oils, TGA data or starting rubbers, and products oils, photographs showing the process from rubber to oils to new rubbers. See ## Introduction The radically induced 1 vulcanization of alkene-containing hydrocarbon polymers, reported by Goodyear in 1844, 2 is a technological advancement that remains an integral part of modern life; sales of automobile tires prepared using this process are expected to reach 3 billion units in 2019. 3 The elastomer products, crosslinked with sulfur oligomers, are incredibly robust products, which pose a major challenge; they are far too stable to be readily recycled. Automobile tires exemplify polymers derived from fossil fuels that are destined for single use. Used tires constitute a significant environmental burden, particularly because of the scale of production. 4 In large part, used tires are simply placed in stockpiles, 5 from which leaching into the environment of their many (toxic) constituents occurs. 6 Dangerous, highly polluting, difficult-to-arrest tire fires at such storage facilities are not uncommon. 7 Some automotive rubber is exploited as fuel in the cement industry; capturing SO 2 during combustion may be problematic. Some tires are turned into crumb and used as fillers, 8 for example, in asphalt, cement or turf replacements from which, however, leaching of contaminants may still occur. 9 There is a longstanding need to recover the organic materials from tires for sustainable reuse. Although the S-S bond strength is only ∼280 kJ mol −1 , 10 practicable processes for S-S cleavage in vulcanized tires have not been reported. Aggressive chemical approaches, for example, reactive reduction with LiAlH 4 11,12 or amines 13 have not proven commercially viable. Reuse strategies therefore typically involve energetically intensive, relatively inefficient pyrolytic conversion into fuel gas, low grade carbon black and other low value materials. 14 The world is in dire need of new efficient methods for recycling waste tire rubber, especially in ways that allow the recovery and reuse of the basic building blocks. The weak Si-H bond in hydrosilanes makes them excellent reducing agents, particularly when strong Si-heteroatom bonds are formed in the process. 15 Thus, reductive hydrolysis/ alcoholysis 16 and CvO hydrosilylation, 17 among others, are efficient processes. Key to the work described here is the Lewis acid-catalyzed (typically B(C 6 F 5 ) 3 = BCF) reduction of carbonyls (Fig. 1a), ethers, silanols, alkoxysilanes (the Piers-Rubinsztajn reaction Fig. 1b), benzylic sulfides, and thioacetals (Fig. 1c) using HSi functional groups. 21 These reactions are normally easy to control, often work at room temperature, and the main experimental issues are associated with managing the co-products when they are flammable gases, including hydrogen or alkanes. We report that, catalyzed by B(C 6 F 5 ) 3 , hydrosilanes effectively reduce S-S bonds of model organic disulfides, tetrasulfides and, more importantly, complex sulfur-crosslinked solid automotive rubbers in the forms of bicycle inner tubes, solid tires or tire crumb in good to excellent yield. The products are polymeric, silyl-protected thiolated organic oils that are readily separated from the accompanying, unreactive solids, such as fillers, fiber and metal reinforcements, pigments, etc. simply by filtration or centrifugation. The products, sulfur-containing polymeric oils, may be converted back into (reinforced) rubbers using simple oxidative or radical processes. ## Model reductions of dibenzyl disulfide and tetrasulfide Model reductions of dibenzyl disulfide 1 (n = 1) were undertaken with bis(trimethylsiloxy)methylsilane 2 (HSiMe (OSiMe 3 ) 2 , MD H M) (the common nomenclature used for silicones is Q: SiO 4/2 ; T: MeSiO 3/2 ; D: MeSiO 2/2 and M: Me 3 SiO∼) as reducing agent in the presence of BCF (Fig. 1d, e, Fig. 2, Table S1, and Fig. S1, ESI †). With less than one equivalent of hydrosiloxane, residual starting material and only product 3 were recovered, demonstrating that the reaction of SiH with S-H, e.g., in compound 4, is faster than that with S-S bonds. Complete reduction of 1 → 3 required 2 equiv. of the hydrosiloxane and occurred in 90% yield using only 0.8 mol% BCF; the other sulfur-based product 5 was removed under reduced pressure. Reduction of the analogous tetrasulfide 6 to 3 using HSiMe 2 OSiMe 3 7 required five equivalents of MD H M and demonstrated that both C-S a -S a and S a -S b -S b linkages undergo efficient reductive silylation (Fig. 1f, 2 and Fig. S2, ESI †). ## Reduction of used automotive rubbers Automotive tires contain a complex variety of constituents, including (spent) catalysts for their formation, antioxidants, colorants, particulate reinforcing agents like carbon black and/ or organosulfur-modified silica, and fibrous reinforcing agents including nylon cord and woven steel. 14 Holding together this complex assortment of excipients is the sulfur-crosslinked elastomer. We reasoned that the (oligo)sulfide linkages in rubber tires could be reductively cleaved using hydrosilanes in analogy to the reactions with the oligosulfide model compounds (Fig. 1d-f). Scrap rubber from automobile tires is available in large quantity in the form of 'rubber crumb'. It is formed by shredding tires from multiple sources to remove metal wires and polyester cord and grinding the resulting product to various crumb sizes. The typical organic constituents in crumb mixtures include isobutene isoprene (IIR), butadiene rubber (BR), styrene-butadiene rubber (SBR), isoprene rubber (IR) and natural rubber (NR). Thermal degradation profiles allow one to determine the constitution of sulfur-crosslinked elastomers; depending on their structure, the polymers degrade between 300-485 °C; 22,23 (Fig. S5, and S6, Table S3 †) while inorganic carbon (carbon black) thermally decomposes from 560-800 °C in oxygen. 24 In the rubbers tested, the organic rubber content was approximately 60 wt% (Table 1, Fig. S11 †). Two separate sources of commercial crumb were compared for their reactivity under the reducing silylation conditions. Much more BCF catalyst (10 wt% compared to the rubber start- S2. † ing material) was required to achieve reasonable yields of reduction with rubbers than with the model compounds (<1 mol%), which is not surprising given the complexity of the mixed rubber starting materials and the fact that they have been exposed to degradation and various environments during use. With 7 as reducing agent, organic oils were recovered in 36% yield from one crumb rubber source and a moderate 56% yield from another (2 lots, Fig. 3g). Use of second and third reduction steps with fresh catalyst and hydrosilane significantly improved overall conversion of elastomer to oils (steps (1) 56% → (2) +29.6% → (3) +2.5% (total 88% organic recovery) the inorganic constituents were removed by centrifugal separation (Tables 2 and 3, Fig. S14 †)). In retrospect, in a practical sense, only the improvement in overall efficiency of the second step could be justified. The reduction process was readily visible by eye, as black dispersions were converted to yellow oils (Fig. 3b → f,g, Fig. S7 and S8, ESI †). Improved recoveries of organic polymers were observed with single composition rubbers. For example, about 60% of an EPDM elastomer (ethylene propylene diene terpolymer, 'pond liner') was converted to a soluble oil using reductive silylation with 7 and B(C 6 F 5 ) 3 in toluene at 60 °C. Automotive rubbers were efficiently reduced to oils in one step in yields ranging from 52-93%. These included: IIR from a (used) bicycle inner tube; IR/NR from the outermost section of truck tire tread; a mixture of IIR, IR/NR and BR from a snow tire side wall; and IR/NR, BR from a snow tire tread (Tables 2 and 3). The process is easily seen from the reduction of a bulk section of snow tire tread (Fig. 3a). Shortly after the reduction reaction started, the reinforcing steel wires separated from the bulk rubber and were collected on the magnetic stir bar (Fig. 3c); accompanied by particulate formation to give a black dispersion (Fig. 3b)the bulk rubber underwent shrinkage, but not complete disintegration (Fig. 3d, Fig. S19 †). Filtration allowed separation of a black solid mass comprised primarily of inorganic excipients and a yellow solution of silylated organic oils in toluene (Fig. 3e, f, Fig. S7 and S18 †). Oligosulfides were converted into silyl thiol ethers during reduction (Fig. 3g and Fig. S7 †). Therefore, the residual solids and recovered non-volatile liquids often exhibited a weight gain when compared to the starting rubber mass (Table S4 †). The product oils 8 (Fig. 3, for 1 H NMR data, see Fig. S15 †) typically exhibited a bimodal distribution of molecular weight, with a low fraction centered near 10 000 g mol −1 and a broad peak centered near 1 million g mol −1 (Table S5, Fig. S16 †). Several factors were manipulated to improve the efficiency of the reduction process. A variety of hydrosilicones are commercially available that vary in the density of SiH groups. Model studies on the reduction of the organic sulfides or auto- motive rubbers were undertaken with 2 or 7, respectively, because the use of small molecules facilitated characterization of the reaction products. Either compound is too expensive for practical use. Attempts to facilitate reduction of rubbers with the inexpensive, high SiH density polymer Me 3 Si (OSiMeH) n SiMe 3 9 were unsuccessful (Fig. S17 †) because the silicone product of the reduction is a network polymer, which led to the formation of intractable tars. By contrast, the use of inexpensive, high SiH density HMe 2 SiOSiMe 2 H 10 led to efficacious, rapid reduction of rubbers (Tables 2 and 3, Fig. S17 †). Unlike the model compounds above, relatively large quantities of the BCF catalyst, 10 wt% against the rubber, were c Organic yield = (total organic-recovered organic)/total organic × 100. Organic composition established using TGA (Fig. S6, S9, S11, Table S3, † and Table 1). d Cumulative yield for 3 process steps. e Process utilized hydrosilicone 10 (1.5 mL) at 100 °C for 30 min. f External road contacting component only. g Cross-section of entire tread from core to external surface. required for the reduction of rubbers to occur efficiently. Preswelling the rubber in commonly used organic solvents like acetone, 25 or an initial Soxhlet extraction using acetone to remove potential catalyst inhibitors, e.g., amines, free sulfur, acetone soluble colorants, antioxidants, processing rubber additives, etc., did not appreciably increase either the rate or yield of the reduction (Table 3, Fig. S14 †). We continue to work on process optimization to reduce the quantities of catalyst required. Several other factors were found to the affect the efficiency of the reduction of rubbers, including surface area, process temperature and reducing agent. Unsurprisingly, the reduction of bulk elastomers, including cylindrical sections of bicycle inner tube (diameter ∼16.9 mm, thickness ∼0.82 mm, IIR, Fig. S7 †), and a cross-section of an automobile snow tire (IR/NR + BR, Fig. 3), were slow to occur and low yielding at 60 °C using 7 (Fig. S6 †). In both cases, the objects underwent significant shrinkage (Fig. S19 †), and increased crosslink density as shown by an increase in Young's modulus (Table S4 †), but maintained their shape. Cryogenic grinding of the starting rubbers to increase rubber surface area (to particle size ∼330 μm, Fig. 4) led to significant enhancements in yield; an increase from 52 to 93% was observed in the case of the inner tube (IIR, Table 3). The tire reductive silylation studies were initially undertaken using relatively mild temperatures because literature suggests that the B(C 6 F 5 ) 3 catalyst undergoes degradation at temperatures above 80 °C in the presence of moisture. 26 Current studies with rubber reduction, however, showed this not to be problematic. A 93% yield of recovered organic polymer (IIR from inner tube) was achieved at 100 °C after 18 hours using 10, but an 87% yield had already been achieved in the first 30 minutes (Table 3). This result suggests that reduction processes at 100 °C or higher could be adapted to a continuous process. Initial studies for reusing/re-crosslinking the recovered oil focused initially on the regeneration of thiols from silyl thio ethers. There are few reports of the reactivity of Si-S compounds. Me 3 Si-S-SiMe 3 is very labile, undergoing rapid degradation simply in the presence of water (vapor) to form H 2 S and Me 3 SiOH. 27 By contrast, the hydrolysis of silicone-based thio ethers was much less facile. Alcoholysis of 3 (Fig. 1g) yielded 12% product only once acetic acid was added to isopropanol solutions (the less sterically hindered thio ether PhCH 2 SSiMe 2 OSiMe 3 underwent rapid, quantitative cleavage under the same conditions). The silylated polymeric oils derived from elastomers 8 were yet less reactive. It was necessary to use more aggressive nucleophiles for silicon, such as Bu 4 NF 28 to regenerate the silyl free thiols 11 (Fig. 3h and Fig. S20 †). Once cleaved, the freed thiols on the organic polymers could be crosslinked into a new elastomer 12 by oxidative coupling using iodine in isopropyl alcohol 29 (Fig. 3i and Fig. S21 †). However, it was also discovered that, if silylated polymers 8 were derived from IR/NR or BR and possessed residual alkenes, re-crosslinking did not require removal of the silyl groups; simply adding a radical initiator such as benzoyl peroxide (BPO) and heating led to new elastomers 13 (Fig. 3j). The ability to create new elastomers from the recovered polymeric oils was demonstrated by creating a new automotive tire (for a child's toy) using 8. A mold of the tire was made in silicone rubber (Fig. 4a and b). Silylated oil 8 derived from IR/NR (tire tread) was placed in the mold in the presence of BPO and heated to give a new, soft elastomer (Fig. 4b + c → e,f durometer Shore OO 68, Table S6 †). Adding to 8 the inorganic excipients (recovered from the production of 8, Fig. 4h), and then curing oxidatively, led to harder, more brittle elastomers (Shore A 91; original rubber Shore A 60). Reductive silylation processes have shown synthetic merit in many arenas. The relatively weak SiH bond, 30 particularly in the in the presence of B(C 6 F 5 ) 3 and related Lewis acids, readily reduces a variety of bonds 15 driven, in large part, by the thermodynamic benefit of forming Si-heteroatom bonds (heteroatomvO, N, S, etc.). 31 We have demonstrated that this type of process works effectively with S-S bonds to form thio silyl ethers. The key finding of this work is that, in addition to clean model compounds, the process works with complex and dirty samplesused automotive tiresto convert sulfur-cured elastomers into polymeric oils, in up to ∼90% yield. The process can be rendered practicable at temperatures as low as 100 °C and the product oils can find new utility in elastomeric objects using at least two different cure modes. We are currently establishing the quality of elastomers that can be produced from different used tire feedstocks. The recovered inorganic mass can also be repurposed as a filler in those new elastomers. Thus, reductive silylation provides a new opportunity to find commercial value in materials that are environmentally problematic. It is not possible to effectively 'reuse' automotive rubber, but reductive silylation is worthy of consideration as a strategy for recycling and reuse. n M) 9, were purchased from Gelest and used after drying over molecular sieves overnight. Dibenzyl disulfide, benzyl bromide, tetrabutylammonium fluoride trihydrate (TBAF), and iodine (I 2 ) were obtained from Sigma Aldrich and used as received. Benzoyl peroxide (BPO) was purchased from BDH. Sodium tetrasulfide was purchased from Dojindo. B(C 6 F 5 ) 3 (BCF) was prepared by Grignard reaction following a literature procedure; 32 we acknowledge with gratitude Prof. David Emslie, McMaster University, for providing this sample. Rubber samples: bicycle inner tube (Chaoyang 700 × 38/45C bicycle inner tube, China), EPDM ( pond liner, purchased at a local garden centre, producer unknown), Crumb-1 (Als-RC, Amazon, Canada), Crumb-2 (Canadian Eco Rubber Ltd, Emterra, Canada), were used as received. Truck tread 1: a piece of truck tread, not part of a complete tire, was found at a local garbage dump (origin unknown). Truck tread 2: (Sailun 225/ 70R19.5). In both cases, samples were cut only from the external, road contacting tread part (Fig. S3 †); both rubbers were based on polyisoprene. The tread and side wall samplescross-sectionswere cut from different parts of a used car tire (snow tire, Cooper 185/65R4, Fig. 5). Naphthalene (internal standard) was purchased from Fisher. Toluene (solvent) received from Caledon (HPLC grade) was dried over activated alumina before use. Deuterated NMR solvents were obtained from Cambridge Isotope Laboratories. Glass apparatus were dried overnight at 120 °C and cooled under a dry nitrogen atmosphere for 30 min prior to use. ## Methods NMR. 1 H, 13 C and 29 Si NMR spectra were recorded on a Bruker Avance 600 MHz nuclear magnetic resonance spectrometer using deuterated solvents chloroform-d and actone-d 6 . For 29 Si NMR, chromium(III) acetylacetonate was used as paramagnetic relaxation agent. GC-MS. Analyses were performed using an Agilent 6890N gas chromatograph (Santa Clara, CA, USA), equipped with a DB-17ht column (30 m × 0.25 mm i.d. × 0.15 μm film, J&W Scientific) and a retention gap (deactivated fused silica, 5 m × 0.53 mm i.d.), and coupled to an Agilent 5973 MSD single quadruple mass spectrometer. One microliter of sample was injected using an Agilent 7683 autosampler in splitless mode. The injector temperature was 250 °C and carrier gas (helium) flow was 1.1 mL min −1 . The transfer line was 280 °C and the MS source temperature was 230 °C. The column temperature started at 50 °C and was increased to 300 °C at 8 °C min −1 , then held at 300 °C for 15 min to give a total run time of 46.25 min. Full scan mass spectra between m/z 50 and 800 mass units were acquired after a solvent delay of 8 min. LC-MS. Analyses were undertaken using an Agilent Technologies 1200 LC coupled to an Agilent 6550 QTOF mass spectrometer. An injection volume of 2 μL was separated on a Phenomenex Luna C18(2) (150 mm × 2.0 mm, 3 μm) column with 100 pore size (Phenomenex, CA, USA). The mobile phases were LC-MS-grade 45/55 water/methanol with 0.5% acetic acid (A) and methanol with 0.5% acetic acid (B) at a flow rate of 300 μL min −1 . The column temperature was maintained at 40 °C, and the autosampler storage tray was set at 10 °C. The mobile phase gradient eluted isocratically with 10% B for 1.0 min followed by a gradient to 100% B over 17 min. The gradient was maintained at 100% B for 2 min and decreased to 10% B over 0.1 min. The gradient was then followed by a 5 min re-equilibration prior to the next injection. The total time for an HPLC run was 25 min. The MS parameters (for LC-MS) chosen were as follows: ESI, gas temperature at 225 °C, drying gas at 13 L min −1 , nebulizer pressure at 20 psi, sheath gas temperature at 400 °C, sheath gas flow at 12 L min −1 , VCap at 3500 V, nozzle voltage at 1000 V, fragmentor at 375 V, and Oct 1 RF Vpp at 750 V. The data were acquired in electrospray positive mode from m/z 50 to 1000 at a scan rate of 1.5 Hz. The mass was auto recalibrated using reference lock mass from Agilent ESI-T Tuning Mix (for Ion Trap). GPC. Molecular weight of recovered organic oil and polydispersity index (PDI) were estimated from gel permeation chromatography (GPC) using a Waters 2695 Separations Module equipped with a Waters 2996 photodiode array detector, a Waters 2414 refractive-index detector, and two Jordi Labs Jordi Gel DVB columns. Polystyrene standards were used for calibration, and THF was used as the eluent at a flow rate of 1.0 mL min −1 . The polymer constituents of rubber samples were estimated from carbon high-resolution magic angle spinning ( 13 C HR-MAS) NMR spectroscopy (Fig. S5 †). TGA: Thermogravimetric analysis (TGA) according to ASTM D 6370-99 (American Society for Testing and Materials) was used to measure the organic carbon (polymer), carbon black content and inorganic residue of the component. 24 A small amount of test sample (2 to 5 mg) was placed into the alumina pan of the calibrated Thermogravimetric Analyzer (Mettler Toledo TGA/DSC 3+). A 100 cm 3 min −1 argon purge was applied and the furnace was heated from 50 °C to 560 °C at 10 °C min −1 . Then, the furnace was cooled to 300 °C and the purge gas was changed to air at 100 cm 3 min −1 . The temperature was allowed to equilibrate for 2 min before the furnace was heated to 800 °C at 10 °C min −1 . The constituents of sulfur-cured rubbers have well-defined thermal decomposition profiles (Table S3 †). Within the thermal decomposition range of the organic polymers, there are further distinctions. TGA and DTA data (Fig. S6 †) show the constituents of the rubber samples tested (Table 1). Young's modulus. The Young's moduli of rubber samples were determined using a MACH-1 micromechanical system (Biomomentum Instruments) with a 0.500 mm hemispherical indenter radius, and Poisson ratio of 0.3. All measurements were conducted at 22 °C and in triplicate, with error bars representing the standard deviation of the replicate measurements. Preparation of powdered rubber. Raw rubber samples of different shapes and sizes were obtained. The rubber crumb samples were of broad dispersity, with average particles sizes of 1.30 ± 0.09 mm for Crumb-1 and 2.14 ± 0.06 mm for Crumb-2. As noted above, more efficient reduction occurred with higher surface area samples. A cryogenic grinding progress was used to obtain rubber powder samples with homogeneous particle sizes for comparable experiments. Liquid nitrogen was employed to cool the rubber samples below their glass transition temperature before they were pulverized with a coffee grinder (KitchenAid) to give powders with an average particle size of ∼330 mm (Fig. 5g, i and Fig. S4 †). ## Preparation of dibenzyl tetrasulfide To a pre-dried 200 mL round-bottomed flask purged with dry N 2 were added sodium tetrasulfide (0.098 g, 0.562 mmol), benzyl bromide (0.209 g, 1.22 mmol) and dry THF (50 mL, 44.45 g, 0.616 mol) as solvent. The mixture was stirred for 23 d and collected by vacuum filtration. The mixture was purified using column chromatography; the low polarity impurity (S 8 ) was removed using hexanes, after which the elution solvent was switched to dichloromethane, to give a yield of 77% (137 mg, based on the different sulfides found in the product. Note: it was not possible to detect tetrasulfide or higher polysulfide linkages in the GC-MS, which may be due to thermal degradation of polysulfide bond (when n > 3). 33 ## Titration of disulfides to establish relative reactivity of functional groups Conversion in the reaction was shown by peak area of the hydrogens on the carbon adjacent to the disulfide bond (-CH̲ 2 SS) in 1 H NMR, which was plotted against different ratios of hydrosilane (SiH̲ ) to disulphide. Analogous techniques were used to follow the reduction of the tetrasulfide (Fig. 2). ## Reduction of rubbers: general procedure The reaction conditions for the reduction of rubber powder are given in Tables 2 and 3. The general experimental procedure is as follows. The cryogenically ground rubber powder was allowed to swell in dry toluene (12 mL) for 30 min. Pentamethyldisiloxane 7 was added to the reaction mixture. The stock catalyst solution was added to initiate the reaction. The suspension was heated in a 60 °C oil bath for 48 h. The residual undissolved rubber powder was washed with toluene and separated by centrifugation (Eppendorf, Centrifuge 5424, at 12 000 rpm for 20 min). The extraction process was repeated twice to completely remove soluble compounds. The supernatants were mixed, and the solvents were removed by rotary evaporation. Any residual (organic) volatiles were removed under a stream of N 2 over 24 h. The recovered organic liquid was characterized by NMR. The recovered rubber powders were dried at 100 °C overnight and then examined by TGA. The bulk sample were reduced by the similar protocol and more details were provided in ESI. † ## Reduction with inexpensive tetramethyldisiloxane M H M H Cryogenically ground inner tube powder (300 mg) was placed in dry toluene (12 mL), followed by tetramethyldisiloxane M H M H (1.5 mL, 8.5 mmol). The stock catalyst solution (600 µL, stock catalyst concentration: 50 mg mL −1 in toluene, catalyst concentration in reaction: 10 wt%/inner tube) was added immediately afterwards to initiate the reaction. The suspension was heated in a preheated 100 °C oil bath for 30 min. The reaction flask was put into a room temperature water bath to quench the reaction and followed by a separation process using the same protocol stated in the former paragraph. The organic yield was 87%. ## Desilylation of product polymeric oils to give 11 The silylated organic oil 8 (based tire tread, derived from IR/ NR, 0.50 g) was desilylated by treatment with TBAF solution (0.5 g TBAF, 1.92 mmol TBAF, dissolved in 10 mL THF containing 0.1 mL methanol) for 24 h at 80 °C. The solvent and siloxane fragments were removed by using a rotary evaporator, followed by kugelrohr distillation; loss of silicone was clearly seen in the 1 H NMR (200 °C, 3 h, Fig. S20 †). ## Crosslinking of recovered oils Oxidative coupling of thiols using iodine. The desilyated organic oil 11 (derived from truck tread IR/NR, 0.2 g) was allowed to react with an iodine solution (50 mM dissolved in 1/1 v/v toluene/isopropanol) for 12 h at room temperature. Solvents associated with the resulting elastomer 12 were removed under a stream of N 2 over 12 h (Fig. S21a †). Crosslinking was confirmed using a swelling test. Compound 12 (0.1 g) was swelled in 10 mL of hexane; the degree of swelling was 209 wt%. The precursor oil 8 dissolved readily in hexane easily while 12 remained an elastomeric solid even after 1 h sonication (Fig. S21b †). Preparation of toy tire using a peroxide initiator (Fig. 4) A silicone mode was prepared with a two-part liquid component kit (Sylgard 184). Two components were mixed at the recommended ratio of 10 parts (10.0 g) base to 1 part curing agent (1.0 g). The mixing process was performed using a planetary centrifugal mixer (FlackTek Inc.) with a duration of 5 min at a speed of 3000 rpm. In order to fabricate bubble free elastomer, the mixed uncured PDMS was thoroughly degassed in a vacuum desiccator at low pressure for 30 min. The right front tire of a toy car was removed from the toy and placed in the degassed, uncured mixture. The mixture was cured in an 80 °C oven overnight. The tire was removed from the cured mold. Used rubber powder from the truck tread (Sailun) (2.0 g) was reduced by pentamethyldisiloxane 7 in a 100 °C oil bath for 18 h to give a polymeric oil 8 (81.5% yield). Solvents in the supernatants, after centrifugation, were removed by rotary evaporation; any residual volatiles and silicone by-products were removed using a stream of N 2 over 24 h. The oil was accompanied by residual undissolved rubber powder that was washed with toluene and separated by centrifugation (Eppendorf, Centrifuge 5424, at 12 000 rpm for 20 min), dried an 80 °C oven for 18 h, and ground into powder using a mortar and pestle (Fig. 4d). The silylated organic oil 8 from the former step (comprised of IR/NR derivatives, 0.707 g) was dissolved in hexanes (10 mL), benzoyl peroxide (BPO, 0.01 g, 1 wt%, 0.0413 mmol) and, optionally, ground residual inorganic solids (from the preparation of 8, 0.3010 g), were added sequentially and mixed to give a homogeneous dispersion. After the solvent was removed by rotary evaporation, the mixture was placed in the silicone mold and degassed under vacuum in a desiccator for 30 min. The curing process was performed at 100 °C for 18 h. The formulations for rubber with different residual solid are listed in Table S6. † ## Conclusions Utilization of the reactions described here to create polymeric oils by the reductive silylation of used automotive rubbers, and subsequent oxidation to new elastomers, permits the completion of a full cycle of use for these organic materials. Mild, efficient reduction of S-S bonds of organic disulfides, including those found in used, sulfur-cured rubbers permits the formation of silylated polymers 8already in 87% yield at 100 °C within 30 minutesthat can be separated from the inorganic constituents. Radical cure of the oils if alkenes are present, or following desilylation, oxidative cure of thiolated polymers leads to new elastomers 12, 13 that, optionally, can be reinforced by the inorganic constituents recovered during the reduction process. The practice of this recycling process has the potential to reduce the environmental impact of used, sulfur-crosslinked elastomers. ## Conflicts of interest There are no conflicts to declare.

chemsum

{"title": "Dissolving used rubber tires", "journal": "Royal Society of Chemistry (RSC)"}

chemical_biology_tools_for_probing_transcytosis_at_the_blood–brain_barrier

## Abstract: Absorptive-and receptor-mediated transcytosis (AMT/RMT) are widely studied strategies to deliver therapeutics across the blood-brain barrier (BBB). However, an improved understanding of the mechanism surrounding trafficking is required that could promote delivery. Accordingly, we designed a flexible platform that merged AMT and RMT motifs on a single scaffold to probe various parameters (ligand, affinity, valency, position) in a screening campaign. During this process we adapted an in vitro BBB model to reliably rank transcytosis of the vehicle library. Our results demonstrate heightened uptake and trafficking for the shuttles, with a structure-activity relationship for transcytosis emerging. Notably, due to their small size, the majority of shuttles demonstrated increased permeation compared to transferrin, with the highest performing shuttle affording a 4.9-fold increase. Consequently, we have identified novel peptide conjugates that have the capacity to act as promising brain shuttles. ## Introduction A major hurdle hindering the diagnosis and treatment of neurological disorders is the difficulty for biotherapeutics to enter the central nervous system (CNS). This is due to the blood-brain barrier (BBB), which is comprised of tightly connected polarized endothelial cells that limit the passage of hydrophilic components and prevents the accumulation of material for transport at the BBB. 1 Despite these limitations, specialized endogenous transport mechanisms exist to allow the transcytosis of nutrients and ions, thus enabling CNS homeostasis. Of these, absorptive mediated-(AMT) and receptor mediated-transcytosis (RMT) pathways are key vesicular based transport systems which have become long-standing approaches for drug delivery to the CNS. 2 While these routes have become widely exploited by conjugating molecules restricted by the barrier to those which have this capacity, transcytosis at the BBB is more complex than initially thought, and delivery of therapeutics and biologicals remains modest. 3,4 Whilst there is limited experimental data surrounding the molecular basis of uptake and trafficking for CNS delivery, it has been demonstrated that dissociation from receptors on the brain side is essential for trafficking mediated by the transferrin receptor (TfR), the quintessential receptor for RMT. 5 Therefore transcytosis is more likely when the overall affinity towards the receptor is moderate to low, 5 or when bivalent engagement of the receptor is discouraged to limit avidity. 6 Here we present the design, synthesis and validation of a diverse shuttle library to identify key physicochemical properties for transcytosis. We based our approach on two essential modular components to develop the chemical tool kit: (i) sweetarrow-peptide (SAP), an isolated sequence derived from the Nterminal proline rich domain of g-zein, and an innate CPP with a defned PPII secondary structure to act as a scaffold, 2,7 and; (ii) a variety of validated RMT ligands to enable targeted delivery, 8,9 Fig. 1 and Table 1. It was anticipated that uptake at the BBB could be probed through exploring ligand type (i.e. targeted receptor, affinity) and arrangement (valency, position) of the ligands fused to the scaffold, since previous studies have indicated that brain exposure is directly affected by these parameters. 3,5,6,10,11 We provide compelling evidence that uptake and brain trafficking can be improved by combining AMT and RMT motifs on a single shuttle and that better understanding of receptor mediated trafficking within the brain endothelium is required at an individual and ligand-by-ligand basis, with ligand type, number and position effecting permeability in endothelial cells. ## Results and discussion Design and synthesis of brain delivery shuttles SAP, the core of our delivery vehicles, is formed from a short repeating sequence of (VRLPPP) 3 that is readily accessible in high yield through general automated solid-phase peptide synthesis (SPPS). It is noteworthy to mention that SAP retains a PPII conformation if 50% of the sequence remains as proline, forming a left-handed helix of 3.0 residues per turn, which affords three distinct 'faces' in aqueous solution, Fig. 1B. 7,12,13 Consequently, spatial parameters of the BBB shuttle could be readily studied through simple mutation of valine, arginine, or leucine amino acids to orthogonal residues at specifc sites. For this, we included lysine (K) or propargylglycine (X) on the PPII scaffold to conjugate relevant RMT motifs through amide bond forming and copper alkyne-azide chemistry (CuAAC) conjugations respectively. The full scaffold library we generated is shown in Table 2. Mutations to the SAP primary sequence follow standard convention and, when conjugated to the specifed ligand, the attachment point on the scaffold is given in brackets and in order of conjugation. All peptide scaffolds were made with the chemically inert C-terminal amide in place of a carboxylic acid and FAM was conjugated to the N-terminus as cargo and to allow in vitro study. To enable targeted brain delivery, we chose to functionalise our scaffold with established ligands that engage the transferrin receptor (TfR) and low-density lipoprotein-receptor related protein (LRP1). The relative abundance of these receptors on brain endothelial, alongside their high transport capacity, made them ideal targets for RMT mediated delivery and both have been actively explored in research. 8, We focused on three main peptide ligands for these receptors, as shown in Table 1. Both TfRL1 and TfRL2 have been shown to interact with TfR at alternative sites compared to transferrin (Tf), and have either nanomolar (TfRL1; 15.0 nM) or micromolar affinity (TfRL2; 440 mM) towards the receptor. 9 Furthermore, branched BBB-shuttles incorporating dimers of TfRL1 have recently shown a non-linear increase in uptake within cellular models of the BBB. 11 Alternatively, Angiopep2 (APep2) displays a high transcytosis potential mediated by LRP1. 8 While less is known about traf-fcking of LRP1, APep2-conjugates have shown demonstrable success in clinical trials of neurological disease models. 8 Peptide ligands were synthesised by general SPPS on rink-amide resin (see Table S1 † for characterisation data and yields). It is important to note that the stability of the peptide vehicles was not considered in this proof-of-concept screen. Recent investigations with retro-enantio sequences of both TfRL1 and APep2 have shown that metabolic limitations can be overcome without the loss of function, offering a plausible solution to degradation. 14,17 This is also apparent for the D-amino acid analog of SAP. 18 Chemoselective synthesis of brain delivery shuttles Four conjugation approaches were used to decorate the scaffold in various formats, Scheme 1A-D. Solution phase CuAAC chemistry with microwave heating was most effective for affording PPII-ligand conjugates in either monovalent or bivalent formats that included identical ligands (Scheme 1A). Alternatively, for the attachment of orthogonal ligands to the scaffold, TfRL2 was frst incorporated via continuation of Fmoc/ tBu SPPS on the lysine side chain of PPII scaffolds, as generalised in Scheme 1B. Following FAM conjugation, unreactive Nterminal PPII was capped to ensure no unwanted reactions, before the lysine protecting group (ivDde) was selectively removed on resin. Resin was re-submitted to automated SPPS to afford the product following deprotection and cleavage. If propargylglycine (X) was also incorporated on the scaffold, bivalent conjugates with alternative ligands were afforded through subsequent reaction by CuAAC (Scheme 1C). While this approach afforded conjugates in high yield, it was somewhat limited, since conjugates of both TfRL1 and APep2 were unattainable through continuation of SPPS on resin. To overcome this limitation, CuAAC could be used to attach one ligand to the PPII scaffold, which was followed by subsequent coupling of small molecule pentynoic acid to lysine under amide bonding forming reactions. A secondary CuAAC reaction afforded bivalent scaffolds with alternative ligands in moderate yield (Scheme 1D). As demonstrated here and elsewhere, CuAAC mediated convergence provided rapid access to versatile bioconjugates that was reproducible. 19 Detailed characterisation data and yields for all conjugates are reported in Table S2. † ## Secondary structure determination of brain shuttles We determined structural integrity of the PPII helix by circular dichroism (CD) spectroscopy, where spectra of unmodifed SAP and the N-terminal FITC-conjugated analogue were collected as controls, Fig. 2A. Importantly, the results suggest preservation of the secondary PPII structure following modifcation to the backbone since all peptides demonstrated strong absorption at 203 nm, Fig. 2B-D. Specifcally, PPII helicity was insensitive to the conjugate linkage chemistry, ligand type and number of additions to the PPII scaffold. ## In vitro binding capacity of shuttles towards TfR To evaluate functionality of the conjugated peptides, in vitro analyses were performed. TfR binding was assessed through an ELISA format, whereby FAM-conjugated shuttles were serially diluted and added to plates that were coated with TfR protein. FAM-labelled vehicles were detected by an anti-FAM HRPconjugated antibody. As predicted, TfRL1-vehicles contained a higher affinity towards the TfR than the equivalent vehicle incorporating TfRL2 (PPII(L3X;L15X)TfRL1: 0.26 AE 0.17 nM; PPII(L3X;V13X)TfRL1: 0.31 AE 0.26; cf. PPII(L3X;L15X)TfRL2: 11.8 AE 1.11 nM; PPII(L3X;V13X)TfRL2: 12.7 AE 1.37), Fig. 2E. 9 Notably, TfRL2 bivalent shuttles contained a similar affinity towards TfR compared to Tf-biotin when assayed in a similar format (EC 50 : 20.9 AE 2.1 nM). It is assumed that an avidity effect is demonstrated by the bivalent vehicles since monovalent TfRtargeting shuttles contained too low-affinity towards TfR to be detected. Further, positioning of the ligand did not affect TfR binding affinity since scaffolds that incorporated ligands on the same face afforded affinity values similar to those with ligands fused on different sides. ## Evaluation of the BBB-shuttle properties of the conjugates in brain endothelial cells Having demonstrated preservation of receptor binding, we next used bEnd.3 cells to screen the uptake capacity of the shuttle library by flow cytometry. We determined peptide uptake and internalisation in a confluent monolayer of bEnd.3 cells which is an immortalised mouse brain endothelial cell line that has similar characteristics to the BBB. TfR is an established receptor for clathrin dependent uptake of Tf, favoured in an iron bound form when at physiological pH (i.e. holo-Tf). 20 Flow cytometry (FC) and microscopy data obtained with human derived Alexa-Fluor 647-conjugated Tf (Tf-A647) confrmed functional, membrane resident TfR (see Fig. S1 and S2 †). ## Evaluation of the internalisation capacity in brain endothelial cells For library screening, cells were exposed to equal concentrations (500 nM) of FAM labelled molecules in media for 3 h to allow binding, internalisation and sorting, and external FAM fluorescence was quenched by addition of trypan blue. 21 Low temperature was shown to inhibit cellular uptake of the vehicles (Fig. S3 †), demonstrating an energy-driven internalisation process. Furthermore, using DAPI as a viability indicator, no cytotoxicity of the vehicles was demonstrated during these experiments and over a concentration range of up to 100 mM, where viability remained over 98% (Fig. S4 †). It can be seen from Fig. 2F that, in general, brain shuttles contained heightened, non-additive, uptake compared to their unconjugated counterparts, implying a synergistic effect for uptake. TfRL1 demonstrated the highest capacity for endocytosis when incubated alone, however appreciable difference was shown for the monovalent vehicle in this format (TfRL1 cf. PPII(R8X)TfR1 P ¼ 0.798; TfRL2 cf. PPII(R8X)TfR2 P ¼ 0.018; APep2 cf. PPII(R8X)APep2 P ¼ 0.002). Notably, when arranged in a bivalent format, TfRL1 modifed vehicles showed a dramatic increase in cellular internalisation that exceeded additive contributions. Interestingly, endocytosis of TfR-targeting bivalent vehicles was not signifcantly influenced by its positional arrangement on the scaffold (PPII(L3X;L15X)TfRL1 cf. PPII(L3X;V13X)TfRL1 P ¼ 0.408; PPII(L3X;L15X)TfRL2 cf. PPII(L3X;V13X)TfRL2 P ¼ 0.050), compared to those targeting LRP-2 which favoured ligands attached to opposite faces of the scaffold (PPII(L3X;L15X)APep2 cf. PPII(L3X;V13X)APep2 P ¼ 0.0059). This phenomenon could be attributed to increased steric strain of the larger APep2 peptide when incorporated on the same face, or possibly linked to the ligands ability to influence receptor clustering on the extracellular membrane. Again, higher uptake was demonstrated for scaffolds that combined alternative transcytosis ligands on the same molecule compared to individual ligands. However, uptake was not improved compared to bivalent scaffolds containing identical ligands. ## Intracellular location of BBB shuttles To obtain information on the intracellular fate of the shuttles, bEnd.3 cells were pulsed with bivalent vehicles, Tf-A647 and unconjugated TfRL1 and APep2 ligands for 3 h, unbound compounds were removed and cells fxed and stained for immunofluorescence microscopy. Images were processed as described in Fig. S5. † Fig. 3A-C shows colocalisation of PPII(L3X;V13X)TfRL1 with DAPI, Tf-A647 or LAMP1 respectively and represents a typical image generated from the screen following deconvolution and segmentation. It can be seen from Fig. 3D that all vehicles gave highest colocalisation with Tf-A647 as opposed to LAMP1, indicating uptake into endosomal vesicles with minimal degradation through lysosomal sorting. In addition, Tf-A647 colocalisation values were higher for vehicles compared to the individual RMT ligand components. Notably, the degree of Tf-A647 and LAMP1 colocalisation was comparable for TfRL1 vehicles incorporating this ligand on the same or opposite face. In addition, Tf-A647 colocalisation for bivalent TfRL1 vehicles and PPII(L3X;L15X)TfRL2 was also comparable. However, increased levels of LAMP1 colocalisation was observed for TfRL2 bivalent vehicle, indicating that the higher affinity TfR ligand is more efficient for non-degradative cellular uptake in this system. Colocalisation of APep2 bivalent vehicles with Tf-A647 were affected by the position of the ligand, demonstrating a similar trend to that shown with endocytotic capacity, Fig. 2F. Trivalent vehicles decorated with TfR ligands demonstrated similar colocalisation with Tf-A647 to those in a bivalent format. Whilst vehicle PPII(L3X;L15K)TfRL2(K) TfRL1(X) demonstrated highest Tf-A647 colocalisation, it also had heightened association in the lysosome. Notably, ## Permeability experiments To study transcytosis potential of the vehicles in vitro we optimised a BBB model using a 3D-transwell format. We found that a bEnd.3/mesenchymal stem cell (MSC) co-culture reproducibly afforded the highest resistance to paracellular diffusion through assessment with transepithelial electrical resistance measurements (TEER) and small molecule permeability studies (Fig. S6 †). This is similar to results shown by others. 16 Immunostaining for TJ protein ZO-1 in the bEnd.3 cell line confrmed its presence and trafficking towards the cell junctions (Fig. S7 †). While this indicates that an adequate and reproducible barrier was formed from the bEnd.3/MSC co-culture model, we found that it was of paramount importance to distinguish transcytosis from the expected background of paracellular flow since others have reported issues distinguishing these values. 22 ## Optimisation and adaptation of BBB in vitro model Often in vitro assays of the BBB overlook contributions of paracellular flux that can lead to overestimations of brain exposure, and offer no valid comparison between the individual molecules being screened. We believed that without a quantifable probe for paracellular diffusion, it would be difficult to accurately screen the transcytosis capacity of our vehicles. It seemed plausible that a similar sized marker (such as TexasRed labelled dextran (dex-TexR; 3000 g mol 1 )) that is incapable of AMT or RMT uptake could serve as an internal standard to quantify paracellular contribution of the vehicles. As exemplifed in Fig. 4A, experiments are performed in parallel with probes (both dex-TexR and FAM-PPII vehicle or Tf-A647) fed to insets containing either the cell monolayer (PS t ) or flter alone (PS f ). The corresponding P app for dex-TexR in both scenarios is frst calculated and the extent in which movement is reduced by presence of the monolayer (defned here as the transport ratio (TR)) is used to determine the expected paracellular diffusion rate of the individual vehicles. We employed Tf-A647 and a similarly sized TexasRed labelled dextran (dex-TexR; 70 000 g mol 1 ) as a positive control for transcytosis. Typical clearance profles when compounds are incubated with and without cells, are shown in Fig. 4B-C. As expected, due to their similar molecular weight, both dex-TexR and Tf-A647 afforded comparable clearance when incubated without cells (PS f ). When incubated in the presence of cells, dex-TexR was detected in the basolateral compartment indicating background paracellular flux. However, Tf-A647 was cleared at a faster rate in comparison, confrming transcytosis of the molecule. Corrected values were subsequently calculated to internally rank the transcytosis capacity of BBB-shuttles. In addition, to verify validity of the model, a selection of high permeability vehicles were selected for further studies at low temperature, where uptake and transcytosis should be inhibited. In agreement, shuttles showed permeability similar to dex-TexR (3000 g mol 1 ), demonstrating negligible true transcytosis in this condition (Fig. S8 †). ## Permeability screen of BBB shuttles Our screen shows that the majority of functionalised vehicles had higher rates of permeability than Tf-A647 with PPII(R8X) TfRL1, PPII(L3X;L15X)APep2 and both bivalent TfRL2 vehicles being the exception, Fig. 4D. TfRL1 modifed vehicles preferred a bivalent format (PPII(L3X;V13X)TfRL1: 3.5-fold increase relative to Tf-A647; PPII(L3X;L15X)TfRL1: 3.0-fold), whereas TfRL2 and APep2 monovalent vehicles experienced higher levels of basal transport (PPII(R8X)TfRL2: 1.4-fold; PPII(R8X)APep2: 4.9- fold), with the latter vehicle showcasing the best permeability of those screened. In agreement with the endocytosis and microscopy studies, TfR targeting bivalent vehicles showed a moderate or no clear positional preference for ligands attached to alternative faces, whereas Apep2 highly favoured ligands attached to alternative faces of the scaffold with PPII(L3X;L15X)Apep2 showing limited transcytosis. In agreement with the microscopy data, trivalent vehicles where both ligands are directed towards the TfR demonstrated improved transcytosis when ligands were incorporated on opposite sides of the scaffold. In addition, the data implies that affinity towards the TfR does not govern transport ability for our peptidic BBB shuttles since trivalent vehicles combining high and low affinity ligands showed lower permeability than the corresponding bivalent high affinity TfRL1 vehicles. Whilst vehicle PPII(L3X;V13K)TfRL2(K);APep2(X) gave the highest level of transcytosis for a highly functionalised shuttle (3.8-fold increase relative to Tf-A647), the increased permeability may be attributed to the ligands interacting independently with their receptors, since both ligands enjoyed higher transport in a monovalent form. In line with this hypothesis, transcytosis for PPII(L3X;L15K)TfRL2(K);APep2(X) was lowest for trifunctional vehicles with ligands merged to the same face (1.8-fold), implying that the ligands are not working in synergy. This result is in contrast to when ligand TfRL1 and APep2 are combined, which showed improved permeability when ligands are arranged on the same face. ## Conclusions Here we report the frst use of CPP SAP as a scaffold for developing targeted BBB penetrable shuttles, which constitutes one of the most restrictive barriers in the body. In this regard, we designed a versatile vehicle library by a convergent approach, strategically introducing mutations within the sequence to graft RMT ligands in a selective manor using either amide bond forming or CuAAC mediated reactions. As noted elsewhere, CuAAC provided a superior and flexible reaction for these modifcations. Notably, SAP retained a helical PPII structure after modifcation, and vehicles screened in biologically relevant assays demonstrated uptake and trafficking of cargo at the BBB. It was shown that AMT and RMT motifs worked in synergy to encourage cellular uptake, with certain molecular characteristics such as affinity, position and valency influencing both uptake and transcytosis for individual ligands. Notably, the majority of vehicles screened demonstrated heightened transcytosis rates compared to Tf in a BBB model. Here we believe the small size of the peptide conjugates, compared to Tf and other macromolecule shuttles, afford them an advantage for targeted transcytosis due to higher diffusion rates. Within our permeability screening campaign, dex-TexR was successfully included as an internal standard for quantifying paracellular and non-specifc movement. Consequently, the results presented demonstrate that PPII derived shuttles represent novel, exciting and promising classes of bioconjugates for enhancing uptake at the BBB. The flexibility of the screening approach could be readily adopted to investigate other ligands for AMT and RMT uptake at the BBB to validate and identify optimal ligands and shuttles for delivery. ## Conflicts of interest There are no conflicts of interest to declare.

chemsum

{"title": "Chemical biology tools for probing transcytosis at the blood\u2013brain barrier", "journal": "Royal Society of Chemistry (RSC)"}

synthesis_of_a_highly_aromatic_and_planar_[10]annulene

## Abstract: As the next neutral structure following Hückels (4n+2)erule, a planar and aromatic [10]annulene is the ideal framework to study the link between ring size and aromaticity. However, the puckered geometry of the parent [10]annulene (1) suggests that the aromatic stabilization energy (ASE) is not sufficient to overcome the ring strain that exists when the system is forced into planarity. 1 It has been shown computationally that this ring strain can be alleviated through the addition of two or more cyclopropane rings to the periphery (5-6), 2 thereby creating theoretically aromatic structures. An alternative strategy to eliminating the issue of ring strain was demonstrated experimentally with the successful preparation of the highly aromatic 1,6-didehydro[10]annulene (2). 3 However, the system rapidly cyclizes at -40 °C to a naphthalene diradical due to the close proximity of the in-plane p-orbitals present in the system. Here we show that cyclopropanating one side of the unstable annulene (2) successfully prevents the destabilizing cross-ring interaction while maintaining a highly aromatic structure. Remarkably, the formed [10]annulene ( 7) is bench stable and can be stored for extended periods of time. Aromaticity is a fundamental concept that permeates many facets of chemistry. 4 Identifying (anti)aromatic motifs in various forms has helped explain the (in)stability of both transition state 5 and ground state structures. 6 Our working understanding and ability to predict its occurrence has also informed the preparation of materials. 7 However, despite these advances, we still lack much in the way of a fundamental understanding. The traditional criteria for aromaticity proposed by Hückel require that a compound be cyclic, planar, conjugated, and contain (4n+2) electrons. 8,9,10 Any compound that can satisfy these criteria is predicted by Hückel molecular orbital (HMO) theory to have some degree of aromatic stabilization energy. 11 Besides the benzenoid archetypes, charged Hückel systems of various ring sizes have also been prepared to explore the effect of ring size on aromaticity. 12 Unfortunately, the systematic study of these charged species is often quite challenging when compared to neutral annulenes. The larger annulene has been prepared and has been shown to be aromatic, 13 though its low-frequency out-of-plane vibrations and flexibility in solution make it difficult to probe its aromaticity experimentally. 14,15 There has also been a considerable amount of interest in very large aromatic systems with nanoscale structures, such as those prepared by Anderson et al. 16 Figure 1: a. The parent annulene (1) has been found to be non-aromatic due to its non-planarity; b. The cyclization of 1,6-didehydro annulene (2) to a biradical (3); c. The structure of 1,6-methano annulene (4) and the unusual anisotropy of the induced current density (AICD) isosurface; d. The expansion of CCC bond angles upon cyclopropanation has been shown to generate theoretically planar annulenes (5-6); e. Cyclopropanation of the thermodynamically stable 2 may afford a kinetically stable and planar annulene (7) that retains its aromaticity. As the next highest Hückel aromatic neutral hydrocarbon, a rigid and planar annulene would provide a desirable framework to study the link between aromaticity and ring size. Because the parent annulene (1) is non-aromatic 1, 17,18 (Figure 1a) it must be constrained to force it to adopt a planar conformation that allows for electron delocalization. Myers and coworkers found success in this regard through the preparation of a 1,6-didehydro annulene derivative (2), which has shown evidence of aromaticity. 3 However, the -bonds orthogonal to the conjugated aromatic system readily interact across the ring, resulting in a rapid intramolecular cyclization above -40 o C (Figure 1b). The first reported alternant aromatic annulene is the bridged 1,6-methano annulene (4). It was prepared by Vogel in 1964 19 and has recently shown potential as an organic material. 7b The structure is stable and aromatic with a slight deviation from planarity (4, Figure 1c). Despite this, it has been shown that the electronic structure of 4 more closely resembles a homonaphthalene due to cross-ring interactions. 20,21 Additionally, the presence of trans double bonds as well as the unusual and non-homogenous magnetically induced current also suggests it is not an ideal material to study the aromaticity of expanded benzene-like planar annulenes. 22 In an effort to identify potential planar annulene frameworks for further study Schleyer et al. explored cyclopropanated derivatives (Figure 1d). 2 It was shown computationally that cyclopropanated all-cis annulenes (5 and 6) could adopt a planar structure due to the increased exo bond angles of cyclopropenes (Figure 1d, left). 2 However, despite their significant predicted aromatic stabilization energy no preparation of a cyclopropanated annulene has been reported to date. Consequently, no experimental information is available regarding their thermodynamic and kinetic stability. In order to definitively study the role of ring size on aromaticity we sought to identify a bench stable all-cis annulene with a sigma framework more reminiscent of an expanded benzenoid. As the Myers annulene was found to be a thermodynamic minimum, it was believed that cyclopropanation of one of the in-plane −bonds of 2 (Figure 1e) would prevent the transannular orbital interaction, thereby creating a kinetically stable structure (7, Figure 1e). This would generate a structure reminiscent of those predicted by Schleyer and one that closely resembles an all-cis and planar annulene. The essential strategy employed involved constructing a suitably oxidized but strain-free 10-membered ring that could be rapidly aromatized (Figure 2a, 11). A slow and stepwise introduction of strain was rejected in favour of the use of a late-stage rapid sequence of highly reliable strain inducing transformations, bypassing the need to carry unstable and non-aromatic intermediates through the synthesis. As kinetic stability of the final product was a concern, the possibility for a mild aromatization procedure had to be considered. A dichlorocyclopropane group was thought to be beneficial for this purpose as it would improve substrate versatility. Elimination of chloride would generate a transient cyclopropene that would isomerize to an exomethylene cyclopropane, irrespective of an aromatic or conjugative driving force (Figure 2b). This aromatization step was shown to work for the preparation of cyclopropabenzene (see SI). 23 The preparation of diene 19 closely followed literature procedures 24 with some minor modifications. 25 The synthesis began with the preparation of 14 via a Diels-Alder reaction between sulfolene (12) and maleic anhydride (13). The adduct (14) could easily be converted to 16 via an ethanolysis and cyclopropanation. LiAlH4 reduction afforded diol 17, which was readily mesylated to provide 18. Elimination to the diene 19 proceeded with a high yield when it was first converted to the corresponding diiodide (S1, see SI). Exposing the diene 19 to tetrachlorocyclopropene (21) in THF for 3 days generated 22 as a 3:1 ratio of cis and trans diastereomers. A subsequent ozonolysis with a standard reductive quench afforded the corresponding diketone 23, which could be isolated on smaller scales. However, this diketone proved very sensitive and a transannular aldol product (24) was typically obtained when the reaction was scaled up, regardless of the quenching reagent and temperature. Fortunately, a sodium borohydride quench of the ozonolysis reaction mixture at -78 °C was found to generate diol 25 as a single diastereomer in 70% yield (Figure 4). Note that typically only cis-22 underwent ozonolysis and all attempts to induce an ozonolysis of trans-22 by prolonging the reaction time or increasing the temperature resulted in the formation of an intramolecular aldol product like 24. All attempts to dehydrate diol 25 by elimination of the corresponding dimesylate only afforded the mono-alkene and decomposition products (not shown). Hoping that the other diastereomer of the diol would yield more favourable results, it was decided to attempt a double Mitsunobu inversion. Unexpectedly and to our delight, the desired diene 26 was directly obtained in high yield. We speculate that the nucleophile could not approach the concave face of the rigid system, resulting in the more favourable elimination pathway instead. The conditions strongly favoured the symmetric diene 26, presumably due to the strong electron withdrawing nature of the adjacent tetrachlorocyclopropane moiety. 26 Careful exposure of 26 to MeLi at 0 °C generated a dichlorocyclopropene (not shown) by a 1,2-didechlorination without noticeable formation of Skattebøl rearrangement products. 27 The intermediate dichlorocyclopropene could be hydrolyzed in situ to generate the bench stable cyclopropenone 27. Decarbonylation of 27 by exposure to 300 nm UV light in dilute dichloromethane proceeded smoothly to yield the dienyne 28. Initial attempts to oxidize 28 or to carry out a double elimination/isomerization sequence with tBuOK at room temperature led to material degradation. Decreasing the temperature eventually produced observable aromatic signals (> 7 ppm). These were obtained alongside a major product in the 1 H NMR spectrum that was identified as 29. It was presumed that the major product 29 was obtained by deprotonation of the allylic proton in 28 with a subsequent cyclopropane ring opening. Interestingly, the crude reaction mixture remained intact over several days and the aromatic signals that were produced persisted despite storage in CDCl3. Unfortunately however, only minute amounts of material could ever be obtained. Purification attempts were further complicated due to the unexpected extreme volatility of 7 and the surprisingly similar behaviour of the two products (7 and 29). The materials were observed to coelute in both normal and reverse phase chromatography and even co-distilled. 28 Reverse phase HPLC showed peak resolution, but the small amounts of material and its high volatility would have rendered material recovery impractical. Eventually it was discovered that doping a PTLC plate with AgNO3 prior to development resulted in a separation that was sufficient to purify the materials. 29 As the alkyne in 7 is heavily delocalized in the aromatic system, it presumably has a weaker affinity to the Ag + dopant than the alkyne in the non-aromatic 29. Using this technique, the compound could be purified and fully characterized. Interestingly, despite the presence of only trace amounts of material it had a very distinct and pleasant smell. It is also notable that the product is remarkably stable, persisting in CDCl₃ for months at -20 °C and weeks at room temperature without noticeable degradation. The experimental 1 H NMR spectrum of the purified annulene 7 displayed four distinct signals, three of which appeared in the aromatic region between 7.6 and 8.2 ppm (Figure 5, top). 30 The signals displayed the expected multiplicities with minor broadening. Second order effects and couplings rendered experimental coupling constant extraction challenging, although a spectrum obtained with computed chemical shifts and coupling constants very closely matched the signals observed in the aromatic region (Figure 5, bottom). 31 The close matching computed 1 H-NMR spectrum obtained in the absence of rovibrational considerations suggests it is a highly rigid structure that maintains both planarity and aromaticity. 32 This is in contrast to other larger aromatic structures such as annulene, whose computed 1 H-NMR spectrum can only be obtained after considering dynamic motion. 14b Nucleus-independent chemical shift (NICS) 33 computations show a highly aromatic system with NICS(1)iso = -12.0 and NICS(1)zz = -31.9 (benzene NICSiso = -10.6 and NICS(1)zz= -30.09). The uniformity of the magnetically induced current isosurface as calculated by the anisotropy of the induced ring current (AICD) method is also quite apparent (Figure 6). 34 Additionally, all computed sp 2 -sp 2 bond lengths fall between 1.37 and 1.41 , further corroborating the high degree of delocalization. Additional minima corresponding to a potential cumulenic structure could not be located, suggesting that the annulene exists as one fully delocalized minimum rather than two interconverting species. Indeed, natural resonance theory 35 analysis shows dominant contributions from the cumulenic and alkynic valence bond structures with a slight preference for the latter. 36 Interestingly, the presence of the cyclopropane ring appears to have no effect on the aromatic system, 37 and induces only a limited amount of strain (see SI). 38 This is consistent with the literature data on benzocyclopropenes where it has been shown repeatedly that such Mills-Nixon effects do not exist. 39 The quest to prepare an aromatic annulene has been a long standing challenge for synthetic chemists. Here we show that it is possible to synthesize the first kinetically and thermodynamically stable all-cis annulene structure by formally cyclopropanating the Myers 1,6-didehydro annulene (2), thus confirming our hypothesis. The planar annulene 7 by all measures is highly aromatic and remains remarkably conjugated despite the presence of both an alkyne and cyclopropane ring. The successful synthesis also supports the indicated mechanism for endo-to-exo cyclopropene isomerization (Figure 2b) 40 and shows the largely untapped potential of this transformation for generating strained systems. The transformation appears to have no dependence on aromaticity and may ultimately prove to be useful in the preparation of other annulenes. ## Associated Content Experimental procedures and characterization data are provided in the supporting information (SI).

chemsum

{"title": "Synthesis of a Highly Aromatic and Planar [10]Annulene", "journal": "ChemRxiv"}

compact_analytical_flow_system_for_the_simultaneous_determination_of_l-lactic_and_l-malic_in_red_win

## Abstract: During the malolactic fermentation of red wines, l-malic acid is mainly converted to l-lactic acid. Both acids should be precisely measured during the entire process to guarantee the quality of the final wine, thus making real-time monitoring approaches of great importance in the winemaking industry. Traditional analytical methods based on laboratory procedures are currently applied and cannot be deployed on-site. In this work, we report on the design and development of a bi-parametric compact analytical flow system integrating two electrochemical biosensors that could be potentially applied in this scenario. The developed flow-system will allow for the first time the simultaneous measurement of both acids in real scenarios at the real-time and in remote way. Miniaturized thin-film platinum four-electrode chips are fabricated on silicon substrates by standard photolithographic techniques and further implemented in a polymeric fluidic structure. This includes a 15 µL flow cell together with the required fluidic channels for sample and reagent fluid management. The four-electrode chip includes counter and pseudo-reference electrodes together with two working electrodes. These are sequentially modified with electropolymerized polypyrrole membranes that entrap the specific receptors for selectively detecting both target analytes. The analytical performance of both biosensors is studied by chronoamperometry, showing a linear range from 5 × 10 −6 to 1 × 10 −4 M (LOD of 3.2 ± 0.3 × 10 −6 M) and from 1 × 10 −7 to 1 × 10 −6 M (LOD of 6.7 ± 0.2 × 10 −8 M) for the l-lactate and the l-malate, respectively. Both biosensors show long-term stability, retaining more than the 90% of their initial sensitivity after more than 30 days, this being a prerequisite for monitoring the whole process of the malolactic fermentation of the red wines (time between 20 and 40 days). The flow system performance is assessed with several wine samples collected during the malolactic fermentation process of three red wines, showing an excellent agreement with the results obtained with the standard method. The malolactic fermentation (MLF) is a process in the winemaking industry in which bacteria convert l-malic acid into primarily l-lactic acid. The aroma and taste of many wines depend on this process, especially in red wines, but also in certain types of white wines. Besides, this process enables the stabilization of the wine colour, and it also allows its microbiological and bacterial control 1 . The control of the MLF during all the process (from 20 to 40 days) is crucial to obtain a highly qualified wine. Current standard methods 2 are applied in decentralized laboratories located far away from the wineries, meaning in long processes with associated high costs. They are based on chromatography and colorimetry and used bulky equipment which has to be used by highly skilled personnel. Enzymatic approaches based on absorbance detection of nicotinamide adenine dinucleotide (NADH) have been also proposed 3,4 , but they are also applied in external laboratories, meaning in more extra steps to carry out the process (sample uptake, sample stabilisation, sample transport and sample storage). This challenge should be solved for enabling on-time and in-situ corrective actions in the MLF process, to correct possible unpredictable problems. The miniaturization of analytical methods could be of high interest for this type of applications because they enable the integration of multiplexed analysis in low-cost and fast-response portable devices by requiring very low volume of reagents 5 . Compact and portable flow-systems should have an associated manufacturing and maintenance low cost to be competitive in the winemaking industry. Besides, the materials used to fabricate OPEN 1 Instituto de Microelectrónica de Barcelona (IMB-CNM), CSIC, Campus UAB, 08193 Bellaterra, Spain. 2 Institut Català de La Vinya i el Vi (IRTA-INCAVI), Plaça gora 2, 08720 Vilafranca del Penedès, Spain. 3 CIBER de Bioingeniería, Biomateriales y Nonomedicina (CIBER-BBN), Jordi Girona 18-26, 08034 Barcelona, Spain. * email: pablo.gimenez@csic.es; cesar.fernandez@csic.es them have to be cheap and tough, but also easy to machine. First approaches done in research for these devices used glass, ceramic and silicon 6 because they are easy to manufacture and very reproducible, but the integration of sensors or other flow elements is very complicated. To solve this challenge, polymers have been proposed from several decades ago for flow-systems because they are very low cost 7 . The most common polymer used in fluidic systems is the polymethyl methacrylate (PMMA) 8 because it is rigid, hard and very easy to manufacture by fast-prototyping techniques (i.e. milling and laser ablation) . Regarding the sensing part of the device, the electrochemical biosensors have been extensively reported as the best chance for the monitoring of analytical processes, including food control. They are also easy to integrate in portable analytical flow-systems for on-site analysis 15 . We previously reported on the development of individual amperometric biosensors for the detection of l-lactate and l-malate in batch. Both biosensors showed long-term working stability of more than 37 days. This is the key factor that enabled the application of these biosensors to monitor the MLF process. The biosensors' architecture included a thin-film electrochemical transducer selectively electromodified with a bienzymatic membrane, based on a three-dimensional matrix of electrogenerated polypyrrole (PPy). For the l-lactate biosensor (Fig. S1a, in the Supplementary Information-SI), the PPy membrane entrapped lactate oxidase (LOX) and horseradish peroxidase (HRP) as enzymes, while ferrocyanide (Fe(CN) 6 4-) in solution was used as redox mediator 16 . In the case of the l-malate biosensor (Fig. S1b, in the SI), the PPy membrane entrapped malate dehydrogenase (MDH) and diaphorase (DP) as enzymes, together with hexaammineruthenium (III) (HAR) as redox mediator, while β-Nicotinamide adenine dinucleotide (NAD + ) in solution was used as co-factor 17 . In both cases, the selected redox media was oxidized or reduced on the surface of the electrochemical transducer, and the faradic signal resulted from this process was associated to the concentration of the analyte in the sample. In this work, the production of a miniaturized flow-system integrating the above described biosensors is addressed for the very first time. The developed cost-effective flow-system could be readily used in field and in an automatic fashion, representing a significant advance in the precise monitoring of the malolactic fermentation in the winemaking industry. The simultaneous determination of l-lactic and l-malic in red wine samples with this flow-system is thoroughly assessed. A silicon chip comprising a platinum four-electrode electrochemical cell (counter, pseudo-reference and two working electrodes) was integrated in the compact flow-system. The working electrodes were selectively and sequentially modified with electrogenerated polypyrrole membranes to construct the corresponding on-chip biosensors. The resulting biochip was integrated in a multi-layered PMMA flow cell fabricated by a laser cutting process, which allowed the simple alignment of the chip with the fluidic reservoir and channels. Then, the bi-parametric fluidic system was applied to the monitoring of l-lactic and l-malic in samples collected along the MLF process of three different red wines and the results were compared with those recorded with the standard methods. ## Experimental Reagents and solutions. High pure (or analytical grade) reagents from Sigma-Aldrich (Spain) were used in this work. Deionized water was used to prepare the solutions. For the mechanical cleaning of the electrodes, ethanol 96% and 6 M sulfuric acid (H 2 SO 4 ) were used. The distillation of the pyrrole (reagent grade, 98%) was carried out once per week and then it was preserved at − 20 °C in the freezer. Potassium phosphate monobasic (KH 2 PO 4 ) was used to prepare a 0.05 M phosphate buffer (PB) solution (pH 7), which was used for the sensors fabrication and their characterization. The bienzymatic l-lactate biosensor was fabricated by using 10-µL aliquots of 1 U µL −1 Lactate oxidase (LOX, from Pediococcus sp., lyophilized powder, ≥ 20 U mg −1 solid), and then they were preserved at − 20 °C. The enzyme horseradish peroxidase (HRP, type VI-A, essentially salt-free, lyophilized powder, 250-330 U mg −1 solid) was preserved in a refrigerator at 4 °C as it was purchased. As redox mediator for the HRP enzyme, the potassium ferrocyanide (K 4 [Fe(CN) 6 ]) was employed. The bienzymatic l-malate biosensor was fabricated by using 15-µL aliquots of 5 U µL −1 Malate dehydrogenase (MDH, from porcine heart, freeze-dried material, ≥ 119 U mg −1 solid, Sorachim, S.A.) preserved at − 20 °C. β-Nicotinamide adenine dinucleotide hydrate (NAD + , ≥ 96.5% enzymatic, from yeast) and diaphorase (DP, from Clostridium kluyveri, lyophilized powder, 3-20 U mg −1 solid) were preserved in a freezer at − 20 °C as they were purchased. Every day, a 1-mL solution containing 0.1 M NAD + was prepared to be used as co-factor for the MDH. Moreover, the reagent hexaammineruthenium(III) chloride (Ru(NH 3 ) 6 Cl 3 , 98%) (HAR) was used as redox mediator for the DP. ## Devices and equipment. A 11 × 9-mm 2 silicon chip formed by four in-parallel platinum (Pt) microelectrodes was designed and fabricated by using standard photolithographic techniques 18 (Fig. S2, in SI). The larger electrode (2 × 2.5 mm 2 ) was used as counter electrode (CE), the two internal ones (1 × 2.5 mm 2 ) were used as working electrodes (WE 1 and WE 2) and the last one (1 × 2.5 mm 2 ) was used as pseudo-reference electrode (p-RE). The separation between adjacent electrodes was 0.6 mm and the contact pads were located 2.9 mm far from the electrode areas. The chip was inserted in a PMMA cell, designed by Corel Draw v.17 software and machined using a CO 2 -laser printer (Epilog Mini 24, Epilog Laser, USA). Two different cell architectures were mechanized. The first one, shown in Fig. 1a, was used for the batch sequential fabrication of both sensors, whereas the second one, shown in Fig. 1b, was used for the sensor analytical characterization under flow conditions. The chip can be used without any encapsulation process and was directly inserted in the cell where an integrated four spring-loaded connector (RS Components, Switzerland) was placed in contact with the electrode pads to enable contacting the chip with the measuring instrument. The distance from the electrodes to the contact pads enabled leaving enough room for the fluidic cell, providing an easy integration of the chip into the flow cell and a proper approach for connecting the cell to the measuring potentiostatic device. Both PMMA cells were formed by a 3-mm-thick PMMA bottom part fixed to a 0.5-mm-thick PMMA layer using double-sided PSA (175 μm thick) as adhesive. The 0.5-mm layer defined a 11 × 9 × 0.5-mm 3 well to host and align the silicon chip. The top PMMA part was 5-mm-thick. The one used for the sensor fabrication (Fig. 1a) defined a 50-µL chamber, which was aligned over the area of the electrodes. A 180-µm-thick PDMS layer was sandwiched between the bottom and top PMMA layers and the three layers were clamped together with four 1-mm diameter screws to avoid fluid leakage. During all the electropolymerization and activation steps, a 2-mm-diameter stainless-steel wire was used as CE and a 1.5-mmdiameter Ag/AgCl (3 M KCl) flexible Dri-Ref (World Precision Instruments, Sarasota, USA) was used as RE. The top PMMA part of the flow cell used for the sensor characterization (Fig. 1b) included several layers, which were fixed using 175-µm-thick double-sided PSA as adhesive. It comprised a 15-µL cell, two fluidic channels (1 mm width, 7 mm length) showing a thickness of 175 μm and 1-mm-diameter holes to enable fluidic connection between consecutive layers. Two fitting threads for connecting the fluidic inlet and outlet with external Teflon tubes (1.0-mm inner diameter, Teknokroma, Barcelona, Spain) were also included. A 180-µm-thick PDMS layer was also sandwiched between the top and bottom PMMA parts and fixed with four 1-mm diameter screws to avoid the fluid leakage. Here, the electrochemical cell comprised both biosensors together with the integrated Pt on-chip CE and p-RE electrodes. An image of the assembled compact analytical flow-system is shown in Fig. 1c. A cross-section of the fabricated bi-parametric compact analytical flow-system is shown in Fig. 1d. Conditioned samples were flowed inside the device by using a peristaltic pump (403U/VM3, Watson Marlow, UK) in this first approach. www.nature.com/scientificreports/ An Autolab workstation (PGSTAT-100 potentiostat-galvanostat, Ecochemie, Uthecht, The Netherlands) was employed to carry out the electrochemical measurements. The potentiostat was controlled by using the software NOVA v2.0 (Metrohm Autolab, Utrecht, Netherlands). ## Electrochemical procedures. Firstly, the electrodes were cleaned and activated as follows: 96% ethanol, 6 M H 2 SO 4 and deionized water were used to mechanically clean the surface of the electrodes, and subsequently they were electrochemically activated by cyclic voltammetry (20 scans from + 0.8 to − 2.2 V at 100 mV s −1 ) in a 0.1 M KNO 3 solution 18 . Then, the surface of both WEs was selectively electro-modified by applying the conditions optimized in our previous works, allowing the integration of the two biosensors on a single chip. A l-lactate biosensor was constructed in the WE 1 and a l-malate biosensor in the WE 2. In both cases, the PPy membranes were electrogenerated by fixing an overpotential of + 0.7 V (vs. Ag/AgCl) in a 50-μL 0.05 M PB solution at pH 7, which also contained 0.4 M pyrrole and 0.1 M KCl (named after generation solution). For the fabrication of the l-lactate biosensor, the generation solution also included 10 U of LOX and 200 U of HRP. The overpotential was applied until reach an accumulation charge of 500 mC cm −216 . For the electrosynthesis of the l-malate biosensor, a first membrane of PPy with an accumulated charge of 250-mC cm −2 was electrosynthesized. The generation solution used in this step also included 10 mM HAR(III) as redox mediator. Then, a second membrane of PPy with an accumulated charge of 500 mC cm −2 was generated by adding to the generation solution 45 U of MDH and 7.5 U of DP 17 . Afterward, the as-produced two-biosensor chips were rinsed with PB solution to remove the (bio)reagents physically adsorbed onto the PPy surface. Finally, they were preserved in a freezer at 4 °C in a PB solution when they were not in use. For ensuring a stable base line for both biosensors through all the measurements, the PPy membranes were overoxidized just after their electrosynthesis. This oversoxidation was carried out by cycling the potential from 0 to + 1 V 60 times at 100 mV s −1 in a PB solution 19 . The overoxidation process only had to be done once after the electrosynthesis of the biosensors, and it was not necessary to repeat the process during the life of the biosensors. The biosensor responses were based on the cascade (bio) reactions depicted in Fig. S1 (in the SI). All analytical measurements were carried out in the electrochemical flow cell and under stop flow conditions. The p-RE was positioned upstream the biosensors to avoid the potential changes caused by the enzymatic reactions on the biosensors. Initially, cyclic voltammograms (CVs) were recorded at 20 mV s −1 by injecting 125 μL of a 0.05 M PB solution (pH 7) containing 0.5 M KCl and all the other reagents required to complete the bi-enzymatic reaction for both biosensors. For the l-lactate biosensor, 1 mM l-lactate and 2 mM K 4 [Fe(CN) 6 ] as redox mediator were added. Regarding the l-malate biosensor, 1 mM l-malate and 5 mM NAD + as co-factor were added to the characterization solution. The solution also included 0.5 M KCl to minimize the potential drop and the hysteresis effects of the electrochemical processes, which were observed when biosensors were measured in compact cells under flow conditions 20 . Once the optimum operational potential applied for chronoamperometric measurements was set, calibration curves were performed in triplicate for both target analytes. l-lactate was measured in a concentration range between 1 × 10 -7 and 1 × 10 -3 M, whereas l-malate was measured in a concentration range between 1 × 10 -7 and 1 × 10 -5 M. The biosensors' performance was assessed in terms of sensitivity, linear range, limit of detection (LOD) and reproducibility of the fabrication procedure by using three biosensors fabricated under the same experimental conditions. Biosensor selectivity was evaluated in PB solutions containing the main interferences in wine samples (glycerol, glucose, gluconic acid, fructose, acetic acid, citric acid, ethanol, l-lactic, l-malic, tartaric acid and ascorbic acid) with a concentration of 5 × 10 -5 M or 5 × 10 -7 M for the l-lactate and the l-malate detection, respectively. A set potential of 0.35 V and − 0.4 V (vs Pt p-RE) was applied for the l-lactate and the l-malate detection, respectively. Concerning the working stability with time over long-times, both bio-chips were tested by calibrating them every 2 or 4 days in a concentration range of 1 × 10 -7 M-1 × 10 -5 M and 1 × 10 -6 M-1 × 10 -4 M for the l-malate and the l-lactate, respectively. All solutions and samples used for the evaluation of the bi-parametric compact analytical flow-system were flowed continuously during 30 s at 0.25 mL min −1 , in order to ensure that the previous solution filling the electrochemical flow cell was flowed out and replaced by fresh solution for carrying out the following measurement. ## Red wine samples from malolactic fermentation. Three different red wines provided by the Catalan Institute of Vineyard and Wine (IRTA-INCAVI) were selected to validate the developed flow-system for the simultaneous determination of l-malic and l-lactic. The wines were collected from the 2013 vintage and their vineyards were harvested in the region of Tarragona (Spain). The MLF process was induced in the samples after the alcoholic fermentation process by a strain of the specie Oenococcus oeni. A set of samples for each of the three wines were collected along the MLF process, in a concentration range from 0 to 8 × 10 -3 M (0-1.2 g L −1 ) for the l-malic, and from 0 to 6 × 10 -3 M (0-0.5 g L −1 ) for the l-lactic acid. The number of the collected samples was 10 of the Wine 1 (along 28 days), 5 of the Wine 2 (along 33 days) and 13 of the Wine 3 (along 45 days). All the samples were analysed with the bi-parametric compact analytical flow-system. The wine samples were diluted to adjust the l-malic acid and the l-lactic acid concentrations to the linear range of the individual biosensors obtained previously in our group (from 5 × 10 -6 to 1 × 10 -4 M and from 1 × 10 -7 to 1 × 10 -6 M, for the l-lactate and the l-malate, respectively) 14,15 . The wine samples were eventually diluted 1:10,000 by carrying out two intermediate dilutions of 1:100 in a 0.05 M PB solution (pH 7) containing 0.5 M KCl to adjust the l-malic acid concentration to the linear range of the biosensor. For the l-lactic determination, the dilution in a PB solution to adjust the concentration to the linear range of the biosensor for Wine 1 and Wine 2 was 1:50, meanwhile for Wine 3 was 1:20. The results obtained with the bi-parametric compact analytical flow-system were compared with them obtained by the standard enzymatic method applied by the IRTA-INCAVI. The standard method is based on www.nature.com/scientificreports/ the enzymatically catalysed reaction between the l-lactate or l-malate and the NAD + to produce NADH, whose concentration is stoichiometrically related to the analyte concentration in the sample. The change of the NADH concentration is measured spectrophotometrically at 340 nm 21 . Regarding the hydrodynamic performance of the bi-parametric compact analytical flow-system, the procedure used for the detection of both parameters is summarized as follows: Firstly, the wine sample was diluted in two 125-μL PB solutions for adjusting the concentration of the parameter to the linear range of each biosensor. Then, for the l-lactic detection, the ferrocyanide was added to one of these diluted solutions and was inoculated in the bi-parametric compact analytical flow-system during 30 s to totally clean the system from the last sample. The same procedure was repeated for the l-malic detection, but inoculating the diluted sample containing the NAD + . ## Results Biosensor fabrication. The use of electropolymerized membranes to produce the biosensor recognition layers on top of the electrodes allowed for the on-chip integration of both l-lactate and l-malate biosensors. The electrosynthesis of both PPy membranes was sequentially carried out in the electrochemical flow cell by applying potentiostatic conditions, as was described in the experimental section. The electrogeneration method applied avoided the cross contamination between the four integrated electrodes of the silicon chip. As can be seen in Fig. 2, both biosensors were selectively electrodeposited in the WE 1 and WE 2. The current profiles recorded under potentiostatic conditions during the potentiostatic electropolymerization of the PPy films are depicted in Fig. S3 (in the SI). The chip architecture facilitates the integration in a flow cell and can easily be connected via the embedded spring-loaded connectors to the measuring instrument. Millimetre-sized electrodes show suitable dimensions to work in a compact flow system like the one shown in this work with the electrode distance and electrochemical cell layout being adequate to avoid any electrochemical cross-talk due to the reactions taking place at the counter electrode on the biosensor devices and Ohmic drop. The biosensor fabrication was carried out sequentially without observing any chemical interference. The pseudo-reference electrode was placed upstream the biosensors to avoid any potential changes caused by the enzymatic reaction. Also, the counter electrode was placed downstream to avoid possible interference on the biosensor responses, as pointed out above. No chemical cross-reactions were observed in the biosensor responses, convincingly demonstrating the suitability of the architecture used to develop the chip. In both biosensors, the PPy membranes were overoxidized in order to obtain a stable base-line signal. This process consisted of carrying out 60 consecutive potential cycles between 0 and + 1 V at 100 mV s −1 in a 0.05 M PB solution (pH 7) 17 . The voltammetric signals (cycles 5, 30, 45 and 60) obtained during the overoxidation of the l-lactate and the l-malate biosensors are shown in Figs. S4a and S4b (in the SI), respectively. As can be observed, the non-faradaic current decreased with the increase in the number of CV until almost stabilization. This is associated with the electrochemical oxidative degradation of the conducting polymer 22 . Although the overoxidation of the polypyrrole membrane causes a loss of its conductivity or electroactivity, the moderate conditions applied in this work enabled to obtain a more stable background signal, which resulted in an improved analytical biosensor performance 23 . ## Analytical characterization of the bi-parametric compact analytical flow-system. The voltammetric responses recorded with the two on-chip biosensors in the electrochemical flow cell were used to fix the applied potential for the chronoamperometric characterisation of both biosensors. In the case of the l-lactate biosensor, after the addition of 1 mM l-lactate to a 0.05 M PB solution at pH 7 which also contained 0.5 KCl, an increase of the cathodic current at around − 0.35 V (vs Pt p-RE) occurred (Fig. S5a-in the SI). This increase was caused by the ferricyanide reduction, which was previously generated by the bienzymatic reaction involving the LOX and the HRP. For the l-malate biosensor, the addition of 1 mM l-malate and 5 mM NAD + to a 0.05 M PB solution (pH 7) containing 0.5 M KCl caused an increase of the anodic current (Fig. S5a, in the SI). This current increase was the consequence of the re-oxidation of the HAR(II) previously generated by the bienzymatic reac- From the obtained chronoamperograms in the electrochemical flow cell, the corresponding calibration curves of the two on-chip biosensors were constructed. Figure 3a shows that larger negative current densities were recorded along 120 s in solutions containing increasing concentrations of l-lactate. In all cases, the recorded currents for each analyte concentration became stabilized at around 90 s. Therefore, the mean value of the current density of the last 30 s was used as analytical signal for plotting the calibration curve for the l-lactate biosensor (Fig. 3b). Concerning the l-malate biosensor, similar response profiles were recorded but here higher anodic currents were recorded when the l-malate concentration increased in solution (Fig. 4a). The corresponding calibration curve was also plotted by using the mean current density value of the last 30 s vs. the l-malate concentration in solution (Fig. 4b). The l-lactate biosensor showed a sensitivity of (− 173 ± 8) × 10 2 µA M −1 cm −2 (r = 0.997, n = 7) in a linear range from 5 × 10 -6 to 1 × 10 -4 M and a LOD (3σ IUPAC criterion) of 3.2 ± 0.3 × 10 -6 M. A linear response for the l-malate biosensor was observed in a range from 1 × 10 -7 to 1 × 10 -6 M, with a sensitivity of (5.53 ± 0.6) × 10 2 mA M −1 cm −2 (r = 0.997, n = 5) and a LOD of 6.7 ± 0.2 × 10 -8 M. For concentrations above 1 × 10 -4 M of l-lactate and above 1 × 10 -6 M of l-malate, the respective enzymatic biosensors showed a saturation behaviour, this being in agreement with a process following a Michaelis-Menten kinetics. The reproducibility of the biosensors fabrication methodology was evaluated for each analyte by calibrating three different on-chip biosensors fabricated in three different silicon chips. The obtained relative standard deviation (RSD) of the biosensor sensitivity was lower than 8% and 6% for the l-lactate and the l-malate, respectively, meaning that the reproducibility of the fabrication method is very good. The selectivity and the working stability of both biosensors were studied in detail and reported in our previous works as is detailed in the experimental section 16,17 . It can be assumed that this is going to be analogous in the bi-parametric compact analytical flow-system reported here. Regarding the selectivity, any of the checked interferences showed a signal that could affect the analyte detection at the applied set potential for both biosensors. www.nature.com/scientificreports/ Concerning the working stability, the biosensors maintained more than 90% of their initial response during 52 days and 37 days for the l-lactate and the l-malate detection, respectively, meaning that these integrated biosensors can be used for the monitoring of long analytical processes, such as the MLF. Malolactic fermentation monitoring with the bi-parametric flow-system. Finally, the bi-parametric flow-system was applied to the monitoring of the malolactic fermentation of three different red wine samples collected along the MLF process. l-lactic and l-malic acids were determined consecutively in the red wine samples. Considering the required time for the electrochemical flow-cell filling (30 s) and the amperometric detection (120 s) carried out twice, the complete analysis of each sample took around 5 min. Figure 5 shows a comparison between the results obtained by the standard method and those obtained by the developed bi-parametric flow-system. There was an ideal agreement between both compared methods, with absolute errors below 0.15 g L -1 . Moreover, it is important to notice that all the experimental values obtained with the bi-parametric flow-system are within the 95% uncertainty range of the standard method. The evolution of both acid concentrations in the samples was as expected: at the beginning of the MLF process the l-malic acid concentration was high and there was not l-lactic acid. Along the process, the concentration of l-malic acid decreased and the l-lactic acid concentration increased. Finally, the MLF process finished when both concentrations were stable. The MLF process can take from a few days to a few weeks. The control of both parameters along the MLF process is essential to ensure the quality of the wine, because it is very important to determine the end-point of the process when the complete duck-out of the l-malic acid occurred. The appropriate control of the end-point avoids the organoleptic deviations caused by the growth of damaged microorganisms, implying a new inoculation of artificial strain from laboratories for restarting the process and consequently, it provokes additional cost and time for the process. Besides, the control of l-lactic acid concentration is also necessary because the maintenance of the ratio between both acids along the process determines that there are no unwanted parallel www.nature.com/scientificreports/ processes in the MLF. Only one bi-parametric fluidic system without re-generating the on-chip biosensors was employed during all the assays (including the calibrations of the system performed before and after the analysis of each wine), resulting in a total number of assays above 80. It was shown that both biosensors kept more than 91% and 93% of the initial sensitivity for the l-lactate and the l-malate, respectively. www.nature.com/scientificreports/ The proposed bi-parametric compact analytical flow-system reduces the number of reagents involved in the determination of both acids compared to the commercial enzymatic methods, because almost all the reagents are immobilized on the biosensor membrane, which only has to be re-generated after 37 days of intensive use. The volume of the reagents consumed is small, leading to less expense associated with the fermentation control. Moreover, the use of microfluidics enables to high dilute the wine samples, avoiding the potential matrix effects in the p-RE that could interfere in the determination of both acids. The flow-system also enables the fast and easy recalibration of the bio-chips before each analysis, meaning in a stable medium along all the assays. It is important to highlight that, in general, along the malolactic fermentation process, the l-malic acid concentration can decrease from around 3 g L −1 (2.2 × 10 -2 M) to 0 g L −1 (0 M). Nevertheless a concentration of l-malic below 0.3 g L −1 (2.2 × 10 -3 M) is considered as the end point of the MLF process in the winemaking industry. Therefore, the l-malic range can be actually considered from 2.2 × 10 -2 to 2.2 × 10 -3 M. Regarding the l-lactic acid, its concentration usually can increase from 0 g L −1 (0 M) up to 1.5 g L −1 (0.01 M). Thereupon, considering the general concentration ranges of these two biomarkers in red wines (from 2.2 × 10 -3 to 2.2 × 10 -2 M and from 0 to 1 × 10 -2 M, for the l-malic and the l-lactic, respectively) we could anticipate that the dilution factors would be between 1000 and 10,000 for all the possible acid concentrations in the winemaking industry. For the analyses of l-lactic, a sample dilution of 1:1000 will be carried out, whereas for the analyses of l-malic the sample will be diluted 1:10,000, in order to work within the biosensor linear concentration range. In this work, a different and more adjusted dilution ratio was used for both biosensors because the concentration range during the MLF process was previously determined by the standard method, but the general dilution factor could have been applied, too. As summary, the results in this work demonstrate the high potential of the developed compact analytical flow-system for the monitoring of both l-lactic and l-malic acids in the field of the winemaking industry. The accuracy of the proposed flow-system is appropriate for the considered application: the on-line control of both parameters in fermentation barrels of wines, which allows applying corrective actions in real time if it were necessary. Besides, the compacted size of the system will allow its easy integration within barrels, which will enable the biosensors calibration and the l-lactic and l-malic detection in an automated mode including the sample collection and dilution, the reagents inoculation, the analysis and the data processing. It results in an easier and more accurate method for the MLF process monitoring in comparison to conventional current methods without the fluidic part. Regarding the calibration of the system and the analysis of the real samples in the field, it will be as follows: the fluidic inlet of the compact analytical flow system could be easily connected to an automated and remotely controlled system of multi-valves, which will enable the flow of the rest of the reagents involved in the determination in an automatic mode. A total of six valves should be connected to 5 reservoirs containing (1) PB solution to dilute the sample, (2) l-lactate and (3) l-malate to calibrate the system every day, and (4) ferrocyanide and ( 5) NAD + to complete the bi-enzymatic reactions. The total volume consumed for every assay is very low (125 μL) and it is used for enabling the perfect renovation of the fluidic microchannels with next solutions, which avoids the memory effect of the sensors from previous assays by removing any trace of l-lactic or l-malic coming from them. From these numbers, and considering one calibration and one assay per day to control the malolactic fermentation process, if 10 mL containers for each solution were connected to the automated system, the microanalytical flow-system could work during all the malolactic fermentation process without any user intervention. Therefore, the proposed analytical flow-system will enable the control of the l-lactic and the l-malic acids on-site, creating a new portable and automated flow-system for the MLF process monitoring with breakthrough attributes. Comparison with other bi-parametric systems. Table 1 shows some characteristics of the developed bi-parametric compact analytical flow-system and other bi-parametric systems based on enzymes for simultaneous determination of l-lactic and l-malic in wines, previously reported. As can be seen, two systems use independent enzymatic reactors 24,25 , one for each parameter, two other systems are based on the immobilization of the enzymes in different membranes 26,27 and another one is based on the entrapment of the enzymes in a solid composite 28 . The proposed bi-parametric compact analytical flow-system in this work is the only one constructed by simultaneously entrapping both enzymes, and even the redox mediator, in an electrogenerated PPy membrane. This fabrication methodology enables the deposition of the required species on the same silicon chip without affecting the other integrated electrodes, meaning in an excellent approach for the fabrication of flow devices integrating microtransducers. As a result of this simplification, there is a reduction of volumes consumed of reagents and samples in comparison to the other works in the literature, which is especially interesting for reducing costs along the monitoring of long-term processes, as the MLF. All the other works use common species in solution for carrying out the detection of both target analytes: fluorimetric detection of NADH generated for l-LDH and l-MDH, amperometric detection of O 2 consumed for LOX, HRP and MDH/DP and the chronoamperometric detection of the redox mediator ferricyanide. This is the reason why all of them are combined with a FIA system to manage the liquids, except one which is performed in batch 28 . In the bi-parametric compact analytical flow-system presented in this work, two specific redox mediators, ferrocyanide and HAR, were used for each analyte. In the case of the l-malate biosensor, the HAR is incorporated in the PPy membrane, allowing a continuous flow analysis. The analysis time per sample is similar in all the cases, this being between 3 and 6 min. In the other hand, the system described in this work outperforms the other approaches in terms of limit of detection (LOD), especially in the case of l-malate. This may be partially related with the immobilization of the biochemical species in the conductive polypyrrole membrane synthesized under biocompatible conditions that may preserve the enzyme activity to a large extent. www.nature.com/scientificreports/ A system applied to the monitoring of the malolactic fermentation must show a long-term working stability under continuous use because the fermentation process takes around 30 days. Some of the systems in Table 1 show good operational stability values as long as 6 months. The system developed in this work maintains 92% of its initial sensitivity after more than 80 measurements in continuous use, with a lifetime of 37 days, being able to monitor the entire malolactic fermentation process. Finally, all the systems have been applied to the bi-parametric determination in finished wines samples, which are commercially available. However, the bi-parametric compact analytical flow-system presented in this work is the only one that has been assessed using real samples collected during the malolactic fermentation of red wines. This compact analytical flow-system integrates for the first time both sensors on a new automated flow platform, resulting in a simply system for the malolactic fermentation monitoring. It represents a significant advance in the precise monitoring of the process in the winemaking industry because this novel research will provide for the very first time a cost-effective and easy-to-use way to determine l-lactic and l-malic on-site and on-line based on this new concept of test for winemaking control with unique and unprecedented attributes. No other previous reports are focused on the on chip integration of both l-malate and l-lactate biosensors in a flow-system that could show the potential of these devices for monitoring the malolactic fermentation of red wines in field and in an automatic fashion. ## Conclusions A bi-parametric compact analytical flow system for the simultaneous determination of l-lactate and l-malate was designed, fabricated and optimized. The development of a bi-parametric compact system included the design and the fabrication of a 11 × 9-mm 2 silicon chip comprising two working electrodes, together with a counter and a pseudo-reference electrode, all made of platinum. These miniaturized electrochemical sensors are cheap, highly reproducible and robust. The working electrodes were modified with polypyrrole-based enzymatic membranes to produce the two on-chip biosensors for l-lactate and l-malate target analytes. The use of an electropolymerization approach enabled the strict controlled deposition of the required enzymes over the selected electrodes. The biochip did not require encapsulation and wire-bonding and thus allowed for the simple integration in a robust miniaturized PMMA flow cell, mechanized by using rapid prototyping techniques. The resulting bi-parametric compact analytical flow-system showed superior analytical and operational features for the determination of l-lactate and l-malate and was eventually assessed in red wine samples collected during the malolactic fermentation process, showing a good agreement with the results obtained for both analytes with the standard colorimetric methods. As summary, the overall novelty of the proposed flow-system is twofold: (1) it has the ability to fully process the sample and the reagents required for the malolactic fermentation control, by using a low-cost and robust miniaturized PMMA flow cell. The design of the flow cell reduces drastically the complexity of current traditional tests, meaning in a method with a total analysis time of 5 min and a required sample volume of 125 μL per analysis. Besides, the automated flow cell avoids the human error and the potential contamination of barrels derived from the manipulation of the sample and reagents. (2) The flow-system integrates both l-lactate and l-malate sensors in the same chip, which allows the reduction of reagents and sample consumption. Moreover, the monitoring of both parameters using a compact system is proposed for the very first time in literature, meaning in a novel system for the fast and total control of the malolactic fermentation in field, in real-time and on-site in the winemaking industry. The proposed flow-system would set a new paradigm for integration of electrochemical sensing in field by solving the limitations of current methods applied in the winemaking industry. This approach would open new opportunities for in-situ and real-time control of processes by making accessible to untrained personnel an easy handling method for fermentation monitoring. It would also facilitate better management to winemakers and the application of corrective actions, if required, allowing unprecedented time and cost savings.

chemsum

{"title": "Compact analytical flow system for the simultaneous determination of l-lactic and l-malic in red wines", "journal": "Scientific Reports - Nature"}

gold–carbonyl_group_interactions_in_the_electrochemistry_of_anthraquinone_thiols_self-assembled_on_a

## Abstract: New anthraquinone derivatives with either a single or two thiol groups (AQ1 and AQ2) were synthesized and immobilized in self-assembled monolayers (SAMs) on Au(111) electrodes via Au-S bonds. The resultant AQ1-and AQ2-SAMs were studied by cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS), which enabled mapping of the gold-carbonyl group interactions and other dynamics in the Au-S bound molecular framework. Understanding of these interactions is important for research on thiol-coated gold nanoclusters, since (I) anthraquinone derivatives are a major compound family for providing desired redox functionality in multifarious assays or devices, and (II) the gold-carbonyl interactions can strongly affect anthraquinone electrochemistry. Based on equivalent circuit analysis, it was found that there is a significant rise in polarization resistance (related to SAM structural reorganization) at potentials that can be attributed to the quinone/semi-quinone interconversion. The equivalent circuit model was validated by calculation of pseudocapacitance for quinone-to-hydroquinone interconversion, in good agreement with the values derived from CV. The EIS and CV patterns obtained provide consistent evidence for two different ECEC (i.e. protoncontrolled ET steps, PCET) pathways in AQ1-and AQ2-SAMs. Notably, it was found that the formal reorganization (free) energies obtained for the elementary PCET steps are unexpectedly small for both SAMs studied. This anomaly suggests high layer rigidity and recumbent molecular orientation on gold surfaces, especially for the AQ2-SAMs. The results strongly indicate that gold-carbonyl group interactions can be controlled by favorable structural organization of anthraquinone-based molecules on gold surfaces. ## Introduction There is presently a great interest in gold nanoclusters, i.e. gold in sub-nanoparticle size in a variety of shapes. 1 These span polyhedral structures containing from a few to hundreds of gold atoms. Gold nanoclusters are intensively investigated due to their potential exploitation e.g. as catalysts 2 and to singleelectron charging and other quantum effects. 3,4 One of the "traditional" methods of gold nanoparticle synthesis and stabilization is the use of coating compounds terminated with thiol groups, which is comprehensively exploited in the creation of gold nanoclusters. 1 The organization of dithiol-type compounds on gold nanoclusters in comparison to compounds with only a single thiol group is further important from fundamental points of view. The interaction and surface distribution of added functional groups is of particular interest, 1 which would be different for mono-and dithiol compounds. Additional functionalization provides other useful approaches for fne-tuning of gold nanocluster properties, and for controlled aggregation of the functionalized Au-clusters into new "smart" materials. 1,4 One feasible step towards assessing such interactions is the use of anthraquinone derivatives. Anthraquinones with two thiol linkers are attractive for molecular electronics, 5 due to their molecular redox functionalities, and ease of tuning properties by adding functional groups through well-established organic syntheses. The putative application of anthraquinone thiol derivatives as molecular linkers or redox probes also relates to the great importance of quinone electrochemistry in bioenergetics. 9 Future devices based on hybrids of anthraquinone thiols with gold nanoclusters could therefore potentially be utilized e.g. in investigation of electron transfer (ET) kinetics in electrochemical systems based on redox enzymes. 10,11 The quinone-to-hydroquinone conversion process also depends strongly on pH, since each ET step is accompanied by proton transfer (PT). Besides the proton concentration and the surrounding medium, the electrochemical conversion of carbonyl groups can also be affected by the nature of the gold surfaces. 15 In-depth understanding of these interactions in the context of anthraquinone-gold systems is thus highly important. In this report we present a study of the electrochemical properties of self-assembled molecular monolayers (SAMs) of in-house synthesized mono-and dithiol anthraquinone derivatives (denoted as AQ1-and AQ2-SAMs, Scheme 1) assembled on Au(111) electrode surfaces. Deposition of the compounds on a single-crystal (i.e. atomically flat) gold surface enabled frst the recording of interfacial faradaic processes of both the carbonyl group and the surface Au-S linking units. Secondly, electrochemical features related to structural reorganization events in the whole AQ-SAMs could be recorded. The assignment of these features to specifc reactions is based on cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS). Particularly, the EIS data combined with equivalent circuit analysis allowed us to assess at the same time changes in pseudocapacitance and polarization resistance (R p ). The calculated pseudocapacitances based on equivalent circuit analysis are consistent with those that correspond to faradaic process as obtained from CV, thus supporting the proposed model. A signifcant change in R p is an indicator of structural reorganization in the layer subjected to electric felds. 16 An overall electrochemical assessment of gold-carbonyl group interactions framed within the quinone-to-hydroquinone interconversion process is provided. We particularly focus on the carbonyl group proximity to gold which triggers specifc structural SAM reorganization, as well as solvent and intramolecular reorganization that accompanies the "elementary" PCET steps. Very notably the latter was found to be unexpectedly small and much smaller than reorganization free energies commonly encountered for electrochemical ET processes. As noted, the understanding of gold-carbonyl interactions is also more broadly important for future electrochemical investigations of anthraquinone compounds assembled on gold nanoclusters. ## CV of the quinone group Cyclic voltammograms of AQ2 SAMs show two reversible electrochemical reactions (referred to as reactions ( 1) and ( 2)), Fig. 1A. Both reactions depend on pH, and their formal potentials shift with pH (ca. 60 mV pH 1 ) following the Nernst equation for a one-electron-one-proton (PCET) process, Fig. 1B. Since the overall quinone/hydroquinone conversion in aqueous buffer is a 2-proton/2-electron process, 13 the observed oneelectron-one-proton features must be part of a two-step process each step with a PCET-type mechanism. 17 Reaction ( 2) is therefore likely to be related to the formation of semiquinones, and reaction (1) to the fully reduced AQ2 hydroquinone state. The peak related to semi-quinone formation has in fact been observed for quinone flms at high pH, 18 or in CV of freely mobile quinones in unbuffered solutions. 14 On the other hand, reaction (2) for anthraquinone derivative SAMs in buffered solutions at moderate pH should not be observed. This is the case with our reference compound (AQ1-SAM), insert in Fig. 6. This feature therefore not only depends on the proton concentration, but is also likely to depend on the particular organization of the attached quinones at the surface. The latter is in turn controlled by molecular interactions within the layer Scheme 1 The reactions and resultant chemical structures of AQ2 (5) and AQ1 (6) compounds. and by specifc interactions with the electrode surface. Signifcantly different peak capacitances for reactions (1) and (2) are therefore observed, which we address further below using frequency response analysis. ## Reductive desorption of the Au-thiol surface bond Different structural organization of the AQ1-and AQ2-SAMs is expected, as AQ1 binds only via a single Au-S bond, while AQ2 can bind either by a single or two Au-S bonds. AQ2 is therefore expected to form a structurally more rigid surface layer than AQ1. Pronounced differences between the AQ1-and AQ2-SAMs on Au(111) electrodes is in fact substantiated by reductive desorption voltammograms recorded for 0.1 M NaOH, Fig. 2. AQ1-SAMs exhibit a single dominating sharp feature at ca. 0.15 V (actual desorption) and a satellite feature at ca. 0.10 V (arising due to electrode roughness), according with typical voltammograms for desorption of alkanethiol monolayers. 19 AQ2-SAMs exhibit less well-defned features, likely to originate from e.g. sequential desorption or more constrained angular degrees of freedom, both arising from the prevalence of two binding sites. The estimated surface coverages (mol cm 2 ) from reductive desorption of AQ1-and AQ2-SAMs are ca. 6.8 10 10 and 7.4 10 10 , respectively, assuming single-ET for AQ1 and two-ET for AQ2 in the desorption processes. This supports that there are in fact two binding sites for AQ2, but only a single binding site for AQ1. The estimated surface coverages are close to the expected coverage of a dense monolayer (with interchain spacing of z5 ) of 7.76 10 10 mol cm 2 on a (111)-facet. 20 ## AQ-SAMs surface imaging Scanning tunnelling microscopy (STM) in both the ex situ and the electrochemical in situ modes were undertaken. Highresolution images could be recorded, but their interpretation remains presently elusive. As can be seen from Fig. S1, † cluster formation was consistently observed, resulting in signifcantly disordered adlayers. Such structures can strongly affect tunnelling current pathways in different ex situ and in situ STM tip/target molecule/gold surface systems. However, below 150-200 , the resolution was not sufficient to distinguish singlemolecule features or surface orientation of the anthraquinone thiols. This could be attributed to a persistent tendency of anthraquinone thiol derivatives for p-stacking competing with Au-S surface bonding. The synthetized molecules have relatively short side chains, and strong p-p interactions are to be expected. We therefore focused further attention on electrochemical kinetic and frequency response analysis in the assessment of gold-carbonyl group interactions. ## Frequency response analysis EIS for AQ1-and AQ2-SAMs was recorded to obtain insight in the nature of reaction (2) and its relation to reaction (1). The EIS experiments were designed to complement optimally the CV data. An example of AQ2-SAM frequency response and equivalent circuit is shown in Fig. 3. The equivalent circuit is analogous to the typical circuit used to describe faradaic impedance of strongly adsorbed molecular entities. 12 The capacitive element related to the faradaic process was modifed by utilizing a constant phase element (CPE). This small modifcation enabled a far more flexible ft to the wide potential range than utilization solely of a capacitor. Additionally, the use of a CPE offers insight into the roughness of the surface. 21 The complex CPE admittance Y(u) can be described by the equation: where j ¼ (1) 1/2 , n is a phenomenological number describing the deviation of the CPE from an ideal capacitor, and Y 0 the magnitude of the admittance (S s n ). The phase angle of the CPE is independent of frequency and has a value of (90 n) . Y 0 is thus an ideal capacitor if n ¼ 1, and an ideal resistor if n ¼ 0. In the context of surface roughness, the change in the value of n from 1 to 0.5 can be interpreted as the change from a perfectly flat to a highly rough surface. 21 Y 0 can be converted to capacitance (C r ) using the following equation: where f mf is the frequency of maximum phase angle. The conversion of Y 0 to C r using eqn (2), is based on the assumption that the imaginary part of the CPE impedance equals the impedance of ftted capacitance for a given frequency range (eqn S1 †). Based on the applied model (inset in Fig. 3 The C r values (after correction for C dl as noted) can be directly used for the validation of the applied equivalent circuit. As seen from Fig. 4, the calculated specifc capacitances for reaction (1) are quite similar to the values derived from CV, thus justifying the application of the proposed equivalent circuit for the description of faradaic impedance of adsorbed anthraquinone thiols. Fig. 5 shows potential induced changes in CPE admittance (and thus in C r ). There is a signifcant admittance rise in reaction (1) for both AQ1-and AQ2-SAMs (0.15-0.20 V), and a much smaller change for reaction (2) (ca. 0.35 V). These changes are consistent with the CV features obtained (inset in Fig. 6). As noted, n gives a rough estimate of the adlayer deformation. The particular numerical values of this parameter are shown in the brackets for selected potentials, Fig. 5. The change in n with applied potential for AQ2-SAMs, is only signifcant for reaction (2) suggesting a major structural reorganization in a narrow potential range. The n values for AQ1-SAMs are signifcantly lower than for both bare and AQ2-coated Au(111) electrodes, making similar assessment challenging. This fnding might be indicative of more complex intermolecular interactions in the AQ1 adlayers. The plot of R p against applied potential implies that two different EC (ET/PT) reaction pathways specifc for AQ1-and AQ2-SAMs operate, Fig. 6. The kinetics for freely mobile anthraquinone-type compounds was found to proceed by EECC (pH z 10), ECEC (at pH 7-4) and CECE (pH 1 and below) mechanisms (where E is ET and C is PT in the sequence). 13 We could not obtain sufficient electroactivity above pH 7 for the SAMs here, and the highest electroactivity was found at pH z 4.5. The latter fnding is reflected in the observed pH dependence of the current densities for both reactions (1) and ( 2), being similar at pH 7 and pH 1, but notably higher at pH 4.5. The apparent pK a of the frst protonation step is around 4 for the ECEC mechanism. 13 The maximum current density can therefore be associated with the maximum concentration of semiquinone at pH 4.5 compared to pH 7. In the case of pH 1, the apparent pK a of the frst protonation step is estimated to be below 1 for the CECE mechanism. 13 It can then be suggested that the gold-carbonyl group interactions may interfere more strongly with the frst CE step than with the corresponding EC step, resulting in the different current densities at pH 4.5 and pH 1. Since the ECEC mechanism likely prevails at pH 4.5, and the relaxation between the ET and the PT steps is fast, the two distinct reorganization events in the AQ1-SAM structure (separated by ca. 0.15 V) can be assigned to two consecutive EC steps, resulting in a single broad CV redox wave. The assigned EC steps (i.e. reactions (1) and ( 2)) of AQ2-SAMs are separated by a larger potential difference of ca. 0.25 V. The change in R p associated with reaction (2) is signifcantly higher than for reaction (1), although the change in admittance and pseudocapacitance for reaction (2) is still very small. ## Electrochemical kinetics analysis We attempted frst to estimate the interfacial electrochemical ET rate constants (k ox /k red ) for AQ1-and AQ2-SAMs by a Laviron analysis and the Butler-Volmer limit of the current/ overpotential (i/h) correlation, Fig. 7: F, R and T have their usual meaning, a is the transfer coefficient, h the overpotential, and k 0 the standard rate constant at h ¼ 0 V. The apparent i/h correlations for the anodic and cathodic processes appear symmetric around h ¼ 0 V, but approach a quadratic form already at very small overpotentials |h| < 0.1 V. In accordance with electrochemical molecular charge transfer concepts and theoretical concepts and formalism introduced by Marcus, Hush, Gerischer, and particularly by Levich, Dogonadze, Kuznetsov and associates, the latter correlation can be represented as: from which the formal reorganization (free) energy (l) of the PCET elementary steps can be estimated. l represents the change in low-frequency solvent and intramolecular structures, and is distinct from the structural reorganization in the SAM detected with EIS discussed above. The summary of this analysis is given in Fig. 7. Slightly asymmetric Tafel plots for AQ1-SAM were obtained, in contrast to AQ2-SAM, suggesting that the layer of the latter is less prone to molecular structural changes in the PCET steps. Notably, l was found to be only ca. 0.05 eV for the AQ1-SAM and 0.02 eV for the AQ2-SAM. l for reaction (2) could not be determined, due to difficulties in reaching the current plateau region, even at high scan rates. The difference in apparent l could indicate that the carbonyl groups are closer to the electrode surface for AQ2-SAMs than for AQ1-SAMs, which is supported by the specifc faradaic resistance (U cm 2 ) of ca. 6.1 10 5 for the AQ1-SAM and 3.5 10 4 for the AQ2-SAM (Fig. S3 †). The values of l obtained are, however, very small and correspond to almost step-like transition from the Butler-Volmer to the activationless overpotential region. More importantly, the emerging limiting slopes at small overpotentials which represent the electrochemical transfer coefficient, a are quite different from the input values in the Laviron forms (e.g. 1.6 vs. 0.7, Fig. S4 and S5 †). These observations prompt alternative considerations given below. Consistent use of the Laviron and Butler-Volmer formalism rests on the notion of strong electronic-vibrational coupling and large reorganization free energies, l [ k B T. The observed current rise from thermal to activationless behavior is far too abrupt to be compatible with the broadly observed much smaller curvature in both simple electrochemical processes and ET processes in homogeneous solution. To account for step-like i/h behavior the notion of weak electronic-vibrational coupling can instead be proposed. In this limit the i/h correlation is dominated by the step-like Fermi function in the electrochemical rate constant rather than by the Gaussian molecular vibrational energy density form as in "normal" electrochemical ET processes. This difference can be illustrated by more detailed rate constant forms that incorporate contributions from all electronic levels of the metal electrode and not only from levels around the Fermi energy. The current density, here cathodic is: where f(3 3 F ) is the Fermi function of the metallic electronic energy spectrum 3, and 3 F the Fermi energy. A is a h-independent constant available from molecular charge transfer theory. 31 f(3 3 F ) and the vibrational distribution function g(3 3 F ; h) depend strongly on 3, and dominate the i/h correlations with contributions from all electronic levels of the metal electrode, at higher overpotentials: f(3 3 F ) has a step-like functional form, changing from unity to exponentially small values over an energy range of a few k B T. The vibrational distribution g(3 3 F ; h) is given the normal Gaussian form. The Gaussian width is D ¼ 2Olk B T z 0.2-0.3 eV for l ¼ 0.5-1.0 eV as in "normal" strong-coupling electrochemical ET. This limit implies that the current varies from quadratic to activationless overpotential dependence over a range of $0.5 V. Since the Gaussian width 2Olk B T signifcantly exceeds the "width" of the Fermi function the current monitors essentially the Gaussian vibrational energy density up to overpotentials around l. In the opposite limit of weak coupling, the Gaussian width approaches the "width" of the Fermi function, say D ¼ 2Olk B T z 0.04-0.07 eV or (1-2) k B T for l ¼ 0.02-0.05 eV, Fig. 7. In the limit of very weak coupling g(3 3 F ; h) would assume a Lorentzian form. 31 Current is not recorded in these limits, until the overpotential has taken the maximum of the, now very narrow Gaussian or Lorentzian g(3 3 F ; h) function up to the Fermi level, with a very narrow hrange changing the current from "normal" to activationless behavior. What is recorded in the i/h correlations then, is essentially the Fermi function f(3 3 F ) and not the vibrational "bandshape" g(3 3 F ; h). Fig. 8 illustrates the difference between the strong-and weak-coupling limits. Kuznetsov has provided a quantitative formalism in the weak-coupling limit both for ET in homogeneous solution and electrochemical ET processes. 32 A detailed formalism for analogous optical electronic transitions, for which the weakcoupling Lorentzian bandshape limit is much more common, is also available. 33 It thus appears that the i/h correlations obtained accord formally with weak electronic-vibrational coupling but poorly with the much more commonly encountered limit of strong electronic-vibrational coupling. The question regarding physical reasons, why the coupling should be weak particularly for the thiol-derived anthraquinones bound to the Au(111)-electrode surfaces via strong Au-S chemisorption then arises. Based on the results from CV and EIS, it can be suggested that the signifcant structural reorganization in AQ2-SAMs in reaction (2) can be attributed to the impeding effect of goldcarbonyl group interaction on the semi-quinone formation. This is different from the nuclear reorganization in the elementary faradaic processes and resembles autoinhibition in electrochemical systems where a mercury electrode surface is saturated with target adsorbate compounds. 34,35 Together with the unexpected voltammetric behavior in the i/h pattern for both AQ1-and AQ2-SAMs (Fig. 7), with a sharp ((1-2) k B T) transition between "normal" and activationless i/h behavior, the major fndings of our study can then be summarized as: (I) The anthraquinone molecules in the AQ2-SAMs are in close proximity to the electrode surface leading to strong goldcarbonyl group interaction, in contrast to anthraquinone molecules in the AQ1 adlayers. (II) Gold-carbonyl interactions create an energy barrier, leading to a split CV redox wave (denoted as reactions ( 1) and ( 2)) and a narrowing of the peak attributed to reaction (1) (inset in Fig. 6). (III) The strong gold-carbonyl group interactions in the AQ2-SAMs are supported by the Nernstian pH dependence of both reactions ( 1) and ( 2) (Fig. 1), as well by the signifcant difference Fig. 8 The dependence of normalized transition probability (with respect to h ¼ 0 V) and the Fermi function on the electronic energy (A), together with calculated normalized i/h relationship (B), for selected lvalues. The transition probability was calculated using E g 8, and the i/h relationships using a reformulation of eqn (6) (eqn S2 †). in the R p -potential dependence between AQ1-and AQ2-SAMs (Fig. 6). (IV) The clear R p changes can be interpreted as structural reorganization events in the AQ-SAMs. (V) Structural reorganization in reaction ( 2) is signifcantly more pronounced than in reaction (1) (Fig. 6). (VI) Pseudocapacitance for reaction ( 2) is almost negligible compared to reaction (1) (Fig. 5). (VII) A very small environmental reorganization (free) energy accompanies the PCET processes in both AQ1-and AQ2-SAMs (Fig. 7 and 8). Regarding point (V), it might be speculated that prior to reaction (2), anthraquinone molecules in the AQ2-SAM interact with gold either via a single or both carbonyl groups. These binding modes could result in SAMs composed of specifc differently organized domains (hypothetical State 1). The STM data offer some support for such a view, Fig. S1 † After completion of reaction ( 2), the resulting semi-quinones maintain interaction with gold via a single remaining carbonyl group (hypothetical State 2). The structural reorganization from State 2 into fully reduced AQ2-SAM might be lower in reaction (1) than in reaction (2), since the molecular orientation in State 1 is more random than in State 2. Point VI suggests that the appearance of reaction (2) in the cyclic voltammograms of AQ2-SAMs is of complex nature. As a comparison, intermolecular interaction of sulfonated anthraquinones gives narrow spike-like CV features. 34 This is in contrast to the observed broad reaction (2) CV features of AQ2-SAMs, and probably associated with greater disorder of State 1 in the AQ2-SAM, than in layers composed of loosely adsorbed anthraquinones at mercury electrodes. 34 Correspondingly the transition from State 1 to State 2 could result in a decrease of AQ2-SAM compactness, which might rationalize that the observed apparent rate for reaction (2) is about twice higher than for reaction (1) (Fig. 7B). The notable observation in point VII remains open. AQ2 layers are expected to be more rigid than AQ1 layers, due to single Au-S binding for AQ1 and binding by either a single or two Au-S bonds for AQ2, reflected in around twofold higher apparent l and about an order of magnitude higher faradaic resistance for AQ1 than for AQ2. On the other hand, the apparent l-value for AQ1-SAMs is also "unexpectedly small". Furthermore, the ET distances for both AQ1 and AQ2 SAMs are small (i.e. below 1 nm), and the possibly of planar (or close to planar) molecular orientation could indeed result in small reorganization energies. It can therefore be proposed that overall AQ-SAMs rigidity and close proximity of anthraquinone molecules to the electrode surface would result in the small l observed. "Small" apparent l-values for "simple" electrochemical ET processes are known for the mammalian heme redox protein cytochrome c 36,37 and the bacterial blue copper protein azurin. 38,39 These values are, however, still signifcantly larger, z0.25 eV or so, than the apparent l-values presently observed. Although structurally "small", these proteins are also still complex molecules compared with AQ1 and AQ2, and offer options for more complex, multi-step electrochemical ET that involve e.g. structural gating, pre-organization in the protein conformational systems, or other overpotential independent elementary steps that could lower the apparent l-values in the overall process. A second rationale for small l-values could be that quinoneto-hydroquinone interconversion involves both ET and PT in an overall PCET process. PCET processes can involve all degrees of coupling between the elementary ET and PT steps. 40,41 ET and PT can be independent, vibrationally fully relaxed events, each involving charge transfer and signifcant environmental reorganization, but the steps are "coupled" in the sense that a given, say ET step affects the kinetic parameters of the subsequent PT step, or vice versa. In the opposite limit ET and PT are fully coupled invoking the character of the quinone-to-hydroquinone conversion as a hydrogen atom transfer process. An electrostatically neutral particle is then transferred, with little solvent reorganization. This expectation also applies when the time sequence between separate ET and PT steps is shorter than the solvent relaxation time (z10 11 s) in the interfacial electrode surface region. All the limits can be considered in the AQ1 and AQ2 processes, but strong coupling between the ET and PT steps is needed to rationalize the small reorganization energies observed. Reorganization in the intramolecular nuclear modes would not be reflected conspicuously in the i/h correlations, as the appropriate C-C, C-H and O-H modes involve high vibrational frequencies represented by nuclear tunnelling in the preexponential factor of the current density forms rather than in the h-dependent activation factors. 31 ## Conclusions We have synthesized new thiol-derived anthraquinones with both a single (AQ1) and two (AQ2) thiol groups linking the molecules in SAMs to a single-crystal, atomically planar Au(111)-electrode surface via either a single or two Au-S bonds. We have explored the electrochemical SAM behavior using electrochemical techniques particularly CV and EIS. The electrochemical studies addressed voltammetry and EIS of both the quinone and the -SH moieties as well as crucial potential dependent structural reorganization events of the surface bound target molecular SAMs. Overarching objectives were, frst to introduce a class of challenging electrochemical probe molecules with prospects as building blocks in new "smart" materials as hybrids with Au-nanoparticles and in other ways. The anthraquinone thiol SAMs on Au(111) electrodes offer, secondly sensitive probes for fundamental structural reorganization studies arising from low single-crystal surface roughness and a direct dependence of redox center proximity to gold on the number of binding sites. The reorganization events can therefore also be probed by potential dependent polarization resistance. Faradaic monolayer CV and EIS analysis based on interfacial capacitance and resistance, and interfacial electrochemical ET rate constants has led to a coherent view of the elementary electrochemical ET processes and other elementary reorganization steps that accompany the conversion between the fully oxidized and fully reduced AQ1 and AQ2 target molecules. It was found, notably that gold-carbonyl group interactions effectively impede formation of semi-quinones, which results in signifcant reorganization events that can be attributed to specifc EC steps. Unexpectedly and also notably, very small apparent lvalues were observed for both AQ1-and AQ2-SAMs, probably associated with the layer rigidity, close proximity of anthraquinone thiols to the electrode surface, as well as strong coupling between the ET and PT steps. The outcome of the study has disclosed novel features of thiol-derived anthraquinones based on different electrochemical techniques targeting the interactions of both the molecular quinone and the thiol moieties with the single-crystal Au(111)-surfaces. The work offers other steps towards understanding of thiol-derived quinones also bound to gold nanoclusters which may have more direct impact in molecular scale electronics than planar electrode surfaces. The apparently weak electronic-vibrational coupling might here hold advantages by reduced thermal broadening and noise in the electronic functions to be targeted. ## Chemicals All reagents and materials were purchased from well-known chemical suppliers and used without further purifcation. Aqueous solutions were prepared with freshly deionized water (18.2 MU cm resistivity) obtained with the Sartorius ultrapure water system. Synthesis of 2,6-bis(3-hydroxyprop-1-yn-1-yl)anthracene-9,10dione (3) The compound was prepared using Sonogashira coupling conditions between a terminal alkyne species and an arylbromide. The reaction was performed under inert conditions to prevent undesired homolytic coupling of the terminal alkyne. In an oven-dried, condenser-equipped and degassed round-bottomed flask, containing a magnetic stirring bar, a 1 : 1 solution of tetrahydrofuran (THF, 44 mL) and diisopropylamine (DIPA, 44 mL), the crystalline compounds 2,6dibromoanthracene-9,10-dione 1 (0.354 mg, 0.968 mmol, 1 equiv.), CuI (9.2 mg, 48.4 mmol, 0.05 equiv.), PdCl 2 (PPh 3 ) 2 (13.6 mg, 19.4 mmol, 0.02 equiv.) and PPh 3 (12.7 mg, 48.4 mmol, 0.05 equiv.) were dissolved, followed by 30 minutes of degassing. Propargyl alcohol (170.0 mL, 2.904 mmol, 3 equiv.) was added dropwise to the solution and the reaction mixture was set to stir for 24 h under reflux. The reaction was quenched with H 2 O (50 mL) followed by three consecutive extractions with Et 2 O (3 25 mL). The organic phase was collected and dried over Na 2 SO 4 , fltered and concentrated in vacuo. The remaining grey precipitate was recrystallized in EtOAc and fltered, isolating 3 (232.6 mg, 76%); 1 H NMR (400 MHz, DMSO-d 6 ) d 8.17 The compound was prepared using Sonogashira coupling conditions between a terminal alkyne specie and an arylbromide. The reaction was performed under inert conditions to prevent undesired homolytic coupling of the terminal alkyne. In an oven-dried, condenser-equipped and degassed roundbottomed flask, containing a magnetic stirring bar, a 1 : 1 solution of THF (36 mL) and DIPEA (36 mL), the crystalline compounds 2-bromoanthracene-9,10-dione 2 (294 mg, 1.03 mmol, 1 equiv.), CuI (9.79 mg, 51.4 mmol, 0.05 equiv.), PdCl 2 (PPh 3 ) 2 (14.4 mg, 20.6 mmol, 0.02 equiv.) and PPh 3 (13.48 mg, 51.4 mmol, 0.05 equiv.) were dissolved, followed by 30 minutes of degas. Propargyl alcohol (178.0 mL, 3.09 mmol, 3 equiv.) was added dropwise to the solution and the reaction mixture was set to stir for 48 h under reflux. The reaction was quenched with H 2 O (50 mL) followed by three consecutive extractions with Et 2 O (3 25 mL). The organic phase was collected and dried over Na 2 SO 4 , fltered and concentrated in vacuo. The remaining grey precipitate was recrystallized from EtOAc, washed with heptane and fltered, isolating 4 (179 mg, 66%); 1 The compound was prepared from 3, using Mitsunobu conditions to carry out a thioesterifcation of the primary alcohols. In an oven-dried and degassed round-bottomed flask, containing a magnetic stirring bar, 3 (200 mg, 0.63 mmol, 1 equiv.), dry THF (50.4 mL), diethyl azodicarboxylate (DEAD, 40 wt% in toluene, 379.7 mL, 1.26 mmol, 2 equiv.) and PPh 3 (328.0 mg, 1.26 mmol, 2 equiv), thioacetic acid (95.2 mL, 1.26 mmol, 2 equiv.) was added and the reaction was left for 48 h. The reaction mixture was concentrated in vacuo followed by purifcation by flash column chromatography (CH 2 Cl 2 ), isolating 5 (180 mg, 61%); 1 The compound was prepared from 4, using Mitsunobu conditions to carry out a thioesterifcation of the primary alcohol. In an oven-dried and degassed round-bottomed flask, containing a magnetic stirring bar, 4 (113.3 mg, 0.43 mmol, 1 equiv.), dry THF (34.4 mL), DEAD (40 wt% in toluene, 195. 9 mL, 0.65 mmol, 1.5 equiv.) and PPh 3 (169.2 mg, 0.65 mmol, 1.5 equiv.), thioacetic acid (49.1 mL, 0.65 mmol, 1.5 equiv.) was added and the reaction was left overnight. The reaction mixture was concentrated in vacuo followed by purifcation by flash column chromatography (CH 2 Cl 2 ), isolating 6 (31.3 mg, 23%); 1 ## Au(111) electrodes and sample preparation procedures In-house made Clavilier-type bead Au(111) electrodes (ca. 0.04 cm 2 ) were annealed at 850 C for 8 h. Compounds 5 or 6 were dissolved in 4 mL of isopropanol (0.1 mmol), mixed with 1 mL of 25% NH 3 (aq.), and kept in sealed container at 100 C for 8 h, using a microwave synthesizer (initiator, biotage), in order to remove acetyl groups. Au(111) electrodes were annealed in a hydrogen flame, quenched in ultrapure water saturated with dihydrogen, and further immersed in the solutions containing target compound for ca. 24 h. Finally, the samples were kept for 40 min overall in ethanol and subsequently in ultrapure water, prior to each electrochemical experiment. ## Instrumentation and electrochemical measurements Both CV and EIS studies were performed using three-electrode one-compartment glass cells and an Autolab potentiostat/ galvanostat (PGSTAT 12, Metrohm) controlled by the Nova 2.0 software. All measurements were performed at room temperature (23 AE 2 C), using bead Au(111) working electrodes and Pt coiled-wire counter electrodes. The reference electrode was a reversible hydrogen electrode (RHE) (flled with the same supporting electrolyte as in the cell) prepared before each electrochemical experiment. The reported values of the applied potential are shown in Fig. 1 after recalculation to standard hydrogen electrode (SHE). EIS studies were performed at different potentials (from 0.55 to 0 V vs. RHE, DE ¼ 0.05 V) in the frequency range of 100 kHz to 0.1 Hz (61 data points per measurement) by using an excitation amplitude of 0.01 V. Before each EIS measurement a constant potential, which corresponded to specifc DE was applied for 2 min. The Nova 2.0 software was used for the ftting of the impedance spectra. Tafel plots were constructed from selected CV experiments at various scan rates, ranging from 0.005 to 40 V s 1 . Prior to each CV or EIS experiment the electrolyte solutions were degassed with Ar gas for 30 min and an Ar atmosphere maintained above the electrolytes for the whole duration of electrochemical experiments. The electrolytes used were the following: 0.1 HClO 4 (pH 1), 0.1 M KH 2 PO 4 (pH 4.5), 0.1 M phosphate buffer (pH 7), and 0.1 M NaOH (pH 13). All glassware was cleaned prior each experiment by boiling in 15% HNO 3 (aq.) for ca. 20 min. ## Conflicts of interest There are no conflicts to declare.

chemsum

{"title": "Gold\u2013carbonyl group interactions in the electrochemistry of anthraquinone thiols self-assembled on Au(111)-surfaces", "journal": "Royal Society of Chemistry (RSC)"}

highly_fluorinated_interphases_enable_high-voltage_li-metal_batteries

## Abstract: Li metal is regarded as the ''Holy Grail'' electrode because of its highest specific capacity and lowest electrochemical potential. However, challenges arising from the low Coulombic efficiency (CE) and dendritic nature of Li metal in carbonate electrolytes remain to be resolved. Here, by increasing LiFSI salt concentration in the carbonate electrolyte, we successfully increased the CE to 99.3% while suppressing Li dendrite formation. An NMC622jjLi cell was paired and showed excellent cycling performance. ## INTRODUCTION Although they dominate the consumer electronics market and are penetrating the electric vehicle market, Li-ion batteries (LIBs) are approaching the upper limit of energy densities that intercalation chemistries can provide. 1 State-of-the-art LIBs with the typical graphite and LiCoO 2 as anode and cathode can deliver a specific energy of $250 Wh/kg, which is an order of magnitude lower than gasoline can deliver. 2 To further enhance the energy density of batteries, more aggressive chemistries are required, one of which is a Li-metal anode. When coupled with a high Ni-content cathode such as LiNi 0.6 Mn 0.2 Co 0.2 O 2 (NMC622), a 500 Wh/kg battery becomes possible. Li metal is considered the ultimate anode material because of its unique combination of the highest specific capacity of 3,860 mAh/g and the lowest electrochemical potential (3.04 V versus standard hydrogen electrode) ever possible from any known materials. However, its notoriety in both chemical and electrochemical stability, as demonstrated by limited Coulombic efficiency (CE) and dendrite formation during cycling, prevents its application in a battery environment. Extensive work has been devoted to stabilizing Li-metal anodes through approaches including protective layers, electrode designs at nanoscale, electrolyte additives, 14, and solid-state electrolytes 15,28 . Among these, ether-based electrolytes present the highest CE and the lowest overpotential, effectively suppressing dendrite growth owing to their low reactivity with Li metal. 2,4,12,14,17,22, Especially, the highest cycling CE of 99.1% was recently realized in 1,2-dimethoxyethane (DME). 4 However, ether-based electrolytes are intrinsically unstable against oxidation on cathode surfaces, as characterized by their typical anodic limits of <4 V, 36 which is much lower The Bigger Picture Interest in Li-metal batteries (LMBs) is reviving because higher energy densities can be enabled by the highest specific capacity and the lowest electrochemical potential of a Li-metal anode. However, challenges arising from the low Coulombic efficiency (CE) and dendritic nature of Li metal in carbonate electrolytes remain to be resolved. Here, by increasing LiFSI concentration in carbonate electrolytes (dimethyl carbonate [DMC], propylene carbonate, and ethylene carbonate/DMC) to 10 M, we achieved a high CE ($99.3%) of Li deposition and stripping, along with an anodic stability of >5.5 V. Pairing a Limetal anode in this electrolyte with LiNi 0.6 Mn 0.2 Co 0.2 O 2 (NMC622) at high loading (2.5 mAh/cm 2 ) created a NMC622jjLi cell, which showed a high capacity retention of 86% after 100 cycles at a high cutoff voltage of 4.6 V. LiF-rich and F-rich interphases formed on the Li-metal anode and cathode surfaces, respectively, are responsible for the successful suppression of Li dendrite growth as well as stabilization of the highvoltage cathode. than those of carbonate-based electrolytes. Thus, ether-based electrolytes can only be applied in low-voltage systems such as Li-S, Li-O 2 , and Li-LiFePO 4 . 18,22,32,34 For high energy density LMBs that require Li metal to be paired with a high-voltage, high-capacity cathode such as Ni-rich cathodes, a non-ether electrolyte that can simultaneously stabilize both Li-metal and cathode surfaces must be developed. Organic carbonates are exclusively used as electrolyte solvents in almost all commercial LIBs today thanks to their intrinsically high oxidation stability (>4.3 V) and their unique capabilities of forming protective interphases on graphite anodes. However, these electrolytes cannot accommodate the much more stringent requirements presented when graphite and mild transition metal oxides (such as LiCoO 2 ) are replaced by the more aggressive Li-metal and Ni-rich NMC (nickel-manganese-cobalt) materials. 2,10,12,24 Here, we report that by simply increasing the Li bis(fluorosulfonyl)imide (LiFSI) concentration in carbonate electrolytes (propylene carbonate [PC], dimethyl carbonate [DMC], ethylene carbonate [EC]/DMC), a significantly high CE of $99.3% can be achieved with an extremely high cycling stability. The LiF-rich solid-electrolyte interphase (SEI) layer, which is formed mainly from the LiFSI salt reduction, effectively prevents dendrite growth and minimizes the sustained electrolyte decomposition. These highly concentrated electrolytes also stabilize the high-capacity Ni-rich NMC cathode at higher voltage (Figure 1). The simultaneous stabilization of both anode and cathode surfaces in these electrolytes leads to a high energy density NMCjjLi cell, which steadily delivers high energy density at a high cutoff voltage (4.6 V) for extended cycles. ## RESULTS AND DISCUSSION CE truthfully monitors utilization and dendrite formation of metallic Li. A low CE indicates significant consumption of Li + and electrolyte, which will irreversibly result in poor cycling stability and the end of cell life. Li plating/stripping CE, defined as the ratio between the amounts of Li stripped and the amount plated on the Cu substrate, was examined in different electrolytes in 2,032 coin cells. Figures 2A-2D show the voltage profiles of the CujjLi cells at a current density of 0.2 mA/cm 2 in LiFSI-DMC electrolytes at varying salt concentrations. The charge-discharge voltage hysteresis and ionic conductivity of different electrolytes are summarized in Figure 2E, and the Li plating/stripping CEs in different electrolytes are compared in Figure 2F. Here, M stands for molar concentration, i.e., mole of salt dissolved in a liter of solvent. In diluted electrolyte (2 M LiFSI-DMC), the CE is only about 20%, indicating significant irreversible consumption of both Li + and electrolyte solvent as a result of the poor protection provided by the interphase, as revealed in the previous studies. 10 As the LiFSI concentration increases from 2 to 6 M, the CE quickly increased from 20% to 98.7%, eventually reaching 99.2% at 10 M, which is the highest CE ever reported for Li-metal anodes. Equally important is the overpotential between the plating and stripping processes, which becomes much lower in the concentrated carbonate electrolytes, despite their lower bulk ion conductivity than that of diluted electrolytes (Figure 2E), revealing that the interfacial resistances are dramatically reduced as a result of the high salt concentration. As a result of the improved CE and reduced interfacial resistances, the cycling stability of the Li anode in concentrated electrolytes improves dramatically. Although the CE in diluted electrolyte (2 M) never reached >40%, a relatively high CE of about 96% could by obtained in 4 M LiFSI-DMC electrolyte during the initial cycles. However, this CE lasted for only about ten cycles and then dropped to about 20%, which is similar to the diluted electrolyte (Figures 2B and 2F). In sharp comparison, 6 M and 10 M LiFSI-DMC electrolytes enabled up to 200 cycles without any deterioration detected. The Li plating/stripping behavior in Li/Li symmetrical cells was also evaluated in 6 M and 10 M LiFSI-DMC electrolytes (Figure S1). Both electrolytes showed good cycling stability. However, the overpotential in 6 M electrolyte increased at a faster pace than that in 10 M electrolyte. After 250 cycles, the overpotential in 6 M electrolyte was higher. The highly reversible nature of Li plating/stripping in concentrated electrolytes is still preserved after DMC is replaced with other carbonate solvents, as shown in Figure 3 (LiFSI-EC/DMC system) and Figure S2 (LiFSI PC system). At 0.2 mA/cm 2 , an overpotential of about 10.2 mV (hysteresis, $20.5 mV) was observed for LiFSI-EC/DMC (Figures 3B and 3C), which increased to $62 mV and 100 mV at a current density of 2 and 5 mA/cm 2 , respectively. These values representing rate capability are comparable with concentrated DME ether electrolyte. 4 The Li plating/stripping CEs in the 1 M EC/DMC and the 10 M EC/DMC are compared in Figure 3D. A significantly higher CE of 97.5% is achieved in the initial cycles in the 10 M LiFSI-EC/DMC, which gradually ramps up to $99.3% after about 80 cycles. Again, this value remains the highest among all the reported Li-metal anode materials. In sharp contrast, the diluted electrolyte shows only a reversible CE of $84% for EC/DMC and $78% for PC (Figure S2). We also assembled symmetric LijjLi cells to further confirm the cycling stability of Li-metal anode in an EC/DMC system. The long-term cycling stabilities of cointype LijjLi cells in 10 M and 1 M LiFSI-EC/DMC electrolytes are compared in Fig- ure 3E. The interfacial resistance of the LijjLi cell in concentrated electrolyte decreased in the first 60 hr, and then remained constant without any overpotential increase in the following 1,000 hr. However, the identical LijjLi cells in the diluted electrolyte cycled at the same current density demonstrated a significant impedance increase after only about 500 hr. After 700 hr, the overpotential of the cell was almost ten times higher than that using the concentrated electrolyte. The stripping/plating capacities (2 mAh/cm 2 ) in the LijjLi cell generally meet the requirements of commercial Li batteries imposed on the anodes in terms of areal capacity, current density, and cycle life. 37 These results further suggest that the better cyclability of Li-metal anodes in concentrated electrolytes could come from a more robust interphase that inhibits Li-metal dendrite growth and minimizes consumption of the electrolytes. Spectroscopic studies were conducted to understand the mechanism of high CE brought about by concentrated electrolytes. Raman spectra in Figures S3-S7 revealed that as the LiFSI salt concentration increases, the free carbonate molecules and free FSI are dramatically decreased. To establish whether a relationship exists between the coordinated carbonate solvent with CE, 8.5 M LiTFSI-DMC, in which almost all DMC molecules were coordinated with Li + (Figure S8), was also studied as a comparison. Expectedly, the DMC structure in this electrolyte is quite similar to that in 10 M LiFSI concentrated electrolyte. However, a significantly lower Li plating/stripping CE arose from the LiTFSI concentrated electrolyte ($30%; Figure S9). Therefore, the different Li plating/stripping behaviors between 10 M LiFSI and 8.5 M LiTFSI electrolytes must come from the different interphase chemistry originating from the salt anion. We also evaluated the CE of Li plating and stripping in a saturated LiPF 6 EC/DMC electrolyte (<5 M) and highly concentrated LiFTFSI (Li[(FSO 2 )N(SO 2 CF 3 )]) EC/DMC electrolyte (8 M). Both electrolytes showed an improved CE for Li plating and stripping. However, because of the lower solubility of LiPF 6 in EC/DMC (Figure S10), the CE for Li plating and striping increased from 82% for 1 M LiPF 6 electrolyte to only $91% at the saturation of LiPF 6 (Figure S11), whereas the CE can reach >98.5% for 8 M LiFTFSI EC/DMC electrolyte (Figure S12), which is significantly higher than the Li plating/stripping CE (30%) in 8.5 M LiTFSI electrolytes. This difference arises from the much higher reactivity of LiFSI and LiFTFSI than LiTFSI, which accordingly results in much higher F content in the SEI on the Li-metal surface. 31 LiFSI concentrations significantly affected the morphology of the cycled Li metals. Scanning electron microscopy (SEM) images in Figures 4A-4D compare the Li-metal surfaces after being cycled at a current density of 0.5 mA/cm 2 and areal capacity of 1 mAh/cm 2 for 100 cycles in different electrolytes (1 M LiFSI and 10 M LiFSI-EC/ DMC). Significant amount of Li dendrites as well as dead Li formed on the surface of the Li-metal anode during cycling in the diluted electrolyte, which is similar to the morphology of Li metal cycled in the other carbonate electrolytes. 26,29 The formation of needle-like dendrites with diameter of 200-400 nm and length of several micrometers leads to an explosive increase in the specific surface area of the Li-metal anode, which accelerates the parasitic reactions with electrolyte and exhausts the Li sources, resulting in not only poor efficiencies of LMBs and extremely low volumespecific capacities of the Li-metal anodes 38 but also severe safety hazards. The Li foil (Figure 4A, inset) and the Cu electrode (Figure S13) after being cycled in 1 M LiFSI-EC/DMC electrolyte were found to be covered by a polymer-like brown film, indicating that the interphase formed in dilute carbonate electrolyte might mainly be contributed by the massive reduction of solvents. 37 This observation suggests that the poor protection by the interphase likely originates from its organic nature, which is ineffective in shielding bulk Li metal against sustained electron tunneling at certain ''hot-spots,'' leading to sustained electrolyte reduction and preferential Li growth, as demonstrated by a CE of <85% (Figure 3D). In contrast, the surfaces of the Li-metal and Cu electrode turn black after cycling in concentrated carbonate electrolyte, indicating that the composition and the morphology of the SEI layer in concentrated carbonate electrolyte differs from the SEI formed in diluted electrolyte. 4 The Li surface cycled in the concentrated carbonate electrolyte displays a round-shaped morphology with a dense and uniform structure (Figures 4C and 4D). Such morphology of Li deposition enabled by concentrated electrolytes should render three notable advantages to the LMBs: (1) The severe safety concern induced by dendritic and dead Li can be significantly relieved. Differing from the needle-like dendrites deposited in diluted electrolyte, which are a few hundred nanometers long and could easily penetrate the separators, the round-shaped structures reduce the possibility of penetration of the porous separators. (2) The surface area of the Li deposited in concentrated electrolyte is much smaller, hence minimized side reactions between the deposited Li and the electrolyte lead to a much higher Li deposition/stripping CE (as shown in Figures 2 and 3). (3) A much higher volumetric capacity would be available from the packing of Li metal upon its deposition from the concentrated carbonate electrolyte. ## X-ray photoelectron spectroscopy (XPS) analysis revealed the chemical compositions of the interphase formed on the Li-metal surface in different electrolytes (Figures 4E and S14 ). The differences in the F1s, S2p, C1s, and O1s spectra between the two Li-metal anodes that were cycled in diluted and concentrated electrolytes are pronounced. One major difference lies in the significant variations in the elemental compositions in these two interphases. The one formed in the concentrated electrolyte is characterized by much higher F and S content but lower C and O content than those of the one formed in dilute electrolytes. The atomic ratio of F:C in the interphase is dramatically increased by more than ten times, from 0.42 in diluted electrolyte to over 5.05 in the concentrated electrolyte. More specifically, the accumulation of LiF species in the interphase is accompanied by the decline of carbon and oxygen species (C-C, C-H, C-O, C=O, etc.) on the Li-metal surface. Therefore, we can tentatively conclude that, because of the high population of FSI anions and their reactivity against reduction, the interphase formed by concentrated electrolyte tends to contain more LiF, which is contributed by the reduction of FSI rather than by solvent molecules. It is this chemical difference that is responsible for the distinct Li deposition/stripping behavior observed. LiF with high interface energy to Li can effectively prevent Li dendrite growth. 3 Molecular orbital energies of EC, DMC, and LiFSI were calculated by density functional theory (DFT) with the purpose of further understanding their respective reaction pathways. Figure 4G shows the energy values of the lowest unoccupied molecular orbital (LUMO) and the highest occupied molecular orbital (HOMO) for those species. LiFSI has a lower LUMO energy (1.70 eV) than those of EC (0.92 eV) and DMC (0.54 eV), indicating its higher tendency to react with Li metal than that of the solvent molecules under standard conditions. The preferential reduction of FSI by Li metal is dramatically enhanced as the molar ratio between solvent and salt decreases from 9.5 in diluted electrolyte to 0.95 in concentrated electrolyte. In addition to thermodynamic considerations, the -SO 2 F group in FSI is also kinetically more reactive than the carbonate solvents toward the Li-metal anode. 39 Splitting F by cleavage of S-F bond leads to a thermodynamically stable LiF first, which precipitates on the Li-metal surface, followed by cleavage of the S-N bond. The broken fragment of F(SO 2 ) 2 N under the strong attraction of oxygen atoms to the Li surface leads to the formation of SO 2 , which promptly leaves the surface as a result of blocking by the LiF dominating layer formed earlier. This could be the main reason for the scarcity of sulfur-containing species in comparison with LiF (Figure 4F). Consequently, a LiF-rich interphase is formed in the highly concentrated LiFSI carbonate electrolyte, as demonstrated by the XPS results (Fig- ures 4E and 4F). LiF plays two critical roles in increasing the CE: (1) LiF itself is a good electric insulator and effectively blocks electron leakage through the interphase, 40 which was believed to be detrimental and one of the critical reasons causing sustained electrolyte decomposition and dendrite formation. 41,42 (2) LiF exhibits much higher interfacial energy to Li metal and meanwhile produces a reduction of as much as 0.13 eV in the activation energy barrier for Li diffusion at the electrolyte/Li-metal electrode interface. 43 Thus, the surface diffusivity should be increased by more than two orders of magnitude, which facilitates Li transport along the interface and promotes the formation of a uniform morphology of the deposited Li metal. This is consistent with the recent reports that LiF can effectively suppress Li dendrite formation. An ideal electrolyte for LMBs should not only provide a high Li plating/stripping CE but also be stable against oxidation on high-voltage cathode surfaces. Figure 5A demonstrates the anodic linear sweep of three different electrolytes (1 M LiFSI in EC/DMC, 10 M LiFSI in EC/DMC, and 4 M LiFSI-DME) from 3.0 V to 6.5 V at a scanning rate of 10 mV/s. At a high potential of 5.0 V, the anodic current density in 10 M LiFSI-EC/DMC electrolyte is only 1/100 that of 4 M LiFSI-DME or 1/7 that of 1 M LiFSI-EC/DMC electrolyte. As indicated by Dahn et al., 48 the oxidation of carbonate solvents (especially EC) at a high voltage is usually responsible for impedance growth and cell failure. The excellent oxidation stability of concentrated LiFSI-EC/ DMC electrolyte can be ascribed to the following two reasons: (1) the less stable organic component in possible cathode interphase is largely reduced because of the much lower presence of solvent in the concentrated electrolytes (Figure S15); and (2) a fluorine-rich interphase, which is mainly from anion oxidation, passivates the cathode surface and stabilizes the electrolyte solvents (Figure S16). According to XPS, the ratio of F:C (4.1) on such a stabilized cathode surface is significantly higher than the cathode surface cycled in 1 M LiPF 6 EC/DMC electrolyte (0.65). These F-containing species, including LiF, CF x , and S-F, form a dense cathode interphase, suppressing the parasitic reactions between the cathode and the electrolyte. The charge-discharge and cycling behaviors of NMC622jjLi cells using different electrolytes were investigated (Figures 5B-5E and S17-S21). To utilize the high capacity of NMC, we adopted a much harsher charge-discharge protocol with a cutoff voltage of 4.6 V, which is much higher than most common charging protocols with a low cutoff voltage of 4.2 or 4.3 V using conventional electrolytes. Such a high cutoff voltage presents a rather rigorous test on the anodic stability of the electrolytes because it applies very severe electrochemical stress to induce side reactions. NMC622jjLi cells in 4 M LiFSI-DME cannot be charged to >4.2 V because of the poor anodic stability of ether solvents (Figure S17) although they possess a high Li-metal plating/stripping CE of $99% (Figure S18). In the diluted 1 M LiFSI-EC/DMC electrolyte, serious corrosion of the Al current collector by LiFSI (Figure S19) also make the first charge of NMC622 impossible (Figure S20). Because the charging of NMC622jjLi cells in 1 M LiFSI-EC/DMC and 4 M LiFSI-DME failed between 2.7 and 4.6 V, the cycling behavior of NMC622jjLi cells in highly concentrated 10 M LiFSI-EC/DMC electrolyte was compared with that of 1 M LiPF6 EC/DMC electrolyte (Figures 5B-5E). Almost identical electrochemical performances can be obtained from the first cycle of NMC622jjLi cells in both diluted LiPF 6 electrolyte and the high concentration LiFSI carbonate electrolytes (Figure 5B). The slightly lower initial CE in 1 M LiPF 6 EC/DMC might be because of the oxidation of the carbonate solvents at the high voltage. Figures 5C and 5D show the charge-discharge voltage profiles at different cycles. Much higher voltage polarization with significant capacity decay was observed for the cell using 1 M LiPF 6 EC/DMC, suggesting that higher resistance was generated at the electrode interfaces as a result of the decomposition of the electrolyte solvent. Cycling performance and the corresponding CE of NMC622jjLi cells using these two different electrolytes are shown for the first 100 cycles (Figure 5E). The cell in 1 M LiPF 6 EC/DMC exhibited a CE of about 99% in the first 20 cycles. However, starting from the 20 th cycle, this CE sharply dropped to only about 95%, accompanied by a faster capacity decay rate. After 100 cycles, the cell in 1 M LiPF 6 EC/DMC retained only $52% of its original capacity. In contrast, the concentrated EC/DMC electrolyte enabled a high-energy-density NMC622jjLi cell with high and stable CE of >99.6% along with a capacity retention of $86% after 100 cycles, which represents a significant improvement for the aggressive chemistry of both Li-metal and NMC622 over traditional carbonate electrolytes, even with such a limited number of cycles. The enhancement in the electrochemical performance of NMC622jjLi cells in concentrated carbonate electrolytes is universal. Similar improvements in cycling performance were achieved for concentrated DMC electrolyte (Figure S21). After 150 cycles, the cell in concentrated DMC electrolyte retained a capacity of $80%. Yet, the cell in the dilute 1 M LiPF 6 DMC electrolyte dramatically dropped to 19% of the initial capacity. ## EXPERIMENTAL PROCEDURES Materials Cathode NMC622 (LiNi 0.6 Mn 0.2 Co 0.2 O 2 ) electrode laminates ($13 mg active material cm 2 ) were supplied by SAFT America. The electrode laminates were punched into discs and further dried at 80 C under a vacuum overnight. All the solvents were purchased from Sigma-Aldrich, including EC, DMC, and PC. All the solvents were dried by molecular sieve (4 A ˚, Sigma-Aldrich) to make sure the water content was lower than 2 ppm, which was tested by a Karl-Fisher titrator (Metrohm 899 Coulometer). Lithium bis(fluorosulfonyl)imide (LiFSI, 99.9%) was purchased from American Elements. LiPF 6 (>99.99%) was purchased from Sigma-Aldrich. ## Material Characterization The morphology and microstructure of the samples were investigated by SEM (Hitachi SU-70). XPS was conducted on a high-sensitivity Kratos AXIS 165 X-ray photoelectron spectrometer with Mg Ka radiation. All binding energy values were referenced to the C 1s peak of carbon at 284.6 eV. Before XPS characterization, the cycled electrodes were washed with the corresponding solvents several times to remove residual salts. ## Electrochemical Measurements Electrolytes are prepared by adding LiFSI or LiPF 6 into various anhydrous solvents (DMC, PC, EC/DMC). The charge-discharge performances of the LMBs were examined by 2,032 coin-type cells. The same coin-type cells were used to investigate the cycling stability of Li plating/stripping in different electrolytes. The CE of Li plating and stripping was calculated from the ratio of the Li removed from the Cu substrate to that deposited in the same cycle. A three-electrode ''T cell'' was utilized to test the stability window of the different electrolytes with polished stainless steel as the working electrode and Li foils as the reference and counter electrodes with a Gamry 1000E electrochemical workstation (Gamry Instruments, USA). All cells were assembled in a glove box with water and oxygen content lower than 2 ppm and were tested at room temperature. The galvanostatic charge-discharge test was conducted on an Arbin battery test station (BT2000, Arbin Instruments, USA). ## DFT Calculations The Vienna ab initio Simulation Package (VASP) was used to perform DFT calculations, 49,50 and Perdew-Bruke-Ernzerhof (PBE) functional of the generalized-gradient approximation (GGA) was used for electron exchange and correlation. 51 Previous works have shown that GGA-PBE yields qualitatively the same trend for the ground state of higher acenes as the B3LYP function and a high-level wave function method. 52 The projector augmented wave method with an energy cutoff of 580 eV was used to describe the ion-electron interaction on a single k point. 53 The convergence condition for the energy was 10 5 eV, and the structures were relaxed until the force on each atom was less than 0.001 eV/A ˚. For all solvents and the salt, a vacuum larger than 15 A ˚was used to simulate the molecules. Visualization of the LUMO and HOMO was done with Molekel software. 54 ## SUPPLEMENTAL INFORMATION Supplemental Information includes 21 figures and can be found with this article online at https://doi.org/10.1016/j.chempr.2017.10.017. ## AUTHOR CONTRIBUTIONS Conceptualization, X.F. and C.W.; Methodology, X.F., L.C., X.J., and K.X.; Investigation, X.F., L.C., T.D., S.H., J.C., J.Z., and F.W.; Software, X.J. and J.J.; Writing -Original Draft, X.F.; Writing -Review & Editing, K.X. and C.W.; Funding Acquisition, C.W.; Supervision, C.W.

chemsum

{"title": "Highly Fluorinated Interphases Enable High-Voltage Li-Metal Batteries", "journal": "Chem Cell"}

vinylimidazole_coordination_modes_to_pt_and_au_metal_centers

## Abstract: The coordination modes of 1-vinylimidazole to platinum and gold were studied. Complexes [PtCl 3 (Hvinylimidazole)]ÁH 2 O (1), [Au(vinylimidazole) 2 ] + [AuBr 2 ] À (2), [Hvinylimidazole] + [AuCl 4 ] À (3), and [Hvinylimidazole] + [AuBr 4 ] À (4) were prepared and structurally characterized. Compound 1 is the first structurally characterized transition metal complex containing a protonated vinylimidazole, which is coordinated through the vinyl group in the side-on position. In compound 2, the neutral ligands coordinate through the imidazole nitrogens to the reduced gold(I) center and the charge balancing counter anion [Au(I)Br 2 ] À has a short Au-Au contact with the cationic part. In 3 and 4, the acidic reaction conditions lead to the protonation of the imidazole nitrogen and an ion pair with tetrahalogenide gold(III)is obtained. The tendency to the different crystallized products is attributed to the combination of the metal and the halogen properties with the reaction conditions. Computational chemistry was used to explain the preference of the vinyl coordination type, as well as in the interpretation of the spectroscopic details and the nature of the intra-and intermolecular interactions present in the solid state. ## Introduction 1-Vinylimidazole or N-vinylimidazole is a two nitrogen containing, five-membered, aromatic compound with the vinyl group at one of the nitrogen atoms. Imidazoles, in general, are important in biomolecules, pharmaceutical applications and in industrial utilization as corrosion inhibitors. 11 Vinylimidazole is particularly known for its homo and co-polymers. The monomer is easily polymerized through the vinyl group by exploiting UV-irradiation, heat or activator agents. Furthermore, vinylimidazole monomers can be quaternized to ionic liquids and polymerized through the side chains. 12,13 The active imidazole nitrogens in these polyionic hydrogels effectively bind metal atoms. Due to the mechanical stability, strength and recoverability of the materials, diverse applications in polymer anchored metal catalysts, 14 drug delivery, 15 and water purification 16,17 are produced. In the context of transition metal complexes, the most probable coordination sites in 1-vinylimidazole are the imidazole nitrogen and the vinyl group. The substituent at the heteroatom blocks up the possibility for tautomerism typical of the original imidazole. The non-coordinated nitrogen atom is slightly basic and protonation of this site is possible. The high tendency to coordination to metal centers and formation of molecular complexes, as in the case of polymeric structures, is a source of diverse applications such as pharmacologically active species. 18,19 A survey of the CCDC database and literature reveals that exclusively nitrogen bound ligands are present in transition metal vinyl imidazole complexes. A quite large selection of complexes for lighter transition metals is published. However, going towards the heavier metals, the number of synthesized and characterized complexes is reduced. In group ten in the periodic table, Ni has several vinylimidazole derivatives. For palladium, tetranuclear planar structures with one to four ligands have been reported, 23,24 while only the octahedral [PtCl 2 (vinylimidazole) 4 ]Cl 2 is crystallographically characterized for platinum. 25 Among the coinage metals, copper has several derivatives of this ligand. The most common coordination type is octahedral, but also tetrahedral complexes are known. Typically, these have two or four vinylimidazole ligands. The catena compounds, where Cu centers with 1-vinylimidazole ligands and ReCl 6 -units are connected through chloro bridges, have shown magnetic properties. 38 Surprisingly, no compounds of Ag or Au have been reported. In group twelve, Zn 29, and Cd 33,41 form MOF complexes with auxiliary linking ligands, and the vinylimidazoles coordinate monodentately. Ruthenium has a unique cluster derivative [HRu 3 (CO) 10 (vinylimidazole)], where the deprotonated ligand acts as a bidentate bridging C,N-ligand. The imidazole carbon between the nitrogen atoms is utilized in coordination, but the vinyl group remains intact. 42 A similar type of ligand, 4-vinylpyridine typically coordinates through the nitrogen atom, as in the complexes of Cu, 43 Pt 44,45 and Zn. 46 The second isomer, 2-vinylpyridine, has been the subject of a relatively limited number of studies on the synthesis, structure and reactivity, which have involved metals Ru, Os, 50, Co, 58 Rh, 59,60 Ir, 61 Pd, 62 Pt 63,64 and Au. 65,66 2-vinylpyridine can form five-membered cyclometalated complexes with the high versatility of applications in the fields of biological, catalytic and luminescence properties. 64 In this work we concentrated on the reactions of vinylimidazole with late transition metals, Pt and Au, to clarify the relative reactivity of the imidazole group or the vinyl group towards the metal ions. The experimental crystal structures of the products were analyzed via X-ray diffraction, NMR, IR and Raman spectroscopy, and computational DFT methods. ## Reactions and structural studies The reaction of 1-vinylimidazole with K 2 PtCl 4 in water solution led to the formation of a white product, which by spectroscopic measurements and elemental analysis was interpreted as the square planar [PtCl 2 (vinylimidazole) 2 ] with nitrogen coordinated ligands. This is in good agreement with the earlier reported reactions in water, ethanol or acetone. 67 On the other hand, the reaction of 1-vinylimidazole with K 2 PtCl 4 in acidic medium generated a novel vinylimidazole derivative type [PtCl 3 (Hvinylimidazole)]ÁH 2 O (1), where the partially protonated ligand is bound to metal through the vinyl group in side-on geometry (Fig. 1a). The Pt(II) center has further three chloro ligands in the perpendicular plane against the CQC bond of 1.395(2) . The Pt-C(6) bond is 2.135(2) and Pt-C(7) is slightly shorter 2.109(2) , showing a strong side-on coordination type of the vinyl group. The sum of the van der Waals radii of Pt and C is 3.45 . The Pt-Cl bonds are 2.2972(5) and 2.2922(5) in the cis-position and 2.3056(5) in the trans-position to the vinylimidazole. Reaction of vinylimidazole with AuBr 3 in organic aprotic THF solvent led to the formation of an ion pair [Au(vinylimidazole) 2 ] + [AuBr 2 ] (2) (Fig. 2). In the cation, two neutral ligands are attached to the Au(I) center via imidazole nitrogens, the vinyl groups remaining unreacted. Anionic [Au(I)Br 2 ] acts as the counter ion. Therefore, reduction of the metal with partial oxidation of the bromines has taken place during coordination. The coordination at both gold atoms is almost ideal T-shape with Au-Au-N bond angles of 90.8(2)1 and 91.0(2)1. The Au-N distances are 2.018(5) and 2.017(5) . The CQC bond lengths in vinyl groups are considerably shorter, 1.306(10) and 1.300 (10) , than in 1. The Au-Au distance is 3.1200(5) , which is slightly shorter than the sum of the van der Waals radii of two gold atoms (3.32 ). The cationic part is planar and the Br-Au-Br moiety lies perpendicular to this plane. The ion pairs form a chain structure in the solid state via hydrogen bonds (Fig. 3). The bromine forms an intramolecular hydrogen bond of 3.0687(8) to H(2). The bromine atoms have also several short contacts of 2.8839(7)-3.0437(8) to hydrogens of neighboring ion pairs. The nature of the hydrogen bonding network was further studied computationally by analyzing the charge density of an extended model by Quantum Theory of Atoms in Molecules (QTAIM). The model with four adjacent ion pairs was cut directly from the experimental crystal structure (see Fig. S1 in the ESI †) and analyzed using the DFT wavefunction. The results for the properties of the electron density are shown in Table 1. The [Au(vinylimidazole) 2 ] + [AuBr 2 ] ion pairs show a large number of relatively weak intermolecular hydrogen bonds to the adjacent ion pairs, explaining the packing of the ions into rows. The properties of the electron density, such as small r, the ratio between potential energy density and kinetic energy density |V|/G o 1, and interaction energy E INT o 10 kJ mol 1 , all point out to typical noncovalent intermolecular hydrogen bonds. In addition, the aurophilic Au(1)Á Á ÁAu(2) interactions at BCP 3 exhibit slightly larger strength with E INT of 26 kJ mol 1 . Notably, the aurophilic interactions to the neighboring cations have comparable strength (E INT = 13 kJ mol 1 at BCP 6), enhancing the self-assembly of the ion pairs. Furthermore, p-p interactions form between vinyl groups and the imidazole rings of the adjacent ion pairs, explaining the torsional behavior of the N-ligands as they stack together. In the reaction of the ligand with either NaAuCl 4 or HAuCl 4 in water solution containing HCl, protonation of the ligand took place giving compound [Hvinylimidazole] + [AuCl 4 ] (3) (Fig. 4) with a disordered structure. Thus, the protonated ligand does not coordinate to the metal like in the case of platinum. This type of ion pair was earlier known for instance for pyridine. 68 In the solid state, a network supported by hydrogen bonds H(3)-Cl(1) of 2.4256(5) and H(7B)-Cl(4) of 2.9059(5) and 2.9139(5) was obtained (Fig. 4). The synthesis of [Hvinylimidazole] + [AuBr 4 ] (4) from AuBr 3 in acidic HBr solution proceeded analogously to that of 3. The crystal structure and spectroscopic data showed a similar structure type to 3, but in this case with no disorder. ## IR and FT-Raman spectroscopy The appearance of the n(N-H) vibration at 3281, 3254 and 3256 cm 1 supports the protonation of the imidazole nitrogen in compounds 1, 3 and 4. The interpretation of the spectra was verified via computational simulation at the DFT level of theory on optimized molecular models. The experimental and simulated spectra are shown in Fig. S6-S8 (ESI †). Vinylic n(CQC) in the free ligand lies at 1630 cm 1 (computational value at 1626 cm 1 ). In 1 it is clearly shifted to a lower frequency of 1437 cm 1 (simulated 1441 cm 1 ), which is attributed to p-bonded CQC-ligands like ethylene and related ligands. 69 In a s-bonded ligand, the shift is typically to higher frequencies. In 2, 3 and 4, n(CQC) appears at 1640 cm 1 (sim. 1632), so the shift is not remarkable. The Raman spectra were measured for complexes 1 and 2 in the solid state and for vinylimidazole in the liquid form, and also simulated at the DFT level of theory using single molecular models to facilitate interpretation. The computational values are given in Table 2, and an example of the interpretation of the most important signals for compound 1 is given in Fig. S4 and S5 (ESI †). The full interpretation of the solid state Raman spectra was possible only for the platinum complex 1, which yielded the best quality experimental spectrum. Fig. S4 and S5 (ESI †) compare the experimental and simulated signals, which were further interpreted by animation of the vibrations. Even though the wavenumbers do not exactly match because all signals in the simulation do not scale similarly, the general appearance of the spectra is similar enough to allow interpretation (Table 2). The coordination mode of 1 and 2 can most easily be seen in the stretching vibration of the CQC double bond of the vinyl group, which is strongly shifted to smaller wavenumbers in the platinum complex 1, but shows a similar value to that in the free vinylimidazoles for the gold complex 2. The same, but less clear trend is seen in the n(C-H) values of the CH 2 protons of the vinyl group, as well as in the scissor vibration of the CH 2 protons. Furthermore, clear Pt-C(vinyl) signals could be obtained in the spectrum of 1, verifying the side-on coordination of the metal. The poor quality of the experimental Raman spectrum of compound 2, resulting from the degradation of the crystals during the measurement, did not allow all the signals to be resolved, most importantly the Au-N stretching frequency. However, indirect evidence on the coordination mode could be obtained by the absence of interaction with the vinyl group, since n(CQC) vinyl and n(C-H 2 ) vinyl were not shifted compared to the free 1-vinylimidazole. Additionally, there was no N-H stretching signal, which verified the coordination of the ligand in the non-protonated form. It has been observed in the previous computational Raman studies of platinum complexes that the effect of temperature and isotopes of chlorine complicate the situation and some contradiction among the assignments exists. 70,71 ## NMR spectroscopy The 1 H NMR spectra of all compounds were measured in d 6 -DMSO solvent and interpreted by comparison with the spectrum of the free ligand. The spectra are presented in Fig. S2 (ESI †). Compound 1 shows two sets of sharp peaks with a very similar pattern to the free ligand shifted to slightly larger chemical shifts. One possibility to the two sets of signals is the existence of another isomer with a non-protonated and possibly N-coordinated ligand. This suggestion is supported by the observation that the imidazole protons (2, 4, and 5) in isomer b show the same chemical shifts as in the spectrum of 2, while the peaks of the crystallized product appear at the same location as in 3. The protonation of the vinylimidazole ligand was computationally found to change the charge distribution especially in the aromatic ring of the free ligand, which could explain the difference in the chemical shift values. The spectrum of 2, as in the case of 1, has also a pattern of two signal sets shifted toward higher chemical shifts. In this case, the small difference in chemical shifts suggests quite similar structures for the isomers, indicating the presence of two compounds in solution, probably due to the rotation around the Au(1)-Au(2), Au(1)-N(3/3A), and N(1/1A)-C(6/6A) bonds. The major compound has reasonably sharp signals. However, the minor compound shows broad signals, which can be assumed to rise from hydrogen bonding of the protons to the near-by quadrupole bromines. Since the rotation in solution can be rather free around single bonds, in the minor isomer the protons can have short contact with bromines. The elemental analysis indicates the presence of one pure product. The NMR spectra of 3 and 4 are similar with one set of peaks describing one pure product without the possibility of fluxionality. Overall, the most notable shift of the signals is observed at the imidazolic H(2) proton signal in those structures, where protonation of the ligand has taken place. This shift can be interpreted to have originated from the protonation of the imidazole nitrogen N3, which was found to modify considerably the calculated charge distribution of the whole imidazole ring and especially the charge of C(2) and H(2). Both proton coupled and proton decoupled 13 C NMR spectra of the vinylimidazole ligand in the basic form as well as HCl treated vinylimidazole to simulate the protonation were measured (Fig. S3, ESI †), but only the proton decoupled spectra of the products (1-4) were obtained. Comparison of the experimental and the calculated spectra is given in Table S1 (ESI †). The spectra show the remaining vinylimidazole assembly in all cases. In complexes 1, 3 and 4, containing the protonated ligands, the signal of C( 6) is slightly shifted to lower values as in the case of the protonated free ligand. Otherwise, the structure of the spectrum remains practically unchanged. The presence of different types of metals and paramagnetic species as well as the solvent effect has earlier shown to cause peculiarities in the spectrum appearance of 1-vinylimidazole and even the relative positions of the signals can change. 75,76 ## Stability of the coordination modes In order to find explanation for the preferred coordination mode in the Pt or Au vinylimidazoles, we performed computational optimization for models mimicking both N-coordination (A) of the deprotonated ligand and the side-on coordination (B) to the vinyl group of the protonated ligand. Since the HOMO-LUMO energy gap can give information on the relative stability of the models, the highest occupied and the lowest unoccupied orbitals along with the corresponding energies of the small molecular models are presented in Fig. 5. There is a clear difference in the frontier molecular orbitals of the N-coordinated cations. With platinum, the strong involvement of the metal d orbitals in both the HOMO and LUMO stabilizes their energy, leading to a small energy gap. In contrast, the LUMO of the gold cation is expanded over the two ligands without contribution from the metal center, which destabilizes strongly the LUMO energy and leads to a large energy gap and hence to very stable coordination. The opposite is true in the CQC coordination mode B, where the LUMO of the Au complex concentrates over the MX orbitals, again stabilizing LUMO energy, whereas in the Pt complex the LUMO is expanded mostly on the ligand p orbitals. According to the frontier orbital energies, platinum would prefer coordination to the vinyl group and gold would prefer N-coordination of the 1-vinylimidazole, provided that the Au(III) center is reduced to Au(I). It should be noted that we also tested the side-on coordination with Au(I), and even though the stability of the product, De = 4.58 eV, was slightly larger than for Au(III), the preferred coordination mode would still be A. Experimentally, we tested also the reaction HAuCl 4 + vinylimidazole in THF solution. The crystals formed showed again the structure of 3. The 1 H NMR spectrum of the reaction mixture verified this, but two minor products were present, thus other coordination modes are possible for gold. ## Materials and methods Commercially available reagents AuBr 3 (99%, Alfa Aesar), HAuCl 4 Á3H 2 O (Au 49.5%, Alfa Aesar), NaAuCl 4 Á2H 2 O (99%, Sigma Aldrich), K 2 PtCl 4 (99,9% Alfa Aesar) and 1-vinylimidazole (99%, Alfa Aesar) were used without purification. The organic solvents were dried using molecular sieves. Infrared spectra were measured from KBr pellets using a Shimadzu IRAffinity-1 Fourier transform infrared spectrophotometer in the range of 4000-400 cm 1 . The elemental analysis was performed on a varioMICRO V1.7. The 1 H NMR spectra were recorded on a Bruker Avance 400 MHz or Bruker AMX 400 MHz spectrometer. Raman spectra were recorded on a Renishaw inVia Raman Microscope with 514 nm excitation laser wavelength. Raman shifts ranging from 3200-100 cm 1 were collected. ## Crystal structure determination The crystals of 1-4 were immersed in cryo-oil, mounted in a Nylon loop, and measured at a temperature of 150 K. The X-ray diffraction data were collected on a Bruker Kappa Apex II diffractometer using Mo Ka radiation (l = 0.71073 ). The APEX2 77 program package was used for cell refinements and data reductions. The structures were solved by direct methods using the SHELXS-2018 78 program with the WinGX 79 graphical user interface. A numerical absorption correction (SADABS) 80 was applied to all data. Structural refinements were carried out using SHELXL-2018. 78 The crystallization solvent in 2 was heavily disordered and could not be resolved unambiguously. The contribution of the missing solvent to the calculated structure factors was taken into account by using a SQUEEZE routine of PLATON. 81 The missing solvent was not taken into account in the unit cell content. The N-H hydrogen atoms in 1, 2 and 4 and O-H water hydrogen atoms in 1 were located from the difference Fourier map and constrained to ride on their parent atoms, with U iso = 1.2-1.5U eq (parent atom). All other hydrogen atoms were positioned geometrically and constrained to ride on their parent atoms, with C-H = 0.95 , N-H = 0.88 and U iso = 1.2U eq (parent atom). The crystallographic details are summarized in Table 3. ## Computational details All calculations were performed by applying the Gaussian 09 software package. 82 The optimized geometry and simulated scaled infrared and Raman spectra of all the complexes were obtained by the PBE0 functional 83 with the 6-311++G(d,p) basis set for non-metal atoms and the Def2-TZVPPD basis set 84 for Pt and Au atoms. To obtain the electronic properties of the solid state structure of 2, we performed topological charge density analysis with the QTAIM (Quantum Theory of Atoms in Molecules) 85 method, which allowed us to access the nature of the bonding via calculating different properties of the electron density at the bond critical points (BCPs). The analysis was done with the AIMALL program 86 13 C NMR (400 MHz, d 6 -DMSO, 298 K, d): 134.8 (C2), 128.9 (C4), 120.9 (C6), 118.6 (C5), 108.6 (C7) ppm. [AuBr 4 ] [Hvinylimidazole] + (4). AuBr 3 (33.6 mg, 0.077 mmol) was dissolved in 5 M HBr (3 ml) and one drop of vinylimidazole was added. After a few days, the red solid was filtered, washed with water and dried. Yield: 35 13 C NMR (400 MHz, d 6 -DMSO, 298 K, d): 134.9 (C2), 128.9 (C4), 120.9 (C6), 118.7 (C5), 108.7 (C7) ppm. ## Conclusions 1-Vinylimidazole has different coordination modes towards platinum and gold. Reaction conditions were found to have a major effect on the obtained crystalline products, which was attributed to the protonation/deprotonation of the 1-vinylimidazole free nitrogen in acidic/basic medium. Under the acidic conditions, the first complex of this ligand with the vinyl group coordinated to platinum metal was structurally characterized. On the other hand, the protonated ligand did not directly coordinate to gold, but formed halide salts with [AuX 4 ] as the counter anion. Under the basic conditions, the gold complex was observed as an ion pair with cation [AuL 2 ] + and anion [AuBr 2 ] connected via aurophilic interactions. According to the DFT calculations, the reduction of the Au(III) to Au(I) center leads to the formation of a very stable cationic N-coordinated complex with 1-vinylimidazole. For platinum, the calculations predict that the side-on coordination of the vinyl group is energetically more favorable than N-coordination, which explains the structure of complex 1. The stability of the molecular models was fully consistent with the experimentally obtained structures, which were verified via X-ray diffraction and IR, Raman, and NMR spectroscopy. ## Conflicts of interest There are no conflicts to declare.

chemsum

{"title": "Vinylimidazole coordination modes to Pt and Au metal centers", "journal": "Royal Society of Chemistry (RSC)"}

kinetic_resolution_of_sulfur-stereogenic_sulfoximines_by_pd(<scp>ii</scp>)–mpaa_catalyzed_c–h_arylat

## Abstract: A direct Pd(II)-catalyzed kinetic resolution of heteroaryl-enabled sulfoximines through an ortho-C-H alkenylation/arylation of arenes has been developed. The coordination of the sulfoximine pyridyl-motif and the chiral amino acid MPAA ligand to the Pd(II)-catalyst controls the enantio-discriminating C(aryl)-H activation. This method provides access to a wide range of enantiomerically enriched unreacted arylpyridyl-sulfoximine precursors and C(aryl)-H alkenylation/arylation products in good yields with high enantioselectivity (up to >99% ee), and selectivity factor up to >200. The coordination preference of the directing group, ligand effect, geometry constraints, and the transient six-membered concertedmetalation-deprotonation species dictate the stereoselectivity; DFT studies validate this hypothesis. ## Introduction The directing group (DG) assisted desymmetrization of prochiral C-H bonds provides a suitable way to construct carbon, phosphorus, silicon, and sulfur centered functionalized chiral molecules. However, this approach requires achiral precursors with two identical enantiotopic groups, which prevents its application for broad synthetic benefts. On the other hand, kinetic resolution (KR) of C-H bonds offers booming advantages for making functionalized enantioenriched molecules. In this regard, Yu's pioneering work on DG assisted chiral amino acid (MPAA) enabled Pd-catalyzed carbon centered KR of arene C-H bonds through alkenylation, arylation, and/or iodination is undoubtedly a breakthrough (Fig. 1A). 4 In spite of this success, the related strategy of Pd-catalyzed heteroatom centered KR of arenes remains to be explored, although exceedingly appealing. Sulfoximines, which are confgurationally stable motifs with S-stereogenecity, are found in molecules of medicinal importance and agrochemicals. 5 Notably, sulfoximines have emerged as chiral auxiliaries and DG for C-H functionalizations. 6 The syntheses of enantioenriched sulfoximines have invariably relied on resolution techniques, stereoselective imination, and oxidation processes. 7,8 Elegant enantioselective and KR routes to sulfoximines have been independently developed by Cramer, Li, Shi, and others, but all these approaches rely on Rh/Ru-catalyzed [4 + 2] annulation of diazoesters/sulfoxonium ylides and aryl-sulfoximines in the presence of specially designed ligands. 9 On our side, we have devised an expedient Pd-catalyzed C-H functionalization method for KR of 2-pyridylaryl sulfoximines, using Pd(II) catalyst and MPAA ligand, via C(aryl)-H arylation and olefnation (Fig. 1B). The concept relies on kinetically regulated concertedmetalation-deprotonation (CMD) step of C(aryl)-H activation (k 1 >> k 2 , Fig. 1B) through preferred coordination of pyridine over imine to Pd-MPAA (Fig. 1C-I) 10 and ligand geometry CMD Pyr-DS over CMD Pyr-DR (Fig. 1C-II). The transformation is general, constructing a wide array of enantiomerically enriched C-olefnated/arylated aryl-pyridyl-S-sulfoximines. ## Results and discussion The study was initiated with the non-substituted N-Boc-phenyl-2-pyridyl sulfoximine rac-1a-1 and ethyl acrylate (2a; 0.6 equiv.) in presence of Pd(OAc) 2 (10 mol%), Boc-L-Phe-OH (L1; 30 mol%), Ag 2 CO 3 (2.0 equiv.) in ClCH 2 CH 2 Cl (1,2-DCE) at 75 C ee c (%) ee (%) a Reaction conditions: rac-1a-1 (0.1 mmol), ethyl acrylate 2a (0.6 equiv.), Pd(OAc) 2 (10 mol%), ligand (30 mol%), Ag 2 CO 3 (2.0 equiv.), 2-Cl-BQ (0. (Table 1a). The desired C2-alkenylation product (S)-3a-1 (18%, conversion after 3 days) along with precursor (R)-1a-1 were obtained in 75% ee and 17% ee, respectively, exhibiting a low selectivity factor (s) of 8. This encouraging result unfolded our curiosity about examining the effect of other ligands. None of the N-Boc, N-acetyl-, and N-imide-protected commercially available a-amino acid ligands (L2-L6) with distinct side chains were effective. Assuming the additional coordination ability of the easily modifable OH group in threonine, various N,Oprotected threonine ligands were tested. The reaction s factor was improved a little for (S)-3a-1 to 12 and 11 when Boc-L-Thr(t-Bu)-OH (L7) and Boc-L-Thr(Bn)-OH (L8) were used, respectively. Electronic perturbation in the O-benzyl moiety did not have any impact on the enantioselectivity (L9 and L10). The use of 2a (2.0 equiv.) in presence of ligand L8 improved the conversion (50%) with (S)-3a-1 (70% ee) (entry 13). To enhance sulfoximine resolution efficiency while maintaining conversion ($50%; Table 1a), we scrutinized the cooxidant effect (Table 1b). 11,12 2-Chlorobenzoquinone (2-Cl-BQ) was found to be the best, providing (S)-3a-1 in 77% ee with 39% conversion (s factor of 13; entry-2, Table 1b). Next, sulfoximines having various N-protecting groups (PG) were screened; the results are shown in Table 2a. None of the N-Me/Piv/Cbzprotected sulfoximines were found effective. Next, we studied the DG effect (Table 2b). Thus, various substituted 2-pyridyl containing sulfoximines were independently subjected to 2a. After several trials, the 3-methyl pyridyl DG was found superior, affording the alkenylation resolution species (S)-3a in 96% ee with s factor of 58, although conversion was limited to 17% (entry 1). On the other hand, 3-Cl/Br substituted pyridyl DG were unsuccessful (entries 2 and 3). While trace of desired olefnation product (S)-3s with 96% ee was noticed from the reaction of 3-phenyl-pyridyl (DG 5 ) bearing a Reaction conditions: rac-1 (0.25 mmol), olefn (2.0 equiv.), Pd(OAc) 2 (10 mol%), L8 (30 mol%), Ag 2 CO 3 (2.0 equiv.), 2-Cl-BQ (0.5 equiv.), 1,2-DCE (2.5 mL), 75 C, 3 days. b Yield of the isolated olefnated product. c Olefn (1.8 equiv.) was used and reaction was performed for 1.5 days. Calculated conversion, C ¼ eeSM/(eeSM + eePR). Selectivity (s phenyl sulfoximine (1a-4) with methyl acrylate (2b, 0.6 equiv.; entry 4). The 4-Me/5-Cl substituted pyridyl DG were ineffective, providing a lower selectivity factor (s) of 12 and 11, respectively (entries 5 and 6). No desired olefnation product was obtained when 3-F substituted (DG 8 ) pyridyl group was used (entry 7). The reaction conversion was improved to 22% when 2a (2.0 equiv.) was employed under the reaction shown in entry 1 (entry 8). The identical transformation with 2b (2.0 equiv.) could enhance the conversion to 27% (entry 9). Finally, a 50 mol% loading of 2-Cl-BQ led to (S)-3b (96% ee, factor of 85 with 34% conversion; entry 10), which was found optimum. The generality of the Pd-catalyzed C-H alkenylative KR of sulfoximines was then surveyed (Table 3). Compound 3b (98.2 : 1.8 er) was isolated in 26% yield. The alkenylation occurred at the less-hindered arene C-H bond and the chiral sulfoximines 3c and 3d were obtained with s factors of 162 and 44, respectively. The catalytic system was compatible with common functional groups, such as ketone, sulfone, and phosphate in the alkene, providing access to 3e (95.6 : 4.4 er), 3f (94.6 : 5.4 er), and 3g (97.6 : 2.4 er). Notably, the reaction of methyl vinyl sulfone with 1b displayed an exceptional s factor of >200 for compound 3h. The reaction of p-(Me/ t Bu/ i Pr)substituted aryl sulfoximines with 2b/vinyl-ketone (2c)/vinylsulfone (2f) smoothly delivered 3i-m in excellent enantioselectivity and s factor of 56 to >200. The m-substituted electron donating (OEt, Me) and chloro-bearing aryl-sulfoximines underwent olefnation with 2b to give the desired products 3n-p with s factor of 24 to 111. Even the sterically hindered m,m 0 -dimethyl substituted aryl sulfoximine 1i reacted well, yielding 3q (36%, 97.9 : 2.1 er, s factor of 107). The reaction of heteroaryl bearing 2-thiophenyl-2-pyridylsulfoximines (1m) with 2b afforded the olefnation product 3r (32%, 96.2 : 3.8 er, s factor of 51). 13 Next, we investigated the feasibility of Pd-catalyzed C-H arylative KR of sulfoximines (Table 4). 14 The reaction of N-Boc-3-methoxyphenyl-2-(3-methylpyridyl) sulfoximine ( 1b) with (4-CF 3 )Ph-Bpin (4a; 2.0 equiv.) was performed under the catalytic conditions of entry 10, Table 2. Pleasingly, the desired product (S)-5a was obtained in 94% ee with s factor of 39 along with the recovery of (R)-1b in 20% ee and 18% conversion (entry 1). The oxidant Ag 2 O played a vital role; the conversion was increased to 51% (entry 2). Carrying out the reaction at 60 C enhanced the s factor to 50 (entry 3). The s factor was raised to 64 with reaction conversion 41% and 94% ee of (S)-5a, when trifluorotoluene (TFT) was used (entry 4). Performing the reaction with 20 mol% L8 improved the outcome (entry 5). Importantly, reaction concentration from 0.1 M to 0.067 M led to (S)-5a (94% ee) and (R)-1b (88% ee) with 48% conversion and s factor of 95 (entry 6); this catalytic system was thus able to provide a balanced outcome. We next probed sulfoximines KR via enantioselective C-H arylation with arylpinacol boronate esters (Table 5). The reaction of 1b with various arylpinacol boronate esters having electron withdrawing groups [p-CF 3 (4a), m-CF 3 (4b), m-COMe (4c), and p-F (4d)], electron donating groups [p-Me (4e) and p-OMe-m-OEt ( 4f))] at the aryl motif independently led to the arylative resolution products 5a (96.1 : 3.9 er, 42%), 5b (96.5 : 3.5 er, 43%), 5c (97.5 : 2.5 er, 41%), 5d (98.4 : 1.6 er, 40%), 5e (97.2 : 2.8 er, 41%), and 5f (96.4 : 3.6 er, 44%), respectively, with s factor of 70-168 and conversion 46-49%. Moreover, the precursor (R)-1b was isolated in 41-46% yield with good enantioselectivity. The labile -Cl group was tolerated under the Pd-catalytic system, making 5g (97.8 : 2.2 er, 39%) with an s factor of 117. Notably, p-conjugated naphthyl-enabled sulfoximine resolution product 5h (99.0 : 1.0 er, s factor of >200) was reliably accessed. Next, the arylation of m-OEt-phenyl bearing sulfoximine 1f with 4a provided 5i (>99% ee) with s factor of >200. Likewise, 5j (97.2 : 2.8 er, s factor of 110) was made from the arylation of 2-naphthyl containing sulfoximine 1j with 4e. The sterically bulky o-tolyl enabled sulfoximines 1k and 1l were successful in undergoing arylation with 4a/4c/4e to afford 5k-n in good enantioselectivity; the moderate s factors of 19-24 and conversions (c ¼ 29-38%) are considered suitable. We performed a theoretical study to unveil the reaction mechanism (Fig. 2 and 3). The MPAA ligand coordination to the metal center lowers the energy barrier of the CMD step, forming a semiplanar fve membered ring. 15 We believe the CMD step could be the main responsible for the kinetic resolution. This hypothesis has been previously validated by Wu et al., who also focused their study on the CMD as determining step. 15 Based on their fndings, and considering the plane defned by the coordination of MPAA to the Pd, the bulky a-side chain of the ligand (above the plane) pushes the N-Boc moiety down to avoid steric hindrance (Fig. 2 and 1-C-II). Thus, sulfoximine phenyl group coordination complex with Pd-MPAA can point upward (U) or downward (D) on the plane, with R or S confgurations. This translates to four possible CMDs: CMD Pyr-UR , CMDPyr-US, CMDPyr-DR, and CMD Pyr-DS . The CMDs adopt a 6-membered palladacycle with twisted boat conformation. In case of upward phenyl group linkage (CMDPyr-UR and CMDPyr-US), the sulfur atom and its substituents are located above the plane; while these substituents are below the plane for CMDPyr-DR and CMDPyr-DS. In agreement with Wu's observations, 15 the C1-N2-Pd-O3 dihedral angle for CMDPyr-UR and CMDPyr-US is ca. 170 , which generates a high steric interaction when compared with the ca. 140 for CMDPyr-DR and CMDPyr-DS. These latter are favored by hydrogen bond interactions, making the combination of steric and electronic effects accounting for a difference of nearly 10 kcal mol 1 in each enantiomer (Fig. 2). The preference for the S confguration by $2.5 kcal mol 1 over the R isomer, lies in a steric clash of the NBoc group with the methyl group from the pyridine moiety and in consequence with the phenyl group causing an energetically demanding a Reaction conditions: rac-1 (0.2 mmol), 4 (2.0 equiv.), Pd(OAc) 2 (10 mol%), L8 (20 mol%), Ag 2 O (2.0 equiv.), 2-Cl-BQ (0.5 equiv.), TFT (3.0 mL), 60 C, 3 days. b Yield of the isolated arylation product. Calculated conversion, C ¼ eeSM/(eeSM + eePR). Selectivity (s arrangement. The coordination of both 'N' atoms in sulfoximine 1a forms int-0 with the displacement of acetic acid, where the S-confguration at sulfur is 1.0 kcal mol 1 more stable than the R one (Fig. 3). Prior to deprotonation, a cis coordination of aryl group to the N-protected moiety of the MPAA-ligated intermediate occurs. This assists the CMD process by establishing the absolute confguration of the sulfur motif. This calculation fully complies with the experimental observations of the resolution selectivity (calc. 98 : 2, exp. 98 : 2; Fig. 3-III). Notably, the experimentally observed S-int-2Pyr is thermodynamically favored over R-int-2Pyr isomer by 6 kcal mol 1 . In retrospect, the CMD transition states of int-1S]N (Fig. 3-I) and int-1S¼O (Fig. 3-II) lie much higher than int-1Pyr (Fig. 3-III), and their respective DGs do not coincide with the experimental fndings. Of note, the CMD process through int-1pyr is endergonic (Fig. 3-III); thus, the calculated R/S ratio is relevant if the next steps display lower free energies of activation than the CMD Pyr transition states. However, the system becomes too large to study the insertion step; simplifcation is therefore needed. Since we aim to distinguish the absolute confguration at the sulfur atom, a monodentate ligand for example, acetyl-Lalanine instead of bulky mono-protected threonine moiety was used for modelling purposes. 11 The olefn insertion with metalated sulfoximine (made by the coordination of S]N and Pyr) is next considered (Fig. 4). The corresponding S]N coordination with R confguration Int-3 S]N is found most stable (Fig. 4). The detailed analysis of transition states (INS) occurred in the CMD revealed that the pyridine directed insertion (INS Pyr ) involves lowest energy barriers (17.41 kcal mol 1 for the S isomer and 21.80 for the R isomer); see Fig. 4. This results a fnal selectivity >99 : 1 (Fig. 4). This exergonic step, thus, funnels the reaction without affecting the ratio earlier dictated by the CMD. Interestingly, both INS Pyr and INS s]N structures are same (ignoring confguration); since both DGs (S]N and Pyr) are coordinated to the metal center in their corresponding products (int-4; Fig. 4). The synthetic potential of chiral sulfoximine was next probed (Scheme 1). The trifluoroacetic acid (TFA) mediated N-Boc deprotection of (R)-1b provided chiral sulfoximine (R)-6 (>88% ee). Next, reduction of (R)-6 led to chiral sulfoxide (R)-7 (90% ee) when exposed to t-BuNO 2 at rt for 2 h. The N-Boc deprotection and intramolecular Michael cyclization to the activated olefnmoiety of (S)-3c smoothly delivered 8 (as a single diastereomer) in 95% ee. A TFA assisted N-Boc deprotection and oxidative intramolecular C-N bond formation of (S)-5d furnished (S)-9 (93% ee, 62% yield). ## Conclusions In summary, a Pd(II)-catalyzed pyridyl substituted KR of sulfoximines through C(aryl)-H alkenylation and arylation has been revealed. The transformation addresses the inherent challenges in the KR of coordinatively active pyridyl-enabled sulfoximines (highly susceptible to TM-catalyst quenching) with no prochiral center in the presence of chiral amino acid MPAA ligands and Pd(II)-catalyst. The common functional groups were tolerated under Pd-catalysis exhibiting good substrate scope for C-H alkenylative and arylative sulfoximines KR products in high enantioselectivity with s factor up to >200. In-depth DFT studies uncover the salient features of coordination selectivity of pyridyl-group over sulfoximine imine.

chemsum

{"title": "Kinetic resolution of sulfur-stereogenic sulfoximines by Pd(<scp>ii</scp>)\u2013MPAA catalyzed C\u2013H arylation and olefination", "journal": "Royal Society of Chemistry (RSC)"}

controlled_sequential_assembly_of_metal-organic_polyhedra_into_colloidal_gels_with_high_chemical_com

## Abstract: Assembling many chemical components into a material in a controlled manner is one of the biggest challenges in chemistry. Particularly porous materials with multivariate character within their scaffolds are expected to demonstrate synergistic properties. In this study, we show a synthetic strategy to construct porous networks with multiple chemical components. By taking advantage of the hierarchical nature of a colloidal system based on metal-organic polyhedra (MOPs), we developed a two-step assembly process to form colloidal networks; assembling of MOPs with the organic linker to the formation of MOP network as a colloidal particle, followed 2 by further connecting colloids by additional crosslinkers, leading to colloidal networks. This synthetic process allows not only for the use of different organic linkers for connecting MOPs and colloidal particles, respectively, but for assembling different colloidal particles formed by various MOPs. The proof-of-concept of this tuneable multivariate colloidal gel system offers an alternative to developing functional porous soft materials with multifunction. ## Introduction Increasing the chemical complexity by assembling different functional components is key for creating complex architectures with not only material properties beyond a single function but also tailorable characters. Such complex systems have attracted great interests particularly for crystalline and ordered porous solids such as metal-organic frameworks (MOFs) and covalent organic frameworks (COFs), in which more than two functional moieties (e.g., ligands with different functional groups or various metal clusters) can be integrated within the same structure. The chemical complexity, together with the decorated pore surface of these multivariate porous frameworks, showed significant benefits in developing functional materials for catalysis , sensing and separation applications . Yet, the lack of predictable arrangement of mixed compositions in the structure makes it difficult to control the positioning of the specific molecular components in the framework scaffold. The multivariate character of these materials is not limited to a single motif at the molecular level but can be integrated at longer length scales in order for the chemical complexity to emerge, for instance, via epitaxial growth of two distinct MOF or the control of the size and shape of MOF particles resulting to different layout of mesostructures (colloidal MOFs) and material behaviors (e.g., photonic or sensing property). Nevertheless, in the first case, it remains challenging to increase the chemical complexity by integrating various types of material due to the requirement for matching the lattice parameters at the crystal interface, while for the second one, the mesostructures of these types of assembled colloidal MOFs are usually not stable due to weak interactions, and it is still challenging to assemble several distinct nanocrystals into a single mesostructure. To tackle these issues, one can consider using a colloidal gel system, in which chemically different mesoscale colloidal particles can be interconnected through electrostatic interaction or chemical bonding to form multicomponent colloidal networks. Interestingly, depending on the interaction between particles, the assembled structures of two chemically distinct colloids can either (i) be self-sorted into a network and form double interpenetrated colloidal networks, or (ii) lead to the preferential interaction between two different colloids (for instance, electrostatic interaction) yielding to the formation of a mixed colloidal network. In recent years, a number of MOF particle-based gels have been reported, proving the possibility of creating new types of multicomponent colloidal gels with versatile functionalities. Metal-organic polyhedra (MOPs) with discrete cage structures present a good alternative to MOFs due to their processability in solution and lability. When used as secondary building blocks, MOPs can be assembled with additional linkers into colloidal particles or colloidal gel networks. However, all of these systems have been assembled from a single MOP component. Therefore, integration of more than two distinct components during the self-assembly process remains a major challenge due to the lack of methodology. Here, we show a coordination chemistry approach to form multicomponent hierarchical structures from interconnected MOPs. Our methodology takes advantage of available coordination sites exposed on the surface of the MOPs, where first a ditopic N-donor ligand leads to the formation of MOP networks as colloidal particles, which can be further crosslinked into a colloidal gel network by the addition of a secondary N-donor ligand. The transfer of MOPs to complex colloidal gel networks using this two-steps assembly strategy will demonstrate the feasibility to install selected functional moieties at the scales of the colloidal particles or colloidal gel networks (Figure 1). This will allow the precise engineering of the supramolecular structure and advance the multi-functionalities in a controlled manner. ## Controlled synthesis of coordination polymer particle (CPP) In this study, we chose the stable rhodium-based MOP, [Rh2(bdc)2]12 (HRhMOP), assembled from 12 dirhodium acetate paddlewheels and 24 benzene-1,3-dicarboxylate (bdc) with cuboctahedral geometry for the formation of colloidal particles. By taking advantage of the reactivity of the external axial site of the rhodium paddlewheel, we previously demonstrated the polymerization of MOPs with bisimidazole derivatives such as 1,4-bis[(1H-imidazol-1yl)methyl]benzene (bix) as linkers. On the one hand, the gradual addition of the ditopic Ndonor ligand bix to the MOP solution resulted in the formation of colloidal spheres (named as coordination polymer particles, CPPs). On the other hand, the generation of kinetically-trapped MOP as RhMOP(bix)12 by adding an excess amount of bix to the RhMOP solution produced a colloidal gel network. Note that the latter strategy leads to the spontaneous formation of a colloidal gel network once the MOP and bix are mixed, which does not allow for further control of the assembly process, including its termination at the stage of colloidal formation. In order to achieve a multicomponent colloidal structure, one needs to use colloidal particles as initial building blocks. Here, we first optimized the reaction condition for the CPP formation based on HRhMOP with different types of ditopic N-donor ligands and various stoichiometric ratios. Then, we optimized the condition to connect these CPP with another bisimidazole crosslinker to form the colloidal gel network. Finally, two chemically distinct but isostructural CPPs (one with HRhMOP and the other with BrRhMOP) were mixed to yield multicomponent colloidal gel networks (BrRhMOP, [Rh2(5-Br-bdc)2]12, 5-Br-bdc = 5-bromobenzene-1,3dicarboxylate). A series of ditopic N-donor bisimidazole ligands, bix, 4,4'-bis(1H-imidazol-1ylmethyl)biphenyl (bibph), and 1,12-di(1H-imidazol-1-yl)dodecane (bidod) were selected as bridging linkers to sequentially connect MOPs or CPPs. In order to process HRhMOP in solution, the cage was solubilized according to a reported coordination solubilizer method, which requires the coordination of monodentate N-dodecylimidazole (diz) to all of the 12 axial sites of Rh paddlewheels exposed on the surface of the molecule to yield soluble HRhMOP(diz)12. Formation of CPPs was then investigated via sequential addition of 0.5 molar equivalents per step of three distinct bisimidazole linkers, bix, bibph, or bidod, into the DMF solution of HRhMOP(diz)12. Replacement of diz by bisimidazole linkers (L) initiates the linkage between HRhMOP(diz)x(L)y cages and leads to a MOP network in the form of CPPs. The use of bix as a linker provided dispersible CPPs, while the bibph and bidod generated undesired agglomerates of CPPs, which make it difficult to be further processed (Figure S5). Therefore, a series of CPPs made with bix as linkers were selected for further investigation. In the MOP network linked by bix, the HRhMOP acts as a node linked up, in principle, to 12 other MOPs. The effective connectivity of MOP can be estimated from the chemical composition of the CPP determined by proton nuclear magnetic resonance measurement ( 1 H NMR, Figure S8) and represented by the branch functionality, f, a parameter that indicates the number of rhodium sites per MOP that are used to connect with neighboring MOPs (see Supplementary Information). In order to correlate the added bix quantity with the branch functionality, the corresponding CPPs were synthesized with different molar equivalents of bix linker (3, 4, 6 and 12 molar equivalents) added. The calculated f increases up to 8 by the addition of 6 molar equivalents of bix but decreases by the addition of an excess amount of bix, which is most likely due to the coordination of bix as a mono-dentate fashion (Table 1). As further addition of bridging ligands is required to allow connections between CPPs to form a colloidal gel network, the obtained CPPs should have enough accessible coordinative sites at the exohedral HRhMOP. Therefore, the CPPs formed with 3 molar equivalent addition of bix and having a relatively low f value were selected for further studies. The size of CPP is expected to affect the macroporosity and connectivity of the corresponding colloidal gel networks. As reported previously, the CPP formation process can be tuned by the molar equivalent of bix added at each step, in which the first addition would generate nuclei and the following addition would contribute to particle growth. Hence, the effects of different concentrations of HRhMOP(diz)12 and stepwise amount of bix added on the resulting CPP size were investigated by dynamic light scattering (DLS) titration experiments (Figure 2). When the addition of bix changed from 0.5 molar equivalents to 3.0 molar equivalents for each step, the size of final CPPs was decreased from 98 ± 14 nm to 55 ± 10 nm as also confirmed by scanning electron microscopy (SEM) images (Figure 3). This size reduction is attributed to the larger numbers of nuclei formed, yielding to smaller CPPs. In addition, the increase in the concentration of HRhMOP(diz)12 from 0.93 mM to 1.86 mM raises the collision probability of particles, leading to a larger size of CPP (189 ± 30 nm). In contrast, lowering the concentration to 0.46 mM led to a smaller CPP particle size (62 ± 10 nm). ## Investigation of colloidal gel formation from CPPs In order to demonstrate the possibility of forming colloidal gels with CPPs, two types of CPPs with different sizes were produced by different addition rates of bix (CPP-large: 98 ± 14 nm and CPP-small: 55 ± 10 nm) were selected for further investigations. We firstly attempted the gelation with various concentrations of CPP-large (4 mg/mL, 8.5 mg/mL, 17 mg /mL, and 42 mg/mL in DMF) by adding 12 molar equivalents bix and reacting at 80 ºC for 2 h. Note that the molar equivalence of added bix here was calculated as a relative molar ratio to HRhMOP in the CPP, which was estimated from the weight and 1 H NMR spectra. Only the high concentration of CPP-large (42 mg/mL) formed corresponding gel, but the CPP with lower concentrations gave suspensions (Figure S10). To gain more insight into the gelation mechanism, the gelation process of 42 mg/mL CPP-large and CPP-small, after the addition of 12 molar equivalents of bix, were monitored by time-resolved dynamic light scattering (TRDLS) measurements at 50 ºC. 44] The timeaveraged scattering light intensity ⟨𝐼⟩𝑇 was found to be almost constant at the early stage of the reaction before fluctuating at 38 min for CPP-large (Figure 4a). This random fluctuation is a clear indication of the sol to gel transition due to the loss of ergodicity during the percolation of colloidal particles. Insight into the gelation mechanism can be extrapolated from the timeevolution of the intensity correlation functions (ICFs), g (2) (τ)-1, which is a function measuring the similarity between scattering light signal at different delay times (Figure 4b). Indeed, ICFs were fitted by a single stretched exponential function and give access to the following parameters: where σI 2 is the initial amplitude of the ICF, τf is the relaxation time related to the diffusion rate of particles, and β is the stretched exponent that reflects the distribution of τf . τf is further related to the correlation length ξ, which reflects the evolution of the density of the network during the reaction time in the solution phase, through Equation 2: where q, kB, T, η are the scattering vector, the Boltzmann constant, the absolute temperature, and the viscosity of media, respectively (see the Supplementary Information for the derivation of Equation 2). In the first 10 min of the reaction, τf and the corresponding correlation length ξ were almost constant, indicating that the CPPs were individually diffusing and not aggregating with each other (Figures 4c and S13h). This was also proven by the values of β close to unity (≈ 0.9), which reflects the relatively narrow size distribution of the CPP as observed by the SEM measurements. After this induction period, τf increased while β decreased toward the gelation point. This simultaneous change of τf and β indicates that the average particle size and size distribution are no anymore uniform, which is most likely explained by the stochastic aggregation of colloidal particles and the formation of a gel network. This agrees with the drastic increase of the correlation length ξ, (Figure S13h). The TRDLS experiment on CPPsmall also showed a similar trend to the CPP-large as shown in Figure S13. Differences in gelation time as a function of temperatures were also observed by TRDLS measurements. At 50 ºC, the sol-gel transition for CPP-small occurs at 107 min after the addition of bix. In contrast, the gel was formed in 19 min for the same system heated at 80 ºC (Figure S12). This is most likely caused by the increase in the diffusion rate of particles and the kinetics of ligand exchange reaction accompanying the colloidal aggregation when increasing the temperature. In addition, the gelation time is dependent on the CPP size as seen from the gelation thresholds of 107 min and 38 min measured for CPP-small and CPP-large, respectively at 50 ºC (Figure S13). This can be explained by the difference of van der Waals interaction between colloidal particles, which is estimated to be proportional to the particle diameter from Hamaker's theory. These examples illustrate the possibility of controlling gelation time by changing reaction temperature and CPP size, while retaining the chemical composition of the resulting colloidal gel network. Moreover, the reduction of the number of bix added from 12 to 6 eq. led to an increase in the gelation time (Figure S14), which is most likely attributed to the reduced coordinative interactions between CPPs that hindered the gelation process. For the purpose of accumulating more data points during the gelation process, the gelation reaction at 50 ºC is used for the following investigation. To further confirm that the CPP gelation is induced by the second addition bix, a control experiment was performed using TRDLS to follow the reaction of CPP-large without the addition of bix (Figure S14). CPPlarge suspension showed almost constant τf and β during the 6 h of heating at 50 ºC. This indicates that CPP is stable itself in DMF and gelation is indeed driven by crosslinking of CPPs mediated by bix molecules. This successful assembly via two-step addition of bix to form CPP-based gel envisaged incorporating different crosslinkers (bibph and bidod) in a second step. However, only the addition of bidod to more than 42 mg/mL CPP-large was able to form a gel under the synthetic condition of 80 ºC for 2 h. In contrast, the addition of bibph gave a suspension after the 2 h reaction at 80 ºC (Figure S15). We attributed this difference in gelation to the variation in crosslinking reactivity of the linkers as: bidod > bix > bibph, which was proven by the difference of gelation time observed by TRDLS (Figure S16). Therefore, to investigate the multivariate colloidal gel, the system with CPP-large as cores and bidod as secondary linkers was selected and its gelation condition was optimized to obtain more mechanically durable colloidal gel. It is expected that the strength of the gels is predominantly affected by the addition of bidod due to the cross-linkage, interaction between CPPs, and the chemical environment. To clarify the effect of crosslinker amount on the strength of the gels, a series of CPP-large/bidod gels with addition of 3, 6, 9, 12 molar equivalents bidod have been synthesized (Figure S18) from which the compositions were determined by 1 H NMR spectroscopy (Figure S19). The average number of diz, bix, and bidod per HRhMOP was estimated using the average integration of selected representative signals (Figure 5a). Interestingly, the relative number of bix molecules to HRhMOP remained almost the same compared to the original CPP itself regardless of the quantity of bidod added. In contrast, the relative number of diz linearly decreased and that of bidod linearly increased as a function of the amount of bidod added. This clearly indicates that bix and HRhMOP molecules are forming a stable coordination network inside CPPs and newly added bidod can only replace diz molecules in CPP. Unlike the CPP formation, branch functionality f is no longer suitable to reflect the connectivity between CPPs since bidod can crosslink MOPs within the same CPP as well as with the neighboring CPPs. Therefore, the relation between the quantity of bidod addition and the stiffness of the gels was further investigated by rheology measurements (Figures 5b and S20). The CPP-large/bidod gels were prepared in situ in the rheometer by heating the mixture of CPP-large suspension and 6 or 12 equivalents bidod at 80 ºC for 20 min, followed by determination of the storage (G') and loss (G'') modulus at 25 ºC. The G' was always higher than G'' and they were almost constant independently of the shearing frequency and strain, which demonstrate their linear elastic property. It was found that the G' and G'' of CPP-large/12 eq. bidod gel are greater than those of the CPP-large/6 eq. bidod gel, which agrees with the increased number of bidod present in the gel network. This result further confirms that the addition of bidod enhances the crosslinking between CPPs and forms a more robust gel. Furthermore, the successful formation of gels from CPP-small with either bix or bidod was confirmed (Figures 6 and S13), which illustrates the possibility of altering the network structures. ## Multivariate colloidal assembly In order to demonstrate the versatility of material composition, isostructural MOPs, with a substitution at 5 th position of benzene-1,3-dicarboxylate with hydrophilic functionality (-OH), To demonstrate the possibility to create a multivariate colloidal gel network formed from the mixing of CPP-Br with 'pristine' CPP based on HRhMOP, the presence of Br component in the colloidal aerogel network was mapped via energy dispersive X-ray analysis (EDX) in transmission electron microscope (TEM) due to its electron-rich property (Figure 7). ## OHRhMOP The EDX image in Figure 7a shows a homogeneous distribution of Br (green) and Rh (red) atoms in the system, which is also confirmed by a 1:1 ratio of Br and Rh from the elemental composition analysis (Figure S22c). Then the possibility to increase the complexity of this ## Conclusion The two-step hierarchical construction strategy developed here has allowed us to fine-tune the properties of the colloidal gel network via the control of the CPP size, gel composition (CPP core and its connectivity of gel network) and crosslinking degree as demonstrated by the combination of TRDLS, 1 H NMR and rheology techniques. In addition, the possibility to i) have control over the colloidal size without altering the core components, ii) modify the mesoscale connectivity with different crosslinkers and iii) install functional moieties via alteration of RhMOPs and CPPs open up a new alternative to develop multivariate colloidal gel via hierarchical assembly to form multi-functional materials for potential applications such as sensing and separation. ## Supporting Information Supporting Information is available from the Wiley Online Library or from the author.

chemsum

{"title": "Controlled sequential assembly of metal-organic polyhedra into colloidal gels with high chemical complexity", "journal": "ChemRxiv"}

cd-driven_surface_reconstruction_and_photodynamics_in_gold_nanoclusters

## Abstract: With atomically precise gold nanoclusters acting as a starting unit, substituting one or more gold atoms of the nanocluster with other metals has become an effective strategy to create metal synergy for improving catalytic performances and other properties. However, so far detailed insight into how to design the goldbased nanoclusters to optimize the synergy is still lacking, as atomic-level exchange between the surfacegold (or core-gold) and the incoming heteroatoms is quite challenging without changing other parts. Here we report a Cd-driven reconstruction of Au 44 (DMBT) 28 (DMBT ¼ 3,5-dimethylbenzenethiol), in which four Au 2 (DMBT) 3 staples are precisely replaced by two Au 5 Cd 2 (DMBT) 12 staples to form Au 38 Cd 4 (DMBT) 30 with the face-centered cubic inner core retained. With the dual modifications of the surface and electronic structure, the Au 38 Cd 4 (DMBT) 30 nanocluster exhibits distinct excitonic behaviors and superior photocatalytic performances compared to the parent Au 44 (DMBT) 28 nanocluster. ## Introduction Metal synergy is of paramount importance as the rationale to modulate the intrinsic properties of metal nanoparticles. 1,2 However, the precise synergistic interaction in an intermetallic nanoparticle has so far been elusive, due to the challenges in determining the atomic-level arrangement of the metal heteroatoms in the nanoparticle. Atomically precise metal nanoparticles (often called nanoclusters) lead to unprecedented opportunities in signalling clear directions to exploit the cooperativity between the two metal elements within a single nanocluster. 3 Thiolate-protected gold nanoclusters, Au n (SR) m , where n is the number of gold atoms and m is the number of thiolate ligands, SR, have gained momentum over the past few years as an exciting area and have opened up new horizons in precise tailoring of the composition and structure to control the physicochemical properties. The Au n (SR) m nanoclusters are typically confgured with an inner gold core (or kernel) and various surface motifs, in which the motifs containing both gold and thiolate resemble staples. Both the gold core and the surface motifs can contribute to the physicochemical properties such as the optical and electronic properties, as well as catalysis. It has been recognized that substituting one or more gold atoms in either the core or the motifs with other metals can tune the overall performances of the parent nanoclusters. Therefore, it has become possible to access the previously inaccessible metal synergy in the bimetallic nanoclusters with atomic-precision. Among the gold-based bimetal nanoclusters, cadmiumcontaining bimetal clusters provide synergistic strategies to adjust the electronic structures and further modulate the physicochemical properties in the clusters, since Cd has one more valence electron than Au. 21,26,27 Cd introduction usually causes surface reconstruction of gold nanoclusters. For example, Au 19 Cd 2 (SR) 16 was obtained through the substitution of two neighboring surface Au atoms with one Cd with the cuboctahedral Au 13 unchanged. 26 Au 19 Cd 3 (SR) 18 was formed by retaining the icosahedral Au 13 core but only changing the surface of Au 25 (SR) 18 . 27 However, the surface reconstruction strategy remains challenging and no examples of bimetal clusters formed without breaking the face-centered cubic (fcc) core of the parent gold clusters have been documented, which might thus impede gaining a higher understanding of how to tailor the surface structure of gold-based nanoclusters and accordingly optimize their synergy. Herein, we report our success in synthesizing a Au 38 Cd 4 (-DMBT) 30 ## Results and discussion X-ray crystallography analysis shows that the parent Au 44 (-DMBT) 28 nanocluster is composed of an Au 26 kernel, six Au 2 (SR) 3 and two Au(SR) 2 staples (Fig. 1a, c and Table S1 †). The formula of Au 44 (DMBT) 28 is further confrmed by electrospray ionization mass spectroscopy (ESI-MS, Fig. S1a †). The structural framework of Au 44 (DMBT) 28 is identical to that of the reported Au 44 (TBBT) 28 (TBBT ¼ 4-tert-butylbenzenethiol) (Fig. S2 †), 28 both of which can be assembled into the layered structures (Fig. S3-S5 †). Notably, a signifcant difference is observed in the layer's interior, where all the molecules of Au 44 (TBBT) 28 in the layer (marked with the same color, Fig. S3 †) are packed along the same direction, while Au 44 (DMBT) 28 molecules are arranged in different directions (Fig. S5 †). Such a difference may be ascribed to the different steric hindrance between TBBT and DMBT. The UV-vis-NIR spectra of the two Au 44 (SR) 28 nanoclusters show only small deviations. As shown in Fig. S6, † the prominent peak at 380 nm for Au 44 (TBBT) 28 is slightly red-shifted to 388 nm for Au 44 (DMBT) 28 , and the broad peaks at 650 and 725 nm become apparent when TBBT is replaced by DMBT. With Au 44 (DMBT) 28 as a starting unit, a Cd-doped nanocluster was further synthesized via an ion-exchange strategy. From ESI-MS data (Fig. S1b †), the prominent peak at 6025.43 m/ z with a +2 charge is assigned to Au 38 Cd 4 (DMBT) 30 (theoretical value: 6025.48 m/z), which is further confrmed by the excellent match between experimental and calculated isotopic patterns (inset of Fig. S1b †). Single crystallography analysis reveals that Au 38 Cd 4 (DMBT) 30 contains a 26-Au-atom kernel, two Au 5 Cd 2 (-SR) 12 staples, two Au(SR) 2 staples and two bridging SR ligands, as shown in Fig. 1b, d, and Table S2. † Note that the retained kernel of Au 38 Cd 4 (DMBT) 30 experiences a slight distortion from "slender" to "stocky" in comparison with that of the parent Au 44 (DMBT) 28 (Fig. 1e-h). Further analysis shows that the Au 26 kernel in Au 38 Cd 4 (DMBT) 30 can be viewed as the assembly of tetrahedral Au 4 units in a double-helical mode, as well as that in Au 44 (DMBT) 28 (Fig. 2). Furthermore, the two nanoclusters have almost identical distances between neighboring Au 4 units, which is clearly manifested in the similar Au-Au bond lengths according to the different positions of the Au atoms (Fig. S7 †). Therefore, Au 38 Cd 4 (DMBT) 30 can be viewed as the gentle surface reconstruction without breaking the double-helical Au 26 kernel based on the parent Au 44 (DMBT) 28 . In addition, Au 38 Cd 4 (-DMBT) 30 is also patterned along different directions in the layer structure (Fig. S8 †). To gain an in-depth insight into the Cd-induced surface reconstruction mechanism, density functional theory (DFT) calculations were performed. Starting from the Au 44 (SR) 28 cluster, as presented in Fig. S9, † To investigate the electronic structure changes induced by Cd-atom surface modifcation, the optical adsorption spectra of the Au 44 (DMBT) 28 and Au 38 Cd 4 (DMBT) 30 nanoclusters were measured. The absorption peaks of Au 38 Cd 4 (DMBT) 30 are mainly centered at 400, 465, 550 and 678 nm (Fig. 3b), which differ from those observed in the parent nanocluster (387, 452, 635 and 725 nm; Fig. 3a). These optical features can be well reproduced by theoretical calculations (Fig. 3a, b and S10 †). The Kohn-Sham (KS) molecular orbital (MO) energy levels and atomic orbital components in each KS MO of Au 44 (SR) 28 and Au 38 Cd 4 (SR) 30 suggest that the absorption peaks mainly involve the Au(sp) / Au(sp) transitions (Fig. 3c and d). In particular, for Au 44 (SR) 28 , the frst absorption peak centered at 734 nm originates from the highest occupied molecular orbital / the lowest unoccupied molecular orbital (HOMO / LUMO) transition, while for Au 38 Cd 4 (SR) 30 , the frst absorption peak centered at 695 nm originates from the HOMO / LUMO, HOMO / LUMO+1, HOMO / LUMO+4, HOMO1 / LUMO, HOMO1 / LUMO+1, and HOMO1 / LUMO+5 transitions. The more complex orbital transitions in Au 38 Cd 4 (SR) 30 than in Au 44 (SR) 28 can be attributed to the dopant Cd. This behaviour can also be observed for other absorption peaks. Moreover, femtosecond and nanosecond carrier dynamics of the two nanoclusters were measured via time-resolved transient absorption (TA) spectroscopy to decipher their potential energyrelated applications. The femtosecond-resolved TA spectra of the Au 44 (DMBT) 28 and Au 38 Cd 4 (DMBT) 30 nanoclusters are provided in Fig. 4a and b. Similarly, both Au 44 (DMBT) 28 and Au 38 Cd 4 (DMBT) 30 nanoclusters showed broad excited state absorption (ESA) signals overlapped with ground state bleaching (GSB) peaks near 675 nm. We selectively extracted the TA spectra at different delay times, combined with the dynamic traces probed at 515 and 675 nm to study the transient evolution and the relaxation dynamics (Fig. 4c-f). A 0.6 ps process at the early stage, which is attributed to the ultrafast internal conversion from higher excited states to lower excited states, 29 was observed in the two nanoclusters (Fig. S11 and Table S3 †). It is worth noting that the major divergence between the two nanocluster systems emerged after a delay of 2 ps. For Au 44 (-DMBT) 28 , the TA spectra remained nearly unchanged after 2 ps (Fig. 4c), which is consistent with the flat decay kinetic traces shown in Fig. 4e. A 19 ps process obtained by exponential ftting was ascribed to the structural relaxation caused by conformational changes after pumping. For Au 38 Cd 4 (DMBT) 30 , interestingly, an obvious spectral transformation was observed and the lifetime of this component was determined to be 57 ps (Table S3 †), which differs from the 19 ps structural relaxation observed in Au 44 (DMBT) 28 and might be related to the charge transfer states between the ligand and the metal core of Au 38 -Cd 4 (DMBT) 30 , The distinguishable electronic and optical properties of the two nanoclusters would apparently impact their catalytic properties. Thus, visible light-driven degradation of methyl orange was selected to explore the photocatalysis of the two nanoclusters. From Fig. 5a and b, within 50 min, methyl orange can be completely degraded on the Au 38 Cd 4 (DMBT) 30 catalyst under visible light illumination, while on the Au 44 (DMBT) 28 catalyst it was completed in 70 min. The plots of methyl orange degradation on the catalysts versus reaction time further indicate the better photocatalytic performance of the Au 38 Cd 4 (-DMBT) 30 catalyst (Fig. 5c). Electrochemical impedance spectroscopy was performed to investigate the interfacial transfer of electrons. In Fig. 5d, the semicircular diameter of Au 38 Cd 4 (DMBT) 30 was smaller than that of Au 44 (DMBT) 28 , which implies faster electron-transfer in the Au 38 Cd 4 (DMBT) 30 system. The photocatalysis difference in the two cluster catalysts is suggested to arise from their different equilibria established between the carrier recombination and the electron transfer influenced by metal synergy. ## Conclusions In summary, we have developed a Cd-driven surface reconstruction strategy for synthesizing a new Au 38 Cd 4 (DMBT) 30 bimetallic nanocluster with the fcc Au 26 core retained from the parent Au 44 (DMBT) 28 nanocluster. The two nanoclusters that exhibit elegant patterns of Au 4 tetrahedra show distinct differences in the electronic structures, optical properties, and photocatalytic performances. Beyond the Cd-mediated surface reconstruction case, we anticipate that this heteroatom-doping mechanism will fnd applications in using gold and other metals in a series of challenging gold-based nanocluster formations and tuning of their intrinsic properties. ## Conflicts of interest There are no conflicts to declare.

chemsum

{"title": "Cd-driven surface reconstruction and photodynamics in gold nanoclusters", "journal": "Royal Society of Chemistry (RSC)"}

purification_of_indium_by_solvent_extraction_with_undiluted_ionic_liquids

## Abstract: A sustainable solvent extraction process for purification of indium has been developed from a chloride aqueous feed solution using the ionic liquids Cyphos® IL 101 and Aliquat® 336. The high affinity of indium(III) for the ionic liquid phase gave extraction percentages above 95% over the HCl concentration range from 0.5 to 12 M. Attention was paid to the loading capacity of the ionic liquid phase and the kinetics of the extraction process. An extraction mechanism was proposed based on the relationship between the viscosity of the ionic liquid phase and the loading with indium(III) ions. Even for loadings as high as 100 g L −1 , equilibrium was reached within 10 min. Due to the very high distribution ratio for indium(III), stripping of indium(III) from the ionic liquid phase was very difficult with water or acid solutions. However, indium could conveniently be recovered as In(OH) 3 by precipitation stripping with a NaOH solution. Precipitation stripping has the advantage that no ionic liquid components are lost to the aqueous phase and that the ionic liquid is regenerated for direct re-use. The extraction of some metal ions that are commonly found as impurities in industrial indium process solutions, i.e. cadmium(II), copper(II), iron(III), manganese(II), nickel(II), tin(IV) and zinc(II), has been investigated. The distribution ratios for the different metal ions show that indium(III) can be purified efficiently by a combination of extraction, scrubbing and stripping stages. This new ionic liquid process avoids the use of volatile organic solvents. ## Introduction Indium is a scarce metal with a 0.1 ppm abundance in the earth's crust comparable to that of silver. 1,2 It is labeled as a critical raw material by the European Commission, due to its high supply risk. 3,4 This risk is due to two supply and demand factors, namely, the increasing demand for applications in photovoltaics, flat screen TVs, laptops and mobile phones as well as the Chinese production monopoly. In 2008, China had a 58% share in the primary global indium production. 5 The two factors are further compounded by low substitutability and low recycling rates, resulting in higher prices and the need to raise production. 3,4 Indium is derived as a by-product from ores and is most commonly found in association with zinc ores such as sphalerite, although it also occurs in lead, copper, iron and tin ores. Indium is accumulated in low concentrations in residues formed during the processing of these ores. Therefore, indium is often regarded as an impurity that increases the production costs. 5 Furthermore, indium recycling is mainly limited to production scrap. To keep up with the rising demand for high-tech applications, the development of industrial processes for the successful recovery of indium from various primary and secondary sources is crucial. Possible sources of indium are by-products of zinc refining, flue dusts, slags and metallic intermediates, e-waste and impure indium (<99%). 1,2 Hydrometallurgical separation methods are very suitable for extracting indium from these sources. They compromise an essential part of extractive metallurgy utilized for treating complex and low-grade materials. Typically, hydrometallurgical solutions are generated by leaching the metals present in concentrates with strong acids or bases. Metals are already separated partially or completely in the leaching step, if one or more of the metals does not solubilize. Further concentration or separation generally takes place via solvent extraction or ion exchange. Once the metals have been separated, pure metals or metallic compounds can be produced by, for example, precipitation, cementation or electrolysis. Solvent extraction (SX) has been widely used as a process for separation, purification and recovery of metals, due to its simplicity of equipment and operation. Solvent extraction is based on selective and efficient transfer of the desired metal species from one phase to another, usually from an aqueous to an organic phase. 6,7 The aqueous phase, in which the metals are present, is brought into contact with a water-immiscible organic phase, consisting of one or more extractants, a diluent and sometimes a modifier. The separation is based on differences in the solubility of the metal ions in both the organic and aqueous phases due to the variations in the strength of the chemical interaction between the metal ion and the extractant. The efficiency of the extraction process depends on several parameters such as the pH, the temperature, the concentration of the metal ions in the aqueous feed and the concentration of the extractant in the organic phase. The traditional solvent extraction process for indium makes use of water-immiscible organic solvents, many of which are flammable, volatile or toxic. 8 Ionic liquids are good alternatives as an extraction medium for the development of sustainable separation processes. Ionic liquids are solvents consisting completely of ions, mostly an organic cation and an inorganic anion. 12,13 The physicochemical properties of ionic liquids can be tuned adequately for a given application. 14 Therefore, it is not surprising that water-immiscible ionic liquids have already been investigated as extractants for metals from aqueous solutions. 11, These ionic liquids often contain fluorinated anions, such as the hexafluorophosphate (PF 6 − ) or the bis(trifluoromethylsulfonyl)imide (Tf 2 N − ) anion. 22,23 Besides their high prices, in some cases, these type of ionic liquids pose a severe risk due to hydrolysis and formation of hydrofluoric acid. 24 Therefore it is better from an environmental and economical point of view to use ionic liquids with long alkyl chains instead of fluorinated anions. 22 Several non-fluorinated hydrophobic ionic liquids have already been used for the extraction of metal ions but generally not in a pure state. These ionic liquids such as Aliquat® 336 and tri(hexyl)tetradecylphosphonium chloride (Cyphos® IL 101) are usually diluted in molecular solvents prior to use. Diluents such as toluene, kerosene or chloroform are added to decrease the viscosity of the organic phase leading to an increase in mass transfer and faster kinetics. However, recently in some cases, the problem of viscosity was overcome by saturating the ionic liquid with water, by working at slightly elevated temperatures and/or by using intermediate metal feed concentrations. Major advantages of ionic liquids for application in solvent extraction processes, in comparison with traditionally used water-immiscible organic solvents, are their negligible vapor pressure and low flammability. Due to their low volatility, they can be considered as more environmental-friendly and safer alternatives to organic solvents. While the use of ionic liquids for metal extraction does offer many advantages, there are some disadvantages as well. First of all, not all types of ionic liquids can be used for extraction. Water-immiscible ionic liquids are required for solvent extraction. Secondly, extraction often takes place through an ion-exchange mechanism which leads to losses of ionic liquid components to the aqueous phase. Ionic liquids are also quite expensive, so that their use for conventional solvent extraction processes cannot be justified solely from an economical point of view. Furthermore, their generally high viscosity often leads to slow mass transfer so that long contact times are required to reach equilibrium. 33,34 In this paper, an efficient solvent extraction process for the purification of indium from a chloride medium was developed based on the use of quaternary phosphonium and ammonium ionic liquids. The feasibility of the quaternary phosphonium ionic liquid trihexyl(tetradecyl)phosphonium chloride (Cyphos® IL 101) and the quaternary ammonium ionic liquid Aliquat® 336 to serve as a possible undiluted organic phase/ extractant for the extraction of indium is investigated. Special attention was paid to the difficult stripping of indium from the ionic liquid phase after extraction. Is it shown that precipitation stripping of In(OH) 3 with NaOH is a convenient sustainable method for recovery of indium and for regeneration of the ionic liquid. ## Chemicals Trihexyl(tetradecyl)phosphonium chloride (Cyphos® IL 101, purity >97%) was purchased from Cytec Industries Inc. (Niagara Falls, Ontario, Canada). Aliquat® 336 (mixture of quaternary ammonium chlorides, with 88.2-90.6% quaternary ammonium content), InCl 3 •4H 2 O (97%), MnCl 2 (98%) and phosphor standard (1000 ppm) were obtained from Fluka (Sigma-Aldrich, Diegem, Belgium). CdCl 2 (99%), CuCl 2 (99%), NiCl 2 (98%), PbCl 2 (99%), SnCl 4 (99.99%), Na 2 EDTA•2H 2 O (99+ %), Na 2 SO 4 (99%, extra pure) and ethanol (99.8+%, absolute) were purchased from Acros Organics (Thermo Fisher Scientific, Geel, Belgium). As 2 O 3 (99.5%) was obtained from Alfa Aesar (Karlsruhe, Germany), FeCl 3 (98.5%) from Carl Roth (Karlsruhe, Germany) and ZnCl 2 (98-100.5%), CaCl 2 •2H 2 O (99.5+%) and lanthanum standard (1000 ppm) from Chem-Lab (Zedelgem, Belgium). NaOH pellets (>99% AnalaR Norma-pur®) were purchased from VWR (Leuven, Belgium). The silicone solution in isopropanol was obtained from SERVA Electrophoresis GmbH (Heidelberg, Germany) and the rhodium standard (1000 ppm) from Merck (Overijse, Belgium). Hydrochloric acid solutions were prepared from HCl (37%, Acros Organics) and Milli-Q® water. All chemicals were used as received, without further purification. ## Instrumentation and analysis methods After each extraction, the mixtures were centrifuged with a Heraeus Megafuge 1.0 centrifuge. Metal concentrations were determined using a bench top total reflection X-ray fluorescence spectrometer (TXRF; Bruker S2 Picofox) for simple matrices (mono-element system) or inductively coupled plasma optical emission spectrometer (ICP-OES; Agilent, type E730) with an axial plasma configuration for the more complex matrices (multi-element system) due to the spectral overlap and matrix effects that were encountered with TXRF. For analysis of the aqueous phase by TXRF, part of the aqueous phase was mixed with a lanthanum internal standard and Milli-Q® water until a total volume of 1 mL was obtained. The quartz glass sample carriers were first treated with 20 μL of a silicone solution in isopropanol, dried for 5 min in a hot air oven at 60 °C, followed by the addition of 5 μL of the sample and drying for 30 min at the same temperature. The metal concentrations in the aqueous phase were measured for 1000 s. For the organic phase, the lanthanum internal standard was added to a small amount of the organic phase (27 mg) and was further diluted with absolute ethanol until 1 mL. The sample carrier pre-treatment, the drying procedure and the measuring time were performed in the same way for the organic phase as described for the aqueous phase but the sampling volume was reduced to 2 µL. For analysis of the aqueous phase by ICP-OES, a calibration curve was prepared with multi-element solutions over a concentration range of 0-10 mg L −1 with a quality control of 2 mg L −1 and scandium as an internal standard. The calibration solutions and the samples were diluted in 10 vol% HCl. The spectra were measured with a power of 1.4 kW, an argon flow of 15 L min −1 and an auxiliary argon flow of 1.5 L min −1 . For the organic phase, the sample was first digested with a mixture of H 2 SO 4 and HNO 3 in a quartz beaker on a heating plate. H 2 SO 4 and HNO 3 were evaporated (∼400 °C) and the residue was dissolved in 10 vol% HCl before measurement. The calibration and measuring procedure have been performed in the same way for the organic phase as described for the aqueous phase. The viscosity of the organic phase was measured using a falling-ball type viscometer (Anton Paar, Lovis 2000 ME), densities were determined using a density meter with an oscillating U-tube sensor (Anton Paar, DMA 4500 M) and pH measurements were performed with an S220 SevenCompact™ pH/Ion meter (Mettler-Toledo) and a Slimtrode (Hamilton) electrode. A Mettler-Toledo DL39 coulometric Karl Fischer titrator was used with Hydranal® AG reagent to determine the water content of the ionic liquid. 31 P NMR spectra have been recorded on a Bruker® Avance™ II 600 MHz Spectrometer, operating at 242.94 MHz. The samples for 31 P NMR measurements were dissolved in toluene-d 8 and measured with respect to a H 3 PO 4 external reference. A delay time (d1) of 60 s was applied in the NMR pulse sequence to avoid saturation effects in the 31 P NMR spectra. All NMR spectra were analyzed with the TOPSPIN software package. The loss of Cyphos® IL 101 to the aqueous phase was determined by ICP-OES. All measurements were carried out using an ICP-OES spectrometer, model Optima 8300 DV (Perkin Elmer) equipped with axially viewed plasma. 31 P was analyzed at 177.434 nm. The internal standard was measured at 343.489 nm (Rh). Standard solutions of 31 P (3, 1, 0.5, 0.1, 0.05 ppm) were prepared in 5 vol% HCl. 1 ppm of Rh was added as an internal standard. 5 vol% HCl was used as a blank solution for calibration. All samples were diluted 10 times with a 5 vol% HCl solution and 1 ppm of Rh was added as an internal standard. All aqueous solutions were prepared with Milli-Q® water. A reagent blank solution of 5 vol% HCl containing 1 ppm of Rh as an internal standard was used for correcting the Ca-free samples. The Ca-containing samples were corrected using a reagent blank having the same Ca concentration as the samples and containing 1 ppm of Rh as an internal standard. The loss of Aliquat® 336 to the aqueous phase was determined using TOC measurements. The total organic carbon content (TOC) was analyzed with a HiPerTOC TOC analyser (Thermo Scientific) using the non-purgeable organic carbon (NPOC) method. The total inorganic carbon (TIC; dissolved CO 2 , HCO 3 − and CO 3 2− ) of the sample (30 mL) was converted to CO 2 by addition of 400 µL of a 10 vol% H 3 PO 4 solution. The formed CO 2 gas was removed from the solution by flushing with O 2 gas. Subsequently, 1 mL of the sample was transported to the UV reactor together with 1.5 mL of an oxidizing solution (120 g L −1 Na 2 S 2 O 8 ) under O 2 atmosphere (99.999%) with an O 2 flow rate of 250 mL min −1 . The UV reactor converted the remaining carbon to CO 2 . Next, the formed gas was dried by a Peltier cooler and separated from the liquid components by a gas-liquid separator. Afterwards, the dried CO 2 was carried to a non-dispersive infrared detector, where CO 2 was measured. A calibration curve was prepared to relate the detector signal to the CO 2 concentration and hence to the corresponding total carbon concentration of the sample. ## Extraction experiments Extraction experiments were performed using two phases: metal chloride salts in an HCl acidified water phase and undiluted ionic liquid comprising the organic phase. The IL was presaturated with Milli-Q® water and hydrochloric acid in a volume ratio of 4.5 : 1, having the same chloride concentration as the aqueous phase before extraction, at 60 °C to prevent co-extraction of Milli-Q® water and hydrochloric acid. Therefore, the ratio of the volume of the aqueous and the ionic liquid phase remains constant. Extractions were performed by intensive stirring at 500 rpm for 60 min at 60 °C with a magnetic stirring bar. Hydrochloric acid was used as the chloride source. After the extraction, separation of the phases was assisted by centrifugation for 15 min at 3500 rpm. ## Mono-element system Distribution ratio studies. The distribution ratios of indium(III) at different HCl concentrations were determined using equal volumes of ionic liquid and an acidified water phase (5 mL) containing 5 g L −1 of indium(III). Further experiments studied the distribution ratios of indium(III) as a function of feed concentration at the optimal hydrochloric acid concentration of 0.5 M and using feed solution concentrations ranging from 5 to 120 g L −1 . The viscosity of the organic phase was determined for the various feed solution concentrations. Kinetics. Reaction kinetics were studied by shaking equal volumes of ionic liquid and an acidified water phase (5 mL) containing 5 to 100 g L −1 of indium(III) for different times ranging from 1 to 180 min. Stripping. First, extractions were performed from a feed solution containing 5 g L −1 indium(III) at the optimal hydrochloric acid concentration of 0.5 M. Afterwards, the loaded ionic liquid phase was stripped with a water phase (5 mL) containing different stripping agents: Milli-Q® water, ethylenediaminetetraacetic acid disodium salt dihydrate (Na 2 EDTA•2H 2 O) and sodium hydroxide. All stripping experiments were executed using equal volumes of ionic liquid and a water phase at 60 °C, except for the stripping with sodium hydroxide which was performed at room temperature. After stripping, separation of the phases was assisted by centrifugation for 15 min at 3500 rpm. Recycling of the ionic liquid for reuse. Extractions were performed from a feed solution containing 5 g L −1 indium(III) at the optimal HCl concentration of 0.5 M. Afterwards, the loaded ionic liquid phase was stripped with an aqueous NaOH solution. The loaded Cyphos® IL 101 and Aliquat® 336 were stripped by 4.5 and 4 equivalents of NaOH, respectively. All extraction and stripping experiments were executed with equal volumes of ionic liquid and water phase. After extraction and stripping the ionic liquid was equilibrated by a 0.5 M HCl solution in a volume ratio 1 : 5. The equilibrated ionic liquid was reused in a second extraction step using a feed solution containing 5 g L −1 indium(III) at the optimal HCl concentration of 0.5 M. Loss of ionic liquid to the aqueous phase. First, Cyphos® IL 101 was washed 4 times with a 0.05 M NaCl solution in a 1 : 25 volume ratio at room temperature until the water soluble phosphor-containing impurities present in the ionic liquid were removed. The P content of the aqueous phase after washing remained constant. A 0.05 M NaCl solution was used as a washing solution instead of Milli-Q® water due to emulsion formation with Milli-Q® water. Subsequently, extractions were performed using equal volumes of ionic liquid and aqueous phase (5 mL) at 60 °C. The solubility of the ionic liquid in the aqueous phase was determined at different HCl (0.5 M, 6 M and 12 M), indium(III) (5 g L −1 , 40 g L −1 and 120 g L −1 ) and CaCl 2 (0.25 M, 3 M and 6 M) concentrations in the aqueous phase. The loss of Cyphos® IL 101 to the aqueous phase was determined by ICP-OES. Aliquat® 336 was first saturated with a 0.05 M NaCl solution in a 1 : 1 volume ratio at room temperature. Afterwards, extractions were performed using equal volumes of ionic liquid and aqueous phase (250 mL) at 60 °C. The solubility of the ionic liquid in the aqueous phase was determined at different HCl (0.5 M, 6 M and 12 M) and CaCl 2 (0.25 M, 3 M and 6 M) concentrations in the aqueous phase. The loss of Aliquat® 336 to the aqueous phase was determined by TOC measurements. Aliquat® 336 is considered to be a mixture quaternary ammonium chloride with a 2 : 1 molar ratio of octyl groups to decyl groups. A molar weight of 432 g mol −1 was used instead of 404 g mol −1 which is valid for pure trioctylmethylammonium chloride. Based on the molar ratio, the following composition of quaternary ammonium chloride was assumed (not including water and other impurities) according to random statistical distribution: trioctylmethylammonium chloride (33 wt%), dioctyldecylmethylammonium chloride (28 wt%), octyldidecylmethylammonium chloride (22 wt%) and tridecylmethylammonium chloride (17 wt%). 41 These assumptions were made to calculate the solubility of Aliquat® 336 in the aqueous phase based on TOC analyses results. ## Multi-element system Distribution ratio studies. The distribution ratio of cadmium(II), copper(II), indium(III), iron(III), manganese(II), nickel(II), tin(IV) and zinc(II) at different HCl concentrations was studied using equal volumes of ionic liquid and an acidified water phase (10 mL) containing 5 g L −1 of each metal added as chloride salts. Additionally, 1 g L −1 of lead(II) as a chloride salt and arsenic(III) as an oxide were added to the water phase. These elements are commonly found in primary and secondary sources of indium. Extraction experiments of the multi-element system were carried out only once. ## Distribution ratio and separation factor The distribution ratio D of a metal M is defined as where [M] IL is the metal concentration in the ionic liquid and [M] aq is the metal concentration in the aqueous phase after extraction. For metals which are strongly extracted to the ionic liquid phase (%E ≥ 50%) only the remaining metal concentration in the aqueous phase was measured after extraction and eqn (1) can be rewritten as where [M] 0 is the initial metal concentration in the aqueous phase before extraction. The metal concentration of the ionic liquid phase was measured for metals which are poorly extracted (%E < 50%), and eqn (1) becomes The percentage extraction (%E) is defined as the amount of metal extracted to the ionic liquid phase over the initial amount of metal present in the aqueous phase: The separation efficiency between two metals can be described by the separation factor α, which is defined as the ratio of the respective distribution ratios of two extractable solutes measured under the same conditions: where D M1 and D M2 are the distribution ratios D of metal M 1 and M 2 , respectively. By definition, the value of the separation factor is always greater than unity. Metals are removed after extraction from the organic phase by a stripping agent. The percentage stripping (%S) in the stripping phase can be defined as the amount of metal stripped from the organic phase to the total amount of metal in the organic phase before stripping: where [M] IL,0 is the metal concentration in the organic phase after extraction or before stripping. ## Mono-element system Distribution ratio studies. In the first series of experiments, the distribution ratios and the extraction percentages of indium(III) between the ionic liquid and aqueous phase were determined as a function of hydrochloric acid concentration (Fig. 1, S1, † and Table 1). Fig. 1 shows bell-shaped curves; i.e., the distribution ratios increase as a function of HCl concentration until a value from where they begin to decrease. Bellshaped curves of the distribution ratio versus the HCl content have also been observed by other authors for extraction of Cu(II), Co(II), Fe(III), Ga(III), In(III), Mn(II) and Zn(II) by quaternary ammonium salts. The average distribution ratio of indium(III) between Cyphos® IL 101 and hydrochloric acid aqueous solution, increases with increasing chloride concentration, with a maximum D In = 4966 ± 262, at a HCl concentration of approximately 5 M. The same trend but with lower values was observed for the distribution ratio for the HCl-Aliquat® 336 system, where a maximum distribution ratio of D In = 340 ± 28 was found at the same 5 M HCl concentration. In both cases, the initial indium(III) concentration in the aqueous phase was 5 g L −1 . The bell-shaped curve of the distribution ratio is tentatively attributed to a change of the indium speciation in the aqueous and/or in the ionic liquid phase in combination with HCl extraction. Narita et al. 46 already described that the indium speciation in the aqueous phase can change as a function of HCl concentration. Co-extraction of HCl by phosphonium and ammonium ionic liquids has also been previously observed by Komasawa et al., 47 Sato et al. 48 and Wellens et al. 49 The distribution ratios of indium(III) for the HCl-Aliquat® 336 system were significantly lower, minimum by a factor of 5, than those for the HCl-Cyphos® IL 101 system over the whole HCl concentration range (Fig. 1). This can be explained by the higher water uptake of Aliquat® 336 (21.27 wt%) compared to Cyphos® IL 101 (13.65 wt%). The hydrophobicity of ionic liquids with a common anion is dependent on the total number of carbon atoms in the alkyl chains attached to the corresponding central atom. Shorter alkyl chains attached to the ammonium cation core lead to a lower hydrophobicity of Aliquat® 336 compared to Cyphos® IL 101, and therefore to a higher mutual miscibility with water. Furthermore, the charge delocalization at the ammonium cations makes the central part of these cations more charged, and thus overall more polar, than the corresponding phosphonium cations. 50,51 Very high values for the distribution ratio were observed over the entire HCl concentration range for the HCl-Cyphos® IL 101 system indicating a strong affinity of indium(III) for the ionic liquid phase. The distribution ratio observed for the quaternary ammonium ionic liquid over the entire HCl concentration range are not exceptionally high. High distribution ratios are not desirable for solvent extraction since they make stripping in many cases very difficult. Following studies regarding the kinetics, maximal loading, the extraction mechanism and the stripping of the mono-element systems are executed at a low HCl concentration of 0.5 M where the distribution ratio is high enough to ensure almost complete extraction of indium(III), but not too high so that stripping difficulties are avoided. Kinetics. For industrial applications, it is more interesting to expand the indium(III) concentration range of the feed solu- tion. The distribution ratios of indium(III) as a function of feed concentration were studied at the optimal HCl concentration of 0.5 M and with feed solution concentrations ranging from 5 to 120 g L −1 for both Cyphos® IL 101 and Aliquat® 336 (Fig. 2). A stirring time of 1 h was used to ensure that equilibrium was reached. Due to the strong affinity between the ionic liquid and indium(III), the organic phase can be loaded with high amounts of indium(III), while it is still possible to obtain a high percentage extraction (Fig. S2 †). Phase inversion was observed for Cyphos® IL 101 and an initial indium(III) concentration of 100 g L −1 and higher and for Aliquat® 336 and an initial indium(III) concentration of 120 g L −1 . At high loading of the ionic liquid phase, its density becomes larger than the density of the aqueous phase so that a phase inversion occurs. The density at 25 °C of the 100 g L −1 indium(III) loaded Cyphos® IL 101 and the 120 g L −1 indium(III) loaded Aliquat® 336 was 1.0242 ± 0.0023 g mL −1 and 1.0303 ± 0.0021 g mL −1 , respectively The density of the corresponding aqueous phases was 1.0078 ± 0.0003 g mL −1 for the Cyphos® IL 101 system and 1.0303 ± 0.0021 g mL −1 for the Aliquat® 336 system at 25 °C. The avoidance of phase inversion is essential to industrial extraction processes, since the design of the contacting equipment is based on preferred direction of solute transfer, giving optimum mass transfer rates. Phase inversion, can cause considerable operating problems, especially in mixer-settlers, in which the change in properties of both the continuous phase and the drop size can lead to a delay in the settling process. Next, the influence of the extraction time was evaluated from 1 min to 60 min at the optimal HCl concentration of 0.5 M, with feed solution concentrations ranging from 5 to 100 g L −1 (Fig. 3 and S3 †). The results show that equilibrium is reached after 10 min, regardless of the indium(III) concentration. However, longer stirring times to reach the equilibrium were expected at higher indium(III) concentrations due to an increased viscosity of the ionic liquid phase at high indium(III) loadings. This increase in viscosity as a function of indium(III) concentration in the aqueous feed solution is shown in Fig. 4. The influence of the viscosity on the reaction time is thus most probably negligible due to the high affinity of indium for the ionic liquid phase. The viscosity of the ionic liquid phase increased only slightly when it was loaded with higher indium(III) concentrations (Fig. 4). This indicates that the chloroindate(III) anion formed during extraction has the same charge as the chloride anion in the pure ionic liquid, i.e. that the [InCl 4 ] − anion is formed instead of higher charged species such as [InCl 5 ] 2− or [InCl 6 ] 3− . The presence of species with a higher charge are known to lead to a sharp increase of the viscosity at higher loadings. 33,34,46,52 The same speciation was also observed in the organic phase for the extraction of indium(III) using long chain alkylamines and quaternary ammonium chlorides from chloride media. 43,44,46,53,54 The proposed extraction mechanism is therefore: Indium(III) is probably extracted as a neutral InCl 3 or the single negatively charged [InCl 4 ] − complex. The more negatively charged chloroindate(III) complexes, [InCl 5 ] 2− and [InCl 6 ] 3− , are more strongly hydrated, thus more difficult to dehydrate and to convert to extractable anionic species. Stripping. The ease of recovery of the metal from the ionic liquid phase and regeneration of the ionic liquid are as important as obtaining a high extraction efficiency. 7 Stripping of the ionic liquid phase after extraction was tested using several stripping agents. Cyphos® IL 101 and Aliquat® 336 are basic extractants (anion exchangers). The driving force for extraction is the presence of chloride anions. Stripping of metals is efficient if the distribution ratio is low. This can be achieved by decreasing the chloride concentration by addition of water. However, the stripping with water did not result in an efficient stripping; 6.1 ± 0.7% indium was stripped from the loaded Cyphos® IL 101 and 29 ± 3% from the loaded Aliquat® 336 phase. Furthermore, the stripping of the loaded Cyphos® IL 101 ionic liquid gave also rise to a difficult disengagement (formation of an emulsion). A small amount of Na 2 SO 4 (±15 mg) was added as a salting-out agent to 5 mL of the aqueous phase. Although the stripping of the loaded ionic liquids with water resulted in a low percentage stripping, changing the temperature or the volume ratio could increase the stripping efficiency. By increasing the volume ratio of the aqueous to the organic phase from 1 : 1 to 3.5 : 1, stripping percentages of 19% and 78% were obtained for Cyphos® IL 101 and Aliquat® 336, respectively. In contrast to stripping with a 1 : 1 volume ratio, a white In(OH) 3 precipitate was found in the aqueous phase after stripping. Ethylenediaminetetraacetic acid (EDTA) forms a very stable, highly water-soluble chelate complex with indium(III) (log β = 24.90). 55,56 Therefore, it was tested to strip indium(III) from the ionic liquid with an aqueous solution of Na 2 EDTA. Although reasonable percentages of stripping were acquired, 52 ± 1% for loaded Cyphos® IL 101 and 67 ± 1% for the loaded Aliquat® 336, the use of EDTA in the industry is not favored due to its low biodegradability and difficulties associated with recovery of EDTA in a continuous solvent extraction process due to its high level of complexing capacity with respect to heavy metals. If wanted, higher percentages stripping can probably be obtained by adjusting the pH with an alkaline solution, neutralizing the HCl present in the ionic liquid phase and removing the remaining protons from Na 2 EDTA. Care must be taken to avoid ionic liquid decomposition in alkaline environment and anion exchange with EDTA. 34 Aside from the complexing agent EDTA, stripping was also tested with sodium hydroxide as a precipitation agent. Precipi-tation stripping from a metal-loaded ionic liquid phase has already been reported for the rare earths using oxalic acid as the stripping agent. Indium is directly stripped from the organic phase with sodium hydroxide forming an insoluble hydroxide. The precipitation stripping reaction can be represented as: As the pH increases, the equilibrium of the hydrolysis reaction is shifted to the right and more indium is precipitated from the aqueous phase. According to Pourbaix, the solubility of In(OH) 3 is minimal at pH 6.79. 68 The loaded ionic liquid phase was stripped with an aqueous phase containing different equivalents of sodium hydroxide (Fig. 5). Using sodium hydroxide as a stripping agent, it was possible to obtain a percentage stripping close to 100%. The 0.5 M HCl-Aliquat® 336 system was easier to strip compared to the 0.5 M HCl-Cyphos® IL 101 system due to the lower indium distribution ratio obtained during extraction. 4 equivalents of NaOH were needed to achieve a percentage stripping over 99% for the 0.5 M HCl-Aliquat® 336 system. The equilibrium pH under these conditions was 4.35. However, at this point not all of the indium was precipitated yet as In(OH) 3 . Part of it is dissolved in the aqueous phase. 4.75 equivalents of NaOH were needed to fully precipitate all indium. A percentage stripping over 99% was reached for the 0.5 M HCl-Cyphos® IL 101 system if 4.5 equivalents of NaOH were used. Unlike the previous stripping experiments with water and Na 2 EDTA, the stripping was performed at room temperature instead of 60 °C, because quaternary phosphonium and ammonium ionic liquids tend to decompose in alkaline conditions, especially at elevated temperatures. Thus, in this context, quaternary phos- phonium ionic liquids can decompose to yield a tertiary phosphine oxide and alkane under alkaline conditions. Quaternary ammonium ionic liquids can easily undergo Hofmann elimination or β-elimination to yield a tertiary amine and an alkene in the presence of a base (Fig. 6). 61,62,64 Trihexyl(tetradecyl)phosphonium chloride stability was investigated by comparing the integration of the degradation products 31 P resonance signal, i.e. phosphine oxide, formed after mixing for 1 h with an aqueous NaOH solution to the integration of the trihexyl(tetradecyl)phosphonium cation 31 P resonance signal at 32.9 ppm. The resonance signal of phosphine oxide was situated at 45.8 ppm. 31 P NMR spectra were measured for the ionic liquid before and after stripping since commercial Cyphos® IL 101 already contains small amounts of phosphine oxide impurities. Peak integration ratios were then compared and it was concluded that no decomposition occurs because no change of the ratio of the integration of the phosphine peak over the integration of the phosphonium cation peak took place when stripping with NaOH at room temperature. Recycling of the ionic liquid for reuse. An important aspect of green chemistry is the recyclability and reusability of the ionic liquid. The recyclability was shown by reusing the ionic liquid in a second extraction step after stripping. The results of each individual step are given in Table 2. The results indicate that the ionic liquid can be recycled for reuse in liquid/liquid extraction of indium(III) without any loss of activity. Loss of the ionic liquid to the aqueous phase. Ionic liquids are often considered as green solvents primarily due to their negligible vapor pressure and low flammability. However, the avoidance of loss to the aqueous phase and the recyclability of the ionic liquid are equally as important in establishing a sustainable process. The loss of Cyphos® IL 101 and Aliquat® 336 to the aqueous phase was determined. First, the phosphor con-taining water soluble impurities present in Cyphos® IL 101 were removed by four consecutive washing steps with a 0.05 M NaCl solution using a 1 : 25 volume ratio at room temperature (Fig. S4 †). If not removed these impurities will insinuate a higher solubility of Cyphos® IL 101 in the aqueous phase. Subsequently, the ionic liquid was brought into contact with several aqueous solutions at 60 °C: HCl (0.5 M, 6 M and 12 M), indium(III) (5 g L −1 , 40 g L −1 and 120 g L −1 ) and CaCl 2 (0.25 M, 3 M and 6 M). The solubility of Cyphos® IL 101 in the different aqueous phases is presented in Table 3. The purity of Aliquat® 336 (88.2-90.6%) is low in comparison with Cyphos® IL 101 (>97%). Therefore, no prior washing steps were executed to remove possible water soluble organic impurities. For result comparison reasons, Aliquat® 336 was saturated with a 0.05 M NaCl solution in a 1 : 1 volume ratio at room temperature. Afterwards, the solubility of Aliquat® 336 was determined in several HCl (0.5 M, 6 M and 12 M) and CaCl 2 (0.25 M, 3 M and 6 M) solutions after contacting the ionic liquid with the aqueous solutions at 60 °C. The solubility results are displayed in Table 3. The solubility of Aliquat® 336 in aqueous phase is higher than the solubility of Cyphos® IL 101 due to it higher mutual miscibility with water. As the salt concentration in the aqueous phase increases, the solubility of the ionic liquid in the aqueous phase decreases due to the salting-out effect. In general, the solubility of the ionic liquids in the aqueous phase is very limited thereby having a low associated economical and environmental impact. ## Multi-element system Distribution ratio studies. A useful purification process needs to have a reasonable selectivity for the targeted metal. All the experiments described above have been carried out on mono-element solutions but industrial solutions are generally complex mixtures of several elements. In a second series of experiments, the separation of indium(III) from arsenic(III), cadmium(II), copper(II), iron(III), lead(II), manganese(II), nickel(II), tin(IV) and zinc(II) was investigated. More elements are added to the feed solution compared to previous experiments, to a more industrial relevant elemental composition: 5 g L −1 of each of the other elements with the exception of arsenic(III) and lead(II) present in a concentration of 1 g L −1 . A concentration of 1 g L −1 for arsenic(III) and lead(II) was chosen due to the limited solubility of arsenic(III) oxide and lead(II) chloride in the aqueous phase. 69,70 Fig. 7 and Fig. 8 show the distribution ratios between the ionic liquid phase and aqueous phase of the elements as a function of HCl concentration for the HCl-Cyphos® IL 101 and HCl-Aliquat® 336 systems. Percentage extraction as a function of HCl concentration for both systems is given in Fig. S5 and S6. † Many elements display a bell-shaped distribution ratio curve. Indium(III) has a maximum distribution ratio at 5 M in both systems in agree-ment with the previous experiments. For copper(II) and manganese(II), the maximum distribution ratio was found at a HCl concentration of 4 M and 6 M, respectively for the HCl-Cyphos® IL 101 system and at 5 M and 7 M, respectively for the HCl-Aliquat® 336 system. For nickel(II) and arsenic(III), the maximum distribution ratio was found at higher HCl concentration of 8 M and 10 M, respectively, independent of the extraction system. Moreover, the distribution ratios of nickel(II) were significantly lower than those of the other elements over the whole HCl concentration range, indicating a low affinity of nickel(II) for the ionic liquid phase. For lead(II) the behavior was distinctly different from that of the other elements. Here the distribution ratios were high at low HCl concentrations and decreased with increasing HCl concentration. It is very likely that only the right side of the bell-shaped curve is represented. The maximum value for the distribution ratios of cadmium(II), iron(III), tin(IV) and zinc(II) could not be determined with ICP-OES as the analysis method due to the lower detection limit of 1 mg L −1 . However, from the partial shape of the distribution ratio curves of cadmium(II), iron(III) and zinc(II), it can be concluded that also a bell-shaped path is followed. The maximum distribution ratio and corresponding HCl concentration for each of the elements are given in Table 4. Advantage can be taken of the low affinity of arsenic(III), manganese(II) and nickel(II) for the ionic liquid phase at lower HCl concentrations to separate these ions from cadmium(II), iron(III), lead(II), tin(IV), zinc(II) and most importantly from indium(III). Due to the lower distribution ratios obtained for most of the elements, that allows an easier stripping of indium(III) from the ionic liquid phase after extraction, Aliquat® 336 is the preferred undiluted extractant. In the following paragraphs, information is given on the capabilities of the 0.5 M HCl-Aliquat® 336 system for purification of indium(III), and possible process steps are suggested. The low distribution ratios of arsenic(III), manganese(II) and nickel(II) in combination with high separation factors between these elements and indium(III), imply an easy separation (Table 5). The contaminating ions will remain largely in the aqueous phase during extraction. A high separation factor was also obtained for the In(III)/Cu(II) couple, but due to the slightly elevated distribution ratio of copper(II) in comparison with arsenic(III), manganese(II) and nickel(II), extraction of copper(II) cannot be avoided completely (Table 5). However, reasonably low distribution ratio enables the scrubbing of copper(II) from the ionic liquid phase together with traces of arsenic(III), manganese(II) and nickel(II) extracted. A solution with a low HCl concentration has to be used for scrubbing to avoid loss of indium(III) to the aqueous phase. After extraction and scrubbing, cadmium(II), iron(III), lead(II), tin(IV), zinc(II) will still be present in the ionic liquid phase together with indium(III). The high separation factors for the couples In(III)/ Cd(II), In(III)/Sn(IV) and In(III)/Zn(II) suggest an easy separation (Table 5). Although they all possess a high affinity for the ionic liquid phase (D In = 30.7, D Cd , D Sn , D Zn > 5.00 × 10 3 ), indium(III) will be easier to strip due to its distribution ratio being 160 times smaller. It was demonstrated previously that indium can be removed from the loaded Aliquat® 336 phase by stripping with water or NaOH. Both methods can be satisfactory, but stripping with water will require larger volume ratios of the aqueous to the organic phase. Also selectivity will play a vital role in choosing a stripping method. Selective stripping of indium(III) from iron(III) and lead(II) will be difficult due to the small separation factors, α Fe(III)/In(III) and α Pb(II)/In(III) (Table 5). Hydrolysis curves have to be constructed for the stripping of the ionic liquid phase with NaOH to get a better understanding about the hydrolysis behavior of the various metals, elaborating a more selective stripping. After stripping, the ionic liquid phase has to be scrubbed so that it can be reused. Scrubbing of the ionic liquid phase with water will have little effect due to the high distribution ratios of cadmium(II), tin(IV) and zinc(II). Scrubbing with a NaOH solution will in most cases be the best option. ## Conclusions It is shown that the commercial ionic liquids Cyphos® IL 101 and Aliquat® 336 are very efficient for extraction of indium(III) from chloride feed solutions. The ionic liquids are used in an undiluted form so that volatile molecular organic solvents can be avoided. The extraction process is selective for indium(III), over many other metal ions that are commonly found as impurities in process solutions of indium refineries. Due to the very high distribution ratios, co-extracted impurities can easily be scrubbed from the ionic liquid phase without affecting the extracted indium(III) ions. Indium could be recovered in the form of In(OH) 3 by precipitation stripping with a NaOH solution. This stripping step also regenerates the ionic liquid. This work indicates that precipitation stripping with NaOH is a very convenient methods for recovering metal ions that show high affinities for the organic phase, such as indium(III) for undiluted chloride ionic liquids. The method is environmentally friendly because NaCl is the sole waste product and the ionic liquids can be reused. system and TOC analyses were executed in the analytical laboratory of Umicore Group Research & Development. a According to IUPAC, by convention the ratio of the respective distribution ratios has to be chosen so that α > 1. 71

chemsum

{"title": "Purification of indium by solvent extraction with undiluted ionic liquids", "journal": "Royal Society of Chemistry (RSC)"}

coal-based_3d_hierarchical_porous_carbon_aerogels_for_high_performance_and_super-long_life_supercapa

## Abstract: Coal-based 3D hierarchical porous carbon aerogels (3D HPCAs) has been successfully fabricated from a freeze-drying method and with subsequent of calcination process, using coal oxide as carbon precursors, and PVA as both cross-linking agent and sacrifice template. The 3D HPCAs, using as electrode materials for supercapacitors, display outstanding electrochemical performance. The optimal sample (HPCAs-0.4-800) presents a high specific capacitance of 260 F g −1 at 1 A g −1 , and exhibits considerable rate capability with the retention of 81% at 10 A g −1 . Notably, HPCAs-0.4-800 shows an excellent cycling stability with 105% of the capacitance retention after 50000 cycles at 10 A g −1 , attributing to its unique hierarchical porosity, high surface area up to 1303 m 2 g −1 , and improved conductivity. This work offers a promising route to synthesize coal-based porous carbon aerogels electrode materials for supercapacitors. Supercapacitors (SCs), also called ultracapacitors or electrochemical capacitors, have caused a large amount of interest owing to excellent electrochemical stability, fast charge/discharge, high power density and environmental friendly [1][2][3][4][5] . Supercapacitors store electrical charge on high-surface-area conductive materials, so its performance mainly relies on the electrode materials. Outstanding electrode materials should possess ion approachable high surface areas for high specific capacitance and fasted electron transfer for excellent rate capacity 6,7 . So it is very crucial for supercapacitors with high performance to prepare electrode materials with proper architecture structure, suitable pore size distribution and high specific surface area (SSA) 8 . Among the numerous electrode materials of supercapacitors, carbon materials have attracted more attention because of their unique physical and chemical properties 9,10 . Carbon aerogels (CAs), as one of carbon materials, show outstanding characteristics, such as low density, developed porosity, and multi-branched network structure [11][12][13] . These structural features can afford the quick transfer channel for ion migration and more active sites, which can lead to the excellent electric double layer performance in supercapacitors. To improve the specific surface area and porosity, most of the CAs are prepared by using pore-forming agents, such as strong bases [14][15][16] , hard templates [17][18][19] , soluble salts 20,21 , soft templates and so on [22][23][24] . Among of them, the soft templates can be directly decomposed during the carbonization process instead of etching procedure using harmful and toxic or corrosive chemicals. Therefore, it has been attracting extensive attention to prepare CAs using soft template for the application of supercapacitors.Currently, the researches of CAs are mainly focused on precursors, such as resorcinol-formaldehyde 25 , polymers 26 , nanotubes 27,28 , graphene 29,30 , and natural precursors such as cellulose and glucose 31,32 . In our previous works, we have fabricated some functional materials on coal of traditional fossil, such as porous spheres 33 , fibers 34 , bamboo-like carbon nanotubes (CNTs) 35 , graphene quantum dots (GQDs) 36 and hierarchical porous carbon 37 . All of them demonstrate that the coal can be used to fabricate functional carbon materials. However, so far, coal-based porous carbon aerogels have been few reported and the preparation processes were very complicated and the yield was low in a few studies. Therefore, it is still a great challenge to design simple and productive approaches for the controllable synthesis coal-based porous carbon aerogels. In our work, we developed an efficient method to construct coal-based 3D HPCAs by carbonization of freezing-dried PVA/coal-based hydrogels, in which coal oxide serves as the carbon source and PVA serves as the sacrificial template and cross-linking agent, respectively. The amount of mesoporous and micropores of the 3D HPCAs can be controlled by tuning the mass ratio of coal oxide and PVA. The performance of the obtained 3D HPCAs are evaluated as the electrode materials of supercapacitors. The optimal sample displays an excellent electrochemical performance. It exhibits a specific capacitance up to 260 F g −1 in the three-electrode system at 1 A g −1 , and a high rate performance of 187 F g −1 at 20 A g −1 , as well as a remarkable cycling stability (105% of capacitance retention after 50000 cycles). More importantly, the specific capacitance measured was 201.1 F g −1 at the current density 1 A g −1 in an assembled symmetrical cell system, and good cycling stability with 108% over 10000 cycles at 4 A g −1 . The excellent electrochemical performance may be attributed to the characteristic of 3D cross-linked structure with SSA up to 1303 m 2 g −1 , hierarchical porous structure and appropriate ratio of micropore volume to total volume of 65.6%. The materials with hierarchical porous structure can be used as potential electrode materials for energy conversion and storage, and this work provides a green way for high-value utilization of coal in energy storage. ## Results and discussion The synthesis procedure of the 3D HPCAs is illustrated in Fig. 1. Firstly, PVA/coal-based hydrogels were prepared by using PVA as a crosslinking agent of coal oxide fragment, and then the 3D network porous structures were formed through a freeze-drying and with subsequent of calcination. To understand the role of PVA and coal oxide in 3D HPCAs, thermogravimetric (TG) analyses of PVA and coal oxide were studied (Fig. S1). The mass loss of PVA is 96.6% of initial weight from 260 °C to 490 °C, and 98.8% when heated to 800 °C, while that of coal oxide is about 62.6% when heated up to 800 °C under an argon atmosphere. The results demonstrate that coal oxide is the primary carbon source in the 3D HPCAs, while PVA is the cross-linking agent for formation of hydrogels and the sacrifice template to fabricate 3D network porous structures. This result is also directly proved by the optical and SEM images of the pure oxidized coal-800, prepared by the same proceduce only without addition of PVA. As shown in Fig. S2a,b, the pure oxidized coal-800 shows powder shape but non-aerogel on the macro level, and block shape on the micro level, with dense surface and no obvious macropores and mesopores. The morphology of fabricated 3D HPCAs were directly observed by scanning electron microscopy (SEM) and transmission electron microscopy (TEM) (Fig. 2). The SEM images show that the pore structure of 3D HPCAs changed obviously as the increase of PVA content. Consistent well with SEM images, the TEM images of HPCAs-0.4-800 display a richer porous structure and more even pore size distribution obviously. The results demonstrate that an appropriate amount of PVA is crucial for the formation of the pore structure of 3D HPCAs. Coal oxide will stack seriously when crosslinking dose of PVA is excessive, on the contrary, too little sacrificial dose of PVA is not conducive for the formation of holes during the carbonization process. In addition, the influence of calcination temperature on the pore structure of 3D HPCAs was also studied. As shown in Fig. S3, the pore size of HPCAs-0.4-700 and HPCAs-0.4-900 increase due to the change of carbonized temperatures. The broken mesopores can be observed in the HPCAs-0.4-900 due to the skeleton collapse at high carbonization temperature. The effect of different components and calcination temperatures on the structure of the HPCAs was investigated by X-ray powder diffraction (XRD) (Fig. 3). XRD patterns of 3D HPCAs all display two weak peaks located at 24° and 43°, associated with diffraction of the (002) and (101) planes carbon. Compare with the peaks of coal oxide (Fig. S4), the peaks around 23° of the HPCAs shift up to a relatively high angle, indicating that the degree of graphitization is reduced during the process of carbonization . In addition, the HPCAs-0.4-800 displays the greatest peak intensity at 23° among of all HPCAs, indicating that the HPCAs-0.4-800 has the highest degree of graphitization associated with conductivity. Raman spectra are shown in Fig. 3b and S3, two obvious peaks around 1350 and 1590 cm −1 are attributed to D and G bands. The D peak is attributed to the sp 3 defects of disordered ones in the hexagonal graphitic layers and sp 2 carbon with O-containing groups and H-sites, as well as domain boundary. The G peak reflects the vibration of sp 2 -bonded carbon atoms . The I D /I G values of HPCAs-0.4-700, HPCAs-0.2-800, HPCAs-0.4-800, HPCAs-0.8-800, HPCAs-0.4-900 and coal oxide are 0.99, 0.97, 0.87, 0.90, 0.97 and 0.81 respectively. Compared HPCAs with coal oxide (Fig. S4), the HPCAs show higher values of I D /I G , because the surface carbon atoms of HPCAs were carried off during the activation operation, thus, leaving the free bond at the surface and forming a disordered carbon structure 1,45,46 . In addition, the HPCAs-0.4-800 has www.nature.com/scientificreports www.nature.com/scientificreports/ the lowest I D /I G value in the samples of different calcination temperatures, which likely resulted from the reduction reaction at high temperature, on the other hand, the defects of materials increase as the structure collapse at a too high carbonization temperature (900 °C). And among the samples of HPCAs-0.2-800, HPCAs-0.4-800, and HPCAs-0.8-800, HPCAs-0.4-800 also exhibits the highest graphitization degree, which acts as a key role in improving the conductivity. This is because excessive PVA decomposition at high temperatures causes the pore structure to collapse, thus forming an amorphous carbon. The above results prove that appropriate carbonization temperature and PVA content are crucial to the pore structure and the degree of graphitization, which finally reflected in its electrochemical performance. The XPS spectra were carried out to evaluate the surface atomic composition of HPCAs-0.4-800. Fig. S5 shows the XPS spectra of samples. The survey spectrum confirmed the existence of C, O, and N elements in the sample of HPCAs-0.4-800 (Fig. S5a). The results are consistent with the FT-IR spectra data analysis reported in our previous work 38 . The N and O in samples are mainly come from the raw coal and nitric acid and sulfuric acid used in the oxidation process. The high-resolution spectrum of C 1s (Fig. S5b) can be divided into three peaks at 288.6, 286.2 and 284.5 eV, which are assigned to C=O, C-O and C-C, respectively 47,48 . The O 1s spectrum (Fig. S5c) consists of four peaks located at 530.9, 531.5, 532.4 and 533.5 eV, which corresponding to the carbonyl groups (C=O), bridge-bonded oxygen (C-O-C), ester groups (O-C=O) and chemisorbed oxygen or water (COOH carboxylic groups or water), respectively. The high resolution spectrum of N 1s (Fig. S5d) shows three peaks with binding energy values of 398.1, 400.5 and 404.3 eV for the pyridinic-N, pyrrolic-N and oxidized-N, respectively 49,50 . The elemental contents (atomic%) of the HPCAs-0.4-800 obtained from XPS data are presented in Table S1. HPCAs-0.4-800 has high contents of oxygen and nitrogen, with a ratio of 7.27% and 1.6%, respectively, which can ameliorate the wettability between the electrode material surface and electrolytes, further facilitate the immersion of electrolyte into the interior of the electrode materials, and ultimately reflect in high capacitance performance. Besides, the pyridinic-N and pyrrolic-N can introduce the faradaic pseudocapacitance in aqueous electrolytes and thus also enhance electrochemistry capacitance 51 . The electric double layer capacitor (EDLC) is a surface regulated phenomenon, so a larger surface area is crucial for acquiring a high capacitance 52 . The N 2 adsorption-desorption isotherms and the pore distribution of samples with the different mass ratios of coal oxide/PVA are presented in the Fig. 4a-f. The isotherms of the 3D HPCAs show classical type-IV curves with an H4 type hysteresis loop in the relative pressure region between 0.45 and 1.0, suggesting that the existence of a silt-shaped pore structure. Compared with the pure oxidized coal-800 (Fig. S2c), the volume of adsorbed N 2 increase steeply at relatively low pressure, which stands for the existence of a large number of micropores, and the remarkable hysteresis loop between N 2 adsorption and desorption branch manifests the existence of mesopores. It is well known that micropores and mesopores are in favor of the improvement of charge storage and ion transport, respectively 53 . DFT pore-size-distribution curves show that the HPCAs have broad micropores size distribution (0.65-2 nm) and a narrow mesopores size distribution (2-10 nm). The sizes of micropores are close to the size of hydrolyzed K + ions (0.331 nm), which are beneficial for their capacitive performance 54 . The data of S BET and pore distribution of 3D HPCAs are shown in Table 1. S BET of HPCAs-0.2-800, HPCAs-0.4-800, HPCAs-0.8-800 are 1018, 1303, and 847 m 2 g −1 , respectively. Distinctly, the contribution of micropores to S BET of HPCAs-0.4-800 (66%) was greatest among all samples (44% to HPCAs-0.2-800 and 55% to HPCAs-0.8-800). In general, the large specific surface area of micropores can result in a high capacity 6,55 . The N 2 adsorption-desorption isotherms and the pore distribution of samples with different calcination temperatures (HPCAs-0.4-700 and HPCAs-0.4-900) are shown in Fig. S6. They show similar types of type-IV curves with HPCAs-0.4-800. The effects of the calcination temperature on distribution of hierarchical porous have also been studied, as shown in Table S2. The S BET of HPCAs was 971, 1303, and 900 m 2 g −1 at 700, 800, and 900 °C, respectively. The contribution of micropores to S BET of HPCAs-0.4-800 (66%) was also larger than that of HPCAs-0.4-700 (54%) and HPCAs-0.4-900 (46%). This is because the high temperature promotes the decomposition of the large amounts of coal oxide and PVA, leading to achieve abundant pores, but too high a temperature (900 °C) will cause the stacking of carbon layers and/or generation of isolated pores, reflecting a smaller S BET and the amount of micropores. Such result is related to their electrochemical performances. The electrochemical performances of 3D HPCAs were evaluated through CV curves at the scan rate of 50 mV s −1 and GCD curves at the current density of 1 A g −1 (Fig. 5a,b). All 3D HPCAs electrode materials display similar rectangles, suggesting that the energy storage type of the 3D HPCAs are the electric double layer capacitor (EDLC). The values of specific capacitance are 232, 260, 218, 210 and 184 F g −1 at the current density of 1 A g −1 for HPCAs-0.2-800, HPCAs-0.4-800, HPCAs-0.8-800, HPCAs-0.4-700 and HPCAs-0.4-900, respectively, which is all are superior to that of pure oxidized coal-800 (35.7 F g −1 Fig. S2e). The 3D HPCAs-0.4-800 exhibits the highest specific capacitance among all the electrode materials due to advantages of the large accessible surface area, more available mesoporous channels and advisable proportion of micropore volume to total volume. As shown in the Fig. 5c,d, specific capacitances of HPCAs-0.4-800 obtained from the discharge curve are 267, 260, 242, 224 and 210 F g −1 at the current density of 0.5, 1, 2, 5 and 10 A g −1 . On account of the insufficient surface contact and hindrance of ions diffusing into the internal pores, the capacitance decreases as the current density increased 56 . The calculated specific capacitances of the HPCAs at different current densities are presented in Fig. 5e. The 3D HPCAs-0.4-800 electrode material displays the specific capacitance of 260 F g −1 at 1 A g −1 , which is higher than HPCAs-0.2-800, HPCAs-0.8-800, and the commercial activated carbon (Kuraray YF-50, 87 F g −1 ) (Fig. S7). Such result is probably ascribed to that sufficient micropores and mesopores channel to be used. The specific capacitance of HPCAs-0.4-800 is 210 F g −1 at 10 A g −1 , which is about 80.7% of the capacitance retention of 260 F g −1 at 1 A g −1 . However, when the current density increases from 1 A g −1 to 10 A g −1 , HPCAs-0.2-800 and HPCAs-0.8-800 have the only 74.8% and 78.9% capacitance retention, respectively. HPCAs-0.4-800 manifests a considerably better rate performance than HPCAs-0.2-800 and HPCAs-0.8-800 samples, because of HPCAs-0.4-800 have uniformed and well-interconnected hierarchical porous structure. On the other hands, compared to the samples of HPCAs-0.4-700 and HPCAs-0.4-900, HPCAs-0.4-800 has the highest SSA and the suitable ratio of micropore volume to total volume of 65.6%, which is good for charge storage. Therefore, it has the highest specific capacitance. Electrochemical impedance spectroscopy (EIS) were measured to compare the electrochemical kinetics of the samples. Nyquist plots of HPCAs electrode materials consist of a vertical line and a semicircle at the low Table 1. BET Specific surface area and porous structure of HPCAs-0.2-800, HPCAs-0.4-800 and 800. a BET surface area. b The total pore volume at P/P o = 0.99, c The mesopore volume calculated using the BJH method based on the Kelvin equation. d Micropore surface area calculated using the V-t plot method. e Average pore size (4V t /S BET ). (2020) 10:7022 | https://doi.org/10.1038/s41598-020-64020-5 www.nature.com/scientificreports www.nature.com/scientificreports/ frequency and the high frequency area, respectively (Fig. 5f, and S8c). The intercept at the real axis of high frequency of all electrodes are nearly the same, indicating their similar ohmic resistance (R s ) of 0.5 Ω. A smaller semicircle at the high frequency, reflects a relatively lower charger transfer resistance (R ct ) 57 . The R ct values of HPCAs are 3.1, 2.5, 1.8, 1.5, and 1.1 Ω for HPCAs-0.4-900, HPCAs-0.4-700, HPCAs-0.2-800, HPCAs-0.4-800 and HPCAs-0.8-800, respectively. The HPCAs-0.4-800 has the smallest R ct among three samples of different carbonization temperature. The result is due to that HPCAs-0.4-800 has the highest degree of graphitization and comparatively abundant pore structure. Besides, the R ct decreased with the increase of the PVA content in samples, which is likely attributed to more available mesoporous channels and more N content coming from PVA. On the other hand, all samples show nearly perpendicular to the imaginary axis in the low frequency region, which indicated that the electrolyte ions had the best diffusion ability in electrode structure. And the straight line demonstrates the ideal EDLC behavior of electrode materials 58,59 . From the above results, the HPCAs-0.4-800 exhibits the excellent electrochemical property. The specific capacitance performance of HPCAs-0.4-800 is superior to some previously reported porous carbon materials (Table S3). As mentioned above, the excellent capacitive performance of HPCAs-0.4-800 can be ascribed to the following aspects: (i) the carbon yields of coal oxide and PVA are different, which leads to the formation of The stable cycling life is an important factor for the practice application of supercapacitor electrode materials. Fig. 6 shows that the specific capacitance of the HPCAs-0.4-800 still reaches 230 F g −1 at the high density of 10 A g −1 , and superior capacitance is maintained up to 105% after 50000 cycles. What impressed us most is that the capacitance presents an increasing trend during the repeating process of cycling at the high current density of 10 A g −1 . On the basis of the pore distribution, this phenomenon has to do with the porous structure of the electrode materials. To be specific, in the beginning stages, only large pores and mesopores are infiltrated by electrolyte, the micropore structure is not fully utilized due to the thin film on the electrode of supercapacitors 60 . However, the K + hydrated ions can gradually penetrate into the micropores and participate in the establishment of electric double layers 61 . The electrode material of 3D HPCAs-0.4-800 tends to a stable capacitance value due to the full utilization of pores as the cycles increasing. Therefore, the sample has a considerable reversibility and satisfactory cycle stability during the repeated charge-discharge process. Simultaneously, the electrochemical performance of the HPCAs-0.4-800 in symmetric cell system was also investigated. Fig. 7a shows that the HPCAs-0.4-800 has rectangular CV curves at different scan rates, indicating distinct capacitive behavior and fine reversibility. Fig. 7b shows the galvanostatic charge-discharge (GCD) curves at different current densities from 1 to 20 A g −1 in the potential range from 0 to 1 V. The HPCAs-0.4-800 shows an excellent specific capacitance of 201.1 F g −1 at 1 A g −1 and 160.0 F g −1 at 20 A g −1 , respectively. And the electrode has a good rate capability with about 80% capactive retention at 20 A g −1 . As shown in Fig. 7c, the HPCAs-0.4-800-based device has a pretty high energy density (7.2 Wh kg −1 at 500 W kg −1 ). Fig. 7d manifests the cycling stability of the HPCAs-0.4-800//HPCAs-0.4-800 cell. The specific capacitance retention up to 108% after 10,000 cycles due to the fully wetting of sufficient pores during the long time charged/discharged process, exhibiting its outstanding stable cycling. To sum up, the excellent electrochemical performances of HPCAs-0.4-800, such as high specific capacitance, and good cycling stability can be attributed to the 3D hierarchical porous and the appropriate microspores and mesopores size distribution as well as improved conductivity. ## Conclusion In summary, we design a low cost and facile strategy to obtain coal-based 3D HPCAs by carbonization of the freeze dried PVA/coal-based hydrogels. The structure and electrical performances of 3D HPCAs are adjusted and optimized by changing the content of PVA and carbonization temperature. Among of all samples, the HPCAs-0.4-800, as an electrode material of supercapacitors, exhibits excellent specific capacitances of 260 F g −1 and 201.1 F g −1 at 1 A g −1 in 6.0 M KOH electrolyte for the three-and two-electrode systems, respectively. It also displays an excellent cycling stability of 105% capacitance retention after 50000 cycles. This work provides a promising route to construct coal-based 3D HPCAs as highly efficient electrodes materials for supercapacitors. ## Experiment Section Materials. Coal was obtained from Heishan, Xinjiang, China. The similar analysis of coal has been reported by our group 38 . Polyvinyl alcohol (PVA) (molecular weight is 44.05 MW) purchased from Sigma-Aldrich. H 2 SO 4 (98%), HNO 3 (63%), KOH were analytical grade. The preparation of 3D HPCA. Coal oxide was firstly obtained by treating raw coal using a mixed acid (VHNO 3 /VH 2 SO 4 = 1:3) as previously reported by our group 38 . Then PVA/coal-based hydrogels were synthesized as following processes. Firstly, coal oxide (0.8 g), PVA (0.4 g) were dissolved in 10.0 mL deionized water and the pH of the solution was neutralized with the KOH, and then stirred at 80 °C continuously for 12 h. Secondly, the product was frozen in refrigerator (−70 °C) for 24 h and freeze-dried in vacuum for 24 h to obtain the xerogels. Finally, the as prepared xerogel was calcined at 800 °C for 2 h with a heating rate of 5 °C min −1 under flowing N 2 atmosphere for formation of 3D coal-based HPCAs, the sample was called HPCAs-0.4-800. Similarly, HPCAs-0.2-800 and HPCAs-0.8-800 were prepared by the nearly same methods with the only different mass ratios of coal Structural characterization. SEM and TEM images were recorded on field emission scanning electron microscopy (FESEM Hitachi SU-4800) and transmission electron microscopy (TEM, JEM-2100F), respectively. Thermogravimetric analysis (TGA) was tested by using a NETZSCH STA449F3-QMS403C instrument under N 2. XRD measurements were carried out on an X-ray diffractometer (XRD, Bruker D8, using filtered Cu Kα radiation). X-ray photoelectron spectroscopy (XPS) spectra and the Raman spectrum were recorded with a Thermo ESCALAB 250 instrument (Al Ka X-ray source) and a Bruker Senterra spectrometer (532 nm) Raman spectrometer, respectively. SSA and pore size distribution of coal-based 3D HPCAs were determined on Autosorb-IQ, Quantachrome by BET method. ## Electrochemical characterization. The electrochemical experiments of coal-based 3D HPCAs were tested in the three-electrode system in 6.0 M KOH electrolyte at room temperature, in which the 3D HPCAs use as working electrode, Ag/AgCl and Pt foil (1 cm 2 ) as the reference electrode and counter electrode, respectively. The working electrodes were prepared by mixing 3D HPCAs, carbon black and [Poly (trafluoroethylene)] at a mass ratio of 8:1:1 in absolute ethyl alcohol (the active mass is about 2.0 mg), and then the mixture was pasted on the nickel form, and dried at 60 °C for 12 h. The electrochemical performances of these electrodes were carried out using the CHI 660D electrochemical workstation (Chenhua, China). The current density of 10 A g −1 was applied in cyclic GCD measurements for over 50,000 cycles (the potential is −1.0 -−0.1 V). The specific capacitance (C S ) of the 3D HPCAs electrode materials were calculated via Eq. (1). ## S Where I, ∆t, ∆V, m is current, discharging time, potential window and the mass of the active material, respectively. In two-electrode system, CV and GCD curves were measured using the product in 6.0 M KOH as the electrode (the total mass of the active on two working electrodes is 4.0 mg). The specific capacitance for the single electrode (C sp ) was obtained via Eq. (2). Where E, P are the energy density (Wh kg −1 ), power density (W kg −1 ), respectively.

chemsum

{"title": "Coal-based 3D hierarchical porous carbon aerogels for high performance and super-long life supercapacitors", "journal": "Scientific Reports - Nature"}

facile_reduction_of_graphene_oxide_suspensions_and_films_using_glass_wafers

## Abstract: This paper reports a facile and green method for conversion of graphene oxide (GO) into graphene by low-temperature heating (80 °C) in the presence of a glass wafer. Compared to conventional GO chemical reduction methods, the presented approach is easy-scalable, operationally simple, and based on the use of a non-toxic recyclable deoxygenation agent. The efficiency of the proposed method is further expanded by the fact that it can be applied for reducing both GO suspensions and large-scale thin films formed on various substrates prior to the reduction process. The quality of the obtained reduced graphene oxide (rGO) strongly depends on the type of the used glass wafer, and, particularly, magnesium silicate glass can provide rGO with the C/O ratio of 7.4 and conductivity of up to 33000 S*cm −1 . Based on the data obtained, we have suggested a mechanism of the observed reduction process in terms of the hydrolysis of the glass wafer with subsequent interaction of the leached alkali and alkali earth cations and silicate anions with graphene oxide, resulting in elimination of the oxygencontaining groups from the latter one. The proposed approach can be efficiently used for low-cost bulkquantity production of graphene and graphene-based materials for a wide field of applications. Graphene continues to inspire interest in various fields of science due to its outstanding physical and chemical properties , even though intensive studies devoted to this unique nanocarbon material 4,5 have been carried out during the last ten years. It seems to have a wide field of applications in different technologies, including fabrication of transparent electrodes, supercapasitors and nanoelectronic devices, gas sensing and catalysis, biochemistry and microbiology . However, the preparation of graphene in large scales remains to be a challenging task. Several strategies have been developed to overcome this challenge, such as epitaxial growth of graphene on silicon carbide (SiC) 10 , growth of graphene on the surface of transition metals by chemical vapor deposition (CVD) 11 , and reduction of graphene oxide (GO) 12 . Among these methods, the reduction proved to be an effective approach to produce graphene with an optimal quality at relatively low cost 13,14 . Numerous approaches are applied to achieve conversion of graphene oxide to graphene, for instance, high-temperature annealing in a reducing enviroment 15 , ultraviolet 16 and laser 17 irradiation of GO films, electrochemical 18 and chemical 19 treatment of graphene oxide suspensions and films. Compared to other techniques, the chemical reduction of GO offers great ease of large-scale production of rGO in various forms such as suspensions, rGO paper or thin films on various substrates. Various chemicals, namely, hydrazine monohydrate 20 , dimetilhydrazine 21 , hydroquinone 22 or sodium borohydride 23 are typically employed in chemical reduction of GO. However, these reducing agents are highly toxic, unrecyclable, highly unstable, or generate hazardous by-products. Moreover, strong acidic or alkaline conditions are commonly required in these reduction processes thus limiting their applicability for preparing graphene-containing polymer composites 24 when the reduction process must be carried out simultaneously with introducing graphene into the polymer matrix. On the other hand, the use of mild reducing agents, such as green molecules presented by different types of sugars (glucose or sucrose) 25 or L-ascorbic acid 26 , results in a drastic increase in time required for the reduction of GO, and usually takes place in hot solutions. Furthermore, effective reduction of GO by these chemicals commonly requires addition of ammonia that limits the subsequent use of the obtained rGO suspensions. As such, the development of new chemical methods able to provide rapid and efficient reduction of graphene oxide under mild experimental conditions with the use of a reusable deoxygenation agent is of a high interest nowadays. In this paper we report a new facile method to convert GO to graphene under low-temperature heating by using sodium, alkali-barium and magnesium silicate glass wafers as reducing agents. The method can be applied for both GO aqueous suspensions and large-scale GO films formed on various substrates prior to reducing. A possible model of the GO reduction using different glass wafers is proposed. Overall, we demonstrate that reduction of GO via low-temperature heating in the presence of glass wafers appears to be a green, efficient and easily scalable method that is based on the use of recyclable non-toxic reducing agent and can be effectively employed for low-cost bulk-quantity production of graphene and graphene-based derivatives. ## Results Optical images and UV-Vis spectra. Figure 1a presents photographs of the initial GO aqueous suspension and GO suspensions (GO) after heating them at 80 °C in the presence of sodium silicate glass (rGO_S-gl), alkali-barium silicate glass (rGO_AB-gl) and magnesium silicate glass (rGO_Mg-gl) wafers. The suspension color change from yellow to black is an obvious visible characteristic of the successful conversion of GO into graphene 27 . The removal of the hydrophilic functional groups is further evidenced by the aggregation of rGO sheets as a result of π-π stacking interactions. The rate of the aggregation rises from the rGO_S-gl to rGO_Mg-gl sample. This fact suggests more complete elimination of oxygen-containing functional groups and larger areas of the graphene network in the latter one. Figure 1b shows the UV-Vis spectra of GO and rGO samples. The initial GO exhibits two distinctive features, the main absorption peak at 230 nm due to π-π* transitions of C=C bonds and a broadband absorption shoulder centered at 300 nm. The latter one is commonly attributed to n-π* transitions of C=O bonds of carbonyl and carboxyl groups 28 . However, it also can arise from optical transitions between π and π* states in the nanometer-size sp 2 clusters remained in the structure of GO after its oxidation 29 . Upon the reduction, the 230 nm absorption peak progressively shifts towards higher wavelengths, and overall absorption in the range up to near-infrared (NIR) region rises significantly due to the restoration of sp 2 -conjugated graphene network. As seen, the UV-Vis spectra of rGO_Mg-gl and rGO_AB-gl are almost similar with the peak of the π-π* transition lying at 265-268 nm, which is a characteristic feature of the high-degree GO reduction 30 . At the same time, the position of the main absorption peak (at 247 nm) and highly non-linear character of the absorption in the visible and NIR regions in the rGO_S-gl spectrum signifies incomplete elimination of the oxygen functionalities in the case of using sodium silicate glass as a reducing agent. FTIR and XPS spectra. Figure 2a presents the initial GO IR spectrum that exhibits a number of characteristic absorption bands related to the oxygen functionalities and interlayer water 31,32 . Particularly, the broadband absorption feature at 3000-3700 cm −1 originating from the set of overlapping bands of O-H stretching in water molecules, hydroxyls and carboxyls is observed. Additionally, distinguishable bands at 1720 cm −1 , 1620 cm −1 , 1415 cm −1 , 1225 cm −1 and 1040 cm −1 are presented. These absorption lines correspond to the stretching and bending vibrations of the carbonyl/carboxyl groups, water molecules, basal-plane hydroxyls, epoxides and edge-located hydroxyls, respectively. The distinguishable features at 980 cm −1 and 1280 cm −1 are related to the presence of the five-membered ring lactols 32 and ethers 33 . The emergence of the prominent absorption band at 1580 cm −1 that corresponds to C=C vibrations within the recuperated conjugated aromatic structure and vanishing of the absorption bands at 3000-3700 cm −1 indicates successful deoxygenation of the rGOs. However, the IR spectrum of the rGO-Sgl sample still exhibits noticeable absorption bands of epoxides (1225 cm −1 ), edge-located hydroxyls (1040 cm −1 ) and carboxyls/carbonyls (1720 cm −1 ). This suggests retention of some amount these functionalities after the reduction. At the same time, the IR spectra for both rGO_AB-gl and rGO_Mg-gl samples demonstrate nearly complete elimination of all oxygen-containing functionalities. The only absorption feature that can be distinguished is related to edge-located hydroxyl groups (phenols). The preservation of these groups is due to their high resistivity to elimination 34 . Figure 2b-d show the survey, high-resolution C 1 s and high-resolution O 1 s core level XPS spectra of the samples, respectively. The presence of prominent peaks of Na 1 s and Na KLL in the rGO_S-gl survey spectrum (Fig. 2b) demonstrates that a certain amount of Na (~4.9 at.%,) retains in the structure of this sample. The observed preservation of sodium in such concentrations can be understood in terms of substitution of the hydrogen ion in the residual hydroxyl and carboxyl groups with a sodium cation. At the same time, the analysis of the survey spectra of rGO_AB-gl and rGO_Mg-gl indicates that concentration of residual alkaline-earth metals in these samples is considerably low, appearing to be less than 0.8 at % and 0.3 at%, respectively. In the C1s XPS spectra (Fig. 2c), six distinct peaks can be discerned. The peak at 283.9 eV is attributed to carbons that are the nearest neighbors of graphene vacancy defects 35 (peak C-V). The peaks 284.6 eV and 284.9 eV are related to sp 2 -bonded carbons of perfect graphene lattice (peak C=C) and to carbon atoms being partially sp 3 -hybridized due to strong graphene network distortion caused by attachment of oxygen-containing groups (peak C-C), respectively 36 . Note that the C=C peak is asymmetric due to the natural asymmetry inherent for C1s XPS spectra of highly sp2-conjugated graphene-like structure observed in highly reduced GO films 37,38 . Other three peaks located at 286.7 eV, 288.2 eV and 288.9 eV correspond to hydroxyl and epoxide groups (C-OH and C-O-C), carbonyl groups (C=O) and carboxyl groups (COOH), respectively 36,39 . Three main O1s components (Fig. 2d) positioned at binding energies of 531.0, 532.5 and 533.6 eV are assigned, respectively, to the C=O bonds, C-O bonds within the basal plane groups (C-OH and C-O-C) and C-O bonds within phenols (C-OH(ph)) and carboxyls (O=(C-OH)) 34 . Table 1 represents the results obtained by quantitative analysis of the deconvoluted C1s XPS spectra. High content of the oxygen-containing functional groups and low calculated C/O ratio give a hint that the initial GO is highly oxidized. After the reduction, the intensities of the C 1s peaks related to the oxidized groups decrease significantly, which is accompanied by a significant rise in the C/O ratios determined to be 4.22, 5.3 and 7.41 for rGO_S-gl, rGO_AB-gl and rGO_Mg-gl, respectively. These values are very close to those of rGO prepared by chemical reduction using common reducing agents, namely, hydrazine, benzylamine, various alcohols and sodium borohydride 19,40 . Beside the difference in the C/O ratio, the rGOs obtained using different glass wafers also exhibit distinct compositions of residual functionalities. The rGO_S-gl sample is characterized by the presence of a high number of carbonyl groups, which is indicated by the prominent 288.2-eV peak in the C 1 s spectrum and domination of the C=O component in the O 1 s spectrum. This suggests that reduction of GO by using sodium silicate glass lead to formation of new carbonyls. At the same time, according to the quantitative analysis of the deconvoluted C1s XPS the content of carbonyls in rGO_AB-gl is nearly the same as in GO, suggesting absence of their elimination during the treatment. This is also indicated by the relatively high intensity of the C=O peak in the rGO_AB-gl O 1 s spectrum, which significantly differs from the rGO_Mg-gl one. The observed difference in the O 1s spectra of these samples is also related to the higher concentration of retained alkaline earth carbonates formed during the reduction. This is evidenced by a higher content of the retained Ca, Ba, and Mg in rGO_AB-gl sample in comparison to the amount of the residual Mg in rGO_Mg-gl as can be seen from the survey XPS spectra. ## Component Defects C=C XRD patterns, Raman spectra and TEM images. Figure 3a shows the diffraction pattern of the initial GO that exhibits a narrow peak at 2θ = 11.3° that corresponds to diffraction reflection from the (00.2) planes with the basal spacing (d basal ) of 7.8 . This value is consistent with the published data 41 . Additionally, the less intense broadened peaks at 2θ = 43° and 2θ = 77.7° characteristic of the (10) and (11) reflections can be observed. These peaks arise from the 2D diffraction from the planar carbon network of GO flakes. The intensity ratio between the crystalline reflections of type (00.l) and lattice reflections of type (hk) indicates that the GO flakes have a lamellar structure, and their lateral size is larger than several micrometers 42 . Further analysis of the SEM images of arrays of GO flakes on the silicon wafer and laser diffraction measurements of the GO aqueous suspension (Figure S1) confirm this assumption, demonstrating that GO flakes have lateral size of up to 100 μm which complies with the highest values reported previously 43 . After the GO reduction, the (00.2) XRD-pattern diffraction peak has shifted to higher angles due to the decrease in the rGO interlayer spacing, which proves elimination of the oxygen-containing groups. However, this peak positions are different for different types of the glass wafers. The interlayer spacing has been determined to be 4.1 for rGO_S-gl, 3.9 for rGO_AB-gl, and 4.5 for rGO-Mg-gl. These values are noticeably lower than the GO interlayer distance, which confirms elimination of interlayer water and oxygen-containing groups. On the other hand, these values are larger than both the graphite interlayer spacing of 3.4 and that of 3.7 published for the reduced graphene oxide 44 . This is due to retention of metal ions and carbonates formed during the reduction process, which cause an increase in the interlayer distance. The diffraction patterns of the rGO_AB-gl and rGO_Mg-gl samples also contain distinguishable asymmetric (10.l) and (11.l) reflections which are superposition of reflections of the (hk.l) and (hk) types. The shapes and positions of these diffraction features coincide with those in microcrystalline graphite 45 and suggest that an average lateral size of coherent scattering regions (CSR) corresponding to the defect-free regions in rGO_AB-gl and rGO_Mg-gl is 200 nm. Figure 3b shows Raman spectrum of the initial GO with a broad G peak around 1595 cm −1 related to in-plane stretching of the graphene lattice and D peak around 1349 cm −1 caused by the lattice disorder, e.g., edges of the sp 2 clusters and boundaries of the flakes 46,47 . After reduction, the frequency of the D band and G band in Raman spectra of all rGOs is equal to that of GO. The intensity ratio of these bands (I D /I G ) is commonly used to evaluate the stacking order and defect density in the obtained graphene samples 12 . Upon the reduction, the I D /I G ratio does not significantly change and remains within the range of 1.1, although commonly applied chemical reduction procedures lead to the significant rise of the I D /I G ratio 39,48,49 . This evidences that the used reduction procedure does not cause considerable structural disorder if alkali-barium silicate glass or magnesium silicate glass are applied. The absence of the observable decrease in I D /I G ratio is related to the high number of layers in the studied films that are about 500 nm in thick. The absence of a high number of defects, e.g., nanosized holes and rips, which commonly arise due to removal of oxide groups 33,47 in the obtained rGO_AB-gl and rGO_Mg-gl samples, is also shown by the obtained TEM images (Fig. 3e,f). The initial GO exhibits a continuous defect-free structure with the absence within GO flakes of any rips or holes with lateral size of more than tens of nanometers. The sharpness of the obtained diffraction spots and ratio between their intensities collectively demonstrate the monolayer character of the GO flakes. After the reduction with the alkali-barium silicate glass and magnesium silicate glass, no nanosized defects are observed in the structure of the rGO platelets (Fig. 3e,f). Moreover, a set of distinguishable hexagonal diffraction patterns rotated relative to each other can be observed in the case of rGO_Mg-gl (Fig. 3f (Inset)). This indicates that the obtained rGO_Mg-gl consists of the lamellar platelets combined in stacks of several layers having well-preserved crystalline structure with the long-range order of minimum several tens of nanometers. Note that this estimation coincides well with the aforementioned CSR area evaluated based on the X-ray diffraction data. In turn, the electron microdiffraction pattern of rGO_AB-gl (Fig. 3e (Inset)) is more ring-shaped (still having the six-fold graphene symmetry). This may be caused by high density of nanowrinkles arising due to the aforementioned retention of metal carbonates on the surface of rGO platelets. The crumpled structure of the rGO platelet is also observed for the rGO_S-gl sample where sodium-containing species have been retained on the layer surfaces. However, in opposite to rGO_AB-gl and rGO_Mg-gl samples, rGO_S-gl exhibits quite defective structure. TEM image of this sample (Fig. 3d) demonstrates that large number of holes 5-10 nm in lateral size distributed within the structure of the layer. The low structural quality of the rGO_S-gl sample is also evidenced by low intensities of (10.l) and (11.l) reflections in the XRD pattern, although the I D /I G ratio for rGO_S-gl is comparable to that for rGO_AB-gl. This discrepancy may be related to the sufficiently large distances between the formed holes and aforementioned large lateral size of the GO platelets, since the overall length of boundaries strongly affecting the intensity of the D band is considerably small in this case. ## Conductivity measurements. The difference in the efficiency of the GO reduction using different types of glass wafers is further evidenced by the conductivity measurements. The values of sheet resistance and corresponding conductivity values are summarized in Supplementary Table S1. The rGO_Mg-gl sample exhibits the highest conductivity of 33000 S*cm −1 , whereas the conductivities of rGO_AB-gl and rGO_S-gl have been determined to be 10500 S*cm −1 and 117 S*cm −1 , respectively. The values obtained for rGO_Mg-gl and rGO_AB-gl are comparable to those of rGO reduced by using borohydrides 49 , metal-acid solutions in the mild conditions 19 , and high-temperature annealing 50 . At the same time, rGO prepared using sodium silicate glass exhibits quite low conductivity due to highly defective nature of the rGO platelets. Reduction of the prior-formed GO films. One of the main issues that limits the use of the liquid-media chemical reduction method for preparing graphene for its further use is restacking of the suspension rGO platelets into graphite-like multilayer aggregates 48,51 . To solve this problem, the liquid-phase reduction of GO is carried out in strong basic or acidic solutions where the electrostatic repulsion of the remained functional groups prevents restacking 52 . Another approach is to modify rGO with various surfactants 53 . However, the use of reducing agents that efficiently convert GO to rGO under mild conditions gives the opportunity to straightforwardly reduce not GO suspensions but GO films formed on various substrates prior to the reduction process. To analyze whether the method under consideration is applicable for effective reduction of GO films without their disruption, an additional series of experiments was performed. Particularly, GO films on quartz and silicon substrates were placed in aqueous media containing magnesium silicate glass wafer with its subsequent heating at 80 °C during 5 hours. Further characterization of the obtained sample by the means of UV-Vis and FTIR spectroscopy (Supplementary Figure S2), as well as elemental analysis (Supplementary Figure S3), indicates that the treated GO films were successfully converted to rGO with a high degree of reduction. This assumption is further confirmed by the conductivity measurements demonstrating that conductivity of the obtained reduced graphene oxide film is about 30000 S*cm −1 . Importantly, the applied reduction procedure also does not result in peeling of the GO film (Supplementary Figure S4) and formation of any observable defects, e.g., rips and tears, as is indicated by the obtained SEM images (Fig. 4). Note that bright areas observed in the SEM images of the rGO film originate from the oxidation of the silicon wafer surface (Figure S5). Thus, the considered reduction procedure can be applied to reduce not only the GO suspension but also GO films formed on the surface of various substrates that are widely used in graphene-based optoelectronic devices. Recyclability of the applied reducing agent. The efficiency of using glass wafers as a reducing agent in converting GO into rGO further improved by their recyclability and simplicity of use. While conventional reducing agents are usually completely consumed during the reduction process, the glass wafer can be simply withdrawn from the aqueous media, washed by deionized water, and used again. The obtained UV-Vis and FTIR spectra (Fig. 5) demonstrate that a single glass wafer may be reused 5 times, providing effective reduction of GO. After the 5th cycle the reduction efficiency begins drastically decrease, and after the 7th cycle no significant elimination of the oxygen functionalities occurs. This is indicated by the shape of the corresponding UV-Vis spectrum (Fig. 5a, magenta curve) and retention of the distinguishable peaks at 1225 cm −1 , 1365 cm −1 and 1720 cm −1 corresponding to basal-plane and edge-located oxygen-containing groups in the FTIR spectrum (Fig. 5b). Nevertheless, glass wafers are highly recyclable as GO-reducing agents and are much easier to be reused than other deoxygenating agents. ## Discussion The observed conversion of GO into rGO both by low-temperature heating in the presence of a glass wafer can be explained by the following mechanism. In heating a GO suspension with a glass wafer immersed into it to 60-80 °C, the wafer begins dissolving due to acid-catalyzed bimolecular displacement reactions . This process results in leaching of the alkali and alkaline-earth cations (Na + , Mg 2+ , Ba 2+ , Ca 2+ ) along with metasilicate (SiO3 2− ) and ortosilicate (SiO4 4− ) anions from the glass surfaces into the suspension . In the process, the pH value rises from 3.4 to 8-8.5. Beside the formation of alkali and alkaline-earth silicates, the leached silicate anions may also interact with the oxygen-containing functional groups of GO. In this case intermediates are formed, composed by the metasilicate or ortosilicate anion that is attached to the GO layer by the oxygen-bridge bond originated from rearrangement of chemical bonds in epoxides (Fig. 6a), hydroxyls (Fig. 6b) or carboxyls (Fig. 6c). The obtained spectra demonstrate that glass wafers as a reducing agent can be reused up to five times without significant loss in the effectiveness. The assembled intermediates can be further eliminated by two possible ways depending on the chemical composition of the used glass wafer. If the alkali silicate glass wafer is used then further redistribution of the electron density in the intermediate may lead to the subsequent removal of the silicate anion. This results in the cleavage of carbon bond in the graphene network with formation of the carbonyl group (Fig. 6d,e). This way for eliminating oxygen functionalities from GO in the presence of alkali silicate glass is well supported by the observed rise in the carbonyl group concentrations and perforation of the carbon network in the rGO_S-gl sample. In turn, the use of glass wafer that contains alkaline-earth metals, such as Mg or Ba, provides an alternative way of the intermediate transformation. In this case, an alkali-earth cation existing in the suspension interacts with the formed intermediate, resulting in their removal from GO with formation of the alkaline-earth metasilicate and alkali hydroxydes (Fig. 6f,g), while the graphene network remains intact. Furthermore, the presence of the alkaline earth cation also provides elimination of the carboxyl and carbonyl groups. In the case of carboxyls, the intermediate formed from the carboxyl groups and ortosilicate anion is removed and the carbonate along with phenol group are formed (Fig. 6h). Note that the supposed increase in the number of phenols groups after the reduction clearly manifests itself in arising of the 533.6-eV component in the O 1 s XPS spectra (Fig. 2d) of the rGO_AB-gl and rGO_Mg-gl samples. Elimination of carbonyls does not require silicate anions and is based on hydration of carbonyl group 60 with formation of two adjusting hydroxyl groups. These groups further interact with the alkaline earth cation, resulting in their elimination and formation of alkaline earth hydroxide. Magnesium exhibit the lowest value of heat of hydroxide formation in the set of Mg, Ca and Ba 61 . This results in more effective removal of carbonyls using magnesium silicate glass then by using alkali-barium silicate glass, as is indicated by XPS data (Table 1). Moreover, Mg cation diffuse out more easily then Ba or Ca cation due to the lower ionic radius of the former one 62 . As a net result, magnesium silicate glass provide more effective reduction of GO. Taking into account the data of the relative concentration of oxygen-containing groups in the initial GO and rGOs we further estimated the number of alkali-, alkali earth-cations and silicate anions consumed for reduction of the studied suspensions. Details of the calculations can be found in Supplementary materials and the obtained results are presented in Supplementary Tables S3 and S4. As seen, for the 200 µl of GO suspension 0.003 wt% in concentration the number of alkali and alkaline earth cations and silicate anions for all the used glass wafers lies within the range of 0.16 µmol, 0.085 µmol and 0.08 µmol, respectively. These values consist with the published data on the concentration of the anions and cations leached during the glass dissolving with the comparable mass and surface area 55,57 . The number of cations and anions required for conversion of 1 mg of GO into graphene was additionally calculated and the obtained values can be found in Supplementary Table S5. To validate the proposed mechanism, a series of control experiments was carried out. In these experiments GO was heated under otherwise identical conditions in the presence of either a quartz wafer (rGO_Quartz), or magnesium sulphate (rGO_MgSO4), or sodium hydroxide (rGO_NaOH), or sodium silicate (rGO_Sil), or a combination of sodium silicate and magnesium sulphate (rGO_Sil-Mg). The degree of reduction and chemical composition determined for the resulting rGOs is indicative of the role of each component (silicate anions, alkali and alkaline-silicate cations) in reducing GO with glass wafers. The low-temperature (80 °C) heating of the GO aqueous suspension in the presence of the quartz wafer does not result in the deoxygenation of graphene oxide, as shown by UV-Vis, FTIR and XPS spectra (Figure S6). The treatment of GO with magnesium sulphate results in some elimination of the oxygen containing groups (Figure S7), but the reduction degree (C/O = 1.47) is drastically low and is not comparable to the values for the rGOs obtained using glass wafers. These results collectively indicate that introduction of both silicate anions and metal cations into the medium as a result of glass dissolving is of paramount importance for the GO reduction. The primary role of the silicate anions in the observed reduction process was further verified by analyzing the additional rGO_NaOH, rGO_Sil and rGO_Sil-Mg samples. Figure 7 demonstrates C1s and O1s XPS spectra of the obtained rGO samples and results of quantitative analysis of these XPS spectra presented in Table S2. Although strong alkaline solution has been reported to deoxygenate exfoliated GO sheets at the temperatures above 55 °C36,39 , the rGO obtained in alkaline solution with pH~8 has the C/O ratio of 2.9 which is significantly lower than those even in the rGO_S-gl samples. On the other hand, the reduction degree of rGO_Sil (C/O = 4.99) is very close to that in rGO_S-gl (C/O = 4.2), and an increase in concentration of carbonyl groups is observed upon the reduction with both sodium silicate glass and sodium silicate powder. Moreover, the formation of nanoscale holes in the rGO_Sil sample is also demonstrated by the obtained TEM image (Figure S8). In turn, the rGO samples obtained either in the presence of glass wafers containing alkaline-earth oxides or by using sodium silicate mixed with magnesium sulphate exhibit almost equal C/O ratios and compositions of the residual groups. Thus, the obtained experimental results confirm the assessment that the studied reduction process originates from the presence of silicate anions. At the same time, the structural parameters and chemical composition of the obtained rGOs are determined by the type of the metal cations, presented in the suspension. In summary, we for the first time have demonstrated that GO can be easily reduced by low-temperature heating in the presence of various glass wafers, namely, sodium silicate, magnesium silicate and alkaline-barium silicate glass wafers. The discussed method can be used to reduce both GO suspensions and GO films formed on various substrates without any considerable effect on the film morphology in the latter case. The additional studies have also confirmed recyclability of the glass wafers used as reducing agents, i.e., the possibility to efficiently reduce GO five times with a single glass wafer. The mechanism of the observed reduction process has been studied as well, revealing that the GO to rGO conversion by using glass wafers occurs due to cooperative interaction of the leached silicate anions and metal cations with oxygen-containing function groups of GO. The advantages of the proposed reduction method, i.e., its simplicity, low reaction temperature, recyclability and non-toxicity of the reducing agent, and the absence of strong acids and bases, make it attractive for the large-scale production of graphene and graphene-based materials for various applications, e.g., fabrication of composite fillers, graphene-based inks, and graphene coatings for optoelectronic devices. ## Methods Formation of GO suspensions and films. Graphene oxide was synthesized by the Hummers method 63 . In brief, graphite powder (4 g) was oxidized by using concentrated H 2 SO 4 , KMnO 4 , NaNO 3 , and H 2 O 2 solutions. The resulting mixture was centrifuged (3500 rpm for 1 hour), and the supernatant was decanted away. The material remaining after this was additionally centrifuged (1500 rpm for 10 min) to obtain aqueous GO suspension as a supernatant. In the process of synthesis, sonication was excluded to prevent damaging of graphene oxide flakes and obtain suspensions with the utmost size of GO flakes (with lateral size of up to 100 μm). To prepare GO films for the subsequent reduction, 200 μL of GO aqueous suspension 0.003 wt % in concentration was drop-casted on silicon and quartz wafers and dried overnight at room temperature. ## Reduction of GO. Three types of glass wafers with different chemical compositions were used as possible reducing agents for GO deoxygenation: sodium silicate glass containing only sodium oxide, magnesium silicate glass containing only sodium oxide and magnesium oxide, and alkali-barium silicate glass, as one of the most common glass types, containing various alkali and alkaline earth oxides. The chemical compositions of the used glass wafers and their price are presented in Table S6. The reduction of GO aqueous suspensions was performed as follows: a piece of sodium silicate glass (10 × 6 mm wafer, 0.135 g), alkali-barium glass (16 × 7 mm wafer, 0.53 g) or magnesium silicate glass (14 × 8 mm wafer, 0.45 g) was immersed into GO aqueous suspension (40 mL) 0.01 wt% in concentration with subsequent stirring of the suspension at 80 °C for 5 hours in a fluoroplastic flask. The obtained rGO suspensions were copiously washed by centrifuging (centrifuge Sigma 3-30KS) at 26,200 rpm (60.600 g) and rinsing the obtained sediment with de-ionized water. The described purification procedure was repeated five times. The obtained rGO samples were denoted as rGO_S-gl (reduced by sodium silicate glass), rGO_AB-gl (reduced by alkali-barium silicate glass) and rGO_Mg-gl (reduced by magnesium silicate glass). The quantity of graphene, obtained from the 0.5 g piece of glass wafer (Sodium, Alkali Barium or Magnesium Silicate) with actual size of 15 × 7.5 × 1.0 mm was determined to be about 50 mg. Magnesium silicate glass wafer was applied for the reduction of graphene oxide suspension up to 7 times with successful conversion of GO into rGO during 5 cycles. As a result, the maximum quantity of the produced graphene from a single glass wafer in the applied conditions was determined to be about 0.25 g. To analyze applicability of the studied method for reducing GO films on substrates, the GO films on quartz or silicon substrates were put into a fluoroplastic flask filled with de-ionized water (40 mL); after that, a piece of magnesium silicate glass was added, and the flask was heated at 80 °C for 5 hours. After the reduction, substrates with the rGO film were carefully withdrawn from the solution, washed several times with de-ionized water, and dried overnight at room temperature. For better understanding of the processes that lie behind the observed deoxygenation of GO, a series of control experiments was carried out. Namely, GO aqueous suspensions were heated at 80 °C during 5 hours in the presence of the quartz wafer, after adding 0.01 mol. of magnesium sulphate powder (obtained from Acros Organics Company), or 150 µL of NaOH solution (0.1 M, obtained from Acros Organics Company), or 0.7 mmol. of sodium silicate powder (obtained from Acros Organics Company), or 0.8 mmol. of sodium silicate together with 0.8 mmol. of magnesium sulphate. The obtained samples were washed according to the aforementioned procedure. To provide the correct alignment and deconvolution of the XPS spectra of GO and whole series of the studied rGOs, an additional rGO sample denoted as rGO_HT was prepared by annealing the GO film at 600 °C during 2 hours. Characterization of the obtained rGO samples. The pH values of the solutions were determined with a Fisher Scientific Accumet Basic AB15 pH meter. The UV-vis absorption spectra of the GO and rGO samples were collected with a Shimadzu-2450 spectrophotometer. Fourier transform infrared spectroscopy was performed on the Infralum-08 FTIR spectrometer equipped with the attenuation of total reflectance attachment. X-ray photoelectron spectroscopy (XPS) measurements were carried out on a Thermo Fisher ESCALAB 250Xi XPS system with a monochromatic Al Kα X-ray source (1486.6 eV). The spectra were calibrated with respect to the Au 4f7/2 line (84.0 eV). A surface charging revealed for low-conducting GO (and some rGOs) was taken into account by the aligning their XPS spectra with respect to the C1s line position (284.6 eV) for a well-conductive rGO-HT sample (see Supplementary Figure S9). The quantification and curve fitting of the obtained XPS spectra were performed by using standard CasaXPS software. The X-ray diffraction (XRD) analysis was carried out using a Bruker Smart Apex Duo installation with a CuKα source and Apex 2D detector. The sample for diffraction measurements was fixed with nitrocellulose lacquer at the end of a cactus needle. Diffraction patterns were measured at various angles between the normal to the detector surface and the X-ray direction, the 2D data being subsequently recalculated to the 2θ configuration. The obtained diffraction patterns were analyzed using the DIFFRAC.EVA (Bruker Cor.) software based on the data from Powder Diffraction File ICCD PDF-2 release [JCPDS-International Centre for Diffraction Data (http:// www.icdd.com)]. Raman spectra were obtained on a Horiba Jobin-Yvon LabRam HR800 installation equipped with a Laser Quantum Torus 532-nm laser 50 mW in output power. The exciting light was additionally attenuated with a filter having optical density of 1 and focused with a 20x objective lens into a spot approximately 30 μm in diameter. The power reaching the sample after passing the lightpath and objective was 0.11 mW. Size distribution of GO and rGO flakes in aqueous solution was determined by laser diffraction measurements using Mastersizer 2000. Transmission electron microscopy (TEM) images were acquired with a Jeol JEM-2100F microscope (accelerating voltage 200 kV, point-to-point resolution 0.19 nm). Samples for TEM were prepared by deposition of aqueous GO and rGO suspensions 7•10−4 wt % in concentration onto conventional lacey carbon films. SEM images were collected with a JSM-7001F, Jeol microscope. Monolayer films for SEM imaging were prepared by the Langmuir−Blodgett method according to the procedures published elsewhere 64 . Surface morphology and thickness of the rGO films were analyzed with a Veeco Dimension 3100 atomic force microscope operating in the tapping mode by using RTESP probes. Electrical conductivity measurements of the GO and rGO samples were performed on the base of two-electrode system. The GO and rGO films were deposited by the drop-casting method onto the surface of quartz substrates with two comb Au electrodes 80 nm thick separated by 500 µm. The electrode comb consisted of 8 pairs of the electrode bars (Figure S10).

chemsum

{"title": "Facile reduction of graphene oxide suspensions and films using glass wafers", "journal": "Scientific Reports - Nature"}

high-throughput_discovery_of_hf_promotion_on_the_formation_of_hcp_co_and_fischer-tropsch_activity

## Abstract: A proxy-based high-throughput experimental approach was used to explore the stability and activity of Co-based Fischer-Tropsch Synthesis catalysts with different promoters on a variety of supports. The protocol is based on XRD estimation of the active phase polymorph, particle size and ratio of crystalline phases of Co to support. Sequential sample libraries enabled exploration of four Co loadings with five different promoters on six support materials. Catalysts stable to aging in syngas, i.e. displaying minimal change of particle size or active phase concentration, were evaluated under industrial conditions. This procedure identified SiC as a support that confers catalyst stability and that a combination of Ru and Hf promotes the formation of hcp Co. Unsupported bulk samples of Co with appropriate amounts of Ru and Hf revealed that the formation of hcp Co is independent of the support. The hcp Co-containing catalyst afforded the highest catalytic activity and C5+ selectivity amongst the samples tested in this study, confirming the effectiveness of the proxy-based high-throughput method. ## Introduction The Fischer-Tropsch Synthesis (FTS) is an important process for the production of long chain hydrocarbons from syngas (a mixture of CO and H2) using catalysts with Co, Fe or Ru as active metals. The development of active and stable catalysts is a key factor for the implementation of FTS in the production of fuels and chemicals . The activity of FTS catalysts is attributed to different factors such as the particle size, active metal phase, the metal surface area, and the interactions with the support material . The presence of different Co polymorphs (fcc and hcp) for example plays an important role in the activity of the catalysts and has been subject to many studies recently, showing that the hcp Co phase is more active than the fcc Co phase in FTS [3,4, . The deactivation of the catalysts can occur through different mechanisms such as sintering, oxidation, formation of metal-support compounds, carbon deposition and poisoning . Realistic investigations addressing the activity and stability of FTS catalysts require extended testing times and diverse catalyst sets to delineate the effect of the different factors. The formation and stabilization of the hcp Co phase for FTS catalysts has been thoroughly investigated [3, 4, 10-12, 15, 16]. The hcp Co phase is destabilized with increasing temperature because of its higher surface energy compared to fcc Co . Thus, above a temperature of approximately 400 °C fcc Co is the main phase observed . De la Pena et al. reported the formation of Co hcp particles by reducing Co3O4 nanoparticles in a H2 and CO gas mixture. The authors observed the formation of a graphitic layer encapsulating the hcp Co particles, which stabilizes the hcp phase up to a temperature of 700 °C by decreasing the surface energy of those particles. To achieve the formation of the hcp phase on a support material, Gnanamani et al. treat supported Co samples under CO flow and moderate temperatures (230 °C) to form Co2C. The reduction of the Co2C at 230 °C under H2 leads to the formation of the hcp Co, which is attributed to structural similarity between these phases as both crystallize in the hexagonal system. No investigation has facilitated the formation of the Co hcp phase by adding promotors and reducing the samples under H2 at temperatures as high as 500 °C. High throughput (HT) experimentation is used to accelerate the screening of large catalyst sets, it has been successfully applied in the development of heterogeneous catalysts , and can be applied for the identification of stable FTS catalysts, by parallelizing the preparation, ageing and characterization of the samples. While HT testing of the catalytic properties of the materials demands access to specialised equipment , the use of a proxy can accelerate the identification of suitable, stable and active materials without having to test all of the elements of a library. In a previous paper we reported the development of a proxy-based method for the accelerated discovery of stable FTS catalysts. The workflow (Supplementary Note 1) developed includes the parallelized preparation, reduction, characterization of the samples augmented by an aging procedure under FTS similar conditions (H2:CO = 3; 230 o C, 1 bar) for 100 h. The core of the proxy is the assessment of the stability and activity of the samples by estimating the change in particle size and the amount of metallic Co with the aging procedure. This is carried out by comparing the Co peak width and area (calculated relative to the support peak area). The last step is to run high throughput TGA on selected samples to estimate the degree of reduction. The use of this method identified the composition of a series of highly stable, high surface area supported Co catalysts promoted by Mg and Ru, and the optimum preparation protocol (i.e. order of addition, calcination steps). The proxy protocol described above is applied here to investigate the effect of different support materials and promotors on the stability of Co based catalysts. The iterative workflow was used to first screen six different support materials (active carbon, Al2O3, SiC, SiO2, TiO2, ZrO2) and five promotors (Ru, Re, Mo, Mn, La). This set of experiments identified SiC as a suitable support material and the screening of further promotors (Ru, La, Ce, Pr, Nd, Hf) demonstrated the positive effects of a combination of Co, Ru, Hf on SiC, which lead to the formation of hcp Co. Catalytic experiments under FTS conditions revealed higher C5+ selectivity for a catalyst promoted with Ru and Hf. A more comprehensive study of the effects of Ru and Hf on the formation and stabilization of the hcp Co without support materials, shows that Ru and Hf alone stabilize this phase, but the effect increases significantly, when both are present, stabilizing the hcp Co phase to a temperature up to 500 o C and 12 h. ## Sample preparation Catalysts were synthesized by incipient wetness impregnation of the support materials. The support materials were weighed by a Quantos automated balance (Mettler Toledo XPE206) (250 mg) into 24 vials. An Eppendorf ep Motion 5075 was used for the liquid deposition onto the support materials; a plate holding an array of 4 x 6 vials containing the weighed support materials, allowed a variation of four different Co loadings and six different promotors on each support material this gives a total of 24 samples with different composition on each of the six support materials (Table S1), i.e., a total of 144 samples. The impregnation of the support materials is a multistep procedure. In a typical synthesis of a material with 20 wt.-% of Co, 0.212 ml of a 4.0 M Co(NO3)2•6H2O solution were added to the 24 vials each containing 250 mg of SiC using the Eppendorf ep Motion 5075, followed by the addition of 0.047 ml of water. After mixing the materials with a spatula, the vials are placed on a shaker for 1 h to obtain a homogenous distribution of the solution on the support material. The samples were dried at 100 °C for 16 h. The impregnation of each promoter (Ru, Re, Mo, Mn, and La in the first library and Ru, La, Ce, Pr, Nd, and Hf in the second library) is performed with the respective solution and water amounts. The exact amount of promoter precursors and solutions used in each impregnation are provided in the Supplementary Information file (Supplementary Note 2). The addition of promoter solution is followed by a drying step at 100 °C for 16 h and in samples with two promotors the same drying step is performed between the additions. Finally the samples are calcined at 300 °C under air for 6 h for the decomposition of the nitrates. The following step is the reduction of the samples under pure H2 flow (50 ml/min) at 500 °C for 12 h in ceramic well plates (128x43x6 mm, Figure S2). The ceramic well plate can hold 48 different samples and three of the plates can be reduced each time, making a total of 144 samples per run. Before taking the catalysts out of the furnace they are passivated with 1 % O2 in N2 (100 ml/min) at room temperature. The ceramic well plates are directly placed in a high throughput XRD (HT XRD, Panalytical X-pert Pro diffractometer with an XYZ stage) for the first characterization of the samples. The aging procedure of the samples was performed using the ceramic well plates under conditions approximating low temperature FTS with the flow of syngas (H2:CO = 2:1, 90 ml/min) at 230 °C and ambient pressure for 100 h. After the aging procedure the samples were passivated with 1 % O2 in N2 (100 ml/min) at room temperature and transferred to the HT XRD for characterization. ## Scale up of samples Samples selected for scale up were prepared manually in an identical manner to the procedure described above. In a typical synthesis, a sample with 20 wt.-% Co, 0.1 wt.-% Ru, and 5 wt.-% Hf supported on SiC (CoRuHf SiC) was prepared as follows: 1.697 ml of the 4.0 M Co(NO3)2•6H2O solution was added to 0.463 ml water and deposited on 2 g of SiC in a porcelain dish (80 mm diameter) placed on a shaker. The sample was dried at 100 °C for 16 h, prior to the next impregnation. The next impregnation with the 0.1 M Ru(NO)(NO3)3 solution is performed in an identical manner, but with the respective solution (0.197 ml) and water (2.083 ml) amounts, followed by a drying step at 100 °C for 16 h. Finally, the impregnation with 0.47 M HfCl2O is performed with the respective solution (1.192 ml) and water (0.974 ml) amounts, followed by a last drying step at 100 °C for 16 h and a posterior heat treatment at 300 °C under air for 6 h for the decomposition of the nitrates. The second step is the reduction of the samples, which was performed in a tube furnace under pure H2 flow (50 ml/min) at 500 °C for 12 h in ceramic boats (80 mm). Before taking the catalysts out of the furnace they are passivated at room temperature with 1 % O2 in N2 (100 ml/min). ## Unsupported samples Unsupported samples were prepared by mixing quantities of the solutions to obtain the desired atomic ratios. In a typical synthesis of a sample with Ru:Co = 0.003 and Hf:Co = 0.08 atomic ratios 1.073 ml of the 4 M Co(NO3)2•6H2O solution, 0.127 ml of the 0.1 M Ru(NO)(NO3)3 solution and 0.722 ml of the 0.47 M HfCl2O solution were mixed together. After mixing the solutions, the drying, calcination and reduction steps are identical as for the supported samples described above. ## Characterization High-throughput X-ray diffraction, XRD, was performed on a Panalytical X-pert Pro diffractometer with an XYZ stage using Co Kα radiation between 38°-55° 2θ, with a 0.039° step size and 2 s/step. The particle size of the Co metal phase was calculated from the line broadening of the most intense fcc Co peak (111) using the Scherrer equation in X'Pert HighScore Plus software using a Si standard to determine the instrumental line width. Relative crystallinity was determined from the Co (111) to support XRD peak area ratio. XRD of bulk samples was performed with a Panalytical X-pert Pro in Bragg-Brentano geometry laboratory X-ray diffractometer with Co Ka1 radiation = 1.78901 . Synchrotron XRD data was recorded on beamline I11 (λ = 0.825972 ) at Diamond Light Source, UK. Temperature programmed reduction (TPR) was measured using a Quantachrome ChemBET 3000 unit with a TPD; ca. 50 mg of the calcined sample were loaded into a quartz cell and heated up to 100 °C for 30 min under He (100 ml/min) to remove moisture and other adsorbed species from the samples. After cooling down to room temperature the sample was heated to 750 °C at 5 °C min −1 under a flow of 5% H2 in N2 (100 cm 3 min −1 ) to perform the analysis. ## Fischer -Tropsch Synthesis The activity tests in the FTS were performed at Drochaid Research Services Ltd in St. Andrews. The high throughput catalyst testing was done using a custom-built 32-tube fixed bed reactor test plant, designed by Integrated Lab Solutions and built by Premex. One unit, consisting of two heating blocks with eight reactors each, was used for the tests. The reactors (L = 300 mm; ID = 2.4 mm; OD = 6.5 mm) were accommodated in a heating furnace with an isothermal zone of 100 mm. The catalyst bed length is of 7 cm with in the isothermal zone. SiC was used as an inert diluent material. The flow of the gases CO (CP grade, BOC), H2 (CP grade, BOC) and Ar (CP grade, BOC) as an internal standard was controlled by mass flow controller and the pressure by a digital back pressure regulator. The concentration of reaction educts and products was measured using online GC techniques (Agilent Refinery Gas Analyser with a FID channel for the quantification of light hydrocarbons (DB-1 column (2m x 0.32mm x 5μm) and HP-AL/S column (25m x 0.32mm x 8 μm)) and two TCD channels; TCD1 for the permanent gases CH4, CO2, Ar, N2, and CO (HayeSep Q column (6 ft., 80/100 mesh) and a Mol Sieve 5A column (6 ft., 60/80 mesh)) and TCD2 for H2 (HayeSep 3 ft.,80/100 mesh and Molsieve 8 ft., 60/80 mesh). The activity of the catalysts is expressed with cobalt time yield, which was calculated based on molar consumption per gram cobalt per second. ## Results of the first library The HT exploration of FTS catalysts is based on the prior assessment of stability of the active phase, metallic Co, on treatment with syngas as proxy for their activity. The applied workflow involves the parallel synthesis, syngas treatment and XRD characterisation of sample libraries. These are prepared by automated incipient wetness impregnation and reduced concurrently by H2 at 500 o C. HT-XRD analysis is employed to determine the phase of Co and its particle size (Supplementary Note 3) before and after the aging of samples under a syngas mixture, H2:CO = 2:1, at 230 o C for 100 h. Large changes in the particle size of Co or in its overall crystallinity, estimated by the peak area ratio of Co to support, are taken as indicators of the lack of stability and predictors of low catalytic activity for FTS. The effects of the support material, the Co loading and the addition of promotors were investigated in the first library, which consisted of 144 samples. Support materials Al2O3 [3,13, , TiO2 [8,13,30, ZrO2 , SiO2 [4,13,30,34, active carbon (AC) , and SiC were used. Despite the reactivity of metal oxides, particularly Al2O3, with Co they were selected as supports for their porosity and thermal stability. Non-oxide supports do not interfere with Co, which can be reduced at relatively low temperatures. Ru, Re, Mo, Mn and La were used as promotors. Ru and Re are known to increase the reducibility of cobalt on support materials and increase the resistance to deactivation. Furthermore, Ru is known to be the most active element in FTS. The transition metals Mo and Mn were selected too, since they have been reported to increase the chaingrowth probability factor to larger hydrocarbons in the product stream and finally the rare earth metal La, which has been reported to increase the dispersion of Co on the support . Based on a first assessment of the HT XRD patterns, samples showing no metallic Co peaks after the reduction or after the aging procedure were immediately ruled out. This is the case for the samples supported on ZrO2 and samples with only 5 wt% Co loading on every other support. The results of crystallite size analysis for aged samples are plotted versus the crystallite size before aging (Figure 1a). The y=x line is used to display the increase, points above the line, or the decrease, points below the line, of the crystallite size after the syngas aging process. Samples prepared on SiO2, AC and TiO2 show an increase in the crystallite size after aging, up to a doubling of the particle size after the aging procedure observed for SiO2, AC and TiO2, while the samples supported on SiC show an increase to a lesser extent, from 27 to 36 nm. For samples supported on Al2O3 a decrease of the particle size from 33 to 16 nm, in the most extreme case, was observed. Figure 1b shows the peak area ratio between Co and support, for the three crystalline supports (Al2O3, TiO2 and SiC), before and after the aging procedure. TiO2 and Al2O3 show a decrease of the peak area ratio after the aging procedure (up to 70 % Co peak area ratio loss on samples supported on TiO2 and up to 60 % Co peak area ratio loss on samples supported on Al2O3), while the area ratio of samples on SiC show the smallest change (between 8 and 23 % Co peak area ratio increase). The effect of the different promotors can clearly be observed for the samples supported on SiC (Figure 1c and d). The samples that show the smallest change in the Co peak area ratio are those promoted with Ru (from 0.63 to 0.75), La (from 0.65 to 0.83) and Mo (from 0.70 to 0.77). ## Fischer-Tropsch Synthesis -First test Since the samples prepared on SiC show the smallest change of Co crystallite size and peak area ratio after the aging procedure, they were identified as hits from the first library. The samples with a Co loading of 15 wt%; promoted with Ru and Re, at 0.1 wt% loading, and Mn, Mo or La, 5 wt% loading were scaled up and tested in FTS under industrial conditions. These samples were named after the active component and the promoter, e.g. the sample CoRu had 15 wt% Co and 0. ## Results of the second library The proxy screening on the first library proved the suitability of SiC as a support material and the positive effect of Ru and La on the stability of Co after the aging procedure. Furthermore, the catalytic test demonstrated the positive effects of the simultaneous promotion of La and Ru enhancing the CO conversion rate and the C5+ selectivity. A second library focussed on the effect of the content of Ru and the inclusion of rare earth elements (La, Ce, Pr, Nd), and Hf as promotors for Co-based materials supported on SiC. While the effect of rare earths on Co-based catalysts has been thoroughly investigated , little is known about the effect of the promotion of Hf and its effect on the phase formation and stabilization of Co particles. Hf was chosen, because it is in the same group as Ti and Zr, whose oxides are known promotors and support materials for FT catalysts and because it is in the same period as the rare earth metals. Many patents mention the use of Hf as a promoter for Co based FTS catalysts, but only one patent could be found, where experimental data showing the positive effect of Hf promotion of Co-based catalysts is presented. The effect of rare earths (La, Ce, Pr, Nd) or Hf addition at different loading levels was investigated for samples loaded with Co and Co and Ru, as shown in Table 3. The results obtained after submitting the samples to the workflow described above can be seen in Figure 3, where peak area ratio of Co after the aging procedure is plotted over the peak area ratio before the aging procedure. Each colour represents a different promoter, the shape of the symbols represents different levels of each promoter and the open symbols represent samples without Ru. Samples with Ru and 2.5 or 5 wt% Hf, along with samples promoted with Ru and Ce and samples with Ru and Pr, show a good stability under aging conditions. Samples promoted without Ru prove to be stable under the aging conditions but show lower Co peak area ratios than samples promoted with Ru. The samples promoted with Ru and Hf are particularly striking, whilst the sample with 5 wt% Hf and 0.1 wt% Ru stabilizes the Co peak area as shown in Figure 3, it also shows a further peak on the XRD patterns, which corresponds to the hcp Co phase (Figure S3). The formation of the hcp Co phase is also observed for the rest of the samples promoted with Ru and different levels of Hf. ## Scale up of samples The samples with 5 wt% loading of promotors displayed the highest stability in the proxy screening of the second library and they were selected for scale up, characterisation and FTS testing. All samples have fixed loadings of Co (20 wt%), Ru (0.1 wt%), and either Ce, Pr, La or Hf (5 wt%) or no promoter. A sixth sample with standard Co loading (20 wt%) alone supported on SiC was made for comparison purposes. The samples are named according to the active component and the promotors in a similar fashion to the first set of samples. The synchrotron PXRD patterns of the six scaled up samples and of the pure SiC are presented in ## Temperature programmed reduction In order to understand the effect of each promoter on the reduction behaviour of Co, the samples were submitted to a TPR analysis under 5 % H2 in N2 (Figure 5). The sample containing only 20 wt% Co shows two peaks, the first one at a lower temperature (300-375 °C) is associated with the reduction of the Co(III) species to Co(II), and the second one, at a higher temperature (400-500 °C) with the reduction of the Co(II) species to Co(0) . The addition of Ru to the samples enhances the reducibility of the samples lowering the temperature of both reduction steps. The addition of both Ru and rare earth metals as promotors leads to an increase of the reduction temperature for both steps, compared to those promoted by Ru alone, as illustrated by the TPR for the samples promoted with Ru and La or Pr. The sample promoted with mixture of Ru and Ce shows similar reduction behaviour in relation to the sample promoted only with Ru. Again it is the sample promoted with Ru and Hf that exhibits the most differentiated reduction behaviour. No reduction is observed up to 330 °C and the peaks for both reduction steps merge together in a single peak between 330-460 °C. ## Fischer-Tropsch Synthesis -Second test The scaled-up materials were tested under industrial FTS conditions in a parallel reactor system at 20 bar and three different temperatures (Figure 6a). At 210 °C the materials CoRu, CoRuLa, CoRuCe and CoRuPr show similar Co time yield and only the material CoRuHf shows higher performance. At 220 °C all the materials display an increase in their activity with CoRuHf being the most active and CoRuPr showing higher yield than the rest of the samples. At 230 °C all the materials show different CO conversion rates in the FTS with the following activity order CoRuHf> CoRuPr> CoRuLa> CoRuCe> CoRu. The material promoted with Ru and Hf not only shows the highest yield at all temperatures, it also displays the highest selectivity values towards C5+ components as Table 4 shows. (11.84 %). Methane, being a raw material for the production of syngas, and CO2, are undesired by-products in FTS . A comparison between the fresh and used samples is shown on Figure 6b. Compared with the fresh catalysts the XRD characterization of the spent catalysts shows no significant loss of the Co phases for any of the catalysts. The CoRuHf shows the presence of the hcp Co phase in the fresh and in the spent material. Both fcc and hcp Co phases are stable under the testing conditions, and the deactivation observed at 230 °C during the catalytic test is not caused by the formation of Co oxides. The higher activity and C5+ selectivity values of the Hf containing material compared to the rest of the catalysts is consistent with the formation of the hcp Co phase. The effect of the different Co phases on the activity of the catalysts has been studied previously [3-5, 11, 12, 25, 43] and it was demonstrated hcp phase is more active the fcc one. ## Unsupported Samples The formation of the hcp Co phase supported on SiC and promoted with Hf and Ru motivated us to further investigate the effect of Hf and Ru on formation of hcp Co. To eliminate the effect from the support material a set of samples with different Hf:Co and Ru:Co atomic ratios were prepared and reduced at temperatures between 300 and 500 o C for different times. Samples showing a reduction time of 0 h were held at the reduction temperature for 5 min before cooling down at the natural rate of the furnace. The PXRD characterization results of the samples with three different Hf:Co ratios are provided in Figure 7. The sample with Hf:Co = 0 is reduced completely at 300 °C and shows the formation of the hcp Co phase. An increase in the reduction temperature to 500 °C leads to the formation of the fcc Co phase, which is clearly observed by the appearance of the (200) peak at 2θ = 60.34°. Hcp Co is further converted to the fcc phase with longer reduction times, 2 and 12 h, at 500 °C. The phase composition of each sample has been obtained by Rietveld refinement of the PXRD patterns (Supplementary Note 4). The final sample of this series, reduction for 12 h at 500 °C, contains hcp Co phase, 48 wt% and fcc Co, 52 wt% (Table S2). The addition of Hf to the samples shifts the temperature window of Co reduction and hcp to fcc conversion, as the PXRD patterns of the samples with Hf:Co = 0.04 and Hf:Co = 0.08 show (Figures 7b and c). The sample with Hf:Co = 0.04 reduced at 300 °C show mainly the peaks of CoO and some hcp Co (8 wt%). At 400 °C only the metallic phases of Co can be observed and hcp Co remains the dominant phase, 80 wt%, even after reduction at 500 °C for 12 h. Higher Hf content, Hf:Co = 0.08, increases further reduction and phase transformation temperatures. At 300 o C only the oxide phases CoO and Co3O4 are the present and their reduction has been completed at 500 °C. After reducing this sample at 500 °C for 12 h the most of Co is present as the hcp phase, 78 wt%, while very broad peaks corresponding to monoclinic HfO2 have been appeared. ## Fischer-Tropsch Synthesis -Third test The promoting effect of Hf on the stability of hcp Co and FTS activity that was observed on the samples of second library motivated the further investigation its effect on catalytic activity when combined with two Ru contents, 0.01 wt% and 0.1 wt%. Five samples were prepared with fixed loading of Co (20 wt%), Ru loading 0.01 and 0.1 wt% with and without Hf (5 wt%). PXRD measurements (Figure S7) confirmed the presence of both Co phases in samples containing Hf while the samples without Hf display mainly fcc Co and only traces of hcp Co. The materials were tested in FTS in four different temperatures (Figure 9) and the catalytic activity of the samples containing Hf at each temperature is higher than the noncontaining Hf counterparts by factor of ~ 1.6. Moreover, the sample 20Co/0.01Ru/5Hf display slightly higher Co time yield than the sample 20Co/0.1Ru at 230 o C and 240 o C. This result demonstrates that promotion with Hf at 5 wt% and the subsequent formation of hcp Co compensate the loss of activity caused by the decrease of Ru content by an order of magnitude. ## Conclusions Application of the described high throughput protocol enabled the efficient screening of different support materials and promotors for the Co based FTS catalysts. The protocol allowed us to identify SiC as a suitable support for active and stable catalysts. Furthermore the promotors Ru and Hf were found to not only increase the stability of the catalysts, but also to enhance the formation of the hcp Co phase, known to be more active in FTS than the fcc Co phase. Samples containing Co, Ru and Hf were scaled up alongside other samples promoted with Ru and rare earth elements and tested in the FTS under industrial conditions (230 °C, 20 bar). Here the material containing Co, Ru and Hf showed higher CO conversion than other catalysts tested and also exhibited the highest selectivity towards desirable C5+ components. The higher conversion and selectivity of this materials is attributed to the enhanced formation of the hcp Co phase. Further investigation of the impact of Hf in the formation of the different cobalt phases demonstrates that Hf shifts the reduction of CoO to higher temperatures and stabilizes the hcp phase at higher temperatures (500 °C) and also during longer reduction times. The addition of Ru to the samples facilitates the reduction of CoO, such that all samples, even those with higher amounts of Hf (up to 0.08 molar ratio) are reduced at temperatures of 300 °C. The simultaneous addition of Ru and Hf to the samples inhibits the transformation of the hcp to the fcc phase, to a greater extent than is achieved by the addition of Hf alone.

chemsum

{"title": "High-throughput discovery of Hf promotion on the formation of hcp Co and Fischer-Tropsch activity", "journal": "ChemRxiv"}

analysis_of_the_ex-vivo_transformation_of_semen,_saliva_and_urine_as_they_dry_out_using_atr-ftir_spe

## Abstract: The ex-vivo biochemical changes of different body fluids also referred as aging of fluids are potential marker for the estimation of Time since deposition. Infrared spectroscopy has great potential to reveal the biochemical changes in these fluids as previously reported by several researchers. The present study is focused to analyze the spectral changes in the ATR-FTIR spectra of three body fluids, commonly encountered in violent crimes i.e., semen, saliva, and urine as they dry out. The whole analytical timeline is divided into relatively slow phase I due to the major contribution of water and faster Phase II due to significant evaporation of water. Two spectral regions i.e., 3200-3400 cm −1 and 1600-1000 cm −1 are the major contributors to the spectra of these fluids. Several peaks in the spectral region between 1600 and 1000 cm −1 showed highly significant regression equation with a higher coefficient of determination values in Phase II in contrary to the slow passing Phase I. Principal component and Partial Least Square Regression analysis are the two chemometric tool used to estimate the time since deposition of the aforesaid fluids as they dry out. Additionally, this study potentially estimates the time since deposition of an offense from the aging of the body fluids at the early stages after its occurrence as well as works as the precursor for further studies on an extended timeframe. Body fluids are recurrently confronted as major evidence in violent crimes 1 . The potential application of different body fluids typically ranging from their identification to the successful extraction of DNA and its profiling 2 . All of these fluids experience instant biochemical change as they come out from the body. This change can be referred as aging . Some of these changes are rapid and the remaining are gradual 3 , but both the changes are significant to reveal one of the most important aspects of forensic examination; 'time since the deposition' of a crime 4 . The estimation of 'time since deposition (TSD) of a body fluid at the crime scene potentially solves the problem of situating the time of an offense 3 . The TSD of body fluids additionally counter a wide range of issues regarding the crime scene investigation 3,4 . Most of the TSD studies are based on bloodstain aging except for a solitary publication on semen very recently 8 . All these studies except one 9 investigated the TSD for an extended period, escaping the initial changes in the fluid. With the inception of the TSD estimation study in the early twentieth century, various researchers have investigated this phenomenon on bloodstains 3,4 . Since then the techniques used for TSD estimation went through a drastic evolutionary transformation. The analytical technology gradually transformed into nondestructive from destructive methods 3,4,10 . A variety of techniques including Gas chromatography 11 , Liquid chromatography , Oxygen electrode 3,15 , microRNA based assay 3, , Color transformation chart 20 , Electron paramagnetic resonance , Reflectance spectroscopy in Ultraviolet, Visible and Infrared (IR) region , smartphone imaging 27 , IR absorption 1,3,4,8,9,28,29 , Raman scattering 4,6,7,11,30 and fluorescence spectroscopy is explored around the world for the TSD estimation of blood (peripheral and menstrual) 31 and semen 8 . As an age prediction tool, most of these methods realistically depict promising results. Yet, few methodological limitations hinder the universal acceptability of one or more of these techniques as the Table 1. Strong, medium and weak peaks observed in all the three body fluids during their ATR-FTIR spectral analysis with their vibrational modes and spectral assignments. 'Bold' marks denote the peaks changed during the transition from phase I to phase II and 'Bold' + 'Italic' mark denotes the age-linked peaks. ## Body fluid Wave number (cm −1 ) (approx.) Spectral assignment ## Semen O-H stretching (Phase I) and Symmetric N-H stretching of Amide A (Phase II) 37,38,39 spectra in the present study showed similarity with previously reported spectra by several researchers. Although, some minor spectral bands are not visible. It would probably occur due to the spectral acquisition timing, as in this study, all the spectra were collected at the earliest stage of the ex-vivo degradation process up to their drying. The minor bands may have occurred at the later stages in the ex-vivo environment 8, . The spectra in other reported articles were collected at least 4-6 h from the time of deposition 39,40 . Although several peaks were identified in the spectra of each fluid during the study, only a few (age-linked peaks) showed linear changes in their absorption intensity with time. At the initial stage, a similar phenomenon has been observed in the drying of every body fluid. For the first several minutes, only two strong absorption peaks were visible (Fig. 1a-f). These two peaks at 3270-3273 cm −1 (O-H stretching) and 1637 cm −1 (approx.) (scissoring of two H atoms bonded with O molecule) appeared due to the high amount of water in all the fresh fluid samples . Similar results were obtained in the study by Zhang et al., on blood 9 . After a certain amount of time multiple significant peaks corresponding to the biochemical profile of the fluids are revealed throughout the fingerprint region of the IR spectra (Fig. 2a-f). Following the trend in each fluid, the whole drying time was divided into two phases. www.nature.com/scientificreports/ Except for the concentrated samples of fluids, diluted samples of 2:3 ratio (40%) were also prepared to investigate the changes in the drying of the fluids. The dilution was kept constant for all the fluids and a similar extended drying time had been recorded. Multiple dilutions can also alter the drying time as these factors can be studied in future studies on these fluids separately with other factors as experimented by Zhang et al., on blood 9 . The dilution of the fluids showed an extended (2-4 min) phase I due to the excess amount of water in the diluted sample. The longest phase I observed in the Semen samples and the shortest in the saliva samples. Phase II was relatively similar in the spectra of both raw and diluted samples. The duration of phase II of three body fluids was relatively the same (10-12 min). Table 2 demonstrates the minimum, maximum, and mean values of both phases. The difference in the drying time of all three fluids is potentially a result of the qualitative and quantitative variability in their biochemical components. Few researchers reported the correlation between the evaporation of distilled water and time 48,49 . Except for similar height, the peak corresponding to O-H stretching was broader than the peak due to H-O-H scissoring. Except for urine, the absorption intensity of these two peaks showed insignificant change throughout phase I in the spectra of semen and saliva (Figs. 1a-f, 3). On the contrary, Zhang et al. 9 found a different result for blood as the peak at 3308 cm −1 showed very weak but linear absorption change during the early stage. Only the spectra of urine (100% and 40%), showed analogous results with the study by Zhang et al. 9 as the peak at 3273 cm −1 showed a significant decline in the mean absorbance with time during phase I (Figs. 2c,f, 3c). www.nature.com/scientificreports/ Phase II is the fast declination stage where a significant amount of water evaporates rapidly and reveals the other peaks and their intensity changes with time in each body fluid. The peaks at 3271 cm −1 (Amide A) and 1637 cm −1 (Amide I) showed no shift during the whole drying (phases I and II) process of semen stain but the former one sharpens with time and rapidly declined during phase II (Figs. 1a,d, 2a,d) as the peak in Phase II appeared due to the N-H stretching of Amide instead of O-H stretching of water 8, . In the phase II drying of semen droplets, one strong (1546 cm −1 : Amide II), two medium (1446 cm −1 : methylene; CH 2 and CH 3 and 1066 cm −1 : Glycosylated proteins: probably prostate-specific antigen) and three weak (2968 cm −1 : CH 3 stretching, 1396 cm −1 : fatty acids and polysaccharides and 1243 cm −1 : Amide III) significant age-linked peaks were observed (Fig. 2a,d). Zha et al. 8 investigated the changes in few similar peaks at marginally different positions i.e., 1539 cm −1 (Amide II), 1448 cm −1 (Methylene: CH 2 and CH 3 ), 1392 cm −1 (Fatty acids and polysaccharides), 1059 cm −1 (Prostate-specific antigen). Few more researchers reported the IR spectra of semen in several body fluid identification research articles . Phase II spectra of saliva showed the shift of strong peaks at 3273 cm −1 (O-H stretching) to 3286 cm −1 (amide A) and 1637 cm −1 (H-O-H scissoring) to 1645 cm −1 (amide I) that indicated the initiation of this phase (Fig. 2b,e). The shifted peak of amide A sharpens following the trend of semen samples. Among others, one strong and sharp (1546 cm −1 : amide II), four weak (1448 and 1403 cm −1 : Methylene, 1078 cm −1 , and 1043 cm −1 : glycosylated proteins) significant age-linked peaks were found (Fig. 2b,e). Including the peak corresponding to amide II, two peaks at 1448 cm −1 and 1078 cm −1 are similar to the peaks at 1446 and 1066 cm −1 in the spectra of semen and placed at marginally different positions. But the intensity of the peak corresponding to glycosylated protein is relatively weaker in the spectra of saliva. The peaks in phase I spectra of urine bifurcated in phase II. The peak at 3273 cm −1 divided into 3346 cm −1 and 3205 cm −1 and 1637 cm −1 divided into 1623 and 1658 cm −1 (Fig. 2c,f). One strong (1658 cm −1 : amide I), one medium (3346 cm −1 : H-O-H stretching), and 2 weak (1156 cm −1 : urea and 1081 cm −1 : Glycosylated proteins) peaks were observed in the phase II spectra of urine samples (Fig. 2c,f) that significantly changes during the drying process. The peak at 1081 cm −1 is similar to the peaks at 1066 cm −1 and 1078 cm −1 of semen and saliva, respectively. Due to the presence of a significant quantity of prostate-specific antigens in semen, the peak corresponding to glycosylated protein is stronger in its spectra than saliva and urine . In several previous literatures on the IR signature of saliva and urine, the above-mentioned peaks were reported by researchers . Amide A, I, II, and glycosylated proteins are the common biochemical components found in all the 3 body fluids. Elkins, Orphanou, and Takamura et al. previously reported the presence of the common biochemicals in all these body fluids in their article on body fluid identification by ATR-FTIR. In Phase II, all the age-linked peaks of each fluid showed a linear relationship between the mean absorbance at each time point and TSD (Fig. 2a-f). ## Statistical results. The regression equation of a line is the representation of a prediction model. Table 3 depicts the slopes and intercepts calculated for all the age-linked peaks of three body fluids with a 95% level of significance. Body fluids show significant ex-vivo degradation when exposed for a relatively longer period. While early changes are very limited as only a few relevant peaks due to aging are visible. Hence, the TSD estimation was www.nature.com/scientificreports/ performed only on the age-linked peaks of three body fluids. Among several strong and medium peaks 6 (semen), 5 (saliva), and 4 (urine) peaks were selected to calculate the TSD of the body fluids. Irrespective of the variation in the numbers of age-linked peaks for the three body fluids, it was evident that the regression models successfully estimated the TSD with very high accuracy. Both Principal Component Regression (PCR) Analysis and Partial Least Square Regression (PLSR) are strong chemometric tools for the estimation of TSD of body fluids as reported in previous studies 1, . The calculated R 2 values for the calibration and prediction of both models are more than 0.9. While the Root Mean Square Error of Cross-Validation and Prediction (RMSECV and RMSEP) values in both the models for all the three fluids except diluted urine, showed low values (Table 4a,b). Although the initial changes in the IR spectra of body fluids are relatively less distinguishable in comparison to the samples exposed for a longer period, the High R 2 and low RMSE values indicate a good prediction of TSD during this timeframe. In diluted urine samples, the RMSECV (PCRA: 0.7659; PLSR: 0.7622) and RMSEP (PCRA: 0.7383; PLSR: 0.7327) (Table 4a,b) in both the models are relatively higher, that potentially interfere in the accurate age estimation. While, diluted semen also showed higher RMSECV (0.7167) and RMSEP (0.6840) in the PCRA regression model, while PLSR predicts the age for the same condition with significantly higher accuracy with RMSECV and RMSEP values of 0.1897 and 0.1546, (Table 4a,b) respectively. The lower RMSE value also depicts that there is very minute inter-donor variation and low standard deviation values (0.00002-0.00004) of the spectral data obtained from the repeated sampling showed minimal intra-donor variation. Additionally, the RPD (Residual Predictive Deviation) values for each model were also calculated and it has been found that all the values are above 3 which indicates an excellent prediction model accuracy (Fig. 4) 50,51 . Comparatively, PLSR showed better efficiency of prediction than PCRA as for every fluid it records relatively higher R 2 and lower RMSE values than in PCRA. Figures 4 and 5 depicts the PCR and PLS plots of actual vs predicted regression lines for age-linked peaks of three body fluids. Finally, one-way ANOVA has also been applied to the age-linked peaks separately. The F-statistic for all the age-linked peaks showed a significant difference between the time intervals (Table 5). In the present study three forensically significant body fluids other than blood i.e., semen, saliva, and urine were considered to explore the instantaneous changes in their ATR-FTIR spectra up to their drying. This study revealed that all the body fluids undergo significant water loss at the initial stage of their ex-vivo degradation. This rapid loss of water significantly divided the drying process into two phases as the first phase consists of slow evaporation with minimal spectral change with a major contribution of water and the second phase depicts relatively faster evaporation. These two phases can be distinguishable from the ATR-FTIR spectra of each fluid which is a significant marker for estimating the time since deposition of the fluid(s). This study also revealed the spectral regions of interest for the TSD estimation of these fluids as saliva and urine are not explored previously for this purpose. Additionally, if a body fluid is accidentally diluted during the deposition on a wet nonporous substrate during the earliest phase post-deposition, the accuracy of the estimation of TSD can be significantly altered as the water evaporation potentially take more time. In the practical scenario, fresh body fluid samples in liquid and semi-liquid (fluid samples with loss of a certain quantity of water since deposition) conditions from any non-porous surface (e.g., glass, metal, tile, etc.) can be easily collected through a pipette-like apparatus. We can acquire the spectra of the freshly collected fluid with a portable IR instrument containing an ATR-FTIR crystal face exclusively dedicated to forensic crime scene investigation purposes. Although the drying process of body fluids is different on porous substrates like, cloth fabrics, carpets, etc. Hence, further experiments on this subject can be framed based on several factors like, different concentrations, quantities, and interference of porous substrates. Despite, relatively low changes in the IR absorption, chemometric tools like; PCR and PLSR successfully estimate the TSD for each fluid during the initial spectral changes with a very low RMSECV and high R 2 values. Hence, the results (spectral and statistical) of this study potentially be used as a reference for the further TSD studies on these three fluids with a longer timeframe including different factors. ## Materials and methods Sampling. Samples of saliva, semen, and urine were collected from randomly selected eight healthy volunteers (28-40 years). All the methods of this study were carried out in accordance with the World Medical Association Declaration of Helsinki. The experimental protocols were approved by the Ethical Committee of Maharani Laxmi Bai Medical College (4647/IEC/2020/SC-1), Bundelkhand University, Jhansi, Uttar Pradesh, India. All the donors were informed about the nature and procedure of the work and written informed consent was taken from each donor. Each body fluid from a donor was collected separately in a glass test tube(s) just before the spectral analysis to avoid any loss of time after its release from the human body. All the fluids were collected by voluntary secretion, ejaculation, and excretion process without using any invasive technique. Spectra of fluids were obtained in two different concentrations i.e., 100% and 40% (2:3) to investigate the effect of water content on the drying time. The dilution concentration was randomly selected for this study to investigate the difference in drying time between concentrated and dilute fluid samples. 40% solutions of all the fluids were prepared by mixing 2 parts of the fluid and 3 parts of distilled water in a test tube instantly after the collection of the fluid. For experimental purposes, saliva and urine samples were repeatedly taken for three consecutive days, and semen was collected three times with an interval of 3 days to observe any intra-variations of their composition. All the repeated samples were collected from the same individuals. Collection and pretreatment of FTIR spectra. IR spectra of all the body fluid samples were collected by a 'Spectrum Two' FTIR spectrophotometer manufactured by Perkin Elmer corporation, equipped with a 2 mm diameter diamond crystal ATR accessory and spectrum software (Version 10.0). The spectrum software is used for the collection of spectra. The crystal face was cleaned with a 70% methanol solution before drop the fluid sample for each spectral measurement. One drop (approximately 50 µL) of each body fluid was separately added onto the crystal from a constant height of 4 cm. The crystal was used as the drying surface to reduce any loss of water from the fluids during the transfer of the samples from the substrates. Throughout the whole study, the volumes of fluids were kept relatively constant. The experiment was performed in the month of January. The approximate temperature and relative humidity during the complete spectral collection varied between 13 and 17 °C and 56-75%, respectively. As the ambient conditions during the study did not vary significantly, the effect of variable temperature and humidity were not considered in this study. The spectra of each fluid were collected with an interval of two minutes immediately after placing the droplet on the crystal within the range of 4000 cm −1 to 500 cm −1 with 12 scans and a resolution of 4/cm. In this study, the term 'drying out' is the time point at which there was a negligible change between the three consecutive scans taken at a 2-min interval 26 . At this point, the droplet transformed into a dried stain. Before analyzing the obtained data, all the spectra of body fluids were preprocessed by using Unscrambler X software with several spectral corrections i.e., baseline offset, spectral smoothing with Savitzky-Golay algorithm including 13 smoothing points and 3 polynomial orders in a symmetric kernel and range normalization 52,53 . Statistical and chemometric analysis. The mean values of the absorbance of all the significant peaks and the drying time were calculated. Peak identification, fitting, and statistical analysis were carried out by using Origin Pro 2016 and Microsoft excel 2019 software. The correlation coefficient (R) and coefficient of determination (R 2 ) between the TSD and changes in the absorbance values for each body fluid was established with a fitting correlation equation. All the equations showed a 'P' value of < 0.05 which is statistically significant. PCR and PLSR are the two most frequently used tools for prediction model creation and estimation studies. Both of these tools decompose the multiple X variables with respect to the values of Y variables and generate single values for each sample and establish a correlation between X and Y.1 In the present study, the time has been considered as Y and the different age-linked peaks with absorbance values are considered as X. All the chemometric analyses were performed by using 'Unscrambler X' (CAMO Analytics) software. RMSECV and RMSEP were calculated to check the consistency and predictive ability of the regression model. Higher R 2 values and lower RMSE values are indicative of a good prediction model. Full cross-validation was performed for each fluid. Eight samples randomly for each fluid were selected for the model creation and two randomly left out samples were applied for external validation purposes.

chemsum

{"title": "Analysis of the ex-vivo transformation of semen, saliva and urine as they dry out using ATR-FTIR spectroscopy and chemometric approach", "journal": "Scientific Reports - Nature"}

thermochemistry_and_kinetics_of_the_thermal_degradation_of_2-methoxyethanol_as_possible_biofuel_addi

## Abstract: Oxygenated organic compounds derived from biomass (biofuel) are a promising alternative renewable energy resource. Alcohols are widely used as biofuels, but studies on bifunctional alcohols are still limited. This work investigates the unimolecular thermal degradation of 2-methoxyethanol (2ME) using DFT/BMK and ab initio (CBS-QB3 and G3) methods. Enthalpies of the formation of 2ME and its decomposition species have been calculated. Conventional transition state theory has been used to estimate the rate constant of the pyrolysis of 2ME over a temperature range of 298-2000 K. Production of methoxyethene via 1,3-H atom transfer represents the most kinetically favored path in the course of 2ME pyrolysis at room temperature and requires less energy than the weakest C α − C β simple bond fission. Thermodynamically, the most preferred channel is methane and glycoladhyde formation. A ninefold frequency factor gives a superiority of the C α − C β bond breaking over the C γ − o β bond fission despite comparable activation energies of these two processes.Limited energy reserves and global environmental impact of fossil fuel burning became a crucial issue pushing to searching for alternative renewable sources of energy [1][2][3][4][5][6] . Biofuels represent a promising alternative renewable source of energy. Biofuels appear in the energy map of many industrial countries 7,8 . Therefore, a revolution occurred in the forums of the production of biofuels from different biomasses.Among biofuels, the most popular bioethanol suffers from some drawbacks such as low internal energy, water absorption, very high ignition temperature, lower combustion efficiency, and high vapor pressure causing massive emissions to the atmosphere 8-10 giving rise to adverse effects on the human health 11 . In order to avoid most of the above issues, bigger oxygenated materials are preferred. For instance, 2-methoxyethanol (2ME) with bifunctional groups namely etheric (O) and hydroxyl group (OH) is proposed as a model for sizeable molecular biodiesel additive hydroxyethers 12 since it can mimic the behavior of the latter in the combustion process. Furthermore, 2ME is an excellent indirect biofuel candidate due to its original synthesis from small bioalcohols like methanol and ethanol. Besides, it can be obtained by modifying ethylene glycol (EG) itself. Ethylene glycol had recently become available from different biomass categories using various procedures with high yield [13][14][15][16][17][18][19] as a biofuel, but there still some concerns related to its low carbon content, low melting point (−13 °C), high viscosity, high toxicity, and high hydrophilic nature 20 . Those issues can be avoided by using alone in the current engine infrastructure. 2ME could function as a biofuel that might be better than ethanol, ethylene glycol regarding lower vapor pressure, higher boiling point, and high energy content. It also shows high miscibility with oils and gasoline besides the expected enhanced ignition behavior due to its high oxygen content (42.1% per mol). These represent some essential useful properties for 2ME as a good biofuel candidate.2ME has a wide range for applications in industrial and pharmaceutical proposes. For instance, it is used in inks, resins, dyes, paints, metal coatings, phenolic varnishes, detergents, cosmetics, cleaners' products, protective X HF CBS HF bx where X = {T, Q, 5} for aug-cc-pVTZ, aug-cc-pVQZ, and aug-cc-pV5Z, respectively 36 . The extrapolation of the MP2 was obtained by the two-parameter polynomial equation: where X = {Q, 5} for aug-cc-pVQZ and aug-cc-pV5Z, respectively. The transition states for different reactions of 2ME pyrolysis have been located with the aid of the eigenvector-following (EF) optimization technique as implemented in the Gaussian programs. Vibrational analyses have been conducted at BMK/6-31+G(d, p) to characterize the nature of the obtained stationary points whether they are minima or transition states with real frequencies or one imaginary frequency, respectively, and for the zero-point vibrational as well as the thermal corrections of energies at 298 K. Vibrational modes have been analyzed using the Chemcraft program 37 . For further confirmation of correct transition states that connect desired reactants and products, minimum energy paths (MEP) have been computed through intrinsic reaction coordinates (IRC) 38,39 . All electronic structures calculations have been conducted using the Gaussian 09 W suite of programs 40 . The atomization energy approach has been exploited to estimate the gas phase enthalpies of formations for 2ME and its released species at the standard state of temperature and pressure, as it is deduced from the well-known 41 enthalpies of formation of the separated atoms. For any molecule M containing X numbers of isolated atoms, the gas phase enthalpy of formation is obtained from where E e (M) and E e (X i ) are the theoretically calculated electronic energy of molecule M, and the i th atom X at the same level of theory, respectively. ZPVE is the zero-point vibrational energy of the molecule. [H 298 (M) − H 0 (M)] and [H 298 (X i ) − H 0 (X i )] are thermal corrections to the enthalpy for the molecule M and the separated atoms X, respectively. The individual atomic enthalpies ΔH °f(X i ) are extracted from the NIST WebBook 41 . Kinetic parameters for different channels of 2ME pyrolysis have been estimated over a wide range of temperatures using the Kisthelp package program 42 , where the classical transition state theory (TST) 43 is coupled with Eckart tunneling correction 44 to compute rate constants (k) for H-atom transfer reactions of 2ME pyrolysis over the applied range of temperatures (298-2000 K). The rate constant reads: where h, k B , and R symbols are Planck, Boltzmann, and universal ideal gas constants, respectively, and χ T ( ) is the Eckart tunneling correction. T is the system's temperature in Kelvin, σ is reaction path degeneracy, p° is the standard pressure (1 atm), and Δ ° †G T ( ) is the standard Gibbs free energy of activation for reaction. Δn takes two value either zero in the case of unimolecular decomposition or 1 in the case of the bimolecular oxidation. The more accurate correction term Eckart tunneling correction χ(T) which obtained through the numerical integrating probability of transmission ρ(E) over Boltzmann distribution of energies. The asymmetric Eckart tunneling correction gives reliable results at low temperatures and previously demonstrated in many previous publications . The transmission probability coefficient χ(T) can deduce from the following equation: where ρ(E) is the probability of transmission through the one-dimensional barrier at energy E. ΔH f 0K is the zero point correlated energy barriers in the forward direction. Equilibrium relation (K eq = k forward /k reverse ) has been used to calculate the rate constant of simple fission reactions. At first, the equilibrium constant (K eq ) was calculated automatically by the assistance of the Kisthelp program 42 then the previous experimental well-known association rate constants have been used as values for k reverse to get the forward rate constants (k forward ) for the selected simple bond fission reaction. All complex fission reactions barrier heights have been investigated using the more accurate ab initio CBS-QB3, G3, and BMK/6-31+G (d, p). The last level of theory has been proven to have a significant efficiency for the structure optimization in previous works 51,52 . ## Results and Discussion Methoxyethanol conformers. 2ME has 12 conformers. Three of them are illustrated in Fig. 1, and the rest of the optimized structures and energies are presented in the Supporting Information (SI). Several studies on 2ME conformers highlighted the effect of the intramolecular hydrogen bond (IHB) between the alcoholic hydrogen and etheric oxygen on molecular properties . Our findings at CBS-QB3, G3, and BMK/6-31+G(d, p) are in mutual harmony, see Fig. 2. The most stable 2ME conformers adopt tGg − and gGg − structures with IHB . However, tGg − is 1.6 kcal/mol more stable than gGg − . On the other hand, the least stable conformer (gGt), among the studied conformers, is 4.38 kcal/mol higher than tGg − at the CBS-QB3 level of theory. Energy of 2ME conformations: Extrapolation to CCSD(T)/CBS level using FPA. Tables 1 and 2 collect the results of FPA for the most and least stable 2ME conformers, while Table 3 shows a comparison of FPA results at MP2/CBS and CCSD(T)/CBS with our obtained values using ab initio methods (CBS-QB3 and G3) and the DFT/BMK/6-31+G(d,p). The CCSD(T)/CBS energies are 1.43 ± 0.15, 2.47 ± 0.19, 4.11 ± 0.04, and 4.25 ± 0.04 for the conformers gGg-, tTt, gTg, and gGt, respectively. The uncertainty term is obtained using δ [CCSD(T)] ± ∆E CCSD(T) − ∆E CCSD . Comparing the obtained results of the FPA with that of our 2ME conformation analysis using ab initio composite methods and BMK/6-31+G(d,p) level shows harmony as appeared in Table 3, while the convergence of the quantum chemical electron correlations methods at the aug-cc-pVTZ basis set is sketched in Fig. 3. ## Bond dissociation energy. In order to assess the strengths of different bonds in 2ME, their bond dissociation energies have been calculated. Figure 4 displays the bond dissociation energies of 2ME using the CBS-QB3 composite method. The results indicate that the C ɤ −O β and C α −C β are the weakest bonds with bond dissociation energies of 86.2 and 86.7 kcal/mol, respectively. The alcoholic O α -H bond is the strongest one which is close to our previous results (104.5-106.3 kcal/mol) obtained for C1-C4 alcohols 52,57 . The C α -H and C β -H hydrogen atoms are the most acidic and are expected to be abstracted easier in the presence of oxidizing agents as compared to the other hydrogen atoms which agreed with similar bifunctional compound 58 . ## Enthalpies of Formation. Enthalpies of formation for 2ME and its released compounds through combustion have been calculated using atomization approach (at CBS-QB3) and isodesmic equations procedures (at BMK/6-31+G (d, p)). The obtained results are collected in Table 4 using experimental enthalpies of formation values of Table 5. The results have been compared with one another and with available experimental data. The comparison shows impressive agreement with a maximum deviation of ±2 kcal/mol which gives confidence in the future experimental determination of unknown species. The current study concentrates on 2ME pyrolysis. The decomposition mechanism can be expanded into nine complex fissions (barrier reactions) and eight simple bond scission reactions (barrierless reactions). Table 1. The valence focal-point analysis (FPA) of energy differences (kcal/mol) of the most stable 2ME conformers (a) gGg-and (b) tTt. Conformer geometries have been optimized at B3LYP/augcc-pVTZ level. aCCD = aug-cc pVDZ; aCCT = aug-cc-pVTZ; aCCQ = aug-cc-pVQZ; aCC5 = augcc-pV5Z; CBS = complete basis set. The symbol δ denotes the increment in the relative energy concerning the previous level of theory, as given by the competing higher-order correlation series: ## Complex fission reactions Values listed in brackets are taken for extrapolation. Equations ( 1) and ( 2) have been used for extrapolation of HF and MP2 energies to complete the basis set, respectively. Final values (in bold) include core correction. Simple bond fission reactions Complex fission reactions are those reactions proceeding by H-atom transfers via cyclic transition state, while simple bonds fission are those occurring by homolytic cleavage of the chemical bonds. We will concentrate here on that formed due to complex ones. Among nine unimolecular complex reactions, the formation of methoxyethene, methoxy methylcarbene, and oxetane occurs by dehydration (R1, R3, and R8), while 2-methoxy acetaldehyde is formed via hydrogen molecule elimination (R6) reactions. Reaction R5 proceeds via three-membered ring transition state producing ethylene glycol and triplet methylene. The other complex fission reactions R2, R4, R7, and R9 are accomplished by 1,3-H atom transfer reactions via four-membered ring transition state to produce www.nature.com/scientificreports www.nature.com/scientificreports/ methanol and vinyl alcohol, formaldehyde and ethanol, formaldehyde and dimethyl ether, and methane and glycolaldehyde, respectively. The optimized structure of 2ME and proposed transition states leads to the formation of methoxyethene (TS1), vinyl alcohol (TS2), methoxymethyl carbene (TS3), ethanol (TS4), ethylene glycol (TS5), 2-methoxy acetaldehyde (TS6), dimethyl ether (TS7), oxetane (TS8), and glycolaldehyde (TS9) given in Fig. 5. Detailed optimized structures of products and bonds lengths variations versus IRC of complex fission reactions are given in the SI (Figs 1S-9S), while the potential energy diagrams of 2ME pyrolysis at the G3 and CBS-QB3 methods are shown in Fig. 6 and the results at BMK/6-31+G(d,p) are listed in Table 8S. The barrier heights and reaction energies of the main favorable routes at CBS-QB3 and G3 methods are tested against the W1 68 , e ref. 69 , f ref. 70 , g ref. 71 , h ref. 72 , i ref. 73 , j ref. 74 , k ref. 75 , l ref. 76 , m ref. 77 , n ref. 78 , o ref. 79 , p ref. 80 , q ref. 81 , r ref. 82 , s ref. 83 . Energies and IRC analysis. Among complex reactions, two reactions (R3 and R5) proceed via three-membered ring transition state, while the rest is passing over the four-membered ring. Almost all reactions are endothermic so that structures of transition states are close to those of products more than reactants according to Hammond postulate 59 . As a result of the high oxygen content in 2ME (42.1% per molecular weight), a theoretical study on 2ME combustion is essential since many oxygenates like ether, alcohols, and carbonyl compounds can be released to the atmosphere during its ignition. ## Species Formation of ethers compounds. Decomposition of 2ME resembles a platform of many ether categories like methoxyethene, methoxymethyl carbene, dimethyl ether (DME), oxetane, and 2-methoxy acetaldehyde. Methoxyethene formation is the preferable kinetic channel on the potential energy diagram of 2ME decomposition with a barrier height and reaction energy of 72.2 and 6.5 kcal/mol, respectively. The reaction can be accomplished via intra-molecular H-atom abstraction from C β by the alcoholic OH (1,2-water elimination) via the four-membered ring transition state TS1. The selected transition state involves inter-rotational of gauche dihedral angle (OCCO) and anti-gauche dihedral angle (CCOH) to be −108° and 102°, respectively. The IRCs of the methoxyethene formation appear in Fig. 1S. Figure 1S shows a fast rapture for the strong C1-O1 bond (BDE = 97.9 kcal/mol) rather than the weakest C2-H3 bond (BDE = 96.7 kcal/mol). The reason is attributed to the high-frequency factor of the C1-O1 bond compared to that of the C-H bonds (see SI). The broken C2-H3 bond at s = −1 amu 1/2 bohr is associated with the O1-H3 bond formation, where the two curves cross each other at the transition state (s = −0.2 amu 1/2 bohr). Formation of the C1-C2 double bond occurs gently during the reaction. Methoxymethyl carbene is an unstable compound that is obtained by 1,1-water elimination of C α via TS3. The reaction requires a preliminary structure conversion from the most stable conformer tGg-to the tGt structure through multi-steps with final reaction energy of 2.9 kcal/mol. The reaction proceeds via a three-membered ring transition state with barrier energy of 82.5 kcal/mol. The transformation process of 2ME to methoxymethyl carbene is displayed in Fig. 2S. Figure 2S shows a superior rapture for the C1-O1 bond than the C1-H1 one. The disintegration of the C1-O1 bond begins at s = −1.2 amu 1/2 bohr, while the formation of the O1-H1 bond progresses simultaneously with cracking of the C1-H1 bond. The two curves cross each other at s = −0.5 amu 1/2 bohr. The slight decrease in values of the C1-O1 bond length after the cracking is a clue for the formation of an intermediate compound with an H-bond linking separated atoms near each other. Oxetane production has the highest barrier energy value among water elimination reactions from 2ME with a barrier height of 98.3 kcal/mol. The high energy barrier can be attributed to the formation of a highly strained four-membered cyclic product. The reaction proceeds by the alcoholic abstraction of the C ɤ hydrogen (1,4-water elimination) with the four-membered ring transition state TS8. The barrier height and the reaction energy of 98.4 and 12.9 kcal/mol are in line with the work in ref. 60 where the barrier height and the reaction energy were 96.0 and 15.7 kcal/mol, at the same level of theory, for the same investigated channel of 1,4-dehydration of n-butanol. Table 6 shows a comparison between 2ME and n-butanol with respect to 1,1-, 1,2-, and 1,4-H 2 O elimination reactions. Oxetane is formed over multi-conversion processes as the most stable tGg-converts to tGt then to g-Gt conformer by a rotational barrier of 0.6 kcal/mol and reaction energy of 0.4 kcal/mol relative to tGt conformer (2.7 and 1.5 kcal/mol, respectively in case of n-butanol 60 ). Figure 3S illustrates a fast cleavage of the strong C1-O1 bond relative to the weakest C3-H7 bond which occurs at s = 1.5 amu 1/2 bohr. The formation of the O1-H7 bond starts at s = −0.1 amu 1/2 bohr. The two curves of C3-H7 and O1-H7 bonds cross each other at s = 0.9 amu 1/2 bohr, while the formation of the single σ covalent bond C1-C3 occurs gradually during the reaction. DME is produced via TS7. The alcoholic H-atom migrates to C β resulting in DME and formaldehyde. The alcoholic H-atom rotates from the gauche dihedral angle of 51° to 0° for facilitating the conversion process. The change of bond lengths for the formation of DME is shown in Fig. 4S. The Figure shows that the weakest C1-C2 bond (BDE = 86.7 kcal/mol) dissociates earlier (at s = −2 amu 1/2 bohr) than the strong alcoholic O1-H8 bond (BDE = 108.1 kcal/mol) rapture at s = −0.8 amu 1/2 bohr. The C2-H8 bond is formed at s = 1 amu 1/2 bohr and the www.nature.com/scientificreports www.nature.com/scientificreports/ carbonyl C1-O1 bond of formaldehyde is formed smoothly during the reaction. The curves of the O1-H8 and C2-H8 bonds cross each other at the transition state. 2-Methoxyacetaldehyde is a direct result for the 1,2-H 2 elimination from 2ME. The reaction proceeds via the TS6 with a barrier height and reaction energy of 88.9 and 20.9 kcal/mol, respectively. Figure 5S shows a variation of selected bonds lengths during the formation of 2-methoxy acetaldehyde. It is clear that breaking the weak C1-H1 bond (BDE = 96.2 kcal/mol) occurs first and then the alcoholic O1-H8 bond (BDE = 108.1 kcal/mol), while the carbonyl C1-O1 double bond formation progresses smoothly during the reaction. Formation of alcohols and carbonyl compounds. Many alcohols such as methanol, vinyl alcohol, ethanol, glycolaldehyde, and ethylene glycol are released through the combustion of 2ME. Vinyl alcohol production occurs via TS2. It is the 2 nd kinetically preferable pathway with a barrier height difference of 0.6 kcal/mol relative to the most stable methoxyethene transition state TS1. The less stable vinyl alcohol (enol) transforms into the most stable acetaldehyde (keto) (TS11) via the 1,3-intramolecular H atom transfer. The reaction barrier is 55.9 kcal/mol and the reaction energy is 11.54 kcal/mol relative to the vinyl alcohol that agrees with our past recorded data 57 and with alkenol -alkanal conversion using CBS composite methods 61,62 . According to Fig. 6S, the weakest O2-C2 bond (BDE = 86.7 kcal/mol) is broken first (at s = 2 amu 1/2 bohr) then the C1-H1 bond (BDE = 96.2 kcal/mol) stretches slowly until rapture at s = 0.9 amu 1/2 bohr. Fission of the C1-H1 bond and the formation of the alcoholic O2-H8 bond occur at the same time and the two curves cross each other at s = 0.3 amu 1/2 bohr, while the formation of the enolic double bond C1-C2 occurs step by step during the conversion process. EG production is the highest endothermic route among all H-atom transfer channels with reaction energy of 90.6 kcal/mol. The reaction proceeds by 1,2-H-atom transfer via TS5 as one of the C γ hydrogen migrates to the O β via a strained three-membered ring transition state. The high recorded reaction energy may be related to the formation of the less stable triplet methylene. The investigation related to the IRC in Fig. 7S indicates a fast breakage of the O2-C3 bond (BDE = 86.1 kcal/mol) at s = 2 amu 1/2 bohr, while the C3-H6 bond stretches and breaks at s = 1.5 amu 1/2 bohr with the formation of the O2-H6 bond. The two curves interrupted at s = 0.8 amu 1/2 bohr. Similar to the methoxymethyl carbene, the variational of the O2-C3 bond length is a clue for the formation of the H-bond which makes the two separated atoms close to each other after the product formation. Ethanol is produced via TS4 with an energy barrier of 86.2 kcal/mol and reaction energy of 6.7 kcal/mol. The reaction occurs by shifting one of the C γ hydrogens to the C β passing over the etheric oxygen O β . Figure 8S reveals that the O2-C2 bond breaks before the C3-H7 bond, which agrees with the bond dissociation values of the two bonds, while the O2-C3 double bond forms slowly during the reaction. Thermodynamically, ethanol formation is preferable than methoxyethene production by 0.2 kcal/mol. Glycolaldehyde is also a bifunctional compound that has alcohol and aldehyde groups. It is formed through TS9 which is the highest energy barrier among all complex channels (100.4 kcal/mol). However, it is the preferable thermodynamic pathway with reaction energy of −1.2 kcal/mol. Figure 9S in the SI shows the earlier rapture of the least energy O2-C3 bond (BDE = 86.2 kcal/mol), while the C2-H4 bond (BDE = 96.7 kcal/mol) stretches gently till it gets broken at s = −0.6 amu 1/2 bohr. Formation of the C3-H4 bond occurs at s = 1.5 amu 1/2 bohr. The two curves of C3-H4 and C2-H4 bonds cross each other at s = 0.7 amu 1/2 bohr, while the O2-C2 double bond is formed gently during the reaction. For liner relations between ln k vs. 1000/ T for reactions R10, R11, and R12, the activation energy and pre-exponential factor can be derived from the two-parameter Arrhenius equation: ## Rate constant calculation. Taking the Natural Logarithm of the two sides ## Plotting ln k TST (T) versus T 1000 shows a straight line with a frequency factor A (s −1 ) = e Intercept and an activation energy Δ = × . † E (cal/mol) slope 1 987. In Fig. 7, the tunneling correction calculated by Eckert method plays a vital role for the curvature of the relation between ln k vs. 1000/T for R1 and R2 reactions at T ≤ 500 K. Therefore, these reactions can fit the In the case of three-parameter Arrhenius equation, the following equation is used: The equation converts to the general form We will get another two equations of the three variables A, n, and Δ † E . The algebraic solution of the three Eqs (3), ( 4) and ( 5) gives values of A, n, and Δ † E . Arrhenius equations for the calculated rate constant (s −1 ) for main channels R1, R2, R10, R11, and R12 in the temperature range 298-2000 K can be summarized as follow:

chemsum

{"title": "Thermochemistry and Kinetics of the Thermal Degradation of 2-Methoxyethanol as Possible Biofuel Additives", "journal": "Scientific Reports - Nature"}

occurrence_of_the_potent_mutagens_2-_nitrobenzanthrone_and_3-nitrobenzanthrone_in_fine_airborne_part

## Abstract: Polycyclic aromatic compounds (PACs) are known due to their mutagenic activity. Among them, 2-nitrobenzanthrone (2-NBA) and 3-nitrobenzanthrone (3-NBA) are considered as two of the most potent mutagens found in atmospheric particles. In the present study 2-NBA, 3-NBA and selected PAHs and Nitro-PAHs were determined in fine particle samples (PM 2.5) collected in a bus station and an outdoor site. The fuel used by buses was a diesel-biodiesel (96:4) blend and light-duty vehicles run with any ethanol-to-gasoline proportion. The concentrations of 2-NBA and 3-NBA were, on average, under 14.8 µg g −1 and 4.39 µg g −1 , respectively. In order to access the main sources and formation routes of these compounds, we performed ternary correlations and multivariate statistical analyses. The main sources for the studied compounds in the bus station were diesel/biodiesel exhaust followed by floor resuspension. In the coastal site, vehicular emission, photochemical formation and wood combustion were the main sources for 2-NBA and 3-NBA as well as the other PACs. Incremental lifetime cancer risk (ILCR) were calculated for both places, which presented low values, showing low cancer risk incidence although the ILCR values for the bus station were around 2.5 times higher than the ILCR from the coastal site.Cancer is one of the major causes of morbidity and mortality globally. In 2012 new cancer cases accounted for about 14 million new cases, with 8.2 million deaths occurred throughout the world. From that, circa 1.69 million deaths in 2012 resulted from lung cancer. However, only less than one third of them were derived from tobacco smoke 1,2 , what indicates there are other routes contributing to lung cancer incidence. Additionally, it is expected the number of new cancer diagnoses to be risen by about 70% over the next two decades, possibly reaching 21.7 million people, and the prediction of 13 million cancer deaths in 2030 2,3 .Most cancer results from the interaction of genetics and the environment. However, hereditary or genetic factors themselves only respond for less than 10% of all types of cancers 4,5 . The remainder is attributed to environmental factors, and among them physical, chemical or biological toxicants, as well as individual susceptibility 4,6 acting to explain the large cancer incidence worldwide. Human environmental and occupational exposure to atmospheric pollutants may be one of the major causes of lung cancer since the main pathway to atmospheric carcinogenic exposition is through inhalation [6][7][8] . It is well known that energy is the single most important cause of emissions of all main pollutants, and air pollution is an energy problem 9 . Carcinogenic and/or mutagenic compounds occurring in vapor phase and atmospheric aerosols, such as unsubstituted polycyclic aromatic hydrocarbons (PAHs) and their nitrated and oxygenated derivatives (nitro-PAHs and oxy-PAHs, respectively) are of major concern in regard to the potential risk of causing cancer.Nitro-aromatic polycyclic hydrocarbons (nitro-PAHs) are ubiquitous airborne particle contaminants, mainly originated from incomplete combustion or pyrolysis of organic matter 10,11 and/or photochemically-formed in atmosphere 12 . Nitro-PAHs are persistent compounds 13 generally regarded as direct-acting carcinogenic and/or mutagenic agents to humans or animals . Even though the nitro-PAH levels are typically one order of magnitude smaller than their unsubstituted congeners in temperate or remote regions, some of their members present high direct-acting mutagenic and/or carcinogenic potency in bacterial and mammalian cells 10,13 . Representatives may be cited, such as mono-and dinitropyrenes, nitrofluoranthenes and nitroketones 18,19 . Although the understanding of the exact mechanisms of cancer incidence derived from atmospheric aerosols remains mostly uncertain 7,20 , it may be pointed out the nitroketone species 2-nitrobenzanthrone (2-NBA) and 3-nitrobenzanthrone (3-NBA) may be important contributors. They are ubiquously present in atmospheric particle samples as well as it has been reported evidences they contribute to the induction of tumors in animal models . 3-NBA is a potent bacterial mutagen generally found in diesel and gasoline directly-exhausted particles. The 3-NBA carcinogenicity is comparable to 1,8-dinitropyrene, which is one of the most potent mutagens 19, . Indeed, 3-NBA is likely to form adducts to DNA molecule 19,24,25,28 augmenting its genotoxic potential in living beings. In turn, the isomer 2-NBA is rather an ambient PM contaminant which is likely to be produced from the reaction of its precursor (benzanthrone, BA) with nitrogen oxides or other oxidants under typical atmospheric conditions . Despite the fact 2-NBA has been more abundantly found in airborne samples, screening assays studies suggest the genotoxic potency of 2-NBA is significantly lower than 3-NBA . Both of them are lipophilic substances (K ow = 3.99 for 2-NBA and K ow = 3.90 for 3-NBA) yet they may be considered persistent in the environment 30 and may be transferred from atmosphere to other environmental compartments by wet or dry deposition. Although their inherent relevance, there is limited information about the occurrence of 2-NBA and 3-NBA in aerosol particles. 2-NBA and 3-NBA have been unevenly and eventually identified in diesel exhaust and ambient air PM worldwide, although they have remained predominantly underdetermined. In part, the reason for finding little information about particle-bound 2-NBA and 3-NBA are their very low concentration levels in atmospheric aerosols (from low ng m −3 to low fg m −3 ), ranging from 0.5-3.5 f mol m −3 (or 0.14-0.96 pg m −3 ) (3-NBA) in Central Tokyo 22 to 6.79 pg m −3 (3-NBA) in other parts of Japan 37 . Even fewer studies have considered 2-NBA. The reported concentration range of 2-NBA in ambient PM is around 49-831 f mol m −3 (or 13.5-229 pg m −3 ) 38 . Their low atmospheric concentrations together to the complex nature of particulate matter demand reliable and efficient sample preparation and analysis methods in order to be able to confidently quantify 2-NBA and 3-NBA in atmospheric samples. Nonetheless, the latest studies regarding 2-NBA and 3-NBA occurrence in atmospheric aerosols and rainwater and possible atmospheric chemistry implications are dated from late 1990s and early and mid 2000s only 22,26,27,30,32,37,39,40 . After that very little has been done in this subject. Consequently, the implications for atmospheric chemistry and their health-related endpoints are underestimated. More studies regarding the 2-NBA and 3-NBA in airborne particles are needed for better understand their role in these fields. In the present study, we determined 2-NBA and 3-NBA in order to study the atmospheric occurrence of these species in ambient PM2.5 samples collected from a coastal tropical site in Northeastern Brazil as well as in samples collected in an underground level of a bus station, where buses exhausted mixtures of biodiesel to fossil diesel (B4) combustion during commuting. Together to that we also report some PAH levels in order to help trace atmospheric sources, which may be contributing to the found levels of 2-NBA and 3-NBA in our study. To date, this is the first time 2-NBA and 3-NBA levels are reported in the Southern Hemisphere. Cancer risk and mutagenic risk assessments from inhalation exposure were also calculated. Results are conveniently presented and critically discussed. ## Results and Discussion Analysis and identification of 2-NBA and 3-NBA. 2-NBA and 3-NBA have been poorly determined in atmospheric aerosols. Their low atmospheric concentrations together to the complex nature of particulate matter demand reliable and efficient sample preparation and analysis methods in order to be able to confidently quantify 2-NBA and 3-NBA in atmospheric samples. In the same way, previous studies regarding 3-NBA and isomers were mostly determined via a derivatization step prior their analysis. Generally, the 2-NBA and 3-NBA derivatization is done through a reduction step which yields 2-aminobenzanthrone and 3-aminobenzanthrone (2-ABA and 3-ABA), respectively. They are further mostly analyzed by HPLC coupled with either fluorescence 22,41,42 , UV 22,39, , chemiluminescence 37,38,44 and/or mass spectrometer 41 detectors. If we consider possible losses or degradation as well as any artifact formations during 2-NBA and 3-NBA derivatization step associated to their very low atmospheric levels all together also may partially answer for the difficulty of finding them in appreciable levels in the atmospheric environment. Recently, Santos et al. 45 , reported, in a companion paper, a novel miniaturized method for the efficient determination of polycyclic aromatic compounds, and among them 3-NBA, by GC-MS with no derivatization or fractionation steps needed. Details about the sample preparation method is described in the Supplementary Information. In the GC-MS system used in the present study 3-NBA is eluted just before 2-NBA, and both of them with retention times between 27.50 min and 27.75 min, as stated in Fig. 1. In this figure, we show two chromatograms of real samples, being Fig. 1a a chromatogram of a real PM2.5 sample from the bus station as well as Fig. 1b being a chromatogram of a real PM2.5 sample from the coastal site considered here. Accordingly, Fig. 2 shows the mass spectra of both 2-NBA and 3-NBA obtained in this study, which are in accordance with Phousongphoung and Arey 27 . As stated in Santos et al. 45 , every analyte was monitored in SIM mode by three different m/z ions, with the addition of 2-NBA in the present study. Indeed, when our group published the results from Santos et al. 45 , 2-NBA was not listed as an analyte since we did not have its authentic analytical grade standard at that time. Nonetheless, in the current study we have added the 2-NBA to our mixed analytical standard solution and we could also quantify it in real samples without any modification of the GC-MS method. In order to distinguish 2-NBA from 3-NBA, we monitored three m/z ions, the ion base and two reference ions, in order to approach unequivocal identification. In this way, 2-NBA was identified (and differentiated from 3-NBA) by using m/z 275 (ion base) and m/z 201 and 245 (reference ions), as shown in detail in Fig. 1. In turn, 3-NBA was identified by m/z 275 (ion base) and m/z 215 and 245 (reference ions). Quantification was further done by considering the ion base signal only. Limits of detection (LOD) and limits of quantification (LOQ) were calculated from the calibration curve data. We considered LOD = 3 s/a and LOQ = 10 s/a, where "s" is the standard deviation of the linear coefficient "b" and "a" is the angular coefficient (inclination) from calibration curve (in the format y = ax + b) 45 . LOD and LOQ concentrations values were converted to the minimum absolute mass either detected (LOD) or quantified (LOQ) by the MS in 1 µL standard solution injected in the GC-MS. LOD in terms of absolute mass, were 2.0 pg and 2.4 pg for 2-NBA and 3-NBA, respectively. Limit of quantification (LOQ) were 6.6 pg (2-NBA) and 8.1 pg (3-NBA). Finally, recovery levels were above 95% for both compounds. In this way, we consider our analytical methodology is adequate for studying 2-NBA and 3-NBA in the atmospheric environment. Occurrence of 2-NBA and 3-NBA associated to fine particles. Both 2-NBA and 3-NBA were found in PM2.5 samples collected in the bus station and coastal site (Table 1 and Fig. 3). 3-NBA and 2-NBA concentrations (±one standard deviation) were calculated as pg m −3 (picograms per cubic meter) and as mixing ratios, in terms of µg g −1 (micrograms per grams of particles). We decided to do in this way since the former considers the total sampled volume air (or a normalization of the mass of 3-NBA or 2-NBA by total sampled volume air) while the latter consider the compounds masses normalized by the collected PM2.5 masses. We consider the measurement of 3-NBA and 2-NBA in terms of µg g −1 more useful for some of the discussions done here since it better represents the intrinsic or inherent characteristics of the PM considered in relation to 2-NBA and/or 3-NBA and makes sites with different levels of particle mass concentrations (due to different emission rates among sources) more directly and easily comparable. This is also advantageous to use this type of concentration unit when trying to address toxicological responses from those species possibly present in PM. Accordingly, we still present our results in terms of pg m −3 since it is needed when comparing our study with other reports found in the literature (Supplementary Information (SI) Table S1) and for risk assessment calculations (ILCR). In this study 3-NBA concentrations were 431 (±183) pg m −3 in the bus station and 59.0 (±16.6) pg m −3 in the coastal site. In turn, 2-NBA concentration was 200 (±18.8) pg m −3 , although it was found in the coastal site samples only. Indeed, this is consistent with the fact 2-NBA is majorly formed photochemically and, therefore, it would not be found in the bus station samples (since this is a nearly indoor site mainly impacted by direct, freshly emitted vehicular particles). On the other hand, 3-NBA was found in both places, which is mainly derived from fuel burning/vehicular fleet present in these sites. ## Frequency of detection (%) mass conc. (µg m −3 ) 91. The present study focuses on the determination of 2-NBA and 3-NBA in PM2.5 samples together to a better understanding of their atmospheric and human health implications. Other related and important compounds (such as 1-NPYR, 2-NPYR, 2-NFLT, 3-NFLT, FLT, PYR, BaA, and BaP) are also considered in order to support discussion and possible conclusions in regard to PAH and Nitro-PAH reactivities and source identification. Our findings show the 2-NBA levels is only about 3.4 times higher than the 3-NBA levels (Table 1). However, the few studies reported both 2-NBA and 3-NBA levels showed 2-NBA is ~35-70 higher than 3-NBA (Table S1) 22 . In our study while 2-NBA concentrations are similar to the literature, 3-NBA levels seem to be higher. This may be happening for some reasons. Considering 3-NBA is mostly emitted during vehicle fuel burning, this may be a consequence from the differences in the Brazilian fuel composition with other places in the world. Indeed, the Brazilian gasoline is 22-26% ethanol, and diesel fuel was actually a mix of 4% (v v −1 ) biodiesel into mineral diesel. Yet light-duty vehicles in Brazil has been set from the manufacturer to run with any ethanol-to-gasoline proportion. So, the 3-NBA emission is likely to differ from any other country in the world. How much is the 3-NBA emission rate from different fuel compositions should be better addressed in future studies. Another point is that previous 2-NBA or 3-NBA studies were done mainly in temperate regions, with climate conditions completely different from our sampling places. In the tropics, the higher ambient temperature and more effective sun incidence could either favor 2-NBA photochemical production or its degradation through photolysis. All those points stated here could be plausible reasons in our study the ratio 2-NBA/3-NBA is lower than seen in literature. Since there are quite a few reported studies regarding both 2-NBA and 3-NBA these differences may also be due to limited understanding about their role in the atmosphere. To our best knowledge, this study is the first to find 2-NBA and 3-NBA in a tropical area around a coast, and also adjacent to a large city (Salvador city and Metropolitan area) as well as to find 3-NBA emitted in the exhaust of fossil diesel-biodiesel mixes (B4) under real conditions. Yet, this is also the first study to investigate these species in fine particles, which brings more direct concern in regard to health-related endpoints, than other reports because they considered larger PM fractions only (such as PM10 and TSP). In this way, our results (Table S1) are not directly comparable with the previous ambient 22,32,38,39,42,43, , nor chamber studies 26,27 , especially concerned to particle size, different concentrations units reported, the lack of information regarding sampling data and the broad sort of analytical methods used. Some selected PAHs and nitro-PAHs levels are stated in Table 1 and Fig. 3 in order to be able to better discuss possible nitro-PAHs photochemical routes and trace 2-NBA and 3-NBA main sources. 1-NPYR, 2-NPYR, 2-NFLT, 3-NFLT, and 3-NBA concentrations determined within the bus station (Table 1 and Fig. 3) represent the primary emission while the 1-NPYR, 2-NPYR, 2-NFLT, 3-NFLT, 2-NBA and 3-NBA concentrations determined in the coastal area represent both the primary and secondary contributions. They may also depend on the concentrations of PAH precursors in this site. Our reasoning in this study in using two different sites, a bus station and a coastal site, was to investigate the emission profiles (which and how much) selected polyaromatics are emitted to atmosphere. In this study, we chose to collect PM2.5 samples in the underground floor of a bus station in order to better understand 3-NBA and related species emissions from diesel/biodiesel burning under real conditions (as opposed to sometimes unrealistic dynamometer studies). On the other hand, in the coastal area, which is affected by different sources including diesel/biodiesel burning, our goal was to evaluate the relative contribution among different sources to the found PAHs and nitro-PAHs levels in PM2.5. Possible differences in the PM2.5 emission rates from the bus station to the coastal site are minimized when using concentrations in terms of µg g −1 , which consider the polyaromatic masses (in µg) normalized by the collected PM2.5 masses (in g). Thereby, the observed differences in the PAHs and nitro-PAHs levels between the sites (Table 1) are mainly derived from different sources relative contributions in each case. The concentrations of FLT and PYR observed in the present study, in the bus station (19.6 ± 11.0 and 37.3 ± 19.7 μg g −1 , respectively) are higher than the concentrations of 2-NFLT (2.47 ± 0.93 μg g −1 ) and 3-NFLT (11.8 ± 3.95 μg g −1 ) as well as 1-NPYR (7.81 ± 2.04 μg g −1 ) and 2-NPYR (2.82 ± 0.58 μg g −1 ). However, 2-NFLT and 2-NPYR concentration levels in the bus station are considered estimates only since they were found in 36% of the samples (which is considered a low detection frequency). Considering 2-NFLT and 2-NPYR are mainly produced photochemically and the samples from the bus station were collected in the underground floor with absence of light, they were unevenly found in this site. In this way, 2-NFLT and 2-NPYR levels from the bus station is no further considered. Hence, the nitro-PAHs (and also PAHs) determined inside the bus station are not associated to the atmospheric reactions. Indeed, our results from the bus station show both PAHs and nitro-PAHs studied here are directly emitted by fossil diesel/biodiesel burning. On the other hand, the results determined in the coastal area showed an opposite trend, the nitro-PAH levels, [2-NFLT (7.06 ± 2.08 µg g −1 ) and 2-NPYR (11.7 ± 4.14 µg g −1 )], are higher than the PAH precursors [FLT (9.24 ± 5.77 µg g −1 ) and PYR (7.71 ± 4.27 µg g −1 )]. These results are in good agreement with PAHs reactivity since PYR is less reactive than FLT in nitration reactions 15,16, . Also, as can be seen by Fig. S1, air masses arriving the sampling site were subjected to long-range transport (oceanic origin). In fact, during the long-range transport probably PAH-rich air masses were gradually being transformed to their respective nitro-PAH congeners. Indeed, 2-NFLT is known to be formed via reaction between FLT and NO 3 or OH radicals while 2-NPYR is the product of the reaction between PYR and OH radicals only 17,22,37, . This route seems to be important for tropical and warm areas, where OH radical formation is likely to be enhanced due to higher sunlight incidence. This implies in the coastal area there are important enrichment of 2-NFLT and 2-NPYR, both photochemically produced, in relation to directly-emitted FLT and PYR by automobiles, as observed when compared with their concentrations found in the bus station. On the other hand, 2-NBA formation is mainly done via heterogeneous reaction between benzanthrone (not measured here) and NO 3 or OH radicals on preexisting particles. According to Abbas et al. 54 , there are three different processes which may lead to nitro-PAHs formation from the parent PAHs. They can be formed by (a) electrophilic nitration within the combustion process (e.g. in the exhaust of diesel or gasoline vehicle engines, wood burning and cooking), (b) by gas-phase reactions with atmospheric oxidants (as a secondary and homogeneous process), and (c) by heterogeneous oxidation of particle-bound PAHs (also a secondary process). For instance, 3-NBA is mainly formed by the mechanism described by (a). Only parent PAHs with less than 4 benzene rings and high vapor pressure would be substantially in gas phase in order to react with oxidants in typical atmospheric conditions to form their respective nitro-PAHs, as in (b). Considering benzanthrone vapor pressure is low and boiling point is high (2.21 × 10 −7 mm Hg and 403 °C, respectively) and log K oa is high (10.378) (Table S2), this species would be principally present in PM rather than gas phase under typical atmospheric conditions. In this way, we argue 2-NBA main formation route would be via heterogeneous reactions. However, the 2-NBA formation through heterogeneous reaction still needs to be addressed in future studies. In terms of source tracing, it is well accepted 2-NBA, 2-NFLT, and 2-NPYR are predominantly generated via photochemical reactions in the atmosphere while 3-NBA, 3-NFLT, and 1-NPYR is mainly emitted by diesel combustion/vehicles 17,22,37,53,54 . Here again, it implies photochemical reactions are important to explain the nitro-PAHs and PAHs atmospheric levels in the coastal area. ## Multivariate Analysis. Figure 4 shows ternary correlations, which are useful for improving source identification. For 1-NPYR, 3-NFLT, and 3-NBA we found high correlation (r = 0.8643, p = 0.0322) for the bus station and (r = 0.8303, p = 0.0538) for the coastal area. Since they are mainly directly emitted by diesel combustion, this high correlation implies this source is relevant for them in both sites. In turn, for 2-NPYR, 2-NFLT, and 2-NBA we also see high correlation (r = 0.8652, p = 0.0317) for coastal area, which demonstrates photochemistry actually is an important source for this ambient site. Although in the second case the p-value is little higher than 0.05, indicating a statistical confidence slightly lower than 95%, we should keep in mind the r-values are high enough to be statistically valid when considering ternary correlations. Principal component analysis (PCA) were run for both sites (Fig. 5 and SI Table S3). For the bus station PCA explains 66.9% of the whole dataset variance. PC1 accounted for 34.7% of the variance, and it had high positive loadings for particle mass and mass concentration as well as high negative loading for 3-NFLT, and moderate negative loadings for 3-NBA, BaP, and 1-NPYR. In turn, PC2 explained 32.2% of the variance, with high positive loadings for FLT, PYR and BaA. PC1 seems to indicate diesel-biodiesel exhaust direct-emitted as source while PC2, which is represented by less reactive species, then they are able to be constituents of particles direct-emitted by buses and deposited on the floor and be suspended again, which may indicate particle aging and/or particle size growing processes happening in this site. For the coastal site, it was necessary to consider three principal components to explain 82.6% of the total variance. PC1 explains 54.1%, with high negative loadings for particle mass and mass concentration as well as high positive loadings for most of the nitro-PAHs (2-NBA, 2-NFLT, 3-NFLT, 1-NPYR, and 2-NPYR), except for 3-NBA represented in the PC2 (which accounted for 14.9% of the variance), together to FLT and PYR. PC3 accounted for 13.6% of the variance and presents high positive loadings for BaP and BaA. For this site, PC1 represents the photochemical origin (as important secondary source tracers, such as 2-NBA, 2-NPYR and 2-NFLT are better explained in this PC). Although in PC1 3-NFLT and 1-NPYR, which are direct-emitted species, are also presented, it seems through particle aging, some isomers may interconvert into the other (2-NFLT to 3-NFLT and 1-NPYR to 2-NPYR or vice-versa) forming complex interactions. PC2 may indicate direct emission of fuel combustion as probable source since 3-NBA, PYR, and FLT are presented by this PC. Finally, PC3, which has high scores for BaA and BaP, is likely to represent wood combustion as sources. Agglomerative hierarchical clustering (AHC) plots (Fig. 6a,b) for the bus station does not show statistically significant dissimilarities for particles collected in the morning, afternoon or night periods nor among species (PAHs and nitro-PAHs). This is due to the fact the bus station does have few different sources contributing to the levels of PAHs and nitro-PAHs in fine particles there, as suggested by the PCA study. On the other hand, for the coastal site (Fig. 6c,d), in the AHC plots there are 3 clusters discriminated. The first cluster is formed by mass concentration, 2-NBA, 2-NPYR, FLT, PYR, BaA, 3-NBA, and 2-NFLT. In turn, the second cluster is only formed by BaP, and the last cluster includes particle mass, 3-NFLT, and 1-NPYR. The first cluster represents the parent PAHs and their nitro derivatives, while the second one is tentatively attributed to wood combustion. Finally, the last cluster is representative of primary emissions. ## Cancer risk from inhalation exposure. Estimates of incremental lifetime cancer risk (ILCR) according to four different age groups, namely infants (<1 year), children (1-11 years), adolescents (11-16 years), and adults (>21 years) in the population are summarized in Table 2. BaPeq considering the carcinogenicity contribution are 1.21 and 0.45 ng m −3 as well as BaPeq due to mutagenicity 0.90 and 0.38 ng m −3 for the bus station and coastal site, respectively. Despite the fact there are several studies showing 2-NBA and 3-NBA high levels of carcinogenicity and/or mutagenicity 11,13,18,21, 29,30,36,55,56 are similar to those ones of dinitropyrenes (which, in turn, are considered the most potent carcinogens and mutagens), there are no TEFs and MEFs for the nitrobenzanthrone isomers. At this point, knowing the 2-NBA and 3-NBA carcinogenicity and mutagenicity potencies and not considering them in the ILCR calculations would be a not acceptable underestimation for such relevant target species for human health. This would be interesting to account the nitrobenzanthrones in the ILCR since representing the air toxicity by considering only the 16 priority PAHs is not enough anymore in face to the recent advances in this area and the scientific community recent discussions 56,57 . Even though knowing this is not ideal, we have decided to tentatively use the 1,3-dinitropyrene (1,3-DNPYR) TEF and MEF values as surrogates for 2-NBA as well as 1,8-dinitropyrene (1,8-DNPYR) TEF and MEF as surrogates for 3-NBA (SI Table S4). In our point of view, this is plausible because the 1,3-DNPYR and 1,8-DNPYR carcinogenicity and mutagenicity in assays using mammalian cells, bacteria and rodents are close to the ones from 2-NBA and 3-NBA in the same screening assays 11,19,21,22, . After that we were able to calculate more realistic BaPeq values (Table 2). In this way, even though we are aware this may be just a proxy to their real values, we propose to adopt 1,8-DNPYR and 1,3-DNPYR TEF and MEF values as 2-NBA and 3-NBA TEF and MEF, respectively. This could be a temporary alternative while studies about evaluation of the real TEF and MEF for the nitrobenzanthrone isomers are not directly measured. Bus station total daily inhalation exposure (E I ) (considering both carcinogenicity and mutagenic contributions), in ng person −1 day −1 according to different age groups, were 14.4, 28.1, 46.3, and 34.7 for infants, children, adolescents and adults, respectively. Total ILCR was 9.48 × 10 −8 (infants), 5.24 × 10 −7 (children), 2.17 × 10 −7 (adolescents), and 9.73 × 10 −7 (adults). These ILCR estimates mean there is a risk of 9.48 infants in a hundred million and ranges up to 9.73 adults in ten million commuting this bus station to develop cancer during their lifetime. In the coastal site total E I were 5.66, 11.1, 18.2, and 13.6 ng person −1 day −1 for infants, children, adolescents, and adults respectively. Total ILCR were 3.73 × 10 −8 (infants), 2.06 × 10 −7 (children), 8.64 × 10 −8 (adolescents), and 3.83 × 10 −7 (adults). Coastal site ILCR is about 2.5 times shorter than the bus station ILCR. In the same way, ILCR estimates for this site means there is a chance of 3.73 infants in a hundred million and ranges up to 3.83 adults in a ten million that may get cancer during their 70 years of lifetime. If we compare these ILCR to other recent studies 15,16,23, our estimates could be considered one or two orders of magnitude lower, which primarily may not raise much concern. This is needed further studies in order to obtain more comprehensive ILCR estimates. On the other hand, it should be kept in mind these estimates only address ILCR from about ten polycyclic compounds in fine particles considered in this study. This is well accepted there are thousands of other compounds constituting fine particulate matter and we actually have investigated the large majority of them and much less is known about their toxicity in regard to carcinogenicity and mutagenicity. ## Concluding Remarks Selected fine particulate PAHs and nitro-PAHs were characterized under typical conditions from a bus station and a coastal site. Among nitro-PAHs, 2-NBA and 3-NBA, which are potent carcinogens and mutagens were determined for the first time in the Southern Hemisphere. The main sources for the studied compounds in the bus station were mineral diesel/biodiesel exhaust followed by floor resuspension (which contributed to the particle growing and ageing). In the coastal site, vehicular emission, photochemical formation and wood combustion were the main sources for 2-NBA and 3-NBA as well as the other polycyclic aromatic compounds. Incremental lifetime cancer risk (ILCR) were calculated for both places, which presented low values, showing low cancer risk incidence although the ILCR values for the bus station were around 2.5 times higher than the ILCR from the coastal site. ## Instrumentation and analysis. In this study, we used the chromatographic conditions stated in Santos et al. 45 Briefly describing, we utilized a high-resolution gas chromatograph-high-resolution mass spectrometer detector (HRGC-HRMS) from Shimadzu (GCMS-QP2010Plus, Shimadzu, Japan) with a Rtx-5MS gas capillary column (30 m × 0.250 mm × 0.25 µm, Restek Bellofonte, USA). Oven temperature programing initiated at 70 °C (2 min), then rising from 70-200 °C (30 °C min −1 , 5 min), and 200-330 °C (5 °C min −1 , 0.67 min). Injector temperature was set at 310 °C and transfer line was 280 °C. Analysis was done in GC-MS-SIM, at electron impact mode (EI) (70 eV). Sample preparation was done using a filter piece of 4.15 cm 2 diameter added to a miniaturized micro-extraction device using 500 µL solvent extraction 45,61 . Sample preparation details are found in Supplementary Information. Sample collection. PM2.5 samples were collected in two different sites: (i) in the underground floor of a bus station (12°58′S, 38°30′W, 52 m altitude), and (ii) in a coastal area (12°58′S, 38°30′W, 52 m altitude) in Northeastern Brazil . The PM samples collected in the bus station were mostly subjected to the exhausts from buses, which remained on (in idle point) while waiting for passengers. No substantial additional sources have contributed to the found levels of polycyclics since the underground level of this bus station is a nearly indoor environment. Yet, no air dispersion system was present. This bus station has been previously studied by our research group 45,61,62,64 and our findings show PM sample sources mainly are biodiesel/diesel burning released by buses and dust resuspension. During the sampling period, buses were using the B4 mix as fuel (4% v v −1 biodiesel to fossil diesel, as established by law at that time). In turn, PM2.5 samples were collected around the Todos os Santos Bay, in the Brazilian Navy Base, located in Salvador Metropolitan Area, State of Bahia, Northeastern Brazil. This site is close to industries and the Aratu Harbor and it is also influenced by vehicular fleet from Salvador City and surroundings 62,64 . During sample collection, ambient temperature ranged from 22.5 to 24.9 °C, relative humidity was 72-84%, solar radiation was 145-340 W m −2 , and wind speed ranged 3.9-8.0 m s −1 (Mkoma et al. 62 ). Backward air mass trajectories were calculated starting 48 h before arrival time (00:00 UTC) and 500 m a.g.l. During the sampling time air mass trajectories were of typically oceanic origin passing through urban and industrial areas around the coast (Fig. S1). PM 2.5 samples were collected using a high-volume (Hi-Vol) sampler (Energetica, Brazil) with an inlet for classifying particles smaller than 2.5 μm aerodynamic diameter (Thermo Andersen, USA). Samples were collected on quartz microfiber filters (22.8 cm × 17.7 cm, Whatmann, USA) over 4-12 h periods (7 AM to 2 PM, 2 PM to 7 PM, and 7 PM to 7 AM, completing a 24 h period per day, at the bus station) or during 24 h at the coast, at 1.13 m 3 min −1 . The sampling campaigns lasted 15 consecutive days in each place. After collection, filters were folded in half face-to-face, placed in an aluminum foil envelope then in a zip lock type plastic bag, and finally placed in sealed plastic containers for avoiding any contamination. Following, samples were transported cool to the laboratory and stored in a freezer (−4 °C) until analysis. Field blanks also were considered in this study. Field blank filters were placed in the same containers where sample filters were transported from laboratory to the sampling site and back to laboratory in order to make any minor contamination, if there was any, traceable. Our procedure included analysis of both sample and field blank filters in exactly same way for having any detectable analyte signal in field blank discounted from sample filter results. In this work, we did not detect any analyte in the field blanks. Backward air mass trajectories and statistical analysis. Backward air mass trajectory frequencies were calculated during the coastal site sampling time by using the NOAA HYSPLIT database 65,66 . Trajectory frequencies (as number of endpoints per squared grid per number of trajectories) were calculated with frequency grid resolution 1.0° × 1.0°, starting 96 h before arrival time (00:00 UTC) and altitudes ranging from 0 to 99999 m a.g.l. During the sampling period, air mass trajectories were of typical oceanic contribution, passing on Atlantic Ocean through urbanized city centers and industrialized areas around the coast before arriving to our coastal sampling collection site (SI Fig. S1). Multivariate statistical analyses, such as Pearson correlation, ternary correlation, Principal Component Analysis (PCA) and Agglomerative Hierarchical Clustering (AHC) were calculated for both dataset by using XLSTAT BASE software package version 19.5 for Microsoft Excel from Addinsoft Ltd (Paris, France). Ternary correlations were done by STATISTICA version 12.0 (Statsoft, USA). Incremental lifetime cancer risk assessment. Carcinogenic and mutagenic risk assessments 15, induced by inhalation of PM2. 2-NBA, and 3-NBA) and PAHs (PYR, FLT, BaP, and BaA) were estimated in the bus station and coastal site samples according to calculations done by Wang et al. 60 , Nascimento et al. 61 , and Schneider et al. 67 PAH and PAH derivatives risk assessment is done in terms of BaP toxicity, which is well established . The daily inhalation levels (E I ) were calculated as: where E I (ng person −1 day −1 ) is the daily inhalation exposure, IR (m³ d −1 ) is the inhalation rate (m³ d −1 ), BaP eq is the equivalent of benzo[a]pyrene (BaP eq = Σ C i × TEF i ) (in ng m −3 ), C i is the PM2.5 concentration level for a target compound i, and TEF i is the toxic equivalent factor of the compound i. TEF values were considered those from Tomaz et al. 15 , Nisbet and LaGoy 69 , OEHHA 72 , Durant et al. 73 , and references therein. E I in terms of mutagenicity was calculated using equation (1), just replacing the TEF data by the mutagenic potency factors (MEFs) data, published by Durant et al. 73 . Individual TEFs and MEFs values and other data used in this study are described SI, Table S4. The incremental lifetime cancer risk (ILCR) was used to assess the inhalation risk for the population in the Greater Salvador, where the bus station and the coastal site are located. ILCR is calculated as: where SF is the cancer slope factor of BaP, which was 3.14 (mg kg −1 d −1 ) −1 for inhalation exposure 60 , EF (day year −1 ) represents the exposure frequency (365 days year −1 ), E D (year) represents exposure duration to air particles (year), cf is a conversion factor (1 × 10 −6 ), AT (days) means the lifespan of carcinogens in 70 years (70 × 365 = 25,550 days) 70,72 , and BW (kg) is the body weight of a subject in a target population 71 . The risk assessment was performed considering four different target groups in the population: adults (>21 years), adolescents (11-16 years), children (1-11 years), and infants (<1 year). The IR for adults, adolescents, children, and infants were 16.4, 21.9, 13.3, 6.8 m 3 day −1 , respectively. The BW was considered 80 kg for adults, 56.8 kg for adolescents, 26.5 kg for children and 6.8 kg for infants 70 .

chemsum

{"title": "Occurrence of the potent mutagens 2- nitrobenzanthrone and 3-nitrobenzanthrone in fine airborne particles", "journal": "Scientific Reports - Nature"}

formation_of_a_mixed-valence_cu(<scp>i</scp>)/cu(<scp>ii</scp>)_metal–organic_framework_with_the_ful

## Abstract: Cu 4 I 4 ) 2.5 [Cu 3 (m 4 -O) (m 3 -I) (pmc) 3 (Dabco) 3 ]$2.5DMF$2MeCN} N (NJU-Bai61, NJU-Bai for Nanjing University Bai group; Dabco ¼ 1,4-diazabicyclo [2.2.2] octane), was synthesized stepwise. NJU-Bai61 exhibits good water/pH stabilities and a relatively large CO 2 adsorption capacity (29.82 cm 3 g À1 at 1 atm, 273 K) and could photocatalyze the reduction of CO 2 into CH 4 without additional photosensitizers and cocatalysts and with a high CH 4 production rate (15.75 mmol g À1 h À1 ) and a CH 4 selectivity of 72.8%. The CH 4 selectivity is the highest among the reported MOFs in aqueous solution. Experimental data and theoretical calculations further revealed that the Cu 4 I 4 cluster may adsorb light to generate photoelectrons and transfer them to its Cu 3 OI(CO 2 ) 3 cluster, and the Cu 3 OI(CO 2 ) 3 cluster could provide active sites to adsorb and reduce CO 2 and deliver sufficient electrons for CO 2 to produce CH 4 . This is the first time that the old Cu(I) x X y L z coordination polymers' application has been extended for the photoreduction of CO 2 to CH 4 and this opens up a new platform for the effective photoreduction of CO 2 to CH 4 . ## Introduction Due to climate change, CO 2 capture and conversion has recently, become one of the greatest concerns. 1 In particular, the photoreduction of CO 2 into value-added chemicals (such as CO, HCOOH, CH 4 , and so on) has attracted great attention, because it can be considered as a promising approach for solar-to-chemical energy conversion by mimicking the natural photosynthetic process to achieve a carbon neutral economy. 2 In the past few decades, diverse photocatalysts have been extensively employed for the photocatalytic CO 2 reduction reaction (CO 2 RR). 3 Homogeneous/molecular catalysts exhibit high selectivity and efficiency, but low activity due to catalyst deactivation, 4 whereas heterogeneous/inorganic catalysts show high activity and efficiency, but low selectivity. 5 Very recently, due to their high surface area, inorganic-organic hybrid nature, structural and functional diversity and tunability, metal-organic frameworks (MOFs) may combine the advantages of the traditional homogeneous/ heterogeneous catalysts and are emerging as promising platforms for the photocatalytic CO 2 RR. 6 Since 2011, 7 many MOFs have been designed for the photocatalytic CO 2 RR targeting to improve their efficiency, activity and selectivity by functionalizing organic ligands, optimizing metal ions/clusters, and making MOF-based composites. 8 Although, some achievements have been made, research on MOF-based photocatalysts to date is still in its early stages. In terms of the reductive products, most reported MOFs predominantly produce the 2e /2H + products of CO/HCOOH. 8a,9 Due to the fact that the photocatalytic reduction of CO 2 into CH 4 is more difficult than with other C1 fuels, because it involves a complex 8e /8H + reduction process, i.e., multiple steps of hydrogenation and deoxygenation reactions, and requiring the highest kinetic barrier of up to 818.3 kJ mol 1 , 10 the reported MOF catalysts capable of producing even low or moderate yields of CH 4 are still rare. Thus, design of MOFs with high selectivity for the reduction of CO 2 into CH 4 is a great challenge. 11 The Cu(I) x X y L z (where X ¼ Cl, Br or I; L ¼ N, P or S containing organic ligands) are almost the oldest coordination polymers with diversifed structures and interesting properties, such as luminescence and semiconductivity, and so on. 12 Very recently, their use has been demonstrated for photocatalytic H 2 evolution. 13 Herein the exploration of these polymers as promising platforms for CO 2 capture and conversion is reported. From a simple hetero-N,O ligand pyrimidine-5-carboxylic acid, a Cu 4 I 4 and Cu 3 OI(CO 2 ) 3 cluster based and semiconductive Cu(I)/Cu(II) mixed-valence MOF (NJU-Bai61) with a full light spectrum, which exhibits good water and pH stabilities and the relatively large CO 2 adsorption capacity (29.82 cm 3 g 1 at 1 atm, 273 K) was successfully constructed. In addition, NJU-Bai61 could photocatalyze the reduction of CO 2 into CH 4 without additional photosensitizers and cocatalysts and with a high CH 4 production (15.75 mmol g 1 h 1 ) and CH 4 selectivity of 72.8%. As far as is known, the CH 4 selectivity is the highest among the reported MOFs in the aqueous solution. Upon light irradiation, its Cu 4 I 4 clusters as photoelectron generators could transfer photoelectrons to the Cu 3 OI(CO 2 ) 3 clusters, whereas the Cu 3 OI(CO 2 ) 3 clusters could provide active sites for adsorbing and reducing CO 2 and act as photoelectron collectors for delivering enough electrons to CO 2 for CH 4 evolution. ## Results and discussion From CuI and the Hpmc ligand and using Dabco as the structural directing agent, like many Cu(I) x X y L z , a Cu 4 I 4 cluster-based copper(I) coordination polymer, {(Cu 4 I 4 ) (Hpmc) 2 } N (NJU-Bai61p) was initially obtained. NJU-Bai61p is a 2D layered and 4connected network with sql topology (Fig. S3, ESI †), in which each Hpmc ligand uses its N-donor center to link to a 4-coordinated Cu(I) in a tetrahedral coordination geometry resulting in a [Cu 4 I 4 N 4 ] moiety, leaving its COOH functional group uncoordinated (Fig. S4, ESI †). Later on, by changing the acid and extending the time, NJU-Bai61p was further transformed into NJU-Bai61 (Scheme 1). Compared with NJU-Bai61p, the Hpmc ligands in NJU-Bai61 were deprotonated, coordinated with Cu(II) ions in a bridging bidentate mode, facilitating the formation of the Cu 3 OI(CO 2 ) 3 cluster. The Cu 3 OI(CO 2 ) 3 cluster is 7-connected and surrounded by one Cu (Fig. 1d and S7, ESI †). The cages A and B connect alternately with each other to form a 1D channel by sharing quadrilateral windows, whereas the B cages connect with each other to form a 1D cage-stacked chain by sharing the facets including a quadrilateral window and a Cu 4 I 4 cluster (Fig. 1e, f, and S8, ESI †). Therefore, these 1D channels and chains are arranged in an alternating fashion to form a 3D porous framework based on the cages A and B ratio of 1 : 3, in which each cage A shares facets with six cage Bs and each cage B shares facets with two cage As and four cage Bs (Fig. 1g and S9, ESI †). From the viewpoint of structural topology, pmc ligands, Cu 4 I 4 and Cu 3 -OI(CO 2 ) 3 clusters could be regarded as 3-connected triangular nodes, 4-connected tetrahedral nodes, and 7-connected single cap octahedron nodes, respectively. Consequently, NJU-Bai61 is a new (3,4,4,7)-connected network with the point symbol {4 3 $6 12 $8 6 } 2 {4 3 $6 3 } 2 {6 3 } 6 {6 4 $8 2 } 3 (Fig. S10, ESI †). The phase purities and thermal stabilities of NJU-Bai61p and NJU-Bai61 were confrmed using PXRD and TG analyses (Fig. S13 and S14, ESI †). As shown in Fig. S15-S17 (ESI †), they are quite stable under water and other organic solvents. Furthermore, they are also stable under the broad variation of the pH values. NJU-Bai61p exhibits a visible light adsorption up to 550 nm due to the Cu 4 I 4 cluster to linker charge transfer (CLCT) transition (Fig. 2a and Table S2, ESI †). Very interestingly, NJU-Bai61 shows the widest absorption band among the reported MOFs with the edge up to 1400 nm, which are mainly dominated by intra metal cluster transfer (ICT), CLCT, and metal cluster-tometal cluster charge transfer (CCCT) transitions (Fig. 2a and Table S3, ESI †). The bandgaps of semiconductive NJU-Bai61p and NJU-Bai61 were estimated to be 2.33 eV and 0.92 eV, respectively, (Fig. S18, ESI †), which could be correlated with the calculated HOMO-LUMO gaps of 2.16 eV and 1.25 eV for the corresponding cluster models, respectively, (Tables S4 and S5, ESI †). The solid state of NJU-Bai61 with a periodic boundary condition (PBC) model for the band gap was further calculated, showing a narrow band gap of 0.65 eV (Fig. S19, ESI †). The Mott-Schottky measurements further revealed that they were typical n-type semiconductors and their conduction bands (CB) were 0.55 V and 0.58 V, which were more negative than the reduction potentials for the conversion of CO 2 to CO and CH 4 (Fig. 2b and S20, ESI †). 8a Thus, they are very promising for the CO 2 photoreduction applications. The photocatalytic reduction of CO 2 over the activated NJU-Bai61 was further investigated. The amount of CH 4 was 1.26 mmol (i.e., 15.75 mmol g 1 h 1 ) after 4 h. Except for the small amounts of CO (0.32 mmol, i.e., 4 mmol g 1 h 1 ) and H 2 (0.15 mmol, i.e., 1.87 mmol g 1 h 1 ), no other products, such as HCOOH, CH 3 OH and HCHO, were detected (Fig. 2c, S22 and S23, ESI †). The NJU-Bai61 exhibited a CH 4 selectivity of 72.8% in aqueous solution, which was the highest among the reported MOFs (Table S8, ESI †). No obvious change of the CH 4 activity occurred during the four continuous runs (Fig. S24, ESI †). The XRD patterns obtained before and after its photocatalytic experiments revealed the structural robustness of the catalyst (Fig. S27, ESI †). The isotopic 13 CO 2 tracing experiment was also performed to confrm that the carbon source of CH 4 did indeed come from the used CO 2 rather than the degradation of organics in the reaction (Fig. 2d). 11b For comparison, the use of NJU-Bai61p as the photocatalyst was also investigated under the same conditions and only CO (1.37 mmol, i.e., 17.13 mmol g 1 h 1 ) and H 2 (1.34 mmol, i.e., 16.75 mmol g 1 h 1 ) were detected after 4 h (Fig. S25, ESI †). This result may reveal that Cu 3 OI(CO 2 ) 3 clusters in NJU-Bai61 could provide active sites for CH 4 evolution. Then in-depth research was carried out to discover the reason underlying the high efficiency of CH 4 evolution. As for NJU-Bai61, the BET surface area was 248.1 m 2 g 1 and the CO 2 uptakes at 273 K and 298 K were 29.82 and 19.69 cm 3 g 1 , respectively, which was helpful for the subsequent CO 2 conversion (Fig. S28-S30, ESI †). The electrostatic potential analysis may further reveal that the Cu(II) centers in Cu 3 -OI(CO 2 ) 3 clusters are the most favorable sites for the nucleophilic attack of CO 2 (Fig. S31, ESI †). The local interactions between Cu(II) sites and CO 2 molecules were investigated using the in situ FTIR technology. The adsorption of CO 2 onto the Cu(II) sites in NJU-Bai61 was a 16 cm 1 red shift of the asymmetric stretching mode of CO 2 (n ¼ 2359 cm 1 ), indicating the stronger binding between the CO 2 and Cu(II) sites (Fig. S33, ESI †). 11b However, for NJU-Bai61p, no shift existed after CO 2 adsorption (Fig. S32, ESI †). Moreover, this experimental phenomenon was explained by the DFT calculations in which the peaks were also red-shifted and the adsorbed CO 2 molecule takes a slightly bent geometry to facilitate the CO 2 activation (Fig. S34 and Table S9, ESI †). 14 Furthermore, its fluorescence was quenched in comparison to NJU-Bai61p, indicating that the photo-excited electrons of the Cu 4 I 4 clusters were transferred to the Cu 3 OI(CO 2 ) 3 clusters, making it act as a photoelectron collector to provide electrons for the adsorbed CO 2 (Fig. S35, ESI †). An energetically feasible reaction pathway was calculated using DFT with the relative free energy, DG, for each step shown in Fig. 3 and S38 (ESI). † Upon light irradiation, the Cu 4 I 4 clusters in NJU-Bai61 may adsorb light to generate the photoelectrons and transfer them to the Cu 3 OI(CO 2 ) 3 clusters, whereas the Cu 3 OI(CO 2 ) 3 clusters could supply electrons to the adsorbed CO 2 for CH 4 evolution. In the frst step, the adsorbed CO 2 molecule accepted an electron and a proton to generate the COOH*. Then the COOH* combines with the second electronproton pair to generate CO*. The CO* was reduced to the CHO* by accepting two electrons and a proton, and further combined with a total of four electrons and fve protons to generate CH 4 . In the photocatalytic process, the Cu 4 I 4 cluster could serve as a photosensitizer and donated the energy of 2.16 eV to the conversion process of CO* to CHO* at the Cu 3 OI(CO 2 ) 3 cluster which was an endothermic process with the DG of 1.2 eV. Moreover, the stronger CO binding affinity on NJU-Bai61 (E b ¼ 20.13 eV) in comparison with that on only Cu(I)-contained NJU-Bai61p (E b ¼ 8.05 eV) may further stabilize the CO@Cu 3 IO(CO 2 ) 3 complex to complete the CO 2 -to-CH 4 conversion (Fig. S39, ESI †). ## Conclusions In summary, a novel Cu 4 I 4 and Cu 3 OI(CO 2 ) 3 cluster based and semiconductive Cu(I)/Cu(II) mixed-valence MOF with the full light spectrum, NJU-Bai61, was successfully produced, which exhibits good water stability, pH stability and a relatively large CO 2 adsorption capacity. NJU-Bai61 could photocatalyze the reduction of CO 2 into CH 4 , without additional photosensitizers and cocatalysts, but with a high CH 4 production and signifcantly high CH 4 selectivity of 72.8% (the highest among the reported MOFs in aqueous solution). It was revealed that the Cu 4 I 4 and Cu 3 OI(CO 2 ) 3 clusters may play the role of photoelectron generators and collectors, respectively. This work frstly expands the old Cu(I) x X y L z coordination polymers' application into the reduction of CO 2 to CH 4 and may open up a new system of MOFs for the reduction of CO 2 to CH 4 with high selectivity. ## Synthesis of NJU-Bai61p A mixture of Hpmc (11 mg, 0.09 mmol), CuI (30 mg, 0.16 mmol), Dabco (6 mg, 0.05 mmol), H 2 SO 4 (10 mL), DMF (1.0 mL), and MeCN (3.0 mL) was sealed in a 20 mL Pyrex tube and kept in an oven at 85 C for 1 day. After washing with DMF, yellow block crystals were obtained. Yield: 2.5 mg (6%). Selected IR (cm 1 ): 3036, 2666, 2554, 1713, 1586, 1441, 1398, 1330, 1297, 1202, 1170, 1119, 1090, 1054, 996, 908, 837, 749, 695, 667, 568. Elemental analysis (%) calcd. for Cu 2 I 2 C 5 H 4 N 2 O 2 : C 11.89, H 0.80, N 5.54; found: C 11.96, H 1.00, N 5.52. ## Synthesis of NJU-Bai61 A single crystal of NJU-Bai61p (10 mg), Dabco (4 mg, 0.036 mmol) and CuI (20 mg, 0.11 mmol) were added to 1.0 mL of DMF and 3.0 mL of MeCN. To this was added 60 mL of HCOOH with stirring. The mixture was sealed in a Pyrex tube and heated to 85 C for 2 d. Dark-red octahedral crystals were obtained and further characterized by PXRD and the results are shown in Fig. S1 (ESI †). Yield: 8.8 mg (25%). Selected IR (cm 1 ): 3392, 3108, 2952, 2883, 2840, 1681, 1652, 1587, 1435, 1377, 1319, 1218, 1170, 1087, 1050, 1000, 924, 840, 805, 764, 700, 612, 583, 468, 420. Elemental analysis (%) calcd. for Cu 13 I 11 C 44.5 H 68.5 -N 16.5 O 9.5 : C 16.66, H 2.15, N 7.20; found: C 16.87, H 2.30, N 6.98. ## Sample activation The as-synthesized sample of NJU-Bai61 was soaked in MeOH for 5 d with refreshing of the MeOH every 8 h. Then, the solventexchanged sample was activated at 70 C and under vacuum for 10 h to obtain the activated NJU-Bai61. ## Photocatalytic reaction The photocatalytic CO 2 reduction experiments were carried out on an evaluation system (CEL-SPH2N, CEAULIGHT, China), in a 100 mL quartz container. A 300 W xenon arc lamp (300 < l < 2500 nm) was utilized as the irradiation source. The 20 mg MOFs (NJU-Bai61p or the activated NJU-Bai61) were dispersed in 50 mL of a solution of triethylamine and water (TEA/H 2 O ¼ 5 : 45 v/v). The suspension was pre-degassed with CO 2 (99.999%) for 30 min to remove the air before irradiation. The reaction was stirred constantly with a magnetic bar to ensure the photocatalyst particles remained in suspension. The temperature of the reaction was maintained at 25 C by a circulating cooling water system. The gaseous product was measured by gas chromatography (GC-7900, CEAULIGHT, China) with a flame ionization detector (FID) and a thermal conductivity detector (TCD). An ion chromatography (LC-2010 Plus, Shimadzu, Japan) was used for the detection of HCOO . The concentration of Cu in the solution before and after catalysis was determined using an ICP-OES system (Optima 5300 DV, PerkinElmer). Before the photocatalytic reaction, the suspension of the activated NJU-Bai61 (220 mg), TEA (5 mL) and H 2 O (45 mL) was pre-degassed with CO 2 (99.999%) for 30 min to remove the air, then 2 mL of the fltrate was removed and a Cu concentration of 0.6 mg L 1 was detected. Thus, the concentration of dissolved Cu ions of the activated NJU-Bai61 was 0.05% before catalysis. After the photocatalytic reaction, 2 mL of fltrate was also removed and the concentration of Cu in the fltrate was determined to be 13.8 mg L 1 . Thus, the concentration of dissolved Cu ions of the activated NJU-Bai61 was 1.1%. The cycling experiment was carried out as follows: at the end of each run, the suspension was centrifuged and the supernatant was removed. Then the recovered catalyst was washed with distilled water and dried in air at 60 C before the next cycle. ## Conflicts of interest There are no conflicts to declare.

chemsum

{"title": "Formation of a mixed-valence Cu(<scp>i</scp>)/Cu(<scp>ii</scp>) metal\u2013organic framework with the full light spectrum and high selectivity of CO₂ photoreduction into CH₄", "journal": "Royal Society of Chemistry (RSC)"}

activity_enhancement_of_defective_carbon_nitride_for_photocatalytic_ammonia_generation_by_modificati

## Abstract: Photocatalytic nitrogen fixation under ambient conditions is currently widely explored in an attempt to develop a sustainable alternative for the Haber-Bosch process. Still, a lack of fundamental understanding of reaction pathways and nitrogen activation mechanisms result in the slow development in this field. In this work we combine defect-rich C3N4, one of the most investigated photocatalysts reported in literature for ammonia generation, with earth abundant and bioinspired FeS2 to improve the activity for photocatalytic ammonia production. By this approach, an activity enhancement of approx. 300 % compared to unmodified C3N4 was achieved. The optimal FeS2 loading was established to be 5 wt.%, with ammonia yields of up to 800 µg L -1 after irradiation for 7 hours. By detailed characterization of the electronic properties of the composites, we deduce that NH3 generation occurs via a novel mechanism involving mainly the reduction of the =N-CN group adjacent to nitrogen vacancies on defect-rich C3N4. FeS2 acts similar to a co-catalyst, improving the activity by π-back-donation from Fe-centers to the imine nitrogen of the defect-rich C3N4, reducing the activation barrier for terminal cyano group reduction upon illumination. ASSOCIATED CONTENTSupporting Information. Additional XRD patterns, XPS data, SEM images, TGA analysis, DRIFT data, more details on salicylate test, hydrogen evolution rate transients, post-catalytic analysis, elemental analysis, TAS transients, PL excitation spectra, PL emission spectra, AUTHOR INFORMATION ## Introduction The Haber-Bosch process is a large-scale industrial process for the production of ammonia (NH3) from hydrogen (H2) and nitrogen (N2) gas at elevated pressures and temperatures. It is a wellestablished and optimized process, that is crucial for the production of fertilizers. Nevertheless, it still suffers from sustainability issues due to high energy requirements and the utilization of natural gas for the supply of H2 in large, centralized power plants. 1 The photocatalytic nitrogen reduction reaction (NRR), that directly converts N2 into NH3 under (sun)-light irradiation at ambient conditions, presents a feasible alternative. 2,3 However, the NRR is a thermodynamically and kinetically unfavorable process due to the stable and inert nature of the N2 molecule. Furthermore, the similar potentials for N2 reduction and H2 evolution impede a high selectivity in aqueous media so far. This imposes strident demands on an efficient and selective photocatalyst, that are not even close to be met even by the state-of-the art materials, thus highlighting the importance of continued research on the topic. 2,4,5 The low NH3 yields further pose additional difficulties when it comes to accurate quantification, necessitating careful experiment control to avoid impurities and erroneous results. Discrepancies in the experimental conditions and a lack of standardized procedures further limit the comparability and reproducibility of photocatalytic NRR. 6,7,8,9 C3N4, is a low cost, non-toxic polymeric material with good light absorption characteristics, due to a band gap of approx. 2.7 eV, which has received a lot of attention as an n-type photocatalyst during the past decade. 10,11 It is commonly synthesized by thermal polymerization of organic precursor molecules, such as urea, melamine or cyanamide. The polymerization conditions, as well as the choice of the precursors have a strong effect on its structure and properties, by influencing the defect concentration, degree of polymerization, optical band gap, and surface area. 12 Assynthesized C3N4 suffers from high recombination rates. Therefore, alteration of the structure by doping or defect engineering and the formation of heterojunctions have been widely explored to improve the photocatalytic performance of C3N4. 13,14,15,16,17,18,19,20,21 Some of the highest NH3 yields in photocatalytic NRR have been reported using defective C3N4. 22,23,24,25,26 Owing to the polymeric structure of C3N4, the number of possible defects is vast and versatile. 27,28,21,20,18 A frequently exploited strategy for activity enhancement in C3N4 is the introduction of nitrogen vacancies, which are of the same size as the nitrogen atoms in molecular N2, and can thus act as efficient adsorption and activation centres. 27 Another important class of defects are cyano or cyanamide groups, that act as electron-withdrawing groups, assisting in charge separation and suppressing recombination. 14,24,29 The NRR is supposed to proceed via a Mars-van-Krevelen mechanism for both defect types, making a distinction between the influence of cyano groups and nitrogen vacancies difficult. Intercalation of potassium was reported to assist in the replenishment of nitrogen in the structure, more specifically of the cyano groups, resulting in an NH3 production rate of 3.42 mmol g -1 h -1 . 25 Treatment of bulk C3N4 with KOH, or direct incorporation of KOH into the synthesis recently emerged as a promising strategy for both the introduction of vacancies and of cyano groups. Zhou et al. treated bulk C3N4 with KOH in ethanol, followed by solvent evaporation and annealing. They observed an abundance of cyano groups that assisted in charge separation and N2 adsorption. 24 KOH has also directly been involved in the thermal polymerization of urea, likewise promoting the formation of cyano groups. 14 A similar result was obtained by Wang et al., using KOH in the thermal polymerization of dicyandiamide. 25 Etching with KOH was reported to mostly lead to the formation of vacancies. Still high NH3 yields of 3.632 mmol g -1 h -1 were reported, together with a quantum efficiency of over 20 %. 22 Another important type of vacancies in C3N4 for NRR are carbon vacancies, graphitized areas and possibly hydroxyl groups. 30,21 Furthermore, although thermal polymerization is an easy and versatile method, residual intermediates were observed the resulting C3N4, causing a high degrees of disorder. 31 Apart from defect engineering, the formation of heterojunctions, preferably of type II, can lead to an improvement of the photocatalytic activity with C3N4. 32 This is due to improved charge transfer characteristics and suppressed recombination, the latter being the dominant cause for the low efficiency in C3N4 materials despite the favorable band gap for visible light absorption. 33 Scientific research on semiconductor materials for photocatalysis is still in its infancy compared to the process optimization in nature. The reduction of molecular N2 has been realized under ambient conditions at the active centers of selected enzymes, termed nitrogenases. Three different classes of nitrogenases are distinguished, based on the composition of the active centers: all contain sulfur and iron atoms, but differ in the nature of additional constituents, one containing vanadium, one molybdenum and one solely iron. 34 This composition led to the exploration of several iron or molybdenum based compounds for the NRR. 35,36,37,38,39,40,41,42,43 One of the simplest iron and sulfur containing compounds is FeS2, which is a non-toxic, stable, and earth-abundant mineral. It shows very high optical absorption and high charge carrier mobility, but its use in photoelectrochemical applications suffers from charge trapping and recombination. 44 Still, FeS2 has been explored for photocatalytic dye degradation. 45,46,47 By itself it is inactive in photocatalytic NRR, due to unsuitable band positions, but it might still offer beneficial contribution in combination with other materials. 48,49 Thus, composites with FeS2 have been employed for dye degradation, H2 evolution, CO2 reduction and NRR. 50,51,52,53,54,55 Apart from improving the light absorption properties in composites, FeS2 can act as an efficient electrocatalyst, which has been shown separately for the H2 evolution, as well as for the NRR. 56,57,58,59,60 There have been some fairly recent reports on the formation of heterojunctions of FeS2 with C3N4 and their application in organic dye and antibiotics removal. 61,62 The reported mechanism and band positions on which the proposed electron transfers are based, are notably different between the reports, requiring further investigation of the interaction between both materials. In a bioinspired approach, we report the combination of two earth-abundant semiconducting materials, namely FeS2 and defective C3N4, for photocatalytic N2 reduction to NH3. Exploiting the good light absorption and N2 activation characteristics of the Fe-S system together with the high activity of defective C3N4, NH3 yields of up to 800 µg L -1 can be achieved in the course of the reaction (7 h), which equals to an activity enhancement by approx. 300 % compared to bulk C3N4. Even the natural pyrite mineral can be used for this approach. By detailed characterization of the electronic structure of the composites, we propose a novel reaction mechanism, revealing that FeS2 decoration weakens the bonds of C3N4-terminating imine groups in the vicinity of nitrogen defects by back donation, facilitating the reduction of terminal cyano groups under light irradiation towards NH3, with H2 being the only by-product. ## Material Synthesis and Characterization C3N4 was prepared via thermal polymerization of melamine. 63 1 g of melamine (Sigma Aldrich, 99%) was calcined at 550 °C for 4 h in a closed crucible in air, using a heating ramp of 5 K/min. The synthesis was repeated several times and the obtained C3N4 powder was ground and thoroughly mixed, before it was used for further modifications and composite formation. Vacancy-rich C3N4 (VN-C3N4) was obtained by dispersing 2 g of the as-synthesized C3N4 in 36 mL of 1 M KOH for 3 h under stirring. Subsequently, the material was collected via centrifugation and washed until neutral with ultrapure water. 22 For the composite formation, respective amounts of commercial FeS2 (Sigma Aldrich, 99.8%, 325 mesh) and VN-C3N4 were ground for 10 min in a mortar, under addition of low amounts of ipropanol (p.a.). Subsequently, the mixture was subjected to heat treatment for 2 h at 200 °C in air, to improve interfacial contact. The composites were analyzed before and after the photocatalytic experiments. Powder X-ray diffraction (XRD) was measured on a Malvern PANalytical Empyrean device with Cu Kα irradiation (λ1 = 1.5406 ; λ2 = 1.54443 ). Acceleration voltage and emission current were set to 40 kV and 40 mA, respectively. Peak assignment was performed with X'Pert Highscore plus. Following reference cards were used for the reflection assignment in FeS2 and FeS, respectively: 00-042-1340, 00-023-1123. The diffraction pattern for C3N4 was calculated with Vesta, using crystallographic data from the group of Irvine. 64 Diffuse-reflectance UV/vis spectra were obtained using a Perkin Elmer Lambda 750 spectrometer with a Praying Mantis (Harrick) and spectralon as white standard. The Kubelka-Munk function was used for the calculation of pseudo-absorption. 65 ## 𝑓(𝑅) = (1−𝑅) 2 ## 2⋅𝑅 For band gap determination, a Tauc plot was used. 66 [𝑓(𝑅) ⋅ (ℎ𝜈)] 1 𝑛 with n = 0.5 for direct band gaps and n = 2 for indirect ones. For diffuse reflectance infrared Fourier transformed spectroscopy (DRIFT) a Bruker Alpha II spectrometer and the software OPUS were used. Sample scans were taken from 400 to 4000 cm -1 , with a resolution of 4 cm -1 . Fluorescence measurements were conducted on a FluoTime 300 spectrometer from PicoQuant, with the software EasyTau2. Emission spectra were recorded at different excitation wavelengths from a 300 W Xe lamp at room temperature in air. Time correlated single photon counting (TCSPC) spectra were measured using 355 nm laser excitation. The software EasyTau was employed for fitting of the decay curves, using a tailfit with three exponentials, according to: ## 𝑖=1 For steady state measurements, the sample was stuck to carbon tape and placed in a holder for solid powder samples. For quantum yield measurements, dispersions of the respective sample in ethanol were dried as a thin film on the inside of a cuvette, which was placed in an integrating sphere. Measurements were conducted for two geometries: "in"signifying placement of the sample film directly in the excitation beam path, and "out"meaning that the film was positioned outside of the direct excitation path. Ideally both would give the same QY, but due to high absorption at ingeometry with relatively thick samples (not all fluorescence will be detected), values for outgeometry are higher and likely represent the more accurate values here. For the calculation of the quantum yield, the intensity of the fluorescence emission was integrated and the area AS divided by the total integral excitation intensity as measured in an empty reference cuvette (ABE) minus the excitation intensity that is not absorbed by the sample, ASE. ## 𝐴 𝑆 𝐴 𝐵𝐸 − 𝐴 𝑆𝐸 X-ray photoelectron spectroscopy (XPS) was performed with a Physical Electronics PHI VersaProbe III Scanning XPS Microprobe device. Monochromatic Al Kα X-ray irradiation with a beam diameter of 100 µm was used, with the beam voltage being set to 15 kV and x-ray power to 25 W. The sample surface was pre-cleaned by Ar-cluster sputtering with a gas cluster ion-beam. To avoid surface charging, samples were continuously flooded with slow-moving electrons and Ar + . For survey scans, pass energy and step size were set to 224 eV and 0.4 eV, respectively. Highresolution spectra were measured with a pass energy of 26 eV, a step size of 0.1 eV and a step time of 50 ms. For data analysis a CASA XPS 2.3.17 software was used. The background was corrected using Shirley subtraction. Peak fitting was done with Gaussian-Lorentzian line shapes, with 30 % Lorentz ratio. For charge correction C 1s was set to 284.8 eV. Transient absorption spectroscopy data was collected in diffuse reflectance geometry with an LP980 spectrometer (Edinburgh instruments). Pump laser pulse excitation was set at 355 nm (3 rd harmonic of an Nd:YAG laser produced by Ekspla, NT340), while for the probe pulse a 150 W xenon arc lamp was used. Prior to the measurements the powder sample was filled in a cuvette, stored under Argon and sealed directly before the measurement. N2 physisorption measurements were conducted on a Quadrasorb Evo device from Anton Paar QuantaTec at 77 K to determine Brunauer-Emmet-Teller (BET) surface areas, using the software ASiQwin for data evaluation. Samples were degassed for 12 h at 120 °C prior to measurements. Due to the small surface area, Kr at 77 K was used for FeS2, in an AS-iQ-MP-MP-AG instrument from Anton Paar QuantaTec. CHNS elemental analysis was performed with an Unicube instrument from Elementar, using sulfanilamide as standard. Approximately 2 mg of the respective sample were weighed into a tin boat, sealed and combusted at temperatures up to 1143 °C in an oxygen/ argon atmosphere. Thermogravimetric Analysis (TGA) with gas evolution detection via mass spectrometry (MS) was conducted with a Netzsch Jupiter STA 449C thermobalance together with a Netzsch Aeolos QMS 403C quadrupole MS, heating the sample at a rate of 5 K/min up to 900 °C in synthetic air. Scanning electron microscopy (SEM) images were recorded on a Zeiss Leo 1530 device with an acceleration voltage of 3 kV after sputter-coating with platinum (Cressington Sputter Coater 208 HR). Energy dispersive X-ray diffraction spectroscopy (EDX) measurements were conducted on the same instrument, using and an acceleration voltage of 20 kV. An ultra-dry EDX detector by Thermo Fisher Scientific NS7 was employed. ## Photocatalysis Photocatalytic NRR was performed in a flow setup using a doped Hg immersion lamp (Z4, 700 W Peschl Ultraviolet) placed in a water-cooled quartz-glass inlay and operated at 350 W. 200 mg of the photocatalyst were dispersed in approx. 30 mL of water by ultrasonic treatment for 10 min. The dispersion was transferred to the glass reactor and diluted to 600 mL by the addition of water and methanol. The total amount of methanol was 20 vol.%. Nitrogen was bubbled through the stirred dispersion at a flow rate of 50 mL/min overnight to flush out residual air. For pre-purification of the inlet gas stream, it was first passed through a 0.1 M KMnO4 solution, followed by a 0.1 M KOH. 8 The dispersion was illuminated for 7 h, during which a constant temperature of 10 °C was ensured by cooling the reactor with the help of a Lauda cryostat (Proline RP845). Evolving gasses were passed through an acid trap containing 10 mL of 1 mM H2SO4, 67 dried and subsequently analysed by a quadrupole mass spectrometer (HPR-20 Q/C, Hiden Analytical). After the reaction, the dispersion was immediately filtered and tested for NH3 using the salicylate test method, a modification of the indophenol blue method. 68,69 For the salicylate test, a stock solution of sodium hypochlorite and a stock solution containing the catalyst, sodium nitroprusside (Carl Roth, >99%), and sodium salicylate (Carl Roth, >99%), were prepared. The solutions were prepared fresh weekly and stored at 4 °C in the dark. For the preparation of the salicylate/ catalyst solution, 2 g of sodium salicylate and 8 mg of sodium nitroprusside were dissolved in 15 mL of ultrapure deionized water, to which 5 mL of a 2 M sodium hydroxide solution was added. For the preparation of the hypochlorite solution, 200 μL of sodium hypochlorite solution (5-10 % Cl, or 12 %, Carl Roth) and 1 mL of 2 M NaOH were given to 18.8 mL of water. In a typical testing procedure, 500 μL of the hypochlorite solution were given to 2 mL of the reaction solution, which was filtered through a 0.2 µm syringe filter beforehand. Then, 500 μL of the sodium salicylate solution were added. The mixture was stored in the dark at room temperature overnight for color development, before being analyzed by UV/vis spectroscopy (Perkin Elmer Lambda 750 spectrometer), using a mixture of the two testing solutions and 20 % aqueous methanol as reference. For the calibration, ammonium chloride (Carl Roth, >99.7%) stock solutions were prepared in a concentration range from 0.1 μg/L to 10 mg/L of NH4 + . Since the NRR was performed in the presence of methanol as a scavenger, 20 % of methanol were present in the calibration as well. Ion chromatography (IC) was used to analyse the reaction solution for nitrate by-products. The reaction solution was filtered with a 0.2 µm syringe filter. A 882 Compact IC plus device from Metrohm, equipped with a Metrosep A Supp 4-250/4.0 column and RP 2 Guard/3.5 pre-column were used for the analysis. 4 mM NaHCO3/ 1mM Na2CO3 were used as eluent, lower detection limit was 0.1 mg/L. For the quantification of hydrazine, the colorimetric method first reported by Watt and Chrisp was employed. 70 For the testing solution, 0.4 g of p-dimethylaminobenzaldehyde (Sigma Aldrich, 99%) were dissolved in 20 mL of ethanol (p.a.), to which 2 mL of concentrated HCl were added. For the calibration curve, standard solutions of hydrazine sulfate (Sigma Aldrich, >99%) in water/ methanol mixtures were prepared. For the measurement, 1.5 mL of the filtered reaction solutions were mixed with 1.5 mL of the testing solution and stored in the dark for 20 min for color development, before analysis of the absorbance with UV/vis spectroscopy against a reference containing only the testing solution and a water/ methanol mixture. ## Material Synthesis and Characterization The X-ray diffraction (XRD) patterns of the KOH-treated C3N4 (VN-C3N4) show the typical two broad reflections at 13° 2θ and at 27.4° 2θ, corresponding to in-plane order and interplanar stacking, respectively, with a d-spacing of approx. 0.326 nm (Figure S2). 64 The structure of C3N4 obtained via thermal polymerization is best described by a model of parallel melon chains connected via hydrogen bonds, as found by Lotsch et al. and confirmed in later studies. The unit cell is orthorhombic with the space group P212121. 71,72,73 This structure model will be assumed in the following. The reduced intensity of the reflection at 13° 2θ (210) upon KOH-etching indicates a reduction of in-plane order. This could be an effect of defect formation and / or consumption of amino groups that lead to partial breakdown and disorder in the heptazine units, which could in turn affect the parallel arrangement of the melon chains. Additionally, the (002) reflection at 27.4° is less intense and shifted to slightly larger diffraction angles, indicating a decrease in interplanar stacking distance and a general loss of order, which has been reported in literature as well and ascribed to either introduced cyano groups or nitrogen vacancies. 13,74,75 A combination of vacancy formation and conversion of amino into cyano groups can well explain the observed changes of both reflections. Powder XRD patterns of the formed composites show both the reflections for FeS2 and for VN-C3N4, with the intensity of the reflections for FeS2 increasing with its content (Figure 1). No additional phases could be observed. For further characterization of the morphology, SEM images were recorded. They show the rather large µm size of the FeS2 particles, distributed over a C3N4 matrix (Figure 2 and Figure S3). The FeS2 particles are well distributed and do not aggregate. In some places they are clearly distinguished from the C3N4 matrix, whereas in others C3N4 seems to wrap around the FeS2 particles. EDX confirmed the particles to be phase pure FeS2 with a ratio close to the ideal value of Fe:S 1:2 (Figure 2 and Table S1) and the surrounding matrix to consist of C3N4 with a C/N ratio of approx. 0.52. This is lower than the ideal value of 0.75 and could be an indication of incomplete condensation and thus free amino groups in the sample, although it is also effected by the low sensitivity of EDX for light-weight atoms. Physisorption measurements were conducted to evaluate the surface area of the composites with the BET model (Table S2). Pristine C3N4 exhibits the highest surface area of 10.6 m 2 /g, which slightly decreases upon KOH treatment. This is in good agreement with the decreased interlayer stacking distance observed in the XRD patterns and otherwise retention of the morphology. The apparent further decrease of the specific surface area upon composite formation can be explained by a difference in material density and particle size between C3N4 and FeS2, because the specific surface area is measured in m 2 /g. ## S K Fe K C K N K The optical properties of a photocatalyst are of utmost importance, since they vastly determine the efficiency in light harvesting. The color of the composites gradually gained a grey tinge with increased FeS2 content, compared to the previously pale yellow coloring of both untreated C3N4 and VN-C3N4 (Figure S1). Diffuse reflectance UV/vis measurements were conducted to elucidate the effect of FeS2 addition on the absorption behavior (Figure 3). C3N4 is an indirect n-type semiconductor. 76,77 Therefore, an indirect Tauc plot was used for a more accurate determination of the band gap and compared to the values apparent in the Kubelka-Munk plots. KOH-treatment results in a slight decrease of the band gap from 2.73 to 2.70 eV and marginally improved absorption in the UV region. The band gap reduction -and thus red-shifted absorption -could be caused by the introduction of cyano groups, whose electron-withdrawing properties were reported to lower the conduction band edge and lead to a narrowing of the band gap. 20,78 The increased UV absorption might be caused by improved charge separation due to the decreased layer distance and the introduction of cyano groups, since transitions in the UV region are commonly ascribed to ππ* transitions in sp 2 hybridized centers of the aromatic system. 25,79 The composites exhibit essentially all the same band gap, which is an indication that the major contribution to the light absorption is given by C3N4. This is expected, since its concentration is much higher than that of FeS2. The band gap of the composites is slightly increased but very comparable to VN-C3N4 with a change from 2.70 eV to 2.78 eV (Figure 3). The UV absorption is increased compared to KOH-treated C3N4. Both effects hint at a change in the electronic structure and mobility of electrons in the π-system. The color change and increased absorption of visible light upon addition of FeS2 is reflected by diffuse absorption at higher wavelengths, visible in an offset of the baseline (Figure S5). The effect of band gap widening is especially pronounced for the composites with a FeS2 ratio of 2.5 to 10 wt.% which might indicate optimal charge separation at medium FeS2 loading. All band gaps derived from the Kubelka-Munk and Tauc plots are summarized in Table S3. DRIFT spectra were recorded to further elucidate possible structure changes upon composite formation (Figure 3). The broad signal between 3000 and 3600 cm -1 can be assigned to O-H and N-H stretching vibrations, elucidating the presence of free amino groups that in turn indicate only partial polymerization. The sharp peaks between 1700 and 1200 cm -1 belong to stretching modes of C=N and C-N in the heterocycles, as well as bridging units, and the band at approx. 808 cm -1 can be assigned to the breathing mode of the s-triazine units. 80,24,81 Additionally, a band at 2150 cm -1 could be observed, that is ascribed to the presence of cyano groups, that appear to be present in both the untreated and treated C3N4, as well as in the composites to varying extent. 81,24,20 KOH treatment increases the amount of both cyano groups, as well as -OH and/ or -NHx groups. Additionally, the signals arising from stretching vibrations in the heptazine units are of slightly lower intensity for the defective C3N4, with the signal at 1720 cm -1 being more pronounced while the one for deformation vibrations at 808 cm -1 is of decreased relative intensity (Figure S6). This is an indication of structural damage inflicted on the heptazine framework, as would be expected by the formation of defects. The spectra for the composites are fairly similar to that of defective C3N4, indicating retention of the structure. An increased absorbance for the vibration of the heptazine units relative to that at 1720 cm -1 was observed for the composites with 10 and 15 % FeS2 in the normalized spectraan opposite effect to that caused by the KOH treatment. The signals for heterocycle-vibrations at higher wavenumbers (closer to 1700 cm -1 ) stem from C=N vibrations, while those at lower wavenumbers (closer to 1100 cm -1 ) arise from C-N vibrations. 82 Therefore, mainly C=N vibrations appear to be affected by the addition of FeS2. We additionally observe a slight shift of the vibrations for the heptazine unit towards lower wavenumbers, that hints at minor changes in the energy of the entire heptazine framework (Figure S6). To further investigate the structural evolutions upon KOH etching and composite formation, XPS measurements were conducted on bulk C3N4, VN-C3N4 and a composite containing 5 wt.% of FeS2. Survey scans show the expected signals for carbon, nitrogen and low amounts of oxygen in several spots (Figure S6). The latter is mainly due to adventitious carbon at the surface and not OH-groups in the defective C3N4, since oxygen was also observed in several spots on bulk C3N4. C/N ratios for bulk C3N4, and VN-C3N4, are 0.62 and 0.71, respectively, after correction for adventitious carbon, with slight deviations depending on the measured spot (Table S4). This is in good agreement to the C/N ratio derived from XPS analysis in literature and with the expected ratio for melon. 83,81 The increased carbon ratio after KOH treatment might indicate the introduction of nitrogen vacancies. The C/N ratio in the composite is with 0.71 identical to that of VN-C3N4, underlining that the structure remains intact during composite formation. Generally, KOH treatment on bulk C3N4 is expected to lead to partial hydrolysis of the structure. Yu et al. proposed a deprotonation of an apex amine group, during thermal polymerization in the presence of KOH which led to a breaking of the topmost cycle of the heptazine unit and cyano group formation. The overall amount of amino groups is retained here. 81 The reaction conditions notably differ from those employed in this work, but similar structural changes, specifically the introduction of cyano groups in addition to possible vacancy formation, are expected. Nitrogen vacancies are generally believed to be introduced at C-N=C sites. 25 These kinds of possible defects are mainly considered in the following. Two main peaks are observed in the C 1s spectra of untreated C3N4, corresponding to adventitious carbon at 284.8 eV and N-C=N species in the aromatic system at 288.2 eV (Figure 4). The existence of defects gives rise to new peaks. Most important is one at an intermediate binding energy of around 286.1 eV. This is usually assigned to carbon adjacent to amino groups. 80,24 Carbon atoms bound to cyano groups are expected to have a similarthough slightly higherbinding energy, that will overlap with this peak. 72,81 A minor shift from 286.1 eV to 286.2 eV is visible in the spectrum for VN-C3N4, hinting at an increase in the amount of cyano groups, as observed in the DRIFT spectra. Additionally, the intensity ratio of that respective signal to that of C-C changes from 0.21 in pristine C3N4 to 0.27 in the KOH-treated C3N4, which is expected since the total amount of amino groups should be maintained, while additional cyano groups are encompassed by the signal. The ratio of the cyano/ amino groups to that of N-C=N changes from 0.037 to 0.045, which can be explained by carbon atoms of the heptazine unit being converted into -CN upon KOH treatment. Another effect is the observation of a new π-π* satellite at 295.3 eV and a growth in the intensity of the π-π* satellite at ~293 eV, indicating the presence of larger numbers of free electrons in the vacancy rich system. This is in good agreement to the observations from UV spectroscopy and to literature, that claims an increase in the amount of unpaired electrons caused by both vacancy formation and introduction of cyano groups. 84,25 Furthermore, the peak at 298.2 eV is slightly asymmetric for untreated C3N4, which is probably an effect of the extended π-system. Upon KOHtreatment, the asymmetry is reduced, indicating damage inflicted on the aromatic system, which is in good agreement with the assumed breaking of some heptazine units. The formation of a composite of VN-C3N4 and FeS2 results in a marked shift for the signals of both carbon in the -N-C=N and in the cyano group towards higher binding energy (Figure 4 and Table 1). This might indicate electron extraction from the entire heptazine framework, resulting in partial oxidation. Four peaks are identified in the N 1s spectra for bulk C3N4 at approx. 398.6 eV, 399.2 eV, 401.4 eV and ~404 eV, in good agreement to the literature (Figure 4). 80,24,81,25 Correlation of the signals with corresponding nitrogen species is challenging, since peak assignment in the literature is ambiguous. 72 The most prominent peak at 398.6 eV in the spectra for bulk C3N4, and VN-C3N4 both can be assigned to -C-N=C species in the heptazine units (marked as N-(C)2 in the following), and a small peak at 404.9 eV corresponds to a π-π* satellite. The signals between 401 eV and 399 eV can be attributed to amino groups and to the nitrogen atom in the middle of a heptazine unit (N-(C)3), although the exact assignment of the signals to the two nitrogen species is discussed controversially. 25,80,72,83 Additionally, the signal for amino nitrogen should further be fitted by two, since the structure is not ideally graphitic, but closer to that of parallel melon chains, which results in the presence of both -NH and -NH2 groups, of which the primary amino group is expected at lower binding energy (Figure 4). 64,72 Cyano groups give rise to a signal at ~400.1 eV, which is indistinguishable from either the signal for the amino groups, or that of N-(C)3. 72 Generally, the nature of the N 1s spectra allows for the possibility of various fits that give good results mathematically but are meaningless in a chemical and physical sense, due to a large number of independent fitting parameters. This is demonstrated in Figure S8. If we assume that the structure of C3N4 obtained via thermal polymerization of melamine lies in between the model of parallel melon chains and of fully condensed sheets, the amount of NH should be smallest, compared to the other nitrogen species. 72 This expectation supports an assignment of the signal at 401.5 eV to NH instead of N-(C)3. Considering, that the signals for primary and secondary amines are likely to be adjacent, we follow assignment of the nitrogen species as: N-(C)2 at 398.7 eV, N-(C)3 at 399.5 eV, NH2 at 400.5 and NH at 401.5 eV, although we stress that a reverse assignment, as suggested in some XPS studies on pristine C3N4 also has its merits. 72,83 To allow for comparison of the spectra for VN-C3N4 and the composites with that for bulk C3N4, we therefore fitted the spectra using several constraints based on structural relationships (Figure S9 and Figure S10). Further information about the fitting process is given in the SI. The binding energies and atomic ratios for both the C 1s and the N 1s spectra with the most reasonable fitting result are summarized in Table 1. Upon composite formation wit FeS2, a marked shift in all signals towards higher binding energies is apparent, similar to the results observed in the C 1s spectra. This supports the assumption of electron withdrawal from the aromatic system. In order to still derive meaningful insights into structural changes from the N 1s spectra, we decided on normalization (Figure S11). Two things become immediately obvious. One is a shift of the main peak in the N 1s spectra from 398.7 to 399.0 eV upon composite formation, indicating an increase in the binding energy of N-(C)2 in the heptazine units. Additionally, the 'shoulder' at ~401 eV is less sharply pronounced, indicating larger amounts of nitrogen species at medium binding energy. The same shift towards higher binding energy is observed in the C 1s spectra, proving, that the energy of the entire heptazine-framework is increased upon decoration with FeS2. Additionally, the ππ*-satellite is of increased intensity. The former would indicate partial electron transfer from the C3N4 framework to FeS2, thereby altering the binding energy (oxidizing effect), whereas the latter would indicate an increased number of free π electrons. The lower amount of cyano groups in the untreated C3N4 is also shown by the lower intensity at 286 eV compared to both VN-C3N4 and the composite. These observations confirm the main conclusions drawn from the N 1s spectra above, without relying on arbitrary fitting results. Concluding these considerations, XPS analysis supports the introduction of both cyano groups and vacancies upon KOH treatment. Furthermore, it elucidates the changes in the electronic structure upon addition of FeS2 to the system, indicating partial electron extraction from the VN-C3N4 matrix. However, especially the number of possible fits for the N 1s spectra necessitates the comparison of the observed changes to the results of other characterization methods. Hydrolysis and introduction of OH-groups similar to what has been observed for hydrothermal treatment with NaOH, involving the breaking of NH-bridging bonds and introduction of OH-groups cannot be totally excluded, but is not likely to have a major influence, since oxygen contents are similar in both C3N4 and VN-C3N4. 73 TGA-MS measurements were conducted on the composite containing 5 wt.% of FeS2 and on both of the constituents, FeS2 and VN-C3N4, to confirm that no structural changes occur during thermal treatment for 2 h at 200 °C, that is part of the composite synthesis (Figure S12). FeS2 was stable up until about 400 °C, above which a gradual extraction of sulfur in the form of SO2 was observed. Notably, the lack of an increase in the mass indicates an absence of significant oxidation during the initial heating phase. For VN-C3N4, a major mass loss is observed starting at 570 °C, which is completed at approx. 720 °C. During the heating in synthetic air, NO, H2O and CO2 were found to be the main combustion products up until approx. 650 °C, after which an increase in the evolution of CO and NO2 was observed (Figure S13). The two steps of the combustion process are also apparent in the DSC curves (Figure S12). The TGA curves for the composite show many similarities compared to that of C3N4. However, both the mass loss curves and the ion currents for the evolving gasses are shifted by almost 100 °C to lower temperatures, signifying that the presence of FeS2 boosts the decomposition, likely acting as a catalyst and activating the heptazine units. Nevertheless, the observed thermal decomposition only occurs at temperatures far above the 200 °C, thus precluding decomposition during the composite formation. The interaction with FeS2 mainly involves the nitrogen in C3N4, as shown in the much decreased ion current for NO2 during the second step of the combustion process, indicating, that more nitrogen is extracted from the structure in the beginning. This is confirmed by the DSC curves, where the sharp peak at the end of the combustion process is much less pronounced (at 586 and 712 °C, respectively). While the ion currents for all gaseous combustion products in C3N4 show one signal with a sharp peak current, those of the composites appear as double peaks. Perhaps they correspond to the combustion of areas in close proximity to FeS2 and to areas for which the influence of FeS2 is less. Since the ion currents for both NO and NO2 are significantly lowered compared to that in C3N4 (also in relation to the ion current for CO2, so this observation is not only due to the lower content of C3N4), new nitrogen containing reaction products might be formed, that were not detected. ## Photocatalytic NH3 Generation Photocatalysis was performed in a flow setup using 20 vol.% of methanol as a hole scavenger. The reaction was investigated for VN-C3N4 and composites therewith containing different amounts of FeS2. After the irradiation period of 7 h, the solution was directly filtered and analyzed for NH3 by the salicylate test. Additionally, the concentration of NH3 in an acid trap located behind the reactor was evaluated. The calibration curves can be found in the SI, Figure S14. The quantification of NH3 was performed after stable color development, which was only obtained after several hours in the dark (Figure S15). The decoration of VN-C3N4 with FeS2 can significantly enhance the NH3 yield by a factor of 2.1 from 239 µg L -1 to 494 µg L -1 , which equals to 1.9 and 3.9 µmol h -1 for 200 mg photocatalyst. An almost equal improvement was observed for a FeS2 content between 1 and 10 wt.% (Figure 5 and Several control measurements were conducted, to elucidate the source of nitrogen and the selectivity of the reaction (Figure 5 and FigureS18). Firstly, a dispersion of a composite photocatalyst in water/ methanol was tested for NH3 after stirring in the dark. No NH3 was detected in this case. Secondly, a dispersion of VN-C3N4 was filtered, or centrifuged and tested for NH3, to exclude amino groups in the photocatalyst interfering with the test. No NH3 was observed in either control measurement. We also tested a physical mixture of separate FeS2 (5 wt.%) and VN-C3N4 for photocatalytic NRR ("Mix 5 wt.%"), without performing the grinding and subsequent calcination steps that establish an interfacial contact between the two constituents. The activity was significantly lower than that of VN-C3N4 itself, due to the lower amount of active photocatalyst and lack of interaction between FeS2 and C3N4. Additionally, the photocatalytic reaction for the composite containing 5 wt.% of FeS2 was repeated in an argon atmosphere. A similar activity was observed compared to the reaction in N2 atmosphere, strongly suggesting that the nitrogen stems from the C3N4 framework, instead of the feed gas (Figure 5). The slightly higher ammonia yields can be explained by the increased gas flow (100 instead of 50 mL/min). Literature for NRR over vacancy-rich and cyano-rich C3N4 likewise propose a N2 conversion pathway following a Mars-van Krevelen mechanism, therefore this observation is not surprising, 25 however significant NH3 yields for C3N4 under argon atmosphere, like we show, were hardly reported. The reaction solutions were additionally tested for nitrogen-containing by-products such as hydrazine and NO3 -. No hydrazine and only trace amounts of NO3around the lower detection limit were found for both VN-C3N4, and the composites (Figure S19, Figure S20). The main side product was H2, with a production rate of around 200 µmol h -1 . The rate was similar for all composites, with a slightly increased H2 formation rate for higher amounts of FeS2 in the sample (Figure S21). FeS2 itself showed a remarkably high H2 production rate of 360 µmol h -1 . Thus, both the activity improvement for NH3 generation and the selectivity are highest for lower FeS2 loadings in the composites. A gradual decrease in the H2 evolution rate over time is in good agreement with a possible Marsvan-Krevelen-type of structural changes during the illumination. When the photocatalytic experiment was performed in an argon atmosphere, similar H2 evolution rates were observed as compared to the results in a N2 atmosphere (Figure 5). In order to elucidate the formation of methanol oxidation products, UV absorption spectra of the reaction solutions were recorded after the photocatalytic experiments. There is a correlation between the amount of NH3 produced and the amount of UV light absorbance by the filtered reaction solutions at ~205 nm (Figure 6). This absorbance can mainly be attributed to formic acid that is formed alongside NH3 (Figure S22). This is in so far a problem, as it has an effect on the salicylate test, resulting in a significant underestimation of the actual NH3 concentration. 85 The absorbance for a 10 mM solution of formic acid was significantly lower than that observed for the reaction solutions after photocatalysis (Figure S22). Hence, its concentration is definitely not negligible, impeding accurate NH3 quantification. 85 Additionally, the spectra clearly show the presence of other ions that absorb UV-light. The similar wavelengths hamper a clear identification of the responsible ionic species, however. Derivation can help in the distinction between different absorption peaks (Figure S23). Still, the overlap of multiple features renders an accurate identification nearly impossible. Nitrate and nitrite additionally absorb UV light at 203 and at 210 nm, respectively. 86 No clear absorption feature could be identified for those two species, though, which is in good agreement to the ion chromatography results, proving, that only trace amounts are present. The broad absorbance from 230 nm towards higher wavelengths could in part be caused by very fine FeS2 particles, since a filtered FeS2 dispersion gives rise to a noticeable absorbance signal up to 450 nm (Figure S22). The observation of significant amounts of formic acid in the reaction solutions, requires a reevaluation of the testing parameters for the salicylate test. Initially we had ensured that we used sufficient amounts of hypochlorite in the test to stay in a concentration range, in which the test is independent of the amount of hypochlorite (Figure S16). Additionally, we always measured a reference sample of known concentration together with the reaction solution to avoid errors based on the testing solutions. However, when we re-measured the reaction solutions for the photocatalytic NRR tests with solutions containing higher amounts of hypochlorite, we found that the determined concentrations were now different, clearly indicating that we previously underestimated the actual ammonia concentrations in the photocatalytic experiments, due to unknown concentrations of formic acid (Figure 7 and Figure S17). Maximum ammonia yields of 819 µg L -1 (6.50 µmol h -1 ) were determined for the composite containing 5 wt.% of FeS2. Variations in the relative ammonia yields could be explained by different amounts of formic acid in the reaction solution and thus a varying degree in the influence on the salicylate test. This observation clearly elucidates that the test for NH3 should be repeated several times, ideally with different testing solutions and parameters, and the testing conditions should be verified for each new photocatalyst system. Critical evaluation of the conditions and evolving organic oxidation products is crucial, since the accuracy of the salicylate test can be influenced by a variety of parameters that are too often not considered. For NH3 quantification with Nessler's reagent, the contribution of organic side-products is even more severe, resulting in considerable overestimation of NH3 concentrations and hence also the NRR activity. 85 All concentrations given in the paper are averaged between several quantification measurements with the salicylate test. Since hardly any difference in the NH3 yield in argon and N2 atmosphere was observed after 7 h for the most active sample, we devised a longtime measurement, to gain further insights into the reaction mechanism and evaluate if nitrogen vacancies are replenished again under the conditions employed in photocatalysis, confirming a Mars-van-Krevelen mechanism. A dispersion of the composite containing 5 wt.% FeS2 was first illuminated for 14 h in either N2 or argon atmosphere. Every two hours, a sample was taken and analyzed for NH3. After 14 h, the lamp was switched off and the solution was stirred in the dark for 6 h under continuous gas flow to allow for nitrogen reincorporation into the structure, before illumination was continued for another 6 h (Figure S24). During the first irradiation period, the generated amount of NH3 increased almost linearly with a rate of approx. 82 µg h-1 in N2 and 90 µg h -1 in Ar atmosphere, respectively. Once the lamp was switched off, the measured concentration slightly decreased, likely due to NH3 carried out of the reactor by the gas flow. After continuing the light irradiation, the NH3 generation was increased again with about the same rate, as before the period in the dark. Afterwards, the determined concentrations seem to level out. This effect is likely caused by a combination of slow degradation of the structure and accumulation of oxidation products, that are impeding the NH3 quantification progressively more severely. A similar effect was observed for the H2 evolution rate, indicating that the observed effect is not solely caused by the accumulation of formic acid (Figure S24). The experiment does not prove a re-incorporation of nitrogen into the structure under the photocatalytic conditions. In addition to the NH3 concentration reported in the reaction solution, NH3 was detected in the acid trap but not included in the concentrations given here. The amounts of NH3 were maximum around 15 µg and 6 µg total for the experiments in an argon and a nitrogen atmosphere, respectively (Table S8). Still, the use of an acid trap is sensible. The pH of a dispersion of 20 mg of VN-C3N4 in 20 mL of H2O was approx. 9.5. Thus, both NH4 + and NH3 species can occur in significant amounts. For C3N4 the pH is with ~8 closer to neutral, perhaps due to residual OHspecies from the KOH etching, even after washing for at least 10 times with plenty of water. A slightly alkaline pH for C3N4 is expected, due to the large number of amino groups. The feasibility of the activity enhancement of VN-C3N4 for photocatalytic NH3 generation by decorating the surface with FeS2 was further underlined by using the natural mineral material of FeS2, that is earth abundant and inexpensive pyrite. Still, composite formation resulted in an increase in the ammonia yield, even though the particles were very large (mm sized grains ground in a mortar) and the surface was likely oxidized due to storage for years in air (Figure S25). Some additional reflections in the XRD patterns even indicated the presence of trace impurities (Figure S28), that do not seem to have a major influence on the photocatalytic activity of the composite. ## Post-photocatalytic characterization Both VN-C3N4 and the composites with FeS2 were thoroughly characterized after the photocatalytic experiments, in order to evaluate the stability and to gain further insights into the reduction mechanism. Post-photocatalytic XRD patterns still show the same reflections for phase-pure C3N4 and FeS2 (Figure S26). The intensity for the FeS2 reflections is significantly decreased after the NRR. This effect is a result of a loss of interfacial contact between FeS2 and C3N4 as verified by dispersing a composite in water/ methanol mixtures and subsequently regaining the material via centrifugation (Figure S27). FeS2 is known for its flotation tendency in mineral separation, due to its relative hydrophobicity. 87 Additionally, we tested the stability of FeS2 during both the formation of the composite (with annealing at 200°C) and storage of the sample in air. No changes in the crystal structure of FeS2 were observed (Figure S28). UV/vis spectra show a significantly increased absorbance of UV light and a decrease of the band gap (Figure S29). Both VN-C3N4 and the composites showed a pronounced darkening after the NRR that decreased again after storage in air (Figure S30). The effect of increased UV light absorption is less obvious for the composite containing 15 wt.% compared to the other composites, for which the activity was also lowest. All other composites that exhibited a similar activity, show a similar increase in UV absorption, which likely correlates to higher degrees of structural change. The increased UV absorption is least pronounced for VN-C3N4 in agreement with the lower activity. The band gap decreases very slightly by 0.03 eV for VN-C3N4. For the composite containing 5 wt.%, the decrease is most drastic and the band gap is experiencing a change by 0.07 eV from 2.78 to 2.71 eV. The red shift of the absorption might be caused by defect formation. DRIFT spectra for the composites after the NRR experiments show a marked decrease in the vibrations for the heptazine units (Figure S31 and Figure S32), especially in relation to the vibration at 1720 cm -1 , that falls into the range for C=N vibrations. This vibration is further shifted back to slightly higher wavenumbers. A decrease in the relative intensity of the 808 cm -1 vibration agrees with this observation, indicating structural damage to the heptazine framework. Both effects were also observed to a lesser extent upon introduction of vacancies and cyano groups upon KOH treatment and thus confirm further breakdown of the heptazine units during photocatalysis. Additionally, a closer look at the vibration at 2147 cm -1 reveals a decrease in the intensity, supporting the assumption that cyano groups are consumed during the photocatalytic experiment (Figure S33). The effect is even more increased for long-term experiments, highlighting the further degradation of cyano groups with prolonged illumination times (Figure 33). XP spectra of the composite containing 5 wt.% of FeS2 after the photocatalytic reaction show a slight shift for the carbon species adjacent to cyano-/ amino-nitrogen and the N-C=N peak towards lower binding energies (Figure S34). The intensity for both peaks corresponding to the C3N4 structure decreases markedly in relation to that of C-C, further indicating structural changes, that does not only extend to nitrogen, but is further inflicted on carbon in the structure. The ratio might, however, also partly be influenced by adsorbed organic residues from the sacrificial agent. Additionally, the intensity for the satellite peaks is decreased, supporting damage to the aromatic system. The calculated C/N ratio from the survey scan was 0.88, without correction for adventitious carbon, because the amount of C-C or C=C bonds possibly present in the structure after partial extraction of nitrogen is unknown. Thus the C/N ratio is increased compared to 0.80 for the same composite before the photocatalytic experiment. This is a strong evidence for nitrogen extraction from the C3N4 matrix. Sample composition was additionally evaluated by elemental analysis before and after photocatalytic experiments (Figure S35). The amount of FeS2 determined in the sample is less than what was expected for all composites. Since the FeS2 particles are rather large and thus limited in number, this effect might simply be due to inhomogeneous distribution over the C3N4 matrix. The amount of sulfur and thus probably also FeS2 was further decreased after the photocatalytic experiment for all composites (Figure S35). This is likely an effect of the imperfect interfacial contact and washing out of the sample, as has already been observed in the XRD patterns. For a more detailed comparison, the C/N ratio was calculated for the composite samples before and after the photocatalytic reaction. For all composites, the C/N ratio was about 0.555, which is significantly different from the compositional value of 0.75 and indicates incomplete polymerization and the existence of many free amino groups. The value is in good agreement to the EDX measurements, though. No significant differences of the C/N ratio were observed for different FeS2 loadings. For all composites, the C/N ratio was visibly increased after the photocatalytic reaction, supporting the extraction of nitrogen from the structure. Compared to the C/N ratios obtained from XPS measurements, the nitrogen content in the bulk is significantly lower, due to adsorbed carbon impurities and possibly nitrogen deficiency at the surface. ## Charge Carrier Dynamics Due to the large variety of defects that can potentially be present in C3N4, the electronic structure is rather complex and should be studied in detail, in order to gain an understanding of the underlying mechanisms in photocatalysis. To investigate the charge carrier dynamics, transient absorption spectroscopy (TAS) in diffuse reflectance geometry was employed. In Figure 8 the ns-TAS measurement of a composite of VN-C3N4 and 5 wt.% of FeS2 in an argon atmosphere is presented. The positive absorption feature between 650 nm and 900 nm can be assigned to photogenerated electrons in VN-C3N4. 88,89,90,91 The same signal is generally apparent in the measurement of the composite, but an increase of the relative absorption intensity in the range between 650 nm and 750 nm is noticeable, in comparison to the main signal between 750 and 900 nm. This might indicate a change in the electronic structure upon composite formation with FeS2 and an increased relative amount of photo-generated electrons in a second excited state. For a better comparison the main signal of the photogenerated electrons in both materials was analyzed regarding the lifetime of these electrons. The lifetimes monitored at 800 nm of both VN-C3N4 and the composite are comparable and in the order of ns (Figure S36). The signal is best fitted with two different lifetimes, that are 3. We additionally used fluorescence spectroscopy to gain insights into the radiative energy relaxation levels of the excited photocatalyst. Defect states introduce new energy levels into the material and fluorescence can occur from either excited defect states or the conduction band to the valence band, or to defect sites located slightly above it. This would in theory give rise to various emission signals of which many will have highly similar energies. The most dominant emission is that into the ground state. Therefore, low photoluminescence is often an indicator for efficient charge carrier separation and low recombination rates. C3N4 as a polymeric material with a structure in between that of molecules and solid crystals, generally exhibits a prominent blue fluorescence, that has been shown to be significantly influenced by the introduction of defects. 30,93 Upon irradiation at wavelengths below the band gap energy, a broad fluorescence signal between 410 and 640 nm is observed for both VN-C3N4 and the composites, with a maximum emission at 470 nm and a slightly less intense emission at 445 nm, as shown in Figure 9 and Figure S38. The samples could be excited over a wide energy range from 250 to 450 nm (Figure S37) for both emission signals, signifying that radiative recombination proceeds from the same energy levels. Notably, the relation between the intensity of fluorescent emission and excitation wavelength is essentially the same for all composites, with VN-C3N4 showing a different behavior. This indicates an alteration of the defect states in VN-C3N4 upon modification with FeS2. To better elucidate these changes, excitation spectra were normalized (Figure S38). The excitation maximum for VN-C3N4 is located at approx. 320 nm which corresponds to band gap transitions according to Gan et al.. 94 The same group ascribes excitation at around 380 nm to transitions from the valence band to lone pairs of -N-(C)3 species. This is in good agreement to the observed increased absorbance for the composites at this wavelength in the UV/vis measurements and also to the maximal fluorescence emission obtained at this excitation wavelength (Figure S39). The ratio of emission intensity at 320 nm excitation to emission intensity at 380 nm excitation in the normalized excitation spectra is significantly decreased for the composites compared to VN-C3N4, supporting the assumption of larger numbers of electrons in lone π and possibly also π* orbitalse.g. of cyano groups -as derived from XPS and UV-light absorption. Addition of >1 wt.% of FeS2 results in a decrease of the fluorescence emission compared to VN-C3N4, especially for the composite containing 2.5 and 5 wt.% of FeS2, which were also the composites that showed the highest activity for photocatalytic ammonia generation, indicating better charge separation compared to defective C3N4 and thereby reduced recombination (Figure 9). The photoluminescence is shifted to lower wavelengths for the composites, which is in good agreement to the slightly increased band gaps and indicates a decrease in interlayer electronic coupling. 88,95 FeS2 itself does not show any fluorescent emission, although a weak emission signal was detected at approx. 460 nm when FeS2 was stored in air for several days, indicating low degrees of surface oxidation. For better evaluation of differences in the signal shape and thus change in the contributions of intrinsic and defect emissions, the emission spectra were normalized (Figure 9 and Figure S38). The intensity for the emission at 445 nm in relation to that at 470 nm is clearly increased for the composites. The most prominent fluorescence is often attributed to band gap emission in literature. 93 In contrast, our observations would indicate that the most intense fluorescence signal at 470 nm is instead attributed to radiative emission from defect states and band gap emission likely occurs at 445 nm, which is in good agreement to the band gap energy determined via absorption spectroscopy. An additional tail towards higher wavelengths includes contributions from defect emissions, such as transitions from lone pair states of nitrogen atoms in the s-triazine unit. 94,30 The significantly increased inner band gap emission in relation to defect emission in the composites might indicate the inhibition of radiative recombination at defect sites, likely due to interaction of VN-C3N4 with FeS2 at the defect sites. Charge transfer might assist in exciton separation and retard recombination. The emission peak for VN-C3N4 and for each composite is largely independent of the excitation wavelength (Figure 9). A decrease in the relative intensity of the emission peak at 445 nm with increasing excitation wavelength is observed for the composites, along with a slight red-shift of the entire emission signal, that is especially pronounced at excitation with light in the order of the band gap energy. A higher excitation wavelength might increase the relative ratio of charges excited to lower level defect states, thus also increasing the relative ratio of emission from these levels. It has to be noted however, that the absolute emission intensity is decreased, since the light absorption is diminished and less charges are excited (Figure S39). The emission signals were fitted with Gauss functions, to allow for detailed investigation of different contributions to the PL spectra. Mostly, emission peaks for C3N4 are decomposed into three signals. 30,74 We additionally used four Gauss functions to account for deviations at higher wavelengths that are apparent in the fit with three species (Figure S40). The determined emission signals were located at ~437 nm, 465 nm and 505 nm, or at 438 nm, 464 nm, 488 nm and 528 nm, respectively. They likely correspond to inner band gap emission, emission from N-(C)3 sites and possibly either transition from electron lone pairs in NH-bridging nitrogen/ N-C bridging nitrogen, or from graphitized areas to the valence band. 94,30,93 The additional defect state emission at 488 nm could perhaps be related to either lone pairs in imine nitrogen in the vicinity of cyano groups/ nitrogen vacancies, or in the cyano groups themselves. The fitted emission signals show significant differences for the composites compared to VN-C3N4. The relative intensity for the direct band gap emission at 437 nm is considerably increased for the composites, as already observed from the general signal shape (Figure S40). It is also slightly shifted towards lower wavelengths for the composites, in correlation to the increased band gap energy. The relative intensity for the second emission signal at 464 nm is decreased for VN-C3N4 for the composites compared to VN-C3N4, especially for low to medium FeS2 loading of 5 wt.% or less. This might be attributed to better separation of charges at N-(C)3 sites. The emission signals at 480 nm and 520 nm are slightly increased in relation to the other components for the composites and shifted towards lower wavelengths. The shift could be a result of a variation in the energy of the entire C-N-framework, as observed in XPS and is especially pronounced for the state emission at 480 nm, which supports an emission from defect states at this wavelength. The intensity of the fluorescence emission can be an indication for the efficiency of charge separation. However, it is dependent on sample preparation and positioning in the instrument. The lifetime of an emission is regarded as a more reliable parameter for judging charge recombination rates. The lifetime was measured for both detection at 445 and 470 nm (Figure 10 and Figure S41). The decay can be fitted with three exponential functions, as is usually done in literature, as well. 14,30 The shortest fluorescence lifetime (τ3) of about 1.6 ns can be attributed to exciton recombination in the aromatic system (Table S10). 74 A second and third one with about 7 ns (τ2) and 25 ns (τ1) in the reference were attributed to charge carrier migration in the plane or between layers (along the π-stacking direction), respectively. 74 The lifetime for recombination in the aromatic system is slightly shorter for the composites compared to C3N4, which might be due to damaging of the πsystem. Larger changes are observed for the other two lifetimes (Figure S42). An increase in the second lifetime, τ2, for the composites compared to defective C3N4 implies improved intraplanar and intrachain charge separation. This might be due to polarization of the heptazine framework and electron transfer process due to interactions with FeS2. The greatest difference is observed in the time constants for τ1, which increases for the composites (Figure S42). Hence, charge separation in the π system is significantly improved by FeS2. The lifetimes are highly similar for all composites. Notably, the highest lifetimes were observed for composites containing 2.5 to 10 wt.% of FeS2, which is in good agreement to the highest photocatalytic activity at these FeS2 loadings. Fluorescence decay is slower at 470 nm longer compared to at 445 nm for all samples, supporting the assumption of the main emission at 470 nm mainly being caused by recombination at defect sides. The physical mixture of defective C3N4 and FeS2 and the composite with 5 wt.% of FeS2 show almost the same lifetimes, slightly higher for the composite. This implies interaction, likely in the form of charge transfer between both constituents, which is surprising considering that no beneficial effect of FeS2 addition could be observed in the photocatalytic reaction. However, during the fluorescence measurements, solid particles of both C3N4 and FeS2 are in direct contact, while in an aqueous dispersion, no interfacial contact is formed. The quantum yield (QY) was determined at three different excitation energies, to ensure that the increased fluorescent lifetimes truly correlate to improved charge carrier separation (Table 2 and ## Table S11). A decrease in the fluorescence QY with increasing FeS2 content is observed, implying that recombination is indeed reduced in the composites. The QY is highest at 320 nm, in good agreement to direct band gap excitation at this wavelength. ## Proposed mechanism for the interaction of FeS2 and defective C3N4 and resulting activity enhancement in photocatalytic ammonia generation The band positions of FeS2 are located at 0.27 V and 1.23 V vs. NHE, 48 respectively, whereas those of C3N4 are generally around -0.8 V and 1.85 V vs. NHE. 93,14 Hence, a type I heterojunction should be formed, that would result in charge accumulation on FeS2. However, we have shown that FeS2 is inactive for ammonia production. Band bending would additionally impede electron and hole transfer from the conduction band of C3N4 to FeS2 (Figure 11). Alternatively, electron transfer could follow a Z-scheme mechanism. Such an arrangement would result in N2 reduction taking place at the valence band of C3N4 and oxidation reactions occurring at the conduction band of FeS2. For a Z-scheme like transfer, the potential difference between the conduction band (CB) of FeS2 and the valence band (VB) of C3N4 is quite large, and the band alignment would be more suitable for hole transfer from C3N4 to FeS2. Additionally, the oxidation However, these considerations do not yet take into account the interaction with additional defect levels that can alter the electronic structure of the composite and promote interactions between C3N4 and FeS2. In order to test this hypothesis, composites containing 5 wt.% of FeS2 and newly synthesized, not KOH treated C3N4 were fabricated and employed in the photocatalytic experiments. No increase in the ammonia yield was observed (Figure S43). It is additionally evident in the DRIFT spectra before and after the photocatalytic experiment, that the signal for the cyano group remains unchanged for bulk C3N4. Based on the observations for structural changes in the VN-C3N4 induced by the presence of FeS2, we deduced a mechanism for the activity enhancement.: since KOH treatment is supposed to introduce defect sites that have two nitrogen species with free electron pairs in the vicinity (iminetype nitrogen in =N-CN units and amino groups), 81 an interaction similar to ligand to metal coordination to Fe 2+ appears to be feasible, that results in an activation of the C3N4 structure (Figure 11). This model suggestion is based on following observations: 1) Increased UV absorbance observed for the composites (more π-π*-transitions) (Figure 3). 2) Band gap widening upon FeS2 addition (Figure 3): possibly due to deformation/ polarization of the structure upon coordination of defects in C3N4 to iron centers due to partial electron transfer (σ-Donor) to FeS2. 3) The shift of both nitrogen and carbon binding energies towards higher values suggests partial oxidation of the structure, thus supporting the σ-donor effect (Figure 4). 4) The increased satellite peaks in the XP spectra suggest higher number of electrons in the πsystem, which might be due to π-back-bonding, (Figure 4). Possibly, coordination of the cyano groups and/ or amino groups to Fe 2+ centers induces a marked electron withdrawing effect and partial charge transfer to iron, reducing the electron delocalization. 5) Fluorescence quenching and decreased QY indicate improved charge transfer (Figure 9). 6) Decreased ratio of defect emission (from lone electron pair states) in fluorescence measurements (Figure 9) indicates interaction with and possibly partial charge transfer from defects in VN-C3N4 to FeS2. 7) Increased lifetimes for the longest lifetime, τ1, for the composite (Table S10): electrons in the π-system might be influenced by the interaction with iron. Thus, electrons excited into π* states of the aromatic system are possibly stabilized due to increased occupation of antibonding orbitals. 11) VN-C3N4 has significantly more free electron pairs than C3N4 that can coordinate to FeS2. 12) π-back-donation from Fe-centers to imine bonds are known, as is back-donation and charge transfer with cyano group. 96,97,98 13) ML charge transfer could be induced upon radiation, increasing the electron density in =N-CN groups. 14) The oxidation products in the photocatalytic reaction are the same for both, composites and VN-C3N4, (Figure 6), suggesting oxidation on VN-C3N4 15) The HER activity is decreased for the composites (Figure S21), even though FeS2 is present, indicating that electrons in FeS2 are used for other redox reactions, such as π-back-bonding and re-oxidation of Fe 3+ to Fe 2+ . Fe 3+ might form during light-induced ML charge transfer. 16) The activity enhancement was not observed for bulk C3N4 decorated with FeS2, indicating an interaction of FeS2 with the vacancies introduced by KOH-treatment. 17) Interaction of the composite with N2 is indicated in the TGA-MS measurements, possibly due to coordination to Fe 2+ . Such an effect could be advantageous for re-incorporation of nitrogen into the structure, similar to what has been observed for K + coordination. 25 The conditions for interaction with N2 (gas atmosphere vs. poorly dissolved nitrogen in aqueous solutions) is significantly different in TGA-MS compared to photocatalytic reaction, though. The coordination as proposed in Figure 11 would activate both, the imine nitrogen and nitrogen in the cyano group. In VN-C3N4, only the nitrogen in the cyano group would be slightly activated. Conversion of the entire =N-CN group would also result in a loss of C as observed in XPS (Figure 4) and the formation of carbon-containing oxidation products (Figure S24). Vacancies would be necessary for the interaction, therefore the effect was absent for composites with bulk C3N4, even though it also contains some amounts of cyano groups. Thus, photocatalytic ammonia generation with defective C3N4 is most likely always a product of self-degradation, which can be enhanced by back-donation with suitable coordinating species. ## Conclusion We have shown that a combination of defect engineering of C3N4 and subsequent composite formation with FeS2 can significantly improve the activity of C3N4 for photocatalytic ammonia generation, resulting in an activity enhancement of approx. 300 % compared to unmodified C3N4. The optimal FeS2 loading was established to be 5 wt.%. The system only employs inexpensive, earth abundant and non-toxic materials. Knowledge about the exact structure of C3N4 and the presence and character of defects proved to be crucial, as they significantly influence the interactions between the two constituents. Charge transfer between FeS2 and C3N4 was established to proceed only at the defect sites, resulting in an electronic activation of the structure. NH3 generation was found to occur via a novel type of mechanism, involving reduction of the =N-CN group adjacent to nitrogen vacancies. A replenishment of nitrogen in the structure could, however, not yet be verified. FeS2 acts akin to a co-catalyst, boosting the activity, although the mechanism is different to that of (metal) electrocatalysts typically employed for photocatalytic HER or OER. Here, π-back-donation from Fe-centers to imine nitrogen and amino groups of the defect-rich C3N4 reduces the activation barrier for the reduction of terminal cyano groups upon illumination. Although there are numerous reports on C3N4 for the NRR, the complexity of the system, with a broad variety of defects, that can theoretically be present, necessitates a strict control of the synthesis parameters and thorough characterization experiments. We deduce that photocatalytic ammonia generation with defective C3N4 is most likely always a product of self-degradation, which can be enhanced by back-donation with suitable coordinating species to reduce the activation barrier.

chemsum

{"title": "Activity Enhancement of Defective Carbon Nitride for Photocatalytic Ammonia Generation by Modification with Pyrite", "journal": "ChemRxiv"}

study_of_ruthenium-contamination_effect_on_oxygen_reduction_activity_of_platinum-based_pemfc_and_dmf

## Abstract: We outline a systematic experimental and theoretical study on the influence of ruthenium contamination on the oxygen reduction activity (ORR) of a Pt/C catalyst at potentials relevant to a polymer electrolyte fuel cell cathode. A commercial Pt/C catalyst was contaminated by different amounts of ruthenium, equivalent to 0.15-4 monolayers. The resulting ruthenium-contaminated Pt/C powders were characterized by Energy-Dispersive X-ray Spectroscopy (EDS), X-ray Photoelectron Spectroscopy (XPS) and Scanning Transmission Electron Microscopy (STEM) to verify ruthenium contamination. A rotating disk electrode (RDE) technique was used to study the influence of ruthenium on oxygen reduction kinetics. Density functional theory (DFT) calculations were performed to estimate the oxygen reduction activity of the platinum surface with increasing ruthenium coverage, simulating ruthenium-contaminated Pt/C. The binding energies of O and OH on the surfaces were used for activity estimations.It was found that the specific activity of the ORR at 0.85V vs RHE exhibited a pseudoexponential decay with increased ruthenium contamination, decreasing by ~45% already at 0.15 monolayer-equivalent contamination. The results of the DFT calculations were qualitatively in line with experimental findings, verifying the effect of O and OH binding energies and the oxophilic nature of ruthenium on ORR and the 2 ability of the chosen approach to predict the effect of ruthenium contamination on ORR on platinum. ## Introduction Hydrogen-fed Polymer Electrolyte Membrane Fuel Cells (PEMFCs) and their closely related Direct Methanol Fuel Cells (DMFCs) are considered to be promising energy generators for electric vehicles (EVs), backup or off-grid power and mobile electronic devices . Despite intensive research over the last three decades, performance, durability and cost issues are still major obstacles to successful widespread commercialization of PEM-type FCs. One of the reasons for cost challenges of PEMFCs and DMFCs is the fact that carbonsupported nano-size platinum (Pt/C) is used as the catalyst on the anode and cathode. In reformate-based PEMFCs, the hydrogen stream to the anode contains CO produced by reforming or partial oxidation of hydrocarbons or alcohols and by a reverse-shift reaction of CO2 . In DMFCs, direct oxidation of methanol on the anode to CO2 progresses with production of mainly CO as an intermediate. Even very low concentrations of CO (10ppm) in a reformate-based H2 poison Pt/C catalysts by strongly adsorbing on the platinum surface, hence reducing the available electrochemically active surface area and seriously inhibiting catalysis of H2 oxidation. Likewise, in DMFCs, CO (an intermediate of methanol oxidation) is adsorbed on platinum, preventing catalysis of methanol oxidation. To solve this CO-poisoning problem, platinum-ruthenium (PtRu) alloys are used as anode catalysts. The ruthenium component provides the oxygen-containing species needed to oxidize CO to CO2 and release the platinum surface for further fuel oxidation . However, PtRu catalysts were found to be prone to preferential dissolution of ruthenium , especially in the presence of methanol . Ruthenium dissolution from the anode catalyst results in a loss of ruthenium and changes to its Pt:Ru ratio. This change leads to reduced CO tolerance and methanol oxidation activity of the catalyst (for reformate-based PEMFCs and DMFCs, respectively) that will translate to higher overpotentials of the anode . Ruthenium ions leaving the anode enter the Nafion membrane and cross it to deposit on the platinum cathode catalyst. Ruthenium crossover and its deposition on the cathode in DMFCs were first reported by Piela et al. of the Los Alamos Group . Their pioneering work showed, with the use of XRF and CO-stripping, the presence of ruthenium in the membrane and the cathode already after MEA humidification (the socalled current-less contamination) and after operating a DMFC under various operating conditions (the so-called current-assisted contamination). Cathode contamination resulted in a voltage drop of 25mV in an H2/air-operated fuel cell across the entire current density range. This voltage drop was ascribed to lowered catalytic activity of oxygen reduction on the Ru-contaminated cathode. Schoekel et al. reported ruthenium dissolution and deposition on the cathode already during fabrication of the MEA by decal transfer process . Rapid ruthenium contamination of the cathode was recorded during early operation time (two hours) of a DMFC, and was attributed to dissolution of highly soluble ruthenium species in the anode catalyst (Johnson Matthey HiSPEC 12100). Following that, a slower contamination process was recorded, attributed to dissolution of less soluble ruthenium oxides or ruthenium from the platinum-ruthenium alloy phase. The negative effect of ruthenium contamination on the catalytic activity of the ORR demonstrated by Piela et al. , is consistent with studies of ORR kinetics on ruthenium and PtRu surfaces. Anastasijević et al. studied the ORR mechanism and kinetics on a ruthenium rod. Their results clearly showed that ruthenium has poor ORR activity at potentials relevant to the PEMFC cathode. A similar conclusion can be drawn from the studies of ORR on electrodeposited ruthenium on a gold disk by Metikoš-Huković et al. and ruthenium nanoparticles by Cao et al. . Stamenkovic et al. demonstrated the poor ORR activity of the polycrystalline PtRu (1:1) alloy electrode in comparison to the polycrystalline platinum electrode. The negative impact of ruthenium presence on ORR catalytic activity can also be seen in the case of Pt/Ru nanoparticles with exposed ruthenium on the surface . Gancs et al. have studied the effect of platinum contamination by ruthenium on ORR catalysis . In their study, different concentrations of ruthenium ions were used to contaminate commercial Pt/C by spontaneous deposition of ruthenium. With the use of different concentrations of ruthenium ions, several degrees of Ru-contamination were produced, thus enabling the study of the effect of contamination degree on CV polarization curves and ORR kinetics. Continuous suppression of Hupd stripping peaks with increased Ru-contamination was recorded, as well as severe decrease in ORR kinetics as evidenced by RDE polarization curves and Tafel plots. Quite close Tafel slopes were recorded for clean and Ru-contaminated Pt/C (-122 mV/dec vs 113 mV/dec respectively), indicating an identical ORR mechanism (at least at low overpotentials) for which ruthenium contamination is not a factor. Interestingly, at ruthenium coverage of 0.18 monolayer (ML), Ru ORR kinetics reached a minimum value and a maximum overpotential of ~160 mV was recorded. In this work we studied the influence of ruthenium contamination on ORR kinetics with the use of a commercial Pt/C catalyst (Johnson Matthey HiSPEC8000) that was Rucontaminated. Various and precisely known amounts of ruthenium were deposited on platinum by electroless deposition at 90°C with methanol as the reducing agent. This resulted in Pt/C catalysts contaminated with different coverage levels of ruthenium. This deposition method was chosen to mimic the existing conditions in a DMFC cathode (approximated working temperature and presence of methanol which had crossed from the anode). The effect of ruthenium on ORR performance was measured by cyclic voltammetry with an RDE. We believe that this simple approach has allowed us to correlate between precisely known ruthenium contamination of Pt/C and its ORR kinetics behavior. To explain the effect of Ru contamination on the ORR activity, DFT simulations were performed for the adsorption of Ru atoms on the Pt(111) surface and their effect on the O and OH binding energies. We then followed the analysis performed by Nørksov et al. to correlate the calculated binding energies with a model estimate for the ORR activity. We showed that the theoretically obtained ORR activity trends have good qualitative agreement with the experimental trend. The combined experimental and theoretical work leads to a deeper understanding of the effect of platinum contamination by a sub-monolayer to a few monolayers of ruthenium on ORR kinetics and potential losses in PEMFCs and DMFCs. ## Catalyst synthesis The Ru-contaminated catalysts were prepared by electroless deposition of ruthenium on commercial 50%Pt/C (HiSPEC8000, Johnson Matthey) with methanol as the reducing agent. For each catalyst, the total amount of deposited ruthenium was equivalent to 0.15-4 (0.15, 0.22, 0.6, 1, 2, 4) monolayers of ruthenium. The calculation of the required amount of ruthenium for each catalyst was made on the basis of the atomic radius of ruthenium and an approximation of spherical platinum nanoparticles with a surface area of 60 m 2 gPt -1 (manufacturer's data). The catalysts were named according to the amount of deposited ruthenium: 0.15ML Ru/Pt, 0.22ML Ru/Pt, 0.6ML Ru/Pt, 1ML Ru/Pt, 2ML Ru/Pt, 4ML Ru/Pt. For the synthesis, 0.2 g of HiSPEC8000 was dispersed by vigorous magnetic stirring in an aqueous solution of 1M methanol at room temperature. The suspension was heated to ~90 °C while being refluxed. A desired amount of RuCl3•3H2O was dissolved in 10 mL of 0.4 M HCl solution and added to the suspension at a rate of 1 mL every 15 minutes while the suspension temperature was maintained at ~90 °C. On completing the addition of ruthenium solution, the mixture was refluxed for an additional 30 minutes and then cooled to room temperature. The catalytic powder was recovered by centrifugation, washed with DI water until no chloride ions could be detected and dried. ## Electrochemical characterization All electrochemical experiments were performed at a controlled temperature (25±1 °C) with the use of a custom-made three-compartment glass cell with an Ag/AgCl/3 M KCl reference electrode in a Luggin-capillary compartment and a platinum wire as a counter electrode. A 0.5 M H2SO4 solution was used as the electrolyte. All potentials are reported on the reversible-hydrogen-electrode (RHE) scale. Measurements of the electrochemically active surface area (ECSA) of Ru/Pt/C were carried out by the Cuupd stripping method , described in detail in our previous publication . The working electrode was a 1 cm×5 cm glassy-carbon rectangle. The catalytic ink consisted of 10 mg of catalyst powder, 5 %(w/w) Nafion solution, 7.5 mL DI H2O, 2.5 mL EtOH and XC72 that was added to obtain a concentration of 0.2-0.3 %(w/w) solids in the ink. This concentration range of solids allowed obtaining a stable ink. The Nafion volume was adjusted to be ~30 %(v/v) of the solids in the inks. The ink was dispersed for 60 minutes in an ultrasonic ice-water bath, with additional fiveminute dispersion by pulse sonication (also in an ice-water bath) with the use of a horn sonicator (Heilscher UP200st). Immediately after sonication, 10 μL of the catalytic ink was applied to the lower part of the working electrode. Nitrogen (99.999% purity) was bubbled through a 0.5 M H2SO4 solution for 30 minutes before electrochemical experiments and then passed over the solution during the entire procedure. Prior to ECSA measurements, the working electrodes were conditioned in order to clean their surface. A common conditioning procedure of Pt and Pt/C catalysts consists of repetitive cycling in deaerated electrolyte over a potential range of 0-1.2/1.4 volts until a stable voltammogram is obtained . Cycling of Ru/Pt/C catalysts above 750mV might lead to substantial ruthenium dissolution that would result in unintended surface modification of these catalysts . Hence, to reduce the possibility of ruthenium dissolution, the conditioning procedure of Ru/Pt/C catalysts was limited to 0-750 mV potential range . For ORR measurements, a 5mm-diameter glassy-carbon RDE (Pine Instruments, USA) with Ageo=0.196 cm 2 was used. The RDE was polished to a mirror finish with a 0.05 µm Al2O3 particle suspension on a moistened polishing Micro-Cloth (both from Buehler). The electrode was mounted on an interchangeable RDE holder connected to an electrode rotator (MSRX electrode rotator, Pine Instruments, USA). The catalytic ink for ORR measurements consisted of catalyst powder, XC72 powder, 5 %(w/w) Nafion solution, DI H2O and IPA. The Nafion volume was adjusted to 30 %(v/v) of the solids in the inks. A catalytic loading of ~20 µgPGM cmgeo -2 on the RDE was used and the weight of catalyst powder, DI H2O and IPA volumes (~30 %(v/v) IPA ) were adjusted accordingly. As in the case of ECSA measurements, addition of XC72 to obtain a concentration of 0.2-0.3 %(w/w) solids in the ink enabled the preparation of a stable ink and also a uniform catalyst coating on the RDE. The ink was ultrasonically dispersed by the same procedure as the ink used for ECSA measurements, afterwards 10 µL of the ink was applied on the RDE. Several studies have shown that a uniform catalytic film has a beneficial effect on the currents obtained during RDE ORR polarization . We have examined different drying procedures of the ink droplet in order to obtain the most uniform film. In our laboratory environment a stationary drying procedure at room temperature with IPA environment consistently produced the most uniform catalytic films. Subsequently, this drying method was used during this research. Nitrogen (99.999% purity) was bubbled through a 0.5 M H2SO4 solution for 30 minutes before electrochemical experiments and then passed over the solution during conditioning and background measurement. iR drop between the working and reference electrodes was measured and consistently found to be ~4 Ω. Working electrodes with Ru/Pt/C were conditioned as mentioned above. Working electrodes with Pt/C were conditioned by cycling over a potential range of 0-1.2 V. For background measurements, the working electrodes were cycled for five cycles over 0-1V (for Ru/Pt/C) or 0.025-1.2 V potential range (for Pt/C) at 20 mV s -1 . Background measurements were also used to measure ECSA of Pt/C by the Hupd stripping method . Before ORR measurements, the O2 (99.999% purity) was bubbled through the electrolyte for 30 minutes and then passed over the electrolyte during ORR polarization. The RDE potential was cycled between 1 and 0 V at 20 mV s -1 while the RDE was rotated at 2500 rpm in the O2-saturated electrolyte. The current at 0.85 V during anodic sweep polarization, after mass transport, background and iR corrections, was taken as a measure of the ORR activity, 𝑗 𝑘 0.85 𝑉 . ## Physicochemical characterization Detailed procedures of EDS and XPS measurements were described previously . The measurements were made at ten (EDS) and four (XPS) points, respectively, in each sample of homemade catalysts and the results showed no significant inhomogeneity. The reported results are an average of the measurements. Transmission Electron Microscope (TEM) imaging was performed with an FEI F20 Philips-Tecnai STEM operated at 200 kV. Samples were prepared by manually pressing the grid (200-mesh grid, EMS) against the sample powder. TEM images were used to construct particle-size distributions of the catalysts by measuring diameters of at least 80 individual particles with ImageJ software . Elemental mapping was performed with the use of an FEI Titan 80-200 STEM equipped with a CS-probe corrector (CEOS GmbH). "Z-contrast" conditions were achieved with the use of a probe semiangle of 25 mrad and an inner collection angle of the detector of 68 mrad. During STEM-EDS elemental mapping, HAADF detector and Pt L and Ru L peaks were used. Samples were prepared by placing a drop of diluted sample on a 400mesh carbon-coated copper grid. ## Theoretical methods and computational details Our total energy calculations were carried out with the use of DFT simulations within the Perdew-Burke-Ernzerhof generalized gradient approximation (PBE-GGA) as exchange-correlation energy functional, and the all-electron projected augmented wave (PAW) method as implemented in the Vienna Ab initio Simulation Package (VASP) code . For the calculations, a plane wave cut-off energy of 500 eV and k-point grids of 10×10×10 and 6×6×1 for the bulk and slab surface cells, respectively, were used. Geometric relaxation was considered to be complete once the atomic forces on each atom were smaller than 0.02 eV -1 , and a total energy convergence of 10 −6 eV for the structural energy minimization was achieved. For the bulk fcc Pt, the calculated equilibrium lattice constant is 3.968 , which is consistent with theoretical findings from the Aflow database and other theoretical and experimental results. The calculated bulk Ru hcp parameters were a=2.721 and c=4.293 , also in agreement with the theoretical and experimental results. ## Modelling of Ru on Pt(111) The adsorption of Ru atoms was modelled with different coverage levels (of 0.11, 0.22, 0.44, 1ML, which correspond to 1, 2, 4 and 9 atoms in the surface unit cell) on Pt(111) applying the repeated slab geometry model with a 3×3 surface unit cell and five layers in the slab separated by a vacuum region of about 25 . A single Ru ad-atom was placed at the hollow and bridge sites over Pt(111) surface as they are considered to be the most favorable positions. Then, the adsorption of additional Ru atoms, such as 2, 4 and 9 Ru atoms, was modelled on the Pt(111) surface. In order to take into account all possible Ru geometries, as a starting point for geometrical relaxation, the Ru atoms were placed as planar and three-dimensional clusters (that is, pyramid-like configurations for Ru4 and Ru9) and separately diffused atoms over the Pt(111) surface. Then, geometrical optimization of the initial structures was performed, allowing all atoms to move but freezing the two bottom atomic Pt layers. Finally, the lowest energy structures were selected according to the total energies of the optimized structures. ## Modelling of O and OH on RunPt(111) The initial structures for mO (m = 1, 2, 3, and 4) atoms on the Run/Pt(111) surfaces were built on the following basis: O atoms were placed at hollow, bridge and top sites over Ru and Pt atoms. Then the systems were allowed to relax, again freezing the two bottommost Pt layers. The initial structures for mOH (m = 1, 2, 3, and 4) species on the Run/Pt(111) surfaces were constructed on the basis of the optimized mO/Run/Pt(111) structures. ## Binding energies, reaction energies and activity The following steps were applied for the calculations. First, the binding energies (Eb) for O atoms on Run/Pt(111) surfaces were calculated as Eb=(E(mO/RunPt(111))- , where the first, second and third terms are the total energies of the mO/Run/Pt(111), Run/Pt(111) and the O atoms in gas-phase, respectively. m is the number of O atoms in the system, the same definition for binding energy was applied also to OH (OH replacing O in all equations). For the estimation of ORR activity, we follow the approach by Nørksov et al. , for completeness we repeat the main principles of this approach here. The oxygen reduction reaction can be written as: As described in , in the simplest way, it is possible to consider the following processes at the surface: Here "*" implies the pure surface, and O* and OH* imply the surface with the adsorbed species. Following , we analyze the reaction energies for the reactions: We assume that the hydrogen evolution reaction: is in equilibrium for a potential 𝑈 0 = 0 relative to the standard hydrogen electrode. It is hence evident that the free energy change of reaction 6 is the minus of the change in reaction 4, and that the difference of the free energies of reactions 6 and 5 yields the free energy change of reaction 3. The reaction energies (∆EO and ∆EOH) for reactions 5 and 6 for 1 and 4 O and OH on all the surfaces were computed on the basis of the total energies of the species as: In the case of m oxygen atoms (or OH species), we normalized Δ𝐸 𝑂 and Δ𝐸 𝑂𝐻 by m. In order to account for the effect of the surrounding water molecules in the environment, the VASPSol solvation model for water was utilized. The free energy difference was calculated as ∆G = ∆EO/∆EOH + ∆ZPE ̶ T∆S, where ∆EO/∆EOH is the reaction energy, ∆ZPE and ∆S are the changes in zero-point energies and in entropy, due to the reaction, respectively. The second and third terms in the expression are calculated by DFT and were taken by us from , where ∆ZPE -T∆S are 0.35eV for (OH* + ½H2) and 0.05eV for (O* + H2). For a general potential 𝑈 0 , ∆Gx(U0) can be calculated as A potential of 𝑈 0 = 1.23𝑉 is assumed for the reaction of Eq. 1 to be in equilibrium. From the discussion above it is clear that Δ𝐺 2 describes the free-energy change in reaction 4, and Δ𝐺 1 describes the change in reaction 3, reaction 2 is partially described by Δ𝐺 0 . In addition, the activation barrier for O2 dissociation at the surface, Ea, was taken according to the universal relation found in . This relation connects the reaction activation energy and the stability of the reaction intermediates, according to Ea = 1.8*∆EO ̶ ̶ 2.89 eV. While this relationship was established for pure surfaces, we have also extended it to our case. The values of ∆G0(U0), ∆G2(U0), ∆G1(U0), and Ea can be used to calculate the different reaction-rate constants according to: 𝑘 𝑖 ∼ 𝑘 0 𝑒 − Δ𝐺 𝑖 𝐾 𝐵 𝑇 . We can assume that the slowest step determines the overall rate of the reaction. Hence, we define the activity, A, by the logarithm of the rate constants, as A is proportional to the logarithm of the lowest reaction-rate constant. ## Structure and composition analysis Weight and atomic compositions based on SEM-EDS and XPS analyses (before and after sputtering) of Ru-contaminated catalysts are shown in Table I. Both SEM-EDS and XPS show an increasing at% of ruthenium as its experimentally planned monolayer number increases, from Ru4Pt96 (SEM-EDS) and Ru7Pt93 (XPS) for 0.15ML Ru/Pt to Ru56Pt44 (EDS) and Ru65Pt35 (XPS) for 4ML Ru/Pt. The only exception to this trend being 0.22ML Ru/Pt for which a lower at% of ruthenium than for 0.15ML Ru/Pt was detected by XPS analysis. We believe this anomaly is related to insufficient sensitivity of XPS at such low at% of ruthenium in both samples. All catalysts show higher at% of ruthenium on the surface compared to the situation after five minutes of sputtering (determined by XPS) and in the bulk (determined by SEM-EDS). The combined SEM-EDS and XPS results are an indication that ruthenium was deposited on platinum nanoparticles during the synthesis and not as separate nanoparticles. Representative TEM images and size-distribution histograms (insets) of Rucontaminated catalysts, and HiSPEC8000 are presented in Figures S1(a-i) in Supplementary Material. The average particle sizes of Ru-contaminated catalysts are similar to HiSPEC8000 up to 1ML Ru/Pt (3.7-3.9 nm) and increasing substantially to 4.5-4.6 for 2ML Ru/Pt and 4ML Ru/Pt (Table I). Several Ru/Pt catalysts show increased particle agglomeration compared to HiSPEC8000. Suspecting that the agglomeration was caused by exposure to hot reflux during the synthesis, a suspension containing HiSPEC8000 and methanol but not ruthenium salt was refluxed for the same time as during ruthenium deposition. Indeed, the resulting powder (named HiSPEC8000_M) shows significant agglomeration in TEM images (Figure S1(i) in Supplementary Material) that also manifests itself in ECSA measurements presented later. This expansion originates from ruthenium oxidation/reduction and high pseudocapacitance of hydrous ruthenium oxides . particle size and their agglomeration. It appears that 1ML Ru/Pt had a lower particle agglomeration that led to a higher ECSA value compared to other Ru/Pt catalysts with similar particle sizes. On the other hand, 0.6ML Ru/Pt had higher particle agglomeration that led to lower ECSA. ECSA values of 2ML and 4ML Ru/Pt are on the lower side compared to other Ru/Pt catalysts. This was to be expected in light of the higher amount of ruthenium contamination in these catalysts that led to higher particle sizes. ## Experimental study of ORR catalytic activity ORR polarization curves of the examined catalysts are shown in Figure 3 Assuming that the decrease in ORR activity is originating from the oxophilic nature of ruthenium that covers the platinum, we used DFT calculation of the binding energies of O and OH to estimate the ORR activity on model Run/Pt(111) surfaces that represent Ru-contaminated Pt/C catalysts. The results of the DFT studies and their correlation to experimental results will be presented in the next sections. ## Ruthenium coverage of platinum surface Initially, the surface energies of the clean Ru(0001) and Pt(111) surfaces were calculated. Our surface energy value for the Ru(0001) was found to be higher than that of the Pt(111) surface, that is 169 meV/A 2 and 95 meV/A 2 , respectively, which is close to experimental data and in agreement with other theoretical calculations from the literature . Next, the adsorption energy of Ru atoms at the clean Pt(111) was calculated. It was found that in the lowest-energy structure, ruthenium ad-atom binds to the hollow site on the Pt(111) surface, which corresponds to the ABCA Pt stacking. In addition, it was S2 and S3. It is known from the literature that there is formation of Ru monatomic layers at low coverages and bilayer islands and three-dimensional clusters at higher levels of coverage . Our surface cell was slightly too small to show the effect of multilayer formation and hence in our simulations, the ruthenium atoms typically tended to form a monolayer. A Bader charge analysis reveals that the Ru atoms tend to give some of their electrons to the Pt surface. For Ru1/Pt(111) we found that the Ru atom had a Bader charge of +0.34e. For the Ru2/Pt(111), the charge was +0.29e per Ru atom (total of +0.57e), for Ru4/Pt(111), the charge was +0.21e per Ru atom (total of +0.85), and for full coverage of Ru9/Pt(111) we found a charge of +0.11e per Ru atom (total of +1.01e for the Ru monolayer). ## O binding on Run/Pt(111) surface As found previously , and also supported by us, the O atom has a greater tendency to bind on hollow sites on both the pristine Pt(111) and Ru(0001) surfaces. On the Run/Pt(111) surfaces it was found that O atoms generally tend to bind on the Ru atoms, which can be explained by the stronger binding of O atom on the Ru(0001) (with calculated binding energy, Eb, of -5.97 eV) compared with that of binding on the Pt(111) (calculated Eb of -4.25 eV). The lowest energy structures for mO on Run/Pt(111) surfaces are presented in Figure 4. Since we examine the thermodynamic limit of lowest energy structures, the binding energy of a single oxygen atom will not be much affected by the Ru-atom coverage level, even if the coverage is one Ru atom per million Pt surface atoms, the oxygen would still tend to bind to the Ru atom. To fully account for an actual scenario, one needs to build a kinetic simulation which takes into account the adsorption on alternative surface sites. Here we use an alternative approach of saturating the surface with oxygen atoms. At some point, the next oxygen atom cannot bind to the Ru atom(s) and binds instead to a Pt surface atom, hence showing Ru-coverage-dependent behavior. On A Bader charge analysis of mO on the Ru1/Pt(111) surface was performed. At ## OH binding on Run/Pt(111) surfaces The binding of OH species to the Run/Pt(111) surfaces follows trends similar to that of the binding of O atoms, namely, OH species first bind to Ru atoms and occupy all possible sites on these atoms. However, in addition, OH can also form hydrogen bonds with either other OH or surface atoms. This can further stabilize some surface-adsorbed structures. The structures for the OH binding on the Run/Pt(111) are presented in Figure S4 in Supplementary Material. ## Binding energies for O and OH The binding energies (Eb) for mO and mOH (m = 1-4) species on the clean Pt(111) and Ru(0001) surfaces, and on the Run/Pt(111) (n = 1,2,4,9) surfaces, are presented in Figure 5. For the pristine surfaces it was found that the Eb of a single O atom on Pt(111) is weaker than on Ru(0001): -4.25 eV and -5.97 eV, respectively, which is consistent with literature results . For the Run/Pt(111) surfaces it is found that the oxygen binding energy, Eb, decreases with the number of adsorbed oxygen atoms. This trend is especially strong for the Ru1/Pt(111) (Eb decreases from -5.86 eV for a single oxygen to -4.87 eV with four oxygens) and Ru2/Pt(111) (from -6.21 to -5.10 eV) surfaces. This can be explained by the saturation of Ru sites which forces some of the oxygen atoms to be adsorbed on Pt atoms and not on the Ru atoms. This decrease in Eb still exists but is less prominent for the case of Ru9/Pt(111) (Eb changes from -6.35 to -5.92 eV), where all the O atoms are uniformly distributed above the full-coverage Ru monolayer. Here, and also on the clean Pt(111) and Ru(0001) surfaces, another mechanism, of electrostatic repulsion between the adsorbed oxygen atoms, can explain the smaller decrease in the binding energy. The Eb for mO atoms on the Run/Pt(111) is stronger than on Pt(111) and closer to the Eb of Ru(0001). This finding can be related to the strain and ligand effects on the Run/Pt(111) systems, that play an important role in controlling the surface reactivity , and is in agreement with the experimentally known oxophilic nature of ruthenium (compared to platinum). For Ru4/Pt(111), the binding energy, Eb, for 4O atoms approaches that of Ru(0001) because all the oxygen atoms tend to adsorb on Ru sites. However, it is statistically possible that one or more of the O atoms can also bind to Pt atoms at a higher energy state. We therefore considered an additional higher energy structure, defined as Ru4/Pt(111)*, where one of the O atoms is adsorbed on the Pt surface and not on the Ru4 cluster. The Eb for OH species shows a trend similar to that of the Eb of O atoms, except for the case of 4OH/Run/Pt(111), in which we found a stronger binding of 4OH relative to 3OH on most surfaces. This can be explained by the contribution of the hydrogen bonds between the hydrogen and Pt/Ru atoms on these surfaces. and 4O/OH (Figure 6 and Figure S6 in Supplementary Material) the addition of Ru atoms is followed by a decrease in Δ𝐸 𝑂 and Δ𝐸 𝑂𝐻 (i.e., increased binding of O and OH) that leads to a decrease in ORR activity. It is also evident from Figures S5 and S6, that the effect of Ru coverage on Δ𝐸 𝑂 , Δ𝐸 𝑂𝐻 and the estimated ORR activity is more pronounced for the case of 4O/OH. The reason for this is that a single O/OH will always have an available Ru site to adsorb on, while in the case of 4O/OH, the Ru sites become saturated at low Ru coverages. ## Analysis of ruthenium effect on ORR The DFT calculations presented above clearly show that indeed the oxophilic nature of ruthenium (compared to platinum), that manifests itself in increased O and OH binding, is the root cause of the inferior ruthenium ORR activity at potentials relevant to PEMFC. According to the Sabatier principle, too-strong binding of O and OH on ruthenium (compared to platinum) reduces the possibility of O and OH hydrogenation that is needed in order to complete O2 reduction to H2O. Moreover, because of its oxophilic nature, ruthenium is oxidized at much lower potentials compared to platinum. Hence, at PEM-cathode operating potentials ruthenium is oxidized, cannot adsorb O2 molecules and facilitate their reduction. Besides being effectively ORR-inactive, ruthenium deposited on platinum (as in the case of Ru-contaminated platinum studied here) masks three platinum atoms, preventing them from adsorbing O2 molecules and effectively reducing the available platinum sites (i.e., available surface area) for ORR. In Figure 7b it can be seen that the deposition of a 0. 7b). In order to answer this question, it is necessary to examine the mechanism of O2 reduction on a platinum surface and the preferred deposition of Ru sub-monolayers on Pt surfaces. Although ORR is a multistep reaction with a quite complex mechanism that is still somewhat in debate , the breaking of an O−O bond (i.e., O2 dissociation) and the formation of O−H bonds involving a four-electron process must occur in order to achieve a complete reduction of O2. Among multiple possible configurations for O2 adsorption on a platinum surface, adsorption on two adjusted platinum sites (the socalled bridge side-on) is generally favored for the promotion of O2 dissociation . As mentioned above, ruthenium adsorbed on platinum masks three platinum sites, preventing those sites from adsorbing O2 in any configuration. We shall name this type of platinum deactivation a direct deactivation. Nine platinum sites surrounding the masked three platinum sites lose some of their potential neighbors for bridge side-on adsorption. This may lead to increased probability for O2 adsorption in a less favorable configuration for O2 dissociation and thus negatively affecting ORR kinetics. We shall name this type of deactivation an indirect deactivation. Hence, one ruthenium atom has the potential to deactivate to some degree, twelve atoms of platinum and not only three atoms. We believe this to be the reason for the massive decrease in the ORR activity for 0.15ML Ru/Pt. We shall now turn our attention to the deposition of ruthenium sub-monolayers on platinum surfaces. As was mentioned previously, ruthenium deposition on platinum has a tendency to create monolayer clusters at low coverages and bilayer islands and threedimensional clusters at higher coverages (Volmer-Weber growth). It is likely that for 0.22 and 0.6 equivalent monolayers such islands will be formed, hence effectively reducing the number of platinum sites deactivated by each ruthenium atom and reducing the decrease rate in ORR activity with increased ruthenium deposition. Additional reduction of deactivated platinum sites for each ruthenium atom (and subsequent reduction in ORR activity decrease rate) is expected as a result of the overlap between the deactivated platinum sites. It is reasonable to assume that such overlap will occur at sufficiently high ruthenium coverage and will grow as ruthenium coverage is increased until the ORR activity will reach a plateau value that is similar to the ORR activity of ruthenium. We believe that the proposal described above provides ## PEMFCs and DMFCs Even though RDE ORR experiments cannot precisely predict an FC cathode polarization, they can be used to obtain a first-order approximation of it. Hence, the RDE ORR polarization data obtained during this research can be used to assess the overpotential penalty of an FC cathode originating from ruthenium contamination. To assess the added overpotential, we shall look at Figure 8, which presents masstransport, background and iR-corrected Tafel plots for HiSPEC8000 and 0.15ML Ru/Pt. The latter was chosen because of the similarity of its ruthenium content to ruthenium contamination found by Piela et al. . In order to approximate the overpotential, we compare the potentials of both RDEs at the same current densities that polarize the RDEs to typical potential ranges of DMFC and PEMFC cathodes during FC operation. The catalyst loadings on RDEs for both catalysts were similar, ~20 µgPGM cmgeo -2 , hence any overpotential for 0.15ML Ru/Pt can be associated with ruthenium contamination. Looking at the potential range typical to operating DMFC cathode, ~0.8V , it can be seen that the approximated overpotential penalty (marked as ηRu at Figure 8) for 0.15ML Ru/Pt is roughly 45mV. Assuming a cell voltage of 0.45V during DMFC operation, this penalty will translate to roughly 10% decrease in power density. The overpotential penalty grows to roughly 75 mV over a potential range typical to an operating PEMFC cathode, ~0.65V. Taking into account the small overpotential of a PEMFC anode and assuming a cell voltage of 0.6V during operation, we can approximate the penalty in power density at slightly more than 10%. ## Conclusions In this work we studied the effect of ruthenium contamination on ORR on platinum, focusing on the implications on PEMFC and DMFC cathode-relevant potentials. To obtain our objective, a commercial 50%Pt/C catalyst was contaminated by precisely known amounts of ruthenium. The contamination range varied from relatively low contamination, equivalent to 0.15ML of ruthenium, to severe contamination, equivalent to 4ML of ruthenium. The contaminated catalysts were examined with the use of physico-chemical methods to verify and quantify ruthenium contamination. It was found that ruthenium was deposited on the surface of platinum nanoparticles in a core-shell-like structure without creating separate ruthenium nanoparticles on the carbon support. Hupd-stripping region analysis showed a gradual suppression of the platinumcharacteristic hydrogen region that was correlated with increased ruthenium coverage of the platinum. While Hupd-stripping peaks from (110) and (100) platinum planes could be seen for 0.15ML Ru/Pt, only small remnants of platinum-characteristic features could be seen for 1ML Ru/Pt, while higher ruthenium contamination showed ruthenium-characteristic features. RDE ORR polarization showed the negative effect of ruthenium on ORR at potentials relevant to the PEMFC/DMFC cathode, exhibiting a progressive shift of onset potentials and mixed-kinetics/mass-transport regions toward more negative potentials and decrease of ORR specific activity with increasing contamination of ruthenium. However, in contrast to gradual changes in Hupd features, ORR specific activity showed a drastic ~45% decrease already for 0.15ML Ru/Pt and, in general, a pseudoexponential decay with increased ruthenium coverage. With the use of our DFT studies and previously published experimental results , we showed that the negative effect of ruthenium on ORR could be attributed to the masking of platinum sites by adsorbed ruthenium atoms, as well as to the oxophilic nature of ruthenium, that was found to bind O and OH much more strongly than platinum, reducing its effectiveness in ORR catalysis. Potentially unfavorable configuration of O2 adsorption on platinum sites that immediately surround the ruthenium-masked platinum sites, the formation of bilayer islands and three-dimensional clusters and overlap between the deactivated platinum sites were proposed as a possible explanation for the exponential-like decay of ORR on ruthenium-contaminated catalysts. The results of this research stress the negative impact of ruthenium dissolution from the anode and its crossover to the cathode in reformate-based PEMFCs and DMFCs. The dramatic reduction in ORR activity -almost 50% -and the subsequent reduction in power density that accompanied the smallest ruthenium contamination, emphasizes the need for development of PtRu catalysts with higher stability, Ru-pre-leaching procedures during catalyst/GDE/MEA preparation and control of anode potential during FC operation in order to avoid or at least to reduce the performance penalty caused by ruthenium crossover.

chemsum

{"title": "Study of Ruthenium-Contamination Effect on Oxygen Reduction Activity of Platinum-based PEMFC and DMFC Cathode Catalyst", "journal": "ChemRxiv"}

boehmite_nanofibers_as_a_dispersant_for_nanotubes_in_an_aqueous_sol

## Abstract: By exploiting the dispersibility and rigidity of boehmite nanofibers (BNFs) with a high aspect ratio of 4 nm in diameter and several micrometers in length, multiwall-carbon nanotubes (MWCNTs) were successfully dispersed in aqueous solutions. In these sols, the MWCNTs were dispersed at a ratio of about 5-8% relative to BNFs. Self-standing BNF-nanotube films were also obtained by filtering these dispersions and showing their functionality. These films can be expected to be applied to sensing materials. TextOne-dimensional materials such as nanofibers, nanotubes, and nanowires have received a great deal of attention in the field of nanotechnology.[1] Depending on their composition, these materials can exhibit interesting properties that are not found in bulk materials owing to their specific shape, which consists of a high aspect ratio and a narrow diameter composed of a limited number of molecules. Because of their unique characteristics, these materials have been the subject of extensive research motivated by not only fundamental scientific interest but also the potential practical applications such as electronic materials, sensors, and fillers for composite materials. Nanofibers and nanotubes can be produced by various methods, such as electrospinning, decomposition of biological materials, and hydrothermal treatment. A common problem encountered while handling these materials is aggregation. Similar to the case of quantum dots and nanoparticles, which readily form secondary particles and lose their interesting physical properties, nanofibers and nanotubes are susceptible to bundle formation. One solution for maintaining the unique functions of nanofibers and nanotubes in macroscale is to disperse them in liquids or polymers, which necessitates the use of dispersants and surface modification, although the optimal method must be determined for each material. For example, various techniques have been proposed for the dispersion of carbon nanotubes in a liquid, such as surface oxidation, polymer coating, and the use of surfactants or an appropriate dispersion medium. Each of these methods for dispersing one-dimensional materials has both advantages and disadvantages, and the most appropriate choice can depend on the specific application. Our group has studied porous materials by using (pseudo)boehmite nanofibers (BNFs) composed of aluminum oxide hydroxide (AlOOH), which has a high aspect ratio of 4 nm in diameter and several micrometers in length. Nanofibers prepared by the solvothermal method can be stably dispersed in concentrated aqueous acetic acid for several months. However, when a base or phosphoric acid is added under appropriate conditions, the nanofibers form a three-dimensional network without bundling, and the dispersed sol becomes a gel. By supercritical drying of the wet gel obtained from this reaction, ultralow-bulk-density transparent porous monoliths (aerogels of 5 mg cm −3 or less) can be obtained, which have potential applications as optical materials. By dispersing functional materials such as fluorescent molecules or nanoparticles in the gel before freeze-drying, a transparent monolith (cryogel) containing the functional material dispersed in the nanofiber network can also be obtained. These materials are expected to apply to the fabrication of sensing materials. As the nanofiber dispersion sol is fluid, it is also possible to disperse the functional material using ultrasonication. However, because the BNFs themselves are rigid materials, they are expected to hinder the aggregation of the functional material. In particular, when the functional material is also a one-dimensional material, the movement in the dispersion is mostly restricted. Indeed, the addition of multi-walled carbon nanotubes (MWCNT) to a BNF dispersion was found to afford a relatively stable dispersion. The results of this study demonstrate the applicability of already dispersed one-dimensional materials as dispersants. To investigate the dispersibility several BNF dispersions with different concentrations were prepared, and MWCNTs were then added and dispersed by sonication. The obtained BNF-MWCNT sols remained black even after ultracentrifugation at 10,000 ×g for 5 minutes, which indicated that the MWCNTs were still dispersed (Figure 1). In contrast, when the MWCNTs were added to water or aqueous acetic acid, ultracentrifugation resulted in pelletization of the MWCNTs and the formation of a colorless and transparent supernatant. These results confirmed that the presence of BNF affected the dispersion state of the MWCNTs under aqueous conditions. Table 1 shows the MWCNT concentrations in the supernatants obtained after ultracentrifugation of BNF-MWCNT dispersions containing various BNF concentrations. MWCNTs were dispersed at a weight ratio of 5-8% relative to BNF. A lower BNF concentration in the dispersion resulted in a reduction in the amount of dispersible MWCNT and an increase in the error. Figure 1c shows transmission electron microscopy (TEM) images obtained after casting the obtained BNF-MWCNT dispersion on a grid. The MWCNTs were kept in the dispersed state by multiple BNFs, and no noticeable aggregation was observed. Interestingly, the addition of the MWCNTs was found to lead to a change in the viscosity relative to the BNF dispersion. In the system with the highest viscosity (sample X2.5 in Table 1), the viscosity decreased from 76 Pa s to 49 Pa s. This reduction was ascribed to the homogeneous dispersion of the MWCNTs between the BNFs, which weakened the hydrogen-bonding interactions. From these results, it is considered that the dispersed MWCNTs became physically entangled between the rigid BNFs in the dispersed state. Even when this BNF-MWCNT dispersion was allowed to stand for one month, almost no change in the dispersion state was observed. In a similar manner to paper, the structure of which consists of intertwined cellulose fibers, one-dimensional materials such as nanofibers, nanotubes, and nanowires are known to be capable of forming films upon removal of the solvent by filtration and evaporation. With respect to BNFs, Kodaira et al. fabricated free-standing films with visible-light reflectivity and thermally insulating properties by preparing nanofiber bundles under appropriate conditions and subsequently depositing them. In the case of the BNF-MWCNT dispersion, it was possible to prepare a black self-supporting film by vacuum filtration through a polycarbonate membrane filter (Figure 2). Scanning electron microscopy (SEM) images confirmed that this film had a structure in which the BNFs and MWCNTs were arranged in layers. As BNFs are composed of aluminum oxide hydroxide and act as an insulator, the obtained film exhibited only the low electrical conductivity of the MWCNTs. Instead of MWCNTs, polydiacetylene nanotubes (PDANTs), which have been reported as a vapochromic material, could also be dispersed with the BNTs to prepare a similar film. Although PDANTs themselves can be dispersed in water or alcohols, the resulting composite films were stable to immersion in these solvents. This observation indicated that the PDANTs in the film were successfully confined in a dispersed state with the BNFs. As with the BNF-MWCNT film, the PDANT-MWCNT film was also found to have a structure in which the nanofibers formed overlapping layers. Exposure of this composite film to water or 2-propanol vapor revealed that it had vapochromic properties (Figure 3, Movies S1 and S2). Although the color change was small and different from that which occurs for pure PDANTs, the basic functionality remained even after film formation. Therefore, the use of BNF as a dispersant/carrier is expected to be a promising method for fabricating functional films when additional one-dimensional materials with unique functionalities are synthesized in the future. By changing the pH of the BNF dispersion, wet gels can be obtained instantaneously. We have reported studies to prepare ultralow-density transparent porous monoliths by drying those BNF gels. In the case of the BNF-MWCNT and BNF-PDANT binary dispersions, wet gels were also obtained without conspicuous aggregations in the internal structure. However, when these wet gels were subjected to supercritical drying, considerable shrinkage occurred, and the reproducibility of the obtained aerogels was low. This was because the addition of the nanotubes reduced the bonding between the BNFs and the networks became non-uniform. At present, the formation of monoliths based on BNF composites remains difficult. In summary, MWCNTs were successfully dispersed in aqueous solution by exploiting the dispersibility and rigidity of BNFs. In these dispersions, the MWCNTs were dispersed at a ratio of about 5-8% relative to BNF. Self-standing BNF-nanotube films were also obtained by filtering these dispersions, allowing the introduction of functionalities such as chromic properties depending on the type of dispersed nanotubes. These dispersions can potentially be applied in the fabrication of functional films and used as inks and fillers. Unlike existing methods such as chemical modification, the application of this method to one-dimensional materials such as single-walled carbon nanotubes, which are susceptible to bundle formation, remains challenging, and further research is required. In future work, the described method for physically creating a dispersed state by utilizing the geometric properties of a one-dimensional material is expected to lead to the development of new dispersants.

chemsum

{"title": "Boehmite Nanofibers as a Dispersant for Nanotubes in an Aqueous Sol", "journal": "ChemRxiv"}

nanofiber_network_with_adjustable_nanostructure_controlled_by_pvp_content_for_an_excellent_microwave

## Abstract: Carbon nanofibers were widely utilized to improve microwave absorption properties since they are a promising lightweight candidate. Adjustable conductive nanostructures of carbon nanofibers were synthesized by electrospinning technique. The conductive network is controlled by the polyvinyl pyrrolidone (PVP) content due to the special hygroscopicity of PVP. The increased adhesive contacts of nanofibers provide more transmission paths for electrons to reduce the effect of air dielectric. Satisfactorily, the carbon nanofibers that carbonized from the polyacrylonitrile (PAN) and PVP (the mass ratio is 6:4) show excellent microwave absorption performance. The minimum reflection loss (RL) value is −51.3 dB at 15.2 GHz and the maximum effective absorption frequency width (<−10 dB) is 5.1 GHz with the matching thickness of only 1.8 mm. Thereby, we believe that this research may offer an effective way to synthesize lightweight carbon nanofibers microwave absorbents. In modern society, electromagnetic (EM) absorbents have become an indispensable part of information equipment, which can reduce the harm to the surroundings and the humans. The principle of EM absorbers is making microwave transform into other types of energy to attenuate EM wave . Traditional microwave absorption materials like metal as well as alloys, ferrites and all varieties of these composites with strong microwave absorption abilities and board effective frequency width develop rapidly. However, it is also significant for actual demand to reduce the density and weight of microwave absorbers at the same time . As famous lightweight materials, carbon nanomaterials have been achieved tremendous attention. Various classifies of carbon nanoparticles such as carbon nanotubes, carbon cloths and so on are of many wonderful characters such as large areas, low cost, good stability and great electrical conductivity . Thereby, carbon nanomaterials are quite befitting for EM wave absorption field 11,12 . Among numerous carbon nanomaterials, carbon nanofibers are of huge interest for EM wave absorption due to their one-dimensional nanostructure that can form conductive network . For instance, Liu et al. successfully fabricated a kind of helical CNFs coated-carbon fibers through catalytic chemical vapor deposition. The minimum RL value was −32 dB at 9.0 GHz and the widest effective frequency width was 9.8 GHz with only 15% filler ratio 16 . Porous carbon nanotubes decorated carbon nanofibers were also achieved with the minimum RL value of −44.5 dB at 10.7 GHz as well as the broad effective frequency width of 7.1 GHz 17 . Chu et al. compared the microwave absorption abilities of different diameters, they believed that complex permittivity improved along with the decreasing diameters since it contributed to the conductive network 18 . Based on their researches, we can discover that the microwave absorption performance closely depends on their design of one-dimensional nanostructures on carbon nanofibers. In principle, morphology change of carbon nanofibers could dramatically influence the transferring path of the electrons as well as the construction of the conductive network 19 . Accordingly, the permittivity and polarization process would be controlled in the range of testing frequency artificially. Furthermore, electrospinning fiber technique can be used accurately to obtain diversiform nanofibers, which is a good choice to synthesize carbon nanofibers 20,21 . In this paper, we added PVP as a structural adjuster to modify PAN based carbon nanofibers by changing the PVP content, followed by electrospinning and next annealing process. PVP, as a high-molecular compound, it is liable to dissolve in water. In addition, the more average molecular weight of PVP, the more chance of it to be agglutinating 22 . Considering with these intrinsic qualities of PVP, we easily gain nanofiber network with adjustable nanostructure controlled by PVP content. Gratifyingly, the microwave absorption performance is enhanced, too. For the carbon nanofibers carbonized from the PAN and PVP (the mass ratio is 6:4 and the filler ratio is 20%), the RL value is −51.3 dB at 15.2 GHz and the maximum effective absorption frequency width (<−10 dB) is 5.1 GHz with only 1.8 mm. This work provides a novel strategy to build the conductive network of carbon nanofibers through adjusting the content of PVP, which may be impulse the development of the lightweight microwave absorbers. ## Results Figure 1 is the entire synthesis schematic diagram of our nanofibers conductive network. The precursor solution of PAN and PVP was transformed into nanofibers via a representative electrospinning technology. After being treated at 800 °C for 3 h under N 2 atmosphere inside tube furnace, carbon nanofibers with adjustable nanostructure were finally fabricated. Detailed nanostructures of different nanofibers are detected using SEM and TEM methods. Obviously, nanofibers of PAN/PVP samples show typical one-dimensional morphology of electrospun fibers in Fig. 2(a-c). There surfaces are smooth. In addition, we can easily see that the nanofibers become more and more thicker since the quality of PAN is constant and the quality of PVP is larger. The average diameters of PAN/PVP-7/3 sample, PAN/ PVP-6/4 sample and PAN/PVP-5/5 sample are 167 nm, 277 nm and 444 nm, respectively. After calcined at 800 °C under N 2 atmosphere, it can be seen that some of nanofibers are bent and bonded in Fig. 2(d-f). Notably, this phenomenon is relatively clearer with the increasing content of PVP. Due to the strong hygroscopicity of PVP, the solvent doesn't volatilize very well so that the nanofibers change into a sticky one. However, nanofibers of PVP (500) sample present a fracture situation. It is very difficult to identify the one-dimensional nanofibers in Fig. 1(f). Moreover, Fig. 2(g,h) displays the TEM pictures of PVP (333) sample. It should be noticed that there is typical carbon margin in Fig. 2(h). Also, physical photos of the white PAN/PVP-6/4 sample and the black PVP (333) sample are provided in Fig. 2(i). The PAN/PVP nanofibers were also heated at 200 °C, 400 °C and 600 °C. As can be seen from Fig. 3, all of these samples present a curved one-dimensional nanofibers morphology. The nanofibers are getting thicker with the increasing content of PVP. Besides, the fusing phenomenon become serious when we added more PVP. The RL values of different carbon nanofibers are provided in Fig. 5(a-f) using the following formulas : www.nature.com/scientificreports www.nature.com/scientificreports/ where Z in is the input impedance, Z 0 is the impedance of free space, f is the frequency, d and c represent the thickness and the velocity of light, respectively. When the thickness increases, the RL values shifts to lower frequency, indicating that EM wave absorption frequency and the thickness of absorbers can be modulated. Among PVP (214), PVP (333) and PVP (500) samples, PVP (214) sample shows the poorest microwave absorption abilities. Different microwave absorption performances have something to do with the content of PVP. As the RL values below −10 dB, effective microwave absorbers will make 90% microwave attenuated 26 . Thereby, when matching thickness are 1. 6(a,b). As seen in Fig. 6, PVP (333) sample has more boarder effective frequency width while the microwave absorption ability of PVP (500) sample is stronger. The electromagnetic parameters (complex permittivity: ε = ε′ − jε″; the complex permeability: μ = μ′ − jμ″) of carbon nanofibers are tested by a network analyzer. Herein, magnetic properties of carbon nanofibers can be neglected since the dielectric loss more critical. Obviously, permittivity shows a sustainable growth with the increasing content of PVP. The ε′ of PVP (214) sample is the lowest among these carbon nanofibers. The ε′ of PVP (333) sample is from 13.0 to 7.2 while PVP (500) sample descends from 14.6 to 8.2. The ε″ values of PVP (333) sample and PVP (500) sample are higher than that of PVP (214) sample. In Fig. 7, some small fluctuations of ε″ curves arise from the multiple nature resonances. According to pervious reports, it was found that the www.nature.com/scientificreports www.nature.com/scientificreports/ complex permittivity would improve with the increasing average diameters of the PAN based carbon nanofibers 18 . Nevertheless, our nanofibers do not suit this situation. This phenomenon may refer to their special adhesive nanostructures. Namely, average diameters could not be the main reason that effects the complex permittivity. As we all know, if there are some gaps between nanofibers, air will reduce the dielectric of nanofibers, bringing about lower permittivity values. Therefore, the adhesive carbon nanofibers would be the dominant reason that increases the dielectric constant. Thanks to the function of PVP, the more contact carbon nanofibers, the electrical conductivity is higher. As studied by other groups, dielectric loss can be caused by conductive loss as well as by polarization loss involved in the relaxation process 27 . Moreover, different polarization process including ionic polarization, electronic polarization and dipole polarization can be considered in this cause 28 . However, ionic polarization and electronic polarization always happen at high frequency such as 10 3 -10 6 GHz, dipole polarization should be the main reason that leads to the polarization process. Using the Eq. ( 3), dielectric loss degree can be evaluated in Fig. 7(d) : e The results manifest that the dielectric loss tangent of PVP (214) sample is around 0.1. The dielectric loss tangent curve of PVP (333) sample show a rising trend followed by a drop condition. PVP (500) sample show a raising trend. So, we suggest that both PVP (333) sample and PVP (500) are of higher dielectric loss. The RL values and frequency dependence on electromagnetic parameters of different sample filler ratios are displayed in Figs 8 and 9. Accordingly, the permittivity goes up with the increasing sample filler ratio. When the sample filler ratios are 30% and 40%, the microwave absorption abilities of PVP (333) sample are much better than that of other two samples. Interestingly, RL values decreases with the increasing content of PVP when the Impedance matching is an important factor to the microwave absorbers. If the nanofibers are of good impedance matching properties, more microwave will go into the nanofibers and will be attenuated 32,33 . To synthesize a satisfactory microwave absorber, it is the key to improve the microwave absorption performance or board the effective frequency width. Figure 10 www.nature.com/scientificreports www.nature.com/scientificreports/ In Fig. 11, PVP (214) sample shows the lowest average conductivity while the average conductivities of PVP (333) sample and PVP (500) can be greatly improved at S, C, X and Ku frequency range with the increasing content of PVP. The electronic impedance spectrum (EIS) for PVP (214), PVP (333) and PVP (500) were measured to evaluate their conductivity. As can be seen in the Nyquist plot in Fig. 11(b), the theoretical prediction of electrical conductivity in the following order PVP (214) < PVP (333) < PVP (500), indicating the higher conductivity with the increasing content of PVP. In this case, more contact positions of nanofibers will make electrons have multiple paths to transfer, which creates the increased conductivity. The carbon conductive network is fabricated simultaneously. Table 1 gives some data of similar carbon nanofibers about their microwave absorption performance. In Table 1, we can easily conclude that all kinds of carbon nanofibers show enhanced microwave absorption abilities. Compared with carbon fibers that heap up one by one, nanofibers with some bonded places are more beneficial to make electrons transfer. When this type of carbon nanofiber is put in EM field, electron is likely to transfer with many paths. Hence, among these different carbon nanofibers, our adjustable shaped carbon nanofibers display the minimum RL value and broader effective bandwidth with the thinnest thickness, which implies that it is very potential to be used as lightweight microwave absorber. ## Conclusion Carbon nanofibers with adjustable nanostructure were gained by electrospinning method. When calcined at the same temperature (800 °C), the PVP content could be the main reason that influence the microwave absorption abilities. The increased adhesive contacts of nanofibers create the more paths for electrons to transfer and can reduce the effect of air dielectric. Furthermore, appropriate impedance matching is also responsible for the excellent microwave absorption abilities. In detail, the minimum RL value is −51.3 dB at 15.2 GHz and the maximum effective absorption frequency width (<−10 dB) is 5.1 GHz both with 1.8 mm. Our work may be helpful for the exploitation of lightweight microwave absorbents in near future. ## Method Materials. All chemicals and reagents are supplied by business supporters and they are all without pretreated. These chemicals are polyacrylonitrile (PAN, M w = 150 000) and polyvinylpyrrolidone (PVP, M w = 5 800). N, N-dimethylformamide (DMF) is needed as well. Synthesis of carbon nanofibers. Carbon nanofibers were obtained by electrospinning technology and following high-temperature carbonization. First, 0.5 g of PAN was added into 5 mL DMF with continuous magnetic stirring. Then, a certain amount of PVP was also introduced to this system with stirring for 24 h to achieve a transparent and syrupy liquid. The mass ratio of PAN and PVP is 7:3, 6:4 and 5:5, respectively. The precursor nanofibers are called PAN/PVP-7/3, PAN/PVP-6/4 and PAN/PVP-5/5, respectively. Second, this liquid was sucked into a 5 mL plastic syringe equipped with a stainless needle. The next process was the electrospinning. Specifically, the voltage parameters were 20 kV, the collection distance was 15 cm and the pushing speed was 0.5 mL/h. After that, the precursor PAN/PVP nanofibers were dried in a vacuum oven for a day and the PAN/PVP nanofibers were calcined at 800 °C for 3 h surrounded with N 2 atmosphere. Herein, the average rate was of 2 °C per minute. And the carbon nanofibers are marked as PVP (214), PVP (333) and PVP (500), respectively. When these PAN/PVP nanofibers were heated at 200 °C, 400 °C and 600 °C, these samples are named as PVP (214-200), PVP (333-200), PVP (500-200), PVP (214-400), PVP (333-400), PVP (500-400), PVP (214-600), PVP (333-600) and PVP (500-600), respectively. Characterization. The crystal structure was measured by X-ray diffraction (XRD) under Cu Kα radiation. The SEM (Hitachi-S4800) as well as TEM (Tecnai G2 F30) were used to observe the microtopography of these nanofibers. A vector network analyzer (Agilent, N5244A) was applied to test the electromagnetic parameters in the range of 2-18 GHz. The samples (20% filler ratio) and paraffin (80% filler ratio) were crushed into a cylinder. The inner diameter and the outer diameter were 3.0 mm and 7.0 mm. When changing the filler ratio, the samples (30%, 40% and 50% filler ratio) are called as PVP (214, 30), PVP (333, 30), PVP (500, 30), PVP (214, 40), PVP (333, 40), PVP (500, 40), PVP (214, 50), PVP (333, 50) and PVP (500, 50), respectively.

chemsum

{"title": "Nanofiber network with adjustable nanostructure controlled by PVP content for an excellent microwave absorption", "journal": "Scientific Reports - Nature"}

a_core–brush_3d_dna_nanostructure:_the_next_generation_of_dna_nanomachine_for_ultrasensitive_sensing

## Abstract: A highly loaded and integrated core-brush three-dimensional (3D) DNA nanostructure is constructed by programmatically assembling a locked DNA walking arm (DA) and hairpin substrate (HS) into a repetitive array along a well-designed DNA track generated by rolling circle amplification (RCA) and is applied as a 3D DNA nanomachine for rapid and sensitive intracellular microRNA (miRNA) imaging and sensing.Impressively, the homogeneous distribution of the DA and HS at a ratio of 1 : 3 on the DNA track provides a specific walking range for the DA to avoid invalid and random self-walking and notably improve the executive ability of the core-brush 3D DNA nanomachine, which easily solves the major technical challenges of traditional Au-based 3D DNA nanomachines: low loading capacity and low executive efficiency. As a proof of concept, the interaction of miRNA with the 3D DNA nanomachine could initiate the autonomous and progressive operation of the DA to cleave the HS for ultrasensitive ECL detection of target miRNA-21 with a detection limit as low as 3.57 aM and rapid imaging in living cells within 15 min. Therefore, the proposed core-brush 3D DNA nanomachine could not only provide convincing evidence for sensitive detection and rapid visual imaging of biomarkers with tiny change, but also assist researchers in investigating the formation mechanism of tumors, improving their recovery rates and reducing correlative complications. This strategy might enrich the method to design a new generation of 3D DNA nanomachine and promote the development of clinical diagnosis, targeted therapy and prognosis monitoring. ## Introduction In living systems, some highly complicated and hierarchical machines have evolved to perform signifcant biological processes with remarkable precision and efficiency. Inspired by natural ingenuity, diverse DNA-based nanodevices have been created to convert chemical energy into mechanical motion, holding great promise for intelligent drug delivery, disease diagnosis and biosensing analysis. Recently, threedimensional (3D) DNA nanomachines based on gold nanoparticles (AuNPs) have attracted extensive attention due to their improved abilities in simulating complex biological operation compared with one-dimensional (1D) 25,26 or twodimensional (2D) DNA nanomachines. 27,28 However, the limited amount of track and the disordered immobilization of the DNA components (DNA walker and substrate) with a heterogeneous nano (Au)-bio (DNA) interface resulted in the low loading capacity of DNA components and the invalid movement of the DNA walker, further restricting the whole executive ability and efficiency of the Au-based 3D DNA nanomachines. On account of these disadvantages, it has long been a challenging goal to develop a new type of 3D DNA nanomachine that could not only greatly enhance the loading capacity of the DNA components in an orderly manner, but also signifcantly improve the executive ability and efficiency. Herein, a core-brush 3D DNA nanomachine with highly loaded and integrated properties was developed for intracellular microRNA (miRNA) imaging and ultrasensitive sensing with rapid kinetics (Scheme 1). Firstly, the extremely long DNA track cross-linked on a magnetic nanobead (MNB) was generated by rolling circle amplifcation (RCA), in which the locked DNAzyme walking arm (DA) and electrochemiluminescent signal tag Rubpy (Ru) labeled hairpin substrate (HS) at a ratio of 1 : 3 were orderly assembled into the DNA track with a repetitive array to construct the core-brush 3D DNA nanomachine (PART A). According to PART B, once the target miRNA-21 was introduced into the 3D DNA nanomachine, the hybridization of miRNA-21 with a locking strand (L) via strand-displacement reaction could release the active DA from the locked DA (L-DA) to hybridize with the adjacent HS. Fueled by the DNAzymeinduced ribonucleotide hydrolysis, the HS was then cleaved by the active DA in the presence of cofactor Mn 2+ to release the Rulabelled segment (S). When S was captured by the DNA hairpin (H) onto a gold nanoparticle-assembled glassy carbon electrode (depAu/GCE) through the Au-S bond, a highly intense ECL signal was generated for ultrasensitive detection of miRNA-21 with a detection limit as low as 3.57 aM (PART C). Impressively, when the DA and HS-1 labelled with a fluorophore (FAM) and quencher (BHQ) were used to assemble our core-brush 3D DNA nanomachine (PART D), the intracellular miRNA-21 could activate autonomous motions of the DA to cleave the HS and release the FAM-labelled segment away from BHQ in the presence of the cofactor Mn 2+ , leading to the quick fluorescence recovery of FAM within 15 min for sensitive imaging of lowabundance target miRNA-21. Compared with the limited amount of track in the Au-based 3D DNA nanomachine, the core-brush 3D DNA nanomachine not only provided a specifc walking range for the DA to avoid inactive and stochastic selfwalking, but also greatly increased the loading capacity and local concentrations of DNA components to signifcantly enhance the executive efficiency of the core-brush 3D DNA nanomachine. Thus, this strategy gives impetus to exploit highperformance 3D DNA nanomachines for specifc biological applications in complex cellular environments, such as biodiagnostics and bioanalysis. ## Results and discussion Characterization analysis of the core-brush 3D DNA nanomachine Polyacrylamide gel electrophoresis (PAGE) was employed to estimate the construction of the core-brush 3D DNA nanomachine. According to Fig. 1A at the bottom band of lane 11, but not in lane 12, which demonstrated that miRNA-21 could active the proposed 3D DNA nanomachine to generate numerous S. Then, we also employed atomic force microscopy (AFM), UVvis spectroscopy, zeta potential analysis and dynamic light scattering (DLS) to verify the formation of the core-brush 3D DNA nanomachine. According to Fig. 1B and S1A, † the AFM characterization of the MNB core exhibited relatively smooth surfaces with a height of 66 nm approximately. After the accomplishment of the core-brush 3D DNA nanomachine, the apparent DNA brushes were observed on the surface of the core (Fig. 1C and S1B †) with an obvious increase in the height (70 nm). Likewise, the MNB core did not have any peak in the UV-vis absorption spectra (Fig. 1D, black line). Accompanied by the successful construction of the core-brush 3D DNA nanomachine, a very distinct characteristic deoxynucleotide absorption peak located at 260 nm was observed (Fig. 1D, red line). Moreover, the corebrush 3D DNA nanomachine possessed a decreased zeta potential of 26.41 mV compared with the MNB core of 10.57 mV owing to the negative charges of the DNA (Fig. 1E). As illustrated in Fig. 1F, the hydrodynamic diameter of the MNB core and corebrush 3D DNA nanomachine measured by DLS were around 68.5 nm and 233 nm, respectively. The hydrodynamic diameter of the core-brush 3D DNA nanomachine was bigger than the actual size owing to the formation of a hydrogen bond between the hydroxyl of H 2 O and DNA. The remarkable comparisons of the above results suggested successful formation of the proposed core-brush 3D DNA nanomachine. ## Execution efficiency of the designed core-brush 3D DNA nanomachine To study the execution efficiency of the designed core-brush 3D DNA nanomachine (Fig. 2A) in contrast with the conventional 3D DNA nanomachine based gold nanoparticles (AuNPs) (Fig. 2B), we used fluorophore (FAM) and quencher (BHQ)labeled HS-1 or FAM-labeled HS-2 to construct the proposed 3D DNA nanomachine and Au-based DNA nanomachine, respectively. In the presence of the cofactor Mn 2+ , the target miRNA-21 could activate the DA to hybridize and cleave the HS for releasing the FAM-labelled segment (S) away from the quencher BHQ or AuNPs and obtaining the prominent fluorescence recovery of FAM. The real-time fluorescence of FAM for the proposed core-brush 3D DNA nanomachine showed a speedy fluorescence intensity growth and reached saturation about 900 s (Fig. 2C, curve a). In contrast, the typical Au-based 3D DNA nanomachine demonstrated a very slow fluorescence signal growth and the fluorescence saturation plateau did not appear within 2500 s (Fig. 2C, curve b). To calculate the initial rate of the DNA nanomachine, we verifed the relationship between the fluorescence intensities and reaction time of the proposed 3D DNA nanomachine and the Au-based 3D DNA nanomachine. According to Fig. 3D, the fluorescence intensities shown the desirable linear relationships with the reaction time in the frst 250 s. The regression equations of the proposed core-brush 3D DNA nanomachine (Fig. 2D, curve a and linear ft for curve a) and the Au-based 3D DNA nanomachine (Fig. 2D, curve b) were I ¼ 6.064t + 77.98 and I ¼ 0.7522t + 311.3, respectively. Based on the above results and related calculations, we found that the initial rate of our 3D DNA nanomachine (8.03 10 11 M s 1 ) was more than 7-fold higher than that of the control Au-based 3D DNA nanomachine (1.11 10 11 M s 1 ). Moreover, by frst-order derivation of the real-time fluorescence increase curves of Fig. 2C, we could obtain the reaction rate curves of the DNA nanomachine as shown in Fig. 2E. The reaction rate of our designed 3D DNA nanomachine was extraordinarily high in the frst 500 s and then became very low or even went to zero (Fig. 2E, curve a). In comparison, the reaction rate and the overall efficiency of the Au-based 3D DNA nanomachine were relatively low (Fig. 2E, curve b). Collectively, these results demonstrated that the core-brush 3D DNA nanomachine possessed obviously higher execution efficiency than the Au-based 3D DNA nanomachine on account of the high loading and integration of the DNA components in an orderly manner on the programmable DNA track. ## Feasibility investigation of the core-brush 3D DNA nanomachine The core-brush 3D DNA nanomachine was composed of multiple components, and thus each component might have a potential impact on its overall executive efficiency. Firstly, we designed a control unassembled nanomachine that only contained a free DA and FAM and BHQ-labeled HS-1 at a ratio of 1 : 3 to confrm that the DA tended to walk along the HS-1 immobilized on the DNA track rather than the scattered HS-1 in the solution (Fig. 3A). Then, to verify that the highly executive efficiency of the proposed 3D DNA nanomachine was due to the ordered array of the DA and HS-1 on the programmed DNA track, the second control by mixing scattered DA with HS-1@MNB was designed to construct the DA-unset nanomachine in Fig. 3B. In this manner, the specifc walking range for each DA was uncertain. Moreover, to ensure that the intense growth of fluorescence intensity was due to the highly loaded and integrated mode of our designed 3D DNA nanomachine, a control unintegrated nanomachine that only included a single DNA track assembled with a DA and HS-1 at 1 : 3 was further designed as shown in Fig. 3C. According to Fig. 3D, the fluorescence signal increase of the unassembled nanomachine (blue curve a) and the DA-unset nanomachine (black curve b) did not reach the saturation state in 2500 s. According to red curve c, although the unintegrated nanomachine reached reaction equilibrium quickly, its saturated fluorescence intensity was relatively weak. As depicted in Fig. 3E, the reaction rate of the unassembled nanomachine (curve a) and the DA-unset nanomachine (curve b) was extraordinarily low and even went to zero. Besides, the initial rate of the unintegrated nanomachine (8.60 10 11 M s 1 ) was roughly close to that of our proposed 3D DNA nanomachine (Fig. 3F). These results might be explained by effective collision theory that the reactants have to collide effectively with each other to make the reaction occur, and the concentrations of the reactants DA and HS-1 are proportional to the collision frequency. In the control unassembled nanomachine, the effective local concentrations of the DA and HS-1 were lower than that of being assembled together on the programmable DNA track, leading to low collision frequency and further decreasing the reaction efficiency. This result suggested that the programmable DNA track was a critical component to improve the effective local concentrations of reactants and further enhance the executive ability of the DNA walking nanomachine. Then, we thought that the chaotic reaction of scattered DA and HS-1 in the DA-unset nanomachine would reduce the utilization of DNA molecular components due to the uncertain walking range for each DA. In this manner, some DA-unset nanomachines might not hybridized with scattered DA, so these machines would not be in the operating state. Therefore, orderly assembling the DA and HS-1 into the programmable DNA track was the key to improving the utilization of DNA components. Furthermore, we also found that the saturated fluorescence intensity of the unintegrated nanomachine was weaker than that of our core-brush 3D DNA nanomachine. This result suggested that the highly integrated mode might greatly increase the effective local concentrations to further enhance the total fluorescence intensity. In principle, by comparing with the above control DNA nanomachines, our 3D DNA nanomachine possessed higher loading capacity and excellent executive efficiency, which gave impetus to ultrasensitive detection and rapid imaging of specifc miRNA in cancer cells. ## Application of the 3D DNA nanomachine for sensitive detection of miRNA-21 The proposed core-brush 3D DNA nanomachine was employed to construct an electrochemiluminescence (ECL) biosensing system for sensitive detection of miRNA-21. To verify the analytical performance of the sensing platform based on the proposed 3D DNA nanomachine, diverse concentrations of miRNA-21 were tested under the optimized experimental conditions (Fig. S2 and S3 †). According to Fig. 4A, the ECL intensity gradually increased with incremental concentrations of miRNA-21 throughout the tested range (10 aM to 100 pM). As presented in Fig. 4B, the calibration plot demonstrated an excellent linear relationship between the ECL response and the logarithm of the miRNA-21 concentration. And its regression equation was I ¼ 1749.29 lg c miRNA-21 + 6078.51 with a correlation coefficient of 0.9990 and detection limit of 3.57 aM (S/N ¼ 3). Moreover, compared with other methods for miRNA detection (Table S2 †), the proposed strategy possessed a remarkably wider linear range and extremely lower detection limit, which manifested that the proposed core-brush 3D DNA nanomachine has potential applications in ultrasensitive biomarker detection and metrology. To validate the specifcity of our proposed 3D DNA nanomachine for sensitive detection of miRNA-21, a series of interfering agents containing miRNA-126, miRNA-141, miRNA-203a, miRNA-155 and miRNA-182-5p were introduced as interference tests. According to Fig. 4C and D, the ECL intensity of interfering agents (10 pM) was almost negligible as the blank one. In contrast, the mixture containing target miRNA-21 was similar to that of miRNA-21 (100 fM), exhibiting an obvious ECL intensity. Meanwhile, the proposed biosensor also showed excellent selectivity for mismatched miRNA-21 (Fig. S7 †). Next, the selectivity of our strategy based on Mn 2+ -dependent DNAzyme was also monitored by PAGE. As demonstrated in Fig. 5A, lane 1 represents the HS, lane 2 represents the DA, and lanes 3-8 represent the cleavage of the HS by the DA in the presence of Mg 2+ , Mn 2+ , Zn 2+ , Ca 2+ , Cu 2+ and Pb 2+ as alternative cofactors, respectively. As expected, the most obvious band of cleavage products (S) appeared in lane 4, indicating that Mn 2+ was the best cofactor to truncate S from the DA. The above results clearly demonstrate the high selectivity of our proposed 3D DNA nanomachine. Meanwhile, the gray scale intensity of the PAGE bands was analyzed by Image J. The highest gray scale value of S and the most obvious band of S were observed in lane 4 of Fig. 5B and C with the cofactor Mn 2+ , which were consistent with the results of gel electrophoresis. Furthermore, the proposed biosensing platform also presented outstanding stability and repeatability selectivity (Fig. S8 †). ## Rapid and sensitive imaging of miRNA-21 in living cells At frst, endocytosis inhibition experiments and MTT assays were implemented to reveal the uptake mechanism of our nanomachine (Fig. S9 †) and demonstrate the low cytotoxicity of our strategy (Fig. S10 †), respectively. Then, MCF-7 cells and HeLa cells were used to evaluate the imaging ability of the developed core-brush 3D DNA nanomachine for intracellular miRNA-21. As presented in Fig. 6A, MCF-7 cells, with high miRNA-21 expression, showed remarkably strong fluorescence with the cultivation of the core-brush 3D DNA nanomachine. Meanwhile, fluorescence that was obviously distinguishable from the background could be observed inside HeLa cells with a low miRNA-21 expression level (Fig. 6C). By contrast, the intracellular fluorescence intensity of MCF-7 cells cultured with the conventional Au-based 3D DNA nanomachine was distinctly weaker than that of with the proposed 3D DNA nanomachine (Fig. 6B). Especially for HeLa cells, only a low intracellular fluorescence intensity was noticed with the cultivation of the traditional Au-based 3D DNA nanomachine (Fig. 6D). Thus, compared with the conventional Au-based 3D DNA nanomachine, the proposed core-brush 3D DNA nanomachine could Fig. 5 (A) The specificity of the proposed strategy toward cofactor Mn 2+ (15 mM) against other cofactors (each at 150 mM). The quantitative gray scale value (B) and the total gray scale analysis (C) of part A. not only make it possible to image low-expressed miRNA-21 with higher sensitivity, but also offer a general strategy for accurate imaging in living cells with higher specifcity. Moreover, we also used the miRNA-21 mimic and inhibitor to further verify the specifc discrimination of intracellular miRNA-21 by the core-brush 3D DNA nanomachine (Fig. S11 †). We believe that the proposed core-brush 3D DNA nanomachine could provide an attractive way for sensitive imaging of intracellular specifc miRNA in low abundance, which could contribute to study the biological processes and mechanisms within the cell system and its application in early diagnosis and treatment of diseases. Furthermore, real-time imaging of MCF-7 cells cultured with the proposed 3D DNA nanomachine at different incubation time points within 60 min is tracked in Fig. 7. Impressively, the brightest fluorescence was observed in about 15 min and no obvious fluorescence growth was noticed within 60 min, indicating that the visualization of intracellular miRNA-21 was achieved within 15 min with the proposed core-brush 3D DNA nanomachine, which was noticeably faster than those of previously reported nanomachines (Table 1). These results confrmed that the rapid imaging strategy based on the proposed 3D DNA nanomachine might offer great potential for rapid point-of-care medical devices and early detection of cancer in clinical practices. ## Conclusions In summary, we presented a core-brush 3D DNA nanomachine by innovatively assembling DNA components (DA and HS) into a repetitive array on a programmable DNA track. Compared with traditional Au-based 3D DNA nanomachines, our strategy had the following combined advantages. First, the DNA track generated by a RCA reaction was well-designed to homogeneously arrange DNA components, which not only ensured a specifc walking range for the DA to prevent stochastic and invalid movement, but also avoided the discorded nano-bio interface of Au/DNA in the Au-based 3D DNA nanomachine. Second, the highly integrated 3D DNA nanomachine possessed an enhanced loading capacity and movement efficiency due to the organized and high local concentration of DNA components. Third, as practical applications, the proposed 3D DNA nanomachine was successfully applied for rapid and sensitive detection and imaging of intracellular specifc miRNA, which could help us explore its biological functions in tumor differentiation and proliferation as well as potential drug screening in early disease diagnosis.

chemsum

{"title": "A core\u2013brush 3D DNA nanostructure: the next generation of DNA nanomachine for ultrasensitive sensing and imaging of intracellular microRNA with rapid kinetics", "journal": "Royal Society of Chemistry (RSC)"}

a_flexible_copper_sulfide_composite_membrane_with_tunable_plasmonic_resonance_absorption_for_near-in

## Abstract: Near-infrared light driven devices for water evaporation are strictly limited by their inflexibility, high cost, complicated fabrication processes, and low energy-conversion efficiency. Here, a flexible copper sulfide composite membrane with tunable plasmonic resonance absorption for an efficient near-infrared light photothermal conversion is proposed. Both the uniformity of the morphology and the proportion of Cu + in the flower-like copper sulfide (CuS) superstructure are easily controlled by adjusting the amount of polyvinylpyrrolidone (PVP), which effectively improves the absorption of the CuS superstructure in the near-infrared region. Furthermore, the flexible CuS/Matrimid composite membrane constructed by combining CuS and polyimide membranes exhibits highly flexible properties, strong NIR absorption, fast heating (10 s), and good thermal stability. A highly efficient photothermal conversion is achieved by nearinfrared light-driven water evaporation. Under 808 nm light irradiation, the water evaporation conversion efficiency is ca. 80% and has excellent evaporation stability. The flexible CuS/Matrimid composite membrane developed in this study could have promising practical applications in near-infrared lightdriven devices for seawater desalination. Environmental signicanceThe seawater desalination driven by solar steam has emerged as one of the most promising ways to address this problem, due to its low energy input, high evaporation efficiency, and easy operation. NIR photothermal conversion materials, which can convert most of the NIR light energy into heat energy, can greatly improve the efficiency of NIR light utilization. Copper sulde, as an easy to fabricate doped semiconductor, shows excellent photostability, adjustable NIR absorption, and outstanding energy transfer efficiency. A NIR light-driven exible hot plate by incorporating doped semiconductors with a exible polymer membrane realizes the potential utility of the NIR light and doped semiconductors in seawater desalination. ## Introduction Water is the source of human life and the foundation of all things. It is because of moist water that the earth appears to be full of vitality, thriving. However, the rapid development of industrialization and modernization has led to the gradual reduction of available clean water resources. 1,2 Seawater desalination driven by solar steam emerges as one of the most promising ways to address this problem, due to its low energy input, high evaporation efficiency, and easy operation. 3 Generally, photothermal materials floating on the water surface or adhered to the container wall absorb sunlight, so as to carry out heat conversion and transfer, and generate water vapor through interface heating. However, the water evaporation efficiency was still inhibited by the high cost, poor stability, complicated fabrication processes, and low efficiencies of photothermal materials for solar light absorption, especially broadband nearinfrared (NIR, 780-2526 nm) light. 4,5 Therefore, it is imperative that novel NIR-driven materials or devices are devised, which have a high NIR utilization efficiency, so as to broaden the possible scope of application. NIR photothermal conversion materials, which can convert most of the NIR light energy into heat energy, can greatly improve the efficiency of NIR light utilization. Doped semiconductors, noble metal nanostructures, organic polymers and carbon materials are four typical kinds of NIR photothermal conversion materials that are widely used in the feld of NIR light conversion. In order to effectively use NIR light, several devices have been built recently based on NIR photothermal conversion materials. For example, Sun et al. constructed NIR light-induced shape memory polymers and silver nanoparticles for healing mechanical damage; 10,11 a carbon nanotube composite 12 and organic polymer 13,14 were used as NIR lightdriven photothermal-electrical and photo-magneto-thermoelectric devices by Wang et al. and Kim et al., respectively. Doped semiconductors, mainly including copper chalcogenide compounds (Cu 2x E, E ¼ S, Se, Te) 15 and transition metal oxides (WO 3x , MoO 3x , Mn x O y , etc.), 16,17 are a class of defect semiconductors with a local surface plasmon resonance (LSPR) effect. Compared with noble metal nanostructures, organic polymers, and carbon materials, doped semiconductors feature the advantages of low cost and stability. 18 Song et al. developed Cs x WO 3 nanoparticles to enhance the upconversion luminescence of monolayer upconversion nanoparticles as a high performance narrowband NIR photodetector, demonstrating the potential advantages of doped semiconductors. 19 Hence, it is meaningful to further explore NIR light-driven devices constructed from doped semiconductors, such as a NIR light-driven flexible hot plate by incorporating doped semiconductors with a flexible polymer membrane, to realize the potential utility of NIR light and doped semiconductors in seawater desalination. Copper sulfde, one of the most easily prepared doped semiconductors, shows excellent photostability, adjustable NIR absorption, and outstanding energy transfer efficiency, 15 and has been broadly applied to solar cells, catalysts, pollutant degradation, supercapacitors, and biomedicine. Matrimid® 5218, a thermoplastic polyimide based on 5(6)-amino-1-(4 0aminophenyl)-1,3-trimethylindane, is soluble in a variety of common solvents and will leave a strong, durable, and high temperature-resistant flexible coating. 24,25 More importantly, it is easy to cut into any shape and can be placed in any location. Thus, once inorganic materials are introduced into the polymer matrix, a combined effect of its polymeric and inorganic components' properties is thought to ensue. Therefore, here we propose a NIR light-driven flexible hot plate based on the introduction of copper sulfde into Matrimid® 5218, whereby copper sulfde can absorb NIR light and Matrimid® 5218 provides flexibility and endows the polymer matrix with high temperature resistance. This integrated device may show potential applications in seawater desalination with the synergy of the two components. The main challenge for this proof-of-concept experiment is to fabricate a uniform copper sulfde-Matrimid® 5218 flexible membrane with strong NIR absorption. To do this, we relied on several steps. Firstly, N-methyl-2-pyrrolidinone (NMP), which is normally used to dissolve Matrimid® 5218, was used as a solvent to prepare the copper sulfde nanocrystals. Thus, both the obtained copper sulfde nanocrystals and Matrimid® 5218 are soluble in the same solvent, uniformly. After the evaporation of NMP, a strong, durable, uniform, and flexible copper sulfde-Matrimid® 5218 membrane was obtained. Secondly, to attain adequate absorption of NIR light, the doping state and morphology of copper sulfde nanocrystals were tuned, accordingly, by simply controlling the PVP content. Finally, once irradiated with NIR light, the copper sulfde-Matrimid® 5218 membrane's temperature rises rapidly, enabling its use as a NIR light-driven flexible hot plate. The applications of this devised NIR light-driven flexible hot plate in photothermal evaporation water were also explored. Due to its impressive photothermal effect and flexible cutting advantages, this novel hot plate can be placed on either the outer or the inner surface of a given device, to prevent salt precipitation from accumulating on the surface, thereby enhancing the overall evaporation efficiency. ## Experimental PVP (M w ¼ 29 000) was purchased from Sigma Aldrich. Copper(II) sulfate pentahydrate (CuSO 4 $5H 2 O), N-methyl-2-pyrrolidinone (NMP), sulfur, and polyimide were all purchased from Sinopharm Chemical Reagent Co., Ltd. ## Preparation of the ower-like CuS nanostructure First, 0.5 g of PVP was dissolved in 15 mL of NMP, and then 0.5 mM (0.125 g) CuSO 4 $5H 2 O was added to the PVP solution under magnetic stirring. This stirring was continued until a bright green transparent solution was obtained, and then 1 mM sulfur powder (0.032 g) was added to the mixed solution. After 10 min of stirring, the above solution was transferred into a 25 mL Teflon autoclave for the reaction at 180 C for 4 h. The blackgreen CuS was centrifuged and washed twice with NMP, and dissolved in 1 mL of NMP for its later use. ## Preparation and characterization of the CuS/Matrimid composite membrane Polyimide (0.2 g) was mixed with 1 mL of CuS in differing amounts (i.e., 1, 2, 5, and 10 mg mL 1 ). After undergoing thorough mixing, an appropriate amount of a given sample was added dropwise onto a glass pane and we manually scraped the 500 mm surface with a membrane coater. Then it was quickly placed in a vacuumdrying oven and dried with a gradient temperature series: 80 C for 12 h, 140 C for 1 h, and 200 C for 1 h. For absorption measurements, a blank PI membrane without CuS served as the baseline. The absorbance of the CuS/ Matrimid composite membrane generated with different amounts of CuS was measured accordingly. For testing its photothermal properties, the blank membrane and CuS/Matrimid composite membrane were cut into small disks (2 cm diameter). An 808 nm laser (1 W cm 2 ) was used to irradiate the top of each membrane, whose temperature change was recorded with a thermal imaging device. ## Characterization The structure of the CuS nanostructure was confrmed by X-ray diffraction (XRD, Rigaku DMAX2000), and its morphology was quantitatively examined by SEM (JEOL JEM-6460A) and TEM (JEOL TEM-2100). The electronic state of Cu in the flower-like CuS was determined by X-ray photoelectron spectroscopy (ESCALAB 250Xi), and the absorbance of the CuS nanostructure and CuS/Matrimid composite membrane was determined by UV-Vis-NIR (Beckman Coulter DU730) and UV-Vis diffuse reflectance spectroscopy (Shimadzu UV-2450). The heating effect was tracked by using a FLIR A300 thermal imaging device. ## Vapor generation by the CuS/Matrimid composite membrane hot plate The saltwater sample (26.5 g L 1 NaCl, 0.2 g L 1 NaHCO 3 , 0.28 g L 1 NaBr, 24 g L 1 MgCl 2 , 3.3 g L 1 MgSO 4 , 0.73 g L 1 KCl, and 1.1 g L 1 CaCl 2 ) and sewage sample (3.5 wt% NaCl, 10 ppm pchlorophenol) were respectively prepared according to previous reports. 4,26 The seawater was collected from the East China Sea, in Fengxian (Shanghai). The CuS/Matrimid composite membrane was cut into a square (1 cm 1 cm) and this was stuck to the outside wall of a cuvette flled with water. The composite membrane was irradiated with an 808 nm laser (2 W cm 2 ) for 900 s, and water temperature was recorded with a thermal imaging device. Two key parameters, the water evaporation rate (n) and water evaporation efficiency (h), were calculated using eqn ( 1) and ( 2), respectively. where H e denotes the total enthalpy of the liquid water-to-vapor phase transition (J g 1 ), n is the water evaporation rate, and Q s is the light density of solar illumination. 28 For the solar water desalination test, the CuS/Matrimid composite membrane was pasted on the inside of a 100 mL aluminum cup to better absorb solar energy. The aluminum cup Environmental Science: Advances Paper was then flled with seawater and irradiated with a xenon lamp at a power density of 1 W cm 2 . Both photothermal imaging and water temperature were recorded with a thermal imaging device (FLIR A310). 3 Results and discussion ## Tuning the NIR absorption of the CuS nanocrystals To obtain copper sulfde nanocrystals with high photothermal conversion efficiency, some rational designs have been adopted to enhance their absorption of NIR light, for example, by tuning the structure, copper defciency, and hybrid composite. Nearly all of these methods are used in hydrothermal or thermal decomposition processes. But in order to prepare a uniform copper sulfde-Matrimid® 5218 flexible membrane, a solvothermal method must be implemented that is able to synthesize the copper sulfde by using NMP as the solvent. In this study, the copper sulfde was synthesized by a solvothermal method that uses PVP as the surfactant and NMP as the solvent (Fig. 1a). The effects of PVP content on the crystal phase, shape, copper valence and NIR absorption of the copper sulfde were analyzed. Previous reports have demonstrated that the crystal phase is critical for tuning the localized surface plasmon resonance (LSPR) absorption for copper sulfde. 29,30 Thus, the effect of PVP content on the crystal phase was frst investigated. The copper sulfde was prepared solvothermally by adding sulfur and CuSO 4 into NMP with differing amounts of PVP (0-1 g). The structural changes of the obtained copper sulfde samples were then investigated by studying the X-ray diffraction (XRD) patterns (Fig. 1b). The main diffraction peaks situated at the 2q angles of 29.28, 31.81, 32.95, 47.93, 52.71, and 59.34 of the copper sulfde prepared without PVP (i.e., 0 g added) agreed well with the (102), (103), (006), (110), (108), and (116) lattice planes of covellite CuS (JACPDS card no. 06-464). No characteristic peak can be indexed to any other phase of copper sulfde, except covellite CuS. When PVP was added in amounts of 0.125 to 1 g, evidently the peak patterns maintained the same structure as that of covellite CuS, while the intensity of characteristic peaks at the 2q angle of 32.95 decreased a little. This change in peak intensity may be attributed to the growth of the (006) lattice plane being inhibited by the strong coordination of PVP and Cu 2+ . The XRD patterns for copper sulfde obtained in the absence and presence of PVP suggest that PVP shows no apparent effects on the crystal phase of copper sulfde. Next, the morphology of the copper sulfde obtained with and without PVP was observed by scanning electron microscopy (SEM). The CuS material prepared without PVP is composed of a large-sized flower-like superstructure and irregular nanoparticles (Fig. 1c), while a more uniform flower-like superstructure smaller in size was obtained when PVP was introduced (Fig. 1d-h). PVP is competent to stabilize the CuS superstructure. In the chemical sintering process, the small particles will form compact solids due to the coalescence and Ostwald ripening behaviors triggered by the detachment of PVP. 31 With more PVP added, the growth of the CuS superstructure can be reliably controlled, and the flower-like superstructures of CuS appear uniform (Fig. S1 †), which ought to improve the photon reflection capacity and the NIR absorption and further enhance the photothermal conversion effect. In addition, the size of the CuS superstructure can also be tuned by changing the PVP content. As the SEM images show, the size of the prepared CuS superstructure decreased 10-fold, from 10 to 1 mm, when the PVP content is increased from 0 to 1 g; this indicated that the higher the PVP content used, the smaller is the size of the CuS superstructure. These results demonstrated that the uniformity and size of the superstructure, which are benefcial to promote the photothermal conversion effect, are governed by the PVP content. Given that the electronic state of Cu is related to the hole density of the nanomaterials, which is crucial for NIR absorption, XPS measurements were conducted to confrm the electronic state of Cu in the prepared CuS nanostructures. The Cu 2p peaks of these CuS featured the typical asymmetric tail of covellite. 32 Without PVP present, the binding energy intensity values of Cu 2p 3/2 and Cu 2p 1/2 were slightly left-shifted to 932.24 and 952.12 eV, with weak satellite peaks at 942.07 observed, indicating the presence of typical Cu 2+ but little Cu + in the obtained CuS nanostructures. As more PVP was used, the two binding energy intensity values of Cu 2p 3/2 and Cu 2p 1/2 were increasingly right-shifted and their satellite peaks gradually decreased, suggesting a greater proportion of Cu + in the CuS nanostructures (Fig. 2a). Moreover, the S 2p band of these CuS also corresponded to the typical "three peaks" of covellite, and the disulfdes eventually disappeared as the amount of PVP increased, thus indicating a gradual shift to chalcocite (Fig. S2 †). To clarify the ratio of Cu 2+ to Cu + in the CuS nanostructures, their Cu 2p bands were adequately ftted to four curve-ftting bands. Using the area of the four curve-ftting bands, the ratio of Cu 2+ to Cu + was calculated accordingly, yielding values of 0.71, 0.34, 0.32, 0.30, 0.27, and 0.28 (Fig. 2b and c). The results suggest that the ratio of Cu + in the CuS nanostructures increases as the amount of PVP increases until it's reduced to saturation, which can be ascribed to the effect of precursor reducibility by PVP. 36,37 The optical properties of CuS prepared by introducing various amounts of PVP were examined by Vis-NIR spectroscopy. To fairly compare the absorption intensity of these differently prepared CuS samples, their absorption at 650 nm was assigned the same value. With more PVP added, the normalized absorption intensity of the obtained CuS nanostructure in the NIR region (700-1000 nm) clearly increased (Fig. 2d). Moreover, absorption at 808 nm is linearly enhanced with increasing PVP content, until it reaches saturation when 0.75 g of PVP is added (Fig. 2e). The NIR absorption of CuS is the LSPR absorption of electrons and holes, which is like the LSPR absorption of electrons in the noble metal. The hole arises from the copper vacancy in the CuS nanoparticles, such that CuS nanoparticles with a higher hole density will exhibit stronger NIR absorption. 38,39 The copper vacancy can be tuned by modulating the proportion of Cu + in the CuS nanostructure, yet the Cu + in the CuS nanostructure can be controlled by the applied amount of PVP in the synthesized system. In addition, the superstructure facilitates the NIR absorption of CuS nanoparticles, likely because photo-absorption can be augmented by the faceted end planes of well-shaped crystals that serve as good light-cavity mirrors. 40,41 This inference is also corroborated by our fnding of the NIR absorption of the CuS superstructure decreasing after undergoing ball milling (Fig. S3 †). Therefore, the strong LSPR absorption of CuS in the NIR region can be simply tuned by the amount of PVP used in the synthesis system. ## Preparation and characterization of the exible hot plate based on the CuS/Matrimid composite membrane Materials with flexibility, ease of cutting, and ease of construction are now receiving extensive attention from researchers. To broaden the application scope of the CuS nanostructure with strong NIR absorption, it would be prudent to develop a CuS flexible membrane with excellent photothermal properties. Matrimid® 5218 is a widely used polymer matrix for various inorganic nanoparticles in microelectronics and the gas separation industry. More importantly, the Matrimid® 5218 membrane has good heat-resistance. Here, we developed a flexible hot plate by incorporating the CuS nanostructure into Matrimid® 5218 uniformly. The CuS nanostructure was prepared in NMP and it can be dispersed in NMP very well, while the NMP performs well as a solvent for the Matrimid® 5218 polymer. Thus, a uniformly dispersed liquid of the CuS nanostructure and Matrimid® 5218 was easily obtained when they were added to NMP with ultrasonic dispersion. Following this step, the CuS/Matrimid membrane was prepared by applying a scrape coating/drying technique (Fig. 3a). The doped The absorption performance of the fabricated membranes with increased levels of CuS doping (0.5 wt%, 1.0 wt%, 2.5 wt%, 5.0 wt%, and 10 wt%) was evaluated from their Vis-NIR spectra (Fig. 3b). The absorption values rose considerably by increasing the CuS doping from 0.5 wt% to 10 wt% for the CuS/Matrimid composite membrane. Its good absorption properties in the NIR region will confer an excellent photothermal performance. We then tested its photothermal performance by recording the temperature change under 808 nm laser (1 W cm 2 ) irradiation for 60 s (Fig. 3c). Compared with the blank Matrimid membrane, the CuS-doped Matrimid membrane exhibited outstanding photothermal performance, in which the temperature is capable of rising from 84.5 C to 145.5 C when the doped CuS content is increased from 0.5 wt% to 10 wt%. Furthermore, the temperature increase of the CuS/Matrimid composite membrane is particularly fast under the laser irradiation, attaining its maximum in just 10 s. After that, the temperature can maintain its maximal value without undergoing signifcant change, which is attributable to the generated heat from the CuS/Matrimid composite membrane under the laser irradiation being equal to the heat diffused into air. These results suggest that the CuS/Matrimid composite membrane has the property of fast heating under laser irradiation. Next, the relevance of light density for the photothermal properties of the CuS/Matrimid composite membrane was investigated (Fig. 3d). For the CuS/Matrimid composite membrane with 2.5 wt% CuS, the maximum temperature reachable by the CuS/Matrimid composite membrane is 55 C to 145 C under irradiation of a laser with a light density from 0.1 to 1.5 W cm 2 ; this implies that light density exerts important effects. Thermal stability is a key parameter that determines whether the CuS/Matrimid composite membrane can be used as a hot plate. To evaluate it, the photothermal circle test was used. After 8 'on/ off' cycles of the laser-each cycle consisting of a laser on time of 1.5 min and a laser off time of 1.5 min-the CuS/Matrimid composite membrane still exhibited similar photothermal properties, indicating very good thermal stability (Fig. 3e), which can be attributed to its structural stability after laser irradiation (Fig. S5 †). Moreover, the prepared CuS/Matrimid composite membrane still possessed its highly flexible properties on par with those of the Matrimid® 5218 membrane without doping the CuS nanostructure, as demonstrated by the photographed Matrimid® 5218 membrane before (upper panel) and after (lower panel) doping the CuS nanostructure (Fig. S6 †). In addition, the as-prepared membrane could be carved into many different shapes (Fig. 3f), without any influence on its impressive photothermal performance (Fig. 3g), endowing it with more potential for realistic applications. The excellent performance of the CuS/ Matrimid composite membrane, characterized by its highly flexible properties, strong NIR absorption, fast heating, good thermal stability and easy cutting, makes it a promising candidate for use in a NIR light-driven hot plate. highly efficient vapor generation by heat localization at the evaporation surface (Fig. 4a). Evidently, with the CuS membrane present, the temperature rose rapidly under continuous laser irradiation and remained stable for the entire 900 s duration. In stark contrast, the temperature in the blank panel showed only a slight difference after its irradiation (Fig. 4b and c). Moreover, the cumulative weight loss was positively correlated with the irradiation time (Fig. 4d). Under 808 nm irradiation, the weight loss over the 900 s period was 5.86 kg m 2 , and the steady-state evaporation rate was calculated to be 23.4 kg m 2 h 1 . This is much higher than the evaporation rate of water in the absence of CuS, which was 3.06 kg m 2 h 1 . Hence, the photothermal evaporation conversion efficiency of the CuS/Matrimid composite membrane (z80%) is nearly 8 times higher than that of the blank membrane (Fig. 4e). Crucially, the steady-state evaporation rate and cumulative weight loss did not change signifcantly over eight cycles of reuse (Fig. 4f). This suggests that the CuS/Matrimid composite membrane is highly stable and can be reused multiple times without a pronounced decrease in its evaporation capacity. The NIR light-driven photothermal evaporation of water was also investigated in stimulated saltwater and wastewater. As expected, under continuous irradiation of the laser, both salt water and wastewater covered by the CuS/Matrimid composite membrane incurred a rapid heating effect (Fig. 5a). Due to the relatively complex nature and high concentration of ions in salt water, its temperature rise is affected to some extent (Fig. 5b). Accordingly, the weight change of salt water after evaporation is not as great as that of wastewater or pure water (Fig. 5c). Nonetheless, much water was still evaporated within 15 min, and the evaporation rate and efficiency of salt water and wastewater are still good (Fig. 5d). More importantly, after water evaporation, the concentration of ions in the collected distilled water had decreased substantially (Fig. 5e). In addition, the evaporation of the composite membrane is relatively stable (Fig. S7 †). Solar seawater evaporation and desalination tests showed similar heating and desalination effects, thus indicating that the CuS/Matrimid composite membrane may be useful for light-driven photothermal evaporation of water (Fig. S8 †). Overall, because of its strong NIR absorption, high photothermal conversion, flexible cutting and localization, the CuS/Matrimid composite membrane shows great photothermal efficiency, which we anticipate will be applied in actual water evaporation and seawater desalination projects. ## Conclusion In summary, a kind of flower-like, self-doped CuS superstructure with tunable plasmonic resonance absorption and photothermal effects was designed, for which PVP was the surfactant and NMP is the solvent. The results show that with a greater amount of added PVP, there is an increased degree of Cu 2+ reduction, generating more copper defects that enhance the absorption ability of the CuS superstructure in the near infrared region, until Cu 2+ is no longer reduced. Furthermore, CuS membranes featuring high-temperature resistance and good flexibility were prepared by combining CuS with polyimide membranes via coating and gradient high-temperature curing. Photothermal performance testing shows that the temperature of the CuS/ Matrimid composite membrane can rise to more than 100 C within just a few seconds under the irradiation of an 808 nm laser, suggesting that it functions as a robust photothermal conversion membrane. The responsive properties of the CuS/ Matrimid composite membrane to vapor evaporation driven by NIR light were explored. Compared with a blank membrane, the composite membrane evinced a heating effect and better evaporation efficiency, both in stimulated saltwater and sewage. This proves that the CuS/Matrimid composite membrane has promising application prospects. This work provides the possibility for further development of an NIR light-driven flexible and tunable absorption semiconductor, which we anticipate will broaden the further application of this kind of device.

chemsum

{"title": "A flexible copper sulfide composite membrane with tunable plasmonic resonance absorption for near-infrared light-driven seawater desalination", "journal": "Royal Society of Chemistry (RSC)"}

isotope_depletion_mass_spectrometry_(id-ms)_for_enhanced_top-down_protein_fragmentation

## Abstract: Top-down mass spectrometry has become an important technique for the identification of proteins and characterisation of chemical and posttranslational modifications. However, as the molecular mass of proteins increases intact mass determination and top-down fragmentation efficiency become more challenging due to the partitioning of the mass spectral signal into many isotopic peaks. In large proteins, this results in reduced sensitivity and increased spectral complexity and signal overlap. This phenomenon is a consequence of the natural isotopic heterogeneity of the elements which comprise proteins (notably 13 C). Here we present a bacterial recombinant expression system for the production of proteins depleted in 13 C and 15 N and use this strategy to prepare a range of isotopically depleted proteins. High resolution MS of isotope depleted proteins reveal dramatically reduced isotope distributions, which results in increases in sensitivity and deceased spectral complexity. We demonstrate that the monoisotopic signal is observed in mass spectra of proteins up to ~50 kDa. This allows confident assignment of accurate molecular mass, and facile detection of low mass modifications (such as deamidation).We outline the benefits of this isotope depletion strategy for top-down fragmentation.The reduced spectral complexity alleviates problems of signal overlap; the presence of monoisotopic signals allow more accurate assignment of fragment ions; and the dramatic increase in single-to-noise ratio (up to 7-fold increases) permits vastly reduced data acquisition times. Together, these compounding benefits allow the assignment of ca. 3-fold more fragment ions than analysis of proteins with natural isotopic abundances. Thus, more comprehensive sequence coverage can be achieved; we demonstrate near single amino-acid resolution of the 29 kDa protein carbonic anhydrase from a single top-down MS experiment.Finally, we demonstrate that the ID-MS strategy allows far greater sequence coverage to be obtained in time limited top-down data acquisitions -highlighting potential advantages for topdown LC-MS/MS workflows and top-down proteomics. ## Introduction Top-down mass spectrometry (MS) has emerged as a powerful technique for the analysis of protein sequence and the detailed characterisation of chemical modifications to protein side-chains. Consequently, top-down MS is a powerful strategy for the comprehensive identification and characterisation of all proteoforms arising from genetic variation, alternative splicing, and post-translational modifications (PTMs). The technique consists of first measuring the intact molecular mass of a protein, followed by gas phase fragmentation of a selected proteoform ion by tandem mass spectrometry. The resulting fragment ions are assigned based on their observed accurate mass. If sufficient numbers of fragments can be assigned, top-down MS can provide a complete description of protein sequence and PTM state. Over the last two decades several fragmentation techniques have been employed for top-down studies. However, electron-based fragmentation techniques such as electron capture dissociation (ECD) and electron transfer dissociation (ETD) have been the most widely applied and they offer improved diversity of backbone site cleaved when compared techniques which rely on vibrational excitation, such as collision induced dissociation (CID) and infrared multiphoton dissociation (IRMPD). Thus top-down ECD and ETD can provide comprehensive sequence coverage for the analysis of small proteins (<20 kDa). However, as protein mass increases, top-down fragmentation efficacy notably diminishes; spectra become increasingly complex, and a series of other compounding factors result in reduced sequence coverage (the challenges of top-down MS have been discussed in depth in several recent publications). One fundamental factor which proves detrimental in top-down analysis is the increasing breadth of the isotopic distribution that accompanies increasing molecular mass. For proteins, the isotopic heterogeneity of the organic elements (particularly the ~1.1% natural abundance of 13 C) results in the ion signal being spread over a distribution of discrete isotopologues (the isotope distribution); with each isotopologue differing in composition by a neutron. As protein mass increases, this isotope distribution widens, and so the overall signal gets more disperse. For example, a 10 kDa protein the isotope distribution will consist of 12 isotopologue signals; whereas for a 50 kDa protein, the number of isotopologues observed can be over 40. This phenomenon reduces the signal to noise ratio (S/N) and can lead to the overlapping of signals for species which are close in mass (e.g. proteoforms of the same protein with similar masses). Furthermore, proteins, or protein fragment ions, over ~10 kDa commonly do not display a monoisotopic signal of sufficient ion abundance to accurately assign. In the context of a top-down fragmentation experiment, these compounding difficulties all reduce the number of fragment ions which can be confidently assigned as protein mass increases. ## 4 One solution to this problem is the production of isotopically enriched/depleted proteins, by recombinant production in hosts grown on carbon/nitrogen sources that are enriched or depleted in specific naturally occurring heavy isotopes. The feasibility of this strategy has been demonstrated by Marshall et al., in 1997. They reported the production and intact mass analysis of the 12 kDa FK506-binding protein, in media depleted in 13 C and 15 N. Using this approach, the authors demonstrated an increase in the mass spectral sensitivity and the detection limit. Despite the publication of this seminal report over twenty years ago, the use of this strategy has been limited; despite theoretical studies highlighting the potential benefits of the approach. The reason for this may be the technical difficulty in producing isotopically depleted proteins, and currently the strategy has only been applied to produce a handful of small proteins (<15 kDa). Herein we detail a robust method for the recombinant production of isotopically depleted protein in E.coli and demonstrate its benefits to top-down protein analysis by producing and characterizing a series of proteins up to 50 kDa. All isotopically depleted proteins displayed dramatically simplified isotope distributions and, as a consequence, we report a reduction in mass spectral complexity and dramatic S/N increases. Using this strategy, termed isotope depletion mass spectrometry (ID-MS), we show that the monoisotopic mass signals can be observed in isotopically depleted proteins up to 50 kDa. This allows direct and accurate determination of molecular mass for large proteins and protein fragment ions, for the first time. Finally, we perform top-down fragmentation of isotopically depleted proteins and demonstrate that the reduced spectral complexity and increased S/N allow assignment of fragment ions with increased confidence, and results in dramatically improved sequence coverage. 5 ## Results and Discussion. We chose three well-characterised proteins as model systems for this study -encapsulated ferritin (EncFtn, 13.2 kDa), carbonic anhydrase (CA, 29.3 kDa), and serine palmitoyltransferase (SPT, 47.3 kDa). These proteins were recombinantly expressed in E. coli using M9 minimal growth media, containing glucose and ammonium sulfate as the sole carbon and nitrogen sources. This allowed isotopically doubly-depleted protein samples to be prepared by using isotopically-depleted glucose (99.9% 12 C 6 ) and ammonium sulfate (99.99% 14 N 2 ) in the cell culture preparation. Full details of the expression protocol can be found in the (Supporting Information, Figure S1). ## ID-MS allows direct determination of monoisotopic mass of intact proteins up to 50 kDa. After protein expression and purification, MS analysis of the intact proteins was performed using high resolution electrospray (ESI) Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS). ESI mass spectra of samples prepared in natural abundance cell culture and double-depleted cell culture exhibited identical charge state distributions (Figure S2). However, dramatically simplified isotope distributions were observed in the mass spectra of proteins produced in isotopically depleted media when compared to natural isotopic abundance proteins (Figure 1). (Top) The observed isotope distribution for samples prepared from natural isotope abundance cell culture and (bottom) the observed isotopic distribution for samples prepared from isotope depleted cell culture. The theoretical isotopic distributions are overlaid on the spectra as scatter plots (natural abundance: 98.89% 12 C, 99.63% 14 N; isotopically depleted abundances 99.90% 12 C, 99.99% 14 N). In each spectrum, the monoisotopic species is highlighted with an asterisk (*). ## 6 For EncFtn (monoisotopic molecular mass 13,186.4 Da, Figure 1A), the width of the isotopic distribution decreased from 14 Da in the natural isotopic abundance protein to 7 Da in the isotopically depleted protein (isotopologues with abundance greater than 0.5% of the base peak); and the monoisotopic peak increased from ~0.06% of the total signal (i.e. below the noise) to ~30% of the total signal, and was the highest peak in the distribution. For CA (29,294.8 Da) and SPT (47,201.8 Da) the monoisotopic peak was not visible in the natural isotopic abundance protein distribution. In contrast, in the corresponding mass spectra obtained from isotopically depleted proteins, the monoisotopic peak was easily identifiable. For isotopically depleted CA, the monoisotopic peak accounted for ~13% of the total signal, and for the isotopically depleted SPT protein, the monoisotopic peak still accounts for ~2% of the total signal. Similar to EncFtn, the isotopically depleted variants of CA and SPT exhibited a reduction in isotopic distribution width of 20 Da to 10 Da and 26 Da to 16 Da respectively. In all case, the S/N also improved as the same number of proteins ions were partitioned between fewer isotopologue peaks. As first proposed by Marshall in 1997, we have demonstrated that 13 C and 15 N double depletion increases the proportion of the monoisotopic isotopologue and results in an observable monoisotopic signals for protein up to ~50 kDa. This allows the unambiguous and immediate assignment of the accurate molecular mass of intact proteins. In contrast, for accurate mass assignment of natural isotopic abundance proteins above ~10 kDa, it is necessary to infer the monoisotopic mass by matching the observed isotope distribution with the calculated theoretical isotope distribution of an 'average' protein -i.e. a repeating polymer of the model amino acid averagine. This 'poly-averagine approximation' method relies on obtaining a statistically reliable experimental isotope distribution and often results in significant mass error (up to 3 ppm) and/or the misassignment of the monoisotopic mass by +/-1 Da, regardless of the resolution achieved in the data acquisition. Consequently, this makes confident detection of low molecular mass PTMs, such as disulfide bond formation or deamidation, at the protein level, particularly challenging. Therefore, the ID-MS strategy may be particularity powerful for the detection and characterisation of these low molecular mass PTMs. In order to demonstrate this, we produced an EncFtn 'deamidated' single-point variant (N58D), in an isotopically depleted form. MS analysis of isotopically depleted N58D EncFtn allowed confident detection of the deamidation at the protein level. Direct detection of protein deamidation was also possible from mixtures of WT and deamidated proteoforms (Supporting Information, Figure S3). ## ID-MS dramatically improves protein sequence coverage in top-down fragmentation. Typically, top-down fragmentation generates many hundreds of fragment ions, with each ion appearing in multiple charge states, and exhibiting its own isotopic distribution. Thus, the resulting spectra are highly complex and consist of many thousands/ tens of thousands of individual peaks, over a wide dynamic range of ion-abundance. As the observed fragment ions fall in a comparatively narrow m/z range (typically m/z 500-2000), fragment ion isotope distributions often overlap; and, even with high resolving power instrumentation, superposition of peaks is common. Therefore, fragment ions can be overlooked or misassigned due to low signal and/or signal overlap. In order to investigate the benefit of the isotope depletion strategy for top-down mass spectrometry, we analysed natural isotopic abundance and isotope depleted forms of both EncFtn and CA, using both CID and ECD fragmentation. Initially, CID was performed on the [M+16H] 16+ precursor ion of natural isotopic abundance and isotopically depleted EncFtn. Both fragmentation spectra were remarkably similar on initial inspection, displaying identical high abundance fragment ions at similar m/z (Figure S4A). However, all fragment ions derived from the isotopically depleted EncFtn exhibited reduced isotope distribution widths, which greatly reduces signal overlap of individual fragments. In addition, the S/N ratio displayed by isotopically depleted fragment ions was dramatically increased (for example, the complementary ion-pairs b 37 4+ and y 78 10+ exhibit S/N gains of 7.0-fold and 4.7-fold in the isotopically depleted spectrum when compared to the natural isotopic abundance spectrum). It was also apparent that, for this 13 kDa isotopically depleted protein, the monoisotopic signal was the base-peak (i.e. the highest signal) in every isotopically depleted fragment ion's isotope distribution. This allowed direct determination of the accurate monoisotopic mass of every fragment ion (Figure S4C). Taken together, these three advantages led to confident assignment of substantially more CID product ions in the isotopically depleted EncFtn CID spectrum. For CID of the [M+16H] 16+ of EncFtn, 110 b and y fragment ions were assigned in the natural isotopic abundance spectrum (39 b-ions, 71 y-ions; 45.7% total sequence coverage); in comparison, 217 b and y fragment ions (84 b-ions, 133 y-ions; 64.7% total sequence coverage) were assigned in the natural isotopic abundance spectrum (Supporting Information, Figure S5). This increase in the observed fragment ion number is similar to that demonstrated by Akashi et al., who reported an 63% increase in the number of assigned fragment ions when performing CID of an isotopically-depleted version of the 10 kDa protein cystatin. However, for CID of both natural isotopic abundance and isotopically depleted EncFtn, the assigned b-and y-ions only constitute only around 20-30% of the total number of observed fragments; and even employing an isotopically depleted strategy with top-down CID, it is clear that there are regions of the protein with limited sequence coverage. Further analysis of the unassigned fragment ions in both CID spectra revealed a substantial number of internal 8 fragments, and widespread neutral loss during fragmentation (-H 2 O, -CO, -NH 3 ). Taking these fragmentation channels into considerations allowed assignment of a total of 448 product ions (a, b, x, y, and y -H2O ions; 82% total sequence coverage) in the CID spectrum of isotopically depleted EncFtn (Supporting Information, Figure S5). The lack of product ion specificity, and the biased nature of fragmentation with CID has been well documented, and this limits the utility of the technique for top-down studies of proteins over 10-15 kDa. In contrast to CID, electron-driven dissociation techniques (such as ECD and ETD, together termed 'ExD') are thought to result in relatively unbiased fragmentation throughout the protein sequence. Thus, potentially higher sequence coverage has been reported (especially in larger proteins) and ExD fragmentation is a far more attractive technique for top-down fragmentation as protein mass increases. However, one drawback of the ExD approach is its relatively inefficient precursor-to-product ion conversion and so ExD characteristically results in c-and z-type fragment ions of low ion abundance. Therefore, we reasoned that the substantial increased S/N evident in top-down ID-MS may potentially be of more benefit when used in conjunction with ExD studies. 9 Figure 2 shows the spectrum obtained after ECD of EncFtn (spectral averaging of 150 acquired transients; magnitude mode). Post-acquisition, Autovectis was used to process the data in absorption mode and assign fragment ions (for details see Supporting Information, Figure S6). ECD of a single charge state of natural isotopic abundance EncFtn yielded 131 c-10 ions and 125 z-ions. This included fragment ions from throughout the protein sequence and represented a total sequence coverage of 84.5% (Figure 2C, left); in our hands, this result is entirely typical for top-down ECD fragmentation of a 13 kDa protein. In comparison to ECD of natural isotopic abundance EncFtn, fragment ions obtained from ECD of isotopically depleted EncFtn displayed reduced isotopic distribution widths with a dominant monoisotopic signal and increased signal abundance (typically ~2-fold to 7-fold S/N increases were observed; dependant on fragment ion molecular mass). These factors allow assignment of many more low abundance ECD fragment ions, and accurate assignment of fragment ions which overlap in the natural isotopic abundance spectrum (Figure 2B and further examples in Supporting Information, Figure S7). In addition, ECD of the isotopically depleted protein allowed accurate assignment of sidechain losses and revealed low abundance ions in 'congested' regions of the spectrum (see Supporting Information, Figure S8). ECD of the isotopically depleted EncFtn yielded 276 c-ions and 220 z-ions fragment ions from this single experimental condition; a total sequence coverage of 97.4% (Figure 2C, right). Cleavages N-terminal to proline are not generally observed in ECD. Remarkably, if this is taken into account, of 114 peptide bonds in EncFtn only 2 possible cleavages were not observed. In addition, complementary c-and zion pairs cover over 85% of the protein sequence. It is clear from our analysis of EncFtn that three characteristics of the isotope depletion MS strategy lead to dramatic improvements for top-down fragmentation -namely, (i) improved overall S/N, (ii) increased monoisotopic signal abundance, and (iii) decreased isotope distribution width. These compounding benefits should be more evident as the precursor protein mass increases over 20 kDa. Therefore, we tested the utility of top-down isotopically depleted MS at higher mass, analysing bovine CA (29 kDa) by top-down ECD. CA has been used extensively to characterise top-down fragmentations technologies by multiple research groups and on multiple MS platforms; [11,23, thus it constitutes an ideal model study. Either the [M+32H] 32+ (m/z 916) or the [M+22] 22+ (m/z 1332) charge state of CA was isolated and subject to ECD (Figure 3). ECD of the [M+22H] 22+ charge state of natural isotopic abundance CA produced highly complex spectra (20,000 peaks with S/N > 2.5), which exhibit overlapping fragment ion isotope distributions throughout the spectrum (Figure 3A,3B, top). In addition, following substantial spectral averaging (300 averaged transients), even more fragment ions were observed with low S/N, this was especially evident as fragment ion mass increased. In total, from this single dataset, 229 c-and z-fragment ions could be assigned, representing 50.0% sequence coverage (Figure 3C, left). Low sequence coverage was especially evident in the central region of the protein. ## 11 As expected, ECD of the isotopically depleted CA resulted in significantly reduced spectral complexity and fragment ion distribution overlap. Fragment ions were observed with increased S/N (~2-to 8-fold increase -similar to previous results, vide infra; Figure 3B and further examples in Supporting Information, Figure S9). Interestingly, compared to the equivalent natural isotopic abundance spectrum, a similar number of individual peaks were observed in the ECD spectrum of isotopically depleted CA, suggesting that substantially more fragmentation channels should be evident. As a consequence, from the ECD spectrum of the [M+22H] 22+ of isotopically depleted CA, 593 fragment ions (377 c-ions, 216 z-ions) were assigned; i.e. approximately a three-fold increase in the number of fragment ions assigned from the natural isotopic abundance CA sample. These fragment ions yielded a sequence coverage of 82.6% for the isotopically depleted protein (Figure 3C, right). Comparable assignment rate increases were possible when analysing the ECD spectra of the [M+32H] 32+ charge state of isotopically depleted CA (Supporting information, Figure S10). If the sequence coverage observed for both charge states are combined, the overall sequence coverage obtained for isotopically depleted CA was over 90% (95.2% if bonds with adjacent proline residues were discounted); i.e. only 12 cleavages were not observed in this 263-amino acid protein -very close to the 'ideal' of single amino-acid level resolution throughout the protein sequence (Supporting Information, Figure S10). To our knowledge, represents the most comprehensive sequence coverage of CA observed to date, irrespective of fragmentation technique or MS platform. 12 The fragmentation maps (protein sequence coverage) achieved after ECD of the [M+22H] 22+ charge state of natural isotopic abundance CA (left; 50%) and isotopically depleted CA (right; 82.6%). ## 13 One striking characteristic of the ECD of isotopically depleted proteins is the ability to assign extended stretches of complimentary c-and z-ions, even in central regions of larger proteins. In effect, allowing comprehensive fragment ion sequence coverage 'deeper' into the protein sequence. Comparison of the mass distributions of the fragment ions assigned after ECD of natural isotopic abundance and isotopically depleted CA show that more fragment ions are assigned in the isotopically depleted ECD spectrum from across all molecular mass ranges (Figure 4A). However, these histograms highlight that the ID MS strategy has the greatest benefit for the assignment of fragment ions of higher masses; where ECD of isotopically depleted CA consistently affords 3-to 8-fold more fragment ions than ECD of natural isotopic abundance CA. For example, in ECD of isotopically depleted CA resulted in 52 fragment ions in the mass range 15-18 kDa; whereas only 10 fragment ions of similar mass were assigned from natural isotopic abundance CA. Not only is sequence coverage improved in isotopically depleted ECD, but fragment ions are also assigned with lower error in the isotopically depleted ECD spectrum. Figure 4B shows the distribution of errors for the assigned ECD spectra of natural isotopic abundance and isotopically depleted CA. Using a 'poly-averagine'-based approach for deconvolution of the ECD spectrum of natural isotopic abundance CA, the resulting fragment ions were assigned with a RMS error of 1.306 ppm. While AutoVectis analysis of the ECD spectrum of the same charge state of isotopically depleted CA allowed assignment of fragment ions with a RMS error of 0.800 ppm. The ability to assign dramatically more fragment ions, especially fragment ions of mass >10 kDa, is a direct consequence of the inherent increase in the S/N which accompanies isotopically depleted MS. In addition, the ability to directly observe the monoisotopic signals in isotopically depleted fragment ions is also highly advantageous, as it removes the requirement to obtain isotopic distributions with sufficient S/N for precise poly-avergine based deconvolution methods. Furthermore, the mass error introduced using the poly-averagine approximation during deconvolution is removed; leading to assignment of fragment ions with lower overall mass error. Therefore, higher confidence in fragment ion assignment can be achieved, which is particularly important for the interpretation of highly complex spectra, such as top-down analysis of large proteins or assigning branched protein ions. ## ID-MS improves top-down ECD on an LC Timescale. One of the overarching goals of top-down mass spectrometry is to achieve comprehensive protein sequence coverage using spectral acquisition times that are compatible with front-end chromatography; potentially allowing top-down protein analysis to be used in an LC-MS/MS proteomics workflow. Significant advances have been made in this field of top-down proteomics in recent years; however, it is particularly challenging to achieve extensive sequence coverage as protein molecular mass increases, and the sequence coverage achieved in top-down LC-MS/MS experiments is often restricted to limited regions at the Nand C-termini of the protein. Although this is often sufficient to provide a 'sequence-tag' and allow protein identification, low sequence coverage is insufficient for confidently mapping protein modifications and full characterisation at the proteoform level. These limitations are due to the time-constraints of the experiment and the inability to perform the extensive spectral averaging required to obtain fragment ion signals of sufficient abundance. In effect, there is a compromise that exists between the signal-to-noise level achieved and the spectral acquisition time. Spectral averaging produces a gain in the S/N ratio that is approximately proportional to the square root of the number of scans averaged. Because of this non-linear relationship, the increased S/N inherent in our isotopically depleted MS approach should be particularly effective for increasing the fragment ion sequence coverage obtainable with limited spectral averaging. In order to investigate this, ECD spectra were acquired in the same fashion for both the natural isotopic abundance and isotopically depleted forms of EncFtn (13 kDa) and CA (29 kDa) using both 20 or 5 spectral averages; which constituted total data collection times of ~25 and ~6 seconds respectively. The resulting spectra were analysed and fragment ions assigned (Supporting Information, Figure S11 and S12) and compared to the longer spectral acquisition time, described above (Figure 5). As expected, for natural isotopic abundance EncFtn reduction in the spectral averaging reduces the obtained protein sequence coverage significantly and with spectral averaging limited to 5 transients, only 48 ions could be assigned constituting 31% total sequence coverage. In contrast, for isotopically depleted EncFtn the reliance on extensive spectral averaging to obtain high sequence coverage is far less pronounced, and 86.2% protein sequence coverage was achieved with only 5 averaged spectra. For the larger protein, CA (29 kDa), it is clear that without extensive spectral averaging the sequence coverage obtained after ECD of the natural isotopic abundance protein is severely limited -28.4% sequence coverage is obtained with 20 averaged spectra and 14.4% sequence coverage is obtained upon averaging only 5 spectra. As discussed above, this phenomenon is well-documented in larger proteins, and is a current bottleneck in top-down proteomics. Dramatic improvements are observed using the isotopically depleted strategy and in-depth sequence coverage can still be assigned under time-limited data acquisitions. ECD of isotopically depleted CA using 20 and 5 spectral averages affords sequence coverage of 61.7% and 47% respectively. These initial findings demonstrate the potential benefit of applying isotopically depleted strategies in top-down proteomic workflows and highlight the possibility of achieving comprehensive sequence coverage of larger proteins on chromatographic timescales. ## Conclusion We have produced several isotopically depleted proteins with molecular masses up to ~50 kDa. We demonstrate that mass spectra of intact isotopically depleted proteins display decreased isotope distribution widths and increased S/N. In addition, direct observation of the monoisotopic signal of isotopically depleted proteins is possible; allowing accurate molecular mass to be directly determined, even in large proteins. Applying ID-MS in conjunction with topdown fragmentation affords reduced spectral complexity, increased S/N and increased mass accuracy; together this allows assignment of dramatically more fragment ions (typically 2-to 3-fold) and consequently increased protein sequence coverage. We also highlight the potential of applying ID-MS for performing top-down fragmentation on a chromatographic timescale for top-down proteomic applications. Finally, we note that this isotope depletion strategy is analogous to the isotope enrichment techniques which have become integral to biomolecular nuclear magnetic resonance (NMR) spectroscopy. Similarly, it is clear that ID-MS has huge promise for many biomolecular MS applications, particularly for proteins (or other biomolecules ) of high molecular mass. Techniques such as hydrogen/deuterium exchange MS, native protein MS, and structural MS will all benefit greatly from the advantages which accompany isotopic depletion. ## Experimental Section Isotopically-depleted proteins were produced by recombinant expression in E. coli using minimal media supplemented with 12 C(99.9%)-glucose and 14 N(99.99%)-ammonium sulfate as the sole carbon and nitrogen sources; see Supporting Information for detailed protocols. MS experiments were performed on a 12T SolariX Fourier transform ion cyclotron resonance (FT-ICR) mass spectrometer equipped with an Infinity ICR cell (Bruker Daltonics, Bremen, Germany). Ionisation was achieved with a TriVersa NanoMate nanoelectrospray robot (Advion Bioscience, Ithaca, NY). Data was processed in magnitude mode, using Data Analysis (Bruker Daltonics, Bremen, Germany); and, in the absorption mode, using an in-house developed, enhanced version of AutoVectis (Nottingham Trent University and Spectroswiss Sàrl, Lausanne, Switzerland). Full details are available in the Supporting Information. ## Datasets All mass spectrometry datasets used in this study are available to download, in their original data formats, at Edinburgh DataShare (www.http://datashare.is.ed.ac.uk/handle/10283/760), using the following link: http://dx.doi.org/10.7488/ds/2446.

chemsum

{"title": "Isotope Depletion Mass Spectrometry (ID-MS) for Enhanced Top-Down Protein Fragmentation", "journal": "ChemRxiv"}

the_effect_of_layer_thickness_and_immobilization_chemistry_on_the_detection_of_crp_in_lspr_assays

## Abstract: The immobilization of a capture molecule represents a crucial step for effective usage of gold nanoparticles in localized surface plasmon resonance (LSPR)-based bioanalytics. Depending on the immobilization method used, the resulting capture layer is of varying thickness. Thus, the target binding event takes place at different distances to the gold surface. Using the example of a C-reactive protein immunoassay, different immobilization methods were tested and investigated with regard to their resulting target signal strength. The dependency of the target signal on the distance to the gold surface was investigated utilizing polyelectrolyte bilayers of different thickness. It could be experimentally demonstrated how much the LSPR-shift triggered by a binding event on the gold nanoparticles decreases with increasing distance to the gold surface. Thus, the sensitivity of an LSPR assay is influenced by the choice of immobilization chemistry.The C-reactive protein (CRP) is an important biomarker for inflammation and infection of the human body 1-4 . CRP is an acute phase protein of the pentraxin family formed in the liver, as a marker of general or post-operative infectious diseases primarily for bacterial infections 5 , acute myocardial infarction, and other diseases 6 . In healthy people, the concentration of CRP in serum is below 10 mg/L. Between 10-40 mg/L is typical for mild inflammation and viral infections, while active inflammation and bacterial infections result in levels of 40-200 mg/ L 3 . Thus, the CRP level correlates with the stage of the diseases and is a decisive criterion for the prescription of antibiotics for the patient 7 . Therefore, diagnostic detection is very important. Additionally, CRP detection allows for a discrimination between bacterial and viral infections 8 . The most used diagnostic methods for CRP are rapid point-of-care tests (POCT) based on lateral flow-assays with a sensitivity of 10 mg/L. Though surface plasmonic resonance sensing for CRP is becoming more common there are still very few such methods close to diagnostic use [9][10][11] . Several plasmonic nanoparticle-based methods are established in enzyme-linked immunosorbent assay (ELISA) platforms. These assays use labeled secondary antibodies in sandwich assays [12][13][14][15][16] or metal-enhanced optical signals by enzymatic deposition 15 or metal-enhanced fluorescence 17 for the signal enhancement.Direct detection of CRP with plasmonic nanoparticles is possible altogether avoiding labels and secondary antibodies. A simple detection is the main advantage of colorimetric assays. The binding of CRP on the particles either stabilizes against salt-induced aggregation, or competes with bound aptamers, leading to destabilization of the nanoparticle solution 18 . Besides colorimetric detection, which yield somewhat qualitative results, plasmonic nanoparticles can also act as transducers. This is accomplished by using the change in spectroscopic properties (resonance wavelength) upon refractive index change (binding of molecules on the surface) enabling quantitative detection. Examples include the direct binding of CRP on anti-CRP antibodies (anti-CRP-AB)-modified gold nanospheres 19 , nanorods modified with single chain variable fragment (scFv) 20 , or silver nanoprisms modified with cytidine 5´-diphosphocholine (PC) 21 . In combination with straightforward optical readout units, LSPR sensors can also provide a new field of application for on-site diagnostics.Because the capture-antibody-immobilization determines the critical analytical parameters such as sensitivity, reproducibility, and robustness, this step has to be adapted and optimized for a given assay on a given technological platform. In the case of LSPR, gold nanoparticles (AuNP) represent the sensors used, and have to be modified with the detection antibodies. A wide range of antibody immobilization approaches have been developed during the past few decades, starting with passively adsorbing the antibodies on the substrate, and subsequently establishing various functionalization and cross-linking strategies, overcoming certain shortcomings of earlier ## Materials and methods Materials. Poly(allylamine hydrochloride) (PAH), polystyrene sulfonate (PSS) were purchased from Sigma-Aldrich Chemie GmbH (Taufkirchen, Germany), N-Hydroxysuccinimid (NHS), glacial acetic acid, sodium chloride (NaCl), sodium hydroxide (NaOH), glycine, 1-Ethyl-3-(3-dimethylaminopropyl)carbodiimid-hydrochlorid (EDC), ethanol, HCl, 3-triethoxysilylpropylamine (APTES), 10 × PBS Buffer and BSA were purchased from Carl Roth GmbH (Carl Roth GmbH + Co. KG, Karlsruhe, Germany). Sodium acetate was purchased from Merck KGaA (Darmstadt, Germany). Human C-reactive protein, anti-hCRP-capture-antibody biotinylated (α-hCRPcb) and non-biotinylated (α-hCRPc) and anti-hCRP-detector-antibody (α-hCRPd) was purchased from Senova GmbH (Weimar, Germany). All proteins were used as received without further purification and all used antibodies are monoclonal. Sodium acetate buffer (NaAc) was prepared by dissolving sodium acetate close to 10 mM in ultrapure water, pH was adjusted with HCl/NaOH to 4, 4.5 or 5 and filled up with ultrapure water to a final concentration of 10 mM NaAc. Glycine-HCl Buffer (Gly-HCl) was pre-pared similarly but with glycine and adjusted to pH 2.5. AuNP-chip preparation. Schott borosilicate wafers were cut into 25 × 16 mm glass substrates. After washing with soap and water by hand, they were cleaned 10 min each under sonication in water, ethanol, acetone, rotisol, ethanol and water. Finally, they were blow-dried with nitrogen. The glass chips were activated by treatment with oxygen plasma etching for 1 h at 380 W in a 200G Plasma System (TePla GmbH, Wettenberg, Germany) and afterwards directly transferred in a preactivated (10 min stirring) 1% APTES solution with 1 mM Figure 1. Scheme of the studied immobilization approaches for anti-CRP antibodies. Left: Biotinylated anti-CRP antibodies are attached to the gold surface by thiolated streptavidin. Right: After a surface modification by a self-assembled monolayer of MUA, EDC chemistry is utilized to attach unmodified anti-CRP antibodies. The inset shows the decrease of the sensor signal with increasing surface distance for layer-by-layer (LbL) deposition with charged polyelectrolyte (PEL) bilayers. acetic acid for 10 min. A 5 min sonication in ultrapure water was the next step before the glass chips were again blow-dried with nitrogen and stored under an argon atmosphere or were used directly. On the respective APTES chips a droplet of 20 µl 10 × concentrated spherical 80 nm AuNP from BBI (British BioCell International, Cardiff, UK) were deposited and incubated for 1 h at room temperature. If coverslips were used, they were treated similar. The produced AuNP chips were then dipped two times in water for washing and carefully blow-dried with nitrogen. Stored in a closed container, the chips can be used from several month up to years. ## LSPR instrument. All microfluidic assays used a custom build LSPR instrument (figure S3) at IPHT-Leibniz. It consists of a halogen light source HL-2000-FHSA (Ocean Optics, USA), an optical fiber connected UV/ VIS linear photodiode array spectrometer USB 2000+ (Ocean Optics, USA), a peristaltic pump (Ismatec Reglo-ICC, Cole-Parmer GmbH, Wertheim, Germany), a 2-way valve (Bio-Chem Fluidics Inc, Boonton, USA) for the waste and a custom designed 3D printed microfluidic chamber (figure S4) and sealing with two inputs and one output capillary at each channel side similar to an earlier published setup . For some measurements singleuse flow cells "Basic sensor platform II" (# 10001354, microfluidic chip shop GmbH, Jena, Germany) were used with adhesive tape gasket for Fl. 1005-rhombic chamber shape # 10001361. The in-and outputs were used in the same way as for the 3D printed chamber by blocking unused channels with a plug from the same company. For the pump and valve control, a custom-built Python program was used which also records the spectral information and calculates the centroid position of the LPSR peak in nm which is then visualized in a sensogram (plot of peak wavelength against time). The evaluation of the measured sensogram was also done via a Python script to extract mean values, standard deviations of mean values and wavelength shift (Δλ) values. Dried AuNP-Chips with or without SAM were attached to 3D printed microfluidic chamber with specific 3D printed sealing or glued in the commercial chamber. If the commercial chamber was used, it is mentioned in the method. All chips were cleaned before use (in the case of SAM before the SAM deposition) 3 min under ozone (UV ozone cleaner UVC-1014 Nano-BioAnalytics, Berlin, Germany). Before and after each measurement, the whole system was flushed with water for at least 15 min. All solutions used in the microfluidic system were filtered with a 0.22 µm Syringe filter (Carl Roth GmbH + Co. KG, Karlsruhe, Germany) and afterwards degassed for at least 30 min under vacuum (Air Admiral, Cole-Parmer GmbH, Wertheim, Germany) in a desiccator (Nalgen, Thermo Fisher Scientific, Waltham, USA). At the beginning of each measurement, a lamp spectrum without an AuNP-chip was recorded for background subtraction and noise reduction. The specific refractive indices of the used buffers were measured with a portable refractometer PAL-RI B331500 (ATAGO, Tokio, Japan) and entered into the software. Afterwards, a calibration script was used to calculate the bulk sensitivity of the chip by alternately flushing two different buffers with different refractive indices and recording the associated centroid wavelength. The bulk sensitivity (S B ) is calculated according to Eq. 1: where Δλ is the wavelength shift of the centroid position and Δn b is the refractive index change for the specific buffer solutions. Three times the standard deviation of a mean over at least 50 s of buffer injection (using as reference or blank) was used as threshold value for the limit of detection (LOD) calculation. Layer by layer deposition. For a pure PAH/PSS assay without any CRP, both 1 mM PAH and 1 mM PSS (with respect to the monomer) were dissolved in 0.1 M NaCl, respectively. The solutions were pumped alternately over the AuNP-chip with a 100 s buffer (0.1 M NaCl) injection before and after each 150 s polyelectrolyte injection. The flowrates were set to 10 µl/min. The surface sensitivity (S S ) is calculated according to Eq. 2: where λ is the wavelength of the centroid positions (Δλ = λ bilayer n − λ bilayer n−1 ) and n l is the refractive index of the layer (Δn l = n bilayer − n buffer ). The decay length (l d ) of the immobilized particles and the refractive index sensitivity (m) was calculated by plotting the recorded plasmon shifts (Δλ) against the layer thickness (d) and fitting the data with Eq. 3 31,32 . The layer thickness of a PAH/PSS bilayer in aqueous environment is ~ 4 nm 33 and the refractive index of the bilayer is ~ 1.5 RIU except for the first two to three bilayer 34 . With respect to the refractive index of the 0.1 M NaCl buffer which was 1.3305 RIU, Δn = 0.1695 RIU. Equation 3: For the measurement of the CRP deposition on three or four PAH/PSS bilayers, coverslips and the commercial flow-chamber were used. Therefore, 2 mg/ml PAH, 2 mg/ml PSS and 31 µg/ml CRP each were dissolved in 0.5 M NaCl buffer. The PAH and PSS solutions were pumped alternately over the AuNP-chip with a buffer injection of 0.5 M NaCl before and after each polyelectrolyte injection up to 3 or 4 bilayers. Then the CRP was injected followed by another buffer step. The flowrates were set to 20 µl/min and all injection times were 150 s. ## CRP-assay design. For all the assay measurements, the flowrates were set to 10 µl/min except for capture and target solutions which were pumped with 5 µl/min. For a clear distinction between bulk signals and binding events, the same buffer was injected before and after all reagents. In general, NaAc-Buffer (pH 4; 4.5 or 5) was used for immobilization and 1 × PBS as running buffer (Supporting Table S1). While running the first step, the channel for the second step was already pre-flushed with 3 µl/min directly to waste as well as further steps analogously. This was ensured by injecting the solutions alternately from different sides of the chamber, and opening www.nature.com/scientificreports/ the corresponding outlet valves. Due to the pre-flow and alternating flow techniques, the selected solutions ran through the channel directly after switching to the corresponding step. The software records all the injection steps accurately. Due to a time resolution of the spectrometer in the second range (1.5-3 s per measuring point), kinetic measurements are also possible. After the regeneration, the chip could be used for further target injections or a negative control. All capture reagents and BSA were dissolved in immobilization buffer. Targets and negative controls were dissolved in running buffer. ## Chemical immobilization of α-hCRPc via EDC/NHS. Carboxyl-groups, which are required for EDC/ NHS coupling, were realized on the AuNP-Chips with MUA as SAM. Therefore, the chips were incubated overnight in 1 mM MUA, 0.5 mM MUA with 5 mM MUD or 0.5 mM MUA with 5 mM 1-OT in ethanol to obtain different SAMs of MUA, MUA/MUD or MUA/1-OT on AuNP-Chips. MUA, MUD and 1-OT were purchased from Sigma (Sigma-Aldrich, Munich, Germany). Before using the chips, they were washed again shortly in ethanol and water. To identify the optimal pH for the immobilization buffer, 0.25 mg/ml α-hCRPc was dissolved in buffers with different pH and an immobilization pH scouting was carried out 35 . Therefore, the capture solutions pH 4, 4.5 and 5 were shortly pumped over the non-activated surface and electrostatically bound molecules were washed away with ethanolamine. The solution with the highest λ shift-pH 5-was chosen for the chemical immobilization. To obtain a covalent binding of the capture antibody to the AuNP-Chip, the carboxyl groups were activated by flushing 0.4 M EDC and 0.1 M NHS parallel with 5 µl/min from the same side in the chamber (finally 0.2 M EDC and 0.05 M NHS). After this activation, the capture solution was flushed over the surface for 400 s and the antibodies were bound randomly via their free amino groups (e.g., lysine). In the next step, the chip was flushed with ethanolamine for 500 s to wash away unspecific bound proteins and to react with the remaining activated carboxyl groups 36 . BSA blocking with 10 mg/ml in NaAc pH 4 for 300 s and a flow rate of 10 µl/min was still necessary. Thiol-streptavidin mediated immobilization of α-hCRPcb. Thiol modified streptavidin (SH-SA) was purchased from Protein Mods LLC (Waunakee, USA) and used as received without further purification or dilution. Blank AuNP-Chips can be easily functionalized with SH-SA by flushing a solution of 1 mg/ml over the surface for 200 s with 5 µl/min. In order to sufficiently block the sensor surface, various concentrations of BSA were tested in different buffers. Starting from 1 mg/ml over 10 mg/ml in PBS up to 1; 2 and 2.5 mg/ml in NaAc pH 4. Flow rates and injection times were also slightly varied between 5/10 µl/min and 200/250 s. After the blocking step a solution 0.25 mg/ml of biotinylated antibody α-hCRPcb was injected for 300 s with 5 µl/min. So, the chip was ready to use for hCRP detection without further blocking steps. For immobilization outside the chamber, one drop each of the SH-SA and the α-hCRPcb solutions were pipetted onto the AuNP chip, one after the other. Each of these were incubated for 1 h at 23 °C with 15-20% humidity, prior to washing 10 min with PBS (150 rpm horizontal shaking) and flushed with ultrapure water and dried in nitrogen stream (here called SH-SA outside). ## Results Signal and sensitivity dependency on layer thickness. LSPR detection yielded the sensor response for attachment or binding events on the sensor surface. The well-established layer-by-layer deposition using PAH and PSS was utilized to demonstrate the sensing approach 34, : The deposition of each additional layer on a gold nanoparticle chip (80 nm spheres) in a microfluidic system resulted in a longer wavelength (red) shift of the resonance peak. This is visible as individual steps in the sensogram, which plots the LSPR response over time (Fig. 2 inset). The first two to three bilayers were improperly formed due to a higher water content, resulting in lower refractive index and lower signal shifts than expected. After the 3rd bilayer, the refractive index of the bilayers was approximately 1.5 34 , the bilayer thickness in the wet state was approximately 4 nm 33 . The centroid of the LSPR-peak was measured over time and the shifts for each bilayer were plotted against the number of layers (Fig. 2). The evidence shows the signals for the bilayers decreased with increasing number of bilayers. This is due to a higher surface distance, where the field decay results in decreasing values. Also, the surface sensitivity was decreasing with increasing distance to the surface of the gold nanoparticles. With Eq. 3 the decay length (l d ) of this AuNP-chip was calculated to be 79.5 nm with refractive index sensitivity (m) = 141.1 nm/RIU. The before measured bulk sensitivity (S B ), calculated with Eq. 1 using the calibration, was with S B = 115.0 nm/RIU, i.e. slightly lower. For 3 bilayers, the S S was calculated to be 66.9 nm/RIU. Although silver particles would provide better plasmonic properties, the higher chemical stability of gold results in a broader use of gold nanoparticles in LSPR sensing 40 . The spherical 80 nm gold nanoparticles therefore used in our investigations were developed over the years 28,41 as a good compromise between larger (and therefore more sensitive, but less homogenous synthesis) and smaller (less sensitive but better reproducible regarding shape and size distribution) sized nanoparticles. Anisotropic shaped particles would provide higher sensitivity, but are more complex to synthesize and usually difficult to biofunctionalize, therefore they were not utilized in this study. ## CRP detection. Immobilization via thiolated streptavidin. It is well established that thiolated compounds form molecular layers on gold surfaces. One example are thiol-alkanes as well-studied model systems in the field of SAMs. On the other hand, biotin/(strept)avidin coupling is a powerful platform for nanoscale fabrication with many different applications in science, medicine, and nanotechnology. Combining these two well-established and straightforward attachment approaches, a scheme using thiolated streptavidin layers on the gold nanoparticle surfaces, and subsequent attachment of biotinylated antibodies on the SAMs, was utilized to prepare sensor substrates for CRP detection (Fig. 1 www.nature.com/scientificreports/ The preparation and subsequent performance of these sensor substrates for CRP detection is documented in Fig. 3. In the beginning (1), immobilized gold nanoparticles in a 10 mM NaAc pH 5-filled fluid cell result in a localized surface plasmon resonance (centroid) of 532.01 nm. Then, the cell was flushed with thiol-streptavidin, the resulting significant increase in centroid wavelength (to 534.388 nm) indicates strong binding on the gold nanoparticle surface (2). After buffer washing steps (3,4) a 1 mg/ml BSA passivation (5) induced only a small signal shift of 0.033 nm, which pointed to a weak BSA absorption on the gold particles in PBS buffer. Now, after a wash step (6), the biotinylated antibody (α-hCRPcb) was binding (7) with a shift of 0.521 nm. Afterwards, again the buffer (8) was injected, and a negative control (9) was conducted by applying the secondary antibody (α-hCRPd), which should bind on CRP only in case of a successful capturing. Because no CRP was present at this moment, this secondary antibody had no specific binding partner. However, the LSPR signal increased, indicating www.nature.com/scientificreports/ the presence of the secondary antibody at the surface. However, the subsequent washing step (10) appeared to remove it completely, the signal decreased to the previous level (cf. levels at 8 and 10). Next a regeneration step was introduced to check for-and remove-loosely bound molecules from the surface (11). The utilized solution seemed to have a lower refractive index as the standard buffer in 10, so the initial steep decrease in 11 can be attributed to this difference. However, afterwards the buffer from 10 was used again (12), allowed for a reproducible measurement: about 0.3 nm decrease between 10 and 12. Comparing 12 with 6 indicates that some of the attached biotinylated antibodies (7) were still on the substrate. After this preparation of the sensor surface, the actual CRP sensing step followed: 310 µg/ml CRP was flushed through the liquid cell leading to a measurable signal increase (13). This slightly decreased in a subsequent washing step (14), so that an overall CRP signal of about 0.2 nm resulted. Now, the already introduced secondary antibody was applied (15), testing on one hand the specificity of the CRP binding and, on the other hand, demonstrating possible signal amplification. There was a significant signal decrease of about 0.4 nm after a subsequent washing (16), which indicates a removal of some of the secondary antibodies. However, when CRP was flushed in again (19), this time one-tenth of the original concentration, the similar increase in signal was observed as before in step 13. This points to the fact that a saturation of the binding capacity was reached, even with the low concentration. A closer look at 13 and 19 reveals that the initial part of the curve was steeper in the case of higher concentration (13) compared to the lower one (19). When the secondary antibody was flushed in again (21), the same signal increase resulted as before in 15, indicating good reproducibility. This sensogram was selected to provide an example to highlight potential problems and challenges such as sensor regeneration, insufficient blocking or non-specific binding. Details will be discussed in the following paragraphs. Limit of detection. Despite a significant lower response of SH-SA mediated CRP binding the measured limit of detection was 0.3 µg/ml (0.3 mg/L). With suitable regeneration solutions, the target molecules could be washed away, so that the sensor chip was available for another sample. Considering three times the standard deviation of a PBS buffer step (blank), the LOD was close to 0.3 µg/ml (Fig. 3 inset) 42 . In addition, the LOD could be improved by using signal amplification with secondary antibodies (sandwich immunoassay). Here, the secondary antibodies were used to evaluate the specificity of the assay. In an example measurement (Fig. 3), blocking with 1 mg/ml BSA in PBS was insufficient since unspecific binding of α-hCRPd in seen in step 9. But, as visible in step 10, the buffer was quickly washing away the "negative control" in the presence of CRP (steps 15/16 & 21/22). We assume that were close to, or above, the upper limit of quantification with the concentrations shown in Fig. 3 for CRP 1:10 (310 µg/ml). We support this claim with the observed shift for 310 µg/ml CRP being slightly lower than for 31 µg/ml CRP. The higher concentration still showed a higher shift during the injection (Step 13), but decreased relatively quickly in Step 14, which could indicate a saturation of the available binding sites. Another explanation for the decreased signal at higher concentrations could be the "hook effect", whereby the effectiveness of antibodies to form immune complexes is sometimes impaired when concentrations of an antibody or an antigen are very high. This can occur in sandwich immunoassays as well as in competitive format at high target concentrations and was also manifested by lower signal at higher concentrations 43 . However, such a high 31 µg/ml CRP signal as shown in Fig. 3 could not be confirmed with any other SH-SA assay. The other measurements were significantly lower, as shown in the comparison of the mean values in Fig. 5 and also in the LOD inset (Fig. 3). The regeneration with 10 mM glycine-HCl pH 2.5 also appeared to be too harsh, which was noticeable by the fact that the buffer steps showed slightly lower centroid signal after each regeneration (steps 11/12; 17 /18 & 22/23). This problem could not be solved. Still after 5 or more regenerations the signal was stable, and the target shift was comparable. Immobilization pH scout for EDC/NHS chemistry. The efficiency of a chemical coupling of a capture molecule to a SAM via EDC/NHS strongly depends on the electrostatic interaction between the molecule and the surface. The pH value of the used immobilization buffer should be preferably adjusted between the isoelectric point (pI) of the capture molecule and the pKa of the SAM 35 . In this way, an electrostatic concentration of the protein to be immobilized will take place on the SAM surface, ensuring optimal binding. It is known from the literature that the pKa of MUA SAMs on particles is significantly higher (pKa = 6.8 for particles of 5 nm diameter and pKa ≈ 10 for flat surfaces) than for MUA in solution (pKa = 4.8) and also increases with increasing particle size 44,45 . Even the concentration and the size of surrounding ions have an impact on the pKa of MUA on nanoparticles, as well as the presence of other SAM-forming molecules such as 1-OT or MUD 44 . The pI value of the α-hCRPc used was not known and therefore immobilization scouting was performed with different pH values (4, 4.5 or 5) in 10 mM NaAc buffer (Figure S1). 1 M ethanolamine hydrochloride pH 8.5 was used to remove electrostatically bound protein from the not yet activated MUA surface. The highest plasmon shift was visible at pH 5. This pH was therefore chosen for the final immobilization. Immobilization via MUA and subsequent EDC/NHS chemistry. The results with the thiolated streptavidinimmobilization chemistry (reported in the previous section) yielded a stable and measurable signal for the studied relevant CRP concentrations. However, as the scheme in Fig. 1 shows, the resulting construct of streptavidin and anti-CRP antibody spans a significant distant away from the sensor surface. As demonstrated in the layer-by-layer adsorption experiments in Fig. 2, an increasing distance from the surface hampered the achieved signal. In order to address this shortcoming, another immobilization approach was considered, which would allow decreasing the distance of the binding target to the sensor surface. It is based on a MUA self-assembled monolayer, binding on one side via thiol to the gold, and providing the means for an EDC/NHS-attachment of the anti-CRP antibody on the other side (Fig. 1, right). www.nature.com/scientificreports/ Blocking for EDC/NHS chemistry. A blocking step was essential to ensure that the "target" molecule under investigation was bound specifically to the immobilized captured molecule and not already adhered non-specifically to the sensor surface. In the initial SH-SA measurements with 1 mg/ml BSA in PBS as blocking reagent, nonspecific binding was clearly visible (Fig. 3). By increasing the BSA concentration to 10 mg/ml and changing the buffer to 10 mM NaAc pH 4, the blocking of the sensor surface was significantly improved. The step 9 in Fig. 4 showed a sufficient blocking with 10 mg/ml BSA, demonstrated by the secondary anti-human CRP antibody (α-hCRPd) as negative control in step 12, which showed no detectable unspecific binding. Only after CRP injection (step 16) was the α-hCRPd able to bind specifically, which resulted in a significant shift (step 18). On the other hand, the observed binding of α-hCRPd in the SH-SA assays (Fig. 3 step 9) could be due to a weak cross-reactivity to SH-SA. This would also explain why the sensogram course in Fig. 3 Steps 16 and 22 (dissociation of α-hCRPd after CRP injection) drops similarly as before CRP in Step 10. Such a dissociation course (obviously no 1:1 binding 46 ) is not observed in the sensograms of the measurements without SH-SA (SAM method, Fig. 4 Steps 12/13 and 18/19). Since this effect was not investigated further, we chose the term 'nonspecific' . ## Discussion When considering a LSPR experiment for the detection of a target molecule one should consider the optimal immobilization method of the capture molecule. It depends on the (physico)chemical nature of the capture molecule (DNA, proteins, lipids and others) and the possible required modifications for its immobilization. Proteins, for example, are known to bind directly on bare gold surfaces. However, this can have an impact on their stability and functionality 30 and can result in weakened target binding. Different immobilization methods have been developed in the past to address this challenge, as described above. Another important factor for an efficient target binding (and therefore signal strength) is the distance of the binding site from the surface. This parameter can be modulated by the choice of the immobilization chemistry, as highlighted in this study. The PAH/PSS bilayer deposition experiments yielded an optimal range for the distance of the target molecules from the sensor's surface. For the AuNP chips used in this study, the first 24 nm from the surface would result in a significant shift (Δλ) of at least 1.5 nm for the binding of a well packed PAH/PSS bilayer with n layer = 1.5 RIU. If the decay length (l d ) would be higher, the signal shift for each bilayer would be smaller in comparison to shifts on particles with lower l d . On the other hand, the decrease of the signal shift for each following bilayer would also be smaller because the total sensing distance is higher, but the absolute maximum shift would be the same. For higher absolute maximum shift the increase of the refractive index sensitivity (m; m = S B if d approaches to infinite 32 ) would be necessary, which can be attained with bigger particles, other materials or particles with anisotropic shapes 32,47 . The calculated m and l d with 141.1 nm/RIU and 79.5 nm for the utilized AuNP-chips differ slightly from the values expected by theory. This is reasonable because the refractive index of the layer and the layer thickness were taken from the literature. In the case of the CRP deposition, the measured plasmon shift was lower than 1.5 nm. This is confirmed in supporting Figure S2, 1.212 ± 0.032 nm/0.918 ± 0.012 nm on the 3rd/4th PAH/PSS bilayer corresponding to 12/16 nm away from the gold nanoparticle surface. Because of the water content the refractive index of a CRP layer is reported to be ≤ 1.45 RIU 48,49 which is also reflected by a lower plasmon shift. It is observed that at least one bilayer more, approximately 4 nm, reduces the signal of the target significantly (by roughly 25%). After the 3rd PAH/PSS bilayer, the total centroid shift was 10.095 nm/10.607 nm, which indicates that the chips behave similarly having comparable surface and bulk sensitivities (S S = 61.0 nm/ RIU and 64.1 nm/RIU for three PAH/PSS bilayer; S B = 97.87 nm/RIU and 115.09 nm/RIU). The methods used for CRP detection in this work (SH-SA and SAM method) are based on covalent bonding to the gold surface. Both MUA (1-OT, MUD) and SH-SA have free thiol groups (-SH) that form a covalent Au-S bond at neutral to basic pH 50 . In the SH-SA method, antibody binding occurs via a biotin-streptavidin bond, which is one of the strongest non-covalent bonds in nature. In the SAM method, the antibodies are again covalently bound to the already anchored MUA via chemical immobilization. In this process, the free carboxyl groups first react with protonated EDC to form active O-acylisourea intermediates, which is then stabilized by NHS to form a succinimidyl ester 25 . Free amine side chains of the antibodies to be immobilized (e.g., lysine) can then react with the formed ester to create an amide bond with MUA 26 . This process is a zero-length crosslinking. Nevertheless, the orientation of the antibodies is most likely similarly random in both cases since the biotinylation of the antibodies also occurs via EDC/NHS in many cases. The binding of the antigen (CRP) then takes place via non-covalent bonds such as hydrogen bonding or hydrophobic interaction, which ensures the regenerability of the sensor 51 . Using different immobilization methods for the capture antibodies, method-dependent differences are visible (Fig. 5). For the samples 'SH-SA outside' , which was prepared outside of the fluid cell and air dried after antibody immobilization, a much smaller shift is observed, apparently air drying has a negative impact on the function of the detection antibody. However, for thiol-streptavidin (SH-SA) mediated immobilization, resulting in a greater surface distance, the plasmon shift for 31 µg/ml CRP was significantly lower than for the SAM method (independent samples T-test: p = 0.046; n ≥ 5 per group) which allows the target binding closer to the nanoparticle surface. On average, due to the thicker SH-SA layer, the target binding should be about 5 nm closer for the SAM method. On the other hand, it takes more time to produce the SAM chips (overnight), and also the EDC/NHS reaction is more time consuming during the assay. SH-SA immobilization and binding of the biotin capture antibody is straightforward and can be done completely in the microfluidic system. Blocking the surface against non-specific binding is necessary for both methods and could be sufficiently optimized for the assay. Regeneration of the sensor for antigen binding with 10 mM glycine HCl pH 2.5 was also successful. However, the LSPR signal after regeneration was always somewhat lower than before antigen binding, pointing to partial removal and/or inactivation of capture molecules, which indicates a need for optimization. The scatter of the CRP response was quite large. This could be due to less homogenous distribution of the AuNP including aggregations. This is also reflected in the bulk sensitivities (S B ) observed during calibration, as they showed a large variance (70.5-225.3 nm/RIU). However, as can be seen from the LbL data, this value cannot be equated with the calculated S B . Also, the measured surface sensitivities (S S ) of the AuNP chips show a significantly lower variance (61.0-66.9 nm/RIU after 3 bilayer of PAH/PSS). With the detection limit of 0.3 mg/L CRP for SH-SA mediated immobilization, this immobilization method is fully sufficient for clinical applications and allows the measurement of several samples in one assay. When further improvement in sensitivity is required, the short-thiol mediated approach could increase the detection limit, but requires a more cumbersome preparation. ## Conclusions The effect of the thickness of the capture probe attachment layer on the LSPR signal has been characterized. Using plain gold nanoparticle chips the straightforward SH-SA method is sufficient to determine CRP in clinically relevant concentrations. The immobilization layer is significantly thicker than with short thiols and thus also provides lower target signals. Using LbL technology, distance layers comparable to both studied immobilization chemistries could be prepared and compared regarding their LSPR signal shift upon binding of detection antibodies. The regenerability of the used sensors and the high time resolution of the utilized setup enables versatile applications such as diverse binding kinetic studies. The studied detection method is well suited for the discernment of protein biomarkers such as CRP in clinical applications allowing the measurement of several samples in one assay.

chemsum

{"title": "The effect of layer thickness and immobilization chemistry on the detection of CRP in LSPR assays", "journal": "Scientific Reports - Nature"}

chemoselective_synthesis_of_uniform_sequence-coded_polyurethanes_and_their_use_as_molecular_tags

## Abstract: Gunay and colleagues synthesized digitally encoded polyurethanes via a facile orthogonal iterative solid-phase approach. The polymers formed exhibited uniform molecular structure and controlled monomer sequences. Furthermore, these coded polyurethanes were very easy to read by tandem mass spectrometry sequencing. Thus, these polymers can be used as readable molecular labels and therefore open up interesting avenues in product-identification and anticounterfeiting technologies. For example, sequence-coded polyurethane tags were included in the present work in polystyrene cast films and 3D-printed polymethacrylate sculptures. ## INTRODUCTION Commodity plastics such as polyesters, polyamides, and polyurethanes (PUs) are usually prepared by step-growth polymerization, 1 which is a straightforward but poorly controlled process. For instance, PUs are generally obtained by reacting a diisocyanate with a diol, leading to polymers containing two oppositely oriented carbamates per repeat unit (Figure 1A). Although such PUs are widely applied as sealants, coatings, and adhesives, 2 they exhibit ill-defined molecular structures with a dispersity value ä of about 2. In addition, polymers formed through a step-growth mechanism exhibit, in general, simple repeating monomer sequences. 3 In contrast, it has been shown in recent years that uniform sequence-controlled synthetic polymers (ä $ 1), which are as molecularly defined as natural polymers such as nucleic acids and proteins, can be prepared using molecular machines, 4 nucleic acid templates, 5,6 or multistep-growth approaches, in which monomers are assembled one by one on a support. 3 This new generation of precision polymers is promising for a wide range of applications 7 and, in particular, for the development of informationcontaining macromolecules. 8 In such materials, information is stored at the molecular level in the form of a coded monomer sequence and can be recovered using a sequencing technique. 9,10 Some types of information-containing macromolecules have already been reported, but their molecular structures differ markedly from those of standard commodity polymers. Here, we report that classic plastics such as PUs can also be used for information storage. Some routes for preparing sequence-defined oligocarbamates have already been reported in the literature, but they involve protecting-group-based protocols inspired by solid-phase peptide The Bigger Picture Product identification is an important topic in modern consumption societies. Indeed, the massive production of counterfeit goods leads to major economic losses and represents a threat to health and the environment. Thus, anticounterfeiting technologies are crucial in many areas such as the pharmaceutical, cosmetics, food, and chemical industries. In particular, methods allowing product labeling at the nano or molecular scale are promising because they may not be easy to mimic. For example, synthetic polymers containing defined sequences of monomers can be used as molecular barcodes for product identification. Here, we report that digital sequences can be written at the molecular level on polyurethanes, which are cheap commodity plastics, using a very simple synthesis procedure. Moreover, the binary sequences formed can be easily deciphered by mass spectrometry. These coded polyurethanes can be used as molecular tags and blended in small amounts in other plastic materials. synthesis. Synthetic chemistry can be greatly simplified by the use of protectinggroup-free strategies. 20 For instance, it has been shown that the preparation of uniform sequence-defined polymers is facilitated by the use of submonomer strategies involving chemoselective coupling steps, but no route of this type has been discovered so far for the synthesis of PUs. In the present work, we demonstrate that uniform sequence-coded PUs can be prepared using a very simple chemoselective multistep-growth process. In addition, it was discovered that the polymers formed are remarkably easy to sequence by negative-mode tandem mass spectrometry (MS/MS). These advantages make this new class of coded plastics potentially interesting for applications in anti-counterfeiting technologies. Indeed, the massive production of counterfeit goods is a major problem in the global world economy. Although different types of nanoscale materials have been proposed for product identification, 28,29 the use of molecular barcodes that can be directly blended in an organic or inorganic matrix is a tempting option. In that regard, cheap sequencecoded plastics could be an interesting alternative to DNA, which is primarily investigated nowadays in such applications. 9,30 Therefore, sequence-coded PUs were tested in the present work for the labeling of different polymer materials. ## RESULTS The strategy for synthesizing uniform polyurethanes relies on the use of two successive chemoselective steps as shown in Figure 1B. Although potentially applicable on a variety of supports, this chemistry was tested in the present work on a hydroxyfunctionalized crosslinked polystyrene resin. This solid support was obtained by modifying a commercial Wang resin with a hydroxy-functional linker (leading after cleavage to a C5 acid chain-end moiety, noted as a in this article). In a first coupling step 1, the hydroxy group was reacted with N,N 0 -disuccinimidyl carbonate (DSC) to afford an unsymmetrical active carbonate. 31 It is known that the adduct formed reacts readily with primary and secondary amines, 31 as well as azides, 32 to form carbamate linkages. Interestingly, it was shown that the reaction with amines is chemoselective and can be performed in the presence of unprotected alcohols. 31 Thus, in step 2, the adduct formed was reacted with an amino alcohol to afford selectively a hydroxy-functional carbamate. The iterative repetition of coupling steps 1 and 2 allows facile protecting-group-free synthesis of uniform polyurethanes. Various monomers were tested in this study to verify this concept (Figure 1C). First, model homopolymers were synthesized using DSC and C5 0 . Different solvents, reactions times, and temperatures were screened to identify the optimal experimental conditions for the synthesis of uniform PUs. The use of acetonitrile and microwave irradiation in step 1 allowed synthesis of uniform polymers with ä $ 1 (P1 in Table S1). For instance, Figure S1 shows single-peak high-resolution electrospray mass spectra obtained for homopolymers of C5 0 (P1 in Table S1). These results imply that steps 1 and 2 are near quantitative and that no side reaction competes significantly with the iterative synthesis. Given these promising results, monomer alphabets were then developed in order to implement readable coded sequences in the uniform PUs. As shown in our previous works, 13,33,34 the use of a simple H/CH 3 molecular variation is sufficient to implement a 0/1 digital code that can be deciphered by MS/MS. Following this established principle, amino alcohols with or without methyl side groups were used to build sequence-coded PUs. In particular, two different binary monomer alphabets, C4 0 /C4 1 and C3 0 /C3 1 (Figure 1C), were tested in the present study. Both languages allowed the synthesis of uniform copolymers (P2-P16 in Table S1). As an example, Figure 2A shows the electrospray ionization (ESI)-MS and MS/MS spectra obtained for a copolymer (P9 in Table S1) synthesized with monomers C4 0 /C4 1 and containing the information sequence a-10001. The ESI-MS spectrum indicates the formation of a uniform structure. It is important to specify that this spectrum corresponds to a final product that was not purified by high-performance liquid chromatography or any other fractionation method. Results of comparable quality were obtained for all polymers (Figures S1-S13), including those synthesized with the alphabet C3 0 /C3 1 (P15 and P16 in Table S1 and Figures S14 and S15). Figures S16 and S17 also show typical nuclear magnetic resonance (NMR) spectra and size-exclusion chromatograms that were recorded for sequence-coded PUs. Sizeexclusion chromatography (SEC) measurements indicated the formation of uniform polymers with apparent ä values of about 1.01. Interestingly, the coupling step 2 seems to lead to similar yields using methylated (1-bit) or non-methylated (0-bit) amino-alcohol monomers. This is a noteworthy advantage over previously reported types of digitally encoded polymers such oligo(triazole amide)s or oligo(alkoxyamine amide)s, 11,13 in which the coupling yields can be affected by the molecular structure of the building blocks. Furthermore, the overall yields of PU synthesis were near quantitative in all cases (Table S1). The formed digitally encoded polymers were also remarkably easy to sequence by negative-mode MS/MS. In positive-ionmode MS/MS, protonated PUs undergo C-O and C-N carbamate fragmentations that lead to relatively complicated spectra (Figure S18). 18 In comparison, activation of deprotonated oligomers in the negative-ion mode leads only to C-O carbamate bond cleavage, and therefore the digital sequences written in the polymers are very easy to read (Figure 2A). It is important to specify that this analysis method permits unequivocal detection of a coded binary sequence. For instance, isobaric oligomers with different sequences can be easily distinguished by negative-mode MS/MS (Figure 2B). Although deprotonation occurs on the terminal carboxylic acid, the analysis of longer coded sequences is also possible, and the quality of sequencing is not affected by chain length (Figures 2C and S13). Hence, sequence-coded PUs seem to be an interesting class of materials for anticounterfeiting applications. Indeed, these polymers can be easily encoded through a facile chemoselective process and decoded in a very short time. These polymers S1). (B) MS/MS discrimination of two isobaric oligomers containing different binary monomer sequences (P10 and P11 in Table S1). (C) MS/MS characterization of a sequence-coded oligomer containing a byte of binary information (P13 in Table S1). were blended in different types of polymer matrices to demonstrate their relevance as molecular tags. Thermogravimetric analysis of these short coded polymers indicated that thermal decomposition starts to occur around 150 C (Figure S19), thus excluding their use in high-temperature processing conditions such as hot-melt extrusion. However, these molecular tags can be dispersed in solid polymer materials via many other procedures, such as film casting, mechanical blending, spray deposition, or in situ polymerization. For instance, Figure 3A shows the analysis of a polystyrene film in which a small amount of a coded PU tag was dispersed. Polystyrene and PUs are, in general, thermodynamically immiscible. 35 However, in this study, the molecular tags were used at very-low-weight fractions and could therefore be dispersed in a polystyrene matrix. The homogeneity of the PU tag dispersion was verified by studying different fractions of the polystyrene films that were cut, dissolved in deuterated tetrahydrofuran (THF-d 8 ), and analyzed by 1 H NMR. In all cases, the spectra contained specific PU signals, and compared with polystyrene peaks, they were integrated more or less the same in all fractions (Table S2). Although the formation of nanostructured PU domains cannot be excluded, this experiment suggests a homogeneous tag dispersion. After being included in the films, the PU tags can also be extracted and studied by MS. To do so, a small portion of the film was cut and immersed in methanol, which is a non-solvent of polystyrene. Even though the sequence-coded polyurethanes are also poorly soluble in methanol, they can be selectively extracted in sufficient quantity to be analyzed by ESI-MS. As shown in Figure 3A, a single-peak ESI spectrum was obtained, and S1). The oligomer barcode could be extracted from the film by selective dissolution in methanol and characterized by ESI-MS and MS/MS. (B) Labeling of a 3D-printed crosslinked poly(methacrylate) structure, in which 0.25 wt % of a polyurethane containing the binary sequence a-000101was incorporated (P12 in Table S1). The oligomer could be found by grinding the skirt of the 3D print into a thin powder, dissolving the powder in methanol, and analyzing the powder by ESI-MS and MS/MS. its MS/MS sequencing permitted easy recovery of the coded binary information. These results confirm that PUs tags can be used efficiently to label polymers films. As a second proof of concept, a sequence-coded PU tag was included in a 3D print obtained through methacrylate-based photopolymerization. As shown in Figure 3B, a sculpture representing a 3D DSC molecule was printed. In 3D printing, a significant amount of polymerized matter (i.e., draft, skirt, and brim) is unused and separated from the sculpture after printing. Small amounts of these residues can be kept attached to the sculpture as an anti-counterfeit strap that can be cut and analyzed. In the present case, some parts of the skirt were ground into a thin powder that was dissolved in methanol. MS analysis of this solution led to a mixture of peaks in which the signal of the PU tag had very weak intensity (inset of MS in Figure 3B). However, this peak can be found and efficiently sequenced by MS/MS. These results indicate that a sufficient fraction of the PU-coded tag survives the photopolymerization process and therefore enables practical 3D-sculpture labeling. ## DISCUSSION A practical new class of information-containing macromolecules that can be easily synthesized and sequenced was identified in the present work. The orthogonal iterative approach introduced here is probably the easiest method ever reported for the synthesis of uniform sequence-defined PUs. The two consecutive coupling steps studied in this work proceed in near-quantitative yields and are chemoselective. As a consequence, uniform macromolecules with a predetermined molar mass can be prepared in high overall yields using this process. Taking into account that the frontier between oligomers and polymers is often assumed to be 1,000 Da, 36 the present method is not restricted to the synthesis of short oligomers but can be used to prepare sequence-coded polymers, as shown in the present work with the synthesis of samples P13, P14, and P16. Moreover, the iterative process is experimentally convenient and can most probably be automatized for the synthesis of longer macromolecules. In addition, these uniform polymers can be molecularly encoded using pre-defined monomer alphabets. 9 In the present work, digitally encoded polymers were prepared using two different binary molecular alphabets in which non-methylated and methylated amino-alcohol monomers were used as 0and 1-bit motifs, respectively. It was observed that coupling step 2 led to high yields with both methylated and non-methylated building blocks. This constitutes an obvious advantage over previously reported classes of information-containing polymers that sometimes rely on the use of coded monomers of different reactivity. 13 The iterative process introduced here is not limited to the synthesis of digitally encoded polymers and could also be used for the synthesis of more complex coded macromolecules (e.g., using ternary or quaternary alphabets). 9 In addition, the coded monomers that can be used in this process do not necessarily have to be methylated or non-methylated building blocks. Many different types of coded side chains can potentially be used in this approach. Furthermore, this new class of sequence-coded PUs, containing a carboxylic acid group at their a-chain end, is particularly easy to sequence by negative-mode MS/ MS. These macromolecules exhibit high ''readability'' because they always deprotonate on their a terminus only and hence undergo solely predictable C-O carbamate fragmentations. This is a clear advantage over positive-ion-mode MS/MS sequencing, in which randomly protonated PUs undergo C-O and C-N carbamate fragmentations that lead to complex spectra. Hence, these PUs can be used as anti-counterfeit tags to label different types of materials. The efficient molecular labeling of cast polystyrene films and photopolymerized 3D polymethacrylate prints provide proof of principle. In both cases, the PU tags were found in the materials, and their coded sequences could be efficiently assessed by negative-mode MS/MS. Here, the use of common plastics such as PUs is an interesting advantage over previously studied information-containing macromolecules. The physicochemical properties of PUs have been extensively investigated and are well documented in the literature. 2 For example, the blending behavior of PUs with other conventional plastics such as polystyrene, polypropylene, and poly(methacrylate)s is well documented. 35 Thus, the use of PU molecular barcodes is a practical and valid option for anti-counterfeit protection of industrial polymer products, as well as valuable objects such as 3D-printed art. ## EXPERIMENTAL PROCEDURES Materials and Reagents The following were used as received: citric acid (Alfa Aesar, >99%), hydrochloric acid (HCl; Sigma-Aldrich, 37%), p-toluenesulfonic acid monohydrate (PTSA; Sigma-Aldrich, 98.5%), trichloroacetic acid (TCA; Sigma-Aldrich, 99%), 4-amino-1-butanol (TCI, 98%), 3-amino-2,2-dimethyl-1-propanol (TCI, 97%), 4-amino-2-methyl-1butanol (TCI, 98%), 5-amino-1-pentanol (TCI, >95%), 3-amino-1-propanol (Alfa Aesar, 99%), 4-(dimethylamino)pyridine (DMAP; Sigma-Aldrich, 99%), N,N,N 0 ,N 0 ,N 00pentamethyldiethylene-triamine (PMDETA; Aldrich, 99%), anhydrous pyridine (Sigma-Aldrich, 99.8%), sodium hydroxide (NaOH; VWR, 99%), triethylamine (TEA; Merck, >97%), acetic anhydride (Alfa Aesar, >99%), ammonium chloride (NH 4 Cl; VWR, R99.5%), 4-benzyloxybenzyl alcohol polystyrene (Wang resin 100-200 mesh; Iris Biotech, 0.94 mmol/g resin), (1-bromoethyl)benzene (Alfa Aesar, 97%), 2-oxepanone (Sigma-Aldrich, 97%), N,N 0 -dicyclohexylcarbodiimide (DCC; Alfa Aesar, 99%), 4,4-dimethoxytrityl chloride (DMT-Cl; ChemGenes), N,N 0 -disuccinimidyl carbonate (DSC; TCI, >98.0%), sodium bicarbonate (NaHCO 3 ; SDS, 99%), anhydrous sodium sulfate (Na 2 SO 4 ; VWR, R99%), anhydrous acetonitrile (dry ACN; Sigma-Aldrich, 99.8%), cyclohexane (Carlo Erba), anhydrous dichloromethane (dry DCM; Sigma-Aldrich, R99.9%, 40-150 ppm amylene), DCM (Sigma-Aldrich, R99.9%), diethyl ether (Carlo Erba), anhydrous N,N-dimethylformamide (dry DMF; Sigma-Aldrich, 99.8%), DMF (Sigma-Aldrich, R99.0%), ethanol absolute (VWR, 99.9%), ethyl acetate (Carlo Erba), and tetrahydrofuran (THF; Aldrich, 99%, stabilized with 2,6-di-tert-butyl-4-methylphenol). Copper(I) bromide (CuBr; Alfa Aesar, 98%) was purified by stirring in acetic acid and rinsing with ethanol and diethyl ether and then dried. Styrene (Sigma-Aldrich, R99%) was distillated over CaH 2 and then degassed by bubbling argon through it. Gray photopolymer resin for Form 1+ (Formlabs, version FLGPGR02) was used for 3D printing. A Monowave 300 (Anton Paar) microwave reactor was used for step 1 of the PU synthesis. Step 2 was conducted in solid-phase extraction (SPE) tubes using a KS 130 basic (IKA) shaker. ## Synthesis of the a-Linker for Modification of the Wang Resin This molecule was synthesized in four steps as shown in Scheme S1. Step 1: Synthesis of 6-Hydroxyhexanoic Acid 2 NaOH (4.4 g, 110 mmol, 2 equiv) was added to a stirred mixture of 2-oxepanone (6.1 mL, 55 mmol, 1 equiv) in water (150 mL). The reaction mixture was stirred at room temperature (RT) for 16 hr before being acidified (pH 3-4) by the addition of HCl and extracted with ethyl acetate (3 3 100 mL). The combined organic layers were dried over Na 2 SO 4 , filtered, and evaporated, affording the targeted compound as a crystalline white solid. Yield: 5.9 g (82%); 1 Step 2: Synthesis of Ethyl 6-Hydroxyhexanoate 3 PTSA (107 mg, 0.56 mmol, 0.013 equiv) was added to a stirred solution of 6-hydroxyhexanoic acid (5.9 g, 45 mmol, 1 equiv) in absolute ethanol (50 mL), and the reaction was stirred at 65 C for 16 hr. The solvent was evaporated, and the resulting oil was dissolved in ethyl acetate (50 mL) and washed with saturated aqueous NaHCO 3 solution (2 3 50 mL) and with brine (1 3 50 mL). The aqueous layers were combined and extracted with ethyl acetate (3 3 50 mL). The organic layers were dried over Na 2 SO 4 , filtered, and evaporated, affording the targeted compound as a yellow oil. Yield: 6.2 g (87%). 1 H NMR (400 MHz, CDCl 3 ): d = 4.12 (q, J = 7.2 Hz, 2H, CH 2 ), 3.65 (t, J = 6.3 Hz, 2H, CH 2 ), 2.31 (t, J = 7.4 Hz, 2H, CH 2 ), 1.55-1.69 (m, 4.5H, 2 3 CH 2 ), 1.40 (m, 2.5H, CH 2 ), 1.25 (t, J = 7.2 Hz, 3H, CH 3 ); 13 C NMR (100 MHz, CDCl 3 ): d = 173.9 (C=O), 62. 8, 60.4, 34.4, 32.4, 25.4, 24.8, 14.4; HRMS (ES+) m/z: [M + H] + calculated 161.1172, found 161.1169. Step 3: Synthesis of Ethyl 6-(Bis(4-methoxyphenyl)(phenyl)methoxy)hexanoate 4 A degassed solution of DMAP (0.8 g, 6.55 mmol, 0.5 equiv), DMT-Cl (5 g, 14 mmol, 1.1 equiv), and ethyl 6-hydroxyhexanoate (2.2 g, 13.7 mmol, 1 equiv) in anhydrous pyridine (25 mL) was stirred overnight for 16 hr at RT. The solvent was evaporated, and the resulting mixture was dissolved in ethyl acetate (100 mL) and washed with water (2 3 100 mL). The combined aqueous layers were extracted with ethyl acetate (150 mL). The organic layers were combined, dried over Na 2 SO 4 , filtered, evaporated, and purified by column chromatography (ethyl acetate/cyclohexane/TEA, 7:93:3) affording the target compound as a white oil. Yield: 6.0 g (95%); 1 H NMR (400 MHz, CDCl 3 ): d = 7.18-7.45 (m, 9H, Ar-H), 6.83 (m, 4H, Ar-H), 4.12 (q, J = 7.2 Hz, 2H, CH 2 ), 3.80 (s, 6H, 2 3 3H), 3.05 (t, J = 6.5 Hz, 2H, CH 2 ), 2.29 (t, J = 7.5 Hz, 2H, CH 2 ), 1.61 (m, 4H, 2 3 CH 2 ), 1.40 (m, 2H, CH 2 ), 1.25 (t, J = 7.2 Hz, 3H, CH 3 ); 13 C NMR (100 MHz, CDCl 3 ): d = 173.9, 158. 5, 145.5, 136.8, 130.2, 128.3, 127.8, 126.7, 113.1, 85.8, 63.3, 60.3, 55.3, 34.5, 29.9, 26.1, 25.1, 14.4; HRMS (ES+) m/z: [M + Na] + calculated 485.2298, found 485.2297. Step 4: 6-(Bis(4-methoxyphenyl)(phenyl)methoxy)hexanoic Acid 5 An aqueous NaOH solution (6 M, 10 mL) was added to a stirred solution of ethyl 6-(bis(4methoxyphenyl)(phenyl)methoxy)hexanoate (3) (2.36 g, 5.1 mmol, 1 equiv) in absolute ethanol (5 mL). The reaction was stirred for 16 hr. Saturated aqueous NH 4 Cl was added to the reaction mixture until a white precipitate formed. The mixture was filtered and the residue was acidified with a solution of citric acid and water (1% m/m, 100 mL) and extracted with DCM (2 3 100 mL). The combined organic layers were washed with water (50 mL), dried over Na 2 SO 4 , filtered, and concentrated under reduced pressure, affording the target compound as a white crystalline foam. Yield: 1.97 g (88%); 1 H NMR (400 MHz, CDCl 3 ): d = 7.18-7.46 (m, 9H, Ar-H), 6.8 (m, 4H, Ar-H), 3.79 (s, 6H, 2 3 3H), 3.06 (t, J = 6.6 Hz, 2H, CH 2 ), 2.34 (t, J = 7.6 Hz, 2H, CH 2 ), 1.62 (m, 4H, 2 3 CH 2 ), 1.43 (m, 2H, CH 2 ); 13 ## Resin Modification Wang resin (300 mg, 0.94 mmol/g, 0.28 mmol, 1 equiv) was placed into an SPE tube and swollen in DCM for 5 min. A solution of DMAP (53 mg, 0.43 mmol, 1.5 equiv), DCC (175 mg, 0.85 mmol, 3 equiv), and 6-(bis(4-methoxyphenyl)(phenyl)methoxy) hexanoic acid 5 (360 mg, 0.83 mmol, 3 equiv) in DCM was then added to the resin at RT, and the solution was shaken at RT for 36 hr. The resin was washed with DMF (103) and then with DCM (23) before proceeding to the next step. A solution of acetic anhydride in dry pyridine (1/5, 6 mL) was added to the SPE tube containing the modified resin, and the solution was shaken for 2 hr at RT. The resin was washed with DCM (53) before proceeding to the next step. A solution of TCA in DCM (3/97, w/w, 5 mL) was added to the SPE tube containing the modified resin, and the solution was shaken for 30 min at RT. The resin was washed with DCM (53). This step was repeated two times. The resin was then washed with diethyl ether (13). General Procedure for the Iterative Synthesis of Polyurethanes General Procedure for Coupling Step 1 N,N 0 -disuccinimidyl carbonate (6 molar equiv) was first solubilized in dry acetonitrile by gentle heating. Then, triethylamine (12 molar equiv) was added to the solution. The mixture was then added to a microwave-suitable glass tube containing a hydroxy-functional solid support (1 molar equiv of hydroxy functions), and the tube was closed under an argon flow. The reaction was then stirred for 1 hr in a laboratory microwave for chemical synthesis (60 C, 8 W). After the reaction, the support was transferred into an SPE tube and washed several times with N,N-DMF before proceeding to step 2. ## General Procedure for Coupling Step 2 A solution of triethylamine (20 molar equiv) and an amino alcohol (10 molar equiv) in dry N,N-DMF was added to an SPE tube containing a solid support activated by N,N 0 -disuccinimidyl carbonate as described in step 1. The tube was then closed under an argon flow, and the reaction was shaken for 20 min at RT. After the reaction, the support was washed several times with DMF and diethyl ether and eventually transferred back to a microwave tube for a further step 1. ## Iterative Repetition of Coupling Steps 1 and 2 Steps 1 and 2 were performed consecutively and repeated a certain number of times to reach the desired polyurethane chain length. At each step 2, the molecular structure of the amino-alcohol building block could be varied to synthesize polymers with controlled sequences of information. ## Cleavage Procedure The cleavage procedure depends on the type of support and linker used for the synthesis. The following describe a solid support containing a labile Wang-type ester linkage. When a polymer with the desired chain length was obtained, the solid support was transferred into a vial. The cleavage was performed by adding a mixture of TFA and dry DCM (5:5 v/v) to the resin. The cleavage reaction was conducted for 2 hr. The support was then filtered off, and the filtrate was collected. TFA and DCM were evaporated under reduced pressure to obtain the desired polyurethane as a white solid. ## Labeling of Materials ATRP Synthesis of Polystyrene CuBr (0.0156 g, 0.109 mmol) was added to a 25 mL flask sealed with a septum, and then the flask was degassed through argon-vacuum cycles. Styrene (5 mL, 43.5 mmol) and PMDETA (0.0227 mL, 0.109 mmol) were poured into the round flask and stirred under an argon flow for 20 min; (1-bromoethyl)benzene (0.0148 mL, 0.109 mmol) was then added, and the reaction took place at 85 C for 5 hr. After the required reaction time, the viscosity of the mixture was increased. THF (10 mL) was added to the flask, and then the mixture was precipitated in 50 mL of cold methanol, filtered, and dried. SEC in THF: M n = 31,500 g/mol, ä = 1.26. ## Preparation of Polystyrene Films Containing a Coded Polyurethane Label Coded polystyrene films were prepared by dissolving polystyrene and the sequence-coded polyurethane P1 in THF (8% w/v) at 40 C where the weight ratio of polystyrene/PU was 99:1. The solution obtained was poured onto a glass plate, and the membranes were formed after 20 hr at RT. The films were further dried under a vacuum until complete removal of the solvent. Homogeneous and transparent membranes were obtained. ## Preparation of 3D-Printed Objects Containing a Coded Polyurethane Label Printing of the 3D object including a PU tag was performed using a Form 1+ SLA 3D printer equipped with a class 1 laser and a diode of violet color of 405 nm wavelength provided by Formlabs. The PU (0.25%, w/v) was dissolved in the photosensitive liquid resin. The tank was filled with the resin mixture, and with the help of the laser, it was cured layer by layer to build the 3D model. After the building of a layer, the laser was raised before continuing with the rest of the layers. The construction of 639 layers took 1 hr 40 min. The volume of the structure obtained was about 8 mL. The final 3D model was characterized from a detailed and high-resolution structure. It was generated on a support. After the end of the building, it was immersed in isopropyl alcohol to rinse the parts and clean liquid uncured resin from the outer side of the 3D model. The 3D object was further cured under UV light of 350 nm wavelength for a final photo-crosslinking curing to completely harden the sculpture. The support was removed using tools, and the surface was sanded to give a smooth finish. ## SEC Two different SEC setups were used in this work. The first was equipped with four PLGel Mixed C columns (5 mm, 30 cm, diameter = 7.5 mm), a Wyatt Viscostar-II viscometer, a Wyatt TREOS light scattering detector, a Shimadzu SPD-M20A diode array UV detector, and a Wyatt Optilab T-rEX refractometer. This setup was used for polymer characterization (1,000-3,000,000 g/mol). The other setup was equipped with a Shimadzu RiD-10A refractometer, a Shimadzu SPD-10Avp UV detector, and four monoporosity PLGel columns (5 mm, 30 cm, diameter = 7.5 mm): 50, 100, 500, and 1,000 A ˚. This setup was used for characterization of oligomers and short polymers (100-20,000 g/mol). In both setups, the mobile phase was THF with a flow rate of 1 mL/min. Toluene was used as the internal reference. The calibration was based on linear PS standards from Polymer Laboratories. ## Thermogravimetric Analysis Thermogravimetric analysis (TGA) was recorded on a Mettler Toledo TGA 2 Star System. The temperature range was from 25 C to 600 C, and the heating rate was 10 C/min in 100 mL/min N 2 . ## ESI-MS HRMS and MS/MS experiments were performed using a QqTOF mass spectrometer (QStar Elite; Applied Biosystems SCIEX) equipped with an ESI source operated in negative mode (capillary voltage, 4,200 V; cone voltage, 75 V). Some experiments were conducted in positive-ion mode with a capillary voltage of +5,500 V and a cone voltage of +75 V. In MS mode, the mass of the ions was measured accurately in an orthogonal acceleration time-of-flight (oa-TOF) mass analyzer using polyethylene glycol oligomers adducted with an acetate anion as internal standards. In MS/MS mode, precursor ions were selected in a quadrupole mass analyzer before entering a collision cell filled with nitrogen, and product ions were measured in the oa-TOF. In this instrument, air was used as the nebulizing gas (10 psi), and nitrogen was used as the curtain gas (20 psi). Instrument control, data acquisition, and data processing were achieved using Analyst software (QS 2.0) provided by Applied Biosystems. Oligomers (1-2 mg) were dissolved in methanol (300 mL) in an ultrasonic bath (15 min). Samples were further diluted (1:100 to 1:1,000 v/v) in a methanol solution of ammonium acetate (3 mM) and injected in the ESI source at 10 mL/min using a syringe pump. Oligomers were ionized in negative-ion-mode ESI, and the deprotonated molecules formed were subjected to collision-induced dissociation. ## General Rules for MS/MS Polymer Sequencing Fragmentation of deprotonated oligocarbamates only proceeds by competitive cleavage of all C-O carbamate bonds as a result of transfer of a proton from the alkyl segment on the right-hand side of the dissociating bond. Because the negative charge remains located in the a termination, only products containing this end group are detected as ions in MS/MS spectra. As a result, the distance between two consecutive peaks in MS/MS spectra is equal to the mass of a coded unit, as shown in Scheme S2 for P6 (MS/MS data shown in Figure S6). Consequently, measuring the m/z difference between consecutive peaks starting from the precursor ion allows the oligomer sequence to be readily reconstructed from the right to the left. Product ions were named after the nomenclature established by Wesdemiotis et al. 37 for synthetic polymers. This nomenclature recommends that a-containing fragments be designated with letters from the beginning of the alphabet (a, b, c, etc., corresponding to product ions obtained after cleavage of the first, second, third, etc., bond in the monomer, respectively). However, in order to adopt a similar nomenclature for all polycarbamates regardless of the alphabet (C3, C4, and C5) used to code information in their structure, this unique ion series was named a i (where i is the number of the entire monomeric unit contained in the product ion). ## MS and MS/MS Analysis of the Polyurethane-Tagged Polystyrene Films A small portion ($1.0 mg) of the PS film was cut and sonicated in methanol containing 3 mM ammonium acetate (200 mL) for 10 min. Because of the low solubility of carbamates in the extraction medium (chosen as a non-solvent for PS), the turbid solution obtained was further diluted (1:10 v/v) in the same solvent before it was injected in the ESI source for MS analysis and MS/MS sequencing. MS and MS/MS Analysis of the Polyurethane-Tagged 3D-Printed Object Small amounts (5-10 mg) of the thin powder obtained by grinding residues cut from the 3D-printed sculpture were sonicated in 400 mL of methanol containing 3 mM ammonium acetate for 15 min (similar results were obtained when using THF as the extraction solvent). The extraction medium was centrifuged (130 rpm), and the supernatant was further diluted (1:10 v/v) in methanol supplemented with ammonium acetate before being injected in the ESI source for MS analysis and MS/MS sequencing. ## SUPPLEMENTAL INFORMATION Supplemental Information includes 19 figures, two tables, and two schemes and can be found with this article online at http://dx.doi.org/10.1016/j.chempr.2016.06.006.

chemsum

{"title": "Chemoselective Synthesis of Uniform Sequence-Coded Polyurethanes and Their Use as Molecular Tags", "journal": "Chem Cell"}

oxygen_storage_in_transition_metal-doped_bixbyite_vanadium_sesquioxide_nanocrystals

## Abstract: Bixbyite vanadium sesquioxide (V2O3) is a metastable polymorph of vanadium oxide that has been shown to have a significant oxygen storage capacity with very low temperature oxidation onset.In this work, bixbyite V2O3 nanocrystals were synthesized with titanium and manganese dopants. Doped materials with varied dopant concentration were synthesized, and all were incorporated as aliovalent metal ions. The oxygen storage capacity of these nanocrystal materials was evaluated over ten oxidation and reduction cycles. It was found that over these ten cycles, the oxygen storage capacity of all the materials fell drastically. In situ X-ray diffraction evidence shows that manganese-doped materials degrade into an amorphous manganese-containing vanadate, while titanium-doped materials form crystalline degradation products. In all cases, this degradation causes an increase in the minimum mass achieved during oxygen release, indicating irreversible oxidation. ## Introduction The bixbyite phase of vanadium sesquioxide (V2O3), a metastable phase of the material, was first reported in 2011 by Weber et. al. 1 In 2015, a method to synthesize the pure bixbyite phase V2O3 as colloidal nanocrystals was described by Bergerud et al.. 2 Shortly after, it was discovered that this material exhibits excellent reversible oxygen storage capacity, with a remarkably low oxidation onset temperature, owing to its intrinsic oxygen vacancies. 3 This property has motivated investigation into ways to maximize the oxygen storage capacity and thermal and environmental stability of this material, to facilitate its potential application as an oxygen storage material (OSM) in automotive catalysts, where it could address problems with cold start. Automotive catalysis has been discussing the cold start problem for quite some time. Catalytic converters were introduced in the United States by General Motors in 1975, consisting of alumina support with small amounts of platinum and palladium catalyst material. 4 Their use was made nearly ubiquitous by emissions standards introduced by the US Environmental Protection Agency in that year, and practically required by updated standards as an amendment to this legislation in 1990. 5 Modern catalytic converters have added cerium oxide, or ceria, to the catalyst bed to promote high conversion and efficiency at low air-to-fuel ratios in the exhaust stream. 6 Ceria, while it has enjoyed decades of use as the OSM of choice in the automotive industry, is not without shortcomings. Below about 300 o C, the ceria surface is not reactive with oxygen. An inactive OSM is one reason that automotive catalysts are virtually inactive until the converter reaches an operation temperature of at least 300 o C. Thus high levels of emission are observed during the first few minutes of operation of an automotive engine, and this is referred to by the catalysis community as the "cold start" problem. 7,8 By contrast, V2O3 nanocrystals have been shown to absorb oxygen from an atmosphere at temperatures as low as 100 o C and have a very similar oxygen storage capacity (OSC) to this industry standard material. It struggles, however, with poor stability upon cycling, owing to the wide variety of structures available to the large number of vanadium oxides that can be formed, reflected in the complexity of its phase diagram. 9,10 Doping is a commonly employed strategy in materials design for applications where stability is desired. In fact, OSMs for automotive catalysts have long used dopants to optimize performance. 11,12 Since at least the 1990s, the ceria OSMs in most auto catalytic converters have included these deliberate impurities, which increase the material's useful lifetime and its OSC. Through experimental and computational methods, it has been shown that zirconium impurities, as Zr 4+ , introduce strain to the ceria lattice that prevents sintering, which causes a decrease in capacity over the material's lifetime, while increasing OSC by causing partial reduction of cerium ions. 13,14 Studies on divalent dopants in ceria show that the interplay between charge imbalance and lattice strain has strong impacts on oxygen storage capacity. 15,16,17 In this work, doped V2O3 nanocrystals are synthesized to investigate the impact of dopants on the OSC and cycling stability of this novel OSM. Titanium and manganese dopants were chosen, as transition metals on either side of vanadium on the periodic table. Incidentally, titanium dopants have been studied before in vanadium oxides, where titanium was shown to stabilize the corundum paramagnetic phase. 18 It has also been shown to increase the monoclinic-to-rutile transition temperature in VO2. 19,20 Likewise, manganese is one of many dopants that have been studied as incorporated into the layered material V2O5 for use in battery and energy storage applications. ## Results and Discussion V2O3 particles were prepared with varied concentrations of titanium and manganese dopants. Changes in the oxidation state of the metal ions caused by doping were evaluated by X-ray photoemission spectroscopy (XPS). The OSC for each sample was measured by thermogravimetric analysis (TGA). The crystallinity and crystal phase for each material was evaluated by ex-situ X-ray diffraction (XRD) before and after processing in TGA. Further, the crystal structure of the materials during oxidation and reduction were followed by in situ XRD, which revealed different degradation mechanisms for materials with different compositions. Titanium doping resulted in particles of size and shape very similar to undoped particles. Both types of nanocrystals have flower morphology and diameter of 20-40 nm, accompanied by some smaller side products, possibly the result of secondary nucleation during synthesis. Molybdenum doping resulted in highly faceted particles with a wide distribution of sizes, ranging from 3 to 20 nm across the longest dimension. Very small, round particles observed here may be from secondary nucleation as well, which can create small crystals that do not have the opportunity to grow. The larger particles are bipyramidal in shape, with rhombohedral projections between 5 and 20 nm across the dark middle, and between 10 and 25 nm across from tip to tip. XPS of the doped vanadium oxides (Figure 2) show the 2p transitions for vanadium in all the samples, along with the 2p transitions for the dopant elements. Survey spectra are shown in Figure S2. Ti-and Mn-doped samples all show a slight shift of the V 2p transitions to higher binding energy. The oxidation state of the dopant atoms can be determined by analysis of their 2p transitions. The Mn 2p spectrum shows a shake-up feature at 647 eV, diagnostic of the Mn 2+ state. 28 To further support this finding, the Mn 3s spectrum was collected (high resolution spectra and fits shown in Figure S4). The 3s spectrum has two peaks, due to correlation with the unfilled 3d 5 shell. The binding energy difference between the 3s 7 S and 5 S multiplet component peaks was found to be 6.5 eV for both the 5% and 10% doped samples, consistent with previous literature observing Mn in the 2+ oxidation state. 29 The titanium 2p peaks have a spin-orbital splitting of 5.7 eV each, which is consistent with reports for TiO2, where Ti is in the 4+ oxidation state. 30 Both of these ions are larger than the native V 3+ ion, at 0.74 and 0.83 nm for Ti 4+ and Mn 2+ with 6-fold coordination, respectively, compared to 0.64 nm for V 3+ , so they are expected to have similar effect on lattice strain. 31 While both dopants are larger than the host ion, they introduce charge imbalance in opposite directions. All the doped samples had higher OSC than undoped nanocrystals for their first six cycles of oxidation and reduction. However, for all samples, the OSC dropped dramatically during ten cycles. The Ti-doped samples both maintained higher OSC during all ten cycles, ending with an OSC 20-30% higher than that of undoped V2O3. Vanadium oxide doped by Mn had lower OSC than undoped nanocrystals after 10 cycles. The OSC is determined by the difference between the maximum weight and the minimum weight during each oxidation and reduction cycle (Figure 4 ab), and tracking those maximum and minimum values provides more information about the decrease in OSC with cycle number (Figure 4 c-f). It is immediately evident that the minimum mass fraction for all samples rose sharply in the first four cycles, indicating irreversible oxidation and suggesting a kinetic barrier to oxygen release during the reduction half-cycle. There is a correlated drop in maximum mass fraction after oxidation for Mn-doped nanocrystals. However, for the Ti-doped samples, the maximum and minimum fractional masses both drop nearly linearly with cycle number after cycle 5, suggesting that the Ti-doped samples are degraded by some separate or additional mechanism. The as-synthesized nanocrystals and the products after the ten oxidation and reduction treatment as described above were analyzed by powder X-ray diffraction. The results are shown in Figure 5. Reference XRD patterns are shown for bixbyite V2O3 in black, the thermodynamically preferred corundum phase of V2O3 in red, rutile VO2 in green, and rutile TiO2 in turquoise, all at the bottom. The samples show good crystallinity before TGA cycling and exist in the bixbyite crystal structure, apart from highly Mn-doped V2O3, which shows only very broad diffraction signal. Two peaks in the pattern for 10% Ti:V2O3, at 2θ of 28.6 o and 47.5 o (marked with a star in Figure 5), were not indexed to a known phase of titanium or vanadium oxide. For samples doped up to 5%, only bixbyite V2O3 peaks are observed. After ten cycles of oxidation and reduction, the crystallinity in both the Mn-doped samples has all but disappeared, indicating degradation to form some noncrystalline product. The undoped and Ti-doped samples lose intensity from the original bixbyite phase and develop new peaks due to crystalline impurities. The major impurities are from the thermodynamically preferred corundum structure of V2O3 and rutile VO2. While diffraction peaks of rutile TiO2 could not be differentiated from those of VO2, the amount of titanium incorporated into the material is quite small, so we expect that these peaks appear due to the formation of rutile VO2. To investigate the different mechanisms of oxidation and reduction for doped V2O3, in situ Xray diffraction was conducted. In Figure 6, the oxidation half-cycle is shown in reciprocal space for q between 3 and 4.5 A -1 . The strongest peak around 3.7 A -1 is the bixbyite (440) reflection, which appears for the ex situ diffraction at 2θ of about 55 o . At time 0, the sample is exposed to synthetic air at 150 o C. Shortly, the bixbyite reflections are seen to move slowly to lower q, indicating lattice expansion that is nearly isotropic. Peaks shown that do not shift are reflections from the sample holder stage. Some weakening in the intensity of those peaks is observed, which may indicate very slight movement of the sample due to thermal expansion of the sample holder during oxidation. We also observe that reflection peaks for the Mn-doped samples are much broader, consistent with ex situ observations showing Mn-doped samples have poorer crystallinity than their Ti-doped counterparts. Notably, the Ti-doped sample in panel b shows a splitting of the (440) peak after about 25 minutes of oxidation. This splitting is attributed to the phase transformation of bixbyite V2O3 to form rutile VO2, as discussed above. By contrast, the peak shifts to lower q for the Mn-doped case are not accompanied by formation of any new crystalline impurity phases. Rather, this transition is partially reversible upon heating to 300 o C under nitrogen. Unabridged in situ diffraction for the oxidation and reduction of the Mn:V2O3 nanocrystals is shown in Figure S6, and for the full q range collected of oxidation of the Ti:V2O3 nanocrystals in Figure S7. All the V2O3 samples show a lattice expansion upon oxidation, as evidenced by bixbyite peaks moving to lower q during oxidation. Scherrer analysis of the ( 222), (440), and (622) reflections was conducted before, during, and after oxidation to assess the extent of lattice expansion. In the fully oxidized state, Ti-doped nanocrystals showed a 1.3% expansion, while Mndoped nanocrystals showed 1.6% expansion, compared to 0.6% expansion for undoped V2O3 seen in a previous study. 3 The main difference in these two doped materials appears to be their degradation mechanism during repeated cycling. These observations support the conclusion that doped V2O3 nanocrystals follow two different degradation mechanisms depending on the dopant. For Mn-doped materials, oxidation causes disordering of the crystalline structure to form amorphous vanadates and manganates, which have much less storage capacity than the original vanadium oxide lattice. In contrast, Ti-doped materials experience an irreversible phase transition to monoclinic VO2 during oxidation that causes loss in OSC. This difference may be due to the Ti 4+ dopant's positive relative charge, which may stabilize oxygen interstitials, facilitating a transformation to the higher oxidation state vanadium oxide. ## Conclusions Vanadium sesquioxide nanocrystals were synthesized with titanium and manganese dopants. These are incorporated as aliovalent dopants Ti 4+ and Mn 2+ . Doped materials were compared to the undoped by thermogravimetric analysis. Cycling these materials between oxidation and reduction ten times revealed that cycling these oxygen storage materials results in degradation of the oxygen storage capacity. This degradation is driven primarily by the irreversible oxidation of the materials during the low temperature oxidation step. Furthermore, we have found that Mndoped and Ti-doped V2O3 nanocrystals degrade by different mechanisms. For undoped and Mndoped materials, the nanocrystals tend to degrade by amorphizing the initially crystalline bixbyite phase material to form an impurity phase of inactive amorphous material. For Ti-doped V2O3, there is an additional degradation method in which the nanocrystals undergo irreversible phase transition to form rutile VO2 during oxidation, perhaps promoted by Ti 4+ , which may promote filling of oxygen vacancies because of its charge imbalance and the fact that Ti itself forms a rutile oxide. This degradation progresses over all ten cycles, resulting in an OSC that declines approximately linearly with cycle number. Aliovalent dopants may drive degradation products to a structure that the oxide of the dopant shares with a vanadate or to amorphous materials in case no favorable crystalline polymorph exists between the dopant and vanadium oxide. Future work might therefore seek a dopant that shares a structure with a vanadate which is a stable oxygen deficient structure. Cerium and indium may be prime candidates for this, since cerium oxide is a known oxygen ion conductor and indium shares the bixbyite structure in common with vanadium sesquioxide. ## Material Synthesis Doped bixbyite-phase vanadium sesquioxide materials were synthesized by modifying the reported vanadium sesquioxide nanocrystal synthesis, replacing some amount of vanadyl acetylacetonate with an equimolar amount of a metal dopant precursor. 2 For the titanium-doped samples, the metal dopant precursor was Ti (IV) oxyacetylacetonate. For the manganese-doped samples, the precursor was Mn (II) acetylacetonate. Nominal dopant percentages indicate the percent of vanadyl acetylacetonate that was replaced with the dopant metal precursor. Synthesis was done with 1 mmol of metal precursor, 4 mmol oleic acid, 4 mmol oleylamine, and 8 ml squalene. Chemicals were loaded into a 50 ml round-bottom 3-neck flask and degassed under dynamic vacuum at 110 o C for an hour before switching to a nitrogen atmosphere and increasing the temperature to the synthesis temperature of 370 o C. This temperature was held for 1 hour before cooling to room temperature. Nanocrystals were washed several times with isopropanol and hexane before use or characterization. Solutions were washed and stored air-free, powder or dry samples stored in nitrogen or in a vacuum dessicator. ## Thermogravimetric Analysis TGA was collected with a Mettler Toledo TGA 2. Samples were dropcast from solution into 100 ul aluminum crucibles and allowed to dry. Multiple depositions were used until the total weight of sample in the crucible was between 5 and 10 mg. ## Inductively Couples Plasma Atomic Emission Spectroscopy Doped and undoped samples were digested with 70% nitric acid and diluted to contain 2% nitric acid. ICP AES was collected using a Varian 720-ES ICP AES. ## Transmission Electron Microscopy TEM images were obtained using a JEOL 2010F electron microscope equipped with a Schottky field emission gun and a CCD camera, operated at 200 kV. ## X-ray Photoemission Spectroscopy As-synthesized colloidal particles were dropcast onto p-type doped silicon substrates, and X-ray photoemission was collected with a Kratos Axis Ultra DLD spectrometer with a monochromatic Al Kα source (1486.6 eV). Resulting spectra were analyzed with CasaXPS equipped with the Kratos library of elemental relative sensitivity factors. Spectra were treated with a simple constant correction to calibrate the main carbon 1s peak to an energy of 284.8 eV. ## X-ray Diffraction In situ X-ray diffraction of doped vanadium sesquioxide materials was performed at the Stanford Synchrotron Radiation Lightsource in Stanford, California. Using the Anton Paar in situ heating cell, powder samples of vanadium sesquioxide materials deposited on silicon substrates were irradiated with an X-ray beam with an energy of 14 eV. The diffraction image was collected with a Pilatus 300k detector in landscape orientation and analyzed using the Nika and Irena X-ray analysis package with Igor. 32,33,34,35 The sample was reduced in situ under constantly flowing nitrogen gas at 300 o C for one hour before the experiment began, to remove any adventitious carbon or water on the sample's surface. The sample was then cooled to 50 o C in nitrogen, at which point the flowing gas was changed to a synthetic air mixture. In the air atmosphere, the sample was heated slowly from 50 o C to 150 o C and held at 150 o C for one hour. Then the gas was switched to nitrogen and the sample was heated to 300 o C and held at 300 o C for one hour. Ex situ X-ray diffraction measurements were obtained using a Rigaku R-Axis Spider diffractometer with a Cu sealed tube source and a large, image plate detector.

chemsum

{"title": "Oxygen storage in transition metal-doped bixbyite vanadium sesquioxide nanocrystals", "journal": "ChemRxiv"}

carrier_dynamics_engineering_for_high-performance_electron-transport-layer-free_perovskite_photovolt

## Abstract: The performance of electron-transport-layer-free (ETL-free) perovskite photovoltaics is far more sensitive to the carrier lifetime of perovskite films than analogous ETL-containing devices. A long carrier lifetime can counterbalance the inferior interface in the absence of a distinct ETL, enabling efficient carrier collection in ETL-free perovskite PV. By using perovskite films with microsecond carrier lifetime, Hu and co-workers achieved ETL-free solar cells with >19.5% power conversion efficiency, which is comparable to the analogous value of ETLcontaining devices. ## INTRODUCTION Perovskite photovoltaics (PV) has attracted tremendous attention because of recent advances in power conversion efficiency (PCE) and diverse processing options. Other than high performance, factors including good stability, simpler device configuration, and low processing cost should be considered in the next-generation perovskite PV. Currently, high-performance perovskite solar cells exclusively employ p-i-n device architectures, wherein distinct electron transport (n-type) and hole transport (p-type) layers are generally viewed as critical components for reliable photogenerated carrier extraction. 9 However, the deposition processes to construct a p-i-n architecture, especially for inorganic electron-transport layers (ETLs), typically require high-temperature conditions (e.g., $500 C for TiO 2 , $500 C for BaSnO 3 , and $250 C for ZnO). 7, Such high-temperature steps compromise the low-cost advantage of solution-based approaches for perovskite film deposition (e.g., generally performed at 25 C-100 C), leading, for example, to increased projected payback time for perovskite PV and other optoelectronics. 13 Additionally, in building more complex optoelectronic devices, such as all-perovskite tandem solar cells, the high temperatures to process ETLs in top cells can damage the perovskite and/or other temperature-sensitive films in bottom cells. The high temperatures can also melt prospective flexible substrates and/or cause The Bigger Picture Interface properties essentially determine the performance of perovskite photovoltaics (PV). Typical electron-transport-layerfree (ETL-free) perovskite PV suffers from significant loss of performance as a result of inferior carrier dynamics at the interface. Here, we determine that the low efficiency of ETL-free perovskite PV is attributed to insufficient photoexcited carrier collection, which originates from the inhibited carrier injection at the interface without a distinct ETL. We found that improving the carrier lifetimes of perovskite films can effectively counterbalance the low-injection-rate interface (e.g., ETL-free interface), enabling efficient carrier collection in ETLfree perovskite PV. By using perovskite films with microsecond carrier lifetimes, we achieved ETLfree PV with >19% efficiency, which makes ETL-free devices comparable to p-i-n-structured perovskite devices. These results indicate a general approach to improving the performance of PV devices with inferior interfaces. ion migration from substrates, limiting the applications of perovskite technology in versatile optoelectronics. More importantly, the existence of some ETLs themselves can detrimentally affect the perovskite devices. For example, ZnO ETLs can cause decomposition of perovskite layers during annealing 15 and TiO 2 ETLs can induce degradation of perovskite films under illumination. 16 Regardless of these performance issues, the inclusion of numerous layers in devices is not desired for effective commercialization. In short, these ETL-induced issues lead to prospects of lower processing yields, increased payback times, faster degradation, and difficulties in constructing more versatile and complex perovskite optoelectronics. To address these issues, building solar cells without distinct ETLs in the architectures is a promising direction for next-generation perovskite PV. In fact, studies have demonstrated that perovskite films possess superior properties for charge-carrier dissociation 18 and transport, 19,20 and that the transparent electrodes (e.g., fluorine-doped tin oxide glass [FTO] and tin-doped indium oxide glass [ITO]) themselves are typically n-type semiconductors. Therefore, distinct ETLs should not theoretically be necessary for high-performance perovskite PV. Unfortunately, all current-generation ETL-free perovskite solar cells suffer from low PCE, with relatively large hysteresis and inferior stability. In this regard, creating deeper understanding and developing effective approaches to improving the performance are keys for the success of ETL-free perovskite PV. In this work, we reveal that the carrier injection process is significantly inhibited at the interface in the absence of an ETL, which leads to insufficient carrier collection and severe interfacial carrier recombination. The recombination decreases the external quantum efficiency (EQE) in both short-and long-wavelength ranges and thereby compromises the performance for ETL-free perovskite PV. Moreover, we demonstrate that improving the intrinsic carrier lifetimes in perovskite films can counterbalance the inferior device interfaces and carrier recombination at the ETL-free interface. Through such carrier dynamics engineering, the carrier collection efficiency in the ETL-free perovskite PV can be remarkably tailored to approach that in ETL-containing devices. Benefiting from this discovery, we exploit perovskite films with microsecond carrier lifetimes to successfully realize ETL-free perovskite PV with a best PCE of 19.52% (18.48% on average), nearly eliminated hysteresis, and good stability. Such a high PCE is comparable to the value (20.7%) achieved for ETL-containing solar cells prepared with analogous perovskite films. Our research provides insights into ETL-free solar cells and points to a promising direction for perovskite PV and analogous optoelectronic devices, offering simultaneously high performance, simplified processing, and better prospects for ultra-low-cost device fabrication. ## ETL-free and ETL-Containing PV To identify the difference between typical ETL-free and ETL-containing PV, we used pristine CH 3 NH 3 PbI 3 (MAPbI 3 ) perovskite films as absorbers. We deposited the pristine MAPbI 3 films by a one-step method employing a precursor solution containing a 1:1 molar ratio of MAI and PbI 2 . ITO-coated glass was used as the transparent electrode substrate. In the ETL-containing PV devices, SnO 2 was chosen as the distinct ETL because of the wider bandgap of SnO 2 (e.g., compared with TiO 2 ), such that the distinct ETL would minimally affect the light absorption within the perovskite films during device operation (as discussed later). 30 The full details of film deposition are described in the Experimental Procedures. Top-view scanning electron microscopy (SEM) images in Figures 1A and 1B present the morphologies of typical pristine MAPbI 3 films on glass/ITO and glass/ITO/SnO 2 substrates, respectively, from which one can deduce that the MAPbI 3 films share similar film morphologies and compactness on both substrates. From X-ray diffraction (XRD) patterns (Figure S1), the MAPbI 3 films exhibit similar structural properties on both substrates. The optical absorption spectra (Figure S2) illustrate that the MAPbI 3 films also have similar optical features (E g = $1.60 eV). These results indicate that glass/ITO and glass/ITO/SnO 2 substrates do not affect the basic structural/optical properties of MAPbI 3 films. In many respects this observation is not surprising, given that both SnO 2 and ITO present a tin-oxide-based surface. To examine the PV performance for such MAPbI 3 films in devices, we fabricated ETL-free and ETL-containing perovskite PV devices with the device architectures of glass/ITO/MAPbI 3 /2,2 0 ,7,7 0 -tetrakis[N,N-di(4-methoxyphenyl)amino]-9,9 0 -spirobifluorene (Spiro-OMeTAD)/Au and glass/ITO/SnO 2 /MAPbI 3 /Spiro-OMeTAD/Au, respectively. Figure 1C presents the statistical distributions of the PCE values, from which the ETL-free solar cells yield PCE from 7.70% to 11.40% (8.68% on average) for 24 devices, while ETL-containing solar cells yield PCE from 15.78% to 17.99% (16.97% on average) for 24 devices. The statistical results indicate that the PCE of a typical ETL-free PV device is significantly lower than that of a typical ETLcontaining device when using the pristine MAPbI 3 films, which is consistent with the earlier reports. 21,22 To ascertain the reasons for the compromised PCE, we compared the EQE for ETL-free and ETL-containing perovskite PV. From the EQE spectra (Figure 1D), the EQE response of the ETL-free device falls significantly lower than that for the analogous ETL-containing device, especially in the short (300-400 nm) and long (500-750 nm) wavelength ranges. Such results indicate that the carrier collection is less effective in ETL-free devices (e.g., owing to enhanced recombination in the ETL-free device structures). Since the glass/ITO and glass/ITO/SnO 2 substrates have nearly the same optical absorption features (Figure S3) and the perovskite films have similar grain structures, the reduced carrier collection in ETL-free devices likely derives from the different interfaces. Previous study attributed this device performance loss to the lack of effective hole blocking. 22 From this opinion, because of the absence of a distinct ETL, photogenerated holes can easily reach the ITO/MAPbI 3 interface and ITO contains a high density of electrons that would facilitate the recombination of photogenerated holes at this interface, eventually leading to the loss of photogenerated carriers and device performance. However, one study showed that the work function of ITO (4.7 eV) is higher than the Fermi level of MAPbI 3 . 31 Therefore, The band-bending direction of MAPbI 3 films at the ITO/MAPbI 3 interface is downward and the built-in electric field formed at the interface points from ITO to MAPbI 3 film, which attracts electrons and repels holes photogenerated in MAPbI 3 films (shown in Figure S4). In addition, ITO is a typical n-type semiconductor whose valence band maximum edge is lower than that of perovskite materials. Such situations make the as-formed ITO/MAPbI 3 interface preferable for electron extracting and hole blocking. Based on this analysis, the performance loss in ETL-free perovskite PV may not be due to the lack of hole blocking. To understand the impact of the ITO/MAPbI 3 interface on photogenerated carriers, we performed photoluminescence (PL) quenching measurements. Figure 1E presents the steady-state PL spectra for the MAPbI 3 films on quartz, glass/ITO, and glass/ITO/SnO 2 , respectively, from which we can observe that the PL signal quenches less on the ETL-free substrates. The quenching results evidently illustrate that the interfacial carrier injection process is substantially inhibited without the assistance of a distinct ETL, which suggests that the compromised performance of ETL-free perovskite PV originates from the inhibited carrier injection on the ITO substrate. One study indicated that the work function/Fermi level of SnO 2 is around 4.36 eV, which is higher than the work function of ITO. 30 Without the SnO 2 ETL in the devices, the interfacial built-in electric field may become weaker because of the relatively lower work function of ITO. The inhibited injection could be derived from this weaker built-in electric field. At or near the interface, if the carriers photogenerated in perovskite films cannot be rapidly collected, these carriers will undergo recombination (e.g., radiative and/or nonradiative recombination) in the perovskite films. From the time-resolved photoluminescence (TRPL) result (Figure 1F), the pristine MAPbI 3 films used in typical ETL-free PV exhibit short average carrier lifetimes (t avg = 1.17 G 0.02 ns). We refer to such lifetimes as ''intrinsic'' lifetimes since they are measured for films on quartz substrates rather than for films within device structures, wherein built-in electric fields will affect the values measured. The photogenerated carriers with such short intrinsic lifetimes will more readily recombine at or near the interface if the built-in electric field cannot cause them to drift to the interface and rapidly across the interface to the contact. For reducing the carrier recombination near the low-injection-rate interface, the typical approach involves adding an additional layer at this poor interface to provide a strong built-in electric field to quickly drift the carriers across the interface for extraction (ultimately, before recombination). However, this approach adds extra layers to the ETL-free architectures and creates an architecture that is essentially identical to that for the ETL-containing devices. Logically, if we are not able to improve the interfacial band alignment to enhance carrier drifting, the next best thing would be to improve the carrier lifetimes within the perovskite films so that the photogenerated carriers can survive long enough to diffuse to the interface and make it across the less effective interface. Based on this analysis and expectation, increasing the lifetimes for the carrier diffusion could be an effective method to compensate the weaker ''drifting ability'' (e.g., built-in electric field), which is highly desired for ETL-free PV (even more so than in ETL-containing PV). In the following section, we explore the effect of intrinsic carrier lifetimes in perovskite films on the performance of ETL-free PV devices. ## Impact of Carrier Lifetimes on PV Performance The effect of the intrinsic carrier lifetimes on the performance of ETL-free perovskite PV was explored by introducing extra PbI 2 into the MAPbI 3 films to tune the film carrier lifetimes. 10,13 Top-view SEM images (Figure S5) show that the MAPbI 3 films with 2.5%, 5.0%, and 7.5% extra PbI 2 on glass/ITO substrates have grain size similar to that in the pristine MAPbI 3 film as shown in Figure 1A. From the atomic force microscopy (AFM) images in Figure S6, the MAPbI 3 films share similar surface roughness, within the range of 6-8 nm, illustrating that adding extra PbI 2 to the MAPbI 3 films does not significantly affect film morphology. The optical absorption spectra (Figure S7) indicate that adding 0%-7.5% extra PbI 2 in MAPbI 3 films does not remarkably affect the absorption characteristics of the films. We performed TRPL measurements for the MAPbI 3 films on quartz substrates to evaluate the optoelectronic properties of the MAPbI 3 films with extra PbI 2 . By comparing the TRPL kinetic traces in Figures 2A and 1F, it can be observed that adding extra PbI 2 significantly increases the intrinsic carrier lifetimes of the MAPbI 3 films, from several nanoseconds to several hundred nanoseconds. To further pursue this point, we also performed steady-state PL measurements on the MAPbI 3 films (Figure 2B), showing significant intensity enhancement with extra PbI 2 addition, which is consistent with the TRPL results. In addition to the intrinsic carrier lifetimes, we also measured the steady-state PL for the MAPbI 3 films with extra PbI 2 on glass/ITO substrates to examine whether extra PbI 2 in MAPbI 3 films may change the carrier injection rate at the ITO/MAPbI 3 interface (as shown in Figure S8). Since the PL intensity of the MAPbI 3 films significantly varies with different PbI 2 levels, we define a PL quenching rate metric (QR PL )-i.e., QR PL = (I 1 I 2 )/I 1 , where I 1 is the PL intensity of the MAPbI 3 film on quartz while I 2 is the analogous value on ITO-to better evaluate the carrier injection at the ITO/perovskite interface. By comparing the QR PL values (Figure S9), it is seen that the extra PbI 2 itself in MAPbI 3 films does not enhance carrier injection yield at the ITO/perovskite interface, as underscored by the extent of PL quenching. This result is reasonable since the relative conduction band position of PbI 2 is higher than that of perovskite materials so that prospective extra PbI 2 accumulation at the ITO/MAPbI 3 interface would not improve the band alignment for electron injection. 32 Taken together, the TRPL and PL results indicate that addition of extra PbI 2 significantly improves the carrier lifetimes in the MAPbI 3 films but does not facilitate the electron-injection rate at the ITO/MAPbI 3 interface. 10,13 To show the correlation between the intrinsic carrier lifetimes and performance for ETL-free perovskite PV, we fabricated solar cells by using extra-PbI 2 -added MAPbI 3 films with the same ETL-free architecture (glass/ITO/MAPbI 3 /Spiro-OMeTAD/Au) described above. For each PbI 2 addition level, 24 devices were used for acquiring the statistical results shown in Figures 2C and S10. It can be observed that the trend of all the PV parameters is similar to the trend of carrier lifetimes as the PbI 2 addition level changes, illustrating close correlation between the device performances and the intrinsic carrier lifetimes, and that long carrier lifetimes substantially benefit the ETL-free device performance. For comparison, ETL-containing PV devices were also fabricated. From the PCE distributions shown in Figure S11, the intrinsic carrier lifetimes are also seen to affect the performance of ETL-containing PV. To explore the degree to which the carrier lifetimes can impact on the ETL-free and ETL-containing perovskite solar cells, respectively, we calculated the correlations between extra PbI 2 addition levels and average PCE values. From Figures 2D and S12, it is found that the performance of ETL-free devices is substantially more sensitive to the intrinsic carrier lifetimes than ETL-containing PV. We also measured the EQE for the ETL-containing and ETL-free PV devices (Figures 2E and 2F) to better understand the impact of intrinsic carrier lifetimes on photogenerated carrier collection. We observed that as carrier lifetimes increased, both the short-and long-wavelength spectral responses substantially improved for ETL-free PV, whereas only the long-wavelength range increased for ETL-containing PV. For ETL-containing PV (Figure 2E), the EQE increase in the long-wavelength range can be attributed to the reduction of bulk-defect-induced recombination, enabling more long-wavelength-excited carriers to travel through the perovskite films for extraction. For ETL-free devices (Figure 2F), the additional enhancement of short-wavelength EQE could be due to reduced carrier recombination at the ETL-free interface, leading to increasing numbers of short-wavelength-excited carriers being effectively collected at the ITO/perovskite interface. The EQE spectra explain the result that the performance of ETL-free PV is more sensitive to carrier lifetimes. Moreover, comparing the EQE spectra shown in Fig- ure 2G, we see that the EQE values for the ETL-free PV progressively converge to that for ETL-containing PV for long carrier lifetimes, totally different from the situation for short carrier lifetimes (Figure 1D). Long carrier lifetimes in perovskite films therefore appear to be important for obtaining high-performance ETL-free PV. In addition, these results also suggest that use of distinct ETLs is beneficial for the device performance if the intrinsic carrier lifetimes in the associated perovskite absorbers are relatively short. ## Perovskite Film with Microsecond Carrier Lifetimes For further boosting the PCE of ETL-free PV, mixed-cation lead mixed-halide perovskite films (Cs 0.05 FA 0.8 MA 0.15 PbI 2.55 Br 0.45 , referred to as CsFAMA, where FA = formamidinium) with microsecond carrier lifetimes were used as the light absorbers in devices (see the Experimental Procedures for deposition details). 33 The top-view SEM image of a CsFAMA film on glass/ITO substrate (Figure 3A) reveals that the grain size is $500 nm, with good compactness and coverage over a large area (Figure S13). According to the AFM image (Figure S14), the CsFAMA films exhibit flat surfaces with roughness on the order of 20 nm. The relatively flat surface reduces the contact area at the CsFAMA/Spiro-OMeTAD interface, leading to reduced interfacial recombination and therefore benefiting the device performance. The XRD pattern (Figure 3B) for a CsFAMA film on glass/ITO substrate demonstrates that the CsFAMA film contains only the photoactive perovskite a phase (black phase), and no nonperovskite d phase (yellow phase) exists to negatively affect the optoelectronic properties. To evaluate the optical properties of the CsFAMA film, we performed optical absorption and steady-state PL measurements (Figure 3C), which demonstrate that the optical absorption onset and PL peak are consistent (E g = $1.59 eV shown in Figure S15). To understand the carrier dynamics, we acquired TRPL data for a CsFAMA film on quartz substrate. The intrinsic carrier lifetimes extracted from the PL dynamics (Figure 3D) are t 1 = 1,231.9 G 27.7 ns and t 2 = 301.9 G 28.3 ns; these values have respective amplitudes A 1 = 71.4% and A 2 = 28.6% (t avg is 966.4 G 27.9 ns). Such microsecond carrier lifetimes, approximately three times the value achieved for the previously shown MAPbI 3 film with 5% extra PbI 2 , are expected to substantially boost the performance of ETL-free perovskite PV, as discussed above. ## High-Performance ETL-free Devices Given the above results, ETL-free perovskite solar cells were fabricated with the device architecture ITO/CsFAMA/Spiro-OMeTAD/Au. From the SEM image of the device cross-section (Figure 4A), the thicknesses of the CsFAMA, Spiro-OMeTAD, and Au layers are seen to be $650, $150, and $80 nm, respectively. The statistical distributions of PCE values (reverse scanning direction) for 24 ETL-free devices (Figure 4B) vary from 17.85% to 19.52% (18.48% on average). For comparison, analogous ETL-containing perovskite solar cells with an ITO/ SnO 2 /CsFAMA/Spiro-OMeTAD/Au structure yielded similar performance levels (Figure 4B), such that PCE values vary from 19.60% to 20.72% (20.03% on average). To evaluate the hysteresis behavior, we measured the best-performing devices by using both forward and reverse voltage-scanning directions. From the reverse (forward) scan, the best-performing ETL-free device (Figure 4C) yielded a PCE of 19.52% (18.84%), open-circuit voltage (V oc ) of 1.061 (1.069) V, shortcircuit current density (J sc ) of 23.61 (23.39) mA cm 2 , and fill factor (FF) of 77.79% (75.37%). The best-performing ETL-containing device (Figure S16) yielded a PCE of 20.72% (20.51%), V oc of 1.100 (1.100) V, J sc of 23.75 (23.51) mA cm 2 , and FF of 79.28% (79.23%). We can see that the ETL-free device hysteresis approaches that of the ETL-containing device. The EQE spectrum of the best-performing ETL-free device (Figure 4D) illustrates high quantum efficiency for energies above the band gap, leading to an integrated J sc of 23.39 mA cm 2 , which is consistent with the results from the current densityvoltage (J-V) characteristics. The steady-state output profile (Figure 4E) shows that the best-performing ETL-free device has a steady-state output current density (J) of $21.37 mA cm 2 under 0.89 V applied bias, corresponding to a stabilized PCE value of $19.02%-i.e., showing good agreement with the J-V measurement. Such a performance level for the ETL-free devices approaches record metrics for any type of perovskite devices and represents the best PCE performance level among all currently reported ETL-free perovskite solar cells to date. The correlations between intrinsic carrier lifetimes and device performance were examined for ETL-free/ETL-containing PV (Figure 4F). The results indicate that the PCE discrepancy between ETL-free and ETL-containing perovskite solar cells is significantly reduced as the carrier lifetime increases, and further suggest that extending the carrier lifetimes of perovskite films (e.g., >>1 ms) may boost the PCE of associated ETL-free perovskite solar cells to the same level as ETL-containing perovskite solar cells. Moreover, the PCE of the best-performing ETL-free device remains $19.1% (from J-V measurement) after 1,000 hr of storage (temperature of $25 C and relative humidity of $25% in the dark), illustrating that the ETL-free devices have good environmental stability (Figure S17). The photostability of the best-performing ETL-free perovskite solar cells was also measured using continuous light soaking (one sun) under ambient conditions (temperature of $25 C and relative humidity of $25%) without encapsulation (Figure S18). Such stability tests suggest that ETL-free PV with CsFAMA perovskite films having good stability can be made. ## Conclusion In conclusion, we use EQE to reveal that typical ETL-free perovskite solar cells with relatively low carrier lifetimes in perovskite films exhibit more substantial photogenerated carrier loss compared with ETL-containing devices. PL quenching experiments show that the injection of carriers from the perovskite to the transparent conducting oxide contact (ITO) is less effective for ETL-free devices. To address this interface issue without changing the ETL-free device design, we tailor the carrier lifetimes of the perovskite films and demonstrate that improved carrier lifetimes can enhance the carrier collection efficiency at the low-injection-rate interface, making the carrier dynamics in ETL-free devices essentially as good as those in ETL-containing devices. On the basis of such an understanding, the use of perovskite films with microsecond carrier lifetimes enables ETL-free perovskite solar cells to realize a best PCE of 19.52% with nearly eliminated hysteresis and good stability. Such high-performance ETL-free perovskite solar cells are comparable to the analogous ETLcontaining devices (PCE: 20.72%). These results offer opportunities for versatile perovskite PV with simple processing, low cost, and high performance. Our results also provide a general approach to improving the performance of PV with low-injection-rate interfaces not only limited to the perovskite PV family. ## EXPERIMENTAL PROCEDURES Perovskite Film Deposition For deposition of the pristine MAPbI 3 films, precursor solutions were prepared with 1.2 M PbI 2 and 1.2 M MAI in the DMF/DMSO co-solvent (V DMF /V DMSO = 9:1). Extra PbI 2 with levels of 0%, 2.5%, 5%, and 7.5% (mol %, relative to stoichiometric MAPbI 3 ) was added into the MAPbI 3 precursor solutions, respectively, to tune the carrier lifetimes in the MAPbI 3 films. For each system, the precursor solutions were stirred at $25 C for 24 hr and filtered with a 0.45-mm PTFE syringe filter before further use. The MAPbI 3 films were then deposited by spin-coating the precursor solution on substrates at 5,000 rpm for 30 s. Chlorobenzene (1.5 mL) was poured on the surface of the MAPbI 3 film $5 s after commencing spin-coating. The as-deposited MAPbI 3 films were annealed at 100 C for 10 min to form the resultant films. For the Cs 0.05 FA 0.80 MA 0.15 PbI 2.55 Br 0.45 (CsFAMA) films, a 1.2 M precursor solution was prepared with 0.06 M CsI, 0.96 M FAI, 0.18 M MABr, 1.02 M PbI 2 , and 0.18 M PbBr 2 dissolved in DMF/DMSO co-solvent (V DMF /V DMSO = 4:1) and 10% extra PbI 2 (mol %, relative to CsFAMA) was added to the CsFAMA precursor solution. 7,34 The precursor solution was stirred at $25 C for 24 hr and filtered with a 0.45-mm PTFE syringe filter before further use. To obtain the CsFAMA film, we spin-coated the CsFAMA precursor solution on the substrates at 2,000 rpm for 10 s and 4,000 rpm for 20 s, respectively. During the second step, 1.5 mL of chlorobenzene was poured on the top surface of the CsFAMA film $5 s before the end of the spin cycle. The as-deposited films were then annealed on a hotplate at 100 C for 10 min to form the resultant CsFAMA films. All the preparation and deposition steps were performed in a nitrogen-filled glovebox. Device Fabrications ITO-coated glass substrates (10 U/sq) were cleaned in soapy water, deionized water, acetone, and isopropanol with sonication. The ITO-coated glass substrates were then subjected to ultraviolet-ozone (UVO) treatment for 10 min. For the ETL-free perovskite solar cells, the perovskite films were directly deposited on the ITO substrates according to the procedure mentioned above. Li-doped Spiro-OMeTAD was then spin-coated as the hole-transporting layers on the perovskite films. A solution consisting of 72.5 mg of Spiro-OMeTAD, 28.8 mL of 4-tert-butylpyridine, 17.6 mL of Li-bis(trifluoromethanesulfonyl)imide (Li-TFSI) solution (520 mg Li-TFSI in 1 mL of acetonitrile), and 1 mL of chlorobenzene was employed with a spin speed of 3,000 rpm for 30 s. To complete the device, Au was thermally evaporated on the hole-transporting layers to serve as the electrode. For the devices with SnO 2 ETLs, the SnO 2 ETLs were deposited by spin-coating a SnO 2 suspension (15 wt % in H 2 O) in air on the UVO-treated ITO substrates and then annealing on a hotplate at 150 C for 20 min to form $20-nm-thick ETLs in air. 30 The glass/ITO/SnO 2 substrates were then treated with UVO for 10 min. Finally, the perovskite, Li-doped Spiro-OMeTAD and Au layers were sequentially deposited to complete the ETLcontaining perovskite solar cells by following the procedures described above for the ETL-free devices. ## Characterizations Morphologies of the CsFAMA and MAPbI 3 films were imaged with a scanning electron microscope (FEI XL-30 SEM-FEG). AFM images were characterized using a scanning probe microscope (Digital Instruments Dimension 3100). XRD measurements were carried out on a PANalytical Empyrean Powder X-ray diffractometer using Cu Ka radiation. The charge-carrier lifetimes were characterized via TRPL experiments using an Edinburgh FLS980 fluorescence spectrometer with excitation wavelength of 510 nm. The steady-state PL was also measured with the Edinburgh FLS980 fluorescence spectrometer with excitation wavelength of 510 nm. Optical absorption measurements were performed on a Shimadzu UV-3600 spectrophotometer. The EQE was taken using a QE-R instrument from Enlitech without bias voltage. The J-V characteristics and steady-state output were measured using a Keithley 2420 source meter. The illumination source was a Newport Oriel 92192 solar simulator with an AM 1.5G filter, operating at 100 mW cm 2 . All devices were masked with area aperture of 0.09 cm 2 to define the active areas. All the J-V characteristics of perovskite solar cells were evaluated with voltage-scanning speed of 1.0 V/s. A standard silicon solar cell from Newport Corp was used as a reference for J-V and EQE measurements. All measurements were performed under ambient conditions with relative humidity level below 30%. ## SUPPLEMENTAL INFORMATION Supplemental Information includes 18 figures and can be found with this article online at https://doi.org/10.1016/j.chempr.2018.08.004.

chemsum

{"title": "Carrier Dynamics Engineering for High-Performance Electron-Transport-Layer-free Perovskite Photovoltaics", "journal": "Chem Cell"}

rapid_mechanochemical_synthesis_of_amides_with_uronium-based_coupling_reagents,_a_method_for_hexa-am

## Abstract: Solid-state reactions using mechanochemical activation have emerged as solventfree atom-efficient strategies for sustainable chemistry. Herein we report a new mechanochemical approach for the amide coupling of carboxylic acids and amines, mediated by combination of (1сyano-2-ethoxy-2-oxoethylidenaminooxy)dimethylaminomorpholinocarbenium hexafluorophosphate (COMU) or N,N,N′,N′-tetramethylchloroformamidinium hexafluorophosphate (TCFH) and K2HPO4. The method delivers a range of amides in high 70-96% yields and fast reaction rates. The reaction protocol is mild, maintains the integrity of the adjacent to carbonyl stereocenters, and streamlines isolation procedure for solid amide products.Minimal waste is generated due to the absence of bulk solvent. We show that K2HPO4 plays a dual role, acting as a base and a precursor of reactive acyl phosphate species. Amide bonds from hindered carboxylic acids and low-nucleophilic amines can be assembled within 90 min by using TCFH in combination with K2HPO4 or N-methylimidazole. The developed mechanochemical liquid-assisted amidation protocols were successfully applied to the challenging couplings of all six carboxylate functions of biotin[6]uril macrocycle with phenylalanine methyl ester, resulting in an 80% yield of highly pure hexa-amide-biotin[6]uril. In addition, fast and high-yielding synthesis of peptides and versatile amide compounds can be performed in a safe and environmentally benign manner, as verified by green metrics. ## INTRODUCTION Amide bond is widespread in both natural compounds and artificial materials. It occurs in molecules fundamental to life, such as peptides and proteins, as well as in synthetic polymers and in a massive array of pharmaceuticals. In fact, amide preparation from carboxylic acids and amines represents the most frequently applied chemical transformation in drug production and comprises about 25% of the current medicinal chemistry synthetic toolbox. 1 As a consequence of its wide usage, the development of sustainable amidation methods was listed among the top green chemistry research priorities by the American Chemical Society Green Chemistry Pharmaceutical Roundtable (ACS GCIPR) in 2007 2 and has been retained in the recent revision. 3 Although direct condensation of carboxylic acids and amines with water as a single by-product can be considered a "green" landmark in the field, it remains impractical because of the process's harsh reaction conditions (T > 100 °C) and limited substrate scope. A robust method of amide synthesis commonly requires prior activation of a carboxylic function to replace OH with a better leaving group. Notably, this is also the case in biosynthetic pathways, including the translation process and non-ribosomal enzymatic transformations. For laboratory and industrial use, vast numbers of amide coupling reagents, performing in situ activation of carboxylic acid, have been developed in the quest for faster, milder, and high-yielding amidation protocols. 14,15 Low atom economy of these reagents and accompanying safety issues are their major drawbacks, which has incited the development of alternative approaches. Important advancements have thus far followed traditional solution-based approaches; however, solvent is actually responsible for 80−90% of mass consumption in a typical chemical process and also plays a major role in overall toxicity. 20 In this way, solvent greatly outperforms the contributions of reagents themselves. Hazardous solvents like DMF and DCM are preferred in industrial amide synthesis, reinforcing both environmental and safety concerns. 17,21 Therefore, the application of solvent-free techniques represents an efficient way to improve the overall process mass intensity and to prevent generation of hazardous waste. Recent advances in mechanochemistry and its related fields have established solvent-free reactions as environmentally friendly tools to perform chemical transformations that are no less efficient than the conventional solution-based chemistry. In the area of amide synthesis, the benefits of solvent-free techniques have not remained unnoticed and have been previously demonstrated in numerous studies (Scheme 1). For example, mechanosynthesis of various amides and peptides has been performed from a series of activated carboxylic acid derivatives, such as N-carboxyanhydrides; 28,29 N-hydroxysuccinimide esters; 30 Nacyl benzotriazoles. 31 N-Acyl imidazoles 32 and acyloxytriazine esters 33 have been produced mechanochemically from carboxylic acids prior to reacting with amines. Notably, even papain enzyme can catalyze the formation of peptides from the corresponding amino acid building blocks under solvent-free conditions. 34,35 In addition, direct coupling of amines with carboxylic acid has been demonstrated by using 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC) as a coupling reagent. 36,37 In general, EDC-mediated transformations have shown remarkably short reaction times (typically within 10-30 min), high yields, and simple work-up protocols. ## Current Work Following these prominent earlier contributions, we aimed to further expand the scope and synthetic utility of the mechanochemical amidation methods. The current research was impelled by three objectives: First, most of the amide coupling reagents are simply not efficient enough for a range of substrates, 8 which require expansion of the established one-step mechanochemical amidation protocols beyond the previously applied EDC; for that purpose, in this work we mapped the coupling efficiency of uronium-type reagents (COMU and TCFH, Scheme 1) on several carboxylic acid/amine pairs. Second, the scope of previously published mechanochemical approaches was evaluated based mainly on peptide synthesis, while the challenging couplings of sterically hindered carboxylic acids and low-nucleophilic amines remained virtually unproven; here we demonstrated that such difficult amide bonds can also be assembled under solvent-free conditions. Implementation of the two objectives mentioned above was required as a prerequisite for the third objective as our ultimate goal. Due to the interest of our group in the synthesis and supramolecular applications of macrocyclic host molecules, we required a robust procedure for amide-functionalization of biotin uril macrocycle (1), to access the family of modified biotin uril hosts. Despite the apparent ease of such a transformation, it also presented a substantial challenge: six-fold stepwise amidation of carboxylate groups in 1 is inevitably accompanied by accumulation of the "failed" underfunctionalized products if incomplete coupling occurs at each step. Limited solubility of 1 in the common organic solvents dictates additional practical inconvenience of the traditional solution chemistry; in fact, only dipolar aprotic solvents like DMF can be used. Here we showed that application of solvent-free techniques, additionally reinforced with the reactive uronium-type amide coupling reagents, allows the desired functionalization of 1 in a high-yielding, scalable, and sustainable manner, avoiding harmful solvents or significant reagent excess. ## RESULTS AND DISCUSSION Development of Mechanochemical Amidations with Uronium-Type Reagents. At the outset, amide coupling of Cbz-protected L-phenyl alanine (2) and ethyl 4-aminobenzoate (benzocaine, 3), mediated by COMU as a representative "green" uronium-type amide coupling reagent, was selected as a model process (Scheme 2). We aimed to screen and compare the results of various reaction conditions, including the evaluation of coupling efficiency for different coupling reagents beyond the COMU itself, to reveal the most promising hits in terms of product yield and green chemistry requirements. The choice of aromatic amine 3 was dictated by its reduced nucleophilicity in comparison with aliphatic amines, additionally attenuated by an electron-withdrawing ethoxycarbonyl group. We expected that suppressed reactivity of 3 in the carbonyl addition reactions would facilitate more reliable differentiation of various coupling conditions. Use of phenyl alanine derivative 2 as coupling counterpart provided an additional opportunity to examine the resistance α-stereocenter towards its possible epimerization, as commonly encountered in peptide synthesis. 9,15 ## Scheme 2. Optimization Experiments The test reactions were run in a Form-Tech Scientific FTS1000 shaker mill operating at 30 Hz by using 14 mL zirconia-coated milling jars, 3 × 7 mm zirconia milling balls and typical solid reactants loading around 0.3−0.4 g (including 0.2 mmol of amine 3 as a limiting substrate). After 30 min milling time, a sample of the crude reaction mixture was treated with CDCl3, followed by separation of insoluble inorganic materials. The conversion of amine 3 into amide 4 was determined by 1 H NMR analysis (see Supporting Information for the details). The amide coupling reagent, base, and amount of liquid additive needed to assist the grinding process (Scheme 2) were identified as the three most crucial parameters affecting the yield of amide 4, as described below. The addition of a small volume of liquid constitutes an efficient method to enhance the performance of solvent-free mechanochemical reactions, known as liquid-assisted grinding (LAG). 22,24 The ratio of the volume of liquid (μL) added to the amount of solid present (mg) is denoted as η (μL/mg). 50 A value of η = 0 generally corresponds to dry grinding, but in a typical LAG process, η is usually between 0 and 1. 24 Although LAG cannot be described as a totally solvent-free technique, it requires a minimal amount of liquid, especially advantageous if a green solvent is used. Among the latter, 20,51,52 ethyl acetate appears to be the most promising and chemically compatible candidate to act as a LAG additive in COMU-mediated amide coupling. In our experiments (Scheme 2, Chart 1), the addition of ethyl acetate indeed showed a pronounced effect on the yield of amide 4, generated in the mixture of solid reactants 2 and 3, with COMU reagent and sodium carbonate (ca. 10 equiv.) as a base. Although dry grinding provided a rather modest outcome (44% conversion), LAG resulted in a markedly improved reaction performance, with the optimal η value in a range of 0.14−0.24 μL/mg, while the further increase of η led to slightly diminished conversion values. The choice of base is also important in amide coupling. State-of-the-art solution approaches commonly apply non-nucleophilic tertiary amines, e.g. N,N-diisopropylethylamine (DIPEA). 15 However, the use of cheap and non-toxic inorganic salts, e.g. NaHCO3, K2CO3, NaH2PO4, 28,30,33,37 insoluble in common organic solvents, can be considered as an additional advantage of mechanochemical reactions. In our hands (Scheme 2, Chart 2), replacement of DIPEA with Na2CO3 gave similar conversion values (72% vs 75%). For further process optimization, a range of readily available phosphate salts, with notably distinct pKa values, were screened. Among them, potassium pyrophosphate K4P2O7 and dipotassium phosphate K2HPO4 provided the best outcomes, especially the latter (96% conversion). Generally, the performance of phosphate salts does not correlate with Brønsted basicity of the respective anions. Although the poor outcome with KH2PO4 (only 18% conversion) in comparison with K2HPO4 (96%) could be probably connected with the significantly reduced base strength of the former (respective pKa values 2.12 vs 7.21; pKa of RCO2H is typically about 4−5 in aqueous media), 53 much more basic K3PO4 (pKa 12.32) also afforded amide 4 with reduced efficiency (72%). Surprisingly, the counter-cation effect (Na + vs K + ) also had a prominent impact on reaction outcome (37% vs 96%, for Na2HPO4 and K2HPO4 respectively). These results clearly indicate that the effect of an inorganic base on a solid-state reaction is more intricate than trivial proton transfer. Finally, amide coupling reagents are essential for attaining high yields. The selection of coupling reagent was governed by considering chemical (substrate scope, reactivity); safety; and environmental issues. Uronium salts are advantageous because of their prominent reactivity and efficient reaction rates, 8,14 but the most commonly applied triazole-based reagents, such as HBTU and HATU, possess dangerous explosive properties 54 and pose significant health risks. 55 COMU was introduced as a safe and "greener" replacement. 47,48,56 To our delight, COMU also noticeably exceeded the coupling efficiencies of HATU and EDC in our experiments (Scheme 2, Chart 2), delivering a high 96% conversion. TCFH can be considered as an even more reactive alternative with better atom economy, affording a high 97% yield of amide 4 within only 10 min. The mechanochemical amidation with COMU/K2HPO4 was also rapid, reaching the maximal conversion within 20 min (Figure 1; see the Supporting Information for further detail), far surpassing the rate of the solution-based process (in DMF-d7, Figure 1). The latter reached the maximal 70% conversion after approximately 20 h (see the Supporting Information). Concurrently, about 30% of COMU reagent degraded due its well-known hydrolytic instability in DMF solutions, which is often referred as the main disadvantage of COMU. 57,58 Evidently, this drawback can be fully eliminated under solvent-free conditions. After achieving these results in the optimization experiments, we formulated the optimal experimental procedure as follows: COMU or TFCH (1.1 equiv.) as coupling reagents; K2HPO4 (3 equiv.) as base; ethyl acetate as LAG additive, and 20 min milling time. The amount of solid base (3 equiv.) was adjusted to keep η within the optimal range (~ 0.2 μL/mg), but not less than 2 equiv. required according to the reaction stoichiometry. Furthermore, an additional equivalent of K2HPO4 was required to release free amine when ammonium salt was used as the starting material. Isolation of pure amide 4 was achieved with a high 96% yield by simple water wash and filtration since all by-products are water soluble. No detectable racemization of the chiral center in 4 occurred during the synthesis, as was established by the chiral phase HPLC chromatography (see the Supporting Information). Green Chemistry Metrics Comparison. The advantages and drawbacks of the developed mechanochemical amidation methods were further revealed and compared with the solution-based reaction by analyzing the respective green metrics (Table 1). The metrics were calculated and assessed by marking them with red, orange, or green flags by following the Clark's unified metrics toolkit (see Supporting Information). 59 Atom economy (AE), reaction mass efficiency (RME), and process mass intensity (PMI) are defined as follows: 59 ## AE = molecular weight of product total molecular weight of reactants × 100 RME = mass of isolated product total mass of reactants × 100 PMI = total mass in a process mass of product First, isolated yields and product purity were much better in mechanochemical reactions, due to the higher conversion and more facile isolation procedure discussed above. Atom economy was a bit higher for the TCFH-mediated reaction because of lower molecular weight of TCFH. RME reflects both product yield and atom economy issues and was lower for the solution-based reaction. Comparison of PMI values clearly shows that mechanochemical reactions produce far less waste. Excluding mass-extensive work-up procedures, solvent occupied 84% of PMI for the solutionbased reaction and only about 15% (LAG additive) for the mechanochemical conditions. Furthermore, sustainable solvents like water and ethyl acetate were used in the latter, in contrast with toxic DMF. To determine the safety risks, a combination of physical, health, and environmental threats must be assessed, which can be done with the help of MSDS 60 and further available safety data. 54 DMF, for instance, is a flammable (H226), acute toxic (H312, 332), as well as a reproductive toxin (H360) and can thus be cited as the main hazard contributor for the solution-based process, which therefore received a red flag. For the mechanochemical reactions, the TCFH-mediated process was given a red flag due to the production of tetramethylurea by-product (reproductive toxin, H360). On the other hand, exothermic decomposition with a thermal onset of 127 °C can be considered as the main hazard of COMU, according to a recent study. 54 However, this property produced an orange flag, since COMU-mediated mechanochemical amidation protocol operates at room temperature. To conclude, although the developed mechanochemical amidation conditions cannot be considered totally safe, the risks are minimal because of its room temperature operation and relatively low amount of produced waste, as opposed to the solution-based reaction (see Supporting Information for additional safety considerations). Substrate Scope for mapping reactivity with COMU and TCHF. Having established the optimal conditions, substrate scope and limitations was briefly examined on a range of amine and acid coupling partners (Scheme 3). The substrate scope included, besides other, N-and C-protected amino acids and pharmaceutically relevant starting materials (e.g. (S)-naproxen, (S)-ibuprofen, benzocaine 3, N-Boc-protected piperazine). In addition to the foremost example of Cbz-masked amide (S)-4 comprehensively described above, its Fmoc-protected analogue (R)-5 was obtained in a high 95% yield by using the COMU-mediated reaction. Following the same protocol, dipeptides 6 and 7 with sterically hindered amino acid residues (phenylalanine and valine) were flawlessly prepared in high yields. No detectable epimerization of the stereocenters was noted in these cases. Coupling of (S)-(+)-6methoxy-α-methyl-2-naphthaleneacetic acid [(S)-naproxen] with amine 3 provided a more demanding test for stereochemical integrity, since 2-arylpropionic acids are prone to easy epimerization. The amide product (S)-8 was obtained from (S)-naproxen with a high 86% yield and excellent stereochemical purity (>99% ee). This was also the case in the TCFH-mediated reaction, which showed high reactivity and a subtle amount of epimerization for (S)-9. On the other hand, amidation of (S)-ibuprofen (98% ee) produced (S)-10 with slightly degraded optical purity (93-94% ee). Crude amides 10, 12, 15 appeared as oils immediately following the milling, which eventually enabled a chromatographic isolation for these cases (see Supporting Information). One advantage of TCFH over COMU-mediated amide coupling is the higher reactivity of the former reagent, which makes it more suitable for less reactive substrates. This property was explicitly revealed during the amidation of sterically hindered ortho-substituted benzoic acids. Thus, coupling of benzoic acid with benzocaine 3 proceeded well under the COMU-mediated protocol, furnishing amide 11 in a 93% yield after 20 min of milling time. Conversely, 2,4,6trimethylbenzoic acid under the same conditions produced only 22% of the target amide 14, without any further improvement, even when a longer milling time (up to 60 min) was applied. After the brief optimization studies (see Supporting Information), we found that a slight excess (1.3 equiv.) of more reactive TCFH and at least 60 min of milling time are required to attain a high 89% yield of 14. Moreover, chromatographic purification of 14 was necessary to separate mesitoic anhydride impurity. Diminished reactivity was also observed for 2,6-difluorobenzoic acid, furnishing amides 15 and 16 in reactions with N-Boc piperazine and low-nucleophilic amine 3 in acceptable yields after milling times of 40 and 60 min, respectively. On the other hand, coupling of the same amines with benzoic and 3,5-bis(trifluoromethyl)benzoic acids proceeded flawlessly, producing amides 12 and 13 with excellent yields and brief reaction times. Activating Effect of Phosphate Salts. During the optimization studies, the enhancement of yields with dipotassium phosphate and potassium pyrophosphate was especially notable (Scheme 2, Chart 2). We speculated that phosphate salts could additionally contribute to the activation of the carboxyl substrate 2 via the formation of acyl phosphate intermediates containing a "high-energy" phosphoester bond, prone to easy nucleophilic amine attack. 65,66 Interestingly, the same pathway is also involved in the ATP-dependent biosynthesis of amide bond-containing biomolecules. 11 The plausibility of our assumption is further supported by existing literature showing that acyl phosphates can be indeed generated in solution by the DCC-mediated coupling of carboxylic acids with phosphate salts. To confirm the credibility of our hypothesis, mechanochemical synthesis of acyl phosphates from carboxylic acids and phosphate salts, mediated by COMU and TCFH, was attempted. As expected, a 20-min ball milling of COMU (1.1 equiv.) with acetic acid (1 equiv.) and K2HPO4 (3 equiv.) yielded 60% of acetyl phosphate 17, which was confirmed by NMR analysis of the freshly obtained reaction mixture in D2O solution (Figure 2). Acetyl phosphate 17 displayed a singlet signal at δ = −2.1 ppm in 31 P NMR, which rapidly disappeared after the addition of morpholine, both in D2O solution and in the solid state (see Supporting Information). In the 13 C NMR spectrum, carbonyl group 17 showed a doublet signal at δ = 168.1 ppm (JCP = 8.8 Hz), due to its coupling with the neighboring phosphorus. 65 Significantly lower yields of 17 were attained with K3PO4 or with TCFH as coupling reagent (Figure 2). The reaction of acetic acid with K4P2O7 produced acetyl pyrophosphate 18 in a 50% yield, according to 31 P-NMR analysis. As a result of the non-equivalence of phosphorus atoms in 18, a pair of doublet signals appeared in 31 P NMR, at δ = −5.0 and −17.9 ppm (d, JPP = 21.7 Hz), thus confirming its structure. 67 As an extra example, the generation of acyl phosphate 19 (50% yield, δ = −7.6 ppm in 31 P NMR) was also successful from Cbz-masked phenyl alanine 2, which was similar to the acetic acid outcome. For example, K3PO4 produced a rather low 15% yield of acetyl phosphate 17 (Figure 2), which also agrees with the lower conversion to amide 4 in the comparison with K2HPO4. The acyl phosphate pathway probably contributes less in the case of the more reactive TCFH reagent, which also produced a rather low 30% yield of 17 (Figure 2). The exact mechanistic sequence leading to acyl phosphates C from COMU, RCO2H and K2HPO4 remains unclear but may include the reaction of acyl uronium intermediate A with HPO4 2anion (Scheme 4) or, alternatively, the initial formation of uronium phosphate 71 by the reaction of COMU with K2HPO4. ## Scheme 4. Plausible Mechanistic Pathways Leading to Amide Product Challenging Amide Bond Formation. As shown above, the coupling of low nucleophilic amine 3 with sterically hindered mesitoic acid could be efficiently mediated by the TFCH/K2HPO4 reagent system (Scheme 3). In accordance with existing literature, 61,72 we selected the coupling of electrondeficient 4-aminobenzonitrile 20 with 2-methyl-2-phenylpropanoic acid 21 (Scheme 5), an even more arduous way to test the performance of mechanochemical amidation protocols. Brief screening of various coupling conditions was undertaken, and conversion to amide product 22 was determined by 1 H NMR analysis after 60 min of milling time (Scheme 5). ## Scheme 5. Mechanochemical Coupling of Hindered Carboxylic Acids and Poor Nucleophilic Amines The use of EDC alone, 36 or the COMU/K2HPO4 system, yielded only a low ~10% conversion. Combination of TCFH and K2HPO4 delivered a noticeably better outcome but still failed to raise the conversion above 22%. According to the recent study of Beutner et al., 61 N-methylimidazole (NMI) and TCFH reagent combined provided a high yield of 22 in solution, due to in situ generation of reactive N-acyl imidazolium ions. To our gratification, same combination of reagents also worked well under the solvent-free conditions, affording respectable 84% conversion after a 60-min reaction time. Finally, a slight excess (1.3 equiv.) of TCFH reagent, along with a bit longer milling time (90 min), allowed us to obtain pure amide 22 in 92% isolated yield after an aqueous work-up (see Supporting Information). Following the same reaction protocol, the coupling of 21 with sterically hindered 2,4,6-trimethylaniline was performed and furnished the corresponding amide 23 with a 92% yield. Notably, high yields of amides 22, 23 were attained in a rather efficient reaction time of 1.5 h, in significant contrast with the solution-based reaction (21 h for amide 22). 61 Surprisingly, the same highly reactive combination of reagents failed to render amide 14 from mesitoic acid with yields exceeding 20%. This was also the case in the CD3CN solution (see Supporting Information). We found that the reaction was stopped due to the formation of sterically bulky and therefore non-planar N-acyl imidazolium D, which, in contrast to the analogous species produced from benzoic acid, was totally inert towards the subsequent reaction with amine 3 (see Supporting Information for further detail). Inertness of D could be explained by the efficient steric shielding of the carbonyl group with both neighboring mesityl and imidazolyl moieties, preventing attack of a nucleophile along the Bürgi-Dunitz trajectory (Scheme 5). This stands in sharp contrast to the successful TCFH/K2HPO4-mediated transformation, where the less sterically crowded intermediate species are expected to form (e.g. mesitoyl chloride, uronium or phosphate). Amide Coupling of Biotin uril. As a part of our ongoing efforts towards the development of new chiral supramolecular receptors, we needed an expedient synthetic procedure for derivatization of biotin uril (1), 44 easily available in multigram quantities by HCl-catalyzed condensation of formaldehyde with D-biotin. The starting macrocyclic molecule, notable for its anion binding properties, common for the cucurbituril family, satisfies 6 carboxylic functions, which could be conveniently coupled with various amines, thus providing facile access to a library of diversely functionalized chiral macrocyclic receptors. Although amide coupling of carboxylates in 1 might appear simple, unencumbered by any steric or electronic influence, full amidation of 1 is challenging because it proceeds via six consecutive steps. For example, if a high 97% yield were produced during each step, the fully functionalized product would eventually generate only a (0.97) 6 •100% = 83% yield, while the rest of the produced material would contain a set of "failed" underfunctionalized molecules, thus necessitating time-consuming, laborious and mass-inefficient chromatographic purification. The situation resembles the synthesis of oligopeptides and oligonucleotides, in which an extremely high coupling efficiency (>99% per coupling step) is required to attain reasonable yields and high purity of long-chain oligomers, and it is customarily achieved by using an excess of highly reactive coupling reagents. 56 The low solubility of 1 in the environmentally benign and volatile organic solvents, compatible with the conventional amidation protocols (e.g. ethyl acetate), constitutes an additional restriction of the solution-based chemistry. We believed that the high coupling efficiency observed under the solvent-free conditions would allow us to perform the desired functionalization in a high-yielding, and scalable manner without using an excess of reagents, toxic solvents, or laborious purification. As a convenient model reaction for this study, we selected the amide coupling of 1 with methyl ester of phenyl alanine 24 (used as HCl salt, see Scheme 6). At its outset, this task required us to explore the performance of different amide coupling conditions. Only a slight excess of amine 24 and a coupling reagent (7-7.8 equiv., which is 1.16-1.3 equiv. per CO2H group of 1) were applied in the optimization experiments. It was expected that more reactive combinations of reagents would deliver higher yields of the hexa-amide product 25. Based on our previous findings, the order of coupling efficiency for the different reagent systems can be roughly plotted as follows: EDC ~ COMU/K2HPO4 < TCHF/ K2HPO4 << TCHF/NMI. Although such generalizations must be made with care since the coupling performance is substrate-dependent 8,78 and exceptions are possible (e.g. case of amide 14 above), this preliminary reactivity plot provided a helpful guide. Outcomes of the test reactions were analyzed by HPLC (Figure 3, see Supporting Information for further detail) and quantified by calculating HPLC area percentage for the hexa-amide product 25 (Srel, Table 2), relative to underfunctionalized compounds. These initial experiments (Table 2, entries 1-4) clearly indicated that complete hexafunctionalization of 1 is difficult to perform. Thus, both the COMU and TCFH/K2HPO4 systems produced a mixture of phenylalanine-derivatized biotin urils, containing all possible products from mono to hexa-amide 25, the latter displaying a rather low 16% contribution (entries 1 and 2; Figure 3A). The use of EDC/DMAP combination (entry 3), 36 was more successful in this case, primarily producing a mixture of penta-and hexa-amides (Figure 3B). The highly reactive TCFH/NMI combination (entry 4) generated hexa-amide 25 as its main reaction product, but it was noticeably contaminated with underfunctionalized compounds (52% HPLC area, Figure 3C). Notably, at least 90-min milling times had to be applied, since samples taken after 30 and 60 min still showed incomplete conversion (see the Supporting Information). Although the FTS1000 shaker mill could hypothetically achieve long milling, we considered any time longer than 1.5 h as impractical; therefore, our next goal was to adjust the reaction parameters accordingly, in order to reach at least 90% conversion within a 1.5-h reaction time. Applying a slightly greater excess of TFCH (1.2 equiv. per carboxylate) and NMI (3.5 equiv. per carboxylate) noticeably improved the yield of the target product, 25 (86% HPLC area, entry 5), and also shortened the reaction time. For further improvement, a screening of optimal η and LAG additive was performed. Since NMI is a liquid, and liquid tetramethylurea is produced, the addition of solid NaCl was attempted to reduce the initial η to 0.16 μL/mg. However, this distinctly reduced the yield of the product (73% HPLC area, entry 6). On the other hand, the addition of a few drops (ca. 35-50 μL) of solvents noticeably improved the outcome (entries 7 to 9) and was best when polar solvents like DMF or EtOAc were added (entries 7 and 9). These results clearly indicate that the nature of LAG additive plays an important role 37 and can substantially increase reaction rate, a probable result of the favorable interactions of the polar reactants with the mobile surface layers of LAG additive and improved mass transfer. 24 The outcome with EtOAc was especially remarkable, providing 25 with the best purity (98% HPLC area, Figure 3D). Since the reaction mixture visibly liquefied as the reaction progressed (due to the generation of tetramethylurea), slurry stirring was also tried instead of the ball milling (entry 10) and resulted in slightly reduced coupling efficiency. Solution-based amide couplings were performed in DMF (entries 11 and 12) for the comparison with mechanosynthesis. Homogeneous solutions were obtained with an amount of solvent (ca. 0.5 mL) comparable to the weight of solid reactants (ca. 0.24 g), what kept η at around 2 μL/mg. In the DMF solution, HATU was another frequently used and highly reactive uronium-based amide coupling reagent that produced a rather modest outcome (entry 11). Conversely, the coupling efficiency of the TCFH/NMI combination in DMF solution (entry 12) was virtually the same as in the DMF-free transformation to a solid state (entry 9). Importantly, a bulk amount of harmful solvent was fully avoided in the latter. Under the optimal reaction conditions (entry 9), the desired hexa-amide product 25 was isolated in a nearly quantitative yield and 95% HPLC-purity (relative to all other peaks) after the simple water wash and filtration. Purity of product was further increased (99% according to HPLC) by following the simple purification protocol (filtration of chloroform solution via Celite ® , and then precipitation with hexane from EtOAc solution [see Supporting Information]). The same amide coupling reaction was also successful at loadings that were 3 times higher (150 mg of 1 per milling jar, 300 mg total), creating 25 in 80% isolated yield and 99% HPLC purity, albeit with a longer milling time (90 min).

chemsum

{"title": "Rapid Mechanochemical Synthesis of Amides with Uronium-Based Coupling Reagents, a Method for Hexa-amidation of Biotin[6]uril", "journal": "ChemRxiv"}

bioorthogonal_two-component_drug_delivery_in_her2(+)_breast_cancer_mouse_models

## Abstract: The HER2 receptor is overexpressed in approximately 20% of breast cancers and is associated with tumorigenesis, metastasis, and a poor prognosis. Trastuzumab is a first-line targeted drug used against HER2(+) breast cancers; however, at least 50% of HER2(+) tumors develop resistance to trastuzumab. To treat these patients, trastuzumab-based antibody-drug conjugates (ACDs) have been developed and are currently used in the clinic. Despite their high efficacy, the long circulation half-life and non-specific binding of cytotoxic ADCs can result in systemic toxicity. In addition, standard ADCs do not provide an image-guided mode of administration. Here, we have developed a two-component, two-step, pre-targeting drug delivery system integrated with image guidance to circumvent these issues. In this strategy, HER2 receptors are pre-labeled with a functionalized trastuzumab antibody followed by the delivery of drug-loaded nanocarriers. Both components are cross-linked by multiple bioorthogonal click reactions in situ on the surface of the target cell and internalized as nanoclusters. We have explored the efficacy of this delivery strategy in HER2(+) human breast cancer models. Our therapeutic study confirms the high therapeutic efficacy of the new delivery system, with no significant toxicity. The HER2 receptor is one of four human plasma membrane receptors from the ErbB tyrosine kinase receptor family 1 . HER2 regulates the cellular proliferation and survival of cells by dimerizing with other ErbB receptors and stimulating tyrosine kinase activity 2,3 . Approximately 20-30% of human breast cancers overexpress HER2 receptors by amplification of the HER2/neu gene, which is a marker of aggressive cancer with an unfavorable prognosis that correlates with tumorigenesis and metastasis . The anti-HER2 humanized monoclonal antibody, trastuzumab (Tz), is a first-line biotherapeutic against HER2(+ ) breast cancer 8,9 . However, recent clinical statistics have revealed that the long-term use of trastuzumab can generate trastuzumab resistance in HER2(+ ) tumors 10 . The causes of trastuzumab resistance are not yet fully understood . The HER2 receptor also exhibits a poor ability to internalize, even after the binding of bioligands, such as anti-HER2 antibodies, HER2-specific antibody fragments, aptamers, and peptides 13 . Poor internalization of HER2 is also a potential drawback to the use of this receptor as a therapeutic target. To overcome the trastuzumab resistance in HER2(+ ) tumors, trastuzumab-based antibody-drug conjugates (ADCs), such as trastuzumab-emtansine (T-DM1), have been developed with the chemotherapeutic drug mertansine directly attached to the antibody, which boosts the cell toxicity. After successful clinical trials, T-DM1 is currently used in the clinic 14,15 . However, by design, ADCs are intrinsically highly toxic and can produce severe side effects due to their long circulatory half-life and non-specific toxicity in healthy tissues 16 . Furthermore, a simple ADC does not provide a mechanism by which to enhance the cellular internalization of therapeutics to maintain a high therapeutic index 17 . To circumvent these issues, we have designed a pre-targeting two-component, two-step drug delivery system driven by bioorthogonal click chemistry between the pre-targeting and delivery components. In this strategy, HER2(+ ) cancer cells are pre-labeled by click-reactive trastuzumab, and subsequently, click-reactive, drug-loaded albumin nanocarriers (Alb) are delivered to enhance the internalization of drug carriers after the in situ bioorthogonal cross-linking of components, as shown in Fig. 1. The pre-targeting approach was used for imaging and therapy in lung cancer . We have also demonstrated the efficacy of HER2 pre-targeted therapy in isolated cells 21 . However, to the best of our knowledge, this is the first demonstration of an anti-HER2 pre-targeted therapeutic strategy in vivo. By definition, bioorthogonal reactions are reactions that can occur in living systems at physiological conditions without interfering with regular biochemical and physiological processes. For example, the Staudinger ligation, the copper-free azide/alkyne click reaction, and the trans-cyclooctene/tetrazine cycloaddition have been explored for in vitro and in vivo imaging. Azide (Az)/difluorocyclooctene (DIFO) click chemistry has been used by Bertozzi's group for in vivo imaging in living systems 22 . Kim's group has demonstrated the in vivo imaging of mouse tumor models using Az and dibenzylcyclooctyne (DBCO) click chemistry between metabolically labeled glycans and liposomes 18 . Trans-cyclooctene (TCO)/tetrazine (Tt) cycloadditions have been extensively studied for in vivo imaging experiments because this reaction is extremely fast (3,100-380,000,000 × 10 3 M −1 s −1 ) compared to the Az/DBCO (0.9-4,000 × 10 3 M −1 s −1 ) and Az/DIFO (7.6 × 10 −2 M −1 s −1 ) bioorthogonal click reactions 23,24 . The TCO/Tt click reaction has been used to image nanoparticles in living systems 25 . We have previously used Az/DBCO bioorthogonal click chemistry in two-component drug delivery system to evaluate the strategy in vitro in HER2(+ ) BT-474 cells 21 . We observed the cluster formation and internalization of nanoclusters with confocal fluorescence imaging. Our in vitro therapeutic experiments confirmed the high therapeutic efficacy of a two-component, two-step drug delivery system. Due to its fast kinetics, TCO/Tt cycloaddition was used in this study for the in situ conjugation of the components of a two-component, two-step drug delivery system in a HER2(+ ) human breast cancer mouse model, as shown in Fig. 1. In this strategy, due to the close proximity of overexpressed HER2 receptors on cancer cells and multiple functionalizations of the pre-targeting and delivery components, multiple cross-linking reactions induce the self-assembling of cell membrane-bound nanoclusters in situ. These nanoclusters can be effectively internalized by clathrin-mediated endocytosis 26 . In the present study, we employed the new delivery system in mouse models of HER2(+ ) human breast cancer, evaluated the strategy, and determined whether this system could enhance therapeutic efficacy. ## Enhancement of cellular internalization by a two-component delivery strategy. The cellular internalization of components by the two-component strategy was evaluated in HER2(+ ) BT-474 breast cancer cells, using a confocal fluorescence microscope (Supplementary Figure S1A). As shown in Fig. 2A, the cell surface was initially labeled by a Tz(TCO) 6 (AF-488) 4 pre-targeting component. Co-localization of the delivery component, Alb(Peg 4 -Tt) 15 (Rhod) 4 , with the pre-targeting components on the cell surface was detected immediately within 15 minutes after labeling. Upon incubation at 37 °C for 4 h, we observed a rapid internalization of co-localized nanoclusters formed by the two components (Fig. 2B). No cellular internalization was detected after incubation at 20 °C (Fig. 2A). Moreover, neither co-localization nor internalization of components was observed when reactive Tz(TCO) 6 (AF-488) 2 and non-reactive Alb(Rhod) 4 were used as the pre-targeting and delivery components, respectively (Supplementary Figure S2). ## Enhancement of tumor uptake of components. We evaluated the tumor uptake and plasma clearance of circulating non-specifically bound components on an in vivo Xenogen optical imaging system (Supplementary Figure S1B). Three groups of mice, including click-treated, mock-treated, and untreated-controls were used in this study. The first two treatment groups, click-treated and mock-treated, were administered pre-targeting components, including a reactive Tz(TCO) 6 (CF-680) 2 (where CF-680 is an NIR fluorophore), and a non-reactive Tz(CF-680) 2 , respectively. Animals were imaged for 12 h to measure the biodistribution of pre-targeting components. Generally, a rapid clearance of both components from the circulation was detected in our models. A significant accumulation of TCO-functionalized and non-functionalized pre-targeting components was detected in the tumors at approximately eight hours. For TCO-functionalized Tz, the degree of functionalization (DOF) of 6 was maintained. There was no significant change in the binding affinity of Tz with HER2 at DOF = 6 (Supplementary Figure S3) 27 . At this time-point, plasma was free of excess pre-targeting components; however, some amount of the component accumulated in the kidneys (Supplementary Figure S4A) and was found in urine, as well. After 24 h, the amount of pre-targeting component in the kidneys had decreased significantly and was completely cleared from the urine (Supplementary Figure S4B). At the eight-hour post-injection, the Alb(Px) 2.6 (Pe g 4 -Tt) 15 (DL-800) 2 therapeutic carrier was administered to the click-treated and mock-treated groups, where Px and DL-800 were paclitaxel and DyLight 800 NIR fluorophore, respectively, while the untreated group received saline. Animals were imaged to track the drug delivery component routinely for two days. During this imaging time-frame, pre-targeting components were still visible in tumors in both click-treated (Fig. 3-i, intensity 1,174 a.u.) and mock-treated (Fig. 3-iii, intensity 1,362 a.u.) mice, and mice in the click-treated group showed a higher tumor uptake of the Alb drug delivery component (Fig. 3-ii, intensity 6,753 a.u.) compared to mice in the mock-treated group (Fig. 3-iv, intensity 3,327 a.u.). ## High cellular internalization of components in vivo. To explore the in vivo cell labeling and cellular internalization of components, we observed the tumor microenvironment using an intravital multiphoton microscope during the treatment (Supplementary Figure S1C). Mice were first administered Tz(TCO) 2 (Rhod) 2 intravenously, and, at eight-twelve hours post-injection, they were imaged after minimally invasive skin-flap surgery (Fig. 4A,B). The surface labeling of cancer cells by pre-targeting components was observed after 10 minutes and did not change significantly for twelve hours (Fig. 4C-i). At the next step, mice were administered Alb(Peg 4 -Tt) 15 (AF-488) 2 intravenously and the imaging was continued for approximately two hours post-injection. The delivery of Alb(Peg 4 -Tt) 15 (AF-488) 2 and the build-up of the fluorescence signal was detected and the signal was co-localized with the pre-targeting agent, Tz(TCO) 6 (Rhod) 2 (Fig. 4C-v). No significant auto-fluorescence was detected in the green channel used to image the drug carrier component (Fig. 4C-ii). The co-localization of components and the internalization of clusters was observed starting at 30 min post-injection, and reached a maximal level after 90 minutes (Fig. 4C-vi). We observed neither co-localization nor internalized nanoclusters in the intravital imaging experiment repeated with reactive Tz(TCO) 6 (Rhod) 2 (Fig. 4D-i) and non-reactive Alb(AF-488) 2 (Fig. 4D-ii). ## Enhancement of therapeutic efficacy with low toxicological effects. We evaluated the therapeutic effect based on the change in relative tumor volumes (RTV = tumor volume at day t, V t /V 0 , initial tumor volume), calculated from the tumor dimensions and measured by a caliper, over 28 days, with two doses of treatments (Supplementary Figure S1D). Mice in the click-treated and mock-treated groups were injected with Tz(TCO) 6 (CF-680) 2 and Tz(CF-680) 2 , respectively, and the untreated control group received saline. After eight hours, the first two groups were administered Alb(Px) 2.6 (Peg 4 -Tt) 15 (CF-750) 2 and the mice in the untreated-control group received saline. Mice received the second dose of therapy after 14 days. During the treatment period, the changes in tumor volumes and body weights were examined for 28 days and plotted as a function of time post-injection. The mice in the click-treated group exhibited a significant inhibition of tumor growth, as confirmed by the lowest RTV over the treatment period (Fig. 5A), compared to the mock-treated and untreated-control groups. The corresponding bar chart was used for the statistical analysis of the changes in relative tumor volumes (Fig. 5B). The relative tumor volume in the click-treated group was significantly low compared to the untreated group from day 8 onward (Fig. 5B), and it was significantly low compared to the mock-treated group after day 12 onward. The Kaplan-Meier analysis curves shown in Fig. 5C were obtained using changes in the terminal RTV by a factor of four from the initial tumor size. The highest RTV was detected in the untreated-control group. Visible changes in tumor sizes in different treatment groups are shown in Fig. 5D. We also performed a comprehensive study of the toxicological effects in mouse models of BT-474 breast cancer. Mice in the click-treated and mock-treated groups received both components at doses of 0× , 1× , 2× , and 5× following the same treatment schedule used for the therapy (Supplementary Figure S1D). Based on the representative results of the study, there was no body weight loss (Supplementary Figure S5), and only mild thrombocytopenia (Supplementary Figure S6) was detected at 24 h post-treatment dose in the high-dose (5× ) treatment group. However, in this group, the platelet count recovered by factor of five at day 28 of the therapy. ## Discussion We chose trastuzumab as the target-specific pre-targeting ligand for HER2, since Tz has high target-specificity and binding affinity to HER2 receptors, even after chemical conjugations on its free amine groups 27 . Furthermore, Tz resistance in HER2(+ ) cancer cells does not decrease the degree of HER2 expression 28 . There is no HER2-specific endogenous ligand in the human body for competitive binding affinity with Tz. As a chemotherapeutic, Px has high therapeutic efficacy against solid tumors and can be easily modified and conjugated with drug delivery platforms without altering cytotoxicity 29 . Paclitaxel albumin conjugates are also used in chemotherapeutic regimens (Abraxan) 30 . Serum albumin was chosen as the drug carrier for several reasons. Albumin is an acidic, hydrophilic, and highly stable globular protein. It is stable in a broad pH range (pH 4-9), in 40% ethanol, and at high temperatures (60 °C) without denaturing 31 . Hydrophobic low-molecular-weight chemotherapeutics can be chemically conjugated with albumin without a significant change in the hydrophilicity of albumin in plasma. Drugs encapsulated in albumin exhibit favorable pharmacokinetics with albumin as a drug carrier. Albumin can also be accumulated in solid tumors by the enhanced permeability and retention (EPR) effect; however, our approach was intended to increase the cellular uptake of drug-loaded nanocarriers rather than to enhance the accumulation of nanocarriers in the tumor extracellular microenvironment 32 . To synthesize drug-loaded nanocarriers, paclitaxel was first derivatized into an amine reactive sulfo-NHS analogue (sulfo-NHS-paclitaxel), and conjugated with albumin. The conjugation ratio of paclitaxel to albumin is a critical parameter that significantly decreases the hydrophilicity of the delivery component. Thus, the degree of conjugation (DOC) of paclitaxel was chosen at ~2.6 to synthesize the Alb(Px) 2.6 precursor. Paclitaxel was conjugated with albumin by an ester linkage, which is stable in vivo and enables the efficient release of drugs after internalization, followed by acidic or enzymatic cleavage. We evaluated the cell surface labeling and the internalization of nanoclusters in vitro by confocal fluorescence imaging. The poor internalization kinetics of Tz in HER2(+ ) cells increases its availability as a pre-targeting component on the cell surface for multiple click reactions with Alb-based delivery components. Multiple bioorthogonal click reactions in TCO/Tt-based delivery strategy were faster than those for previously used with Az/DBCO system 21 . The co-localization of two components on the cell surface provides efficient click reactions in physiological conditions. We also observed the formation of nanoclusters and their internalization after four hours incubation at 37 °C. The control Alb(Rhod) 4 delivery component, which lacked functional Tt groups, did not react with the pre-targeting component and none of the components were internalized upon incubation (Supplementary Figure S2). We explored the tumor uptake of components based on the in vivo fluorescent intensities of the components. Images acquired by Xenogen IVIS optical imaging suggested a high accumulation of pre-targeting Tz components in the tumor. Interestingly, we observed higher tumor uptake of non-reactive Tz(CF-680) 2 in the mock-treated mice compared to the reactive Tz(TCO) 6 (CF-680) 2 in the click-treatment. We suggest that this observation is, in part, due to the slight decrease in the binding affinity of trastuzumab functionalized with TCO (Supplementary Figure S3). In addition, specific click reactions between the components resulted in the formation of cross-linked complexes, which may have further reduced fluorescence by quenching. Finally, cell internalization of the complexes resulted in their fast degradation and clearance of the fluorescent marker observed in the bladder of the click-treated mice (Fig. 3-i). Clinically, a relatively long mean serum half-life was reported for unmodified Tz and Tz-based ADCs, T-DM1, of 5.83 days and 4 days, respectively 33 . Both TCO-functionalized and control pre-targeting Tz components exhibited a relatively short circulatory half-life in the plasma in our studies. While the Tz conjugation strategy did not saturate the available amino groups in the antibody, possible changes in the lipophilicity and surface charge of the component would depend on the DOF and the nature of the conjugation groups 34 and might affect the clearance and circulation time of the antibody. In addition, the nude mice used in our study have a functional complement system and mature B-cells. Therefore, humanized trastuzumab antibodies can be recognized and cleared by the host. The in vivo cell surface labeling, cluster formation, and internalization were assessed using intravital multiphoton imaging, which is a powerful technique with which to study dynamic processes in living animals 35 . This technique is particularly appropriate when observing cancer cells, the tumor microenvironment, and the microvascular architecture in tumor models in vivo 36 . Traditional dorsal window chambers are a common accessory used for intravital in vivo imaging studies due to the convenience of focusing and choosing a suitable FOV. However, these chambers allow a narrow time-window for tumor growth and imaging, and create an artificial microenvironment for the tumors. These chambers can be optimized for imaging in subcutaneous tumor models, but are difficult to use with orthotopic tumor models 37 . Therefore, in this study, we used a custom-made mouse-holder equipped with a compression window plate that can fix orthotopic tumors to allow imaging without motion artifacts. This set-up facilitates intravital imaging of cancer cells, the tumor microenvironment, and the vascular architecture after a minimally invasive skin-flap surgery (Fig. 4A). This imaging system was used for the real-time visualization of perfusion and extravasation of drug delivery components in the tumor microenvironment. Intravital imaging demonstrated high uptake and internalization of drug-loaded components in the delivery system driven by bioorthogonal click chemistry. After approximately two hours of skin-flap surgery, the rate of the blood flow in the tumor site was reduced, possibly due to the local thrombosis in the area. Mice in the click-treated group showed a higher therapeutic response, likely due to the tumor uptake of drug-loaded nanocarriers by cluster formation and subsequent cellular internalization (Fig. 5A). The mock-treated delivery system included non-reactive variants of both components and did not facilitate tumor uptake, cluster formation, and internalization driven by bioorthogonal click chemistry. However, we did detect an improved therapeutic response in mock-treated mice compared to untreated controls, and this observation was presumably attributable to the combination of nonspecific accumulation of the cytotoxic nanocarrier through the EPR effect and the therapeutic effect of Tz. The therapeutic efficacy in the click-treated group was significantly higher than the treatment effect in the untreated group (after day 8 onward) or in the mock-treated group (after day 12 onward). At the end of the therapeutic study, the click-treated group showed significantly low tumor sizes compared to both control groups (Fig. 5B). The Kaplan-Meier curves based on the time taken by tumors to reach four times the initial size (Fig. 5C) demonstrated a significant effect of the targeted therapy compared to mock-treated and untreated controls. The typical appearance of tumors that shrank in the click-treated group (left) compared to the mock-treated group (right) is shown in Fig. 5D. The nanoclusters of pre-targeting and drug-loaded delivery components were rapidly internalized by tumor cells through a receptor-mediated endocytosis and led to controlled-release of paclitaxel by enzymatic or acidic cleavage of the linker. During the course of the treatment, we observed no significant change in body weights in any group of mice. The analysis of in vivo images also confirmed the liver and kidney uptake of a trace amount of pre-targeting components and a medium amount of drug-loaded nanocarriers at the eight hours post-injection, with no toxicological effects. The liver and kidneys are major sites for metabolism and excretion in the animal body. Drug-induced toxicity in the liver and kidneys could disrupt their functionality and alter the composition of platelets (PLT), lactate dehydrogenase (LDH), blood urea nitrogen (BUN), alanine transaminase (ALT), and aspartate aminotransferase (AST) in the blood. During the treatment in this study, no significant drug-induced toxicity was detected in long liver and kidney blood panels, even at treatments with 3× and 5× doses. Pre-targeted therapy was originally proposed for radioimmunotherapy using biotin-avidin conjugation chemistry . A pre-targeting approach was also used for molecular imaging of cancer using specific antibodies and a biotin-avidin in situ multi-step conjugation 43 . Other investigators reported a pre-targeting approach using antibodies and bioorthogonal chemistry for imaging and radioimmunotherapy . Compared to those previously published reports, our strategy uses a unique combination of specific pre-targeting and cluster formation by multiple click reactions between the pre-targeting and therapeutic carrier components to achieve rapid internalization and delivery of the therapeutic cargo to cancer cells. We also suggest that the cross-linking between neighboring receptors is only plausible on HER2-overexpressing tumor cells; hence, the cluster formation and internalization of drug-loaded nanocarriers are likely not possible in healthy cells. In this approach, the size, location, and stage of tumor can be determined by imaging of the pre-targeting component before the administration of the drug-loaded delivery component. Since trastuzumab is used in this delivery system only as the target-specific molecule, the efficacy of therapy could be high for trastuzumab-resistant tumors as well. This approach also provides direct delivery of therapeutics via HER2-mediated endocytosis, and thus, avoids multi-drug resistance (MDR) transporters 47 . The low pH of subcellular compartments and the enzymatic environment of the cytoplasm trigger the hydrolysis of drug-linkers and controlled-release of Px. In summary, we have developed a two-step, two-component drug delivery system driven by bioorthogonal click chemistry and evaluated it in preclinical systems, in vivo. In this strategy, a trastuzumab-based first component was used to pre-label the HER2(+ ) tumor cells and a paclitaxel-loaded albumin carrier was subsequently delivered as the cytotoxic treatment component. This new two-component delivery system showed high accumulation of delivery components in the cancer cells and enhanced therapeutic efficacy. Image guidance can be provided by labeling the components with appropriate imaging agents, and can be used for cancer staging (location, size, and HER2 status) by tracking the low/non toxic pre-targeting component with noninvasive imaging. The results of pre-labeling can be used to make clinical decisions regarding the administration and timing of the cytotoxic drug carrier component with suitable chemotherapeutics to maximize efficacy and minimize non-specific toxicity and side effects. ## Materials & Methods Biotherapeutics, chemotherapeutics, and chemicals. ## Formulation of components. Trastuzumab (500 μL of 10 mg/mL in PBS) was treated with TCO-NHS ester (300 moles equiv. in 10-20 μL of dry DMSO) by gently stirring for one hour. Samples were purified by ultracentrifugation followed by HPLC. The resulting Tz(TCO) 6 (500 μL of 10 mg/mL in PBS) was treated with NHS-AlexaFluor 488, NHS-Rhodamine or NHS-CF-680 (10 moles equiv. in 10 μL of dry DMSO) and stirred for one hour. The degree of labeling (DOL) of the fluorophore was maintained at 2-4. The products were purified by ultracentrifugation followed by HPLC. The resultant Tz(TCO) 6 (AF-488) 4 , Tz(TCO) 6 (Rhod) 2 and Tz(TCO) 6 (CF-680) 2 were used as the pre-labeling component in vitro confocal fluorescence imaging, intravital multiphoton confocal imaging, in vivo imaging and therapeutic studies, respectively (Supplementary Figure S7A). For imaging and therapeutic experiments, albumin or drug-loaded albumin 21 (2 mL of 10 mg/mL in PBS) was treated with NHS-Peg 4 -Tt (25 moles equiv. 4.0 mg in 20 μL of dry DMSO) and gently stirred for one hour (Supplementary Figure S7B). Modified albumin was labeled with fluorophores by treating with NHS-Rhodamine, NHS-AlxaFluor 488, DyLight 800, or NHS-CF-750 (10 moles equiv. in 10 μL of dry DMSO). The final products, Alb(Peg 4 -Tt) 15 (Rhod) 4 , Alb(Peg 4 -Tt) 15 (AF-488) 2 , Alb(Px) 2.6 (Peg 4 -Tt) 15 DL-800) 2 , and Alb(Px) 2.6 (Peg 4 -Tt) 15 (CF-750) 2 , were used for in vitro confocal fluorescence imaging, intravital multiphoton confocal imaging, in vivo imaging and image-guided drug delivery studies, respectively. The molecular masses of intermediates at each step were measured by MALDI-TOF (Supplementary Figure S7C). The products were purified by ultracentrifugation followed by HPLC. For therapeutic experiments, albumin was first conjugated with paclitaxel, followed by functionalization with Peg 4 -Tt and labeling with CF-750 according to the procedure described above. Ultracentrifugation and HPLC purification. Amicon ultra centrifugation filter units (0.5 mL, 3 kDa, and 15 mL, 30 kDa) were used to concentrate samples and to remove the unreacted low-molecular weight reagents after each step of the conjugation reactions. The samples were further purified by a Waters binary pump/dual absorbance HPLC system equipped with a YMC-Pack Diol-300 (300 × 8.0 mm I.D.; particle size, 5 μm; pore size, 30 nm) size exclusion column, using 0.1 M PBS with 0.2 M NaCl (pH 7.2) as the mobile phase. ## MALDI of components. The molecular weights of modified proteins were determined by MALDI-TOF (Mass Spectrometry and Proteomics Facility, The Johns Hopkins University School of Medicine). The DOF of the functional group and the DOC of paclitaxel were calculated based on the change in molecular weights (Supplementary Figure S7C). The DOL of fluorophores was determined following manufacturers' protocols. Cell lines. The HER2-overexpressing BT-474 cell line was purchased from the American Type Culture Collection (ATCC). The cells were grown in 46-X medium supplemented with 10% FBS and 1% penicillin-streptomycin according to the manufacturer's protocol, and maintained at 5% CO 2 in a humidified incubator at 37 °C. Cells were confirmed to be free of mycoplasma infection. ## Human breast cancer mouse models. A pellet of 17β-estradiol (0.72 mg/90 day release, Innovative Research of America) was implanted in the subdermal space of each healthy, four-to-six-week-old, female Nu/Nu mouse. After approximately 24 h, BT-474 cells at 70-80% confluency, in fresh medium for < 24 h, were collected and prepared with 5 × 10 6 in 50 μL of 46-X medium:Matrigel (1:1) for each inoculation and maintained at 4 °C. Cells were orthotopically inoculated into the 2 nd mammary fat-pads. When tumor volume reached 100-150 mm 3 , mice were used for in vivo and intravital imaging and therapeutic experiments. At the end of the experiments, mice were euthanized according to the protocol. All animal experiments were carried out in accordance with protocols approved by the Johns Hopkins University Animal Care and Use Committee, and were conducted in strict compliance with all federal and institutional guidelines. In vitro confocal fluorescence imaging. BT-474 cells at the third or fourth passage (5 × 10 5 cells/well in 0.5 mL of 46X medium) were placed in four-well chamber slides and grown for 24-48 h to 70-80 confluency. Cells were first incubated with Tz(TCO) 6 (AF-488) 4 (20 μg/mL, 130 nM) in PBS+ (PBS supplemented with 0.5% BSA) at room temperature for 20 min. Pre-labeled cells were treated with reactive Alb(Peg 4 -Tt) 15 (Rhod) 4 or non-reactive Alb(Rhod) 4 and incubated at room temperature for 15 min. Treated cells were then incubated in fresh 46-X medium for 3 h at room temperature, or 37 °C, and fixed by 4% PFA in PBS. Cells were counterstained by Hoechst 33342 (1 μg/mL in dH 2 O) and wet-mounted for confocal microscopic imaging on a Zeiss Axiovert 200 system equipped with an LSM 510-Meta confocal module. In vivo Xenogen optical imaging. Mice were injected intravenously with either Tz(TCO) 6 (CF-680) 2 or Tz(CF-680) 2 (0.2 mg in 200 μL of sterile PBS), and imaged using a Xenogen IVIS 200 Optical Imaging system at 5, 15, 30 min, 1, 2, 4, and 8 h post-injection. At 8-10 h post-injection time, there were significant amounts of tumor uptake in the pre-targeting components, while the targeting components were below the detection limit in the systemic circulation. At this time point, mice were injected with a drug-loaded nanocarrier, Alb(Px) 2.6 (Peg 4 -Tt) 15 (DL-800) 2 (2.0 mg in 200 μL of sterile PBS), and we continued to image them to see the tumor uptake of the delivery component. ## Minimally invasive surgery and intravital imaging. Mice were injected intravenously with Tz(TCO) 6 (Rhod) 2 (0.2 mg in 200 μL of sterile PBS). After 8 h, mice were anesthetized with ketamine and acepromazine (i.p. ketamine 100 μg/g body weight; acepromazine 10 μg/g body weight) and placed on the custom-made mouse-holder in the supine position (Fig. 4A). The holder was maintained at 37 °C using a feedback-controlled heating pad. The tumor was disinfected using alcohol-prep and the dorsolateral skin of the tumor was pulled and cut using a forceps and micro surgical scissors. The "U"-shaped skin-flap, with a ~0.5 cm width, was immobilized using stitches. The skin opening treated with saline was covered with a compression plate and imaged through the window on the compression plate, which was parallel to the longitudinal axis. Intravital images were obtained under 10× or 25× magnification using an Olympus FV1000MPE multiphoton laser-scanning microscope. The drug-loaded nanocarrier, Alb(Peg 4 -Tt) 15 (AF-488) 2 or control Alb(AF-488) 2 (2.0 mg in 200 μL of sterile PBS), was injected through a catheter secured to the tail vein and was continuously observed in real-time. Therapeutic procedure. Three groups of female Nu/Nu mice, orthotopically inoculated with BT-474 human breast cancer cells, were administered Tz(TCO) 6 (CF-680) 2 , Tz(CF-680) 2 (both intravenously at a dose of 10 mg/kg) or saline, and were considered click-treated, mock-treated control, and untreated control groups (n = 5 each group), respectively. After 12 h, mice in the click-treated and mock-treated groups were administered Alb(Px) 2.6 (Peg 4 -Tt) 15 (CF-750) 2 at a dose of 25 mg/kg, while mice in the untreated group were given saline as a control. The second therapeutic dose was given after two weeks. The sizes of the tumors were measured with a caliper every fourth day. The tumor volume was calculated using the formula (L × W 2 )π/6, where L is the longest diameter (the major axis) and W is the width dimension, which is perpendicular to the major axis. ## Statistical analysis. The statistical analysis between treated and untreated groups was performed using JMP 12.1.0 Statistical Discovery TM from SAS. The significance of therapeutic effects in each pair was analyzed by the nonparametric, multiple comparisons Wilcoxon each pair test. A p value of less than 0.05 was considered significant (*p < 0.05, **p < 0.005).

chemsum

{"title": "Bioorthogonal two-component drug delivery in HER2(+) breast cancer mouse models", "journal": "Scientific Reports - Nature"}

solution-phase_synthesis_of_the_chalcogenide_perovskite_barium_zirconium_sulfide_as_colloidal_nanoma

## Abstract: Chalcogenide perovskites such as BaZrS3 have promising optoelectronic properties. Methods to produce these materials at low temperatures, especially in the solution phase, are currently scarce. We describe a solution-phase synthesis of colloidal nanoparticles of BaZrS3 using reactive metal amide precursors. The nanomaterials are crystallographically and spectroscopically characterized.The search for efficient and low-cost materials for thin-film photovoltaics has in recent years been dominated by a focus on hybrid lead halide perovskites due to their low cost, facile processing, and high efficiency-exceeding 29% when combined with silicon in a tandem device. 1 However, concerns about the stability, toxicity, and potential environmental effects of these lead-based materials have already begun to drive substantial research efforts towards the development of related materials with higher stability and lower toxicity. 2-4 Among the many proposed materials, chalcogenide perovskites and related materials with general formula ABQ3 (A = Ca, Ba, Sr; B = Zr, Hf, Ti; Q = S, Se) show particularly strong promise based on their optoelectronic properties and excellent stability. [5][6][7] Of these materials, BaZrS3 has drawn the most attention because of its distorted perovskite structure and useful optical properties (Figure 1A). In particular, at ~1.8 eV its bandgap is higher than the ideal for a single-junction solar cell but competitive to replace perovskites in tandem applications; moreover, the bandgap could be lowered closer to the ideal by alloying or the use of related Ruddlesden-Popper phases. 6, Despite the theorized potential of BaZrS3, development and testing of it in thin-film devices has been largely hindered by the lack of low-temperature methods to deposit it as a thin film. 6 The first bulk syntheses of BaZrS3 required temperatures near or above 1000 o C. 11,12 Further tuning of the chemistry and stoichiometry eventually lowered this to 450 o C. 13,14 Initial efforts to generate BaZrS3 thin films have relied on sulfurization of oxide films or co-sputtering approaches, but high temperatures (>900 o C) were usually needed to complete the reaction and achieve crystalline materials. Recently, epitaxial film growth has been achieved at lower temperatures (>700 o C) using pulsed laser deposition. 18 Most promisingly, film growth and crystallization at 600 o C has been achieved using a sputtering/sulfurization approach. 19 Many of these techniques also require more complex and expensive equipment compared to the solution-phase growth and processing that is possible with the hybrid lead halide perovskites. In cases where direct solution-phase growth of a thin film is not feasible, an alternative approach is to use a colloidal suspension of nanocrystals as a precursor ink. 20,21 For this reason, among others, there has been some interest in the preparation of BaZrS3 as colloidal nanomaterials. There has been one report of the successful preparation of colloidal BaZrS3 nanoparticles by grinding bulk BaZrS3 to a fine powder and treating it with appropriate solvents/ligands to extract a population of small (40-60 nm) colloidal particles, which were successfully processed into thin-film devices. 22 However, methods for the direct solution-phase synthesis of BaZrS3 nanomaterials are still lacking. In this communication, we report the synthesis of BaZrS3 as colloidal nanoparticles using reactive metal amide precursors in oleylamine solution, with N,N'-diethylthiourea as the sulfur source, using a procedure adapted from that we previously reported for the synthesis of BaTiS3 nanomaterials. 23 The reaction was successful at temperatures ranging from 365 o C to as low as 275 o C. However, nanomaterials synthesized at the lower temperatures showed structural distortions and a more pronounced platelet-like morphology as compared to those synthesized at the upper end of the temperature range. Briefly, in a typical synthesis, Ba[N(TMS)2]2(THF)2, 24 Zr[N(CH3)2]4, and N,N'diethylthiourea are combined in a 1:2:60 mole ratio in rigorously dried oleylamine, at a concentration of 0.08 M in Ba 2+ (Figure 1B). The reaction is carried out in a Schlenk reaction tube under inert gas using a set-up similar to that we previously reported for the heat-up synthesis of BaTiS3. 25 The reaction mixture is heated to the desired reaction temperature (e.g., 365 o C) and maintained at this temperature for 30 minutes before being allowed to cool to room temperature. During heating, the reaction mixture takes on a deep red-brown color and remains homogeneous in appearance. Following precipitation and washing of the nanomaterials from the reaction solution using anhydrous chloroform and ethanol, BaZrS3 nanoparticles are isolated as an orange-red powder. We found that a large excess of the sulfur precursor and a high concentration in solution were both important for the success of the reaction; otherwise, impurity phases were commonly observed. Additionally, the use of the readily soluble and reactive metal amide precursors greatly facilitated the successful production of crystalline BaZrS3. Attempts to synthesize this material using simple chloride and acetate salts of Ba 2+ and Zr 4+ have thus far been unsuccessful in our hands, as have attempts to use alternative sulfur precursors including CS2 and (Me3Si)2S; in these cases, binary phases such as BaS and ZrS2 are frequently observed, or no detectable crystalline phase is observed at all. Although the synthesis itself must be carried out under rigorously anhydrous conditions due to the reactive nature of the precursors, the resulting nanomaterials are quite stable to air and moisture (vide infra). However, the initial nanoparticle purification must also be carried out under air-and water-free conditions in order to avoid contamination of the sample with amorphous oxide byproducts (e.g. ZrO2). Figure 2 shows characterization data for a sample of nanoparticles obtained at the highest readily accessible temperature-approximately 365 o C, corresponding to vigorously refluxing oleylamine. Figure 2a compares the powder diffraction pattern of the nanoparticles to the predicted pattern based on the reported orthorhombic distorted perovskite structure of bulk BaZrS3; the diffraction peak positions are reasonably well-matched, with the expected sizerelated broadening. 26 TEM imaging shows nano-sized particles which appear to have a platelike morphology and tend to be highly aggregated, making precise size distribution measurements challenging. Typically, the particles are non-uniform and polydisperse with the majority of particles falling within a lateral size range from approximately 10 to 40 nm. Lattice at lower temperatures showed some apparent structural changes (Figure 4); we refer to these material as LT-BaZrS3. We note that this result was also sometimes observed in nanoparticle samples synthesized at 365 o C for reasons that are not currently clear. Powder X-ray diffraction analysis of these nanomaterials showed evidence of deviations from the reported bulk structure, as illustrated in Figure 4A. Key differences include the presence of additional diffracted intensity around 2θ = 34 o and 22.5 o , along with small shifts of the other major diffraction peaks towards higher angles and the appearance of a broad shoulder on the lowangle side of the peak near 44.5 o . The Ruddlesden-Popper Ba3Zr2S7 and Ba2ZrS4 phases do not appear to be a better match to the experimental data (Figure 4A). 27,28 It is difficult to completely rule out that the samples could actually be a mixture of related phases, although they do appear uniform by TEM (Figure 4B-C). TEM imaging (Figure 4B-C) of LT-BaZrS3 shows nanoplatelet-like particles. Elemental analysis (EDX) data is similar to that measured for HT-BaZrS3, although the samples tend to have a higher barium/zirconium ratio. Given the platelet-like morphology of the particles, this discrepancy could be partly accounted for if the lateral surface planes are Ba/S-rich. However, we cannot conclusively determine if this is the case. To better understand the origin of the apparent structural distortions, and to further confirm the identity of the material as a distorted perovskite BaZrS3, preliminary structural refinement was carried out using pair distribution function (PDF) analysis of synchrotron-based X-ray scattering data (see Supporting Information for details and fit parameters). The resulting data and fits are shown in Figure 5. Data from the high-temperature particles (HT-BaZrS3) are shown in Figure 5a, overlaid with a fit to the bulk BaZrS3 structure. During the fitting, refinement of the lattice parameters and isotropic atomic displacement parameters was allowed while other structural parameters were fixed to those of the reported bulk structure. Overall the PDF data appears consistent with a Pnma distorted perovskite BaZrS3 structure. comparison, attempted fits of this PDF data to the reported structures of Ba3Zr2S7 and Ba2ZrS4 are shown in Figure 4C and by visual inspection are less successful than the fit to BaZrS3; these fits also have higher Rw values (see SI). Therefore, the PDF data suggests that these nanoparticles may still possess the perovskite-like structure of bulk BaZrS3 at least on the local level. The exact nature of the structural distortions suggested by PXRD and PDF analysis, and their relationship to the particle morphology and synthesis temperature, is not yet known; investigations into this question are ongoing and will be elaborated in future reports. Figure 6A shows an extinction spectrum for a colloidal solution of the HT-BaZrS3 nanoparticles. The spectrum shows a shallow sloping onset around 750 nm and a steeper slope commencing around 600 nm. A broad shoulder is detectable at 460 nm, and an additional peak at 300 nm. The absorption onset is in line with those reported in the literature for bulk BaZrS3, and is also reflected by the reddish-orange color of the materials (Figure 6B-C). 6 The exact value of the intrinsic band gap for BaZrS3 is not well-established, with reports ranging at least from 1.75 eV to 1.94 eV. 17,29 Unfortunately, we were unable to detect significant luminescence from our materials; it has been noted that the luminescence of BaZrS3 nanomaterials can be highly dependent on surface treatment, so further optimization of the surface termination could elicit emission in the future. Finally, we tested the stability of the nanoparticles to exposure to ambient atmosphere and to direct immersion in water, to determine if the excellent stability observed for bulk BaZrS3 might hold true on the nanoscale (Figure 7). For a sample of HT-BaZrS3, the PXRD pattern did not change significantly over the course of 9 weeks of exposure to ambient atmosphere. Upon immersion in water, a small unidentified impurity peak was observed in the sample after 30 minutes. Therefore, although the stability of these nanomaterials against water is higher than that of lead halide perovskite nanocrystals or BaTiS3 nanocrystals, 25 they may be less resilient than bulk BaZrS3. In conclusion, we have demonstrated that colloidal suspensions of BaZrS3 nanomaterials can be obtained using a low-temperature, solution-phase process. The approaches reported here may be generalizable to other related materials.

chemsum

{"title": "Solution-Phase Synthesis of the Chalcogenide Perovskite Barium Zirconium Sulfide as Colloidal Nanomaterials", "journal": "ChemRxiv"}

excited_state_dynamics_for_visible-light_sensitization_of_a_photochromic_benzil-subsituted_phenoxyl-

## Abstract: Visible-light sensitized photoswitches have been paid particular attention in the fields of life sciences and materials science because long-wavelength light reduces photodegradation, transmits deep inside of matters, and achieves the selective excitation in condensed systems. Among various photoswitch molecules, the phenoxyl-imidazolyl radical complex (PIC) is a recently developed thermally reversible photochromic molecule whose thermal back reaction can be tuned from tens of nanoseconds to tens of seconds by rational design of the molecular structure. While the wide range of tunability of the switching speed of PIC opened up various potential applications, no photosensitivity to visible light limits its applications. In this study, we synthesized a visible-light sensitized PIC derivative conjugated with a benzil unit. Femtosecond transient absorption spectroscopy revealed that the benzil unit acts as a singlet photosensitizer for PIC by the Dexter-type energy transfer. Visible-light sensitized photochromic reactions of PIC are important for expanding the versatility of potential applications to life sciences and materials science. ## Introduction Photochromism, which is defined as the reversible transformation of a chemical species between two structural isomers by light, has been extensively studied over decades . Recently, visible-light sensitized photochromic materials have been paid particular attention in the fields of life sciences and materials science because long-wavelength light reduces photodegradation, transmits deep inside of matters, and achieves the selective excitation in condensed systems . General strategies for the sensitization of the photochromic reactions to visible light are to extend the π-conjugation and to utilize photosensitizers. Especially, triplet photosensitizers, which form the triplet state of a molecule by the triplet-triplet energy transfer, have been frequently used in photoresists, photodynamic therapy, and photocatalysts because the lowest triplet excited (T 1 ) state can be formed by light whose energy is smaller than that of the optically active transition . However, photochromic reactions of some systems do not proceed via the T 1 state. For example, it was reported that the photochromic reaction of hexaarylbiimidazole (HABI), which is a well-known radicaldissociation-type photochromic molecule , is not sensitized by triplet photosensitizers . On the other hand, it was reported that singlet photosensitizers effectively sensitize the photochromic reaction of HABI to the visible light . While the S 0 -S 1 transition of HABI is located at the visiblelight region, the transition is optically forbidden. Therefore, the photochromic reaction of HABI without singlet photosensitizers occurs via the S 0 -S n transition, which is located at the UV region. On the other hand, singlet photosensitizers efficiently transfer the visible-light energy to the optically inactive S 1 state of HABI, and thus the photochromic reaction of HABI proceeds with visible light. The phenoxyl-imidazolyl radical complex (PIC, Scheme 1) is one of the recently developed rate-tunable T-type photochromic compounds which reversibly generate an imidazolyl radical and a phenoxyl radical (biradical form) in a molecule upon UV light irradiation . The great advantage of PIC is the tunability of the thermal back reaction from tens of nanoseconds to tens of seconds by simple and rational molecular design . The wide ranges of thermal back reactions of photoswitches expand the potential applications of photochromic materials such as to dynamic holographic display , switchable fluorescent markers , and anticounterfeit inks. However, PIC is photosensitive only in the UV region, which limits the application fields. It was reported that the S 0 -S 1 transition of PIC is optically forbidden and is located at the visible-light region as similar to that of HABI . It is expected that the photochromic reaction of PIC occurs via the optically forbidden S 1 state as similar to other radical dissociation-type photochromic molecules such as HABI and pentaarylbiimidazole (PABI) . Therefore, if we could substitute a singlet photosensitizer unit to PIC, the visible-light sensitivity could be achieved by singlet-singlet energy transfer. The visible-light sensitization of PIC expands the versatility of the rate-tunable photoswitches of PIC systems. In this study, we synthesized a novel PIC derivative conjugated with a visible-light photosensitizer (Benzil-PIC, Scheme 1) and investigated the excited state dynamics. We used a benzil framework as a photosensitizer unit because aryl ketones have been widely used as visible-light photosensitizers . While most of aryl ketones were used as triplet photosensitizers, the benzil unit in the present study acts as a singlet photosensitizer. The detail of the sensitization processes was investigated by wide ranges of time-resolved spectroscopies. ## Steady-state absorption spectra The synthetic procedure of Benzil-PIC is described in the Experimental part. Benzil-PIC has two structural isomers (isomer A and isomer B) as shown in Scheme 1. These isomers were separated by high-performance liquid chromatography (HPLC), and each isomer was characterized by steady-state absorption spectra and time-dependent density functional theory (TDDFT) calculations as shown below. Figure 1 shows the steady-state absorption spectra of the two isomers of Benzil-PIC and PIC in benzene at 298 K. While the absorption of PIC appears only at wavelength shorter than 350 nm, those of the two isomers of Benzil-PIC are extended to the visible-light region. The simulated absorption spectra by TDDFT calculations (MPW1PW91/6-31+G(d,p)// M05-2X/6-31+G(d,p) level of the theory) are also shown as the vertical lines in Figure 1. The simulated absorption spectra well explain the experimental absorption spectra of the two isomers. Therefore, the absorption spectra of isomers A and B were assigned as shown in Figure 1. The absorption band at 357 nm of isomer A is assigned to the electronic transition from the molecular orbital distributed around the triphenylimidazole unit (highest occupied molecular orbital: HOMO) to that around the benzil unit (the second lowest unoccupied molecular orbital: LUMO+1) (Figure S14, Supporting Information File 1). On the other hand, the absorption band at 375 nm of isomer B is assigned to the electronic transition from the molecular orbital distributed around the triphenylimidazole unit (HOMO) to that around the benzil unit and the phenoxyl unit (mainly the lowest unoccupied molecular orbital: LUMO and LUMO+1, Figure S15, Supporting Information File 1). While the HOMOs of isomer A and isomer B are very similar, the LUMO and LUMO+1 of isomer B are more delocalized than the LUMO+1 of isomer A, suggesting that the LUMO and LUMO+1 levels of isomer B are lower than those of isomer A. This would be the most plausible reason why isomer B has an absorption band at the longer wavelength than isomer A. PIC generates the biradical species upon UV-light irradiation and shows the broad transient absorption spectrum over the visible-to near infrared-light regions. The half-life of the thermal back reaction of the biradical in benzene is 250 ns (the lifetime is 360 ns) at 298 K. To investigate the difference in the photochromic properties between two isomers of Benzil-PIC, we measured the absorption spectra and nanosecond-tomicrosecond transient absorption dynamics of isomer A in benzene upon repeated irradiation of 355 nm nanosecond laser pulses (355 nm, 7 mJ pulse −1 , Figure S8a, Supporting Information File 1). The absorption band at 357 nm of isomer A gradually decreases upon irradiation of the nanosecond laser pulses and the absorption edge alternatively shifts to the longer wavelength. It indicates that the irradiation of the UV pulse induces the photochromic reactions (breaking of the C-N bond) and interconverts between isomer A and isomer B. The system reaches the photostationary state (PPS) within 696 shots of the laser pulses. The ratio of isomer A and isomer B is estimated to be 22:78 by the curve fitting of the absorption spectrum at the PPS with those of pure isomer A and isomer B (Figure S9, Supporting Information File 1). Figure S8b (Supporting Information File 1) shows the nanosecond-to-microsecond transient absorption dynamics of isomer A probed at 650 nm under repeated irradiation with the 355 nm nanosecond laser pulses at 298 K. While the transient absorption dynamics of isomer A accumulated by 8 shots are slightly fluctuated most probably because of the low signal-to-noise ratio, the decay kinetics do not change by repeated irradiation with UV-light pulses. It shows that both isomers generate the same biradical form by UV-light irradiation as shown in Scheme 1, indicating that the excited state dynamics of the two isomers of Benzil-PIC after the bond breaking are identical. Therefore, the mixture solution of the two isomers was used for further time-resolved spectroscopic measurements. ## Nanosecond-to-microsecond transient absorption spectra To investigate the photochromic properties of Benzil-PIC, the nanosecond-to-microsecond transient absorption measurements were conducted by the randomly interleaved pulse train (RIPT) method . Figure 2a shows the transient absorption spectra of Benzil-PIC in benzene (2.9 × 10 −4 M) under argon atmosphere at room temperature excited with a 355 nm picosecond laser pulse (pulse duration = 25 ps, intensity = 30 μJ pulse −1 ). At 0.5 ns after the excitation, two broad transient absorption bands are observed at 660 and <450 nm. The spectral shape is more or less similar to that of the biradical form of PIC , indicating Benzil-PIC generates the biradical by 355 nm light irradiation. The transient absorption spectra gradually decay with a time scale of hundreds of nanoseconds and another absorption band at 580 nm remains after 900 ns. The transient absorption dynamics at 590 nm was fitted with a biexponential decay function and the lifetimes are estimated to be 260 and 820 ns (Figure 2c). On the other hand, while the transient absorption spectra of Benzil-PIC in benzene under air show the same transient absorption spectrum as under argon at 0.5 ns, the transient absorption band at 580 nm is not observed in the time scale of microseconds. The transient absorption dynamics at 590 nm can be fitted with a single exponential decay function and the lifetime is 220 ns (Figure 2d), which is almost identical to that of the fast decay component under argon atmosphere. Because the transient absorption spectrum at 0.5 ns is similar to that of PIC and because the fast decay component does not depend on the molecular oxygen, the fast and slow decay components can be assigned to the biradical form generated by the C-N bond breaking and the T 1 state of Benzil-PIC, respectively. It is worth mentioning that the T 1 state of Benzil-PIC would be formed by some portions of the S 1 of the benzil unit where the energy transfer did not occur to the PIC unit (discussed below). ## Femtosecond-to-nanosecond transient absorption spectra To investigate the sensitization process by the benzil unit of Benzil-PIC in detail, we performed femtosecond transient absorption measurements using a 400 nm excitation pulse. The instrumental response function is ≈170 fs. It is noted that the excitation wavelength for femtosecond transient absorption spectroscopy (400 nm) is slightly different from that for nanosecond transient absorption spectroscopy (355 nm). The difference may affect the ratio of isomer A and isomer B at the photostationary state (PSS) and initial relaxation kinetics at subpicosecond time scales. Benzil was used for a reference sample. Figure 3a shows the time evolution of the transient absorption spectra of benzil in benzene (6.8 × 10 −2 M). At 0.3 ps after the excitation, a transient absorption band is observed at 546 nm. The transient absorption band continuously shifts to 531 nm and a shoulder is observed at 500 nm. It was reported that the spectral shift of the transient absorption spectra of benzil at the sub-picosecond time scale was assigned to the structural change from the skewed structure to the planar structure . Solvent and vibrational relaxations would also take place in this time scale. After the rapid spectral shift, the transient absorption spectra are preserved until 100 ps. This signal can be assigned to the excited state absorption from the lowest vibrational level of the S 1 state. The transient absorption band at 531 nm gradually decreases with a time scale of nanoseconds and another transient absorption band appears at 485 nm. The transient absorption band at 485 nm was assigned to the T 1 state according to previous studies . The quantum yield of the formation of the triplet excited state was reported as 92% , indicating that most of the S 1 state is converted to the T 1 state in benzil. Figure 3b shows the transient absorption spectra of Benzil-PIC in benzene (2.2 × 10 −3 M) excited at 400 nm with a femtosecond laser pulse. The signal around 800 nm was omitted because it was perturbed by the second order diffraction of the excitation pulse around 400 nm. At 0.3 ps after the excitation, two transient absorption bands are observed at 520 and 563 nm, which are most probably assigned to the transient absorption of the benzil unit of Benzil-PIC. The absorption is slightly shifted to the red as compared to those of benzil probably due to the extended π-conjugation of the benzil unit connected to the PIC unit. The two peaks continuously shift to the shorter wavelength (503 and 543 nm, respectively) with a time scale of picoseconds as similar to that of benzil, which supports that these bands are originated from the benzil unit. In addition to the two bands, a broad absorption band over the visible-light region is also observed at 0.3 ps. Because the spectral band shape of this absorption band is similar to that observed in Figure 2, this absorption band is ascribable to the biradical form of PIC, which was directly excited at 400 nm and underwent the rapid radical formation in the sub-picosecond time range. The instantaneous formation of the biradical form under these excitation conditions suggests that a peak at ≈430 nm at 0.3 ps would be most probably assigned to the S 1 state of the PIC unit. In addition to this rapid appearance of the biradical form, the gradual increase of the absorption due to the biradical is observed in picoseconds to tens of picoseconds region, together with the decay of the S 1 state of the benzil unit. This slow process of the biradical formation indicates the energy transfer from the benzil unit to the PIC unit. The amplitude of the increased biradical form with a time scale of tens of picoseconds is larger than the instantaneously generated biradical form at the early time scale, indicating that the energy transfer process is dominant for the photochromic reaction of Benzil-PIC under the excitation with 400 nm. In the nanoseconds time region, the absorption around 580 nm slightly increases with a time scale of nanoseconds. To elucidate the details of the reaction dynamics, we performed global analyses with singular value decomposition (SVD) with the Glotaran program (http://glotaran.org) . We tentatively used the three-state sequential kinetic model for benzil (Equation 1) and the five-state sequential kinetic model for Benzil-PIC (Equation 2) convolved with Gaussian pulse. The detail of the SVD analyses are shown in Supporting Information File 1. (1) (2) The evolution associated spectra (EAS) thus obtained indicate the resolved transient absorption spectra into each component of the kinetic models. Because the time window of our measurements was limited to 2 ns, it was difficult to determine the time constant of nanosecond time scale exactly. Therefore, the lifetimes of the intersystem crossing (ISC) of benzil and the benzil unit of Benzil-PIC were fixed to a reported value of benzil (2.5 ns) . The lifetime of the T 1 state of benzil was fixed to 2.0 μs according to the nanosecond-to-microsecond transient absorption spectroscopy. In the benzil system, time constants of three EAS are revealed to be 420 fs, 2.5 ns (fixed), and 2.0 μs (fixed), respectively (Figure 3c). Each EAS species (A to E in the Equation 1 and Equation 2) is denoted as EAS1 to EAS5 in the order of the time constants as shown in Figure 3c and Figure 3d. The fastest time constant of benzil reflects the structural change from the skewed structure to the planar structure and solvent and vibrational relaxations. However, it should be noted that the lifetime of 420 fs is the apparent lifetime because the conformational change from the skewed to the planar structure at sub-picosecond time scale induces the continuous spectral shift. Because the present SVD global analyses do not consider the continuous spectral shift, it is difficult to extract the exact time constant at the early stage of the transient absorption spectra. The EAS with time constants of 2.5 ns and 2.0 μs are safely assigned to the absorption spectra of the S 1 and the T 1 states, respectively, because of the similarity of the spectra to those reported previously . In the Benzil-PIC system, the time constants of five EAS were obtained to be 160 fs, 1.4 ps, 38 ps, 2.5 ns (fixed), and 240 ns (fixed), respectively (Figure 3d). EAS1 has 4 peaks located at 430, 520, 582, ≈710 nm, respectively. The absorption bands at 430 and ≈710 nm are ascribable to the S 1 state of the PIC unit and the biradical generated instantaneously, respectively. It indicates that the biradical was also formed by the direct excitation of the PIC unit with 400 nm light. The spectral evolution from EAS1 (160 fs, grey line in Figure 3d) to EAS2 (1.4 ps, red line in Figure 3d) shows the C-N bond cleavage of the PIC unit and the spectral shift due to the benzil unit (from 582 nm to 556 nm). In PABI, which is a similar photochromic molecule to PIC, it was reported that the C-N bond fission occurs with the time constant of 140 fs and the broad absorption assigned to the biradical form was formed with a time constant of ≈2 ps . The similarity of the time constant of the bond breaking to that of EAS1 supports that the C-N bond is cleaved by the direct excitation of the PIC unit. The spectral evolution from EAS2 (1.4 ps) to EAS3 (38 ps, green in Figure 3d) shows the continuous spectral shift due to the benzil unit and the increase in the absorption due to the biradical form (660 nm). Because the continuous spectral shift due to the benzil unit is still observed in EAS2 (1.4 ps), it is suggested that the structural change of the benzil unit of Benzil-PIC is somehow slightly decelerated as compared to that of benzil (420 fs). However, it should be mentioned that it was difficult to resolve the structural change of the benzil unit and the formation process of the PIC unit by the present SVD analysis. The spectral evolution from EAS3 (38 ps) to EAS4 (2.5 ns, fixed, blue line in Figure 3d) shows the decay of the S 1 state of the benzil unit and the alternative increase in the biradical form of the PIC unit. This result clearly shows that the energy of the S 1 state of the benzil unit is used for the photochromic reaction of the PIC unit. It is important to note that the S 0 -S 1 transition energy of PIC, which is optically forbidden, was reported to be 2.8 eV (≈440 nm) . These results suggest that the energy transfer occurs from the S 1 state of the benzil unit to the ground state of the PIC unit with the time constant of 38 ps. Since the bond-breaking process from the S 1 state of the PIC unit would be much faster than this time scale (hundreds of femtoseconds), the time constant of 38 ps reflects the singlet-singlet energy transfer process from the benzil unit to the PIC unit. It should be noted that the fluorescence quantum yield of benzil was quite low (<0.001) and the PIC unit has no absorption in the emission wavelength of the benzil. Accordingly, the effective energy transfer by the Förster mechanism is not plausible. The energy transfer of the 38 ps time constant is probably due to the Dexter mechanism at weak or very weak coupling regimes owing to the overlap of the wave functions of the benzil and the PIC units in the excited state. The spectral evolution from EAS4 (2.5 ns, fixed) to EAS5 (240 ns, fixed, purple line in Figure 3d) shows the increase in the absorption around 580 nm. Although both lifetimes of EAS4 and EAS5 are longer than the measured time window (≈2 ns), the spectral difference around 580 nm at 10-100 ps and that at nanosecond time scales enable to resolve these spectra. The increased absorption band is similar to the transient absorption band assigned to the T 1 state of Benzil-PIC (Figure 2a). It indicates that the spectral evolution over nanosecond time scale is ascribable to ISC of the benzil unit. It should be noted, however, that this slow rise of the T 1 state of the benzil unit by ISC indicates that some portions of the benzil unit do not undergo the effective energy transfer to the PIC unit because the S 1 state of the benzil was deactivated with the time constant of 38 ps. Although the clear mechanism is not yet elucidated at the present stage of the investigation, the reason for the two relaxation pathways (energy transfer and ISC) from the S 1 state of the benzil unit of Benzil-PIC might be due to the difference in the mutual orientation of benzil and PIC units including the structural isomers (isomer A and isomer B). As was discussed above, the energy transfer is due to the overlap of the wave function of the both units, of which mechanism might be sensitive to the difference in the mutual orientation. ## Effect of triplet-triplet energy transfer Ultrafast spectroscopy revealed that the benzil unit acts as a singlet photosensitizer for Benzil-PIC by the Dexter-type energy transfer. It was reported that benzil was often used as a triplet photosensitizer because the quantum yield for the T 1 state formation is 92% . To investigate the possibility for the triplet-triplet energy transfer process in Benzil-PIC, we performed two experiments. Firstly, we measured the phosphorescence spectra of benzil and PIC in EPA (diethyl ether/isopentane/ethanol 5:5:2) at low temperature to estimate the energy levels of the T 1 states of benzil and PIC. In the conventional emission measurement setups at low temperature, both fluorescence and phosphorescence are observed upon irradiation of excitation light. To extract the phosphorescence spectra, the excitation light (continuous wave laser, 355 nm, 1 mW) was chopped at 1 Hz and the afterglow emission under blocking the beam was accumulated as the phosphorescence spectra. Figure 4 shows the phosphorescence spectra of benzil in EPA at 77 and 100 K. While the phosphorescence spectrum of benzil at 77 K is broad and observed at 500 nm, that at 100 K becomes sharper and the peak is shifted to 567 nm with a vibrational fine structure at 625 nm. The spectral shift with the increase in temperature is most probably due to the rigidity of the environment of molecules. At 77 K, it is expected that the solvent is too rigid for benzil to change the conformation in the excited state, namely, the conformation of benzil is fixed to the skewed conformation. On the other hand, it is expected that the increase in the temperature to 100 K softens the rigid matrix and allows the benzil to form the planar conformation at the T 1 state. The energy level of the T 1 state of benzil was estimated from the phosphorescence at 100 K because the T 1 state of benzil in solution forms the planar conformation. The energy level of the T 1 state was determined by an edge of the high energy side of the phosphorescence, where a tangent line crosses the x-axis. The energy level of the T 1 state of benzil is estimated to be 53 kcal mol −1 , which is consistent with a reported value (53.7 kcal mol −1 ) . On the other hand, the phosphorescence of PIC was only observed at 77 K and the signal is very weak. Because the conformation of PIC is relatively rigid, we tentatively estimated the T 1 state energy level from the phosphorescence at 77 K. The T 1 state energy level of PIC is estimated to be 63 kcal mol −1 . It suggests that the T 1 state energy level of benzil is slightly lower than that of PIC. Moreover, the triplet photosensitization was examined by the microsecond transient absorption measurements of the mixture solution of benzil and PIC in benzene (3.7 × 10 −3 M and 2.8 × 10 −5 M for benzil and PIC, respectively). A 450 nm excitation pulse was used to selectively excite benzil. The transient absorption dynamics of the mixture solution of benzil and PIC probed at 500 nm is identical to that of benzil, which is assigned to the T 1 state (Figure S13, Supporting Information File 1). It indicates that the triplet-triplet energy transfer is negligible between the benzil and PIC units. The plausible reason for the negligible triplet-triplet energy transfer is the lower energy level of the T 1 state of the benzil unit than that of the PIC unit. While PIC absorbs light of wavelength only shorter than 350 nm, the introduction of the benzil unit extends the photosensitivity of the photochromic reaction to the visible-light region. When Benzil-PIC absorbs visible light, the conformation of the benzil unit, which is the skewed structure in the ground state, quickly changes to the planar structure with a time scale of picoseconds and the S 1 state of the benzil is formed. While the photochromic reaction partly proceeds via the direct excitation of the PIC unit, most of the photochromic reaction is induced via the Dexter-type singlet-singlet energy transfer from the benzil to the PIC units with the time constant of 38 ps. The triplet photosensitization does not occur in Benzil-PIC most probably because the triplet energy level of the PIC unit is higher than that of the benzil unit. The clarification of the visible-light sensitization mechanism of PIC is important for expanding the versatility of potential applications of PIC in life and materials sciences. ## Experimental Synthetic procedures All reactions were monitored by thin-layer chromatography carried out on 0.2 mm E. Merck silica gel plates (60F-254). Column chromatography was performed on silica gel (Silica Gel 60N (spherical, neutral), 40-50 μm, Kanto Chemical Co., Inc.). 1 H NMR spectra were recorded at 400 MHz on a Bruker AVANCE III 400 NanoBay. DMSO-d 6 and CDCl 3 were used as deuterated solvents. Mass spectra (ESI-TOF-MS) were measured by using a Bruker micrOTOFII-AGA1. All reagents were purchased from TCI, Wako Co. Ltd., Aldrich Chemical Company, Inc. and Kanto Chemical Co., Inc., and were used without further purification. The synthetic procedure of Benzil-PIC is shown in Scheme 2. The synthetic procedure is analogous to that of PIC . ## 4'-Hydroxy-[1,1'-biphenyl]-2-carbaldehyde (1) Compound 1 was prepared according to a literature procedure . ## 1-(4-(2-(4'-Hydroxy 4'-Hydroxy-[1,1'-biphenyl]-2-carbaldehyde (1, 0.088 g, 0.44 mmol), 1,4-bisbenzil (0.176 g, 0.51 mmol) and ammonium acetate (0.240 g, 3.12 mmol) were stirred at 110 °C in acetic acid (2.7 mL) for 6 h. The reaction mixture was cooled and neutralized by aqueous NH 3 . The precipitate was filtered and Scheme 2: Synthetic procedure of Benzil-PIC (analogous to synthesis of PIC in ). washed with water. The crude product was purified by silica gel column chromatography (CH 2 Cl 2 /AcOEt 20:1 to 3:1), to give the desired product as a mixture of two structural isomers as a yellow solid, 0.0674 g (0.130 mmol, 29%). 1 ## Benzil-PIC A solution of potassium ferricyanide (0.968 g, 2.94 mmol) and KOH (0.741 g, 13.2 mmol) in water (3.3 mL) was added to a suspension of 2 (70 mg, 0.14 mmol) in benzene (7.3 mL). After stirring for 3 h at room temperature, the resultant mixture was then extracted with benzene and the organic extract was washed with water and brine. After removal of solvents, the crude product was purified by silica gel column chromatography (AcOEt/ hexane 2:3) to give the desired product as a yellow powder, 42 mg (0.081 mmol, 58%). Two structural isomers were separated by HPLC (eluent: CH 3 CN/H 2 O 7:3). 1 ## Steady-state measurements Steady-state absorption spectra were measured with an UV-3600 Plus (SHIMADZU) at room temperature with 1 cm quartz cuvette. Phosphorescence spectra were measured by home-build millisecond time-resolved emission spectrometer at 77 K with nitrogen cryostat (OptistatDN2, Oxford instruments). Briefly, the cooled samples in EPA (diethyl ether/isopentane/ ethanol 5:5:2) under argon atmosphere were excited with a 355-nm continuous wave (CW) laser (Genesis CX355 100SLM AO, Coherent) and the emission was detected by EMCCD (Newton DU920P-OE, Andor Technology). The excitation light was blocked with 1 Hz by an optical shutter (76992 and 6995, ORIEL) and the time evolution of the emission spectra was measured to separate the fluorescence and phosphorescence. The shutter was controlled by LabVIEW. ## Nanosecond transient absorption measurements The laser flash photolysis experiments were carried out with a TSP-2000 time resolved spectrophotometer system (Unisoku Co., Ltd.). A 10 Hz Q-switched Nd:YAG laser (Continuum Minilite II) with the third harmonic at 355 nm (pulse width, 5 ns) was employed for the excitation light and the photodiode array was used for a detector. Transient absorption measurements on the nanosecond to microsecond time scale were conducted by the randomly interleaved pulse train (RIPT) method . A picosecond laser, PL2210A (EKSPLA, 1 kHz, 25 ps, 30 μJ pulse −1 for 355 nm), and a supercontinuum (SC) radiation source (SC-450, Fianium, 20 MHz, pulse width: 50-100 ps depending on the wavelength, 450-2000 nm) were employed as the pump-pulse and probe sources, respectively. A 355 nm laser pulse was used to excite the samples. The measurements were performed in a benzene solution placed in a 2 mm quartz cell under stirring at room temperature. We used the mixture solution of isomer A and isomer B as was obtained by the synthesis and irradiated a 355 nm pulse laser during the measurements. By considering the duration of the measurements (usually it takes one hour) and the total photon numbers, the system probably reaches the PSS. The ratio of isomer A and isomer B at the PSS upon excitation with the 355 nm pulse is 22:78. ## Femtosecond transient absorption measurements Transient absorption spectra in the visible-light region were measured using a home-built setup. The overall setup was driven by a Ti:Sapphire regenerative amplifier (Spitfire, Spectra-Physics, 802 nm, 1 W, 1 kHz, 100 fs) seeded by a Ti:Sapphire oscillator (Tsunami, Spectra-Physics, 802 nm, 820 mW, 80 MHz, 100 fs). The output of the amplifier was equally divided into two portions. The first one was frequencydoubled with a 50 μm β-barium borate (BBO) crystal, and the generated second harmonics was used for excitation of the sample. The second portion was introduced into a collinear optical parametric amplifier (OPA, TOPAS-Prime, Light Conversion) and converted into the infrared pulse at 1180 nm. This 1180 nm pulse was focused into a 2 mm CaF 2 plate after passing through a delay stage, so as to generate femtosecond white light continuum for the probe pulse. The probe pulse was divided into signal and reference pulses. The signal pulse was guided into the sample and then the both pulses were detected using a pair of multichannel photodiode array (PMA-10, Hamamatsu). The chirping of the white light continuum was evaluated by an optical Kerr effect of carbon tetrachloride and used for the corrections of the spectra. The FWHM of the cross correlation between the excitation and probe pulses was ca. 170 fs. The polarization of the excitation pulse was set to the magic angle with respect to that of the probe pulse. The typical excitation power was 100 nJ pulse −1 at the sample position. During the measure-ment, the sample solution was circulated with a home-made rotation cell with 1 mm optical length. Steady-state absorption spectra were recorded before and after the transient absorption measurement to examine photodegradation of the sample and no permanent change in absorbance was observed. We used the mixture solution of isomer A and isomer B as was obtained by the synthesis and irradiated a 400 nm pulse laser during the measurements. By considering the duration of the measurements (usually takes several hours), the system probably reaches the PSS. Under the irradiation of the 400 nm laser, the ratio of isomer A and isomer B at the PSS depends on each absorption coefficients and the efficiency for the bond cleavage. The absorption coefficients of isomer A and isomer B at 400 nm are 2.1 × 10 3 M −1 cm −1 and 4.1 × 10 3 M −1 cm −1 , respectively.

chemsum

{"title": "Excited state dynamics for visible-light sensitization of a photochromic benzil-subsituted phenoxyl-imidazolyl radical complex", "journal": "Beilstein"}

direct_catalytic_conversion_of_cellulose_to_liquid_straight-chain_alkanes

## Abstract: High yields of liquid straight-chain alkanes were obtained directly from cellulosic feedstock in a one-pot biphasic catalytic system. The catalytic reaction proceeds at elevated temperatures under hydrogen pressure in the presence of tungstosilicic acid, dissolved in the aqueous phase, and modified Ru/C, suspended in the organic phase. Tungstosilicic acid is primarily responsible for cellulose hydrolysis and dehydration steps, while the modified Ru/C selectively hydrogenates intermediates en route to the liquid alkanes. Under optimal conditions, microcrystalline cellulose is converted to 82% n-decane-soluble products, mainly n-hexane, within a few hours, with a minimum formation of gaseous and char products. The dominant route to the liquid alkanes proceeds via 5-hydroxymethylfurfural (HMF), whereas the more common pathway via sorbitol appears to be less efficient. High liquid alkane yields were possible through (i) selective conversion of cellulose to glucose and further to HMF by gradually heating the reactor, (ii) a proper hydrothermal modification of commercial Ru/C to tune its chemoselectivity to furan hydrogenation rather than glucose hydrogenation, and (iii) the use of a biphasic reaction system with optimal partitioning of the intermediates and catalytic reactions. The catalytic system is capable of converting subsequent batches of fresh cellulose, enabling accumulation of the liquid alkanes in the organic phase during subsequent runs. Its robustness is illustrated in the conversion of the raw (soft)wood sawdust. Broader contextA novel one-pot catalytic approach is presented that is able to directly transform cellulose into straight-chain alkanes (mainly n-hexane). The carbon-based yields are high (up to 82%) and the process completes in less than 6 hours at only 493 K. The so produced and thus bio-derived light naphta fraction is an ideal green feedstock for existing processes that produce aromatics, gasoline or olens. Considering the vast and cheap amounts of cellulosic residue and the absence of its pretreatment for this process, this catalytic one-pot approach seems highly promising en route to more sustainable chemicals and fuels. ## Introduction Interest in lignocellulosic biomass as a renewable feedstock for fuels, chemicals and materials has increased tremendously in recent years. The high oxygen-to-carbon ratio of cellulosic biomass creates ample opportunities to produce chemicals and polymer building blocks with high chemical functionality, which cannot be produced as cheaply from fossil feedstock. 5-hydroxymethylfurfural (HMF), 18 (vinyl) glycolic acid, 19,20 lactic acid 21,22 and levulinic acid 23,24 are four examples of such chemicals, for which synthesis directly from cellulose is under investigation. Targeting fuels with biomass feedstock primarily concentrates on depolymerization and defunctionalization strategies to produce molecules with high heating value like alkanes and aromatics. As the value of a fuel per tonnage is usually low, but the targeted volumes are enormous, process and energy cost should be kept to an absolute minimum. There are elaborate examples in literature describing the production of new generation biofuels from sugars, sugar alcohols or other platform molecules such as HMF and levulinic acid, 3,10,25, but research on the direct route from low cost cellulose to alkanes is still in its infancy. Although high temperature hydropyrolytic routes from biomass towards mixtures of gasoline and other compounds are promising, there is room to improve the carbon efficiency to liquid alkanes. Due to its high natural abundance 50 and uniform chemical structure with repeating C 6 sugar units, cellulose should be the ideal precursor for selectively making C 6 alkanes (and thus light naphtha) as C-C bond breaking and forming are not required. The main challenge is to selectively break C-O in presence of C-C bonds. The prospect of renewable n-hexane directly made from cellulose, as outlined in Scheme 1, is exciting as this alkane has many uses such as technical solvent, 51 fuels and building block for chemicals. For use as a transportation fuel, viz. gasoline, nhexane needs to be isomerized to branched hexanes with higher motor octane numbers (MON) like 2,2-dimethylbutane (MON 95). 52 It is well known that, unlike C 7+ alkanes, light (C 4 , C 5 and C 6 ) alkanes can be selectively isomerized with minimal cracking (e.g., in the Hysomer Process of Shell), and mixed in with gasoline. 52,53 Since highly branched alkanes, possibly mixed with some ethers, constitute the environmentally most friendly gasoline, 52 bio-based isomerized light naphtha (with or without ethers) may be an interesting option to improve the renewability of gasoline in short term. Besides fuel and solvent use, n-hexane may also serve as ideal feedstock for bio-benzene production 54 and for bio-ethylene and propylene production via steam or catalytic cracking. 55 During submission of this manuscript, the group of Tomishige reported the frst selective one-pot conversion of cellulose to n-hexane using Ir-ReO x /SiO 2 and H-ZSM-5. 56 The reaction proceeds via the hydrolytic hydrogenation of cellulose to sorbitol, which is subsequently converted to n-hexane through consecutive hydrodeoxygenation cycles. Apart from this report, only multistep processes 57 and the conversion of cellobiose 36 (94.8% n-hexane yield) and methylcellulose 58 into nhexane (80% total yield) were demonstrated. The major obstacles for direct cellulose conversion are its poor solubility in conventional solvents and high chemical recalcitrance. 59 These issues necessitate severe reaction conditions in terms of acidity and/or temperature, which can lead to unwanted side reactions. This paper reports a direct, fast and selective conversion of cellulose into liquid straight-chain alkanes, mainly n-hexane, by tuning the hydrogenation selectivity of a commercial Ru catalyst in a biphasic liquid system. The surface modifcation steers the reaction via a novel pathway, forming liquid alkanes through intermediate HMF (see Scheme 2). ## Experimental A typical modifcation of commercial 5 wt% Ru/C proceeded as follows: Ru/C (1 g), tungstosilicic acid (TSA) hydrate (0.25 g) and water (40 ml) were loaded into a 100 ml stainless steel batch reactor (Parr Instruments Co.). The reactor was flushed with N 2 and subsequently pressurized with 5 MPa H 2 . The mixture was stirred at 700 rpm and heated to 483 K at an average rate of 10 K min 1 and kept at this temperature for 1 h. The reactor was then cooled, depressurized and opened. The synthesized catalyst (htTSA(2)Ru/C) was fltered, thoroughly washed with distilled water and dried to constant weight. In a typical catalytic experiment, microcrystalline Avicel PH-101 cellulose (2 g), TSA hydrate (5 g), htTSA(2)Ru/C (0.5 g), water (20 ml) and n-decane (20 ml) were loaded into a 100 ml stainless steel batch reactor (Parr Instruments Co.). The reactor was flushed with N 2 and subsequently pressurized with 5 MPa H 2 . The mixture was stirred at 700 rpm and heated to 493 K at an average rate of 12 K min 1 from room temperature to 423 K and further to 493 K at a fxed rate of 0.5 K min 1 . The mixture was kept at 493 K for an additional 40 min. After reaction, the reactor was cooled, depressurized and opened. Samples were taken from both the water and n-decane phases and centrifuged before GC and TOC analysis. For determination of cellulose conversion and catalyst reuse experiments, centrifuged particles were added back to the reaction mixture. The reaction mixture was subsequently fltered, thoroughly washed and dried to constant weight. Complete experimental procedures are provided in the ESI. † ## Rationale From literature on n-hexane production from sugar and sugarderived feedstock, 31,32 and the recent one-pot approach from Tomishige and co-workers, 56 one may deduce one major pathway, which proceeds via a combination of various reactions including hydrolysis, hydrogenation, dehydration and hydrodeoxygenation (HDO), catalyzed by bifunctional acid/redox catalytic systems (Scheme 2). There is general agreement that the route involves the initial formation of sorbitol as key intermediate towards n-hexane. 32,36,60,61 However, we propose here an alternative pathway that runs through HMF. Deep HDO of HMF to n-hexane, e.g. by direct metal-catalyzed C-O hydrogenolysis or acid/metal-catalyzed dehydration/hydrogenation cycles, 60,62 has not been demonstrated experimentally. The use of cellulose rather than sugar solutions signifcantly complicates the balance of reaction rates required for selective n-hexane formation. As fast cellulose hydrolysis generally requires strong acidic conditions or high temperatures, sorbitol produced from glucose may undergo rapid dehydration to sorbitan and isosorbide. As a remarkable stability of isosorbide in the presence of acid/redox catalysts at high temperatures was encountered, 63 isosorbide formation may be a signifcant hurdle for the low energy conversion of cellulose to n-hexane. Isosorbide formation can be prevented as long as glucose dehydration to HMF is kinetically favored over glucose hydrogenation to sorbitol. Additionally, subsequent and fast hydrogenation of HMF to e.g., 2,5-dihydroxymethylfuran (2,5-DHMF, Scheme 2) and 2,5-dihydroxymethyltetrahydrofuran (2,5-DHMTHF) should be promoted to avoid HMF degradation into levulinic acid and humin (char). From that point on, a series of HDO cycles of the furanic species should ensue. A recent example of Buntara et al., showing a selective conversion of 2,5-DHMTHF to 1,6-hexanediol, underscores the potential of an HMF route that ultimately leads to n-hexane. 64 The main challenge seems the integration of the acidic hydrolysis of cellulose with a selective hydro(deoxy)genation of acid-sensitive HMF in presence of glucose. We anticipate that these demands can be fulflled by (i) compartmentalization of the acidity in an aqueous phase and redox activity in an organic phase and (ii) modifcation of the redox catalyst to increase its selectivity towards HMF hydrogenation instead of glucose hydrogenation (vide infra). The biphasic system is essential to extract acid-sensitive intermediates from the acidic aqueous phase into the organic phase, while the organic phase should be a favorable medium for hydrogenation and dehydration reactions due to a higher hydrogen solubility and more efficient dehydration in organic solvents, respectively. The important role of catalysis at water-oil interfaces in biomass conversion has been suggested in other work as well. 65,66 The benefcial effect of a biphasic solvent system on the selective hydrogenation of HMF to 2,5-DHMTHF for instance has been investigated and confrmed by Alamillo et al. 67 and Yang et al. 68 Furthermore, glucose hydrogenation is suppressed in favor of HMF hydrogenation by the use of a hydrophobic hydrogenation catalyst, which predominantly resides in the organic phase. Since HMF traverses phase boundaries, the latter will result in selective hydrogenation of HMF, avoiding sorbitol formation. ## Tuning the hydrogenation properties of Ru/C Since Ru/C is commercially used to hydrogenate glucose to sorbitol, 69 it seems at frst sight an unlikely catalyst choice, but its selection here was primarily based on its high affinity for the organic phase (see ESI, Fig. S1 †) and its commercial relevance. In order to suppress its glucose hydrogenation ability and favor HMF hydrogenation, Ru/C was modifed. Modifying the chemoselectivity of metal redox catalysts is usually accomplished by adding promoters. 70 Below, we show that the hydrogenation selectivity of commercial Ru/C is drastically changed in favor of HMF hydrogenation by hydrothermal treatment (ht) in presence of tungstosilicic acid (TSA, H 4 SiW 12 O 40 ). The modifcation was carried out under H 2 pressure (5 MPa at room temperature) at 483 K for 1 h in water in presence of varying TSA concentrations. Despite the harsh treatment, we barely noticed Ru leaching during the hot water treatment in presence of TSA: elemental analysis of the fltrate demonstrated the presence of 2.5 ppm Ru, corresponding to 0.3 wt% of the initial Ru content. The change in hydrogenation selectivity was evidenced in a kinetic study. A frst series of experiments with glucose was carried out in water in presence of unmodifed Ru/C, ht-treated Ru/C and Ru/C ht-treated in a 2 and 135 mM TSA solution, denoted as Ru/C, htRu/C, htTSA(2)Ru/C and htTSA(135)Ru/C, respectively. The TSA loading on Ru/C after drying, studied by gravimetric analysis, correlates to the TSA concentration in the pretreatment mixture (Table S1 †), in agreement with the strong adsorption of heteropoly acids on carbon supports. The fnal TSA modifed catalysts htTSA(2)Ru/C and htTSA(135)Ru/C contain approximately 9 and 27 wt% TSA (on dry basis), respectively. The kinetic profles are presented in Fig. 1. As expected, glucose is selectively converted to sorbitol by each catalyst. The data show a signifcant decrease in activity after modifcation with TSA, and this decrease is more pronounced with the higher TSA loading. Comparison of the initial conversion rates of htTSA(2)Ru/C and htTSA(135)Ru/C versus pristine Ru/C showed a remarkable three-and six-fold activity loss, respectively, whereas a hydrothermal treatment in absence of TSA only shows a minor impact on the hydrogenation activity. A similar set of kinetic experiments was carried out for the hydrogenation of HMF (Fig. 2). All reactions formed 2,5-DHMF as main product. Interestingly, the hydrothermal modifcation of Ru/C both with and without TSA results in a hydrogenation activity increase. Although the fundamentals behind the selective modifcation of Ru/C with TSA are unclear, CO chemisorption (see ESI †) showed a decreased number of total active sites upon modifcation and this decrease correlates linearly with the initial glucose hydrogenation activity of the different catalysts (Fig. S2a †). Unless CO is selectively probing glucose adsorption sites, this observation is indicative of a structural change of Ru e.g. Ru sintering. Calculation of the turnover frequency (TOF, s 1 ) shows that modifcation with TSA has little impact on glucose hydrogenation, while there is a signifcant increase in the TOF (s 1 ) (calculated as mol converted HMF per mol surface Ru per second) of HMF hydrogenation (Fig. S2b †). The modifed Ru surface thus seems to beneft the planar adsorption of HMF, with strongly adsorbed C]C and parallel-oriented C]O bonds, 62 likely on atomically smoother Ru surfaces of the sintered Ru. Investigation of the physicochemical properties of the TSA-modifed Ru/C catalyst is ongoing. ## Exploring cellulose to liquid alkanes conversion with modied Ru/C The modifed htTSA(135)Ru/C catalyst, with its altered hydrogenation selectivity, was used to explore the one-pot conversion of cellulose to liquid alkanes. A biphasic water/n-decane (50 : 50 vol%) solvent mixture was initially chosen. Microcrystalline cellulose was used and its conversion to liquid alkanes was initially tested with htTSA(135)Ru/C at temperatures ranging from 483 K to 503 K at 5 MPa H 2 pressure. The reaction uses an additional amount of water-soluble TSA catalyst to accelerate cellulose hydrolysis. Unlike most inorganic solids like alumina and silica/alumina, TSA is a strong Brønsted acid, and most importantly, it shows a high selectivity to glucose during cellulose hydrolysis. 63,75, The catalytic results are summarized in Table 1. The table includes the main reaction products found: n-hexane, methylcyclopentane (MCP), n-pentane, 2,5-DMTHF, 1-hexanol and some hexitols (sorbitol, mannitol and their anhydrides like sorbitan and isosorbide). A frst experiment, in presence of 135 mM of soluble TSA and htTSA(135)Ru/C (Table 1, entry 1), showed appreciable yields of n-decane-soluble products (42%, including 22% n-hexane yield), while only 6% hexitol yield was obtained in one hour at 483 K, reached by rapidly heating the reactor (see conditions in Table 1). This frst result validates the concept of directly converting cellulose to liquid alkanes in the biphasic liquid conditions using hydrothermally TSA-treated Ru/C. The 41% carbon defcit indicates signifcant losses in form of gaseous and insoluble polymeric products. To minimize these side reactions, a slower heating rate of 0.5 K min 1 instead of 5.5 K min 1 was applied from 423 K onward (Table 1, entry 2). This stepwise heating protocol resulted in a notable yield increase of n-decane-soluble products to 60%, including 34% n-hexane and a 78% yield of identifed liquid phase products. Insignifcant amounts of gaseous products were detected in this experiment (mainly methane, see ESI †). TSA in the aqueous phase plays a key role in the conversion of cellulose to liquid alkanes. As expected, low yields of liquid phase products are observed in absence of soluble TSA (Table 1, entry 6), since the acid is responsible for cellulose hydrolysis. An increase of TSA concentration from 15 to 135 mM considerably enhances the total liquid alkane yield, mainly at the expense of oxygenates like DMTHF and 1-hexanol (Table 1, entries 2-5). This yield increase is in line with the strong dehydration property of TSA, required to efficiently carry out series of bifunctional HDO reactions. Reactions with only TSA and no hydrogenation catalyst should obviously be avoided, as it leads to pronounced char formation. Interestingly, reactions at higher temperatures require less acid (Table 1, compare entries 3, 7 and 8). 83 For instance, by increasing the reaction temperature to 493 K, 71 mM TSA is sufficient to completely convert cellulose to 65% n-decanesoluble products of which more than half is n-hexane. The total product yield from both liquid phases accounts for 80% of the carbon balance. Based on the amount of TSA in the aqueous phase, and assuming cellulose hydrolysis, various acid-catalyzed rearrangements and dehydration steps to break C-O bonds en route to n-hexane, a catalytic turnover of about 12 can be estimated for each proton, showing a catalytic contribution of TSA. The effect of modifying Ru/C with TSA, as predicted in the rationale of this contribution, is apparent from the experiments in Table 2. Unmodifed Ru/C (Entry 8) led to signifcantly less alkane formation, while the hexitol yield considerably increased (from about 9 to 29%). Main compound in the hexitol fraction is isosorbide (with 18%), followed by sorbitan, isoidide and isomannide. This difference in product distribution is in line with the well-known glucose hydrogenation ability of commercial Ru/C. Hydrothermal treatment in absence of TSA (htRu/C, entry 7) partly decreases the hexitol fraction, but this decrease is not as efficient as with the TSA-modifed Ru/C catalysts. Entries 1 to 6 illustrate the catalytic results with different htTSA(x)Ru/C catalysts, where x represents the TSA concentration during hydrothermal pretreatment (ranging from 2 to 135 mM). The highest carbon efficiency and liquid alkane yield, viz. 90% and ## View Article Online 60% respectively, were obtained with the lowest TSA modifcation (entry 6). ## Reaction network study The previous data displayed a wealth of intermediates and endproducts, with the liquid alkanes being the desired ones in this work. To gain more insight into the reaction network, a systematic catalytic study was carried out by feeding the major reaction intermediates into the reactor under identical conditions. The data are collected in Table 3. In contrast to previously reported pathways to alkanes, 32,60,84,85 sorbitol and isosorbide turned out to be fairly unreactive (see Table 3, entries 1-4): only 30% of sorbitol carbon (or 14% in fed-batch mode) and 19% of isosorbide carbon were converted into n-decane-soluble products. Sorbitol was mostly dehydrated to isosorbide (here also referred to as 'hexitol'). These observations confrm a kinetically less favorable route from cellulose to alkanes via sorbitol. In line with our hypothesis, it predicts that hydrogenation of glucose should be slow compared to its dehydration to HMF in order to avoid yield loss to hexitols and their anhydrides. Interestingly, modifcation of Ru/C with TSA fulflls this particular role. Selected key intermediates, which were analyzed in the previous experiments and are likely involved in the alternative HMF route, are glucose, fructose, HMF, 2,5-DMTHF, 2,5-hexanedione, 2,5-hexanediol, 1,2-hexanediol, 2-hexanol and 1hexanol. Scheme 3 collects these chemicals in a tentative reaction network. Reactions with these molecules are presented in Table 3 (entries 5-18). Before giving a detailed description of the data and a network analysis, it can already be concluded from the data that high yields of n-decane soluble products were attained from all the above mentioned molecules, validating the proposed n-hexane pathway via HMF. Conversion of glucose and fructose with htTSA(2)Ru/C in presence of 71 mM TSA yields an insignifcant amount of hexitols (in line with Fig. 1), while the alkane yield is highest for glucose (Table 3, entries 5-8). The formation of a signifcant amount of n-pentane is apparent. A fed-batch approach was used next to the batch reactions to imitate the gradual release of glucose from cellulose, as was done successfully in the direct conversion of glucose to ethylene glycol. 86,87 The product distribution indeed changed with reactor type, the batch reactor systematically leading to higher alkane yields, in agreement with the preferred high contact time to form the alkane end-products. Indeed, test reactions with n-hexane and n-decane showed negligible conversion (data not shown), while longer reaction times provides higher liquid alkane yields, as will be demonstrated below. Study of the HMF conversion, although the key molecule in the new n-hexane pathway, was somewhat problematic in batch mode due to its high reactivity (Table 3, entry 10). A low content of n-decane soluble products was observed and large quantities of polymers (char) and degradation products like levulinic acid were noticed. To obtain high liquid alkane yield from cellulose, gradual formation of HMF and subsequently fast HMF hydrogenation is thus important. Fed batch conversion of HMF is more efficient, yielding 22% alkanes and 23% 2,5-DMTHF (Table 3, entry 9), which will be further converted into alkanes upon longer reaction time. Indeed, reaction with 2,5-DMTHF shows an almost quantitative conversion to n-hexane, in agreement with our proposed HMF pathway (Table 3, entry 11). Presence of acidity in the aqueous phase is essential for the latter reaction. A similar reaction without TSA in the aqueous phase, results into a low 2,5-DMTHF conversion and showed additional formation of 2-hexanol (Table 3, entry 12). The acidcatalyzed ether bond hydrolysis of 2,5-DMTHF, likely frst to 2,5hexanediol, followed by a dehydration/hydrogenation to 2-hexanol, is thus an essential step en route to n-hexane (Table 3, entry 13 and Scheme 3). A suitable amount of acidity in the aqueous phase is thus crucial not only to hydrolyze cellulose to glucose and to dehydrate glucose to HMF, but also to achieve fast ring-opening hydrolysis of 2,5-DMTHF to 2,5-hexanediol. 2,5-Hexanediol is indeed very selectively converted to n-hexane, as demonstrated in entry 13 of Table 3. 2,5-Hexanedione was also occasionally analyzed in the cellulose experiments, especially at low contact time (see later). As this dione is reported to result from an acid-catalyzed ring opening hydrolysis/HDO of furans like MHMF (2-methyl-5hydroxymethylfuran) and DMF (2,5-dimethylfuran), 88,89 such reaction happens when ring hydrogenation is slower than hydrolysis. Interestingly, the occurrence of the reaction imposes no decrease of the n-hexane selectivity as 2,5-hexanedione is almost quantitatively converted into n-hexane in reaction conditions (Table 3, see entry 14). One may conclude at this point that 2,5-hexanediol, either resulting from 2,5-DMTHF or A similar ring opening hydrolysis, followed by dehydration/ hydrogenation, occurs with 2,5-DHMF, mainly forming 1hydroxy-2,5-hexanedione. 67,88,89 Accordingly, this intermediate is prone to convert ultimately to 1-and 2-hexanol through a family of diols such as 1,2-hexanediol (see Scheme 3) in our harsher reaction conditions. To better understand the reactivity and reaction pathways of the primary alcohol, a series of catalytic experiments was carried out with 1-and 2-hexanol and 1,2hexanediol. The data are reported in Table 3, entries 16 to 18. Whereas 2-hexanol nearly quantitatively converts to n-hexane, 1,2-hexanediol and 1-hexanol yield remarkably lower n-hexane amounts (40% and 37%, respectively). Surprising amounts of npentane (53% and 49%) were obtained instead. The n-pentane is thus formed via C-C splitting of a primary alcohol under the reaction conditions, likely proceeding through a sequential dehydrogenation/decarbonylation reaction mechanism on the modifed Ru/C. 60 This reaction should form CO as by-product, which was indeed analyzed (as methane) in gas phase analysis in equimolar amounts with n-pentane (Fig. S12-S13 in ESI †). Since formation of 1-hexanol entities entails a signifcant loss of carbon yield in the liquid alkane fraction, ring opening of 2,5-DHMF should be delayed in favor of C-O hydrogenolysis and ring-hydrogenation. Besides n-pentane and n-hexane, the liquid alkane fraction also contains signifcant amounts of methylcyclopentane (MCP). The presumable formation route proceeds through the acid-catalysed Piancatelli rearrangement from 2,5-DHMF or MHMF, 90 but this suggestion needs further confrmation. The reaction has been reported in the formation of cyclopentanone from furfural in presence of NiCu based catalysts under reducing conditions in water. 91 Optimizing the cellulose to n-hexane reaction Knowledge of the reaction network indicates that a minimum amount of redox catalyst htTSA(2)Ru/C is necessary to achieve efficient conversion of cellulose to n-hexane (Table 4, entry 1-3). Otherwise, char formation from HMF and water solubles like diols and triols will form, decreasing the content of n-decane solubles. Interestingly, reducing the original amount of catalyst twofold did not result in a signifcant change in total product yield, indicating that the hydrogenation activity in the biphasic system is still sufficient. A fourfold reduction of htTSA(2)Ru/C causes a drop in total carbon yield in the organic phase. In previous experiments the reaction mixture was analysed at a fxed reaction time, showing in some occasions signifcant amounts of intermediates like 1-hexanol and 2,5-DMTHF in the sampling mixture (Table 4, entry 4). As these molecules ultimately lead to n-hexane and n-pentane according to the results of Table 3, prolongation of the contact time is an obvious option to further increase the liquid alkane yields (see for instance an HPLC analysis of aqueous phase at various reaction times, Fig. S9 †). Fig. 3 plots the product distribution and total cellulose conversion in function of time (with an indication of the reaction temperature at each time). The plotted data indeed confrm the increase in liquid alkane yield from cellulose with time. During the reactor heating stage between 423 and 453 K, the conversion of cellulose proceeds rapidly, reaching 40% and 90% at 30 and 60 minutes, respectively. During this interval, 2,5-DMTHF, glucose, some hexitols and their anhydrides and other water-solubles including HMF, 2,5-hexanedione and 1,2-hexanediol were formed (Fig. S9 †). The accumulation of glucose and low hexitol (sorbitol, sorbitan and isosorbide) yield are indicative of the reduced glucose hydrogenation activity of the TSA-modifed Ru/C, in agreement with the data presented in Fig. 1. The inability of this catalyst to rapidly hydrogenate glucose opens up a fast cascade route, involving glucose to HMF Scheme 3 Proposed reaction pathways from cellulose to n-hexane and n-pentane through HMF with TSA and htTSA(2)Ru/C, partially based on Liu et al., 88 Alamillo et al., 67 Li et al. 60 and Yang et al. 92 Intermediates tested in this study are indicated in blue. The most selective reaction pathway from cellulose to n-hexane is indicated with bold arrows. HDO, hydrodeoxygenation; HG, hydrogenation; DH/ DC, dehydrogenation/decarbonylation. conversion, which is mainly hydrogenated to 2,5-DMTHF and some DMF was analyzed as well. The water phase contains a family of alcohols, mainly the most stable primary alcohols like 1-hexanol and 1,2-hexanediol, and also the secondary alcohols like 2-hexanol. After 1.5 h contact time, when the reactor temperature is in the range of 453-473 K, n-hexane is formed in expense of the reactive secondary alcohols through dehydration/hydrogenation cycles, 60 while the primary alcohols remain largely untouched. 2,5-DMTHF and 1-hexanol are abundantly present, while also MCP is mainly formed in this period. Conversion of 2,5-DMTHF and the alcohols continues with longer reaction times (and increasing reaction temperature up to 493 K) until they are almost completely converted. At this temperature, n-pentane is formed, while 1-hexanol is completely converted. Thus, after a short reaction time of about 6 hours, high n-decane soluble product yields (about 82% based on carbon) and C 5 -C 6 alkane yields (up to 75%, including 52% n-hexane) were obtained. In these conditions, the catalytic turnover based on surface Ru atoms can be estimated at about 200, assuming the consumption of 7H 2 molecules for n-hexane production per glucose unit (and thus 7-metal catalyzed turnovers). The defciency in the mass balance is due to some insoluble products (gas and solid, about 7% total estimated yield), while 11% carbon is present in the water phase as hexitols and some oligomeric products. Note that the hexitols were already formed very early in the reaction (after 30 min), but largely survived the reaction conditions, again proving the importance of the novel HMF route and differentiating the current biphasic system with the known pathways via sorbitol. ## Process robustness: converting real wood feedstock and catalyst and reuse The direct conversion of softwood sawdust (from a local sawmill) to n-hexane was investigated in the aforementioned optimal reaction condition to assess the robustness of the catalytic system. Softwood was deliberately chosen here due to its high polyhexose content 93 (here: 58%). Apart from cutting, no other pretreatment of the wood sample was foreseen as to omit biomass pretreatment costs and energy. Irrespective of that, an appreciable C 5 -C 6 alkane yield of almost 60%, including approximately 40% n-hexane, was attained at full conversion of the polysaccharide component. Besides the use of real lignocellulosic feedstock, multiple catalyst reuse is of vital importance to a heterogeneous process and thus two types of reuse strategies were put forward to test the resilience and durability of the TSA-modifed catalyst. At frst, the catalyst was recovered from the reaction medium by View Article Online fltration, washed and re-suspended in a fresh reaction medium after drying. The results of two such consecutive recycling runs are summarized in Fig. 4. Some loss of catalytic activity was noticed, which could originate from catalyst loss during fltration. The possibility of reusing both the heterogeneous htTSA(2) Ru/C catalyst and the soluble TSA co-catalyst in two successive runs was also investigated, by adding fresh cellulose to the batch reactor after each run and starting a new reaction, while accumulating the products. The results of this recycling are presented in Table 4 in entries 4 to 6, with yields based on the total amount of cellulosic carbon added. The catalytic system is acceptably reusable, not withstanding the harsh reaction conditions and high concentrations of products potentially inhibiting active metal sites. A small decrease in alkane and hexitol yield and an increase in oxygenate yield was monitored. ## Conclusions This contribution has demonstrated the feasibility of a one-pot conversion of cellulose to alkanes. In contrast to recently reported hydroprocessing processes, this biphasic liquid approach at moderate temperatures mainly produces straightchain alkanes with n-hexane and n-pentane as major components. The process allows an easy recuperation of alkanes, floating on top of a separate water phase, while hydrogen selectivity is high as almost no gaseous products are formed. A thorough reaction network study showed the dominant pathway, which deviates from the currently accepted sorbitol-toalkane route. 36,56,60 Instead, the major pathway proceeded via hydrolysis of cellulose to glucose, followed by dehydration into HMF. The latter needs to be hydrogenated quickly and leads to 2,5-DHMF and subsequently, via ring hydrogenation/hydrogenolysis, into 2,5-DMTHF. This cyclic ether is selectively converted into n-hexane via consecutive ring-opening hydrolysis and dehydration/hydrogenation cycles. Contribution of C]O hydrogenation/hydrogenolysis of HMF to the methyl-furans, DMF and MHMF, followed by furan ring opening constitutes a productive parallel pathway to n-hexane through 2,5-hexanedione. The ring-opening of 2,5-DHMF on the other hand leads to the formation of linear primary alcohols such as 1hexanol, and this path leads to a mixture of n-hexane, n-pentane and methane. The fast hydrogenation of 2,5DHMF to 2,5-DMTHF or hydrogenolysis to DMF or MHMF is thus an essential step in the cellulose-to-n-hexane reaction. MCP is proposed to be the result of a rearrangement reaction of 2,5-DHMF, but requires further confrmation. The critical elements of the presented catalytic system are: (i) the use of a biphasic reaction solvent system -with redox activity in the organic phase and acidity in the aqueous phase -to partition reactive intermediates and to provide the best conditions for the different reactions to occur; (ii) controlled reactor heating to gradually release glucose and to form HMF in the right temperature zone to avoid their degradation; and (iii) modifcation of the hydrogenation selectivity of commercial Ru/ C to steer the reaction from glucose to HMF hydrogenation to avoid sorbitol formation. The catalytic system proved appreciably reusable and was applicable on raw softwood sawdust (almost 40% n-hexane yield). Future improvement in n-hexane yield is envisioned through a more selective formation of 2,5-DMTHF (or HMMF and DMF) to circumvent the n-pentane production. Identifcation of the modifying role of TSA on Ru/C, optimization of the stability of the catalytic biphasic system and decreasing the carbon content in the water phase are several focus points for future research.

chemsum

{"title": "Direct catalytic conversion of cellulose to liquid straight-chain alkanes", "journal": "Royal Society of Chemistry (RSC)"}

development_of_a_new_biocathode_for_a_single_enzyme_biofuel_cell_fuelled_by_glucose

## Abstract: In this study, we reported the development of Prussian blue (PB), poly(pyrrole-2-carboxylic acid) (PPCA), and glucose oxidase (GOx) biocomposite modified graphite rod (GR) electrode as a potential biocathode for single enzyme biofuel cell fuelled by glucose. In order to design the biocathode, the GR electrode was coated with a composite of PB particles embedded in the PPCA shell and an additional layer of PPCA by cyclic voltammetry. Meanwhile, GOx molecules were covalently attached to the carboxyl groups of PPCA by an amide bond. The optimal conditions for the biocathode preparation were elaborated experimentally. After optimization, the developed biocathode showed excellent electrocatalytic activity toward the reduction of H 2 O 2 formed during GOx catalyzed glucose oxidation at a low potential of 0.1 V vs Ag/AgCl, as well as good electrochemical performance. An electrocatalytic current density of 31.68 ± 2.70 μA/cm 2 and open-circuit potential (OCP) of 293.34 ± 15.70 mV in O 2 -saturated 10 mM glucose solution at pH 6.0 were recorded. A maximal OCP of 430.15 ± 15.10 mV was recorded at 98.86 mM of glucose. In addition, the biocathode showed good operational stability, maintaining 95.53 ± 0.15% of the initial response after 14 days. These results suggest that this simply designed biocathode can be applied to the construction of a glucose-powered single enzyme biofuel cell. With the growing demand for green electrical energy generation technologies, scientists are making great efforts to develop fuel cells (FCs), which are considered as one of the most promising alternative sustainable energy sources due to their renewable and environmental protection characteristics 1 . Unlike conventional FCs, which utilize the oxidation of fuels (H 2 , ethanol, or methanol) on an anode and reduction of an oxidant on a cathode employing a noble metal catalyst 2 , biological fuel cells (biofuel cells, BFCs) convert chemical energy into electrical energy by using organic fuels (sugars, alcohols, organic acids) produced during metabolic processes and a biological catalyst, which is usually either a microorganism or an enzyme. An enzymatic biofuel cells (EBFCs) are type of BFCs that use purified redox enzymes immobilized on an anode and/or on a cathode to achieve electrocatalytic reactions 3 . Enzymes are highly specific to their respective substrate and typically operates in mild conditions. Therefore, BFCs are an attractive alternative when it is not possible to use high temperatures or where harsh reaction conditions are undesirable. Moreover, enzymes immobilized on the electrode surface allow membrane-less configuration of EBFCs, opening up opportunities for the development of miniaturized systems for powering electronic devices 4 and self-powered electrochemical biosensors, the main advantage of which is a simplified two-electrode system without an external power supply 5 . In addition to lactate 6 , cholesterol 7 , ethanol 8 and other EBFCs, special interest in recent years has been focused on the development of membrane-less EBFCs that can deliver electrical energy using oxidation of glucose at an anode and O 2 9,10 or H 2 O 2 11,12 reduction at a cathode. Glucose and O 2 are an ideal source of fuel and oxidizer because they are found in all organic tissues and can be constantly replenished during metabolic processes 13 . EBFCs that use enzymatic reactions on both electrodes have also been researched and published 14,15 . Researchers expect that in the future, miniature membrane-less EBFCs will supply energy to implantable medical devices, like insulin pumps, hearing devices, bone stimulators or pacemakers, and will also be used as self-powered biosensors, which, using an analyte as a fuel, will be able to supply themselves with energy, and at the same time to determine the amount of an analyte 16 . Implanted self-powered biosensors could be used to measure various substances that cause heart disease or cancer, as well as blood glucose 17,18 . To use EBFCs for these purposes, they should be small and light, operate at Over the last decade, the performance of EBFCs has greatly improved by using various nanomaterials, such as carbon nanotubes (CNTs), graphene oxide (GO), noble metal nanoparticles or conjugated polymers (CPs). These materials often have good biocompatibility, also electrical conductivity and large surface area. Their use allows to improve the efficiency of electron transfer and a magnitude of the generated electrical signal and often provides a stable matrix for enzyme immobilization. Due to the large surface area, nanomaterials can increase enzyme loading, furthermore, to improve the activity and stability of immobilized enzymes, thus improving the performance of EBFCs. Among the nanomaterials mentioned, CPs and CPs-based nanocomposites have gained the considerable attention of many scientists. For example, Haque and co-workers 19 reported a glassy carbon electrode modified with a conducting composite consisting of chitosan, reduced GO, polyaniline (PANI), ferritin and glucose oxidase (GOx) as a potential bioanode for glucose EBFC. The bioanode was capable to generate a current of 3.5 mA/cm 2 at 20 mM of glucose. The performance was improved due to the porosity and large surface area of the composite material, which allows the immobilization of a larger amount of enzyme and facilitates the diffusion of glucose. Although the bioanode generated a lower current signal, it nevertheless had a high operational stability and maintained 95% of the initial response after one week. Kang and co-workers 20 proposed a glucose/O 2 EBFC based on glassy carbon electrodes modified with a novel three-dimensional PANI and CNTs composite with rhizobium-like structure. The composite was prepared by in-situ polymerization of aniline monomers around and along the functionalized CNTs and then carbonized at a high temperature was used as a substrate for immobilization of GOx (anode) and laccase (Lac) (cathode). The EBFC was performed with a maximum power density of 1.12 mW/cm 2 at 0.45 V. Moreover, three fabricated EBFCs connected in series were able to light up a yellow light-emitting diode (LED) whose turn-on voltage was about at 1.8 V. Later Kang and co-workers 21 reported glucose/O 2 EBFC based on GOx and Lac immobilized on carbonized rectangular polypyrrole tubes. A nickel foam was utilized as the substrate electrode. The open-circuit voltage The opencircuit of the designed EBFC reached 1.16 V and the maximum power density was measured to 0.350 mW/cm 2 at 0.85 V. Three of the fabricated EBFCs connected in series were able to light up a white LED whose turn-on voltage was about at 2.4 V for more than 48 h. Most of the glucose EBFCs utilize glucose-oxidizing enzyme (GOx or glucose dehydrogenase) on a bioanode combined with oxygen reducing enzymes (commonly bilirubin oxidase or Lac) on a biocathode. Biocathodes based on immobilized peroxidase (PO) 14,22 and biocathodes in which GOx is combined with PO that catalyses the reduction of H 2 O 2 , produced during glucose oxidation on GOx modified electrodes 23 , have also been published. Such systems have a drawback: the use of two enzymes, which makes the system more complex and increases the cost. In addition, enzymes used may have different optimal operating conditions. These drawbacks can be avoided by designing a so-called single enzyme EBFC, in which the same enzyme is used for the anodic and cathodic reactions. The present paper describes the fabrication and investigation of a novel biocathode in the construction of which an "artificial PO" Prussian blue (PB) was used instead of PO. According to the mechanism of H 2 O 2 reduction on PB modified electrodes, PB is electrochemically reduced to form Prussian white (PW), which catalyses the reduction of H 2 O 2 at low potential 24 and PW is oxidized to PB again. Due to the reversible electrochemical redox ability of PB, it acts as a renewable catalyst throughout the electrochemical process 25 . Although PO-like property of PB has been studied for the design of biosensors 5,26 and FCs 27,28 , it has not yet been used in the construction of a biocathode for a single enzyme EBFC, whose anodic and cathodic reactions would be both based on the processes biocatalysed by GOx. This enzyme immobilized on the bioanode and biocathode would catalyze the oxidation of glucose to H 2 O 2 , which would be reduced on the surface of the biocathode. To design such biocathode, graphite rod (GR) electrode was coated with a composite of PB particles embedded in the PPCA shell (GR/PB-PPCA) and an additional layer of PPCA (GR/PB-PPCA/PPCA) by cyclic voltammetry (CV). Finally, GOx was covalently linked to the carboxyl groups of the PPCA (GR/PB-PPCA/PPCA-GOx). Immobilized GOx acted as a catalyst that oxidizes glucose by molecular O 2 , PB, meanwhile, acted as an electrocatalyst for the reduction of H 2 O 2 formed during the enzymatic reaction. To achieve the best performance of the biocathode, preparation conditions were optimized by evaluating the generated signal to glucose. After optimization, the performance of the biocathode was investigated. ## Materials and methods Chemicals. GOx from Aspergillus niger (freeze-dried powder 360 U/mg protein), iron (III) chloride (FeCl 3 ) and N-(3-dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride (EDC) were purchased from Carl Roth GmbH. Potassium hexacyanoferrate (III) (K 3 [Fe(CN) 6 ]) were from Sigma-Aldrich. Pyrrole-2-carboxylic acid (PCA) and D-(+)-glucose monohydrate (C 6 H 12 O 6 × H 2 O) were obtained from Alfa Aesar GmbH. Hydrogen peroxide (H 2 O 2 ) was obtained from Chempur and N-hydroxysuccinimide (NHS) from Merck. All chemicals were of analytical grade. All aqueous solutions were prepared in ultra-high quality (UHQ) water, which was obtained using the DEMIWA rosa 5 water purification system (WATEK, Czech Republic). In addition, glucose solution was prepared at least 24 h before use to allow glucose to mutarotate and to reach equilibrium between α-and β-forms. 40.0 mg/mL solution of GOx was freshly prepared in sodium acetate-phosphate buffer solution composed of 0.05 mM CH 3 COONa, 0.05 mM Na 2 HPO 4 and 0.05 mM NaH 2 PO 4 (A-PBS) and rapidly used. 0.5 M solution of PCA was prepared in ethanol absolute. ## Instrumentation. All electrochemical experiments as well as electrochemical synthesis of the PB-PPCA/ PPCA composite were performed using a computer-controlled potentiostat/galvanostat Autolab PGSTAT30 (Eco Chemie, Netherlands) driven by NOVA1.9 software. The voltammetric and amperometric cell was com-GR electrode pre-treatment and preparation of the biochatode. GR electrodes were prepared by breaking a graphite rod (15.0 cm in length, 3.0 mm in diameter and 99.999% purity, Sigma-Aldrich) into smaller rods of the required length. The broken rods were mechanically polished using very fine (P320) and finally ultrafine grit (P2000) sandpaper until the working surface of the electrode is completely smooth, and then were polished using a sheet of paper, washed with UHQ water, then ethanol and dried at room temperature. Finally, the side surface of the rod was isolated with a silicone tube so that only the working surface of the GR electrode was in contact with the solution in the electrochemical cell. The working surface area of GR electrodes thus prepared was 0.0707 cm 2 . Pre-treated GR electrodes were used to perform electrochemical synthesis of the PB-PPCA/PPCA composite using CV. Under the optimized modification conditions, the pre-treated GR electrode, together with the reference and auxiliary electrodes, was immersed in an electrochemical cell filled with 5 mL of a solution consisting of 100.0 mM HCl, 100.0 mM KCl, 1.0 mM FeCl 3 , 1.0 mM K 3 [Fe(CN) 6 ] and 35.0 mM PCA. The potential was then scanned for 50 consecutive cycles in the range of potentials from − 0.4 to + 1.0 V at a scan rate of 0.1 V/s. During this process, a composite of PB particles embedded in the PPCA shell (PB-PPCA) was formed on the GR electrode surface (GR/PB-PPCA). After synthesis of the PB-PPCA composite, the GR/PB-PPCA electrode was washed well with UHQ water and immersed in an electrochemical cell filled with 5 mL of A-PBS buffer solution with 0.1 M KCl additive (A-PBS-KCl), pH 6.0, and containing 200.0 mM PCA. The potential was then scanned for 5 consecutive cycles in the range of potentials from − 0.4 to + 1.0 V at a scan rate of 0.1 V/s. During this process, an additional layer of PPCA was formed on the GR/PB-PPCA electrode (GR/PB-PPCA/PPCA). To modify the GR/PB-PPCA/PPCA electrode with GOx, the electrode was immersed in a tube filled with a mixture of 0.4 M EDC and 0.1 M NHS solutions in a ratio of 1:1 and left for 30 min at ambient temperature. The electrode was then removed from the mixture, washed with UHQ water, and immersed immediately in a 40 mg/mL GOx solution in A-PBS, pH 4.0, stirring the solution gently from time to time. The GOx was attached covalently by an amide bond to the electrode surface (GR/PB-PPCA/PPCA-GOx). Finally, the GR/PB-PPCA/ PPCA-GOx electrode was washed with UHQ water and to remove non-covalently bound enzyme was conditioned in A-PBS-KCl, pH 6.0, for 15 min, stirring the solution gently from time to time. The prepared GR/ PB-PPCA/PPCA-GOx electrodes were stored in closed test tubes above a drop of A-PBS-KCl, pH 6.0, at + 4 °C temperature until used in the experiments. ## Electrochemical measurements. Electrochemical characterization of bare GR and modified GR electrodes was performed by CV, amperometric and potentiometric techniques. For half-cell measurements, threeelectrode cell was used for CV and amperometric measurements, while two-electrode cell was used for potentiometric measurements at open circuit or at an external load of 476 kΩ. A-PBS-KCl buffer solution, pH 6.0, was used as the electrolyte solution. The solution in the cell was continuously stirred with a magnetic stirrer at 450 rpm during amperometric and potentiometric measurements. Meanwhile, CV measurements were performed without stirring; stirring was turned on only after the addition of glucose to mix the solution. Amperometric current dependence of the biocathode on glucose concentration was studied at + 100 mV vs reference electrode. After stabilization of the background current or potential (base line), in the amperometric and potentiometric measurements respectively, a solution of glucose was injected in the electrochemical cell. The biocathode-generated signal was expressed as the change in cathodic current (ΔI) or the change in potential (ΔE) calculated from the signal recorded by the addition of glucose minus the baseline signal. During the CV measurements, the potential was scanned in the range of potentials from − 0.2 to + 0.5 V at a scan rate of 0.1 V/s or other as required, and the peak current was monitored. The results of all experiments were represented as a mean value of three independent measurements. ## Results and discussion In this research, a novel GR/PB-PPCA/PPCA-GOx biocomposite based biocathode was developed. CV was used for the electrochemical synthesis of the composite consisting of PB particles embedded in a PPCA shell (PB-PPCA) and for the formation of an additional PPCA layer over PB-PPCA (PB-PPCA/PPCA). To synthesize PB-PPCA on top of a GR electrode (GR/PB-PPCA), the electrode was immersed in an electrochemical cell filled with a solution consisting of HCl, KCl, FeCl 3 , K 3 [Fe(CN) 6 ] and PCA, and polymerization was performed. The GR/PB-PPCA electrode was then immersed in an electrochemical cell filled with A-PBS-KCl buffer solution containing PCA and an additional PPCA layer (GR/PB-PPCA/PPCA) was synthesized. Finally, using activation of carboxyl groups of PPCA by a mixture of EDC and NHS, GOx molecules were linked to the PPCA via amide bonds (GR/PB-PPCA/PPCA-GOx). The design concept and operation of the biocathode are shown in Fig. 1. The operation of the biocathode can be explained by the electrocatalytic activity of PB towards to the reduction of H 2 O 2 formed during GOx catalyzed oxidation of glucose. Fe(III) of PB after receiving the electron is electrochemically reduced to form PW, which has a high reduction activity 25 . The H 2 O 2 formed during enzymatic reaction is reduced by PW, and the PW is reoxidized to PB. Due to the reversible redox activity of PB, it acts as a renewable catalyst throughout the bioelectrochemical process. The amount of H 2 O 2 formed during GOxcatalyzed reaction depends on the glucose concentration, thus the current or potential signal generated by the biocathode due to H 2 O 2 reduction dependent on the glucose concentration. biocathode, its preparation conditions were optimized by estimating the magnitude of the generated current signal to glucose. Since the GR electrode was coated layer by layer with PB-PPCA/PPCA composite, the electrochemical synthesis conditions of PB-PPCA were first optimized. During optimization, PB-PPCA electrodes prepared under different conditions were modified with an additional layer of PPCA and immobilized GOx under constant conditions. The amperometric response of the GR/PB-PPCA/PPCA-GOx biocathodes to glucose in A-PBS-KCl buffer solution, expressed as cathodic current change (ΔI), was then investigated. Figure 2A shows the experimental results obtained during the optimization of FeCl 3 and K 3 [Fe(CN) 6 ] concentration. As can be seen, the current signal increased with increasing equimolar concentrations of FeCl 3 and K 3 [Fe(CN) 6 ] up to an optimal concentration of 1.0 mM. Meanwhile, the optimal PCA concentration was found to be 35.0 mM (Fig. 2B). It is likely that when the concentrations of FeCl 3 and K 3 [Fe(CN) 6 ] are too high and the PCA is too low, PB particles from the resulting PB-PPCA composite can diffuse into the solution, thus reducing the current generated by the biocathode. On the other hand, at too high concentration of PCA, a thick layer of PPCA is formed. Due to the low conductivity of this layer, the current generated is also reduced. The current response to glucose was also depended on the potential scan rate (Fig. 2C) and the number of potential scans (Fig. 2D). The highest current response of the biocathode after addition of glucose was registered when a potential scan rate of 0.1 V/s and 50 potential scans were used. During polymerization, a thick polymer shell is formed by applying more potential scans and higher scan rate. Meanwhile, with less potential scans and lower scan rate, the electrode surface may be inefficiently coated by PB-PPCA composite. This, as can be seen from the results, also affects the current generated by the biocathode. During the optimization of the deposition of the additional PPCA layer, the dependence of the current signal on PCA concentration and potential scan rate was also observed. The highest current signal was registered when 200 mM of PCA (Fig. 3A) and 5 potential scans (Fig. 3B) were used. Such results are related to the thickness of the resulting additional PPCA layer. The higher the PCA concentration and the number of potential scans, the thicker the additional PPCA layer. The thicker layer causes a decrease in the current generated by the biocathode. On the other hand, with a fuller and more uniform coating, more GOx molecules can be attached to the electrode surface. Based on the results of this study, 200 mM PCA and 5 potential scans were selected as optimal conditions for the electrochemical polymerization of the additional PPCA layer on the GR/PB-PPCA electrode. The concentration and pH of the GOx solution used for biocathode preparation were also optimized. In this work, a covalent amide coupling technique using a mixture of EDC and NHS was used for GOx immobilization. To find the most suitable pH, after activation of the PPCA carboxyl groups, the GR/PB-PPCA/PPCA electrodes were immersed in GOx solution in A-PBS buffer with a certain pH from 4.0 to 7.0. The current response of the prepared biocathodes to glucose was then investigated. As can be seen from the results presented in Fig. 3C, the magnitude of the current signal was dependent on the pH, and the highest current signal was registered at pH 4.0. These results were consistent with those obtained for GR electrodes modified with a nanobiocomposite composed of poly(1,10-phenanthroline-5,6-dione), PPCA, gold nanoparticles and GOx 29 , and can be explained by the pre-concentration of the enzyme near the electrode surface at this pH, resulting in a higher amount of immobilized GOx. Very similar results were obtained when the influence of the concentration of GOx solution used during immobilization on the magnitude of the current signal generated by the biocathode was investigated. The increase in the enzyme concentration resulted in an increase in the current response of the biocathode to glucose and the highest response was recorded then the highest GOx concentration of 40 mg/mL was used (Fig. 3D). Because 40 mg/mL is a sufficiently high concentration, the effect of higher concentrations was not studied and this GOx concentration was chosen for use in biocathode preparation. Electrochemical behaviour of the biocathode. Electrochemical behaviour was studied by recording cyclic voltammograms after an appropriate step of the biocathode preparation process. Figure 4A shows the corresponding cyclic voltammograms registered for GR, GR/PB, GR/PB-PPCA, GR/PB-PPCA/PPCA, and GR/ www.nature.com/scientificreports/ PB-PPCA/PPCA-GOx electrodes. As can be seen, in the voltammogram recorded for the GR electrode, there are no oxidation and reduction current peaks (redox peaks) in a potential range between − 0.2 and + 0.5 V vs Ag/ AgCl. Meanwhile, for the other electrodes studied, a pair of well-defined redox peaks due to the electrochemical reaction of high-spin ferric ions in PB (Fe 2+ /Fe 3+ transition) 24 were recorded in this potential range. The positions of the characteristic redox peaks and the distances between them are given in Table 1. The occurrence of characteristic redox peaks and the increase in peak current compared to the GR electrode indicate successive deposition of PB on all modified electrodes. The observed redox peaks potential values were similar to those described in the literature for PB modified electrodes . In addition, as can be seen from Fig. 4A, the cyclic voltammograms of the GR/PB-PPCA, GR/PB-PPCA/PPCA and GR/PB-PPCA/PPCA-GOx electrodes showed a significant increase in redox peaks intensity compared to the GR/PB electrode. These results demonstrated that the presence of PCA in the electropolymerization solution increases the amount of PB on the electrode surface due to its distribution inside the polymer matrix. Meanwhile, the decrease in the intensity of the GR/PB-PPCA/PPCA redox peaks compared to GR/PB-PPCA could be explained by the formation of an additional PPCA layer. Because electropolymerization was carried out in an aqueous medium without removal of oxygen, the PPCA layer formed is low conductive or even non-conductive 29 . Therefore, the formation of an additional layer of PPCA causes a decrease in the intensity of the redox peaks. For the GR/PB-PPCA/PPCA-GOx, an even greater decrease in redox peaks intensity was observed due to nonconducting nature of enzyme 33,34 immobilized on the surface. In addition, as shown in Fig. 4A, the presence of 10.0 mM glucose in A-PBS-KCl buffer solution caused an increase in GR/PB-PPCA/PPCA-GOx reduction peak current by approximately 29 μA (black line). This indicates that the H 2 O 2 formed during the enzymatic glucose oxidation reaction was electrochemically reduced on the surface of the biocathode. The cyclic voltammograms of the GR/PB-PPCA/PPCA-GOx biocathode in A-PBS-KCl, pH 6.0, at different potential scan rates are shown in Fig. 4B. As can be seen, the intensity of the redox peaks varied with the potential scan rate and was directly proportional to the scan rate (Fig. 4C). The linear relationship between oxidation (I ox ) www.nature.com/scientificreports/ and reduction (I red ) current peaks and potential scan rate and the ratio of I ox /I red almost equal to unity, revealed the quasi-reversible and surface-confined 32 electrochemical behaviour of the PB in PB-PPCA/PPCA-GOx biocomposite, in which PB is reduced to PW and is re-oxidized to PB. ## Morphological study. The surface morphology of bare GR and GR at different stages of the modifying process (GR/PB, GR/PB-PPCA, GR/PB-PPCA/PPCA, and GR/PB-PPCA/PPCA-GOx) was studied using SEM at 3 kV accelerating voltage, 50,000 magnification and 0.8 nA current. The SEM images in Fig. 5 clearly demonstrate changes in surface morphology during GR modification. As can be seen, the GR surface is quite smooth with minor defects occurring during surface polishing. (Fig. 5A). Meanwhile, a completely different surface morphology was observed for the other electrodes. Cubic PB structures of approximately 100 nm in size are observed on the surface of GR/PB (Fig. 5B), similar to those published by other authors 32,35 . However, as can be seen, the surface coating is very uneven with large, uncoated GE areas. Not uniform coating may be related to the removal of PB particles from the surface during synthesis and electrode washing after synthesis. Meanwhile, the presence of PPCA in the synthesis solution resulted in a much better and more uniform coating with a higher amount of PB on the electrode surface (Fig. 5C). However, the characteristic cubic structure of PB was no longer visible. This change in morphology may be related to the disruption of the growth of PB particles into the cubic framework due to the spatial limitations resulting from the embedment of PB particles into the growing PPCA shell 36 . PB-PPCA coating showed an irregular globular morphology, which was in agreement with other reports for PB and polymer composites 36,37 . A rougher coating with larger structures compared to GR/PB-PPCA as well as globular surface morphology was observed for GR/ PB-PPCA/PPCA (Fig. 5D). This indicates that PB-PPCA was coated with an additional layer of PPCA. Similar surface morphology, with slightly larger structures was observed for GR/PB-PPCA/PPCA-GOx (Fig. 5E). The results of surface morphology studies confirmed the assumption that PB was incorporated into PPCA during the formation of the PB-PPCA layer, and an additional PPCA layer was formed on the surface of the PB-PPCA. As can be seen, a linear increase in biocathode response was observed with increasing glucose concentration up to about 10.00 mM. With further increase in glucose concentration, a non-linear increase in the biocathode response was observed up to 98.86 mM, and then the recorded signal became saturated because the catalytic reaction was inhibited at higher glucose concentrations. As a result, the amount of H 2 O 2 formed during the enzymatic reaction became constant. The maximal current density of 85.86 ± 6.30 μA/cm 2 , the potential of 221.03 ± 13.90 mV, and the OCP of 430.15 ± 15.10 mV at 98.86 mM glucose were recorded. Meanwhile, at 10.00 mM glucose, the recorded current density was 31.68 ± 2.70 μA/cm 2 , the potential was 150.73 ± 6.70 mV and the OCP was 293.34 ± 15.70 mV. The performance of the developed biocathode was comparable to the previously reported biocathodes. Some of them are listed in Table 2.. ## Stability study of the biocathode. To investigate the stability of the developed biocathode, the current signal generated by it was monitored over a period of 14 days. The measurements were carried out in A-PBS-KCl buffer solution, pH 6.0, containing 10.0 mM of glucose. Between measurements, the biocathode was stored at + 4 °C in a closed vessel above a drop of A-PBS-KCl buffer solution, pH 6.0. As can be seen from the experimental data presented in Fig. 7, current generated by the GR/PB-PPCA/PPCA-GOx biocathode after addition ## Conclusion In this study, a graphite rod electrode modified with a biocomposite composed of Prussian blue, poly(pyrrole-2-carboxylic acid) and glucose oxidase (GOx) was proposed as a potential biocathode for glucose-powered single enzyme biofuel cell. PPCA allowed covalent immobilization of GOx, PB, meanwhile, exhibited electrocatalytic activity to the reduction of H

chemsum

{"title": "Development of a new biocathode for a single enzyme biofuel cell fuelled by glucose", "journal": "Scientific Reports - Nature"}

thermodynamic_insights_into_the_entropically_driven_self-assembly_of_amphiphilic_dyes_in_water

## Abstract: Self-assembly of amphiphilic dyes and p-systems are more difficult to understand and to control in water compared to organic solvents due to the hydrophobic effect. Herein, we elucidate in detail the self-assembly of a series of archetype bolaamphiphiles bearing a naphthalene bisimide (NBI) p-core with appended oligoethylene glycol (OEG) dendrons of different size. By utilizing temperaturedependent UV-vis spectroscopy and isothermal titration calorimetry (ITC), we have dissected the enthalpic and entropic parameters pertaining to the molecules' self-assembly. All investigated compounds show an enthalpically disfavored aggregation process leading to aggregate growth and eventually precipitation at elevated temperature, which is attributed to the dehydration of oligoethylene glycol units and their concomitant conformational changes. Back-folded conformation of the side chains plays a major role, as revealed by molecular dynamics (MD) and two dimensional NMR (2D NMR) studies, in directing the association. The sterical effect imparted by the jacketing of monomers and dimers also changes the aggregation mechanism from isodesmic to weakly anticooperative. ## Introduction Self-assembled nano-and mesoscale structures play a major role in nature and particularly in living organisms. The sophisticated functionalities of these structures is imputable to the interplay of hydrogen bonds (H-bonds) 5 as well as solvophobic effects, 6 derived to the unique role of water as a solvent. 4,7 The desire to mimic and understand such naturally occurring self-assembled systems has prompted the investigation of various amphiphilic/bolaamphiphilic molecules consisting of non-polar hydrophobic cores attached with water solubilizing side chains. 8 Through these investigations, a wide variety of nanometric supramolecular aggregates of different morphologies (tubular, fbrillar, micellar, vesicular) has been prepared via exploring solvophobic effects, H-bonding, electrostatic screening, 16,17 metal-ion coordination, variation of hydrophilic/hydrophobic balance, 21,22 and co-solvent modulation. However, different from the very intensively conducted studies on the enthalpic and entropic contributions that govern supramolecular host-guest complex formation in water, studies devoted to an in-depth understanding of the thermodynamic profle of self-assembly processes of p-amphiphiles in water remain scarce. 29 Nevertheless, such an understanding is warranted not only from a supramolecular design perspective, but also in therapeutic, and materials sciences. In this direction, we have identifed p-conjugated cores of perylene bisimide dyes appended with six oligoethylene glycol (OEG) chains as very useful amphiphilic molecules, whose selfassembly processes can be easily followed by various spectroscopic and microscopic techniques. But only very recently we unveiled our serendipitous discovery that the self-assembly of OEG substituted perylene bisimide (PBI) derivatives in water is not driven by enthalpic dispersion and electrostatic forces as in organic solvents, 31 but by entropic factors, albeit the process can be shifted to an enthalpic route by the addition of only small amounts of an organic co-solvent. 34 We attributed this intriguing behaviour to the exclusion of water molecules from the OEG side chains which leads to a dominant entropic contribution to the self-assembly in pure aqueous environment, which was also later demonstrated for other dye assemblies by Ghosh et al. 35 Whilst these interesting results warrant further studies, our PBI systems aggregated too strongly in water, evading an in-depth thermodynamic characterization including isothermal titration calorimetry (ITC). Due to the smaller pcore, naphthalene bisimides (NBIs) appeared to be more promising because their self-assembly requires higher concentrations which is benefcial for methods like NMR and ITC. Herein, we report our detailed studies on the self-assembly of a series of naphthalene bisimides functionalized with OEG chains of different glycol units (NBI 1: n ¼ 2, NBI 2: n ¼ 3, and NBI 3: n ¼ 4) in water (Fig. 1a). By means of UV-vis spectroscopy and ITC studies, we have achieved the dissection of entropic and enthalpic contributions to their self-assembly. Remarkably, we found that enthalpy, entropy and free energy changes of NBIs 1-3 aggregation in water strongly depend on the interaction of water molecules with the ether oxygens and accordingly on the OEG chain length (Fig. 1b). Additional structural investigations by molecular dynamics (MD) and 2D-NMR techniques revealed back-folding of glycol chains with sequestration of the NBI p-cores from water, to be of importance as well. ## Molecular design and synthesis In our previous work, the larger p-core of PBI prevented the quantifcation of aggregation parameters in pure water due to its pronounced association tendency. 34 Hence, for the current study, we have employed the smaller naphthalene homologue in lieu of perylene to assert moderate aggregation constants in water. Following the same design principles, bolaamphiphilic derivatives NBI 1-3 were prepared with OEG-substituted brushes at both ends of the p-core (Fig. 1a). The number of glycol units per chain was systematically increased from three to four and fve for NBI 1, NBI 2 and NBI 3, respectively. The synthesis of amphiphilic brush substituents was carried out by a two-fold, one-pot Sonogashira reaction via coupling of 2-bromo-1,3-diiodo-5-nitrobenzene with respective glycol chain substituted with a propargyl unit. 34 Subsequent reduction of the triple bonds and the nitro group in H 2 atmosphere at high pressure in the presence of 10% Pd on carbon yielded the corresponding amino derivatives. Finally, these compounds were treated with naphthalene-1,4,5,8-tetracarboxylic dianhydride in acetic acid to obtain the bolaamphiphilic NBI derivatives. The synthetic details and characterization data for all new compounds are reported in the ESI. † ## Temperature-dependent self-assembly and morphology of NBI aggregates To investigate the aqueous self-assembly behaviour of NBIs 1-3, initially we performed temperature-dependent UV-vis experiments below the cloud point (vide infra). For comparison, frst we measured the absorption spectra of the NBIs in CHCl 3 (Fig. S1 †). In an organic solvent of intermediate polarity like CHCl 3 , NBI 1 exists in monomeric form and shows an absorption maximum at 381 nm. In water, even at a dilute concentration of 9.7 10 5 M, this absorption maximum is shifted to 364 nm, suggesting an H-type aggregated state (Fig. 2a). While monitoring the temperature-dependent UV-vis spectra of NBI 1 in a range of 10 C to 50 C in water, we observed a decrease in the ratio of these two vibronic bands, with a concomitant hypochromic shift. This clearly indicates an increase in degree of aggregation with increasing temperature. 13 Similarly, both NBI 2 and NBI 3 exhibit an inverse temperature response where the aggregation is favoured at elevated temperatures (Fig. S2 †). Unlike the majority of supramolecular systems which disassemble upon heating, we have previously observed that bolaamphiphilic PBIs attached with OEG brushes show an inverse temperature response, where aggregation is favoured at higher temperatures. 34 This unique thermodynamic signature is now also manifested in the current NBI series as corroborated by temperature-dependent UV-vis experiments. It is noteworthy that the spectral changes upon aggregation are by far less pronounced for NBIs compared to PBIs due to the much smaller transition dipole moment of their S 0 / S 1 transition. At higher temperatures, we observed the phase separation of the NBIs from the aqueous phase. This is attributed to the lower critical solution temperature (LCST) phenomenon which is typical for OEG appended systems. 36 The specifc temperature denoting the onset of this precipitation, called cloud point (CP), can be determined by monitoring the transmittance at a wavelength where the molecule does not absorb (here 800 nm). The phase separation from the binary solution is characterized by an abrupt drop in transmittance. The knowledge of CP is quintessential for our self-assembly studies since it sets the upper limit for the temperature window where aggregation can be monitored. Furthermore, it gives clue towards the amount of water molecules forming H-bonds to OEG chains, as the CP increases with extent of hydration. NBI 1, containing the shortest glycol chain, shows a CP of 59 C at a concentration of 1 10 3 M in water, while NBI 2 and NBI 3 show phase separation at 78 and 88 C, respectively, at the same concentration (Fig. 2b). Since the clouding is mainly associated with the dehydration of glycol units, an increase in the CPs suggests an increase in the extent of hydration with elongation of glycol chains. In order to characterize the morphology of the aggregates formed below CP, stock solutions of NBIs in water at 22 C were spin coated onto silicon wafer treated with argon plasma and visualized using atomic force microscopy (AFM). The microscopy images for NBI 1 obtained by tapping mode reveal short nanorods with a diameter of $2 nm and an average size distribution of 20-45 nm, suggesting a one dimensional (1D) self-assembly (Fig. 2c). The presence of anisotropic aggregates was further confrmed via DLS measurements which showed size dependence upon variation of the scattering angle (Fig. S3a †). 37 Similarly, morphological investigations performed on NBI 2 as well as NBI 3 suggested nano-rod like selfassembled species with a diameter of $2 nm (Fig. S4 †). ## Thermodynamic proling of NBI self-assembly In an attempt to obtain a comprehensive thermodynamic profle for the self-assembly of NBIs 1-3 in water, we explored concentration-dependent UV-vis studies below CP to monitor their transformation from monomers to 1Daggregates. Fig. 3a displays the spectral changes observed in our concentrationdependent experiment performed on NBI 1 at 25 C. It was observed that with an increase in concentration, the absorption maximum shifts to 364 nm compared to the monomeric absorption maximum (381 nm), correlating to the spectral changes observed in temperature-dependent measurements. This suggests the formation of an H-type excitonically coupled stack. 38 Moreover, the transition from the monomeric to aggregated state is characterized by the presence of two isosbestic points (324 nm and 394 nm), implying an equilibrium between monomeric and aggregated species. Fig. 3a inset shows the corresponding plot of the degree of association (a agg ) versus the logarithm of concentration. It was observed that the best ft for the data points was obtained with an isodesmic model, i.e. an equal association constant for each monomer addition. 39 From this, the logarithm of the association constant, log K ass ¼ 3.8, and the standard Gibbs free energy of association ðDG ass Þ of 21.9 kJ mol 1 was estimated for NBI 1 (at 25 C). From the concentration-dependent UV-vis studies, the critical aggregation concentration (CAC) of 0.33 mM was also determined for NBI 1 at 25 C (Fig. S9a †). 14 To delve deeper into the understanding of thermodynamic parameters associated with the self-assembly, we performed the same experiment at different temperatures, from 10 to 50 C (Fig. S5 †). Previously, the van't Hoff equation has been successfully utilized to derive standard enthalpy ðDH ass Þ and standard entropy ðDS ass Þ changes of self-assembly by assuming a linear relationship of the natural logarithm of aggregation constants with respect to temperature. 40,41 However, this method is only valid when the enthalpy and entropy changes remain constant with changes in temperature. 42 Processes in water, however, are usually associated with wide fluctuations in these parameters, thus impeding an accurate description of the self-assembly process. 29,43,44 This limitation can be surpassed by taking the heat capacity changes into account. One such modifcation is Clarke-Glew method, where the isobaric temperature dependence of rate constants is described around a reference temperature, q. 45,46 This approach allows the calculation of the change in heat capacity at constant pressure, DC p , which is inaccessible by the van't Hoff equation due to its inherent assumptions. According to the simplifed form of Clarke-Glew method (also referred as extended/integrated van't Hoff equation), the change in association constant with respect to temperature can be expressed by eqn (1). ln ½KðTÞ ¼ ln ½KðqÞ þ DHðqÞ R where ln [K(T)] is the natural logarithm of the equilibrium constant at temperature T, ln [K(q)] is the natural logarithm of the equilibrium constant at the reference temperature q, DH(q) is the enthalpy change at the reference temperature, and DC p is the change in heat capacity at constant pressure. While plotting the natural logarithm of the association constant versus the inverse of temperature, indeed a much better ft is obtained with the non-linear Clarke-Glew equation as compared to the van't Hoff equation (Fig. 3b). Accordingly, a standard enthalpy of 11.2 kJ mol 1 and a heat capacity change of 289 J mol 1 K 1 can be calculated for the self-assembly of NBI 1. With an elevation in temperature, an increase in aggregation strength is observed, quantitatively supporting our temperature-dependent UV-vis measurements. Furthermore, the negative slope of the curve suggests the endothermic nature of self-association over a broad temperature range, which is hence enthalpically disfavoured. Similarly, concentration-dependent UV-vis experiments were conducted for NBI 2 and NBI 3 at different temperatures (Fig. S6 and S7 †). In both cases, we observed that the mechanism of selfassembly differs from the isodesmic model and is better described by a weak anti-cooperative process with a cooperativity factor of s ¼ 2 and s ¼ 3 for NBI 2 and NBI 3, respectively. By ftting the data according to the Goldstein-Stryer model 47 utilized for (anti)cooperative aggregation processes, a logarithm of the association constant, log K ass ¼ 3.3, and a standard Gibbs free energy, DG ass ¼ 18:8 kJ mol 1 was determined for NBI 2 at 25 C, suggesting a weaker aggregation tendency as compared to NBI 1. Using the Clarke-Glew plot, a standard enthalpy change of 18.1 kJ mol 1 is calculated, which shows that the selfassembly of NBI 2 is enthalpically more disfavoured than NBI 1 (Fig. S8a †). NBI 3 exhibited the weakest aggregation tendency of all three derivatives, with log K ass ¼ 2.8 and a standard Gibbs free energy, DG ass ¼ 16:4 kJ mol 1 at 25 C. The selfassembly, in this case, is disfavoured by a standard enthalpy ðDH ass Þ of 23.2 kJ mol 1 (Fig. S8b †). Furthermore, the CAC estimated for NBI 2 (1.6 mM) and NBI 3 (3.5 mM) at 25 C confrms the decreasing tendency of aggregation while increasing the glycol chain length from NBI 1 to NBI 3 (Fig. S9b and c †). The thermodynamic signature at 25 C obtained for the three derivatives is represented in Fig. 4, which depicts that the selfassembly for all the NBI derivatives in water is enthalpically disfavoured and entropically driven. Furthermore, this penalty in the standard enthalpy of association ðDH ass Þ and the gain in standard entropy of association ðDS ass Þ is augmented as the OEG chain length is increased from NBI 1 to NBI 3. Since our CP measurements suggest an increase in hydration with chain elongation, this trend can be attributed to the increased number of water molecules that are removed for well-hydrated monomer units upon aggregation. On the other hand, the aggregation tendency decreases with an increase of glycol units as reflected by increased DG ass values. In order to validate the thermodynamic parameters obtained by our UV-vis experiments, we resorted to an independent technique to derive the enthalpy, entropy and free energy of association. This technique is given with an ITC dilution experiment that allows direct determination of enthalpy and gathers insight into its temperature dependency, which is inaccessible via other methods. Even though ITC is well established for natural 48,49 and synthetic host-guest interactions, 28, the advent of this technique to probe self-assembly is quite recent. 34, In a typical ITC dilution experiment, aliquots of a concentrated solution of the aggregated species is titrated into the pure solvent taken in the cell. The dissociation of the aggregate is then accompanied by non-constant heat signals along with constant heat of dilution. 53 From this, enthalpy and other thermodynamic parameters can be determined. Fig. 5a shows the evolution of heat per injection of a concentrated aqueous NBI 1 solution (c ¼ 5.2 10 3 M) into pure water at 25 C leading to its disassembly, which depicts an exothermic heat flow, i.e., the dis-assembly process is enthalpically favoured. The corresponding enthalpogram could be well ftted to an isodesmic model (Fig. 5b). 56,57 A standard enthalpy change of 13.8 kJ mol 1 for dis-assembly (or +13.8 kJ mol 1 for the corresponding self-assembly) and logarithm of the association constant, log K ass ¼ 3.8 at 25 C was determined for NBI 1, which is indeed in good concordance with the previously obtained values from UV-vis experiments (vide supra). Also, a CAC value of 0.21 mM was deduced for NBI 1 from the aforementioned ITC dilution experiment (Fig. S10 †). 35 The accompanying heat of dilution estimated from the overall heat evolved during injection of NBI 1 is provided in Table S1. † Accordingly, different from our previous study of a strongly aggregating PBI, 34 here we could for the frst time quantify the entropically driven self-assembly thermodynamics in water and derive values for DH ass and K ass of high accuracy. The thermodynamic parameters obtained by both these methods are tabulated in Table 1. Successively, to understand the influence of temperature on the enthalpy of self-assembly, we repeated the ITC dilution experiment at different temperatures, from 10 to 50 C (Fig. S11 †). It was observed that with increasing temperature, the enthalpy of association for NBI 1 is concomitantly decreased (Fig. 5c). It is expected that an elevation in temperature decreases the H-bond strength between the water molecules and OEG chains, 58 thus reducing the enthalpic penalty associated with the dehydration of water molecules during selfassembly. The increased aggregation tendency of these systems at higher temperatures could be traced to this easiness in the release of H-bonded water molecules. The resulting heat capacity change for NBI 1 aggregation was quantifed as 280 J mol 1 K 1 using eqn (2), Similar dilution experiments in pure water were also performed for NBI 2 and NBI 3 at 25 C (Fig. S12 and S13 †). Here also the dilution experiments revealed exothermic signals for disassembly, accordingly the self-assembly is endothermic. Intriguingly, in both cases, we observed heat signals associated with two distinct processes (Fig. S12b and S13b †). Such two-step processes with similar heat signature have been previously reported for host-guest studies of ions with macrocycles which follow negative cooperative mechanism. 59 We assume that since the aforementioned derivatives aggregate via a weak anticooperative mechanism, the frst injections might represent the dissociation of fully aggregated aliquots into monomers whereas the latter injections show the dissociation into the dimeric species. Unfortunately, the currently available model was not able to describe these processes and hence hampers the accurate determination of aggregation parameters for NBI 2 and NBI 3. Furthermore, the lack of saturation at the end-point of dilution experiment due to lower aggregation tendency impedes the estimation of CAC for NBI 2 and NBI 3 via ITC. Thus, we could independently confrm by both UV-vis studies and ITC dilution experiments that the self-assembly of NBIs 1-3 in water is entropically driven and primarily attributable to the release of water molecules from the glycol units. Here the length of the OEG side chains plays a prominent role for both the enthalpic and entropic contributions to the aqueous self-assembly of our NBI series. In order to obtain deeper insights into the role of molecular structure in orchestrating this specifc aggregation trend in water, structural attributes, especially the conformational nature of glycol units have to be investigated in detail. ## Structural characterization via molecular dynamics (MD) and 2D NMR studies After procuring quantitative information about the thermodynamic signature associated with the self-assembly, we pondered upon the role of molecular structure in directing the association process. For this, we employed all atoms molecular dynamics (MD) simulations in pure water on NBI 1 and NBI 3 which have the shortest and the longest OEG chains, respectively, in the series. Interestingly, both NBIs show a back-folded conformation of glycol chains around the naphthalene core in the monomeric form (Fig. 6a and d). A similar observation was previously discussed by Meijer and Pavan et al. and has been attributed to the shielding of the hydrophobic surface from the surrounding bulk water. 12,60 Fig. 6b and e show the density profles of C-atoms of OEG chain over the aromatic cores for NBI 1 and NBI 3. As seen clearly, the preferred orientation of side chains during the MD regime resides over the core instead of extending into the bulk water. Next, two such pre-equilibrated monomers were immersed into a periodic simulation box flled with explicit water molecules and allowed to equilibrate over MD regime. The distance between the two monomers (0.4 nm) suggests an explicit p-p stacking, with a rotational offset of 10 (Fig. S15a and S16a †). A snapshot from the trajectory of NBI 1 stacking depicts that the glycol chains still prefer a back-folded orientation in the aggregated state (Fig. S14a †). However, the tail density is now more distributed around the p-core, suggesting that some of the back-folding was replaced in order to accommodate the incoming monomer (Fig. 6c). This release of ordered chains might contribute to the conformational entropy of side chains, aiding overall entropy of the association, along with the removal of hydrated water molecules. Similarly, for NBI 3, stacking interactions were studied via MD simulations (Fig. S14b †). Here we see again that the glycol chains are folded over the naphthalene core, in both monomeric and dimeric form. However, due to the increased length of OEG units, the density of back-folded conformation is concomitantly higher as compared to NBI 1 (Fig. 6f). The rotational offset for the NBI 3 stack ($60 ) is signifcantly larger compared to NBI 1, which could be rationalized by the steric hindrance of back-folded glycol chains (Fig. S16b †). To experimentally verify the presence of back-folding as suggested by MD simulations and to unravel the aggregate structure, we conducted detailed one-dimensional (1D) and two dimensional (2D) NMR studies. The 1 H NMR spectrum of NBI 1 in CDCl 3 shows well resolved sharp signals corresponding to the monomeric state (Fig. 7a). In contrast, the naphthalene core protons are signifcantly broadened as well as up-feld shifted in D 2 O, indicating an aggregated state aided by p-p stacking. Insights into the aggregate structure were probed subsequently via 1 H-1 H Rotating Frame Overhauser Effect Spectroscopy (ROESY). Fig. 7c and d show selected regions of superimposed ROESY and COSY spectra of NBI 1 in D 2 O. Nuclear Overhauser Effect (NOE) correlations could be observed between the naphthalene core protons (H a ) and the glycol protons (H e /H e 0 ) which is in compliance with the back-folded conformation of side chains. The coupling between the naphthalene core protons and phenyl protons suggests a slightly rotated offset between NBI 1 monomers in the stacked conformation as predicted by MD simulations. A tentative assignment of NOE correlations with a snapshot from MD regime of NBI 1 is given in Fig. 7e. Similarly, for both NBI 2 and NBI 3, through-space interactions could be traced between glycol chain protons (H e / H e 0 ) and core protons (H a ), thus corroborating the presence of back-folded conformations in these systems (Fig. S17 and S18 †) and validating the structural predictions from MD simulations. In the thermodynamic analysis of the current system, we have observed that the elongation of glycol units from NBI 1 to NBI 3 is associated with a nearly ten-fold decrease in association constant and a concomitant drop in the magnitude of free energy. Combined results from MD simulations and NMR studies suggest that the back-folding of glycol chains is orchestrating this effect. Furthermore, the change of aggregation mechanism from isodesmic to weak anti-cooperative can also be attributed to the more pronounced jacketing of monomer and dimer species by the longer OEG chains. In addition, our results also relate to the studies on biomolecules. Here, previously it was observed that the substitution of proteins with polyethylene glycol (PEG) results in a decrease in binding affinity due to the interactions between PEG chains and the active site of the protein. 61,62 Meijer et al. predicted that this could be ascribed to the back-folding of the glycol chains operative in water. 63 Our current studies prove that the back-folding indeed interferes with the association process by shielding the hydrophobic surface from the surrounding bulk water. Accordingly, we can conclude that OEG and PEG chains play a pivotal role in directing the thermodynamics of aggregation in water. ## Conclusions In this contribution, three archetype bolaamphiphilic naphthalene bisimides were studied to derive an understanding of the different factors that contribute to the entropically driven self-assembly of bolaamphiphilic moieties substituted with OEG units in water. By utilizing UV-vis and ITC dilution experiments, we have successfully dissected the thermodynamic parameters of the aggregation process. The entropically favoured nature of the self-assembly is attributed to the release of water molecules from the glycol units which is enthalpically penalized. Further, we were able to show that a thermodynamic tuning of p-core aggregation in water can be achieved by modulating the length of solubilizing OEG chains. With elongation of the side chains, the enthalpic as well as entropic parameters also increase, attributed to an increment in dehydrated water molecules upon aggregation. However, this augmentation in their length hinders the self-assembly via a back-folding process as revealed by MD simulations and 2D-NMR studies, resulting in a decrease of the magnitude of Gibbs free energy and deviation from the isodesmic mechanism. Our current study sheds light into the fundamental aspects of bolaamphiphilic aggregation in water and opens up a strategy for more predictable aqueous self-assembly processes of oligo-and polyethylene glycol functionalized amphiphilic molecules. ## Conflicts of interest There are no conflicts to declare.

chemsum

{"title": "Thermodynamic insights into the entropically driven self-assembly of amphiphilic dyes in water", "journal": "Royal Society of Chemistry (RSC)"}

hydrofluoromethylation_of_alkenes_with_fluoroiodomethane_and_beyond

## Abstract: A process for the direct hydrofluoromethylation of alkenes is reported for the first time. This straighforward silyl radical-mediated reaction utilises CH 2 FI as a non-ozone depleting reagent, traditionally used in electrophilic, nucleophilic and carbene-type chemistry, but not as a CH 2 F radical source. By circumventing the challenges associated with the high reduction potential of CH 2 FI being closer to CH 3 I than CF 3 I, and harnessing instead the favourable bond dissociation energy of the C-I bond, we demonstrate that feedstock electron-deficient alkenes are converted into products resulting from net hydrofluoromethylation with the intervention of (Me 3 Si) 3 SiH under blue LED activation. This deceptively simple yet powerful methodology was extended to a range of (halo)methyl radical precursors including ICH 2 I, ICH 2 Br, ICH 2 Cl, and CHBr 2 F, as well as CH 3 I itself; this latter reagent therefore enables direct hydromethylation. This versatile chemistry was applied to 18 F-, 13 C-, and D-labelled reagents as well as complex biologically relevant alkenes, providing facile access to more than fifty products for applications in medicinal chemistry and positron emission tomography. ## Introduction The introduction of fluoroalkyl groups has garnered signifcant interest in medicinal chemistry, enabling the modulation of biological and physicochemical properties of lead candidates for drug discovery. Whilst the felds of radical tri-fluoromethylation and difluoromethylation have been extensively explored, the fluoromethyl radical has received far less attention. This is unexpected as the fluoromethyl group features frequently in pharmaceutical drugs, more often to improve metabolic stability by serving as a bioisosteric replacement of functional groups responsible for poor performance. 14,15 In recent years, several reagents for the generation of the CH 2 F radical have been developed. Often, efficient activation of these reagents requires harsh reaction conditions, such as elevated temperatures, strong oxidants, or strong reductants. Furthermore, many of these reagents are either expensive, highly toxic or non-commercial, requiring multistep syntheses for their preparation. As part of our growing interest in developing "minimalistic" procedures for the late-stage hydrofluoroalkylation of alkene-containing biologically active molecules, we sought to develop an operationally simple method for the direct hydrofluoromethylation of alkenes, as an attractive strategy for the introduction of this motif to C(sp 3 )enriched backbones (Scheme 1). In 2020, an indirect method for the hydrofluoromethylation of alkenes was developed by Aggarwal and co-workers; 13 this elegant multi-step procedure starts with the conversion of alkenes into boronic esters, subsequent treatment at low temperature (78 C) with in situ formed fluoroiodomethyl lithium to generate fluoroboronic esters, and a fnal protodeboronation. Our aim was to develop a one-step method that avoids operational complexity and over-engineering, ideally using fluoroiodomethane which is a non-ozone depleting, easy to handle and inexpensive commercial CH 2 F radical precursor. We noted that fluoroiodomethane has found applications as an electrophilic or nucleophilic fluoromethylation reagent as well as in cross-coupling reactions, but has not been explored in the context of radical chemistry. The high reduction potential of CH 2 FI (E red ¼ 2.19 V vs. saturated calomel electrode (SCE) in MeCN), 28 much closer to MeI (E red ¼ 2.39 V vs. SCE in MeCN) 28 than CF 3 I (E red ¼ 1.22 V vs. SCE in MeCN), 29 encouraged the implementation of an activation pathway exploiting instead the favourable bond dissociation energy (BDE) of C-I (BDE (FH 2 C-I) ¼ 233 kJ mol 1 ) versus C-F (BDE (IH 2 C-F) ¼ 460 kJ mol 1 ). 30 Since the pioneering work of Chatgilialoglu, 31 tris(trimethylsilyl)silane (TTMSS) has found ample applications as a powerful tool for mild radical generation via the activation of alkyl halides. In addition, TTMSS has valuably complemented Giese-type reactions, a commonly exploited platform for late-stage functionalisation, by providing a suitable alternative to traditional toxic tin-based reagents. 37 Consequently, we envisioned that the supersilyl radical (TMS) 3 Sic would be well suited to release cCH 2 F from CH 2 FI. Subsequent Giese-type addition of cCH 2 F to the electron-defcient alkene would generate a carbon-centered radical intermediate. Hydrogen-atom transfer (HAT) between this electrophilic species and hydridic (TMS) 3 SiH would afford the desired hydrofluoromethylated product, and (TMS) 3 Sic entering chain propagation. Initiation for this process would be triggered by photolytic C-I cleavage of FH 2 C-I. 38 This method offers the prospect of being applicable to a range of other halocontaining alkyl radicals, provided that competitive hydrogen atom abstraction with (TMS) 3 SiH does not occur prior to Giese addition. Herein, we report the realisation of this strategy with a wide range of haloiodomethanes for the direct hydrohalomethylation of electron-defcient alkenes including biologically relevant molecules. The method was extended to 18 F-hydrofluoromethylation and hydromethylation with iodomethane along with fve of its D and 13 C isotopomers. ## Results and discussion Preliminary experiments were conducted with N-phenyl acrylamide (1a) (Table 1). 39 Various combinations of silanes and solvents revealed that the desired hydrofluoromethylated product (2a) was obtained in 71% with (TMS) 3 SiH in MeCN at room temperature under blue light irradiation for 16 h (entry 1). 40 The addition of fac-Ir(ppy) 3 (0.5 mol%) did not lead to signifcant improvement (entry 2). The simpler protocol was therefore retained for further investigations. Control experiments indicate that the reaction was not effective in absence of light (entry 3), and unsuccessful in absence of silane or in presence of the radical scavenger TEMPO (entries 4 and 5). No deuterium incorporation was observed in the product when the reaction was performed in CD 3 CN. 40 These data corroborate our proposed radical chain propagation mechanism, initiated by blue-light homolysis of the CH 2 F-I bond. 38 Giese addition of the fluoromethyl radical to an electron-defcient alkene furnishes an electrophilic carbon-centered radical intermediate, capable of undergoing HAT with (TMS) 3 SiH. The resulting silyl radical enables chain propagation by abstracting iodine from CH 2 FI to afford (TMS) 3 SiI along with cCH 2 F. 40 With the optimised reaction conditions in hand, we sought to explore the scope of this hydrofluoromethylation protocol (Scheme 2A). Various functional groups, such as methoxy, nitrile, halide, ketone, ether, amide, ester, aniline, and sulfone were tolerated. The addition of fac-Ir(ppy) 3 (0.5 mol%) led to higher yields for selected substrates. 40 N-Aryl acrylamides bearing electron-withdrawing and electron-donating groups afforded the desired products in moderate to excellent yields (2a-d). The hydrofluoromethylation of N-heteroaryl acrylamides, such as pyridyl and benzothiazyl was also successful (2e, 2f). Alkenes substituted with sulfones and esters were competent substrates generating 2g and 2h in moderate yield. As deuteration can improve metabolic stability, 41 we investigated the hydrofluoromethylation of a deuterated alkene (1i) that was successfully converted into [D 3 ]2i. The gem-disubstituted alkene 1j provided 2j in 64% yield. Pleasingly, the internal alkene 1k was reactive under our reaction conditions and afforded fluoromethylcyclobutane 2k in moderate yield. This result is signifcant as 1,2-disubstituted fluoroalkyl cyclobutanes currently require multiple steps for their preparation. 42 A non-Scheme 1 Hydro(per)fluoromethylation of alkenes. This work: direct silyl radical-mediated hydrofluoromethylation of electron-deficient alkenes and extension to numerous hydro(halo)methylation reactions. cyclic trisubstituted alkene afforded the product in 57% yield (2l). Styrene derivatives such as 1m and 1n afforded the desired products in synthetically useful yields (2m, 2n). Our protocol is amenable to scale-up as demonstrated by the 10 mmol scale hydrofluoromethylation of N-benzylmaleimide affording 2o in 88% yield. The synthesis of fluorinated pyrrolidine 2p, amine 2q, alcohol 2r and carboxylic acid 2s was performed in two steps, offering a pathway to diversify the range of products within reach from CH 2 FI. The late-stage hydro-fluoromethylation of complex biologically active molecules was considered next. The anti-cancer drug ibrutinib as well as estrone, tyrosine and ethacrynic acid derivatives afforded the desired hydrofluoromethylated products in good yields (2t-w). The tolerance of functional groups was investigated with a robustness screening. 40 These experimental data provide an overview of the many heteroarenes (e.g. pyridazine, 1,3,5triazine, indole, benzothiazole or oxazole) that are tolerated under the optimised reaction conditions. Whilst additives containing nucleophilic functional groups such as alcohols and anilines were tolerated, side reactivity arising from nucleophilic substitution was observed. 40 Competitive alkylation was suppressed when using 1.0 equivalent of CH 2 FI, albeit at the expense of reduced yield for the hydrofluoromethylated product. Aliphatic amines were tolerated but yields did not exceed 30%. 40 The hydro-fluoromethylation of alkenes not bearing electron-withdrawing groups was possible albeit signifcantly less efficient. 40 With a protocol relying on the favourable C-I bond dissociation energy and considering the importance of homologation in medicinal chemistry, 43 we considered the generation of products from a series of homologated fluoroiodoalkanes (Scheme 2B). 44,45 Hydrofluoroalkylation of alkenes 1g, 1j and 1l provided effortlessly the homologous series of products 3d-i. Specifcally, the fluoroethyl radical was efficiently generated applying similar silyl radical activation, and 3a was isolated in good yield. The introduction of the fluoroethyl radical was successfully performed on linear terminal, gem-disubstituted, and trisubstituted alkenes (3d, 3f, 3h). The method was further extended to fluoroiodopropane as shown with the synthesis of 3e, 3g, and 3i. Precursors featuring additional fluorine atoms were less suitable with the difluoroethylated product 3b isolated in 30%, and no product observed when attempting to prepare the hydrotrifluoroethylated product 3c. Increased fluorine content enhances radical electrophilicity, thereby encouraging undesired H-atom abstraction from (TMS) 3 SiH. 40 Given the success of our protocol, we further investigated the applicability of our method for the generation of [ 18 F]CH 2 F radical from [ 18 F]CH 2 FI (Scheme 2C). Compounds labelled with the radioisotope F-18 are important for applications in Positron Emission Tomography (PET). The synthesis of [ 18 F] CH 2 FI in high molar activity (A m ) is well-established and has been automated. To date, this labelled reagent is mainly employed for the electrophilic 18 F-fluoromethylation of phenols. 58,59 We now demonstrate that [ 18 F]CH 2 FI is well suited for [ 18 F]CH 2 F radical chemistry. Specifcally, Ibrutinib, an estrone, a tyrosine, and an ethacrynic acid derivative underwent 18 F-hydrofluoromethylation in radiochemical yields up to 81% ([ 18 F]2t-w). This reaction was best performed for 20 minutes at ambient temperature in the presence of fac-Ir(ppy) 3 under bluelight irradiation. This method offers an alternative to nucleophilic 18 F-fluorination with [ 18 F]fluoride for precursors that are either unstable, require complex multiple steps synthesis, or lead predominantly to elimination products. Haloiodomethanes other than fluoroiodomethane were also considered as they would allow for the one-step introduction of reactive halomethyl groups to alkenes (Scheme 2D). Controlled activation of reagents such as ICH 2 X (X ¼ Cl, Br, I) would enable their use for example as cCH 2 + synthon. To date, only few examples for the generation and use of halomethyl radicals have been reported. When diiodomethane was employed under the standard reaction conditions, N-benzylmaleimide underwent hydroiodomethylation in 62% yield (4a). Similarly, hydrobromomethylation (from dibromomethane or bromoiodomethane), hydrochloromethylation (from chloroiodomethane), and hydrobromofluoromethylation (from dibromofluoromethane) provided the corresponding halomethyl alkanes in moderate yields (4b-4d). 23,65 Other alkenes afforded the hydrochloromethylated products in moderate yields (4e-4g). Although full conversion of starting material was observed for these reactions, purifcation via silica gel chromatography led to elimination, which is reflected in the lower yield for these compounds upon isolation. Competition experiments were performed to calibrate the reactivity of fluoroiodomethane versus other alkyl iodides (Scheme 3). When equimolar amounts of iodomethane and fluoroiodomethane were subjected to the standard reaction conditions, product resulting from fluoromethyl radical addition was obtained in 74% yield (2n), along with 25% of the hydromethylated product 5a. When the reaction was carried out with equimolar amounts of iodoethane, products 2n and 6 were formed in close to 1 : 1 ratio. Additional competition experiments showed that the iso-propyl and tert-butyl adducts (7, 8) were formed preferentially over the hydrofluoromethylated product. The reactivity of these alkyl iodides therefore decreases in the following order: tBuI > iPrI > CH 2 FI $ EtI > MeI. A notable outcome of this study was the observation that net methane addition across the double bond took place with iodomethane. Currently, protocols for the generation of the methyl radical from iodomethane (BDE CH3-I ¼ 239 kJ mol 1 , E red ¼ 2.39 V vs. SCE in MeCN) remain underdeveloped. 34,66 In recent years, the methyl radical has been generated from numerous precursors. The formation of the methyl radical often requires harsh reaction conditions, limiting the applicability of these protocols. Furthermore, the use of the methyl radical towards application to isotopic labelling is far from trivial. Iodomethane, on the other hand, can provide effortless access to a variety of useful isotopologues that would otherwise be beyond reach. The straightforwardness of our protocol prompted us to optimise the hydromethylation of alkenes using iodomethane as methyl radical precursor (Scheme 4). We noted signifcant gas release when applying our reaction conditions, attributed to methane resulting from competitive HAT between the methyl radical and MeCN (BDE NCCH2-H ¼ 389 kJ mol; BDE CH3-H ¼ 439 kJ mol 1 ). 69 A screen of solvents, reactants stoichiometry and photocatalysts allowed for hydromethylation to occur in up to 93% yield (5a). 40 Under the optimised reaction conditions consisting of 4.0 equivalents of MeI, 3.0 equivalents of (TMS) 3 SiH and 1,2-difluorobenzene as solvent, in combination with photocatalyst MesAcrBF 4 (0.5 mol%), the hydromethylation of various alkenes took place in good to excellent yield (5b-f). Considering that bioactive compounds containing stable heavy isotopes are useful for example as mass spectroscopy standards, 41,72 the hydromethylation of an ethacrynic acid derivative was performed with CH 3 I, CH 2 DI, CHD 2 I, CD 3 I, 13 CH 3 I, and 13 CD 3 I. All six isotopologues (5h-5m) were obtained in moderate yield. ## Conclusions In conclusion, the frst direct hydrofluoromethylation of a broad range of electron-defcient alkenes has been developed using fluoroiodomethane. Mechanistically, the process harnesses known principles; so its core value is rooted in its immediate synthetic power. With the current global necessity "to do more with less", this minimalistic and mild chemical method stands out as it is operationally simple with the supersilyl radical precursor (TMS) 3 SiH being the only chemical required in addition to the reaction partners. The mild reaction conditions are compatible with complex biologically active molecules such as Ibrutinib. The methodology was successfully adapted for the 18 F-labelling of complex alkenes, and offers a new C-CH 2 18 F disconnection strategy for radiotracer development. The method was extended to additional fluoroiodoalkanes enabling facile product homologation, as well as multiple (halo)methyl radicals including the methyl radical itself and fve of its D and 13 C isotopomers.

chemsum

{"title": "Hydrofluoromethylation of alkenes with fluoroiodomethane and beyond", "journal": "Royal Society of Chemistry (RSC)"}

a_photoacoustic-fluorescent_imaging_probe_for_proteolytic_gingipains_expressed_by_porphyromonas_ging

## Abstract: Porphyromonas gingivalis is a keystone pathogen in periodontal disease. We herein report a dual-modal fluorescent and photoacoustic imaging probe for the detection of gingipain proteases secreted by P. gingivalis. This probe harnesses the intramolecular dimerization of peptide-linked cyanine dyes to induce fluorescence and photoacoustic off-states. Upon proteolytic cleavage by Argspecific gingipain (RgpB), five-fold photoacoustic enhancement and >100-fold fluorescence activation was measured with detection limits of 1.1 nM RgpB and 5.0E4 CFU/mL bacteria in vitro. RgpB activity was imaged in the subgingival pocket of porcine jaws with 25 nM sensitivity. The diagnostic efficacy of the probe was evaluated in gingival crevicular fluid (GCF) samples from subjects with (n = 14) and without (n = 6) periodontal disease, wherein activation was correlated to qPCR-based detection of P. gingivalis (Pearson's r = 0.71). The highest activity was seen in subjects with the most severe disease. Finally, photoacoustic imaging of RgpB-cleaved probe was achieved in murine brains ex vivo, demonstrating relevance and potential utility for animal models of general infection by P. gingivalis, motivated by the recent biological link between gingipain and Alzheimer's disease. ## Introduction Periodontitis is a chronic inflammatory disease that affects 46% of adults in the United States and generates billions of dollars per year in direct costs . The pathogenesis of the disease remains an active research topic; however, it is principally associated with a dysbiotic oral microbiome and the accompanying immune response . Periodontitis-associated bacteria reside in the subgingival crevice, and their presence in biofilms and gingival crevicular fluid contribute to degradation of host tissue and deepening of the periodontal pocket . When untreated, periodontitis causes oral pain, tooth loosening, and tooth loss. Furthermore, the long-term loading of the immune system has been linked to increased risks for cardiovascular disease , pre-term birth , cancer , and even dementia . Periodontal health is measured via periodontal probing and clinical examination with metrics that include the pocket depth, clinical attachment level, bleeding on probing, tooth mobility, and inflammation. Together, these metrics are used to form a diagnosis. In general, this established practice is functional and affordable, but pocket depth and clinical attachment level measurements suffer from relatively high inter-examiner error due to differences in probing force/angulation while also causing patient discomfort. Moreover, these techniques largely assess the effects of disease rather than using molecular diagnostics for precision health. Therefore, new techniques to detect disease at the point-of-care-particularly with utility for imaging and identification of disease at the molecular levelremain an unmet need in the field of oral health. Many of the periodontal pathogens that have been linked to disease are anaerobic, such as Tannerella forsythia, Treponema denticola, and Porphyromonas gingivalis . Among this "red complex", P. gingivalis is the most well-characterized: Its presence in subgingival plaque has been correlated with disease progression in longitudinal human studies . As a function of their anaerobic metabolism, these pathogens secrete protease virulence factors that degrade extracellular proteins and modulate the host immune response . P. gingivalis, in particular, is known to secrete proteases called gingipains that exhibit trypsin-like activity . Indeed, P. gingivalis and gingipain proteases have attracted attention both as diagnostic and therapeutic targets. A variety of naturally derived and synthetic gingipain inhibitors have been reported in the literature while demonstrating evidence for potential treatment of periodontal disease though clinical trials have been relatively rare . Intriguingly, evidence of gingipains has been identified in the post-mortem brains of patients with Alzheimer's disease (AD) and are the target of an ongoing AD clinical trial for a small molecule gingipain inhibitor . A parallel research effort is targeting P. gingivalis directly with an antibody therapy . From a diagnostic perspective, advances in gingipain detection have included the development of substrates and paper-based assays for in situ analysis , a plasmonic nanosensor , and a gingipain-responsive/drug-loaded hydrogel . The goal of this study was to develop an activatable probe for gingipains with utility for in vivo imaging-such work was motivated by its potential as a clinical tool for periodontal diagnosis and as a research tool for investigation of the role of gingipains in periodontitis and other diseases. Photoacoustic imaging is particularly attractive because it augments the existing strengths of ultrasound-good tissue penetration, low cost, and real-time monitoring. It can use both exogenous and endogenous contrast based on optical absorption. Many small molecule and nanoparticle contrast agents have been engineered for photoacoustic imaging and activatable probes for molecular imaging are particularly desirable . Further, the applications of acoustic imaging and nanoscale materials have been expanding but they have not yet been combined for oral imaging. In previous work, we introduced a dye-peptide scaffold that exploits the intramolecular coupling of cyanine dyes to generate photoacoustic and fluorescent signal upon proteolysis by trypsin. Here, we leveraged this approach to create an activatable photoacoustic and fluorescent molecular imaging agent for gingipain proteases. To select a gingipain-cleavable peptide substrate, we first applied a structural model of peptide-protein affinity to screen a series of pentapeptides for their affinity to the Arg-specific cysteine protease gingipain R (RgpB, PDB: 1CVR) . The RgpB protease is composed of a 435-residue, single-chain polypeptide that forms a catalytic domain and an immunoglobulin-like domain . The peptide candidates were generated with three constraints: a five-residue length, an arginine at the third residue (P1), and a lysine at the fifth residue (C-terminus, P2'). The peptide length was restricted to facilitate intramolecular interaction between N and C terminal dyes while reducing the likelihood of cleavage by off-target proteases. The central arginine was necessary for cleavage by RgpB, and the Cterminal lysine was chosen for its reactive free amine. These conditions allowed us to generate 8,000 possible sequences that were screened for affinity to RgpB using an open-source structural model (PepSite 2.0) based upon a library of known peptide-protein complexes [20b] . The results were plotted as the inverse p-value to signify relative affinity (Fig. 1A) where the p-value represents the statistical significance for the overall score of a given binding site defined by Petsalaki et al. . Of these peptides, the top result that did not contain a cysteine (excluded to reduce effects from dithiol coupling) was APRIK (p-value 0.0266) and was selected for probe synthesis. Additionally, the median result (TTRIK (p-value: 0.1866)) and last result (EEREK (p-value: 0.6872)) were synthesized and 3 conjugated with dyes to serve as experimental controls for the model predictions (Fig. 1B). Visualization of the APRIK-RgpB interaction demonstrated that the peptide was predicted to bind the catalytic domain of RgpB (Fig. 1C). The three candidate peptides were used to synthesize homodimer probes [Cy5.5]2[APRIK], [Cy5.5]2[TTRIK], and [Cy5.5]2[EEREK], referred throughout as C2A, C2T, and C2E, respectively (Fig. S1). RP-HPLC retention times for the conjugates decreased slightly from C2A (11.8 min) to C2T (11.7 min) to C2E (10.9 min), corresponding to the increasing hydrophilicity of the residues in each peptide (Fig. S2A-C); the structures of the probes were confirmed with ESI-MS (Fig. S2D-F). The absorbance maxima of the conjugates in water were blueshifted relative to their spectra in DMSO (Fig. 1E)-a solvatochromic effect indicative of aromatic dye stacking . This blue shift confirmed intramolecular dye coupling, i.e., DMSO promotes intramolecular separation of the dyes by neutralizing their attractive π-π interactions, thus mimicking the effect of proteolytic cleavage of the peptide linker. Indeed, the fluorescence of the intact conjugates was also selfquenched but was activated upon incubation with RgpB: We measured the fluorescence from each conjugate at a range of concentrations with constant RgpB and observed stronger foldenhancement for C2A/C2T than C2E (Fig. 1F). While C2A and C2T performed similarly in this comparison, we selected C2A for further development due to its higher predicted affinity for RgpB and higher signal to background ratio at concentrations > 6 μM. The decreased activation at higher concentrations was caused by increased selfquenching of the probes, though this is dependent upon the amount of DMSO in the mixtures. Upon incubation of C2A with RgpB, the absorbance maxima of the dyes at 680 nm were recovered with increasing concentrations of protease (Fig. 2A). The fluorescence emission at 700 nm was also proportionally enhanced (Fig. 2B). The fluorescence limit of detection was 1.1 nM (linearity 0 -5 nM) (Fig. 2C). Additionally, the photoacoustic intensity of the samples excited at 680 nm was proportional to their absorption (Fig. 2A, D), and the photoacoustic limit of detection was 10 nM RgpB. To further verify the probe's sensitivity and selectivity for gingipains associated with P. gingivalis, we grew and isolated bacterial supernatants from both P. gingivalis and another oral anaerobe, F. nucleatum (Fig. 3A). F. nucleatum is a good negative control because it is also commonly identified in the gingival sulcus but is a saccharolytic and commensal bacterium known to not secrete gingipains . These anaerobes were first grown on blood agar and enumerated from liquid suspensions via optical density after development of standard curves (Fig. S3, Fig. S4). The presence of Arg-specific gingipain in the P. gingivalis cultures was confirmed with a commercially available enzyme-linked immunoassay (ELISA) kit (Fig. S5); in addition, activity was measured by incubation with a commercially available fluorescent substrate, Boc-Phe-Ser-Arg-AMC, as previously described (Fig. S6). Then, upon incubation of the C2A probe with P. gingivalis supernatant, we directly observed cleavage of intact C2A (TR = 21.2 min) into Cy5.5-APR (TR = 17.2 min, [M+2H] 2+ = 412.91 m/z) and IK-Cy5.5 (TR = 18.2 min, [M+2H] 2+ = 454.93 m/z) fragments with HPLC and ESI-MS (Fig. S7), thus demonstrating the expected activity of Arg-gingipain in the bacterial supernatant and intended cleavage of C2A. Indeed, the probe activated fluorescence 135-fold over the course of 2 hours, corresponding to enhanced emission at 700 nm and absorbance at 680 nm (Fig. 3B, Fig. S8); this activation was reduced by 97% upon coincubation with leupeptin-a known gingipain inhibitor (Fig. 3B). The fluorescence was not activated by F. nucleatum. As with fluorescence, we observed an increasing trend for the photoacoustic intensities of the samples excited at 680 nm, thus demonstrating selective photoacoustic imaging of gingipains from P. gingivalis (Fig. 3C-D). The limits of detection for the bacteria were tested via serial dilution of the supernatants in broth and determined to be 4.4E4 CFU/mL via fluorescence and 4.1E5 CFU/mL via photoacoustics (Fig. ## 3E-F, Fig. S9). To date, reported strategies for measurement of gingipain activity have used in vitro detection methods, including a nanobody immunoassay [15a] , an electrochemical biosensor , fluorogenic dipeptides [15b] , peptide-functionalized magnetic nanobeads , and refractometry of protein-functionalized gold nanoparticles . These have reported detection limits of 7.81E6 CFU/mL bacteria, 5E5 CFU/mL bacteria, 1E5 CFU/mL bacteria, 49 CFU/mL bacteria, and 4.3 nM Kgp (CFU/mL not reported), respectively. While the C2A probe has comparable sensitivity to these in vitro sensors (fluorescence LOD: 4.4E4 CFU/mL and 1.1 nM RgpB, photoacoustic LOD: 4.1E5 CFU/mL and 15 nM RgpB), it is the first reported gingipain probe suitable for photoacoustic imaging while also achieving a dual-modal fluorescence readout, with applicability for in vivo oral photoacoustic imaging, a technique that is gaining preclinical traction . The added value of imaging is the monitoring of disease progression or response to therapy with the spatial integration of anatomic markers of disease. Indeed, to characterize the imaging performance of the C2A probe in relevant oral anatomy, it was used to resolve the periodontal pocket/gingival sulcus of intact porcine jaws with photoacousticultrasound imaging (Fig. 4). Here, buffer, C2A, and C2A + RgpB (25 and 50 nM), were irrigated sequentially into the gingival sulcus of the second molar of a porcine mandible (n = 3). 3D photoacousticultrasonographs of the tooth/gingiva were generated (Fig. 4A, left) and anatomical markers were readily resolved in the midsagittal cross sections (Fig. 4A, right), including the gingival margin (GM, pink) and alveolar bone crest (ABC, teal). The uncleaved C2A probe did not possess significantly more photoacoustic signal (red) than buffer alone (Fig. 4B). However, C2A activated with 25-50 nM RgpB generated clear and increasing subgingival photoacoustic signal (Fig. 4C-E, yellow boxes), representing the subgingival distribution of RgpB-cleaved probe. In addition, spectral imaging could distinguish the imaging signal from cleaved C2A (< 750 nm) from the relatively flat spectra from supragingival signal caused by tooth staining (Fig. 4F-G). Overall, this experiment demonstrates the ability to image the spatial distribution of subgingival gingipain activity in relation to key landmarks of oral anatomy while achieving low nanomolar sensitivity. In a study by Guentsch et al., ELISA was used to identify micromolar concentrations of gingipain in gingival crevicular fluid (GCF) collected with paper point sampling from patients with periodontal disease . This is well above the low nanomolar detection limits of C2A for RgpB: Therefore, to evaluate the diagnostic efficacy of the C2A probe in clinically relevant samples, we collected GCF from 40 tooth sites in a set of 20 subjects, comprising both healthy patients and individuals with symptoms of periodontal disease sampled at a dental clinic. The GCF samples were assayed with both qPCR and C2A via fluorescence to measure the number of P. gingivalis cells and proteolytic gingipain activity, respectively. Of these, 25% (10/40) contained P. gingivalis via qPCR and these were considered positives (Fig. 5A). Gingipain activity via C2A fluorescence was correlated with the PCR results (Pearson's r = 0.71, Fig. 5B), albeit with lower sensitivity: Fluorescence activation was observed in 5/10 of the positives and 2/30 of the negatives, corresponding to a detection rate of 50% and a false positive rate of 6.67% (Fig. 5A). However, the higher sensitivity of qPCR was expected given its inherent signal amplification mechanism. Another difference is that while qPCR may reflect the amount of live and dead cells, it is not a measurement of the active gingipain activity that is evidenced to play a direct role in the pathogenic process of periodontal disease . The activity data was also analyzed with respect to disease severity for each tooth site (Fig. 5C). Interestingly, gingipain activity was primarily observed in the GCF from Class III sites (with the greatest total facial CAL). Though half of these sites did not exhibit gingipain activity, these results support the hypothesis that local gingipain activity may contribute to more severe periodontal damage. Lastly, the potential role of P. gingivalis and gingipains in neurological pathologies, especially Alzheimer's disease, is of mounting research interest . Photoacoustic imaging is well-suited for real-time imaging and monitoring of murine disease models, and thus we performed proof of concept imaging of cleaved and uncleaved probe in extracted murine brains (fixed in 1% agar). The C2A probe was first incubated with RgpB at increasing probe concentrations to confirm cleavage at sufficient concentrations for imaging in animal tissue (Fig. S10A-B), and the highest tested concentration (30 μM) was chosen for injection (Fig. S10C). Subsequently, aliquots of buffer, C2A, and C2A + RgpB (pre-incubated and monitored for 2 h) were injected into the lambda points of respective brains (Fig. S10D)these were then imaged in 3D with a photoacoustic-ultrasound scanner at 680 nm using sonography gel for acoustic coupling. Negligible photoacoustic signal was detected in the buffer-injected brain (Fig. 6A), while minor background was observed for the uncleaved probe (Fig. 6B). The strongest signal was detected from the brain injected with C2A + RgpB, visible in axial, coronal, and sagittal cross-sections of the tissue (Fig. 6C). Further, spectral photoacoustic imaging of the injected brains allowed signal from C2A to be distinguished from background by its characteristic absorption/photoacoustic peak in the near infrared (Fig. S11). These experiments demonstrate that the C2A probe could have value as a research tool for gingipain imaging in more complex models of infection for Alzheimer's disease pathogenesis. Future studies may integrate the probe with in vivo models of P. gingivalis infection, though potential limitations include issues that affect many smallmolecule photoacoustic probes, including low signal to background ratio in blood at low concentrations and photoinstability associated with the dissociation of conjugated π electrons following absorption . Nevertheless, proof-of-concept imaging utility was demonstrated in the oral cavity and brain parenchyma using resected porcine jaws and murine brains, respectively. Lastly, in future efforts to improve sensitivity to P. gingivalis, a lysine residue could be included in the peptide linker for cleavage by Lys-gingipain (Kgp), in addition to D-amino acids for increased bacterial specificity . In summary, a molecular imaging probe, C2A, was designed and synthesized to harness the intramolecular dimerization of peptide-linked cyanine dyes to induce fluorescence and photoacoustic off-states. Upon proteolytic cleavage by Arg-specific gingipain (RgpB), 5-fold photoacoustic enhancement and >100-fold fluorescence enhancement was achieved with detection limits of 1.1 nM RgpB and 4.4E4 CFU/mL bacteria. RgpB activity was imaged in the subgingival pocket of porcine mandibles with 25 nM sensitivity. The diagnostic efficacy of the probe was evaluated in gingival crevicular fluid (GCF) samples from subjects with (n = 14) and without (n = 6) periodontal disease; activation correlated to qPCR-based detection of P. gingivalis (Pearson's r = 0.71), and activity was highest in subjects with the most severe disease progression. Lastly, photoacoustic imaging of RgpB-cleaved probe was demonstrated in murine brains ex vivo, thus demonstrating future utility for imaging studies of general infection by P. gingivalis.

chemsum

{"title": "A photoacoustic-fluorescent imaging probe for proteolytic gingipains expressed by Porphyromonas gingivalis", "journal": "ChemRxiv"}

multi-target_dopamine_d3_receptor_modulators:_actionable_knowledge_for_drug_design_from_molecular_dy

## Abstract: Local changes in the structure of G-protein coupled receptors (GPCR) binders largely affect their pharmacological profile. While the sought efficacy can be empirically obtained by introducing local modifications, the underlining structural explanation can remain elusive. Here, molecular dynamics (MD) simulations of the eticlopride-bound inactive state of the Dopamine D3 Receptor (D3DR) have been clustered using a machine learning-based approach in the attempt to rationalize the efficacy change in four congeneric modulators. Accumulating extended MD trajectories of receptor-ligand complexes, we observed how the increase in ligand flexibility progressively destabilized the crystal structure of the inactivated receptor. To prospectively validate this model, a partial agonist was rationally designed based on structural insights and computational modeling, and eventually synthesized and tested. Results turned out to be in line with the predictions. This case study suggests that the investigation of ligand flexibility in the framework of extended MD simulations can assist and inform drug design strategies, highlighting its potential role as a powerful in silico counterpart to functional assays. ## Introduction The pharmacological properties of a drug are commonly considered to be continuous functions in chemical space: small changes in chemical structure lead to small differences in a compound's pharmacological profile. However, recent evidence has highlighted the existence of discontinuities: sometimes, small structural changes lead to large differences in one or more features. Activity cliffs are the best characterized form of discontinuity, but this concept can be extended to the study of other relevant properties. GPCRs have emerged as a target class whose modulators explore rugged chemical landscapes. Small variations in GPCR binders can lead to significant changes in the efficacy, with or without affecting binding affinity. This is particularly important in drug discovery because agonists, antagonists, or inverse agonists can all be therapeutically relevant, depending on the receptor and pathological framework. In some cases, the efficacy can be tuned by adopting empirical synthetic strategies. However, the underlying structural mechanisms have often remained elusive. GPCRs exist in an equilibrium ensemble of metastable conformations whose stabilization, following ligand binding, is crucial to eliciting a particular response. The energy difference among states is often minimal, which likely explains how small structural variations in a ligand could affect the receptor's conformational ensemble. Here, we tested the possibility of using MD simulations and cluster analysis in a comparative fashion, to rationalize and predict how local structural variations affect the efficacy of four modulators which we had previously reported as multi-target binders of Fatty Acid Amide Hydrolase (FAAH), an enzyme involved in the endocannabinoid signalling pathway, and dopamine D3 receptor (D3DR). Both proteins have been independently studied for nicotine addiction. Despite these targets being biochemically and structurally unrelated, we were able to conceive molecules with high affinity for both of them. Yet, accurately predicting efficacy at D3DR remained problematic. In this context, computational methods improving the design of multi-target directed ligands (MTDLs) holds great potential toward the development of efficient drugs against tobacco dependence. Here, we selected a training set of four compounds that, despite their high structural similarity, show increasing levels of efficacy at D3DR without any relevant change in affinity. Compound 1 (Figure 1) is a neutral antagonist that bears: i) a 2,3-dichloro substitution on the pendant aromatic ring of the piperazine; ii) an unsaturation in the butyl/(E)-but-2ene linker, and iii) a carboxamide substituent in the distal ring of the biphenyl group. In the presence of a saturated linker, 2 behaves like a partial agonist, eliciting only 56% of the inhibitory response. The removal of the carboxamide in 3 does not alter the efficacy (65%) profile with respect to 2 (58%), despite a moderate decrease in activity. Compound 4 differs in features i) and ii), and is therefore an almost full agonist with 88% efficacy compared to the effects of 300 nM of dopamine on cAMP inhibition . To derive our model and to understand which structural features affect efficacy, we attempted to interpret how the conformational behavior of each ligand within the binding site induces or stabilizes different interactions in residues H349 6.55 , Y365 7.35 and S193 5.43 , at the same time destabilizing the inactive crystal structure of D3DR (the superscript number indicating conserved positions according to Ballesteros and Weinstein numbering) . H-bond networks involving conserved pairs of amino acids in positions 6.55, 7.35 and 5.43 have been subject of several studies on the D2-like dopamine receptors sub-family (D2, D3 and D4). In the homologous (78% sequence homology) dopamine D2 receptor (D2DR) in complex with agonists and partial agonists, these interaction networks have been associated with low-energy patterns and functional bias. Three conserved functional serine residues on TM5, i.e., S192 5.42 , S193, 5.43 and S196 5.46 are crucial in the GPCR activation pathway that involves the formation of H-bonds between ligand, water, and receptor. The inward movement and anticlockwise rotation (from the extra cellular point of view) of TM5 is required to orient the serines toward the binding site. However, the serine in position 5.43 is only secondarily involved in catechol binding and has been found to establish favorable agonist-induced H-bonds with H393 6.55 in D2DR. [10, In H393 6.55 A and H393 6.55 F mutants, a 28-fold drop in dopamine binding affinity has been observed, which correlated with reduced efficacy and confirmed the role of an aromatic residue with H-bonding capabilities in that position. In the eticlopride-bound D3DR crystal structure, Y365 7.35 is fixed in a stable interaction network involving H349 6.55 and I183 on the extracellular loop 2 (EL2). This interaction is absent in the D4 Dopamine Receptor (D4DR) structure complexed with nemonapride, where a valine substitutes the tyrosine in position 7.35. Y365 7.35 V D3DR mutants show decreased constitutive activity of the receptor, while V430 7.35 Y cross-mutation causes the opposite effect in D4DR, highlighting that a H-bond network involving the two residues plays a functional role in regulating specific response at D3 receptor subtype. Furthermore, in microsecond-long MD simulations on an active D2DR homology model in complex with dopamine, the χ1 torsion (C-CA-CB-CC) of H349 6.55 mainly adopted three specific dihedral angle values, namely -60°, 60°, and 140°. While these angles could all be induced in the presence of the natural substrate, MD simulations with known partial agonists revealed that, in that case, the χ1 dihedral angle of H349 6.55 was mainly stabilized at 60° or 140°. These different orientations influence the interactions and the dynamics of the D2DR-G protein interface. Therefore, the H-bond patterns established by H349 6.55 with Y365 7.35 and S193 5.43 , hereafter referred to as interaction pattern 1 and interaction pattern 2, respectively, have been proposed to play a crucial role in modulating the response of D2-like sub-family of dopamine receptors in presence of molecules with different pharmacological profiles. Based on the insights gained from our studies, we designed, synthesized and tested a novel D3DR partial agonist (compound 5 in Figure 1) with dual FAAH/D3DR affinity and the sought efficacy profile. ## Ligand Docking and MD simulations Receptor coordinates of the human D3DR crystal structure in complex with eticlopride (PDB ID: 3PBL) were retrieved from the PDB and used for docking and refinement procedures implemented in the ICM software suite. Hydrogen atoms were added. Polar hydrogen atoms and the positions of asparagine and glutamine side chain amidic groups were optimized and assigned the lowest energy conformation. After optimization, histidines were automatically assigned the tautomerization state that improved the hydrogen bonding pattern. Finally, the original ligand was deleted from the holo structure. Binding sites were identified with the IcmPocketFinder tool as implemented in ICM3.7.3. The tolerance value was set equal to 4.6. The macro provides a mesh associated with every detected pocket. The graphical object closest to the co-crystallized ligand position was selected. All the residues with at least one side chain non-hydrogen atom within 3.5 of the selected mesh were considered part of the pocket. Ligands were assigned the right bond orders, stereochemistry, hydrogen atoms, and the most abundant protonation states predicted at pH 7.4. Each ligand was assigned the MMFF force field atom types and charges. The docking engine used was the Biased Probability Monte Carlo (BPMC) stochastic optimizer, as implemented in ICM3.7 (Molsoft LLC, San Diego, CA -USA). The ligand binding site at the receptor was represented by precalculated 0.5 spacing potential grid maps, representing van der Waals potentials for hydrogen and heavy atom probes, electrostatics, hydrophobicity, and hydrogen bonding. The van der Waals interactions were described with a smoother form of the 6-12 Lennard-Jones potential with the repulsive contribution capped at 4.0 kcal/mol. The electrostatic contribution was buffered, artificially increasing the distance between oppositely charged atoms to avoid their collapse when the electrostatic attractive energy prevailed on the softened van der Waals repulsion. The molecular conformation was described with internal coordinate variables. The adopted force field was a modified version of the ECEPP/3 force field with a distance-dependent dielectric constant. Given the number of rotatable bonds in the ligand, an adaptive algorithm (thoroughness 3.0) calculated the basic number of BPMC steps to be carried out. The binding energy was assessed with the standard ICM empirical scoring function. For each ligand, the best scoring pose was selected as representative of the ligand-bound conformation. The D3DR in the inactive crystallized structure was properly refined and minimized after ligand placement, and chosen as a starting structure for MD simulations. Here, we focused on the ability of our compounds to destabilize the crystallized inactive state of the D3DR, triggering the onset of early local events connected to full structural transitions. In fact, large scale conformational changes remain out of reach for an antagonist-bound initial state, even with microseconds of MD trajectories. By comparing different ligands bound to the same inactive structure, we did not have to model an active state by homology. Furthermore, small differences (5-10 folds) in energy have been found between D3DR conformations with high and low affinity for agonists. In the crystal structure, the long intracellular loop 3 (IL3) (R222-R318) involved in G-protein binding is not solved but substituted by T4-lysozime for stability reasons. As in previously reported long MD simulations of β2-adrenergic GPCR, we did not attempt to model IL3, since IL3 does not affect ligand binding. In this study, we adopted the same strategy for four reasons: i) ligand binding to D3DR is not significantly modified by the presence of guanyl nucleotides (G-shift), meaning that G-protein binding and activation has only a weak influence on the orthosteric binding pocket; ii) D3DR expressed in E. Coli has shown similar ligand binding capabilities in the presence and absence of Gi/o; , and iii) modeling extended protein loops does not ensure the reproduction of salient features of these domains. Using the membrane embedding procedure implemented in BiKi LifeSciences (http://www.bikitech.com/), we enclosed each complex in a simulation box of 8x8x10 nm containing 182 molecules of palmitoyl oleoyl phosphatidylcholine (POPC) lipids. Systems were prepared as described in the Supplementary Material. Rather than simulating multiple replicates, we preferred to carry out single MD simulations accumulating 3.05 µs on each system. This choice allowed us to increase the probability to observe rare events in the inactive structures of the receptor, enabling systematical comparisons of ligand-induced states on longer time-scales. To avoid any memory of the initial structure, we discarded the first 50 ns from each production run; the analyses were performed on 30000 snapshots extracted from each trajectory. ## Cluster analysis Conformers were pooled together based on their shared scaffold (Figure S1) and clustered by k-medoids algorithm (see description of the algorithm in the Supplementary Material). To simplify the analysis and get more interpretable and robust results, three clustering rounds were performed, merging compounds 1, 2, 4 in the first run, and 1, 3, 4 in the second run. Similarly, compounds 1, 5, 4 were merged in a meta-trajectory to study ligands partitioning in presence of the predicted ligand 5. ## Principal Component Analysis (PCA) in the space of dihedral angles To get a per ligand retrospective characterization of the space sampled by compounds 1-5, we built a dataset such that each line represented a ligand, while the columns (the features) were obtained by a processing of dihedral angles time data (30000 values x ligand). In detail, we first computed the sinuses and cosines of each angle for each ligand to correctly take into account the periodicity of the variables. Next, for each dihedral angle time series (either the sin/cos values) we computed a histogram with nb bins. This binning allowed us to get a discrete version of the distribution function that, at equilibrium, must be stationary. Hence, each ligand was represented by a column of ni entries where ni = 3*nb*2. The multiplier by three derives from the number of angles analyzed, while the multiplier by two derived from the sinus and cosine representations. On this matrix we performed PCA analysis. ## Synthesis and Pharmacological evaluation of Compound 5 Description of synthesis and structural characterization of compound 5 are reported in the dedicated section of the Supplementary Material. The pharmacological profile of the molecule was assessed based on the ligands' inhibiting effects on cAMP accumulation via activation of Gi protein, as described in details in ref. . ## Analysis of the clustering graph in the training set (Compounds 1-4) Compounds 1-4 were docked into the binding site of D3DR inactive structure (PDB ID: 3PBL) (see Material and Methods). Similarly to the eticlopride-bound structure, all the best-scoring ligand poses reproduced the driving interaction between the basic nitrogen of the piperazine ring and the side chain of D110 3.32 . As expected, given the similarity, all molecules docked consistently and in qualitative agreement with the binding mode proposed by Chien and colleagues for D3-selective derivatives. In Figure 2, the clustering results on the 1, 2, 4 set are reported as a graph. Bent and extended conformations of the ligands were isolated from each trajectory but with varying frequencies. In the graph, clusters display a selective enrichment in one or more ligands (Table S1) with different efficacy profiles and are connected through a heterogeneously populated hub node mainly characterized by bent poses (cluster 0, Figure 2c). The insets 2a, 2b, 2d-2f in Figure 2 highlight global differences in ligand scaffold orientations in the binding site (cyan licorice) relatively to the representative conformation of the hub node (gold licorice). Cluster 1 and cluster 3 were almost exclusively populated by compound 1, with 10145 and 7624 members (Figure 2a-b). The substituted 2,3dichlorophenyl ring and the trans double bond in the linker led to extended rearrangements that could be found in these clusters and that were stable in time. For example, the conformation observed in cluster 1 only appeared after 1.85 µs but was thereafter stably preserved until the end of the simulation. Lacking both the unsaturation in the linker and substituents in the pendant aromatic ring, 4 sampled the broader portion of the conformational space, showing substantial enrichment in node 7 (2005 members, bent conformation, Figure 2f) and almost exclusively populating cluster 5 (Figure 2d) and cluster 8. In the trajectory, the bent conformation associated with cluster 7 appeared after 100 ns and lasted for an additional 400 ns. Agonist-specific conformations associated with cluster 5 (9672 members) and cluster 8 (6778 members) appeared late in the simulation (around 2 µs) and were stably preserved. Compound 2 mostly interconverted between bent conformations of cluster 4 (10305 members) and cluster 6 (11773 members) (Figure 2e). Through MD simulations and cluster analysis carried out on the second dataset (compounds 1, 3, 4) we: i) assessed the robustness of the algorithm in reproducing agonist-selective and antagonist-selective clusters obtained in the first group; and ii) investigated the influence of the carboxamide substitution at the biphenyl group on the dynamics of 2 and 3. The topology of the graph obtained for the set formed by 1, 3, and 4 (Figure S2) is consistent with the one reported in Figure 2, robustly returning a similar partition (Table S2), in line with the overlapping efficacies of 2 and 3 (see discussion in the Supplementary Material). Taken together, our results suggest that different efficacy profiles could be linked to preferential stabilization of ligand-specific bent and extended conformations. Whereas compound 1 stabilized extended rearrangements, the two partial agonists 2 and 3 preferentially bound the receptor with exclusive bent conformations (clusters 4 and 6) without never transitioning into agonist-/antagonist-specific nodes. Similarly, the full agonist 4 was able to enrich exclusive clusters, but it preserved the unique feature of switching between selective bent (cluster 7) and extended conformations (clusters 5 and 8). At the receptor level, ligand-induced changes involved functional residues as H349 6.55 , Y365 7.35 , S193 5.43 and extracellular loops 2 and 3 (EL2, EL3) conformations. To compare our results with relevant findings on D2-like sub-family of DRs, we monitored the status of ligand-induced networks and local interaction patterns (1 and 2) involving the phenyl ring of the ligands, H349 6.55 , Y365 7.35 , and S193 5.43 , in each system (Figure 3). Also, we collected the values of χ1 dihedral angles in H349 6.55 and S193 5.43 , which recent studies have linked to the earliest stages of the activation process. The fluctuations of these internal variables are summarized in Figure 3 and they will be separately discussed and compared in the light of the experimental efficacies of the individual compounds. Cluster graph of the conformations explored by 1 (magenta), 2 (green), and 4 (blue). Each node represents a conformation. The size of each node is proportional to the cluster population. Each node is colored according to the relative cluster enrichment. Edges define transitions between clusters. In the insets (a-f), representative medoids (cyan) from each cluster are shown in complex with their corresponding D3DR conformation, and compared to "hub" medoid pose (gold) in cluster 0. Red circles on TM3, TM5, TM6, and TM7 indicate diagnostic residues D110 3.32 , S193 5.43 , H349 6.55 , and Y365 7.35 . EL2 connects TM4-TM5; EL3 connects TM6-TM7. (a, f) H-bond distance H349 6.55 (N)-Y365 7.35 (H); H-bond distances indicating the interaction pattern 1 are calculated between the hydrogen on the epsilon nitrogen atom of H349 6.55 and the phenolic hydrogen atom of Y365 7.35 ; (b, g) aromatic interaction distance H349 6.55 -phenyl ring (ligands 1-5) calculated between rings' centers of mass. (c, h) H-bond distance H349 6.55 (N)-S193 5.43 (H); H-bond distances connoting the interaction pattern 2 are calculated between the epsilon-bound hydrogen atom of H349 6.55 and the oxygen atom of S193 5.43 side chain; (d, i) χ1 dihedral angle of S193 5.43 . (e, l) χ1 dihedral angle of H349 6.55 . Color codes for ligands 1-5 are consistent with Figures 2 and S2. ## Antagonist-induced conformations Antagonist-specific clusters 1 and 3 were both characterized by an extended conformation of the common core (Figure 4) and stabilized D3DR in a closed state due to concerted motions of TM6-TM7 toward TM1-TM2 (compare TM7 in Figures 2a-c). This shift caused EL2 to come in close contact with EL3 (Figure 2a). H349 6.55 and Y365 7.35 remained around 6 apart. In line with recent work on antagonist-bound D3DR structures, H349 6.55 and S193 5.43 were found at around 8 apart, indicating that the antagonist stabilized longer TM5-TM6 interface distances. Therefore, neither interaction pattern 1 (Figure 3a, magenta line) nor interaction pattern 2 (Figure 3c, magenta line) were ever observed. Instead, they were replaced by stable H-bond gating bridges established by H349 6.55 and Y365 7.35 with EL2 residues, as I183 and S182 (Figure 4 and Figure 5a, c). In cluster 3, the antagonist further stabilized Y365 7.35 orientation via T-shape p-p interactions with the biphenyl moiety (Figure 4b). The antagonist limited the fluctuations of EL2, stabilizing the loop in a conformation that sealed the binding site from above (Figure 2a), as also observed from the lowest average number of waters surrounding the ligand along the trajectory (Table S3 and Figure S3). This shielding process was enhanced by the van der Waals interactions established by I183 side chain in the phenylpiperazine binding site. Our findings are in agreement with recently reported MD simulations and mutagenesis studies, which recognize EL2 as a crucial element in GPCR activation, and I183 as an important residue for antagonist binding. The key difference between the two antagonist clusters was in the orientation of the 2,3-dichlorophenyl ring relative to membrane plane. While cluster 3 identified a perpendicular orientation (Figure 4a), cluster 1 displayed a parallel one (Figure 4b). The 2,3-dichlorophenyl group established very weak aromatic and hydrophobic interactions with H349 6.55 (distance ≈ 6.5 , magenta line in Figure 3b). After 2 µs, the transition from cluster 3 to cluster 1 occurred and the distance between the imidazole ring in H349 6.55 and the ligand further increased, reaching an average value of 7.5 . This change was induced by the rotation of the C(Cl)CNC dihedral angle (see Figure S1 and Figures 2a, 4b) observed in cluster 1, where the unfavorable interaction with H349 6.55 was replaced by a T-shape aromatic stacking with the side chain of F188 5.38 . This phenylalanine made an inward rotation into the binding site, which was followed by the formation of p-p interactions with 1. Contacts established with position 5.38 have been associated with β-arrestin activation in 5-HT2B and D2DR. When the antagonist stabilized in cluster 1, rotation of the phenylpiperazine core also induced F346 6.52 side chain to shift outward at the interface of TM5-6, where optimized parallel stacking between the ring of 1 and the side chains of F197 5.47 , F338 6.44 , and F346 6.52 was observed. In agreement with our observations, such outward rotation opens a "cryptic pocket" which was found to be crucial in explaining the efficacy of a D3DR antagonist. In contrast, Michino and coworkers have recently observed that F346 6.52 rotation toward TM5-TM6 interface facilitates the inward motion of TM6 and can be considered a signature of partial agonists-driven destabilization of the inactive state of D3DR. The authors observed how the aromatic residue pointed toward the receptor core in presence of an antagonist, hindering TM6 movement. Our results showed that F346 6.52 sampled both orientations, pointing toward receptor core when 1 was in cluster 3 and toward the TM5-TM6 interface when the antagonist populated cluster 1. On the same line, along the trajectory, the χ1 dihedral angle of H349 6.55 showed small fluctuations around 60° (Figure 3e) while S193 5.43 mainly oriented outward (-50°) (Figure 3d). In recent MD studies on dopamine-bound active model of D2DR, H349 6.55 χ1 value could indeed be sampled, but with much lower frequency with respect to values conducive to a fully substrate-activated state of the receptor (-60°), whereas S193 5.43 χ1 value stabilized around 160°. ## Partial agonist-induced conformations In 2 and 3, a partial agonist efficacy profile was associated with a different behavior, as a consequence of the increased ligand flexibility. Namely, bent conformations of 2 observed in nodes 0, 4, 6, and 7 were stabilized by T-shape stacking between the ligand 2,3-dichlorophenyl ring and F346 6.52 . This residue pointed toward the receptor core, as in the receptor crystallographic structure, without rotating outward. This is particularly relevant in light of the results recently reported by Ferruz and coworkers, who have shown how inward/outward conformations of F346 6.52 could have a crucial role in D3DR response to ligands of varying efficacies (see the Supplementary Material for detailed discussion of partial agonist 3). In our simulations, partial agonist binding dynamics did not lead to antagonist-induced H-bond bridges (Figure 5). Conversely, we could observe EL2 displacement (Figure 2e and Figure S2a), enhancing water influx into the orthosteric binding pocket relatively to 1 (Figure S3 and Table S3). Compounds 2 and 3 facilitated this process preserving a dynamic coupling of TM6-TM7 interface. Indeed, distances between interacting atoms in H349 6.55 and Y365 7.35 were substantially shorter in these trajectories than in the antagonist-bound receptor, dropping on average by 2 to 4 . The 2,3-dichlorophenyl ring of 2, as also seen for 1, made only weak T-shape p-p interactions with H349 6.55 (Figure 3b). Furthermore, in contrast to the antagonist, bent conformations of compounds 2 and 3 formed a parallel stacking interaction with the side chain of Y365 7.35 , promoting the breaking of the H349 6.55 -I183 and Y365 7.35 -S182 Hbonds and the formation of the interaction pattern 1 (Figure 6 and Figure S4). For compound 2, the break occurred more frequently in the pair Y365 7.35 -S182 (Figure 5c). Indeed, in cluster 4, this ligand stabilized also a partially open state of the gate, retaining H349 6.55 in proximity of both Y365 7.35 and I183 (Figure 6a). Accordingly, the H349 6.55 -S193 5.43 distance was on average longer for partial agonists-bound receptor, with values fluctuating in the range of 4-10 (Figure 3c). Partial agonists could block progression toward a fully active state by preventing stable TM5-TM6 interface contacts. Thus, none of the simulated partial agonists was able to establish significant H-bond interactions between residues involved in interaction pattern 2. Our results agree with recent MD simulations of D3DR-ligand complexes, where bent poses have been associated with partial agonism. In line with these findings, in selective clusters 4 and 6 (Figure 2 and Figure S2) the distance between TM5 and TM6 increased, while tightening interactions at TM6-TM7 interface could be observed. The formation of the interaction pattern 1 in partial agonists simulations stabilized this inter-helical rearrangement. Inward rotation of the χ1 angle of the H349 6.55 in cluster 6 did not break pattern 1 (Figure 3a, e, 6b) and induced substantial side-bending of TM6. The analyzed binding modes were in very good agreement with the effect induced by the partial agonist FAUC350 on the same dihedral angle of H349 6.55 in the active ternary model of D2DR, where the ligand promoted coupling of TM6-7 and formation of the H349 6.55 -Y365 7.53 H-bond. Interestingly, this study reports that differential stabilization of inter-helical interaction patterns in the G-protein-bound model of D2DR is responsible for loosening intracellular coupling between the G-protein and D2DR, likely identifying structural patterns at the basis of partial agonism at the highly homologous D2DR. ## Agonist-induced conformations In analogy with partial agonists, the almost full agonist 4 (hereafter simply referred to as agonist) initially induced interaction pattern 1 in the representative bent conformation of cluster 7 (Figure 3a and Figure 7a), while also inducing TM6-TM7 coupling. The gating H-bond interaction between Y365 7.35 and S182 was not formed when 4 populated this cluster, again indicating that TM6-TM7 coupling promoted receptor opening (Figure 5c). Consistently, EL2 displacement was observed and this was, in turn, conducive to a pronounced increase in solvation (Figure S3 and Table S3). These results are in line with NMR studies on rhodopsin in which H-bond network reorganization between EL2 and TM4-6 has been coupled to EL2 displacement from the binding site during substrate-induced activation. In cluster 7, the pendant phenyl ring in the arylpiperazine moiety of 4 and the side chain of H349 6.55 formed an aromatic T-shape p-p interaction which was uniquely preserved by the agonist along the whole trajectory. Indeed, the distance between the two rings was stably preserved at approximately 5 (Figure 3b). The dihedral angle of S193 5.43 frequently rotated inward, around 160° (Figure 3d), i.e. the value observed for dopamine in the active model of D2DR. Our simulations revealed that interaction pattern 1 destabilized after 500 ns (Figure 3a) and the trajectory evolved toward the conformations that populate cluster 5 and cluster 8. These agonist-selective extended conformations caused the largest increase in the H349 6.55 -I183 distance (Figure 5a), which further allowed waters to reach the pocket (Table S3 and Figure S3). Besides the aromatic interactions with H349 6.55 , the phenyl ring of 4 established tight parallel stacking with F346 6.52 . While H349 6.55 side chain maintained its initial orientation (χ1 ≈ 60°, Figure 3e), F346 6.52 side chain underwent an exclusive rotation into the binding site in the opposite direction with respect to what was observed for 1, and pointed toward the intracellular region of the receptor. The ability to maintain stable interactions with H349 6.55 and F346 6.52 was a unique feature of 4. The ligand-induced rotameric state of F346 6.52 was observed only in response to the shift of the phenyl ring of the agonist at the TM5-TM6 interface, where it preserved an orientation perpendicular to membrane plane (Figure 7b). Inward rotation of F346 6.52 observed in our simulations minimized the steric hindrance that hampers TM6 inward motion, which is a crucial event in the destabilization of the receptor inactive state. The biphenyl ring interacted with Y365 7.35 pushing TM7 toward TM5-6. Furthermore, stable interactions of the phenylpiperazine ring of 4 with H349 6.55 at TM5-TM6 interface brought the histidine to point toward TM5 with higher frequency. In this conformation, H349 6.55 and S193 5.43 were only 2 to 4 apart, a rearrangement never observed in ligands with lower efficacy (Figure 3c and Figure 7b). The interaction pattern 2 first appeared in the trajectory around 1 µs, remained stable for roughly 100 ns, and was later transiently re-visited, as observed in Figure 3c (spikes in the blue line). The agonist, probably due to the persistent memory of the initial inactive state of D3DR, was not able to induce a complete rotation of H349 6.55 and S193 5.43 χ1 torsions toward -60° and 160°, respectively, which are the values that fully optimize the H349 6.55 -S193 5.43 interaction in dopamine-bound D2DR. However, even without stably preserving interaction pattern 2, in our simulations of D3DR in complex with 4, a series of agonist-specific changes in the interaction networks, which did not occur in antagonist-and partial agonist-bound complexes, could be observed. In particular, compound 4 selectively enriched bent and extended clusters, in which it was the only ligand able to preserve tight interaction with the H349 6.55 . In doing so, the agonist stabilized the interaction pattern 1 for 500 ns and also induced transient formation of the interaction pattern 2 for a maximum time of 100 ns, complementing the shift of H349 6.55 at TM5-TM6 interface. Such changes promoted abrupt opening of the extracellular portion of receptor and temporarily increased the contacts between TM5 and TM6. Maintaining aromatic interactions with F346 6.52 in a uniquely inward-rotated state contributed to this scenario, reducing the hindrance at the interface. ## Design, MD analysis and biological evaluation of the predicted compound 5 Taken together, our results provided the structural basis for understanding the varying efficacy of the 1-4 series. The 2,3-dichloro substitution and the butyl/(E)-but-2-ene linker were both needed to obtain a full antagonist, i.e. a molecule that was unable to promote any of the agonist-stabilized interaction patterns and that preserved D3DR in a closed configuration. Conversely, the removal of chloride atoms and the introduction of a flexible butyl linker led to an almost full agonist activity for 4. Compound 4 was able to uniquely establish long-lasting contacts with functional histidine in position 6.55, orienting this residue to establish both of the investigated interaction patterns, albeit to different extents. Interestingly, the partial agonist properties of semi-flexible 2 and 3 (bearing the 2,3-dichloro substitutions, but a saturated butyl linker) were explained according to their ability to stably induce only one of the intermediate interaction schemes, thus hampering but not completely blocking the cascade of events that concurred in the perturbation of the inactive receptor state. These results are in good agreement with the structure-efficacy relationship previously reported for other series of compounds. To gain confidence on this model, we designed a new compound introducing on the shared scaffold local modifications that, based on our understanding of the structureefficacy relationship, were likely conducive to a partial agonist profile. Two main aspects were considered in the design of the new ligand. First, our MD studies tried to provide a rationale for previous studies on a series of structurally related phenylpiperazine derivatives, which showed that the simple replacement of a butyl linker with a butyl/(E)but-2-ene in two identical ligands could transform an agonist into a partial agonist, and a partial agonist into a full antagonist. We reasoned that the introduction of a butyl/(E)but-2-ene linker in 4 could reduce ligand flexibility, and in doing so, it could prevent the stabilization of both patterns. Second, we reduced the molecular weight of the new compound by removing the carboxamide substituent in the meta position of the biphenyl group. Our SAR studies on O-aryl carbamate derivatives revealed that this substituent influences the affinity for the receptor but not the efficacy of the ligand. Insights gained from our simulations revealed that removing this moiety allowed 3 to sample a wider set of conformations, which, however, resulted in the induction of the interaction pattern 1, that is, the hallmark of partial agonists' profiles (see the Supplementary Material and Figure S2). The designed compound 5 is shown in Figure 1. The conformations obtained from 3.05 µs of MD simulations were then subjected to the previous analysis, merging the trajectories of compounds 1, 5, and 4 to assess how the new ligand conformations partitioned in the presence of our reference agonist and antagonist. The obtained clusters are reported in Figure 8. As in previous cases, the graph highlighted a hub cluster, cluster 0, which was almost equally populated by the three ligands (Table S4). In analogy with 3, thanks to the increased flexibility in the analyzed dataset (1, 5, 4), the hub node was characterized by extended and not bent conformations (Figure 8a). The featuring binding modes of 5 grouped in cluster 4 and cluster 6 (Figure 8b-c), with 9688 and 8429 conformations, respectively. An interesting difference between the partitions was the relative abundance of 5 in the agonist-selective cluster 7 (Figure 8). This was three times greater than 4 (6463 vs. 2173 members) (Table S4). In the previous analysis, this cluster contained only 569 conformations of 3 and 126 of 2. Our model identified this medoid as a crucial intermediate for establishing interaction pattern 1 in the agonist simulation. Interaction scheme 1 was never formed in cluster 4 (Figure 9a). It only appeared after 1.2 µs when 5 transitioned first to cluster 7 (refer to medoid in Figure 7a) and then to cluster 6 (Figure 9b). Indeed, this pattern was stably preserved until the end of the simulation (Figure 3f), with H349 6.55 -Y365 7.35 distance fluctuating in the range 2-4 . While the histidine kept a stable χ1 angle of 60° (Figure 3l), following the formation of the interaction pattern 1 and population of the agonist-like cluster 7, we observed inward rotation of S193 5.43 in TM5 (from 60° to 180°) (Figure 3i). In line with results on 2 and 3, the interaction pattern 2 was never formed in presence of 5 (Figure 3h), as the distance between H349 6.55 and S193 5.43 was stably over 4 for the whole trajectory. In clusters 4, 6 and 7, the receptor was found in an open state with TM6-7 being dynamically coupled through stabilization of the interaction pattern 1 (see also TM6-TM7 in Figure 8b-c). The average number of solvent molecules around the ligand was comparable to 2-3 and in between the values calculated for 1 and 4 (Table S3 and See Figure 2 for graph description. In the insets (a-c), representative medoids (cyan) of each cluster are shown in complex with their corresponding D3DR conformation and compared to the most populated "hub" medoid pose (gold) in cluster 0. Indeed, bending of compound 5 in cluster 4 caused the breaking of the H-bond between Y365 7.35 and S182. Later in the trajectory, the ligand stabilized in the bent orientation of cluster 6, where both gating interactions were broken (Figure 5b, d) and interaction pattern 1 was stably induced (Figure 3f and Figure 9) In analogy with 3, in cluster 6, the lack of the carboxamide substituent in the distal phenyl ring of the biphenyl group induced the ligand to drift deeper into the pocket (Figure 8c and Figure 9b), compromising the aromatic interaction between the ligand phenyl ring and H349 6.55 (Figure 3g). Ligand 5 interacted with F338 6.44 , F197 5.47 , and F346 6.52 at the interface of TM5-TM6, inducing outward rotation of F346 6.52 side chain and causing an increase in distance between these two helixes. To release the steric clash between F346 5.47 and F338 6.44 , TM5 slightly rotated clockwise, inducing EL2 to partially extend over the binding site (see relative orientations of TM5 and EL2 in cluster 4 to 6 in Figure 8b, c). These results are in agreement with the recently proposed mechanism of D3DR activation by a phenylpiperazine series of partial agonists. In contrast, recent MD studies on D3DR antagonists from Ferruz and colleagues have associated the outward rotation of F346 6.52 side chain to the formation of the cryptic pocket responsible for antagonist-like responses. Our observations suggest that the interaction of partial agonists, like 3 and 5 with this cryptic site could actually be responsible for antagonistlike properties, resulting in hampered activation of D3DR and partial agonism profiles. However, MD simulations of partial agonist 2 showed that this compound stabilized the rotameric state observed for F346 6.52 in the eticlopride-bound D3DR, suggesting that both conformations are likely to reduce the efficacy depending on the preferred bent and extended binding mode. In other words, we propose that an antagonist would likely elicit no response if an outward rotation of the F346 6.52 occurs in D3DR. When a rigid ligand binds the receptor in an extended mode, it blocks solvent access, preventing the formation of any interaction pattern. In contrast, in presence of semi-flexible partial agonists, which stabilize a more open conformation of the binding site, an outward rotation of the active site phenylalanine could still concur to destabilization of the inactive state. Interestingly, when our flexible agonist increased receptor solvation, assuming bent and extended conformations, F346 6.52 rotated in an unexpected direction, orienting its side chain toward the center of the helical bundle. Tight binding of H349 6.55 was also found to be uniquely preserved in the full agonist simulation, where both interaction patterns were visited. In contrast, this residue seems not crucial for antagonism. In partial agonists simulations, weak interactions with this residue were responsible for the induction of just one of the functional interaction schemes involved in D3DR activation. Importantly, the partition for compounds 1 and 4 was very robust. This is because, despite changing the initial set of conformations, medoid positions along the trajectory did not change. Indeed, we found that clusters 1, 3, 5, 7, and 8 were consistent with those already identified in the two previous sets of compounds. According to our model, compound 5 behaved as a partial agonist, destabilizing the inactive state, promoting receptor opening via TM6-TM7 coupling and binding site solvation. In line with other partial agonists, it induced only one of the two interaction patterns associated with agonist-like properties. To validate our hypothesis, 5 was synthesized and its experimental efficacy was tested in vitro for its ability to inhibit accumulation of cAMP. Compound 5 revealed partial agonist properties when compared to the effects elicited by 300 nM of dopamine (D3DR efficacy: 60%). Moreover, dropping the terminal carboxamide led to a D3DR 23 nM EC50. As expected, this is weaker than that of 2 and 3. However, it is a promising value in light of the reduced molecular weight. ## Dihedral PCA To further characterize our results, we also performed a retrospective compoundbased PCA analysis of torsional angles (Figure 10). Namely, we attempted to identify key torsional angles in the analyzed series of compounds. The dataset was obtained by extracting the values of three dihedral angles from 1-4 (Figure S1) over the entire 12 µs (120000 data points) of aggregate production runs. Next, data were binned in a reduced number of representative histograms. We performed a dimensionality reduction through dihedral principal component analysis (PCA) on this space and projected on the three main components to help visualize our ligands. Compound 5 was then embedded in the space identified by the four known compounds. The two distances obtained by averaging those from 1 and 4 were 1.43*10 4 and 1.46*10 4 for compound 2 and compound 5, respectively, making them equidistant from the two reference points. These results were robust to changes in the initial number of bins. Compound 5 actually stabilized in a specific orientation, which, as for 2 and 3, was somewhere in between the agonist and antagonist ones. This suggests that 5 could possess the same functionality of 2 and 3 at D3DR. Overall, although limited to a dataset of only four compounds, we found this (retrospectively applied) analysis useful in obtaining a quick and concise understanding of the simulations, recapitulating insights obtained with more complex analytical frameworks. The highly symmetric behavior of 2,3 and 5 when compared to 1 and 4 suggests that this could be an effective vector space where the efficacy profile of new compounds belonging to this series could be prospectively characterized. This also make us optimistic that a similar framework could be duplicated for other series and, possibly, other targets. ## Conclusions In summary, we used MD simulations to rationalize the experimentally observed efficacies of O-aryl carbamate derivatives. Our comparative analyses on the destabilizing effect of our ligands on the D3DR inactivated structure, we got evidence that structurally similar molecules can engage in subtly different interaction patterns and that these are, in turn, conducive to different efficacy profiles. The conformational changes reported from our simulations could be only related to the destabilization of the D3DR reference state, albeit known to be connected to initial steps in the concerted process of receptor activation. We found that the extent of these conformational changes was helpful in discriminating between ligand efficacies, and could therefore be of great help in designing a new ligand with a tailored pharmacological profile. Compound 5 was designed based on SER data to further test the consistency of the simulative outcome and was eventually synthesized and tested. As expected, this ligand behaved as a partial agonist. In due time, and in light of the ever-increasing computational power available to the scientific community, this work could pave the way to a more systematic application of MD as the in silico counterpart of functional assays, much as docking and free-energy methods can be regarded as the in silico counterpart of binding assays. ## Molecular Dynamics Setup The membrane protein complex was then solvated with an average of 12300 water molecules (TIP3P model). Force fields available in AMBER 14 were used to parameterize protein, lipids, and ligands, corresponding to ff14SB, lipid14, and GAFF, respectively. Point charges for ligands were derived from the electrostatic potential calculated after geometry optimization at the Hartree-Fock level of theory with a 6-31G* basis set, following the RESP procedure as implemented in Antechamber. Simulations were performed on GPU-equipped workstations with Gromacs 4.6.7 MD engine. In detail, the MD protocol encompassed three steps: minimization, equilibration, and production. Each system was minimized for 5000 steps and then thermalized to 300 K in different phases. Temperature was raised to 300 K in 300 ps within the NVT ensemble, in three consecutive increments of 100 K lasting 100 ps each. Then, volume and density were allowed to equilibrate in NPT ensemble at 300 K and target pressure of 1 bar for 200 ps. Lipids, ligands, and water molecules were equilibrated first, applying position constraints only to protein backbone (1000 kJmol -1 nm -2 ) in NVT steps. During the NPT equilibration, protein structure constraints were removed to allow relaxation at 300 K. Production runs were performed in NPT conditions with semi-isotropic pressure control, using Parrinello-Rahman barostat; temperature was kept at 300 K using v-rescale thermostat. A cut-off of 11 was used to switch off van der Waals interactions, while Particle Mesh Ewald was used to calculate electrostatics of the system, with a spacing of 1.6 . Finally, a 2 fs time-step was used to accumulate 3.05 µs of simulated time for each of the five systems, for a total of 15.25 µs. Dumping time was set equal to 100 ps. Our analysis covered the last 3 µs of collected statistics for each of the investigated complexes. ## Cluster Analysis We used a variant of k-means algorithm, namely k-medoids, as implemented in the BiKi LifeSciences suite. Generally, k-means generates an artificial mean structure, identified by coordinates that have minimal sum of squared deviations from a cluster center. The algorithm minimizes a distance-based cost function which is the sum of squared errors (SSE) as defined in Eq. (1): In Eq. ( 1), k is the number of clusters, n is the total number of conformations, x is the ith element of the cth cluster and m is the mean of the cth cluster. In each step, centroids are randomly chosen, closest objects are grouped around them, and SSE is calculated. Then, new arbitrary k medoids are chosen, clustering is performed again, and the new SSE is compared with the previous one in an iterative fashion until the difference between the previous and the present cycle cannot be further reduced, and medoid positions do not change anymore. While efficient in terms of computational time, k-means clustering is sensitive to outliers and to the initialization procedure for the random search of medoids. K-medoids is similar to k-means, as it is a medoid-centered algorithm, but instead of taking means as the centroid of the cluster, k-medoids assigns to centroids a physical meaning, identifying them as real objects in the data set. The new medoids are the most centrally located objects of each cluster. This modification introduces the possibility of returning to minimize the real sum of dissimilarities (distances) between the objects x in a cluster and their medoid m, which is a real representative conformation of the cluster. In other words, in Eq. ( 1), the difference is an absolute distance from a reference point and not a distance from the mean. Moreover, it overcomes some limitations of the classical k-medoids algorithm, which, as k-means-like algorithms, randomly select the initial medoids. This procedure affects computational efficiency and makes the results dependent on the choice of k. This k-medoids version provides a method to select the initial k medoids. The distances dij are first calculated between every pair of i and j objects, and a distance matrix is created once. Then, a variable for each j object, vj, is calculated as in Eq. ( 2): The values obtained for each j object are sorted in ascending order and the first k medoids with minimal values of v are considered as initial cluster centers. This makes the algorithm deterministic because the initial k medoids are always those that minimize the total distance to all other objects i.e. the most central ones. Also, medoids are updated, finding a new one for each cluster that minimizes the total distance to any other objects in the cluster. The main advantages of this procedure are the ability to work with medoids that can be associated with sampled conformations, and to use an objective function based on absolute distances to refine the quality of the clustering. Moreover, the algorithm is robust to outliers because the most centered conformations are selected as the initial 5 medoids. In our systems, trajectories were concatenated based only on the common scaffold shared by the five ligands (Figure S1). The choice of the representative medoids was performed based on RMSD-based threshold between medoids in a given partition. The number of clusters was considered meaningful of sampling diversity if the difference between the medoids was more than 3. Automated column chromatography purifications were performed using a Teledyne ISCO apparatus (CombiFlashTM Rf) with pre-packed silica gel columns of different sizes (from 4 g to 120 g). ## General methods and abbreviations Mixtures of increasing polarity of cyclohexane and ethyl acetate or dichloromethane and methanol were used as eluents. Preparative TLCs were performed using Macherey-Nagel pre-coated 0.05 mm TLC plates (SIL G-50 UV254). Hydrogenation reactions were performed using H-CubeTM continuous hydrogenation equipment (SS-reaction line version), with disposable catalyst cartridges (CatCartTM) preloaded with the required heterogeneous catalyst. Microwave heating was performed using ExplorerTM-48 positions instrument (CEM). NMR experiments were run on a Bruker Avance III 400 system (400.13 MHz for 1H, and 100.62 MHz for 13C), equipped with a BBI probe and Zgradients. Spectra were acquired at 300 K, using deuterated dimethylsulfoxyde (DMSO-d6) or deuterated chloroform (CDCl3) as solvents. Chemical shifts for 1H and 13C spectra were recorded

chemsum

{"title": "Multi-Target Dopamine D3 Receptor Modulators: Actionable Knowledge for Drug Design from Molecular Dynamics and Machine Learning", "journal": "ChemRxiv"}

photocrosslinking_and_photopatterning_of_magneto-optical_nanocomposite_sol–gel_thin_film_under_deep-

## Abstract: This paper is aimed at investigating the process of photocrosslinking under Deep-UV irradiation of nanocomposite thin films doped with cobalt ferrite magnetic nanoparticles (MNPs). This material is composed of a hybrid sol-gel matrix in which MNP can be introduced with high concentrations up to 20 vol%. Deep-UV (193 nm) is not only interesting for high-resolution patterning but we also show an efficient photopolymerization pathway even in the presence of high concentration of MNPs. In this study, we demonstrate that the photocrosslinking is based on the free radical polymerization of the methacrylate functions of the hybrid precursor. This process is initiated by Titanium-oxo clusters. The impact of the nanoparticles on the photopolymerization kinetic and photopatterning is investigated. We finally show that the photosensitive nanocomposite is suitable to obtain micropatterns with submicron resolution, with a simple and versatile process, which opens many opportunities for fabrication of miniaturized magneto-optical devices for photonic applications.Magneto-optic (MO) effects refer to phenomena which modify the light polarization according to an external magnetic field applied to a MO active material 1-3 . One of these MO effects is the Faraday rotation (FR), which is defined as the change produced in the plane of polarization of the light transmitted through a material when a magnetic field is applied: the plane of polarization is rotated. In reflexion mode, this effect is known as the magneto-optic Kerr effect. Amongst the most attractive properties of the magnetic transparent compounds are those related to the magneto-optical (MO) effects and their scientific and industrial applications in areas such as data storage 4 , three-dimensional (3D) imaging 5 , magnonics 6,7 , sensing 8,9 and photonics [10][11][12] . MO active materials are ubiquitous in photonic devices, but they are still lacking in integrated photonic platforms although they are essential components for optical communication systems (optical isolators, optical circulators, optical switches, magneto-optic (MO) modulators) [13][14][15][16][17][18] and high performance magnetic field sensors [19][20][21] .In the past, magneto-optical materials have been mainly developed by physical methods such as sputtering 22,23 , pulse laser deposition 24 , and molecular beam epitaxy [24][25][26] . More recently, nanocomposite approach has been proposed to simplify the synthesis of the material and its shaping into devices 27 . The principle is based on the synthesis of magnetic nanoparticles then their incorporation into a matrix which can be a polymer [28][29][30][31][32][33] , or an inorganic matrix prepared by sol-gel for example 34,35 . The materials can thus be processed by the usual processes adapted to liquid formulations. Solution-based processes appeared with the major advantage of allowing simpler processes without the need of sophisticated equipment 36 . In this context, sol-gel chemistry emerges as a very relevant and versatile solution. The principle of the sol-gel process is based on successive reactions of hydrolysis and condensation of precursors that form networks of metal oxides. One of the interests of this route lies in its compatibility with a large number of precursors. Among numerous commercially available precursors, alkoxides (alkoxysilanes zirconium, titanium, aluminium, etc.) are the most commonly used. Other derivatives are also used (chlorides, nitrates, …). The sol-gel pathway thus offers a number of advantages in terms of energy harvesting, processing versatility and a wide range of final properties . The solutions can be deposited as thin films by simple means such as spin-coating, dip-coating or spray-coating. Moreover, the final material can reach good transparency, good mechanical properties and high refractive index. These advantages account for a wide use of sol-gel coating for optical applications. We recently introduced a process that relies on the synthesis of a photocurable sol-gel matrix in which magnetic nanoparticles (MNP) can be introduced 40 . Cobalt ferrite (CoFe2O4) MNPs were chosen because they exhibit large Faraday rotation in the 1400-1550 nm spectral range 41 , an important criterion for potential applications in telecommunication devices and photonic integrated circuits. The major advantage of using MNPs incorporated in a non-magnetic host matrix is that the magneto-optical properties are then obtained in the final material without the need for additional thermal post-treatment. The main challenge is to manage and avert the NPs aggregation during the different steps of the process in order to preserve the magnetic properties of the individual MNPs and to avoid light scattering by aggregates in the nanocomposite. The photocurable sol-gel matrix was based on a hybrid precursor that can be crosslinked by light (namely 3-methacryloxypropyltrimethoxylsilane MAPTMS). Photocrosslinking of such matrix can be obtained by Deep-UV laser irradiation (193 nm), which opens the doors towards micro and nanopatterning. Indeed, solubility switch can be induced by laser irradiation which allows defining by laser direct writing the patterns. Other materials and patterning strategies have been proposed. In 42 , Lai et al. prepared a photopatternable material based on a commercial SU8 photoresist doped with magnetite nanoparticles and achieved high resolution patterns and 3D structures by laser direct writing. In the present paper, we aim to investigate the photocrosslinking mechanism involved in the nanocomposite and further exemplify the application for photopatterning in the sol-gel matrix that we developed. The optimized synthesis of CoFe 2 O 4 nanoparticles and hybrid sol-gel host matrix is first described. Conditions are defined to reach high concentration of MNPs. The kinetic of photopolymerization is then studied by Fourier transform infrared (FTIR) spectroscopy, in order to investigate the photoinduced mechanisms allowing the photocrosslinking of the material. In particular, we highlight an original photo-induced mechanism based on the excitation of Ti complexes. On this basis, a mechanism can be proposed. In a second part, DUV photopatterning is demonstrated and the effect of the composition on the final results is discussed. ## Experimental Synthesis of magnetic nanoparticles. The magnetic nanoparticles embedded in the matrix are cobalt ferrite nanoparticles (CoFe 2 O 4 ). The preparation of these nanoparticles was performed by precipitation of cobalt chloride and ferric chloride, based on the method of Massart and Tourinho 43 . The different stages of ferrofluid synthesis can be summarized as follows: First, iron chloride (FeCl 3 ) and cobalt chloride (CoCl 2 ) are mixed in aqueous solution, with a molar ratio Fe/Co equal to 2. Concentrated NaOH (10 mol/L) is then added to the mixture to form the hydroxides of each metal (Fe(OH) 3 and Co(OH) 2 ). The solution is then heated to 100 °C for two hours to convert the hydroxides into cobalt ferrite. After several washing steps in water, an acid treatment is applied (overnight) by adding a 2 mol/L solution of nitric acid. This acid treatment removes hydroxides that would not have been converted into cobalt ferrite. It also makes it possible to shift from a negatively charged surface (charge due to Ogroups on the surface with Na + counter-ions) to a positively charged surface. As cobalt ferrite particles are not stable in an acidic environment, it is necessary to protect them by a surface treatment. This treatment is carried out by adding an almost boiling solution of ferric nitrate (concentration 0.3 mol/L) to the solution containing MNP. Finally, to obtain the ferrofluid, washing steps with acetone and then ether were done and the MNP of cobalt ferrite are dispersed in water. The resulting ferrofluid is an acidic ferrofluid: the MNPs are positively charged at the surface with NO 3 − . The concentration of iron and cobalt ions was determined by atomic absorption spectroscopy method after degradation of the MNPs in a highly concentrated acidic medium. The concentration was confirmed by recording the magnetization curve of the ferrofluids, measured with a vibrating-sample magnetometer (VSM, Quantum Design PPMS). The size dispersion and morphology of the particles were determined by analysis of images obtained by transmission electron microscopy (TEM). Transmission electron microscopy (TEM) was performed using a JEOL ARM-200F microscope operating at the 200 keV accelerating voltage. The chemical analysis was performed using a JEOL Centurio detector. The samples were prepared for observation using a Leica ultramicrotome, model EM-UC7, operating at RT. To observe the film cross-section, it was deposited on the surface of a thermanox substrate. The slide thickness was ~ 100 nm. Highest concentration in MNPs in the final materials (> 2 vol%) were reached by using a concentrated ferrofluid, obtained by dialysis. This technique consists of concentrating the ferrofluid by reverse osmosis, by immersing a porous dialysis tube filled with the ferrofluid to be concentrated in an aqueous solution of polyethylene glycol (PEG) with a mass concentration of 20,000 g/mol. Typically, dialysis of 20 mL of ferrofluid initially concentrated at 1.45 vol% for 24 h with agitation results in a highly viscous ferrofluid with an MNPs concentration of about 10 vol% (corresponding to a concentration of a factor of ~ 7). By introducing dialyzed ferrofluids into a matrix, it has been possible to achieve doping levels of up to 20 vol%, which is comparable to the highest level of MNPs in nanocomposites described in the literature 44 . The advantage of using a dialyzed ferrofluid is to introduce more magnetic nanoparticles without adding more water to the mixture. ## Sol-gel matrix and nanocomposite solution. 3-methacryloxypropyltrimethoxylsilane (MAPTMS), titanium isopropoxide (TTIP), methacrylic acid (MAA), Hydrochloric acid (HCl), 1-propanol, and cyclohexanone were purchased from Sigma Aldrich. Deionized water was used throughout the reactions. The main stages of material preparation are shown in Fig. 1. It consists in 5 main steps: 1. 0.01 mol/L concentrated hydrochloric acid is added to MAPTMS (molar ratio MAPTMS/ H 2 O of 1.3:1). The solution is placed under magnetic stirring for one hour to give a colorless emulsion. The objective of this step is to pre-hydrolyze the MAPTMS 45 and favor a homogeneous introduction of Ti precursor. 2. Methacrylic acid MAA is added to Ti precursor in a molar ratio MAA/Ti = 2.2. This molar ratio was chosen to have a large excess of MAA to ensure a full complexation of Ti precursor. Indeed, according to the literature, a minimum ratio of 1.2 is necessary for total complexation of the titanium precursor 46,47 . The mixture is magnetically stirred for five minutes. This reaction is exothermic. Subsequently, 1-propanol is added to the solution in a ratio of 0.9 molar MAA/1-propanol. The solution is homogenized with a magnetic stirrer for 10 min. A clear, bright yellow solution is obtained. 3. The solutions of MAPTMS and complexed Ti precursor are mixed together. Water is then added with a molar ratio MAPTMS/H 2 O = 0.4. This third step completes the hydrolysis-condensation steps of the sol-gel chemistry. 4. A given volume of ferrofluid containing the MNPs is added to the sol-gel solution. The doped material is homogenized by ultrasonic treatment. The stirring time varies from ten minutes to one hour depending on the amount of MNP introduced. The introduction of MNP has been considered at different stages of the process. The most stable formulations are obtained when the introduction is made at the end of the sol-gel formulation preparation steps and this process was retained (Fig. 1). 5. The doped solution is diluted with 1-propanol. The purpose of this dilution is twofold. On the one hand, the viscosity of the material must be adjusted, which will make it possible to change the thickness of the thin film during spin-coating. On the other hand, dilution ensures the stability and homogeneous dispersion of MNP in the sol. Dilution is particularly important in the case of heavily doped matrices. The molar ratio of Si/Ti can be varied in the range 3/1 to 20/1. The reference matrix corresponds to a composition of Si/Ti in a molar ratio of 6/1. Solutions prepared are stable for several months and can be stored at room temperature. www.nature.com/scientificreports/ Thin films preparation and characterization. Substrates (Silicon wafer or glass slides) were first cleaned by rinsing with ethanol and then placed in an UV-ozone cleaner to remove the organic pollutant and increase the polarity of the substrate for good adhesion of the thin film. The formulation was filtered through 0.2 μm PTFE filters. Homogeneous films were obtained by spin coating, with typical thicknesses between 200 and 500 nm, depending on the dilution factor and rotation speed during deposition. The photopolymerization kinetics were followed by real time-FTIR with a Thermo Scientific Nicolet 8700, coupled with a Hamamatsu high intensity mercury-xenon lamp equipped with a light guide (Lightnincure series LC8 lamp). Si wafers (thickness = 0.25 mm) were used as substrates and the irradiance was fixed to 70 mW/cm 2 . With this configuration, the sample can be irradiated in situ, which is more convenient than the ex situ laser irradiation and justifies that we used the DUV lamp for polymerization kinetics. Absorption measurements were performed with a Lambda 950 UV/Vis (Perkin Elmer). The Faraday spectra were acquired in the wavelength range 600-1700 nm with a homemade polarimetric optical bench based on the modulation technique combined with an ellipsometric-type calibration method (see 48 for details). The sample (nanocomposite film or ferrofluid cell) is placed perpendicular to the incident beam in the air gap of an electromagnet. The magnetic field can be varied in the ± 0.8 T range. The light from a xenon white light source combined with a monochromator passes through a polarizer, the sample, a photoelastic modulator, an analyzer, a detector, and a lock-in amplifier (LIA). This optical arrangement is suitable for analyzing the polarization state by means of the first (ellipticity) and second harmonic (rotation) signals of the LIA 49 . The calibration method 50,51 allows to measure the absolute value of the polarization rotation with the detection limit of 0.001°. Photopatterning. The photopatterning setup relies on a Braggstar (Coherent) nanosecond ArF Excimer laser emitting at 193 nm 40 . The beam section measures 3 × 6 mm 2 . An attenuator located after the laser is used to tune the power. A shutter allows controlling the exposure time. A beam expander makes it possible to enlarge the beam spot by a factor of 5 and thus increases the exposed material surface. Its role is also to homogenize the beam, reduce its divergence and thus increase spatial coherence. At the exit of the expander, the beam reaches a semi-reflecting blade allowing 25% of the power to pass through and returning the remaining 75% of the power to the sample at 90°. The 25% passing through the blade allows the laser power to be measured in real time using a power meter. The sample is placed on a motorized stage in x, y and z. Displacements in x and y position the sample under the beam and the z stage adjusts the sample-interferometer distance (i.e. the focal point). Measurements of the films thickness were done by ellipsometric spectroscopy. The measurements were performed on a UVISEL ellipsometer from Horiba-Jobin-Yvon (spectral range 190-830 nm). Data were fitted with the software from the UVISEL ellipsometer. The photopatterned films were characterized by Atomic Force Microscopy (AFM), in tapping mode, with a PicoPlus 5500 System model from Agilent. ## Results and discussion Structural, magnetic and magneto-optical properties of Cobalt ferrite CoFe 2 O 4 nanoparticles were first studied. Nanoparticles were characterized by TEM. A typical TEM image is presented in Fig. 2a. The NPs size distribution fitted by a log-normal distribution gives a NPs average size of 8.4 nm with a standard deviation σ = 0.6 (Fig. 2b). The VSM hysteresis curve of the ferrofluid (given in Fig. 2c) shows the classical behavior of superparamagnetic nanoparticles. The magnetization is 800A/m and the concentration of Fe and Co ions measured by flame spectrophotometry is [Fe + Co] = 1.8 mol/L. Ferrofluids prepared under these conditions have excellent stability (several months). Cobalt ferrite nanoparticles have been selected for their interesting magneto-optical properties. The nanocomposite material is intended for use in MO devices operating at the wavelength 1550 nm. For this reason, the spectral response of the Faraday rotation of ferrofluid has been investigated. Figure 2d shows the Faraday rotational spectral behavior of the ferrofluid synthesized according to the protocol described above. The Faraday rotation of the CoFe 2 O 4 MNPs is particularly important around 750-850 nm, as well as in the 1500-1600 nm region, where it reaches values higher than 200°/cm. These CoFe 2 O 4 MNPs are therefore good candidates for obtaining MO properties in these two operating windows, at especially at the 3rd telecom wavelength (1.525-1.625 nm), interesting for telecom applications. The magneto-optical response of the thin-film material after Deep-UV curing is also shown in Fig. 2d to be compared to the response of the ferrofluid used to incorporate the MNPs. The evolution of the Faraday rotation is measured at normal incidence of a thin layer (Si/Ti: 6/1), doped to 2.2 vol% in nanoparticles. The sample thickness was 1.9 μm, deposited on a glass substrate. It was obtained by deposition of 4 layers of 480 nm, deposited and subsequently irradiated with an energy of 1.5 J/cm 2 by the Excimer laser, which is enough to ensure that the film is dry after exposure. Crosslinking molecular phenomena will be discussed below. The thickness of MO material allows to improve the signal-to-noise ratio. As shown in Fig. 2d, the Faraday rotation of the composite material has the same spectral behavior as that of NPs ferrofluide. This result confirms that the integrity of the nanoparticles is preserved during their incorporation into the sol-gel matrix and the UV cross-linking of the thin film. The DUV cured material presents thus interesting MO properties. The structural characterization of the nanocomposite material was completed by a TEM and EDX analysis (Fig. 2e). The TEM image illustrates the good repartition of the MNPs (bright spots) within the sol-gel matrix. Si and Ti are well distributed within the film thickness, showing that the synthesis strategy is suitable to obtain an homogeneous material, which is needed for optical applications. The next sections are aimed at investigating the Deep-UV induced modification of the material leading to the crosslinking of the material and then to show how this molecular behavior can be used to direct laser write micro-and nanostructures with MO properties. As mentioned before, we observed that the film deposited with spin-coating from the sol-gel solution doped with CoFe 2 O 4 MNPs could be crosslinked by laser DUV irradiation. After laser irradiation, the film is tack-free www.nature.com/scientificreports/ and resistant to etching by organic solvents such as alcohols or cyclohexanone. As stated in the introduction, there is no study of the photoinduced phenomena in such nanocomposite materials, which is shown hereafter. Figure 3a displays the typical evolution of the FTIR spectrum of a nanocomposite thin film (450 nm thickness, Si/Ti = 6/1, 0.4 vol% of MNP) under DUV irradiation. Before irradiation, the absorption bands were assigned according to previous works 52 : • The 1638 cm −1 band corresponds to the vibration modes of the sp2 carbons (C=C double bonds) of the material. Its presence shows that the C=C bands are not affected by the sol-gel reaction and thus available for forming the polymer network by photopolymerization. • C=O from carboxylic acids and methacrylic functions appears at several wavenumber, according to their environment: free C=O from the methacrylate are visible at 1740-1700 cm −1 . This position is expected for a methacrylate function, which confirms that the methacrylate can be polymerized. A second band located between 1500 and 1550 cm −1 corresponds to the vibrations of the C=O (methacrylic acid) complexed with titanium. This band shows the presence of the complex, formed in the early stages of the sol-gel synthesis, in the thin film after spin-coating. • The sol-gel reaction is also confirmed by the presence of bands in the region 800-1250 cm −1 that can be assigned to Ti-O, Si-O and combinations thereof. Figure 3a also shows the evolution of the FTIR spectrum during DUV irradiation. Several modifications were recorded during irradiation. The most obvious change appears at 1638 cm −1 (enlarged in Fig. 3b). This band gradually disappears during the irradiation, showing the consumption of the C=C double bounds. This demonstrates the polymerization of the methacrylate functions and explains the crosslinking of the material under DUV irradiation. The conversion of the C=C can be plotted as a percentage of the C=C consumed in reference to the initial quantity of C=C bounds (Fig. 3c). The shape of the curve is classical with a maximum polymerization rate (defined as the slope of the curve) at the beginning of the irradiation and a progressive decrease of the polymerization rate with time up to a maximum conversion ratio. It confirms the very good yield of polymerization achievable with this system, despite the presence of the MNP, with a total final conversion close to 100%. The excellent final conversion is important to guarantee good mechanical and optical properties to the thin film. www.nature.com/scientificreports/ Additionally, we confirmed that the decrease of the C=C double bounds (1638 cm −1 ) is not due to the loss of volatile compounds (free methacrylic acid for example). For this purpose, we followed also the evolution of the C=O bound. We observed a shift of the position of the C=O. This can be explained by the loss of the conjugation between the C=O and the C=C as the C=C is consumed by polymerization. If the area of the corresponding band is plotted versus time, the value keeps constant during polymerization, showing that there is no significant loss of material during photopolymerization. One of the reasons explaining the excellent conversion yield of the C=C bounds in the nanocomposite is linked to the limited contribution of the MNPs to the light absorption through the film at the irradiation wavelength. Figure 4a illustrates the value of absorbance and transmittance that were determined by UV spectroscopy at a wavelength (210 nm) close to the irradiation wavelength (193 nm), for several concentrations of MNPs. The data were collected from samples with different thicknesses and the optical properties were calculated for a film thickness of 100 nm. The increase of the concentration of the MNPs only slightly increases the absorption at the irradiation wavelength, the absorption being mainly linked to the host matrix. For the highest load of MNP (20%), the contribution of the MNPs to the total absorption is about 30%. Interestingly, absorption is not linear towards MNP concentration, which can be explained by the partial aggregation of the MNPs when the concentration is increasing. The impact of the MNP on the polymerization kinetic was evaluated for several film thickness (Fig. 4b,c). Interestingly, for the lowest MNP concentration (0.4 vol%), the photopolymerization kinetics is only slightly dependent on the film thickness and the polymerization rate and final conversion are very good for thicknesses up to 530 nm. At higher MNP concentration (10 vol%), the polymerization rate is decreased but the final conversion (80%) is high enough to insure good adhesion and mechanical properties of the thin film directly after irradiation, without any further curing. This result confirms that DUV irradiation is effective to trigger the photopolymerization of the nanocomposite. For the maximum MNP concentration achievable (20 vol%), the final conversion was 75% (for 70 J/cm 2 ). As mentioned before, the photopolymerization process under deep-UV irradiation (193 nm) relies on the crosslinking of the organic part of the hybrid nanocomposite material. However, since there is no organic photoinitiator added in the matrix to start the polymerization, there are questions arising about the light induced mechanism accounting for the photocrosslinking. As proposed in previous studies 46,53 , metal alkoxides, when exposed to DUV light, can decompose to produce free radical species that are able to start the free radical polymerization of the hybrid matrix. We would like to confirm this mechanism for the nanocomposite material. Figure 5 displays the influence of several parameters involving the Ti complexes in order to demonstrate their central role in the photopolymerization mechanism. We first investigate the influence of the concentration of metal alkoxide on the photopolymerization kinetic (Fig. 5a). In Fig. 5a the photopolymerization kinetic was followed by FTIR spectroscopy, with the same method as before (conversion calculated from the decrease of the C=C band at 1638 cm −1 ). This graph clearly shows that the photopolymerization is accelerated when the concentration of metal alkoxide is increased. This confirms the central role of the metal alkoxide complexed with methacrylic acid in the photopolymerization process. We www.nature.com/scientificreports/ observed that there is no significant improvement of the polymerization kinetics between 6/1 and 3/1 for Si/Ti ratio, which justifies the atomic ratio Si/Ti = 6/1 used in this study. Moreover, we observed that the stability of the formulation was not guaranteed after 10 days at highest load of Ti. One reason may be due to ligand exchange on the MNP surface by free methacrylic acid that is used to stabilize the metal. Also, we recorded similar kinetics after adding the MNP at low concentration (0.4 vol%), for all Si/Ti ratio, which means that there are no significant interactions between the MNP and the Ti complex acting as a photoinitiator. Zirconium was also evaluated as a metal for photocatalyst, instead of Ti. (Fig. 5b). With Zr as a metal, the polymerization proceeds with a rate equivalent to the formulation with a very low load of Ti, which means that the photoinitiating efficiency of the Zr complex is much lower than the one of the Ti complex. Since the absorption of both Ti and Zr complexes were found to be close in the DUV range, we concluded that this difference in reactivity can be explained by a difference in redox power between the two species. The interest of Ti as a metal precursor to induce the free radical polymerization of the hybrid sol-gel is thus demonstrated. Finally, the nature of the ligand used to complex the titanium alkoxide precursor was also investigated. Figure 5c presents the FTIR kinetic study of the polymerization for different ligands in Ti complexes. Four different ligands were used to complex TTIP: three carboxylic acids (acrylic acid, isobutyric acid and methacrylic acid) and a β-diketone (acetylacetone). In each case, the ligand/Ti ratio was maintained at 2.2 molar % (without MNPs). No significant difference in polymerization kinetics was observed for the different complexing agent, with comparable polymerization rate and final conversion. The ligand chosen to complex the titanium therefore has little influence on the polymerization of the composite formulation. In Fig. 5d, we propose a simplified mechanism to summarize the role of the Ti-complexes in the elaboration of the nanocomposite material. It is admitted that the chosen experimental conditions lead to the incorporation of the Ti complexes as Ti-oxo clusters, as schematized in Fig. 5d. 42 The decrease of the band located between 1500 and 1550 cm −1 that corresponds to the vibrations of the C=O (methacrylic acid) complexed with titanium demonstrates the photolysis of the Ti-oxo clusters under DUV. This result is relevant with the results proposed in previous study, for comparable cluster, but in different conditions. The reactive species are able to trigger the free polymerization of the acrylate functions. The very good polymerization yield suggests a free radical mechanism. Note that the free radical formation of such clusters under DUV irradiation was already proposed in previous studies 46 . Such a property of the Ti complexes suggests that there is no interest to add any organic photoinitiator into the formulation to improve the photopolymerization efficiency. Several commercial photoinitiator known for their efficiency in free radical photopolymerization were added with a concentration of 2 wt% (Irgacure 184, 369 et 819 from Ciba). The corresponding polymerization kinetics revealed only a slight increase of the polymerization rate. This improvement is minor and thus the addition of an organic photoinitiator inside the formulation is not justified. In this final part, we discuss the potential of this formulation to be used as a negative tone photoresist to produce sub-micrometric patterns at room temperature with magneto-optical properties. Indeed, as shown before, a very efficient Photocrosslinking can be obtained by DUV irradiation, which can be used for photopatterning the MO material. For this purpose, two home-made photolithography setups were used, as depicted in Fig. 6, in order to show the performance and versatility of this material for photolithography applications: for the lowest resolutions (typical feature lateral size superior to 1000 nm), a proximity printing setup was used, consisting in binary masks (chromium patterned deposited on fused silica substrates) placed close to the sol-gel film (Fig. 6a). In this configuration, the metal lines cut the DUV lights and prevent the Photocrosslinking reaction to occur. For highest resolutions, in order to overcome diffraction problems, an interferometric lithography setup was used (Fig. 6b). With this setup, the period of the patterns can be varied accordingly to the phase mask used. In the present study, we focused on patterns with period of 500 nm generated by a phase mask having a period of 1000 nm. The interference approach allows higher resolutions but, in this case, the light pattern is sinusoidal, since the contrast is generated by the interference between the two diffracted beams. In this configuration, there is thus no 0-light area, which can have some consequences on the shape of the structures, as shown later. The coatings were prepared by spin-coating on a substrate cleaned with UV-ozone cleaner. The spin-coating rotation speed and dilution factor of the solution were adjusted to obtain the desired thickness. After irradiation with one of the two configurations described above, the sample is directly developed in a solvent to dissolve the non-irradiated parts. The material behaves like a negative resin, with irradiated parts becoming insoluble. The nature of the solvent as well as the development time have been optimized. Water cannot dissolve the unexposed parts thus cannot be used as a developer for this material. Alcohols (ethanol, methanol) and acetic acid are too strong developers and dissolve the exposed parts and are thus not suitable neither. Cyclohexanone was proved to constitute a good candidate and was chosen in the following as a solvent for development. Well-defined structures could be obtained after 10 s. development in cyclohexanone. No thermal annealing is required after development for stabilization of the sample as a post-treatment. We observed that the time between sample preparation and irradiation shall not exceed 10 min, otherwise, the development is more difficult to carry out. This is due to the condensation reaction that can occur at room temperature, because of atmosphere moisture. This condition is not limiting since the typical irradiation times are shorter than this value (few sec. to few tens of sec.). Figure 6c,d show typical patterns obtained in both photopatterning configurations. In both cases, photopatterning could be demonstrated. For the proximity printing lithography (period 1600 nm with line width of 800 nm), well-defined patterned with free substrate between lines and low line edge roughness could be demonstrated (Fig. 6c). The patterns height was 125 nm for a deposited film thickness of 150 nm. We attributed the loss of height to the shrinkage occurring in the material upon DUV irradiation (partial loss of organic moieties). These results illustrate that patterns with width of 1000 nm or more can be obtained. In interference lithography, patterns were obtained with higher resolutions but as shown in example in Fig. 6d, the pattern height was lower. Indeed, in this example, though the film thickness was decreased to 80 nm, patterns height was only 50 nm. This result was interpreted as a residual layer remaining between written lines. In order to further investigate the behavior of the material in these conditions, a systematic study of the DUV photopatterning was conducted. Results are exposed in Fig. 7. The period of the patterns in Fig. 7 was 500 nm (corresponding to a line width of 250 nm with a space of 250 nm). The heights of the structures were measured by AFM. They are plotted in Fig. 7e. Figure 7a-e show the AFM images of surfaces irradiated with respectively 2.5, 5, 7.5 and 15 mJ/cm 2 . Figure 7f gives a schematic interpretation of the evolution of the pattern structure with the irradiation dose. For the lowest doses (less than 2.5 mJ/cm 2 ), no pattern was observed. We explain this response by a too low conversion within the irradiated parts and thus the crosslinking of the matrix is not enough to promote the adhesion of the material on the substrate during development (case i) in Fig. 7f). This behavior corresponds to a too low conversion of the C=C bond in the organic part of the hybrid matrix, especially at the resin-substrate interface due to the internal filter effect. At 2.5 mJ/cm 2 , the conversion of the organic matrix is sufficient to give rise to a very thin layer of crosslinked material at the substrate surface. It explains why in Fig. 7a, patterns are observable but with a height much smaller (a few nm) than the initial film thickness (80 nm). From 2.5 mJ/cm 2 to 5 mJ/cm 2 , the measured height increases rapidly with the dose as crosslinking proceeds more and more efficiently in the photoresist. However, the case depicted in Fig. 7f-iii is never reached since the maximum height (50 nm, Fig. 7b) was always significantly lower than the initial film thickness (80 nm). This explains the apparition of a residual layer between lines (Fig. 7f-iv) due to the irradiation in dark fringes of the interference pattern. This assumption is confirmed by the gradual decrease of the pattern height with dose (Fig. 7c,d) that corresponds to the increase of the thickness of the layers between lines. Such behavior is partially explained by the interference pattern irradiation configuration. Indeed, one of the drawbacks of this configuration is that the light intensity is sinusoidal so it is not null in the dark fringes. As shown in a previous study 40 , the concentration of the MNP has an impact on the photopatterning. Patterns with various concentrations of MNP (between 0 and 20 vol%) were prepared using interference lithography (period 600 nm). Photopatterning can be obtained in this wide range of MNP concentration but there is a strong impact of the MNP load on the quality of the patterns and their height. In particular for the highest loads of MNPs we observed the apparition of roughness at the sample surface after development that can be linked to partial aggregation of the MNPS that occurs at the surface of the patterns. These aggregated nanoparticles may create bridges between close structures, which may account for the difficulty to conduct development in these conditions, which results in remaining material between lines. In Fig. 8, we plotted the typical maximal height and optimal dose for the different MNP concentrations. Interestingly, Fig. 8 reveals that the optimal dose is increased, as expected, with the content of MNP, but only slightly in a 80 nm thin film. In conclusion, photopatterning with submicron resolution is achievable with MNP concentrations as high as 20 vol% but in this case, www.nature.com/scientificreports/ a significant decrease of the pattern modulation, due to the presence of a residual layer between written lines is observed. However, for many applications as gratings or in guided optics, such residual layer is not a problem for practical applications since it can be taken into account in the design of the optical design to produce a given optical function. Finally, in Fig. 9, we show the impact of the Ti complex concentration to confirm that the patterning is indeed triggered by the Ti complex, as suggested by the polymerization kinetic studies shown previously, and to evaluate the impact of the Ti complex concentration on the patterns. Two concentrations of Ti complexes were used (Si/Ti = 3/1 and 6/1), with the same concentration of MNP (0.4 vol%). In both cases, a residual layer between lines is obtained after photolithography. As expected, the dose needed to achieve the photopatterning of the material is lower for the higher content of Ti, which confirms the role of Ti complex as a photoinitiator of the crosslinking reaction within the material. However, the maximum pattern height was obtained for the lower Ti content, which confirms the interest to use a molar ratio Si/Ti = 6/1, as mentioned previously. Increasing the content of Ti allows decreasing the exposure time for a given power but finally, no significant improvement of the maximum height was observed. In conclusion, we have shown in this paper that DUV photolithography (193 nm) is an extremely interesting tool for the micro and nanostructuring of thin films with magneto-optical properties. Starting from solutions whose composition can be easily adapted to modulate the properties, the DUV photolithography step allows to cross-link the material and to structure it at submicrometer scales. No additional step (in particular no thermal annealing) is required to obtain the magneto-optical properties, which opens perspectives for the integration of these materials in devices, on glass, on silicon, but also on plastic.

chemsum

{"title": "Photocrosslinking and photopatterning of magneto-optical nanocomposite sol\u2013gel thin film under deep-UV irradiation", "journal": "Scientific Reports - Nature"}

leaded_aviation_fuel_may_present_long-term_effects_on_campus_life_from_the_adjacent_albert_whitted_a

## Abstract: Propeller planes and small engine aircraft around the United States, legally utilize leaded aviation gasoline. The purpose of this experiment was to collect suspended particulate matter from a university campus, directly below an airport's arriving flight path's descent line, and to analyze lead content suspended in the air. Two collection sets of three separate samples were collected on six separate days, one set in July of 2018 and the second set in January 2019. The collection procedure began in the morning and continued into the afternoon. Samples were collected with an air abatement monitor, borosilicate glass fiber filters. The negative and positive control samples were collected in sterile conditions; the negative being devoid of trace metal particles and the positive saturated in In and Zn. The fiber filters were digested in in a 2.06M nitric acid and extracted through sonication in an 80°C water bath. They samples were measured on an Induced-Coupled Plasma Mass Spectrometer, utilizing a linear standard. The experiment showed that levels of lead all exceeded the Environmental Protection Agency's federal regulated standards of 0.15µm/m³. Lead levels exceeded 3.2µm/m³ within a 6-hour collection time and reached as high as 6.3µm/m³. ## INTRODUCTION The University of South Florida, St. Petersburg is placed in an ideal location. Located right on the southside edge of downtown St. Petersburg area and crested on a small inlet of Tampa Bay. The USFSP campus is located next to Albert Whitted Airport (AWA), a private airfield that focuses on charter planes, small jets, single and double piston engines and medical emergency flights. A study in Michigan found that child blood levels declined in the months following 9/11. 3 This independent study was undertaken to check the amount of Pb that is in suspended particulate matter (PM) in air and compare to the EPA's regulated PM of Pb, 0.15µg/m³, according to the 2008 National Ambient Air Quality Standards. The EPA has noted that airports have potential to exceed the ambient Pb concentration set in the NAAQS. 1 The proximity of a small airport and a university campus is not something many realize until hearing an older plane or large engine moving overhead while on its decent. Jet engine aircraft utilize jet fuel for flight, though small engine aircraft use gasoline which still contains lead (Pb), called Tetraethyl lead (TEL). The EPA phased out leaded gasoline from vehicles by the end of 1995. Though piston engines for propeller planes and small aircraft have legally continued use of leaded gas for flight. The Clean Air Act banned the sale of leaded fuel still available in some parts of the country as of January 1 st , 1996, for use in on-road vehicles. 5 Though in 2011, piston engine aircraft used 225 million gallons. 1 AWA landing runway is less than 100 yards from the east side of DAV building, the south side of DAV opens to large grass covered areas, walkways to/from Tampa Bay shoreline, and several locations for students to congregate. It is this area that p lanes arriving at AWA descend over to touch down. Since the EPA regulates that ambient air should not have more than 0.15µg/m³ Pb, it needs to be determined if this small airfield exceed the NAAQS. Dependent on the number of arrivals per given day, the detection of Pb should show different levels. The potential health risk to a university population surely warrants investigation. While forming the experimental procedure, it was predicated that related to the number of arrivals would be directly correlat ed with levels of lead detected on any set day, therefore it is believed that these levels fluctuate. ## Sample Collection Samples were obtained using a BDX II Abatement Air Sampler, including polyurethane collection tubing and glass funnel . Two collection sets of three samples were taken. The two sets were separated by six months, each provided three individual days for each set's collection of SPMs. The BDX II pumps monitor for asbestos and lead. 8 Using Whatman Binderless borosilicate glass fiber filters, 31mm diameter and 0.45µm pore size, each sample was taken over a 6-hour interval. The BDX II was placed on a two-story rooftop at the University of South Florida, St. Petersburg's Davis (DAV) building. The experimental collection set had two controls, a negative control, consisting of a blank filter and a positive control, provided with an excess amount of trace metal grade Zn and In. All four controls ran for a 6-hour interval under sterile hood conditions. To provide a significant signature, two clean glass stir rods were used to administer the positive controls saturation of Zn and In delivery at hours 2 and 3, respectively for each collection set. Each of the sample's and control's filter paper were removed following standard laboratory safety, under a sterile hood, using Personal Protective Equipment (PPE) and folded in half, placed in polyurethane bags and stored at ambient temperature until digestion and extraction. ## Digestion and Extraction Graduated cylinders were first rinsed with 1% HNO₃, followed by DI water. Using 50mL conical tubes, each folded sample from the polyurethane storage bag was removed by inverting the bag and setting the filter into its corresponding tube. Tubes were filled with 15mL of 2.06 M HNO₃ and placed in a rack, then submerged in an 80°C water bath and ultrasonicated for 60 minutes. After sonication, an additional 15mL of DI water was added for total volume of 30mL. Sample tubes were then centrifuged at 8,000 rpm for 5 minutes. Tubes were then decanted into new corresponding 50mL conical tubes and stored at 4°C until ICP-MS analysis. ## Analysis Allowed samples to reach ambient temperature and then a second centrifugation of 10,000 rpm for 10 minutes. Using an Agilent 7500 series Inductive Coupled Plasma -Mass Spectrometer, generated a linear gradient using industry standards, Germanium (Ge) and Bismuth (Bi). Analyses was performed on the controls and samples, as well as the two standards Ge and Bi. The linear gradient began at 6.25ppb, increasing two-fold at each interval standard until 200ppb. ## Collection Set I The two standards that were used with the analysis was Germanium (72) and Bismuth (209), kept at a controlled 50 ppm. The linear standard that was generated began at 6.25ppm and doubled thereafter up to 200 ppm. Control A showed the absence of In and had 1459 ppb of Zn. Pb reflected 2.821ppm in control A. Control B showed In at 217.5 ppb and Zn at 1711 ppb, while Pb was 2.90 ppb. Samples 1, 2, 3 showed no trace In, suggesting that environmental In is undetectable. The samples showed that Pb levels were 6.328ppm, 4.76ppm, 4.51ppm, respectively. This corresponds to the number of arrivals that landing for each of the three samples collected at 19+, 13+ and 10+ respectively. Control A and B had an average of 2.864 ppb of Pb, the samples had an average of 5.20 ppb. Experimental samples were saturated in Zn. It was determined that not only was there an abundance of Zn suspended particulate in the experimental environment but that the zinc plated screws to hold the glass fiber filter paper was adding considerable amounts of trace particles. ## Table 1. Collection Set II Again, two controls were used, administered in the same manner as Collection Set I. Control A was void of In and had 9331 ppm of Zn. Control B showed 56.92 ppb of In and over 10, 000 ppb of Zn. All three samples also showed levels of Zn exceeding 10,000 ppb. This is most likely due to the removal and replacing the four zinc plated screws when changing the filter between samples and controls. In was again not present in the samples collected, affirming a lack of SPM within the vicinity. The lead measured in the three samples collected for Collection Set II was 3.216, 3.76. and 3.35 ppb respectively. While the arrivals on the ## DISCUSSION When the EPA selected a new standard for suspended Pb in 2008 which was ten times more rigorous, monitoring data showed that out of all US counties only 18 would violate the new standard. 7 St. Petersburg is within Pinellas county, which was not one of the eighteen the EPA noted. Though within this general area of USFSP, it seems that five to six thousand students, staff, and university employees are exposed to levels that are even above the prior standard that the EPA changed in 2008. Information on arrival flights were obtained through flightaware.com, which provides data on arrival and departure flights for small airports across the US. 10 The BDX II's filter externally mounted in place at the front of the equipment. It is held by a molded, hard plastic that allows the tubing to inject the air flow. To remove the filter and place a new one inside, four small metal screws must be removed from the plastic that attaches it to the equipment. These screws are small black zinc-plated, as are the other screws used on the equipment. There are mainly two types of screws; zinc plated or galvanized, which is an outer coating used to combat rust from the natur al oxidation of steel and stainless-steel, which is rust proof but more expensive. Kolle describes galvanizing as a Tootsie pop and that the zinc does eventually give way to the steel screw, though there are variations of thickness. 9 The Zn levels are most likely do to fine particulate matter released when removing and replacing these screws, as well as the naturally high levels of Zn in the atmosphere. During Collection Set I, samples one and three experienced a random error of significant mention just prior to analyses. The second centrifugation process was performed in equipment that had been solely operated for microbiologic procedures, the 10, 000 rpm was too severe on sample one and three. This led to a crack in the conical tubes from the 15mL level down to the top of the coned bottom. The exterior tubes that held the conical tubes were leaked into. Upon completion of the 10-minute run, these exterior tubes were immediately removed and poured into two new sterile 50mL conical tubes and labeled samples one and two. After investigating what potential contaminants may be present from the exterior tubes, the information obtained was that basic microbiological materials may be present. This ranged from EtOH, DI water, NaCl solution, to Taq-polymerase, live bacteria, and DNA. None of the materials associated with the potential contaminants would directly apply to the specific trace metals that were analyzed. Nevertheless, due to this reason the data provided by sample two and controls A and B are free of experimental error. Even still, therefore a secondary experiment, Collection Set 2, was done and measured, as to support the findings of the first. The data provided is supportive of the original hypothesis, increased arrival flights of small engine/propeller planes show increased lead levels of suspend particulate matter. It is also assisted by the high Zn levels that were detected, which did not fluctuate from the random error. However, since the random error is potentially a source of contamination the experiment was repeated for conformational data. Aside from the error, the levels of lead that are present in sample two of CS1 is 4.76 ppb. This is 17.86 times greater than the EPA regulation and warrants an in-depth investigation of SPM of lead levels around the USFSP campus. While the average amount of Collection Set 2, which experienced no random errors, was 3.62 times greater than the EPA regulation. Having small samples sizes limits the reproducibility at that point, though if rigorously approached in the same manner as CS I and CS II, further experimentation indicates significant results. Though the EPA website notates that levels of Pb maybe higher in areas of immediate surroundings of airports with small aircraft, having a university campus located direct at the start of the arriving flights runway could be negatively impacting thousands of individuals on an annual basis. It is suggested that t he airport and the city that regulates it, at a minimum, change the approach pattern of arriving flights. This should at least decrease the levels that directly fall over the campus.

chemsum

{"title": "Leaded Aviation Fuel May Present Long-Term Effects on Campus Life from the Adjacent Albert Whitted Airport", "journal": "ChemRxiv"}

manganese_complex-catalyzed_oxidation_and_oxidative_kinetic_resolution_of_secondary_alcohols_by_hydr

## Abstract: The highly efficient catalytic oxidation and oxidative kinetic resolution (OKR) of secondary alcohols has been achieved using a synthetic manganese catalyst with low loading and hydrogen peroxide as an environmentally benign oxidant in the presence of a small amount of sulfuric acid as an additive. The product yields were high (up to 93%) for alcohol oxidation and the enantioselectivity was excellent (>90% ee) for the OKR of secondary alcohols. Mechanistic studies revealed that alcohol oxidation occurs via hydrogen atom (H-atom) abstraction from an a-CH bond of the alcohol substrate and a twoelectron process by an electrophilic Mn-oxo species. Density functional theory calculations revealed the difference in reaction energy barriers for H-atom abstraction from the a-CH bonds of Rand Senantiomers by a chiral high-valent manganese-oxo complex, supporting the experimental result from the OKR of secondary alcohols. ## Introduction The selective oxidation of organic substrates using earthabundant transition metal catalysts (e.g. manganese and iron) and environmentally benign oxidants (e.g. molecular oxygen and hydrogen peroxide) is fundamentally important in enzymatic/biomimetic reactions and immensely useful in organic synthesis. 1,2 Therefore, tremendous efforts have been devoted to elucidating the biomimetic oxidation reactions and developing highly efficient, selective (asymmetric) oxidation reactions using earth-abundant transition metal catalysts and environmentally benign oxidants under mild conditions. As a result, a great advance has been achieved recently in catalytic (asymmetric) epoxidation and hydroxylation reactions using synthetic iron and manganese catalysts and aqueous H 2 O 2 as an environmentally benign oxidant in the presence of carboxylic acid as an additive. In these reactions, it has been proposed that high-valent metal-oxo intermediates are the active oxidants that affect the (asymmetric) epoxidation and hydroxylation reactions, and that the role of the carboxylic acid is to facilitate the heterolytic O-O bond cleavage of putative metal-hydroperoxo species to form high-valent metal-oxo intermediates. Very recently, we reported that a manganese complex bearing a tetradentate N4 ligand is an efficient catalyst in the asymmetric epoxidation of olefns by aqueous H 2 O 2 in the presence of a small amount of H 2 SO 4 , affording high product yields with excellent stereo-and enantioselectivities. 7 In the latter reaction, it was shown that carboxylic acid can be replaced by H 2 SO 4 for activating H 2 O 2 by manganese complexes, generating highvalent manganese-oxo species as active oxidants, 7 although the role of H 2 SO 4 remains elusive. Another important research area in oxidation reactions is the oxidation of alcohols to aldehydes or ketones. 8,9 Recently, nonporphyrinic manganese complexes have been employed as catalysts in the development of efficient catalytic systems for alcohol oxidation reactions, especially in those using H 2 O 2 as an environmentally benign oxidant in the presence of carboxylic acid. 9 Mechanistic studies have been performed to elucidate the alcohol oxidation reactions using synthetic metal-oxo complexes. 10 In alcohol oxidation chemistry, the oxidative kinetic resolution (OKR) of racemic secondary alcohols has attracted much attention for developing efficient catalytic systems to obtain enantio-enriched alcohols, since chiral secondary alcohols are valuable synthetic intermediates in the pharmaceutical and fne chemical industries. In the OKR of racemic secondary alcohols, chiral metal complexes (e.g. the Mn(III) salen complex) with artifcial oxidants (e.g. iodobenzene diacetate and sodium hypochlorite) have been used to produce chiral secondary alcohols. 12d,13 However, to the best of our knowledge, the OKR of secondary alcohols has never been explored using synthetic mononuclear manganese catalysts and aqueous H 2 O 2 as the terminal oxidant. Herein, we report that manganese complexes bearing tetradentate N4 ligands, such as Mn(II)(P-MCP)(OTf) 2 (1) and Mn(II)(Dbp-MCP)(OTf) 2 (2) (Scheme 1A), 7 are highly efficient catalysts for the oxidation of alcohols by aqueous 30% H 2 O 2 in the presence of a catalytic amount of H 2 SO 4 (Scheme 1B). Furthermore, to the best of our knowledge, the present study reports the frst example of the use of a chiral manganese catalyst with low loading, aqueous H 2 O 2 as the terminal oxidant, and a catalytic amount of H 2 SO 4 as an additive in the OKR of racemic secondary alcohols (Scheme 1C). Density functional theory (DFT) calculations reveal that the energy barriers for the a-CH bond activation of chiral 1-phenylethanol (R-and S-enantiomers) by a high-valent manganese-oxo complex are signifcantly different, explaining the experimental observation of high enantioselectivity in the OKR of racemic secondary alcohols. ## Results and discussion Firstly, the reaction conditions for the catalytic oxidation of alcohols by manganese complexes and aqueous H 2 O 2 in the presence of H 2 SO 4 were optimized using 1-phenylethanol as a model substrate (see the Experimental section). Among the tested manganese catalysts (see Scheme 1A for the structures), Mn(II)(P-MCP)(OTf) 2 (1) and Mn(II)(Dbp-MCP)(OTf) 2 (2) exhibited high catalytic activity (Table 1, entries 1 and 2), whereas Mn(II)(MCP)(OTf) 2 (3) was a poor catalyst (Table 1, entry 3). In the absence of the manganese catalyst, the oxidation of 1-phenylethanol to acetophenone was not observed (Table 1, entry 4). Since 1 can be prepared easily and cost-effectively, 1 was used as the catalyst to determine the optimal reaction conditions by varying the amount of catalyst (Table 1, entries 5 and 6), H 2 O 2 (Table 1, entries 5 and 7) and H 2 SO 4 (Table 1, entries 7-10). In addition, as reported in the olefn epoxidation reactions by nonporphyrinic Mn catalysts and H 2 O 2 , 7 other Brønsted acids, such as HClO 4 , H 3 PO 4 , HCl and CF 3 SO 3 H, turned out to be poor additives (Table 1, entries 11-14). After optimizing the reaction conditions (Table 1, entry 9), we investigated the substrate scope for the oxidation of secondary alcohols by 1 and H 2 O 2 in the presence of H 2 SO 4 (Table 2). 1-Phenylethanol derivatives with electron-donating and -withdrawing substituents at the para-position of the phenyl group were oxidized to their corresponding ketones with good yields (e.g. >80%) (Table 2, left column). However, the product yields were moderate (e.g. $50%) for the oxidation of 1-phenylethanol derivatives with steric hindrance at the ortho-position of the phenyl group (Table 2, left column). Similarly, increasing the chain length and steric hindrance on the methyl side of the 1phenylethanol derivatives, such as 1-phenylbutan-1-ol, 2-methyl-1-phenylpropan-1-ol and 2,2-dimethyl-1-phenylpropan-1-ol, Scheme 1 (A) Schematic structures of manganese complexes bearing N4 ligands: Mn II (P-MCP)(OTf) 2 (1; P-MCP ¼ (1R,2R)-N,N 0 -dimethyl-N,N 0 -bis-(phenyl-2-pyridinylmethyl)cyclohexane-1,2-diamine and OTf Entry Catalyst (mol%) Additive (mol%) H 2 O 2 (equiv.) Yield (%) decreased the product yields (Table 2, middle column). On the other hand, a series of diphenylmethanol derivatives were converted to the desired products with yields of >80% (Table 2, middle column). Unactivated aliphatic secondary alcohols were also oxidized to their corresponding ketones with moderate to high yields (Table 2, right column). We then investigated the OKR of secondary alcohols utilizing the manganese catalyst, H 2 O 2 oxidant and H 2 SO 4 additive system. Firstly, we optimized the reaction conditions by varying the amount of catalyst, oxidant and H 2 SO 4 , and the reaction temperature (see Fig. 1; also see Tables S1 and S2, ESI †). Obviously, the conversion of 1-phenylethanol would increase with an increasing amount of H 2 O 2 , as shown in Fig. 1. Due to the preference for oxidation of the S-enantiomer in the OKR of racemic secondary alcohols using the sulfuric acid-enabled manganese system, the ee value improved when increasing the number of equivalents of H 2 O 2 from 0 to 0.80, while no signifcant change occurred when further increasing the number of equivalents of H 2 O 2 up to 1.0. Therefore, the oxidant amount was chosen to be 0.80 equiv. for the OKR of 1-phenylethanol, as a result of the excellent ee with lower conversion. Besides, 2 was chosen as the catalyst since 2 afforded a higher enantiomeric excess (ee) value (90%) than 1 (65%) in the oxidation of 1-phenylethanol (Table 3, entry 1 and footnote c). Under the optimized catalytic conditions, we obtained high ee values (>90% ee) irrespective of the substituents on the phenyl group of the benzylic alcohols (i.e. no effect from steric hindrance or the electronic nature of the substrates) (Table 3, entries 2-6). Importantly, 1-phenylpropan-1-ol derivatives, which were reported to be poor substrates in most manganese salen systems, 12d,13,14 also worked well in this sulfuric acidenabled manganese system, with ee values of >90% (Table 3, entries 7-12). In addition, increasing the steric hindrance and chain length in the 1-phenylpropan-1-ol derivatives did not affect the ee values either (Table 3, entries 13-15). In order to gain mechanistic insight into the manganesecatalyzed alcohol oxidation reactions, we frstly investigated the effect of para-substituents on the reactivity of the benzyl alcohol by carrying out competitive alcohol oxidation of the benzyl alcohol against para-substituted benzyl alcohols (see the Experimental section). A good linear correlation was obtained when the k rel values were plotted against the Hammett parameters of the substituents (Fig. 2); the small but negative r value of 0.58 indicates that the active intermediate possesses electrophilic character, as reported in the oxidation of benzyl alcohol derivatives by synthetic metal-oxo complexes. 10 Secondly, when the intermolecular competitive oxidation of 1phenylethanol or its a-deuterated compound (1-deuterated 1phenylethanol) was carried out together with 1-(p-chlorophenyl) ethanol as a mediator, a kinetic isotope effect (KIE) value of 1.8 was obtained (see the Experimental section for the detailed method), suggesting that hydrogen atom (H-atom) abstraction from an a-CH bond may be the rate-determining step, as observed in other manganese complex-catalyzed alcohol oxidation reactions. 8b,c It is also notable that the KIE values determined in C-H bond activation reactions by synthetic metal-oxo complexes under stoichiometric conditions (e.g. KIE values of > 10) 10,15 are much higher than those obtained in metal complex-catalyzed oxidation reactions under catalytic conditions (e.g. KIE values of < 4). 8b,c It would be of interest to understand the reason for the difference in the KIE values obtained from the stoichiometric and catalytic reactions (e.g. the involvement of different metal-oxygen intermediates in the catalytic oxidation reactions). Thirdly, since cyclobutanol has often been used as a substrate probe to distinguish one-electron and two-electron processes in alcohol oxidation reactions, 10a,c,16 we performed the oxidation of cyclobutanol and found that cyclobutanone was yielded exclusively and the ring-opened Table 2 Substrate scope for the oxidation of secondary alcohols a,b,c a Reaction conditions: a CH 3 CN (0.50 mL) solution containing 30% H 2 O 2 (1.2 equiv.) was added dropwise to a CH 3 CN (1.0 mL) solution containing the substrate (0.50 mmol), 1 (0.30 mol%) and H 2 SO 4 (0.30 mol%), using a syringe pump at 25 C for 1 h. b Yields were determined by GC. c Numbers in parentheses are product yields. d 1.0 mol% H 2 SO 4 was used. product, 4-hydroxylbutyraldehyde, was not detected (Scheme 2). Based on the mechanistic studies discussed above, we conclude that alcohol oxidation by an electrophilic manganese-oxo species is a two-electron process. This reactive manganese-oxo intermediate has been proposed previously in the asymmetric epoxidation of olefns by the manganese catalyst (2) and H 2 O 2 in the presence of H 2 SO 4 . 7 Density functional theory (DFT) computations were then carried out to explore the enantioselectivity in the a-CH bond activation of the chiral 1-phenylethanols (both R-and S-enantiomers) by a chiral [(P-MCP)Mn(V)(O)(SO 4 )] + complex (I), which was proposed previously as the reactive intermediate in the reaction of 1 and H 2 O 2 in the presence of H 2 SO 4 . 7 The computational results reveal that H-atom abstraction from the a-CH bond of the S-enantiomer, with an energy barrier of 7.7/ 5.5 kcal mol 1 for the triplet/quintet spin state, is easier than that from the a-CH bond of the R-enantiomer, with an energy barrier of 10.6/6.8 kcal mol 1 for the triplet/quintet spin state (Table S3, ESI †). This reactivity trend was obtained by comparing the geometric character of these two transition states, in which TS S with r C-H ¼ 1.203 has a smaller elongation of the C-H bond, while TS R with r C-H ¼ 1.240 has a larger elongation of the C-H bond for the quintet ground state (Fig. S32, ESI †). In addition, the energy barrier difference for the two isomers might be due to the non-covalent anion-p interaction between the phenyl group of the substrate and the sulfuric acid anion ligand, which can stabilize the transition state, lowering the energy barrier. This kind of interaction only exists for TS S , with a distance of ca. 3.4 between the two groups. Thus, we may conclude that the S-enantiomer is an easier substrate than the R-enantiomer for oxidation by the high-valent Mn-oxo intermediate, regardless of the spin states; therefore the R-enantiomer remains in the reaction solution. Based on the Arrhenius equation, an energy barrier difference of 1.3 kcal mol 1 will give a rate constant ratio (k (R) /k (S) ) of 0.112, which corresponds to a high ee value ($80%), as obtained from the experiments (vide supra). From inspection of the spin densities (Table S4, Fig. S32 and S33, ESI †), we can see that for the quintet spin state, the sulfuric acid group has a spin density of ca. 0.5 in the 5 I+ sub species and ca. 0.0 in the transition state. a Reaction conditions: a CH 3 CN (0.50 mL) solution containing 30% H 2 O 2 (0.70-0.90 equiv.) was added dropwise to a CH 3 CN (1.0 mL) solution containing the secondary alcohol (0.50 mmol), 2 (0.20 mol%) and H 2 SO 4 (1.0 mol%), using a syringe pump at 0 C for 1 h. b Conversion yields and ee values were determined by GC with a CP-Chirasil-Dex CB column. c When 1 was used as a catalyst under identical reaction conditions, the conversion yield and ee value were 66% and 65%, respectively. d Conversion yields were calculated from the isolated products and the ee values were determined by HPLC with an IA column. This indicates that the sulfuric acid group should be noninnocent to the reaction, and the non-innocence of the sulfuric acid group may make the reaction on the quintet spin state approachable. In addition, the low energy barrier indicates the high electrophilicity of the high-valent manganese-oxo species, which may originate from the existence of a low lying s* orbital that can accept electronic density from the C-H bond (Fig. S33, ESI †). ## Conclusions In summary, we have reported the frst example of sulfuric acidenabled chemoselective oxidation of secondary alcohols by manganese catalysts and hydrogen peroxide. 17 Secondary alcohols were oxidized to their corresponding ketones with good yields and an efficient OKR of racemic secondary alcohols was achieved with excellent enantioselectivities (>90% ee). Mechanistic studies revealed that the active manganese oxidant possesses electrophilic character, H-atom abstraction from an a-CH bond of the alcohol substrate is the rate-determining step, and alcohol oxidation occurs via a two-electron process. DFT calculations revealed that the difference in reaction energy barriers for H-atom abstraction from the a-CH bonds of the Rand S-enantiomers by a putative high-valent manganese-oxo intermediate is signifcant (i.e. 1.3 kcal mol 1 ), affording the enantioselectivity in the OKR of racemic secondary alcohols. Future studies will focus on the improvement of the catalytic activity as well as the enantioselectivity in the OKR of secondary alcohols, using synthetic nonheme iron and related manganese catalysts and environmentally benign oxidants such as molecular oxygen and hydrogen peroxide. ## Materials All chemicals were purchased from Aldrich, Alfa Aesar and TCI, which were of the best available purity and were used without further purifcation unless otherwise indicated. Solvents were dried according to published procedures and distilled under argon prior to use. 18 PhCD(OH)CH 3 was prepared from acetophenone according to the literature method. 19 The ligands, MCP, P-MCP and Dbp-MCP, and their corresponding Mn II complexes, Mn II (MCP)(OTf) 2 , Mn II (P-MCP)(OTf) 2 and Mn II (Dbp-MCP)(OTf) 2 , were prepared according to the reported methods. Typical procedure for the oxidation of 1-phenylethanol All reactions were performed under an Ar atmosphere using a dried solvent and standard Schlenk techniques. The catalyst (0.30 mol%), 1-phenylethanol (0.50 mmol) as a substrate and H 2 SO 4 (0.30 mol%) were added into a Schlenk tube containing CH 3 CN (1.0 mL) at 25 C. Subsequently, a solution of CH 3 CN (0.50 mL) containing 30% H 2 O 2 (1.2 equiv.) was added dropwise using a syringe pump for 1 h. Then, the reaction solution was quenched with NaHCO 3 and Na 2 S 2 O 3 , and n-decane was added into the solution as an internal standard. The yields were determined by GC. Typical procedure for the OKR of 1-phenylethanol All reactions were performed under an Ar atmosphere using a dried solvent and standard Schlenk techniques. The catalyst (0.20 mol%), racemic 1-phenylethanol (0.50 mmol) as a substrate and H 2 SO 4 (1.0 mol%) were added into a Schlenk tube containing CH 3 CN (1.0 mL) at 0 C. Subsequently, a solution of CH 3 CN (0.50 mL) containing 30% H 2 O 2 (0.80 equiv.) was added dropwise using a syringe pump for 1 h. Then, the reaction solution was quenched with NaHCO 3 and Na 2 S 2 O 3 , and ndecane was added into the solution as an internal standard. The yields were determined by GC (see Fig. 1 and Table 3; see also Tables S1 and S2, ESI †). ## Determination of the relative rate constants (k rel ) A solution of CH 3 CN (0.50 mL) containing 30% H 2 O 2 (0.40 equiv.) was added dropwise into a CH 3 CN solution (1.0 mL) containing a mixture of 1-phenylethanol (0.50 mmol) and parasubstituted 1-phenylethanol (0.50 mmol), the catalyst (1, 0.30 mol%) and H 2 SO 4 (0.30 mol%) using a syringe pump at 25 C for 1 h. Then, the reaction solution was quenched with NaHCO 3 and Na 2 S 2 O 3 , and n-decane was added into the solution as an internal standard. The yields were determined by GC. The k rel values were calculated using eqn (1), 19a where [R] i and [R] f are the initial and fnal concentrations of para-substituted 1-phenylethanol, respectively, and [H] i and [H] f are the initial and fnal concentrations of 1-phenylethanol, respectively. ## Determination of the KIE value A solution of CH 3 CN (0.50 mL) containing 30% H 2 O 2 (0.40 equiv.) was added dropwise into a CH 3 CN solution (1.0 mL) containing a mixture of substrate (0.50 mmol; 1-phenylethanol or 1-deuterated 1-phenylethanol) and 1-(4-chlorophenyl)ethanol (0.50 mmol), 1 (0.30 mol%) and H 2 SO 4 (0.30 mol%) using a syringe pump at 25 C for 1 h. Then, the reaction solution was quenched with NaHCO 3 and Na 2 S 2 O 3 , and n-decane was added into the solution as an internal standard. The yields were determined by GC. The KIE values were calculated using eqn (2)-( 4 D] f are the initial and fnal concentrations of 1deuterated 1-phenylethanol, respectively. ## Computational details Density functional theory calculations were performed using Gaussian 09 software. 21 The high-valent [(P-MCP) Mn 5+ (O 2 )(SO 4 2 )] + species (I) was chosen as the reactive intermediate, and two enantiomers of the chiral 1-phenylethanol (S-and R-enantiomers) were used as the substrates. The spin-unrestricted B3LYP (UB3LYP) functional 22,23 was employed with two basis sets: (1) the LACVP basis set for Mn and the 6-31G* basis set for the rest of the atoms, denoted as B1, were used to optimize the minima and transition states; (2) the LACV3P basis set for Mn and the 6-311+G** basis set for the rest of the atoms, denoted as B2, were used to obtain the single point energy corrections. 24,25 The transition states and optimized minima were ascertained by vibrational frequency analysis with only one and zero imaginary frequencies, respectively. All calculations were performed in acetonitrile solvent using the self-consistent reaction feld (SCRF) in the conductor-like polarizable continuum model (CPCM). ## Conflicts of interest There are no conflicts to declare.

chemsum

{"title": "Manganese complex-catalyzed oxidation and oxidative kinetic resolution of secondary alcohols by hydrogen peroxide", "journal": "Royal Society of Chemistry (RSC)"}

the_"decisive"_role_for_secondary_coordination_sphere_nucleophiles_on_hydrogen_atom_transfer_(hat)_r

## Abstract: Although it has been reported that some radical reactions are possibly promoted by external ions, the origin of this phenomenon is unclear. In this work, several hydrogen atom transfer (HAT) reactions in the presence of anions were studied by density functional theory (DFT) calculations, electronic structure analysis and other methods, and it is concluded that both the electrostatic interaction and polarization of the transition state (TS) by the electric field generated by anions play a fundamental role in the TS stabilization effect, whereas the "charge shift bonding" that was previously presumed to be a major contributor is ruled out. Although the stabilization toward TSs in terms of electronic energy (and thus enthalpy) is significant, it should be noted that the effect is almost completely cancelled by entropy and solvation, and further cancelled by the formation of stable resting states. Thus there is still a long way for this effect to be used in actual catalysis. ## Introduction The "electrostatic catalysis" or "salt effect" is a long-standing and well-established concept. Early in 1990s, the catalytic effect of ions that seem inert at the first glance toward organic chemical transformations has been studied by Craig Wilcox . Later on, the promotion of cobalt-carbon bond dissociation by a nearby charge was found in a biochemistry-related Vitamin B complex 4 . In the recent years, the catalytic effect of charged groups toward Diels-Alder reaction was studied by Michelle Coote 5,6 , and Kendall Houk 7 . The catalytic effect of charged groups is believed to have an electrostatic nature, proceeding through interaction between the dipolar moment of transition states (TSs) and the electric field generated by nearby charges, and thus is closely relevant to the external electric field effect in chemical reactions, which has been documented extensively in many cases . It is noteworthy that hydrogen atom transfer (HAT) reactions have also been reported to be affected by metal ions and ligands , which is believed to be a field-induced phenomenon (chargeinduced catalysis). On the other hand, however, Thomas Cundari and coworkers reported that external anions provide "decisive" stabilization to the TS for the hydrogen atom transfer (HAT) reaction between methane and hydroperoxyl radical very recently 18 . The authors concluded that the interaction between anions and the HAT TS is due to "charge shift bonding 19,20 " originating from a 2-center-3electron interaction, which is a brand new explanation for the influence of external ions. Thus it is interesting how much role charge shift bonding plays in the reported reactions, and also in other examples that were previously believed to be field-originated. In this work, we conducted a more detailed investigation on the "salt effect" for HAT reactions, which will provide new understanding toward this long-standing concept. ## Computational Methods The geometry optimization of all structures were performed with the Gaussian 16 program 21 , at M11/6-311+G(d,p) level , if not specially mentioned. DLPNO-CCSD(T) calculations were carried out with the ORCA 4.2 program 27,28 , in combination with the aug-cc-pVTZ basis set . All electronic structure analysis, including but not limited to bond critical point (BCP) properties, electron localization function (ELF), electron density Laplacian, were performed using the Multiwfn program 32 , based on the wavefunction obtained at M11/ma-def2-TZVPP level 33 . The GAMESS-US program 34 was employed for LMO-EDA calculations 35 . It is found that the GAMESS-US program gave wrong results with the M11 functional, and thus M06-2X/6-311+G(d,p) level 36 was selected to perform energy decomposition with the M11-optimized geometry. The SMD implicit solvation model 37 was used for calculations with solvation effect, and the solvation free energies were obtained by G(M05-2X 38 /6-31G(d), with SMD(DMSO)) -G(M05-2X/6-31G(d), gas phase). The final Gibbs free energies were obtained by the sum of DLPNO-CCSD(T) single point energy, M11/6-311+G(d,p) correction to free energy, and solvation free energy (the last term only for calculations in DMSO). Particularly, the solvation free energies for Cl -, Brand proton are taken from experimental reports 39 , and those for F -, HOand HSwere derived from experimental pKa of HF, H2O and H2S in DMSO, which is 15.0 40 , 31.4 41 and 13.7 42 respectively. ## Results and Discussions Although the meta-GGA M06-L functional 43 was employed in Cundari's report, it was found from a benchmark study involving M06-L, B3LYP-D3BJ 44,45 , M06-2X, wB97xD 46 and M11 that the performances of density functionals parallel their Hartree-Fock (HF) components (Table S1), and thus the range-separated functional M11 with a large HF component was chosen to be the functional used for geometry optimization in this work. The TSs were located for the HAT reactions between methane and hydroperoxyl radical (Scheme 1a), in the presence of various anions X -. The energetics and optimized C-X bond lengths at DLPNO-CCSD(T)/aug-cc-pVTZ//M11/6-311+G(d,p) level are listed in Table 1. Note that at this stage only the TSs, but not preactivation complexes or any other resting states are discussed. The full Gibbs free energy surface will be discussed later. Scheme 1. The reactions studied in this work, and the definitions for some quantities discussed. 1 that all anions provide significant stabilization (i.e. negative E) to TS1 in terms of electronic energy, although largely cancelled by entropy. The second-row anions, fluoride and hydroxyl anion, are among the most stabilizing ones, whereas the HCOOwith delocalized negative charge exhibits much less stabilization. Interestingly, the heavier anions, Cl -, Brand HSprovide similar stabilization energy at ~8.5 kcal/mol, while the E value becomes surprisingly much more negative upon replacement of the hydrogen in HSwith the methyl group. Overall, it is concluded that the combination of anions to TS1 is exothermic in most cases, and next we are about to figure out the reason. As suggested by Cundari's work, we firstly examined the existence of charge shift bonding, which originates from the interplay between two resonance structures shown in Figure 1a. It has been proposed in the early research on the charge shift bonding in silicon-halogen bonds that resonance structures close in energy might lead to larger resonance energy 47 , and thus stronger charge shift bonding. Herein the energy difference between resonance structure 1 and 2 could simply be characterized by the difference in the vertical electron affinities (EA shown in Scheme 1) for the anion and the "bared" TS1(X=none). In addition, the magnitude of the involvement of 2 is reflected by the spin population on X. It is expected that a EA close to zero, as well as large spin population on X, should indicate strong charge shift interaction. The spin population values are listed in Table 1, and it is seen that for most anions the spin population on Xis negligible, except for OH -, HSand MeS -, which also exhibit EAs closest to 0. Furthermore, the TS stabilization energy E correlates with both EA and spin population on X terribly. Thus it is questionable whether charge shift interaction could explain the observed energy change. (f) The ELF contour for F2 molecule, a molecule with typical charge shift bonding. (g) The RDG isosurface for TS1(X=HO -), in which weak noncovalent interaction is shown in blue (stronger ones) and green (weaker ones). Traditionally the existence of charge shift bonding is characterized by the properties at bond critical point (BCP), such as the electron density, electron density Laplacian, and electron localization function (ELF). Typical charge shift bonds, such as that in F2 molecule, exhibit large electron density, positive Laplacian, and slightly accumulated ELF at BCP. The electron densities at the BCP located between the carbon atom and anion are recorded in Table 1, and all TSs exhibit negligible electron densities at BCPs. The contours for ELF and electron density Laplacian for TS1(X=HO -) clearly show that there is no accumulation of ELF, and only near-zero Laplacian in the C-X interatomic region (Figure 1d and e). Furthermore, the reduced density gradient (RDG) analysis 48 , a method that directly shows the region with noncovalent weak interaction and filters covalent interaction, directly shows that the C-X interaction lies in the region of weak interaction, even with OHas the anion. All these results indicate that there is negligible chemical bonding between C atom and anions. The ELF, Laplacian and RDG analysis results for TS1(X=MeS -) are shown in Figure S3, and there is no different conclusion. However, it still cannot be concluded that there is no charge shift interaction that contributes to the TS stabilization, although charge shift bonding has been ruled out. Next, however, we will focus on alternative ways to explain the observed energy change, and then reconsider the existence of charge shift interaction. There are three possible types of interaction that may contribute to the TS stabilization besides charge shift bonding, namely hydrogen bonding, intrinsic dipole-anion interaction, and polarization effect due to the electric field generated by anion. It has been reported by Tian Lu et al that the electron density at the BCP between a hydrogen acceptor and the corresponding hydrogen atom is a good indicator of hydrogen bond strength 49 . Unfortunately, in all TSs studied there is no BCP between X and H atoms. Thus we turned to core-valence bifurcation (CVB) 50 , another well-known hydrogen bonding indicator. The CVB values for all anions except HSand OHindicate very weak hydrogen bonding (Table S2). Based on these observations, it is proposed that hydrogen bonding contributes only little to the total E. In order to clarify how much role electric field plays in the TS stabilization, a uniform electric field model was employed (Figure 2). In this model, the bared TS1(X=none) is placed in a uniform electric field simulating the influence of external anions, and an "effective" field strength is defined by the electric field at the midpoint of TS1(X=none) along z-axis. TSs were re-optimized in the presence of varied external field (Figure 2b) and the stabilization energies of external field along zaxis at M11/6-311+G(d,p) level exhibit quadratic correlation with field strength. Interestingly, field along y-axis, which is perpendicular to the hydrogen atom transfer path (and also the intrinsic dipole moment of TS1), affords stabilization effect in similar magnitude, and thus it seems that the stabilization originates from polarization, but not intrinsic dipole interaction. The effective field strength resulted from each anion is calculated from the restrained electrostatic potential (RESP) atomic charge 51 on the electronegative atom, and the E predicted using the fitted relationship in Figure 2b is in very good consistence with the M11-calculated E in the presence of anions (since the relationship is fitted with M11 data, it is considerable that it is more comparable with Es at M11 level), with only X=MeSas an exception. Since the field model above is able to provide pretty good explanation for the observed E, the contribution of charge shift interaction is further precluded, and it is suggested that it is the "field effect" that plays the major role in the TS stabilization. Further investigation on the role of spin population delocalization is performed for the MeScase (Figure 2d). Upon re-optimization of TSs at elongated C-S distance, the spin population on S atom drops to zero rapidly, and the energy raises by about 2.5 kcal/mol. It is then concluded that the radical delocalization, or "charge shift interaction", contributes around 2.5 kcal/mol for the stabilization in TS1(X=MeS -), in consistence with the ~3 kcal/mol underestimation of E by electric field model. Since TS1(X=MeS -) is the one with largest spin population on X among all cases studied, it is concluded that charge shift bonding plays only a minor role in "salt effect" for TS1, and its contribution should be even less for anions other than MeS -. On the other hand, however, the full Gibbs free energetics give a different scene of the salt effect, although it is established that external anions combine with TS1 significantly. All anions combine even more strongly with the OOH radical by hydrogen bonding, affording stable resting states, and increasing overall barrier, which is not mentioned in Cundari's work. The overall Gibbs free energy surface is shown in Figure 3. The energies for X-HOO complexes are raised upon solvation, but the overall barriers for all reactions still raise. Moreover, the solvation effect further cancels the TS-stabilization effect. As a result of all the effects above, the reaction is actually inhibited by external anions. Figure 3. The Gibbs free energy surface for reaction a in the presence of anions. Although it has been revealed that anions cannot catalyze the HAT from methane to hydroperoxyl radical in the above work, the reaction studied above gives good insight into the physical picture of "salt effect". After establishment of the field-originated nature, we next moved to the HAT from methane to phenyl and methyl radical (reaction b and c in Scheme 1). The absence of strongly hydrogen bond donating hydroperoxyl radical is expected to avoid the formation of stable resting states. It is noticed that both TS2 and TS3 adopt different geometry: the anions are no longer collinear with the methane carbon center and the coming radical, but form a triangle shape to maximum the hydrogen bonding, which could be understood considering the largely reduced intrinsic dipole in these two cases. However, it is believed that there is no particularity for TS1, and further comments could be found in Supporting Information. Again, all anions provide strong stabilization in terms of electronic energies, while most of them are cancelled by entropy. For both TS2 and TS3, there are no longer spin densities on all anions, and the involvement of charge shift interaction is further precluded. BCPs appear in the interatomic region between anions and nearby hydrogen atoms, and thus the contribution of hydrogen bonding is easily estimated from Lu's work, indicating that HB contributes about 50% of E for these cases. The total E for TS3 can again be predicted by the uniform electric field model (while the effective field strength is hard to define for TS2). Since both TS2(X=none) and TS3(X=none) have negligible dipole moments, the field effect must come from either polarization or interaction with partly charged atoms, whereas the latter is also the origin of hydrogen bonding. In addition, energy decomposition methods further support the major role of electrostatic and polarization interaction (Supporting Information). The full Gibbs free energy surfaces for reaction b and c are shown in Figure 5. Despite the fact that the combining of anions with methyl or phenyl radical is much weaker than in reaction a, the overall barriers are still not lowered with only X=OHfor reaction b as an exception. Therefore, for the three reactions examined in this work, the catalytic effect of anions for HAT reactions previously reported by Cundari et al does not exist considering the full Gibbs free energy surface. ## Conclusion With computational methods, we studied the influence of external anions on the HAT reactions from methane to hydroperoxyl radical (reaction a), phenyl radical (reaction b) and methyl radical (reaction c). Although it was previously proposed by Cundari that it was charge shift bonding that made the major contribution to the stabilization effect toward HAT TSs, detailed electronic structure analysis shows that there is no chemical bonding between anions and methane carbon atom, and the spin delocalization is negligible for most anions. For the case with MeS -, which the anion bears the largest spin population among all reactions studied, the spin delocalization only contributes around 2.5 kcal/mol to the total energy. In contrast, a uniform electric field model gains quantitative success in interpreting the transition stabilization energy. By combining all these results, as well as energy decomposition study, it is believed that the electrostatic and polarization effect due to charged species play a major role. On the other hand, although TSs are stabilized in terms of electric energy (and thus enthalpy), the effect cannot compensate the unfavorable entropy. As a result, the overall barrier, which is further influenced by stable resting states formed by substrates and anions, is not lowered or even raised in the presence of anions. In a summary, in contrast to the well-known existence of salt effect for polar cycloaddition or other reactions with polar TSs, our results are strongly against the presence of the Cundari-type "salt-effect" in HAT reactions. Strategical design, such as spatially constrained anions in order to compensate the entropy effect, is highly in need to further make use of the salt effect in the design of real "salt-catalyzed" HAT reaction in the future. ## Supporting Information Benchmark data, CVB values, further discussion on TS geometry and the uniform electric field model, results for energy decomposition analysis, electronic structure analysis for TS1(X=MeS -), and all geometries involved in this work can be found in Supporting Information.

chemsum

{"title": "The \"Decisive\" Role for Secondary Coordination Sphere Nucleophiles on Hydrogen Atom Transfer (HAT) Reactions: Does it Exist and What is its Origin?", "journal": "ChemRxiv"}

the_min-max_test:_an_objective_method_for_discriminating_mass_spectra

## Abstract: Deciding whether the mass spectra of seized drug evidence and a reference standard are measurements of two different compounds is a central challenge in forensic chemistry. Normally, an analyst will compute a mass spectral similarity score between spectra from the sample and reference and make a judgment using both the score and their visual interpretation of the spectra. This approach is inherently subjective and not ideal when rapid assessment of several samples is necessary. Making decisions using only the score and a threshold value greatly improves analysis throughput and removes analyst-to-analyst subjectivity, but selecting an appropriate threshold is itself a non-trivial task. In this manuscript, we describe and evaluate the min-max test -a simple and objective method for classifying mass spectra that leverages replicate measurements from each sample to remove analyst subjectivity. We demonstrate that the min-max test has an intuitive interpretation for decision-making, and its performance exceeds thresholding with similarity scores even when the best performing threshold for a fixed dataset is prescribed. Determining whether the underlying framework of the min-max test can incorporate retention indices for objectively deciding whether spectra are measurements of the same compound is on-going work. ## Introduction "Is it a fentanyl?" Drug analysts are routinely tasked with answering questions of this nature when presented with seized drug evidence. The most commonly employed analytical technique towards confirmatory tasks in seized drug analysis is gas chromatography mass spectrometry 1 ; gas chromatography is used to physically and temporally separate the case sample into constituent components, and electron ionization (EI) mass spectrometry is used to propose the molecular structure (identity) of each component. A mass spectrum can be thought of as a roughly reproducible a representation of a compound's structural information. In some cases, a mass spectrum includes near complete structural information, allowing it to be directly interpreted and the compound to be uniquely identified (see † To whom correspondence should be addressed: arun.moorthy@nist.gov a We introduce the phrase roughly reproducible to indicate that we expect replicate measurements to be self-similar, but we never expect replicate measurements to be identical. introductory examples in references among other textbooks). In most cases, however, mass spectra contain incomplete and non-unique structural information that render interpretation impractical without comparison to reference spectra. And while many mass spectral libraries 5,6 and interactive software tools 7,8 are available to assist in the interpretation process, the burden of decision-making still lies with the analyst. 9 In an application area like seized drug analysis, where decisions must be made rapidly and analyst subjectivity can be of significant consequence, having an explainable numerical approach for deciding whether mass spectra are measurements of different compounds is an obvious need. We use the term similarity score to represent any numerical index that estimates the similarity between a pair of mass spectra. Although inconsistent in terminology and notation, several similarity scores have been explored in the literature. The most well-known in mass spectrometry is the dot product, 11 or more commonly known as the cosine similarity in other pattern recognition applications. The cosine similarity between mass spectra will evaluate to a real number between 0 and 1 arbitrary units (au), inclusive, with 0 au indicating that the spectra share no common peaks, and the value 1 au indicating that the spectra are identical. A refinement of cosine similarity is the identity match factor, 17 based on the composite score described in the seminal paper by Stein and Scott. 11 The identity match factor considers the ratio of relative intensities of adjacent peaks when estimating similarity, thus capturing subtle information about isotopic patterns that is not necessarily reflected in cosine similarity. In most software implementations, identity match factors have been scaled to 100 (e.g., AMDIS 18 ) or 999 (e.g., NIST MS Search 19 ) and are reported as integers; all similarity scores discussed in this manuscript remain unscaled values between 0 and 1 au. While not explicitly recommended b , we can use a threshold similarity score for deciding whether two mass spectra are measurements of different compounds -a task we refer to as negative confirmation in this manuscript. For example, if the identity match factor between the mass spectrum of a case sample and the spectrum of a reference standard is 0.3 au, the case sample is unlikely to be the same compound as the reference. Formally, we can think of this process in terms of binary classification and define the similarity score test as where 𝑀 is a similarity score between two spectra, 𝜏 ! is a threshold similarity score, and class prediction 0 implies that the spectra are measurements of different compounds. Class prediction 1 implies that the spectra are measurements of the same compound, but we know that confirming whether two samples are the same compound (positive confirmation) with mass spectral comparisons alone is problematic as discussed later in the manuscript. The challenge with implementing the similarity score test for negative confirmation is that there is no obvious choice for a threshold value, especially not one that is universal across all classes of drugs and all varieties of similarity scores; equivalent similarity scores could be computed in very different situations (Figure 1). Algorithms and numerical approaches that produce counter-intuitive results and require analysts to make critical decisions are unappealing in forensics applications. 20 The min-max test was initially formulated during our recent work developing targeted gas chromatography-mass spectrometry methods for identifying synthetic cannabinoids 21 and cathinones. 22 We needed to determine if we could discriminate several pairs of closely eluting compounds based on their mass spectra. We wanted to avoid the ambiguity of threshold setting with a similarity score test and the subjectivity of visual comparison with manual interpretation. The min-max test was able to meet these requirements by leveraging replicate mass spectra to characterize spectral self-similarity within a sample and effectively remove subjectivity from the analysis. In this manuscript, we present the method more completely, with updated notation and refinements to reflect what we learned from our initial experimentation. We discuss how the minmax test, by construction, has an intuitive interpretation for deciding whether spectra are measures of different compounds (negative confirmation), and demonstrate how it out-performs the similarity score test for general classification using EI mass spectra of assorted drugs. ## The min-max test At the core of the min-max test are replicate measurements. By computing similarity scores between replicate mass spectra of individual compounds, we have context for decisions about pairs of compounds. Assume we have two samples to compare: the first is an unidentified compound isolated from seized evidence, and the second is a standard reference compound. Let 𝑺 𝟏𝟏 and 𝑺 𝟐𝟐 be sets of intra-sample similarity scores computed between two or more replicate mass spectra of samples 1 and 2, respectively. Let 𝑺 𝟏𝟐 be the set of inter-sample similarity scores computed between the spectra of the two samples. Using these sets of values, we can formulate a spectral comparison index (𝛿 $ ) that follows the general form where 𝑓(⋅), 𝑔(⋅), and ℎ(⋅) are functions chosen to reduce the input sets of data into single representative values. In the min-max test, we decide whether two samples are different compounds by comparing the most conservative estimate of intra-sample spectral similarity to the most generous estimate of inter-sample similarity. Formally, we compute the min-max index (𝛿 !! ) as where the functions min(⋅) and max (⋅) denotes the minimum and maximum values contained in the specified sets, respectively. For ease of reading, we will drop the subscript from 𝛿 !! and refer to the min-max index with the symbol 𝛿 as no other spectral comparison indices are computed in this manuscript. As presented in Eq. ( 2) and with score sets 𝑺 𝟏𝟏 , 𝑺 𝟐𝟐 , and 𝑺 𝟏𝟐 constructed using similarity scores that evaluate between 0 and 1 au, the evaluated 𝛿 will be a real number between -1 and 1 au with practical values ranging between -0.1 and 0.9 au. The most intuitive employment of the min-max index is to assess whether 𝛿 > 0 au and infer that the compared sets of spectra are measures of different compounds; the larger the value of 𝛿, the more certain we are of the claim. If 𝛿 ≤ 0 au, there is at least some overlap in the observed intra and inter-sample spectral similarity and so we cannot confidently claim the samples are different compounds. A transformation such as 𝛿 % = 1 − max(0, 𝛿) allows us to compare min-max indices more readily to similarity scores. The transformed min-max index evaluates between 0 and 1 au with uncertainty of a negative confirmation increasing with index values due to increased spectral similarity, as is the case with similarity scores. We can define the min-max test as where 𝜏 &% is a min-max index threshold value, and prediction classes 0 and 1 implying that the spectra are measures of different compounds and the same compound, respectively. An intuitive employment of the min-max test in this formulation is to set 𝜏 & ! = 1 au. For the remainder of this manuscript, all discussion of min-max indices will reference these transformed values. ## Evaluation methodology To demonstrate and evaluate the min-max test, we collated previously published and newly measured mass spectra into a single collection. The collection contained 10 replicate measurements of 144 illicit drug standards (comprised of synthetic cannabinoids, cathinones and opioids), totaling 1440 mass spectra labeled with names and molecular formulae. With these mass spectra, we computed two min-max datasets, the first with min-max indices computed using cosine similarity as the representative similarity score, and the second using identity match factors. These datasets were constructed by computing the min-max indices between all possible pairs of compounds using 3 randomly selected replicate spectra, repeating the experiment 100 times. In cases where the same compounds were being compared, we ensured the selected replicates were not overlapping. The resulting datasets contained 2073600 min-max indices each, approximately 0.7 % of which were computed between the same compound and the rest computed between different compounds. We also generated two similarity score datasets to evaluate the similarity score test for comparison. Each of these datasets were constructed by computing all possible nontrivial cosine similarity scores and identity match factors between the 1440 spectra, thus each dataset contained 2072160 total scores. Approximately 0.6 % of the similarity scores in the datasets were computed between spectra of the same compound and the rest between different compounds. The raw mass spectra, computed datasets, and source code and scripts used to analyze results are available for review. 23 Performance measures: For convenience, we use 𝑝 to denote an index and 𝜏 ' to denote a threshold value associated with 𝑝. A positive prediction is when 𝑝 ≥ 𝜏 ' , and a negative prediction is when 𝑝 < 𝜏 ' . For a set of indices that can be mapped to binary classifications, the number of True Positives (TP) is the count of positive predictions associated with positive classification (i.e., the compared spectra were replicates of the same compound). True Negatives (TN) are the negative predictions associated with negative classification (i.e., the compared spectra were measurements of different compound). False Positives (FP) are positive predictions that should have been associated with a negative classification, and False Negative (FN) are the negative predictions that should have been associated with positive classifications. Several standard performance measures can be derived from these quantities. In this manuscript, we considered accuracy, true positive rate (TPR) or recall, specificity, precision, and false positive rate (FPR) as described by Fawcett 24 and summarized in Table 1. Determining optimal thresholds: There are several options for determining optimal threshold values. One approach is to simply select the threshold value that optimizes the objective function (e.g., maximizes test accuracy) for the entire available dataset. While this method is likely to identify a unique threshold for each objective function, we gain no insights about how the identified threshold will perform with new data. A second approach is to use several subsets of the data to determine a range of threshold values, removing some data dependency and shedding some light on what an ideal threshold might look like for completely new data. In this manuscript, we obtained optimal thresholds using complete datasets and an iterative subset selection approach. For iteratives subsets, we randomly selected 10000 indices, ensuring that exactly 60 were values from replicate mass spectra, and repeated the process 1000 times. We determined threshold indices that maximized (1) accuracy, and (2) the difference between recall (TPR) and FPR. ## Numerical Results To give a general overview of how similarity scores and min-max indices are distributed across our datasets, we generated box and whisker plots with indices distinguished as estimates of either different compounds or the same compound (Figure 2 recall. The observed accuracy of the similarity score test using the identity match factors dataset (Fig 2b) and a threshold of 0.9 au is approaching 100 %, but with a recall of only 50 %. With the two min-max datasets (Figs 2c and d), the effect of similarity score selection is more subtle, and the indices computed between replicate measurents of the same compound hover close to the 1 au decision-making threshold we intuitively expected. The accuracy of the min-max test with a threshold of 1 au is approaching 100 % with recall values also just under 100 %, using either dataset (Figs 2c and d). A comprehensive performance assessment of both tests using both similarity scoring selections and several fixed threshold values is provided in the Supplementary Information. Figures 2a and b also confirm that confident negative confirmations can be made using just the similarity score test. For example, there was not a single false negative (incorrect negative confirmation) using a threshold of 0.7 au with either of the similarity score datasets in this manuscript. However, using such a low threshold is a very inefficient filter. We define the gray area as the range of similarity scores that will falsely characterize true negatives as false positives using the similarity score test. The lowest observed cosine similarity score for a true positive in the cosine similarity dataset is 0.88 au. We refer to this value as the lower gray area threshold. There were 370 pairs of different compounds where at least one of the computed cosine similarity scores was greater that then the lower gray zone threshold, and so would be falsely classified as the same compound if the lower gray zone threshold was used as the threshold in a similarity score test. Of these pairs, 100 % would be correctly classified as true negatives using the min-max test with a 1 au threshold. Similar results were observed with the gray area false positives from the similarity score test being correctly classified as true negatives using the min-max test when identity match factors were used to estimate spectral similarity (see Table S5 in Supplemental Information). The optimal decision-making thresholds for both maximizing the accuracy and maximizing the difference between recall and false positive rates using all combinations of tests and similarity scores are summarized in Table 2. The optimal threshold value for using the min-max test is always around 1 au, regardless of similarity estimate and objective function. There is at least one instance where the threshold that maximized the difference between recall and false positive rate was 0.975 au. Because we used randomly selected subsets of data, we cannot easily disaggregate the exact conditions that led to that threshold value. The optimal threshold values for the similarity score tests varied more substantially with choice of similarity estimate and objective, but the tests performed well. That said, the objective values at the optimal threshold with the similarity score test were always less than the objective values using the min-max test with intuitive threshold 1 au. ## Discussion To begin, we must acknowledge that our seemingly large datasets, with over 2 million data points each, represent a small fraction of the potential chemical space one might explore using EI mass spectrometry -we only considered mass spectra of 144 synthetic cannabinoids, cathinones and opioids. Additionally, our manuscript only evaluated the effect of two different similarity score choices. That said, the numerical results presented in the previous section and as supplemental information support our notion that the min-max test can be an excellent method for objective negative confirmation. The original motivation for developing the min-max test was to overcome the known limitation of the similarity score test -a good decision-making threshold is difficult to select. While using cosine similarities and identity match factors to estimate spectral similarity is mature in its application, the rough reproducibility of mass spectra and non-linearity of the computations underlying these similarity estimates make them difficult to interpret without visual appraisal, hence limiting their objectivity. In our study, we were able to identify a range of optimal thresholds to maximize similarity score test accuracy and the difference between recall and false positive rate using both cosine similarities and identity match factors. These performance results give us a new data-driven basis from which to select decision-making thresholds for similarity score tests goingforward, yet these values lack the type of intuitive grounding necessary to be completely satisfying. And as with any quantity derived from data, there is concern that the dataset used to determine optimal thresholds was inadequate or inappropriate for the next application of the similarity score test. Selecting a threshold value for the min-max test is intuitive. A min-max index of 1 au means that the minimum self-similarity observed within the sets of replicate spectra is equal to the maximum similarity observed between the spectra from either set, implying that the sets of spectra are not discernible. The performance results for the min-max test using the full dataset and the threshold of 1 au were excellent, with accuracy, recall and specificity all at least 99.8 %, regardless of similarity score imposed on the method. Using a slightly more conservative threshold value of 0.977 au, we have slightly worse accuracy and specificity, but observe 100 % recall. The obvious limitation of the min-max test is that it requires replicate mass spectra of each of the compared samples. In an application like seized drug analysis, this requirement is likely not an issue as there is often enough seized evidence and reference standard to take several replicate measurements. However, it is not impossible that a lack of material or analysis time make taking replicate measurements impractical. One useful strategy in these scenarios is to run the similarity score test first, followed by the min-max test if the result falls within the gray area as demonstrated in the numerical results. other more subtle limitation of the min-max test is that poor quality spectra (e.g., spectra containing contaminants, or measured to inconsistent mass limits) will produce unreliable results. The effectiveness of the min-max test stems from our ability to quantify the conceptual notion of self-similarity. If a poor replicate is included in the test, our understanding of self-similarity is misrepresented in the min-max index, and the utility of the min-max test is essentially void. Alternative spectral comparison indices can be formulated using the general framework in Eq. ( 1) that are less susceptible to failing when provided a single incorrect measurement (e.g., choosing function 𝑔(⋅) to select the median value contained in the set). However, these alternate function choices may lack the simple and intuitive justifications that make the min-max test so satisfying. The similarity score test is also susceptible to failure if provided with poor quality mass spectra; this limitation is far more self-evident and is presently defenseless. Our manuscript focused on the utility of the min-max test for negative confirmation tasks. Positive confirmation with mass spectrometry is appreciably more difficult. The performance metric of interest for positive confirmation tasks would be precision; the fraction of positive predictions that are true positives. With our datasets and a threshold of 1 au, the precision of the min-max test precision was greater than 85 % (using either cosine similarity or identity match factors) with nearly 100 % recall. This means that up to 15 % of positive predictions were false positives. Given that our dataset consists of several pairs of isomeric drugs, with near identical mass spectra, this high false positive rate is not surprising. A well-known strategy for enhancing the value of similarity scores is to combine them with measures from orthogonal technologies such as retention indices obtained from gas chromatography. Exploring how we can effectively combine retention indices within on our framework of spectral comparison indices using replicate measurements for the purposes of objective positive confirmation is on-going work in our lab. ## Conclusions Forensic chemists are routinely required to make consequential decisions as quickly as possible. One example of a major decision is confirming that seized evidence does not contain an illegal drug (negative confirmation). The traditional method employed for this task is gas chromatography mass spectrometry followed by mass spectral interpretation. While effective, this approach is inherently subjective. An alternative is to compute numerical estimates of mass spectral similarity and use these values to make decisions. Unfortunately, setting decision making thresholds is a non-trivial challenge, especially across various classes of drugs. In this manuscript, we introduced the min-max test as an alternative method for negative confirmation using mass spectra. We discussed how selecting a threshold for the min-max test is intuitive and demonstrated that the method outperforms automated decision-making using spectral similarity estimates and threshold values, even when the threshold has been optimally chosen for a fixed dataset. In addition to being effective, algorithms and software tools must be objective and explainable to be of any practical use in a forensic science setting. We constructed the min-max test with these two considerations in mind. We believe the min-max test will be an indispensable tool for forensic chemists performing negative confirmation tasks using mass spectra, and that this manuscript provides a template for further developing objective and explainable methods in the forensic sciences. ## Acknowledgements The authors would like to thank Drs. Kearsley, Mallard, Stein and Wallace of the National Institute of Standards and Technology for sharing insights that were instrumental in developing the min-max test. The authors would also like to thank Amber Burns of the Maryland State Police Forensic Sciences Division for sharing practical perspectives on analytical challenges faced by forensic laboratories. ## Author Contributions A.S.M. and E.S. conceived the methods and designed the research plan. E.S. conducted laboratory experiments. A.S.M. performed computational analyses. A.S.M. and E.S. co-wrote the manuscript. ## Competing Interests The authors declare no competing interests. ## Disclaimer Official contribution of the National Institute of Standards and Technology (NIST); not subject to copyright in the United States. Certain commercial products are identified in order to adequately specify the procedure; this does not imply endorsement or recommendation by NIST, nor does it imply that such products are necessarily the best available for the purpose.

chemsum

{"title": "The min-max test: an objective method for discriminating mass spectra", "journal": "ChemRxiv"}

reversible_dispersion_and_release_of_carbon_nanotubes_<i>via</i>_cooperative_clamping_interactions_w

## Abstract: Due to their outstanding electronic and mechanical properties, single-walled carbon nanotubes (SWCNTs) are promising nanomaterials for the future generation of optoelectronic devices and composites. However, their scarce solubility limits their application in many technologies that demand solution-processing of high-purity SWCNT samples. Although some non-covalent functionalization approaches have demonstrated their utility in extracting SWCNTs into different media, many of them produce short-lived dispersions or ultimately suffer from contamination by the dispersing agent. Here, we introduce an unprecedented strategy that relies on a cooperative clamping process. When mixing (6,5)SWCNTs with a dinucleoside monomer that is able to self-assemble in nanorings via Watson-Crick base-pairing, a synergistic relationship is established. On one hand, the H-bonded rings are able to associate intimately with SWCNTs by embracing the tube sidewalls, which allows for an efficient SWCNT debundling and for the production of long-lasting SWCNT dispersions of high optical quality along a broad concentration range. On the other, nanoring stability is enhanced in the presence of SWCNTs, which are suitable guests for the ring cavity and contribute to the establishment of multiple cooperative noncovalent interactions. The inhibition of these reversible interactions, by just adding, for instance, a competing solvent for hydrogen-bonding, proved to be a simple and effective method to recover the pristine nanomaterial with no trace of the dispersing agent. ## Introduction Carbon nanotubes (CNTs) are nanomaterials with impressive electronic and mechanical properties and are considered as strong candidates for the next generation of transistors, photovoltaics, and (bio)chemical sensors. However, they are insoluble in most solvents and infusible at any temperature, due to strong bundling caused by numerous van der Waals interactions, which hampers many of their applications. 4,5 Covalent and noncovalent surface modifcation with molecules, leading to homogeneous CNT dispersions in various solvents or bulk materials, is now common practice to facilitate their processing and to attach specifc functions. The covalent approach grants more stable dispersions, but the modifcation of the p-conjugated CNT surface with grafted molecules has a detrimental impact on their (semi)conducting and mechanical properties. That is the reason why noncovalent strategies, achieved by promoting strong interactions between CNT sidewalls and p-conjugated or apolar moieties in adequate solvents, are often considered more suitable. However, the weak and dynamic nature of supramolecular interactions frequently produces less durable dispersions of individual CNTs because bundling, which ultimately leads to CNT reprecipitation, remains as a strongly competing supramolecular process in solution. Bundling can be reduced by increasing the number of surface-binding functions in the dispersing agent and, importantly, by inducing wrapping interactions around single CNT cylinders (interaction mode II in Fig. 1a below). A large number of composites with p-conjugated polymers, 12,13 DNA, synthetic peptides, or, in general, oligomers featuring multiple aromatic units have been tested to stabilize CNTs selectively in solution by specifc helical wrapping conformations around the tube. A main drawback that originates from such a robust marriage is that polymer desorption via washing processes is often difficult to achieve, and CNTs purifed or processed in this way may remain contaminated with the dispersing agent. 25,26 In order to address this issue, recent strategies have been reported that focus mainly on: (1) switching between tightly and loosely bound polymer conformations, which can be triggered thermally, photochemically 27 or by a change in solvent 28 or pH, 29 and on (2) inducing depolymerization processes, which can also be made reversible by introducing supramolecular or dynamic covalent bonds 35 along the polymer main chain. Here, we introduce a novel approach that is instead based on clamping discrete self-assembled nanorings around the tube cross-section to efficiently and reversibly produce durable dispersions of individual single-walled CNTs (SWCNTs) in organic solvents. The encapsulation of SWCNTs within well-defned macrocycles, 19, forming rotaxane-type ensembles, has been recently explored using covalent cyclization reactions. The strategy followed by some of us relied on two extended aromatic binding sites to promote the supramolecular association of U-shaped molecules to SWCNT, followed by "clipping" through ring-closing metathesis to produce the mechanically interlocked species. Depending on the chemical nature of the recognition motif, we have also observed the formation of oligomers that wrap around the SWCNTs. 38 The approach followed herein, in contrast, profts from a dynamic, strongly cooperative noncovalent macrocyclization process. Our design focuses on a dinucleoside monomer (GC1; Fig. 1b) that has been rationally synthesized to comply with these objectives (see the ESI † for synthetic details). GC1 structure consists of complementary guanosine (G) and cytidine (C) DNA bases connected by a rigid, linear central block. We demonstrated recently that related molecules self-assembled into cyclic tetramer species 42 through Watson-Crick G-C Hbonding interactions 43 exhibiting record chelate cooperativities, which allow the nanorings to be formed quantitatively in nonpolar solvents within a wide concentration range. The ribose groups in GC1 feature multiple long alkyl chains that, upon cyclic assembly, would point in all directions towards the exterior medium, which should beneft individual CNT debundling and thus afford high solubility to the fnal composites. The monomer-CNT interaction strength is an additional key factor in our design. A p-conjugated dialkoxyarene central block with modest affinity for the CNT surface has been installed between the lipophilic bases. The idea is that stable dispersions would only be obtained if the rings are able to embrace the tube, so that its sidewalls can interact with several monomers in the nanoring cavity (mode IV in Fig. 1), and not through external binding with sections of the rings (mode III) or with individual monomers or linear oligomers (modes I/II). We also explore herein the special synergy of the GC1-SWCNT marriage, which simultaneously brings an enhanced solubility to the nanotubes (when compared to a related molecule that cannot cyclize: CC1) and an increased stability to the macrocycles. This synergy originates from the cooperative action of G-C H-bonding and monomer-nanotube van der Waals interactions, so we reasoned that the disturbance of any of these distinct noncovalent interactions would cause collapse of the supramolecular ensemble, thus facilitating the recovery of the extracted CNTs. ## Initial experiments and sample preparation Prior to their combination with CNTs, we confrmed that this novel GC1 monomer displayed a similar self-assembly process to the one already reported by us with closely related dinucleosides (see the ESI † for further details). 47,50 NMR and optical spectroscopy experiments indicated that GC1 tetrameric rings are indeed formed close to quantitatively in apolar chlorinated solvents within the 10 1 to 5 10 4 M concentration range. These cyclic assemblies are characterized by red-shifted and low intensity emission maxima at ca. 505 nm, and by the presence of a characteristic negative Cotton effect, with maxima at 340 and 387 and a minimum at 428 nm. At lower concentrations, they dissociate gradually into monomeric species, which display distinct emission maxima at 421 and 445 nm and null CD signals (vide infra). The addition of polar cosolvents that can compete for H-bonding, like DMSO or DMF, also results in monomer dissociation. The 1 H NMR spectra of GC1 recorded by modifying the CDCl 3 : DMF-D 7 volume ratio (Fig. S1 †) revealed a strong all-or-nothing behavior: no signifcant participation of any other H-bonded oligomer but the tetrameric macrocycle is detected in solution. This is in agreement with the formation of stable ring species with remarkably high chelate cooperativities, as determined in our previous work. An initial theoretical study using DFT calculations (see below and the ESI † for further details) served to properly design the interacting ring-tube system and experimental measurements. We decided to use (6,5)-enriched SWCNTs with a mean diameter of 0.7 nm and very narrow polydispersity (we observe (6,5) SWCNTs exclusively in photoluminescence experiments; see Fig. S2 †) because they should ft adequately within the GC1 nanoring cavity. Thus, a dispersion of (6,5)SWCNTs in CHCl 3 (0.2 mg mL 1 ) was produced frst by ultrasonication followed by centrifugation, in order to remove large SWCNT bundles and any other carbonaceous impurities. These non-stabilized dispersions are rather short-lived and signifcant precipitation of (6,5)SWCNTs was clearly observed after a few hours (see central image in Fig. 2a). Immediately after centrifugation, GC1 (or CC1) solutions in CHCl 3 (different concentrations were tested, from 5.0 10 4 to 10 6 M) were added to the supernatant (6,5)SWCNT dispersion and the mixtures were stirred at room temperature. The (6,5)SWCNT-GC1 suspensions produced (right image in Fig. 2a) were kept under dark and checked at different periods of time. A few spectroscopic changes were observed within the frst hours after mixing, but after ca. 12 hours spectral properties remained constant for weeks and are reproducible. ## Spectroscopic and thermogravimetric measurements The (6,5)SWCNT-GC1/CC1 composites were analyzed by absorption, CD, Raman, and emission spectroscopies and by TGA, and compared with pristine (6,5)SWCNTs and GC1/CC1 in the same conditions (Fig. 2). The absorption spectra (Fig. 2b) showed the typical features of both components in the mixture and only minor deviations in absorption maxima were noted. The typical scattering phenomena, evidenced by a baseline rise, was notorious in the original (6,5)SWCNTs dispersions and in mixtures with CC1 or with low GC1 content. However, as the GC1/(6,5)SWCNTs ratio is increased, scattering is concomitantly reduced until it is no longer perceptible in the absorption spectrum, which underlines the high dispersing power of this agent. As a matter of fact, the GC1-(6,5)SWCNTs mixtures produced are clear suspensions (Fig. 2a) that show no evidence for nanotube precipitation along several weeks. CD can be considered as one of the most powerful techniques for stereochemical analysis: it is sensitive to the absolute confguration as well as to conformational features, which are often completely obscured in ordinary absorption spectra. 51 Regarding this, the CD spectral shape of GC1-(6,5)SWCNTs composites does not exhibit important differences with GC1 in the same conditions (Fig. 2c). This is a strong indication that the cyclic tetramer structure is preserved in the presence of (6,5) SWCNT, and no dissociation or structural change in the Hbonded assembly was noted. In order to obtain additional proof from a theoretical perspective, we calculated the CD spectrum for the optimized structure of cGC1 4 (see Fig. 3g below), which is displayed as a dotted red line in Fig. 2c. Our calculations revealed a strong dependency of the CD spectra with the symmetry of the macrocycle. Only when the C 4 axis was maintained in the cGC1 4 ring structure, we obtain a good match between the experimental and the theoretical results (with the exception of the vibronic structure of the positive Cotton effect). If the C 4 symmetry axis is lost, due for instance to other binding modes of GC1 to the CNT sidewalls, the CD spectrum is signifcantly perturbed. As an example, we show in Fig. S3 † the structure and calculated spectrum for a cGC1 4 structure, which is close in energy, but has lost the C 4 symmetry. We can therefore conclude that this C 4 symmetry is conserved upon the formation of (6,5)SWCNT-cGC1 4 conjugates, which is a strong indication that the SWCNT is inside the cavity of cGC1 4 ring. Another important proof that demonstrates GC1-(6,5) SWCNTs interactions came from comparing the emission spectra (Fig. 2d) with and without CNTs. As a matter of fact GC1 fluorescence is considerably quenched in the presence of (6,5) SWCNTs. This is very characteristic in noncovalent SWCNT assemblies 22 and is presumably caused by an energy transfer process between the dinucleoside p-conjugated system and the SWCNT when they are in close contact. On the other hand, TGA studies showed that macrocycle loading increases with GC1 concentration, to yield 17% and 28% loading at 10 6 and 10 4 M GC1 initial concentration, respectively. Finally, no signifcant (6,5)SWCNT electronic perturbance was noted in Raman experiments when GC1 or CC1 were added, which is a good indication of the formation of noncovalently bound composites. ## AFM, TEM and DFT analysis Exploration of the GC1-SWCNT suspensions under atomic force microscopy (AFM) was consistent with a picture in which the SWCNTs are encapsulated within GC1 macrocycles. AFM images were obtained upon spin-coating the SWCNT suspensions on mica, and show individualized SWCNTs decorated with objects of about 2 nm in height (Fig. 3a and b), in full accord with the DFT-modelled size of the GC1 tetramer (Fig. 3h). We also explored our samples under transmission electron microscopy (TEM), where we observe mostly rebundled SWCNTs, most likely due to sample preparation issues. However, wherever individualized SWCNTs were located, they showed heavily functionalized sidewalls (Fig. 3c and d). Unfortunately, and in contrast to covalently linked macrocycles, 41 attempts at obtaining higher resolution TEM images were precluded by the instability of the H-bonded organic macrocycle in the conventional transmission electron microscope at 200 kV. High resolution images were obtained in an aberration corrected microscope at 60 kV (Fig. 3e and f). We observe structures of around 2.0 nm surrounding the SWCNTs that are consistent with the picture provided by spectroscopy, but even under low voltage (see also Fig. S5 †), the macrocycles were observed to quickly reorganize and eventually decompose under the electron beam irradiation, probably due to radiolysis damage associated to hydrogen. Further theoretical studies were performed in order to gain a deeper insight into the structure and interactions in the (6,5) SWCNT-cGC1 4 composites. As explained above, the optimized geometry of cGC1 4 belongs to the C 4 point group of symmetry and is depicted in Fig. 3g. The inner cavity has a mean diameter of 2.1 nm and the central aromatic moieties are bent at an angle of 30 , so that planarity is not conserved, allowing the carbon nanotube to maintain a stronger interaction with the ring. In order to have a second proof of the arrangement between the two moieties, we additionally calculated the intensity of the interaction between SWCNTs and cGC1 4 ring with close p-p contacts in the range of 3.2-3.5 . Among the different possible confgurations (see Fig. S4 †), we observed that the (6,5)SWCNT and the cGC1 4 tetramer interact strongly (E int ¼ 22.7 kcal mol 1 ) when the macrocycles are clamping the nanotube and the macrocycle tilts slightly with respect to the tube axis, in order to maximize non-covalent interactions (interaction mode IV in Fig. 1). In contrast, all external binding modes investigated (interaction mode III in Fig. 1), result in positive (i.e. disfavorable) interaction energies ranging between +0.7 and +42.0 kcal mol 1 . The geometry of the lowest energy confguration is shown in Fig. 3h. ## Analysis of the synergistic interactions Concentration-dependent experiments, where we recorded absorption, CD and emission spectra, provided a deeper understanding of the mutual interaction between CNTs and Hbonded nanorings. Two parallel dilution measurements were conducted from GC1 CHCl 3 solutions: one in the absence (Fig. 4a and c) and the other one in the presence of (6,5)SWCNTs (Fig. 4b and d). The spectral changes with concentration were monitored by CD from 4.0 10 4 M (blue lines) down to 3.0 10 6 M (CD) or down to 2.0 10 7 M (emission) (red lines). The trends obtained from both techniques are compared in Fig. 4i. A frst experimental observation that came from comparison of both dilution experiments is that the presence of (6,5) SWCNTs preserved the characteristic cyclic tetramer features along a wider concentration range. This is unambiguously observed in the evolution of the CD spectra as a function of concentration. While GC1 CD features readily disappear below ca. 10 5 M (Fig. 4a), indicating cyclic tetramer dissociation, they remain distinct down to 3 10 6 M in the presence of (6,5) SWCNTs (Fig. 4b). The degree of cyclotetramerization, that is, the molar fraction of GC1 molecules associated as cyclic tetramers (horizontal equilibria in Fig. 1a), can be calculated from each dilution experiment by integrating CD intensity. The comparison of both trends, represented with solid blue and green circles in Fig. 4i, suggests that GC1 nanorings are signifcantly stabilized when mixed with (6,5)SWCNTs. Fitting these data to cyclotetramerization processes (blue and green lines in Fig. 4i) afforded cyclotetramerization constants (K T ) in the order of K T ¼ 3.2 10 14 M 3 for GC1 and K T ¼ 4.0 10 16 M 3 for GC1 + (6,5)SWCNTs. That is, chelate cooperativity greatly benefts from the presence of SWCNTs and nanoring stability is considerably increased. We then turned our attention to the evolution of the fluorescence features in these dilutions experiments. The emission spectra were analyzed in two different ways, attending to: (1) their relative intensity, which can be correlated to the fraction of CNT-bound molecules (vertical equilibrium in Fig. 1a), or (2) the spectral shape and emission wavelength, which reports again on the degree of cyclotetramerization (horizontal equilibria in Fig. 1a). In the frst case we compared GC1 emission intensity in the 400-650 nm range in the absence (I GC ) or presence (I GC-CNT ) of (6,5)SWCNTs at each concentration (see Fig. S6A †). The degree of emission quenching, defned as 1 I GC-CNT /I GC , was then calculated and plotted as red squares in Fig. 4i. Since energy transfer from the photoexcited dinucleoside molecules to the CNT is close to quantitative when they are closely interacting, as determined from the fully quenched emission of the samples at high concentration, these curves actually reveal the molar fraction of GC1 that is bound to (6,5)SWCNTs at each concentration. Our results show that GC1-(6,5)SWCNTs association is virtually quantitative at concentrations above 10 4 M and then falls in the 10 4 to 10 6 M range. This nonlinear trend suggests a cooperative GC1-(6,5)SWCNT interaction that seems to be coupled to the cyclotetramerization equilibrium. In order to compare our GC monomer with a related dispersing agent that does not associate in cyclic systems, we carried out another set of parallel dilution experiments with CC1 in the absence and presence of (6,5)SWCNTs. The degree of emission quenching was likewise calculated at each concentration and the results are shown as orange squares in Fig. 4i (see also Fig. S6B †). The CC1-(6,5)SWCNTs association is clearly weaker and no longer detected below 10 5 M, a concentration where ca. half of the GC1 molecules are still bound to CNTs. In the second case, the degree of cyclotetramerization was estimated by analysing the shape of each normalized fluorescence spectra during the dilution measurements shown in Fig. 4c and d, as we described in our previous work (see also the ESI †). 47,50 As commented above, when the monomeric species, showing emission maxima at 421 and 445 nm in CHCl 3 (see red spectra in Fig. 4c and d), associate in cyclic tetramers, a red-shift to >500 nm and a reduction in emission intensity is noted. The results obtained at each concentration are also included in Fig. 4i as solid green and blue triangles for the GC1 samples without and with CNTs, respectively. The cyclic tetramer association trends calculated by CD and fluorescence spectroscopy (green circles and triangles) display a quite decent correlation for GC1. That is not the case when (6,5)SWCNTs are present (please compare blue circles and triangles), and this is because CD and emission spectroscopy are not reporting on the same GC1 population. Whereas the measured CD spectra is representative of all GC1 molecules in solution, the emission spectra primarily provide information of the fraction of molecules that are not bound to the CNTs, since the emission of the GC1 molecules that are bound is strongly quenched. As a consequence, the shape evolution of the emission spectra when the CNTs are present exhibits again a strong coupling between the horizontal cyclotetramerization and vertical GC1-(6,5)SWCNT association equilibria in Fig. 1a. At concentrations above 10 4 M most GC1 molecules are bound to the CNTs and the residual emission recorded is extremely weak in intensity and representative of the cyclic tetramer in shape and emission wavelength (see blue spectrum in Fig. 4d as an example). However, when GC1-(6,5)SWCNT association is no longer quantitative in the 10 4 to 10 6 M range, the emission spectra becomes rapidly dominated by the fraction of GC1 molecules that are not bound to CNTs, which have stronger monomer-like features (red spectrum in Fig. 4d) because their actual concentration is lower. This is reflected in a very sharp transition around 10 4 M in the blue-triangle trend in Fig. 4i. In other words, in the 10 4 to 10 6 M range, the shape of the emission spectrum is more shifted to the monomer features in the presence of CNTs, since the actual concentration of emissive GC1 molecules (i.e. not bound to CNTs) is lower than in the absence of CNTs. Only when the GC1-(6,5)SWCNT interactions become no longer important, below 10 6 M, samples with and without CNTs display similar emission intensity and shape (see also Fig. S6A †). We believe the graph in Fig. 4i provides a rather faithful description of the self-assembly of this two-component mixture in solution as a function of concentration. The results gathered from these dilution experiments in CHCl 3 clearly indicate that each species benefts synergistically from the presence of the other, as it is schematically represented in Fig. 5. On one hand, the nanorings are more stable in the presence of nanotubes, a gain that is represented by the blue area in Fig. 4i. On the other, the nanotubes can host a higher number of dispersing agent molecules and thus enjoy enhanced solubility along a broader concentration range when the monomer cyclizes, a gain that is represented by the red area in Fig. 4i when comparing GC1 and CC1. In short, H-bonding between complementary bases and van der Waals dispersion forces between monomer and the p-conjugated CNT sidewalls are noncovalent interactions that work here cooperatively to build strongly-associated GC1-(6,5)SWCNTs composites. It is interesting to note that smaller nanorings, where the nucleobases are directly connected by a triple bond (see GC3 in the ESI †), whose cavities cannot host (6,5)SWCNTs, do not exhibit the same synergistic effects as GC1, but rather behave as a regular dispersing agent, like CC1. On the other hand, GC1 cannot solubilize as efficiently multi-walled nanotubes of larger diameter. Both observations are in agreement with the clamping interaction mode IV, shown in Fig. 1a and 5, being the most likely and abundant in CHCl 3 solutions (Fig. 1a). We should nonetheless consider that, when the nanorings are assembled around the tube, local concentration effects may occur, so that a rearrangement into linear polymers may take place when several rings coincide locally. However, these ring-to-chain rearrangements should lead to a reduction in CD intensity at high concentrations that we did not observe, so we can discard this is a relevant situation. As we reduce concentration, these effects are even less likely to occur, because the rings are favoured entropically. ## Recovery of the pristine SWCNTs At this point we reasoned that the whole assembly could be demolished by addressing just one of the supramolecular interactions, which could be a simple and straightforward strategy for pristine (6,5)SWCNTs recovery. For instance, an increase in solvent polarity should disrupt H-bonding interactions without strongly affecting monomer-(6,5)SWCNTs van der Waals interactions. In order to prove this, we performed now the same dilution experiments from GC1 samples with and without CNTs by gradual addition of DMF instead of CHCl 3 , and recorded again nanoring dissociation by CD (Fig. 4e and f; open blue and green circles, respectively, in Fig. 4i), and the degree of emission quenching by fluorescence spectroscopy (Fig. 4g and h; open red squares in Fig. 4i). In line with the previous observations, the cycles are more stable and the CD features resist a higher amount of DMF when (6,5)SWCNTs are present (please compare open blue and green circles). Together with the disappearance of the CD signals, GC1 monomer emission was recovered in a narrower concentration range when the parent GC1-(6,5)SWCNTs dispersion was diluted with DMF instead of CHCl 3 (please compare solid and open red squares), which suggests much weaker interactions with the CNTs in this polar solvent. Interestingly, several minutes after DMF addition a clear precipitate emerged from the original dispersion that, after washing, showed no residual GC1 spectroscopic features. Despite the strong association and remarkable endurance of the GC1-(6,5)SWCNTs solutions in CHCl 3 , which can last for several weeks maintaining the original optical quality, CNT rebundling and precipitation occurred rapidly in DMF, likely due to the lower efficiency of the dissociated monomer as a dispersing agent (see Fig. 5). It is also interesting to note that cGC1 4 dissociation can also be induced by increasing the temperature in CHCl 3 or CHCl 2 CHCl 2 solutions. After being subjected to a heating-cooling cycle the spectra of the initial and fnal samples differed considerably and clear precipitation of the CNTs was noted. However, the results are highly dependent on the concentration and not as reproducible and efficient as the addition of DMF to release and recover the pristine (6,5)SWCNTs. ## Conclusions In conclusion, we have explored herein an unprecedented approach to solubilize SWCNTs in apolar solvents that relies on dynamic macrocycle clamping around the nanotube sidewalls, which allows for efficient SWCNT debundling, and on cooperative noncovalent interactions, which supplies the required reversibility to simply and effectively recover the pristine material. The combination of theoretical DFT-based (6,5)SWCNTs (bottom). At high concentrations in CHCl 3 , GC1 ring assemblies are formed quantitatively and establish strong interactions with the CNTs. Upon dilution with CHCl 3 the cyclic tetramers gradually dissociate, but such dissociation occurs to a lower extent when CNTs are present, due to the stronger (6,5) SWCNT-cGC1 4 clamping interactions. On the contrary, upon dilution with DMF, the cGC1 4 rings are fully dissociated at relatively high concentrations, which eventually produces CNT precipitation due to the weaker interaction of GC1 monomers with the (6,5)SWCNTs. In this last situation, a simple filtration and washing protocol allows to separate efficiently the (6,5)SWCNTs from the GC1 monomer. methodologies, spectroscopic techniques, as well as AFM and TEM microscopies, provide solid evidence for a preferred association mode where the H-bonded nanorings are embracing the tube (mode IV in Fig. 1). Furthermore, a comparison between dilution experiments performed on GC1 in the absence or presence of (6,5)SWCNTs provided a deep insight into the mutual benefts offered by the combination of these two species. On one hand, nanoring stability is unambiguously enhanced in the presence of the CNTs, likely due to the interaction of more than one GC1 monomer with the tube sidewalls. On the other, the GC1 molecule exhibited an extraordinary solubilizing power, in comparison with a regular dispersing agent that cannot cyclize, like CC1. When comparing these two monomers, we observe dramatic differences in terms of the quality of the dispersions produced (that only show very low scattering in the case of GC1; see Fig. 2b), their durability (the dispersions produced from CC1 show a CNT precipitate after a few days, while those from GC1 were seen to resist for many weeks), and the resistance of the monomer-SWCNT interaction to dilution (as determined from the experiments shown in Fig. 4). We believe that these remarkable differences, which are translated in the generation of clear, long-lasting (6,5) SWCNTs dispersions along a broad concentration range, can only be explained if the dispersing agent is able to surround individual nanotubes, as in mode IV in Fig. 1. Future efforts will be directed to study if our self-assembled nanorings can selectively extract CNTs as a function of diameter and chirality.

chemsum

{"title": "Reversible dispersion and release of carbon nanotubes <i>via</i> cooperative clamping interactions with hydrogen-bonded nanorings", "journal": "Royal Society of Chemistry (RSC)"}

linear_atomic_cluster_expansion_force_fields_for_organic_molecules:_beyond_rmse

## Abstract: We demonstrate that fast and accurate linear force fields can be built for molecules using the Atomic Cluster Expansion (ACE) framework. The ACE models parametrize the Potential Energy Surface in terms of body ordered symmetric polynomials making the functional form reminiscent of traditional molecular mechanics force fields. We show that the 4 or 5body ACE force fields improve on the accuracy of the empirical force fields by up to a factor of 10, reaching the accuracy typical of recently proposed machine learning based approaches. We not only show state of the art accuracy and speed on the widely used MD17 and ISO17 benchmark datasets, but also go beyond RMSE by comparing a number of ML and empirical force fields to ACE on more important tasks such as normal mode prediction, high temperature molecular dynamics, dihedral torsional profile prediction and even bond breaking. We also demonstrate the smoothness, transferability and extrapolation capabilities of ACE on a new challenging benchmark dataset comprising a potential energy surface of a flexible drug-like molecule. ## Introduction The efficient simulation of the dynamics of molecules and materials based on first principles electronic structure theory is a long standing challenge in computational chemistry and materials science. There is a trade-off between the accuracy of describing the Born-Oppenheimer potential energy surface (PES) 1 and the length and time scales that are accessible in practice. A convenient way to measure this trade-off is by considering the total number of simulated atoms, which can be a result of either generating a few configurations consisting of many atoms, or many configurations (e.g. a long molecular dynamics trajectory) each consisting of fewer atoms. Explicit electronic structure simulations are extremely accurate and systematically improvable. They can treat on the order of a million simulated atoms in total using either cubic scaling methods and molecular dynamics, or linear scaling algorithms on larger systems. Alternatively, in order to simulate many orders of magnitude more atoms, the PES can be parametrized in terms of the nuclear coordinates only. In this way, the electrons do not have to be treated explicitly, which simplifies the simulations con-siderably. These methods can routinely model a trillion (10 12 ) or more simulated atoms. When parametrizing the PES, it is natural to decompose the total energy of the system into body ordered contributions, which can then be resummed into local atomic (or site) energies. The site energy of atom i is written as where indices j, k run over all neighbors of atom i (either unrestricted, or within a cutoff distance r cut ), z i denotes the chemical element of atom i and r ij = r j − r i the relative atomic positions. The traditional approach to the parametrization of the body ordered terms for molecular systems is to use physically motivated simple functional forms with few parameters, leading to "empirical force-fields". These models typically require a pre-determined topology, meaning that the parameters describing the interactions of a certain atom depend on its neighbors in the bonding graph that is specified before the simulation and is not allowed to change. The potential energy is then written as a sum of body-ordered bonded and non-bonded terms, for example: where r, θ and φ describe the intramolecular bond lengths, angles and dihedral angles in the molecule, and E non−bonded contains a Lennard-Jones (LJ) term accounting for van der Waals and short-range repulsive interactions and a Coulomb term to describe the long-range electrostatics. The bonded terms can be made equivalent to the body order in eq (1) by rewriting the sum over atom-tuples into sums over sites. The advantage of the simple functional form of the bonded terms is very fast evalua-tion and ease of fitting due to the small number of free parameters. 2, On the other hand, this simplicity limits the achievable accuracy 9 and requires significant modification to incorporate reactivity. 10 Note that while in the most widely used force fields, the non-bonded interactions are two-body, this is not the case for polarizable force fields, such as Amoeba. 11 Moreover, the direct evaluation of terms beyond 3-body contributions is computationally expensive, in general growing exponentially with the body order, which severely limits the possibility of systematically improving force fields by adding higher body order terms. Over the past ten years a new approach has emerged, employing machine learning (ML) methods to parametrize the PES. Instead of the body order expansion, the site energy is approximated by a neural network or a Gaussian process regressor (GPR) both of which are extremely flexible functional forms, proven to be universal approximators. 25 Due to this flexibility there is no need to specify topology or atom types beyond the identity of the chemical element, and much higher model accuracy can be achieved given an appropriate (typically rather large) training set. On the other hand, this flexibility comes also at a cost: there is no guarantee that the behavior of these ML models remains chemically sensible in regions of configuration space where there is not enough training data. Spurious local minima or even wildly wrong atomization energies are par for the course. 26 The most prominent examples of ML models are Atom Centred Symmetry Function based feed forward neural networks introduced by Behler and Parinello 27 that also includes the family of ANI force fields 17,28 and DeepMD, 18 the atomic neighborhood density based GPR models like Gaussian Approximation Potentials (GAP) 14,29 and FCHL, 15 the gradient domain kernel based sGDML, 16 and message passing graph neural network based Schnet, 23 Physnet, 24 DTNN, 30 and DimeNet 21 and most recently the covariant or equivariant neural network based Cormorant 22 and PaiNN. 19 There is also a third family of methods, which expands the PES as a linear combination of body-ordered symmetric polynomial ba-Table 1: Comparison of different force field fitting approaches. Molecular mechanics (e.g. AMBER, 2 CHARMM 12 and OPLS 13 ), machine learning: Kernels (GAP, 14 FCHL 15 and sGDML 16 ), Neural Networks (ANI, 17 DeepMD, 18 PaiNN, 19 GMsNN, 20 DimeNet, 21 31,32 (Permutationally Invariant Polynomials (PIPs)), which approximated the PES of small molecules to extremely high accuracy, albeit with exponential scaling in the number of atoms. Introducing finite distance cutoffs reduces this scaling to linear, and the resulting atomic body-ordered permutationally invariant polynomials (aPIPs) have been shown to achieve high accuracy and better extrapolation compared to the above nonlinear machine learning based approaches in both molecular and materials systems. 26,33 The main limitation of the aPIPs approach is that the evaluation time of the site energy increases quickly with body order, making it essentially impossible to go above body-order 5 (certainly when the five atoms are of the same element). More recently, the Atomic Cluster Expansion (ACE) 34,35 (and the earlier Moment Tensor Potentials 36 ) are formulations of symmetric polynomial approximations that remove the steep scaling of the evaluation of the site energy with the number of neighbors independently of body order, resulting in highly efficient interatomic potentials for materials. 37 Table 1 compares the main features of the classical force fields, machine learning based potentials and the linear Atomic Cluster Expansion force fields. In one sense, the linear ACE constitutes a middle ground between the other two: it retains the chemically natural body order, but lifts the limitations of fixed topology and inflexible functional form embodied in eq (2). The purpose of the present paper is to demonstrate the performance of linear ACE force fields for small organic molecules. After briefly reviewing the general ACE framework and outlining the necessary choices that go into fitting our linear models, we start with the MD17 38 and ISO17 23 benchmark data sets. We are particularly interested in going beyond the RMSE (or MAE) of energies and forces (the typical target of the loss function in the fit), because practically useful force fields have other desirable properties too: chemically sensible extrapolation, good description of vibrational modes, and accuracy on trajectories self-generated with the force field, just to name a few. The insufficient nature of mean error metrics has been pointed out before. In addition to the above data sets, we also demonstrate the use of ACE on a slightly larger, significantly more flexible molecule that is more representative of the needs of medicinal chemistry applications. The programme of tests as we outlined is designed to explore the capabilities and properties of different approaches to making force fields. We emphasize here that we are not making or testing force fields that are in and of themselves generally useful to others. That is a significant undertaking and it is to be attempted once we better understand these capabilities and properties, and are able to select which approach has the best prospects. Therefore, in addition to quoting literature results for recently published ML schemes, we refit a number of them, where the necessary software is available (sGDML, ANI and GAP in particular), so that we can show their performance on our tests. We also refit a classical empirical force field (eq (2)) to exactly the same training data to more rigorously quantify the anticipated accuracy gains of the ML and ACE approaches. ## Atomic Cluster Expansion basis functions The atomic cluster expansion (ACE) model 34,35 keeps the body ordering of terms defined in eq (1), but reduces the evaluation cost by eliminating the explicit summation over atomtuples. This is accomplished by projecting the atomic neighbor density onto isometry invariant basis functions. This idea, detailed below, is referred to as the "density trick", and was introduced originally to construct the power spectrum (also known as SOAP) and bispectrum descriptors 14,42 (which are in fact equivalent to the 3-and 4-body terms in ACE, respectively, so in a sense the ACE invariants can be considered a generalization of these to arbitrary body order). We start by defining the neighborhood density of atom i as where ρ z i denotes the density of atoms of element z in the neighborhood of atom i. This density is projected onto a set of 1-particle basis functions, which we choose to be a product of a radial basis and real spherical harmonics: Here the "1-particle" refers to the single sum over neighbors, with the central atom i serving as the center of the expansion. There is considerable flexibility in the choice of the radial basis; the specifics for this work are documented at the end of this subsection. We then define the atomic base as the projection of the neighborhood density onto the 1-particle basis functions where the index z i refers to the chemical element of atom i. For notational convenience, we collect the rest of the 1-particle basis indices into a multi-index, From the atomic base A z i v , we obtain permutation-invariant basis functions, which we will call the "A-basis", by forming the products, The product containing ν factors gives a basis function that is the sum of terms each of which depends on the coordinates of at most ν neighbors, and we refer to it either as a ν-correlation or as a (ν +1)-body basis function (the extra +1 comes from the central atom i). A graphical illustration of this construction is shown in fig 1 for the special case where the two factors are the same. For many (different) factors, taking products of the atomic base (left side of fig 1) takes a lot less time to evaluate than the explicit sum of all possible products (right side of fig 1). This is the key step that we referred to as the density trick. The A-basis is not rotationally invariant. We therefore construct a fully permutation and isometry-invariant overcomplete set of functions, which we call the B-basis (technically not a basis but a spanning set), by averaging the Abasis over the three dimensional rotation group, Figure 1: Construction of high body order invariant basis functions. A graphical illustration showing how higher body-order basis functions can be constructed as products of the projected neighborhood density. The evaluation cost of the basis functions scales linearly with the number of neighbors rather than exponentially by doing the density projection first and than taking the products to obtain higher order basis functions. The figure (and expression) also makes explicit the occurrence of self-interaction terms in the ACE basis. They are automatically corrected through the inclusion of lower-order correlations in the basis. O(3), where the matrix of Clebsch-Gordan coupling coefficients C vv is extremely sparse. Many of the resulting basis functions will be linearly dependent (or even zero), but it is relatively straightforward to remove these dependencies in a pre-processing step, to arrive at an actual basis set. We refer to Dusson et al. 35 for the details of the procedure outlined up to this point. The B-basis in eq (8) is complete in the sense that any function of the neighboring atoms that is invariant to permutations and rotations can be expanded as a linear combination of the basis functions. We therefore write the site energy of ACE as The above equation makes it clear that the model is linear in its free parameters, the c coefficients. The B-basis functions are polynomials of the atomic coordinates, and in order to show that the explicit body ordering has been retained, we can switch back to using the Abasis (with the product explicitly written out), where the c can be obtained as linear combina-tions of the c coefficients appearing in eq (10), using the transformation defined in eq (9). Now the body-ordering is readily identified. Each term corresponds precisely to a sum of ν-correlations, i.e. (ν + 1)-body terms as in the traditional body-order expansion, eq (1). In practice, we use a recursive scheme 35 that leads to an evaluation cost that is O(1) per basis function, independent of body-order. The number of basis functions does grow with body order, at a rate that has an exponent ν. The construction outlined so far yields infinitely many polynomials B z i v , which can be characterized by their correlation-order ν, and their (modified) polynomial degree D = ν t n t + w Y l t , where n t and l t come from the multi-index v t and the weight w Y is used to trade-off the radial and angular resolution of the basis set. When it comes to defining a model in practice the expansion is truncated both in the body-order and in the maximum polynomial degree at each body-order. ## Choice of radial basis In the models in this paper we will not use much of the flexibility of the ACE framework, and simply take R r → x(r) is a one dimensional radial transformation, f cut is a cutoff or envelope function and p n are orthogonal polynomials. For the radial transform we take which amplifies the effect of neighbors closer to the central atom. For the cutoff function we specify both inner and outer cutoffs, r in < r out , and define The polynomials p n are then defined recursively by specifying that p 0 (x) = 1, p 1 (x) = x, and the orthogonality requirement where we have used the inverse of the radial transform, x → r(x). Eq (15) implies that the radial basis R n and not the polynomials p n are orthonormal in x-coordinates. The introduction of an inner cut-off is necessary to prevent wildly oscillating behaviour in high energy regions of configuration space where pairs of atoms are very close to one another and little or no training data is available. Alternatively, one could introduce such training data, but that would unnecessarily complicate the construction of training data sets and this inner cutoff mechanism is sufficient. To ensure short range repulsion we augment the large multi-body ACE basis by a small auxiliary basis set, consisting only of low-polynomial-degree pair interaction (two-body) functions. The construction is exactly the same as before, but we change the cut-off function to ## Basis Selection Before we can parametrize the ACE force field we need to select a specific finite basis set chosen from the complete ACE basis constructed in the previous section. There are three approximation parameters: the cutoff radius (r cut = r out ), the maximum correlation order ν max , and the maximum polynomial degrees D max ν corresponding to order ν basis functions. We have already specified the cut-off radius in the definition of the radial basis in eq (12). The basis is then chosen as (a linearly independent subset of) all possible basis functions B iv with correlation order at most ν max and polynomial degree at most D max ν . In all models for molecules with three or fewer distinct elements we take ν max = 4, which corresponds to a general 5-body potential. In models for molecules with four or more distinct elements we reduce this to ν max = 3 (4-body potential). The weight w Y specifies the rela-tive importance of the radial and angular basis components; here we choose w Y = 2. The maximum polynomial degrees D max ν can be adjusted to balance the size of the basis set against fit accuracy and evaluation time; the precise parameters we choose for each molecule are given in Table S1. The basis truncation we specified here is just one, rather simple, way to obtain a finite basis. There may very well be more sophisticated methods to choose an optimal subset of the complete basis. ## Parametrization of the linear ACE potentials We define the total energy of a linear ACE model with parameters c corresponding to a spatial configuration of atoms (denoted by X, e.g. a molecule in a particular configuration) as the sum of the site energies where E i is a site energy defined in eq (10). Optimal parameters are obtained by minimizing the loss function where the E QM and F QM are energies and forces, respectively, in the training data, obtained from electronic structure calculations. The sum is taken over all configurations in the training set, and w E X , w F X are weights specifying the relative importance of energies and forces. Since the model energy and force are both linear in the free parameters, the loss can be written in a linear least squares form, where the vector t contains the QM energy and force observations, and the design matrix Ψ contains the values and gradients of the basis evaluated at the training geometries. Ψ has a number of rows equal to the total number of observations (energies and force components) in the training set, and a number of columns equal to the total number of basis functions. The least squares problem has to be regularized, especially when the basis contains high degree polynomials. 33 One option is to apply Tychonov regularization, where the loss function is modified as This is widely used to regularize linear regression, often by taking Γ as just the identity matrix, or alternatively in the case of kernel ridge regression (and Gaussian process regression) as the square root of the kernel matrix. 43 In the present case, we use a diagonal Γ with entries corresponding to a rough estimate for the p-th derivative of the basis functions, where n t and l t are part of the elements of the multi-index vector v (cf. eq (6)). This scales down high degree basis functions, encouraging a smooth potential, which is crucial for extrapolation, and is loosely analogous to the smooth Gaussian prior of GPR. The actual solutions are then found using the standard iterative LSQR solver, 44 for the details see the SI. In the other approach we used for solving the least squares problem the same Γ matrix is introduced, but without a Tychonov term, and the solution is found using the rank revealing QR factorisation 45 (RRQR), in which we perform a QR factorization of the scaled design matrix ΨΓ −1 , and truncate the small singular values below some tolerance parameter λ. For more details of the exact implementation see Refs. 26,45. We found that when the linear system is not underdetermined, RRQR gave somewhat better solutions than LSQR. All parameters of the optimization (w E X , w F X , p, λ) are given in the SI. The last modelling choice that needs to be made is the 1-body term, that is the energies of the isolated atoms of each element in our model. One can use the energy of the isolated atoms evaluated with the reference electronic structure method, which ensures the correct behavior of the model in the dissociation limit. In other words, that the force field is modelling the binding energy of the atoms. An alternative approach, often used in the ML fitting of molecular energies, is to take the average energy of the training set, divided by the number of atoms in the molecule, and assign the result to each element. In this case, the fitted model has zero mean energy. This usually improves the fit accuracy slightly, by reducing the variance of the function that we need to fit, in case the data spans a narrow energy range around its average, e.g. because it came from samples of moderate temperature molecular dynamics. A third option is to not use any reference potential energies for the fit, but only forces. Once the coefficients are determined, the potential can be shifted by a constant energy chosen to minimize the training set energy error. In the current work, we evaluated all three strategies for ACE and found that using the isolated atom energies for the 1-body term gives slightly higher RMS errors, but leads to far superior extrapolation. The other two strategies (using the average energy for the 1-body term, and fitting only to forces) result in similar somewhat lower test set errors, but inferior physical extrapolation properties. As mentioned in the introduction, we view tests on data sets such as MD17 and ISO17 as proxies: the models thus created are not useful for any scientific purpose. The promise of ML force fields is greatest when the intention is to describe a very wide variety of compounds and conformations, perhaps including chemical reactions. With this in mind, the most natural choice for the 1-body term is to choose it to match the energy of the isolated atom in vacuum. This choice is independent of any particular data set, and the apparent advantages of the other choices in terms of lower errors are expected to diminish in the limit of a large and wide ranging data set. ## MD17 The original MD17 benchmark data set consists of configurations of 10 small organic molecules in vacuum sampled from density functional theory (DFT) molecular dynamics simulations at 500 K. 38 It has recently been recognized, that some of the calculations in the original data set did not properly converge, in particular, many of the forces are noisy. A subset of the full data set was recomputed with very tight SCF convergence settings and is called the rMD17 (revised MD17) data set. 46 We have used this new version of the data set and the five train-test splits as reported in Ref. 46. These revised training sets consist of 1,000 configurations to avoid the problem of correlated training and test sets: when more than 1,000 configurations are used from the full published trajectory, some of the test set configurations will necessarily fall between two neighboring training set data points that are separated by a much smaller time difference than the decorrelation time of the trajectory, resulting in an underestimation of the generalization error. 46 Table 2 shows the Mean Absolute Error (MAE) of the different force field models trained on 1,000 configurations. For comparison, in Tables 2 and 3, we include a wide selection of models from the various classes of force field fitting approaches that we discussed in the Introduction (Table 1). They include ML approaches such as feed forward neural networks (ANI, GMsNN, GMsNN), Gaussian Process regression models (sGDML, FCHL, GAP) and graph neural network based models (DimeNet, Schnet, Physnet, Cormorant and PaiNN). The models on the left of Table 2 were trained by us, (except for FCHL) using the exact traintest splits of rMD17, whereas the models on the right of the solid vertical line are from the literature and were trained on the original MD17 data set using different train-test splits. The precise details of the fitting procedures and parameters for each of the models can be found in the SI. Of the descriptor based models, sGDML, 17 This was crucial for achiev-ing the errors shown. When the weights were initialized randomly, the errors are higher by factor of 2 (Table S2). The GAP model, using SOAP features to describe the atomic geometry (which are similar to ANI's features), achieves similar errors to the ANI model with pre-training. The fact that ANI is only competitive with GAP if it is pre-trained can be rationalized by the relative sample efficiency of kernel models compared to neural networks. The FCHL kernel models also use 2-and 3-body correlations as features, but they have been more carefully optimized for molecular systems and hence are able to achieve very low errors. 15 The classical force field (FF) refers to a reparametrization of the GAFF functional form 2,47 using the ForceBalance program 6,47 and the rMD17 training set. This model gives at least an order of magnitude higher errors compared to the ML force fields. This is not a huge surprise, but is nevertheless a quantitative characterization of the limitations of the fixed functional form for a situation in which the empirical force fields are designed to do well. For completeness, in Table 3 we show the MAEs of the neural network models reported in the literature that were trained on 50,000 structures from the original MD17 trajectories. The test set errors of these models are probably underestimating the true generalization error, because the large training set contains configurations that are correlated with the test set, as discussed above. 46 It is still interesting to note that the Cormorant equivariant neural network 22 achieves very low energy errors compared to PaiNN, even though it was trained on energy labels only, but the force errors for this model were not reported. On the other hand, the PhysNet 24 graph neural network achieves remarkably low force errors compared to the other models. But similarly to the other equivariant graph neural network models, this comes at the expense of having close to 3 times larger energy errors compared to ACE and FCHL. ## Learning curves The first property to consider beyond the raw energy and force errors is the learning curve, showing how a model's performance improves with additional training data. For kernel models such as FCHL and sGDML, the "kernel basis" grows precisely together with the training data, which is why these methods are universal approximators. Subject to the radial cutoff, the infinite set of Atomic Cluster Expansion basis functions forms a complete basis for invariant functions, so in principle they can also be used to approximate the potential energy surface to arbitrary accuracy. 35 In this case however, the size of the training set and the size of the basis are decoupled. One advantage is that the evaluation cost is independent of training set size, but we have to choose a finite basis set to work with by selecting a maximum body order and the truncation of the 1-particle basis. In order to motivate our choice, we show in fig 2 the force accuracy of ACE as a function of basis set size and the corresponding evaluation time, trained on 1,000 azobenzene configurations (the largest molecule in MD17). The timings were obtained using a 2.3 GHz Intel Xeon Gold 5218 CPU. For context, we show the accuracy and evaluation time of the other ML models we trained, each called in their native environment: ACE in julia, GAP via the fortran executable, and sGDML and ANI directly from their respective Python packages. (Note that in the case of ANI considerable speed up could be achieved using a GPU when multiple molecules are evaluated simultaneously, see fig S1, though our single molecule results are in agreement with the timings reported in the original ANI paper 17 ). The solid part of the ACE curve corresponds to 4-body potentials (ν = 3) and we varied only the polynomial degrees, whereas for the last point (dashed), we increased the body order to 5, because the 4-body part of the curve showed saturating accuracy. Increasing the body order further is likely to bring the error down even more, however, the cost of evaluation would also grow unacceptably if all basis functions for the given body and polynomial degree are retained. In the future, effective sparsification strategies need to be developed that would allow the inclusion of some high body order basis functions without the concomitant very large increase of the overall basis set size. For the purposes of the present paper, for each molecule in MD17 we selected a basis set size such that the evaluation cost was roughly comparable with the other ML models. (Note however that in a real ML force field application, one might very well choose a much smaller basis, e.g. 10K, to take advantage of the sub-millisecond evaluation times.) In fig 3 we show the learning curves for linear ACE and sGDML (the best models we trained from Table 2) and compare to the literature results of FCHL. 46 The low body order linear ACE is equal or better than the other manybody kernel models in the low data limit, but with additional training data the kernel models overtake ACE in several cases. The latter also saturates, showing the limitations of the relatively low body order model. The learning curves for the forces are given in fig S2, and show a broadly similar trend, with less pronounced saturation for ACE. In the case of ACE the number of basis functions is shown in parentheses. The classical force field has a timing of about 1 µs, which would not fit on this scale. For the ANI model we show both the CPU and GPU timings. ## Normal mode analysis The normal modes and their corresponding vibrational frequencies characterize the potential energy surface near equilibrium. This is interesting in the context of the MD17 models because their training set contains geometries sampled at 500 K which means they are, in general, far from the equilibrium geometry. The ability of the models to describe the minima of the PES, even if it is not in the training set, is particularly important when considering larger systems with potentially many local minima, where finding all the different local minima at the target level of theory can be infeasible. To test how well the different models infer the normal modes we took the DFT optimized geometry of each of the 10 molecules and rerelaxed them with the force field models. At the force field minima we carried out a vibrational analysis to find the normal modes and their corresponding vibrational frequencies. Fig 4 shows the errors in the predicted normal mode vibrational frequencies for each of the 10 MD17 molecules. The ACE model achieves the lowest error for all 10 molecules, surprisingly even for those for which sGDML has lower errors based on the 500 K MD test set of Table 2. For example, for toluene sGDML has both lower energy and force errors, but at the same time the ACE model has significantly lower errors in predicting the vibrational frequencies, achieving a MAE of 1.0 cm -1 compared to sGDML with an error of 1.4 cm -1 . Observing the individual molecules in Fig 4 it is notable that the ACE model has the lowest fluctuation in the errors of the normal modes, achieving nearly uniform accuracy across the entire spectrum. The case of benzene also shows the limitations of characterizing the models by the force MAE alone. The linear ACE model has only slightly lower force MAE than sGDML (0.5 meV/ compared to 0.8 meV/ ) but the normal mode frequency prediction is more than 3 times more accurate: 0.2 cm -1 compared to 0.7 cm -1 . The linear ACE model has very low errors for all normal modes, whereas sGDML has much higher errors for the high frequency modes. Similarly, in the case of aspirin, even though the ANI model has lower MAE on the test set both for energies and forces than the GAP model, its vibrational frequency error is significantly larger than those of GAP (8.3 cm -1 compared to 6.4 cm -1 ). We also compared the models to the accuracy of a classical force field. The normal mode frequency errors of the empirical FF are about 10 times higher than the errors of the ML force fields. These errors do not fit on the scale of ## Extrapolation in temperature When building a new force field for a molecule, beyond high accuracy, we also need robustness, by which we mean that there should not be areas of accessible configuration space where the model predictions are unphysical or nonsensical. Sometimes called "holes" in the potential energy surface, these can be remedied by regularization 33 or by iterative fitting 39 and additional data. 48 In the context of the MD17 benchmark, with its fixed training set, we test the robustness of the models we fitted by run- Where a point is missing, the model hit a hole in the potential and the MD run was terminated. This happened most often with the GAP model, indicating that this potential was the least regular. The linear ACE and ANI models can also be prone to hitting holes in the potential at the highest temperatures. Of all models sGDML was the most stable, it al-ways kept the molecule intact even at 950 K for the duration of the simulations. Such extreme stability is not necessarily chemically realistic (see the next section on extrapolation to bond breaking). Looking at the increase in errors with temperature for the different models we can see that the linear ACE often keeps the errors low with a small slope whereas the other models show a clearer increase as the temperature increases. This can be best observed for ethanol, malonaldehyde and uracil. It is notable that the model that works best at lower temperatures (in the training regime) also works best at higher temperatures confirming that the models are able to smoothly extrapolate away from the training data. Furthermore, we can see a good agreement of the test set force MAE in Table 2 with the force MAEs estimated from the models' own trajectories. This hints that the models explore similar regions of the configuration space as the original ab initio trajectories. ## Extrapolation far from the training set To test the extrapolation properties of the different models further we looked at two tests probing the torsional profile of azobenzene We carried out these tests with several different versions of the linear ACE models differing in the definition of their 1-body terms, because we expect this choice to make a significant difference in how chemically reasonable the fitted models are far from the training set. We denote the ACE models fitted using force data only by ACE F. This has the lowest force error on the test set (comparison shown in Table S3). For the other two ACE models, energies were also included in the training. They differ in the 1-body term only, the model using average per-atom training set energy is denoted as ACE AVG, whereas the model using the isolated atom energies as the 1-body term is de-noted ACE E0. The third option is the natural choice, as this ensures that if all atoms are separated from each other the predicted energy will correctly correspond to the sum of the isolated atom energies. Fig 6(a) shows the torsional energy profile of the azobenzene molecule. The ACE E0 model with the isolated atom 1-body term is able to extrapolate furthest, somewhat overestimating the energy, while the ANI and sGDML models also extrapolate smoothly, but slightly underestimate the energy. The linear ACE model with the average energy 1-body term and the GAP model fail to extrapolate and predict a completely nonphysical drop in energy for smaller values of the dihedral angle. 2. geometry of ethanol. The only force field that shows qualitative agreement with DFT is the ACE E0 model. (Note that we do not expect any of the fitted models to quantitatively reproduce the DFT energy profile, even when the isolated H atom is described correctly by design, because the C 2 H 5 O • radical is not.) We attribute this success to the explicit body ordered nature of the linear ACE model, including using the isolated atom as the 1-body term, and careful regularization -as was the case in a similar test for other polynomial models. 26 Fig 6 (c) shows a detailed comparison of the different ACE models together with their test set MAE value. This shows that having the lowest possible test set error does not coincide with the most physically reasonable model, and using stronger regularization can lead to much smoother extrapolation. The more strongly regularized ACE models with relatively higher force error are still significantly more accurate than sGDML, ANI, GAP or the classical force field. Interestingly, having the isolated atom as the 1-body term is not sufficient for good extrap-olation. This is shown by the two different GAP models in fig 6(b), which show essentially no difference to the extrapolation, presumably due to the very poor description of the radical. GAP is not an explicitly body ordered model. ## Fitting multiple molecules Apart from sGDML, whose descriptor is tied to a given molecule with fixed topology, the models under consideration can all be fitted to multiple molecules simultaneously. Therefore, having evaluated their capacity to approximate individual potential energy surfaces one by one, it is interesting to see how the they cope with describing all of the rMD17 data set pooled together. Table 4 shows the energy and force errors for the combined fit with linear ACE, GAP, ANI and the empirical force field. GAP and ANI errors only go up by around 30%, reflecting the fact that these are very flexible functional forms. The ANI model (which is pre-trained by starting from ANI-2x neural network weights) is now distinctly better than GAP. The empirical force field error increases by even less. In this case that is due to the use of atom-types, which help to separate the energy contribution of different functional groups. The increase in the error is largest for ACE, about a factor of two, although for most molecules it is still the combined ACE model that has the lowest error amongst these models. In addition, we also show the performance of the original unmodified ANI-2x model (its energies and forces were tested against values recomputed with exactly the same electronic structure method and parameters that were used in its fitting 49 ). Its energies and forces are better than those of the empirical force fields by factors of around 2-3 and 5, respectively. (The exception is azobenene, for which its energies are worse). The difference between ANI-2x and the re-trained ANI is about a factor of 2-4 for energies (the average over all the molecules is at the high end) and a factor of two for forces. The other commonly used benchmark data set for machine learning based molecular force fields that contains multiple molecules is ISO17. 23 The full data set contains 5000-step ab initio molecular dynamics simulation trajectories of 129 molecules, all with the same chemical formula C 7 H 10 O 2 . The standard task is to train a force field using a randomly selected 4000 configurations of 103 molecules (so about 400K configurations altogether, although these are highly correlated) and evaluate it on the remaining 1000 structures of the trajectory ("known molecules") and on the full trajectories of the "unknown molecules". We note that when all 400K training configurations are used, the conformations of "known molecules" that are usually reported as a test set are very close to the training set, at most 1 or 2 MD steps away on the trajectory from the actual training set, so the error measured on these is essentially the same as the training error. We trained a linear ACE model on only a total of 5,000 configurations and a GAP model on only a total of 10,000 configurations sampled uniformly from the training set and evaluated them on both the known and unknown molecules. The results in Table 5 show that the linear ACE model performs significantly better than GAP, achieving errors in the same ballpark as the other methods for the unknown molecules, but using orders of magnitudes less training data. In particular, the ACE model matches the energy error of the state of the art GM-sNN 20 on the unknown molecules, demonstrating its excellent extrapolation capabilities. For all the neural network models, the error on known molecules is quite a bit lower than that for the unknown molecules, which we consider to be a sign of overfitting. For ACE and GAP, the error is still lower but by a much smaller factor, helped by the explicit regularization. Tellingly, the most similar ratio is for GM-sNN, which is a shallow neural network. ## Flexible molecule test: 3BPA Finally, noting that all the MD17 molecules are rather rigid, our last test is to assess the capabilities of the different force field models on a more challenging system that has relevance for medicinal chemistry applications. We created a new benchmark data set for the flexible drug-like molecule 3-(benzyloxy)pyridin-2amine (3BPA). 50 Though smaller than typical drug-like molecules, with a molecular weight of 200, this molecule has three consecutive rotatable bonds, as shown in fig 7. This leads to a complex dihedral potential energy surface with many local minima, which can be challenging to approximate using classical or ML force fields. 51 ## Preparation of the data set To prepare a suitable training data set we started by creating a grid of the three dihedral angles (α, β and γ) removing only the configurations with atom overlap. From each of the configurations corresponding to the grid points, we started short (0.5 ps) MD simulations using the ANI-1x force field. 28 This time scale is sufficient to perturb the structures towards lower potential energies, but is not enough to significantly equilibrate them. In this way we obtained a set of 7000 configurations as shown in the left panel of Fig 7 . From the distribution of dihedral angles, five different densely populated pockets were identified in the space of the three dihedral angles. One random configuration was selected from each of the 5 pockets and a long 25 ps MD simulation was performed at three different temperatures (300 K, 600 K, 1200 K) using the Langevin thermostat and 1 fs timestep. We sampled 460 configurations from each of the trajectories starting after a delay of 2 ps. In this way the final data set of 2300 configurations was obtained. The configurations were re-evaluated using ORCA 52 at the DFT level of theory using the ωB97X exchange correlation functional 53 and the 6-31G(d) basis set. (These settings are similar to that used in the creation of the ANI-1x data set 49 ). From the total data set we created two training sets, one using 500 randomly selected geometries from the 300 K set, and another one, labelled "mixed-T", selecting 133 random configurations from each of the trajectories at the three temperatures. The rest of the data in each case makes up the three test sets, each corresponding to a different temperature. The right hand panels of Fig 7 show the distribution of dihedral angles in the test sets. At 300 K the separate pockets of the configuration space are sampled mostly individually, whereas at 1200 K the distribution widens significantly, and the sampling connects the pockets across multiple barriers with ease. ## Comparison of force fields models We trained linear ACE, sGDML, ANI and GAP force fields, and re-parametrized the bonded terms of a classical force field (FF), using the 300 K and the mixed-T training sets. Table 6 shows the energy and force RMSEs of the different models alongside the general purpose ANI-2x force field errors on the same configurations. Just as before, the weights of the re-trained ANI model were initialized form the ANI-2x weights, giving it a considerable advantage over the other models, especially because the DFT functional and basis set that we use are the same as that of the underlying DFT method of the ANI-2x model. For the case of training on the 300 K configurations the linear ACE and sGDML models are able to achieve very low errors when tested at the same temperature, but the ACE model shows significantly better extrapolation properties to the configurations sampled at higher temperatures. The model extrapolating most accurately to 1200 K is the re-trained ANI force field, but the linear ACE is not far behind, especially considering how poor the extrapolation of the other models are. Just as for the smaller molecules, the fitted empirical force field shows much higher errors, about a factor of 2-4 for energies and a factor of 4 for forces compared with the ANI-2x force field. Only at 1200 K does ANI-2x become competitive with the ACE trained at 300 K. Training on the mixed-T training set leads to a significant drop in the errors at the higher temperature test sets for all ML models, but not for the empirical force field. The linear ACE model achieves the lowest error in every case, showing approximately 40% decrease in the error for the high temperature test set. The other ML models improve also, by even bigger factors (because their extrapolation power was less). The gains over the general ANI-2x force field, nearly a factor of two in energies for all three test sets, show the potential scope for parametrizing such custom force fields in medicinal chemistry applications. The errors in the empirical force field are mostly unchanged, quantifying the limitations of the fixed functional form when describing the anharmonic high energy parts of the potential energy surface. To look beyond the energy and force RMSE, we performed a constrained geometry optimization using the different force field models and DFT to map out the dihedral potential energy surface of the molecule. The complex energy landscape is visualized in S8. The energy landscape of the empirical force field has most of the features of the DFT landscape and is even correctly predicting the position of the lowest energy minimum in the β = 120 • plane. Some of the potential energies on this plane are clearly too high however. On the other hand the landscape of the GAP model is quite irregular, some of the most basic features are either missing or blurred together. The ANI landscape is also quite irregular, somewhat less than GAP, and some of the high energy peaks are too high and too broad. This is an example where the fixed functional form of the classical force field gives better extrapolation behavior to parts of the configuration space where there is little training data. The RMSE results clearly do not give a full, perhaps not even a very useful distinction between these models. The ACE and sGDML models reproduce the landscape much more closely (and indeed these are the models with the lowest RMSE as well). Some differences include the sGDML getting the position of the lowest energy minimum wrong and ACE having too high a peak at α = 230 • , γ = 150 • . ## Conclusions In this paper we have demonstrated how the Atomic Cluster Expansion framework can be used as linear models of molecular force fields. We showed that body ordered linear models built using the ACE basis are competitive with the state of the art short range ML models on a variety of standard tests. Furthermore we carried out a number of "beyond RMSE" tests to compare the ML approaches, and to study the smoothness and extrapolation properties of the fitted force fields: vibrational frequencies, forcefield driven molecular dynamics and extrapolation to bond-breaking. We also introduced a data set on a flexible drug-like molecule, with the idea that testing the performance on it is more predictive of the quality of the model for medicinal chemistry applications. The linear ACE model was signifi-cantly smoother than other transferable models and was able to extrapolate to higher potential energy regions than all other models. We showed that the ACE framework allows us to build accurate force fields with very low evaluation cost. Together with competing approaches that are in the recent literature and in our comparison tables, the prospects are good for being able to carry out large scale simulations of systems ranging from biomolecular applications to other complex molecular systems such as polymers with electronic structure accuracy in the near future. A number of bottlenecks remain for ACE, which include the steep increase in the number of basis functions as new chemical elements are added to the model. This can be tackled via sparsification strategies, which is the focus of our future work. Furthermore the inclusion of long range electrostatics and charge transfer are essential for the simulation of biomolecular systems and an integration of these into the ACE framework is also underway. Currently ACE is implemented in the Julia language, but can readily be called from Python via the Atomic Simulation Environment (ASE). The fitted models can also be evaluated via LAMMPS. models, a comparison of different 1-body energy ACE models and energy landscapes for 3BPA with β = 150 • and β = 180 • are shown in the Supporting Information. The 3BPA dataset is available as a .zip file free of charge via the Internet at http://pubs.acs.org

chemsum

{"title": "Linear Atomic Cluster Expansion Force Fields for Organic Molecules: beyond RMSE", "journal": "ChemRxiv"}

pd/xiang-phos-catalyzed_enantioselective_intermolecular_carboheterofunctionalization_under_mild_cond

## Abstract: A mild and practical Pd/Xiang-Phos-catalyzed enantioselective intermolecular carboheterofunctionalization reaction of 2,3-dihydrofurans is developed, leading to various optically active fused furoindolines and tetrahydrofurobenzofurans. The key to this transformation is employing two newly modified N-Me-Xiang-Phos ligands ((S, R S )-N-Me-X4/X5) as chiral ligands under mild conditions. Moreover, this synthetic methodology can be efficiently applied to a variety of complex polysubstituted heterocycles with high chemo-, regio-, and enantio-selectivities via introducing diverse substituents on furan rings, which were hard to access by other routes. ## Introduction Benzofused heterocycles are ubiquitous moieties in natural products, pharmaceuticals, dyes and herbicides, in which furoindolines and tetrahydrofurobenzofurans are prevalent as key core structures (Fig. 1). 1 These derivatives have shown signifcant anticancer, antimalarial and antimicrobial activities, as well as antioxidant properties, for instance, Makomotindoline and Aspidophylline A have shown a distinct effect on mammalian cells. 2,3 The frst enantioselective total synthesis of Aspidophylline A was described by Garg via a reductive interrupted Fischer indolization. 4 You and co-workers developed a copper-catalyzed intermolecular dearomative cascade reaction of indoles, which also provided a powerful synthetic method for the construction of furoindolines. 5 E. J. Corey achieved a short, asymmetric total synthesis of Aflatoxin B 2 via an aromative cascade reaction. 6 Despite these seminal reports, hetero-annulation of alkenes developed by Catellani and Larock has become a classic and useful strategy for the construction of various heterocycles from readily available starting materials. 7 Although various methods have been developed to construct these two skeletons, it still remains a considerable challenge to extend the substrate scope of asymmetric variants, particularly those that enable access to poly-substituted benzofused heterocycles. Over the past two decades, palladium-catalyzed carboheterofunctionalization of alkenes has been proved to be a reliable and efficient method for the synthesis of a variety of poly-cyclic heterocycles. 8 The majority of these reactions proceeded through a crucial hetero-palladation of alkenes with aryl halides along with N-or O-nucleophiles. 9,10 However, the development of an enantioselective version, especially under mild conditions, poses a considerable challenge due to the lack of any suitable robust chiral catalyst. Recently, Mazet and coworkers reported the frst asymmetric Pd-catalyzed syn-carboe-therifcation and syn-carboamination of 2,3-dihydrofurans (2,3dhfs) at 110 C by utilizing two different chiral ligands (Scheme 1a). 11 Inspired by the good performance of our chiral sulfnamide phosphine (Sadphos) ligands in the asymmetric construction of C-C and C-X bonds, 12 we wondered whether Sadphos could realize the highly enantioselective carboether-ifcation and carboamination of 2,3-dhfs under mild conditions and also address the low enantioselectivity issue of the carboamination reaction. Herein, we report a highly chemo-, regio-, and enantioselective palladium-catalyzed carboheterofunctionalization of 2,3-dhfs employing two newly modifed Xiang-Phos ligands as chiral ligands, which can give direct access to enantioenriched poly-substituted functionalized furoindolines and tetrahydrofurobenzofurans in moderate to high yields with high enantio-selectivities at a reduced reaction temperature (Scheme 1b). ## Results and discussion With the use of our developed chiral sulfnamide phosphine ligands as the chiral ligands, the carboamination reaction of 2-bromoaniline derivative 1a and 2,3-dhf 2a was investigated. It was found that Ming-Phos M1, PC-Phos PC1, Xu-Phos Xu1 and Xiang-Phos X1 did not efficiently deliver the desired product. As observed in our previous work, the N-H bond in ligands could greatly affect the reactivity as well as enantioselectivity in some cases. 16 Several representative N-Me sulfnamide phosphine ligands lacking the hydrogen-bonding site were further investigated. We were pleased to fnd that the desired product 3aa could be obtained in 81% yield with a 48% ee value in the presence of (S, R S )-N-Me-X1 and CH 3 ONa, albeit with a small amount of the Heck byproduct 4aa. Other chiral ligands such as Ming-Phos N-Me-M1 and Xu-Phos N-Me-Xu2 showed less efficiency comparatively, leading to a lower yield and enantioselectivity along with a poor regioselectivity (Fig. 2). Under the conditions of Xiang-Phos (S, R S )-N-Me-X1 utilized as the chiral ligand, NaOPh appeared to be the optimal base, affording 3aa in 78% ee albeit with a 2 : 1 regioselectivity ratio (r.r.) (Table 1, entries 1-5). Solvent screening showed that 1,2-DCE gave a better yield (81%) and r.r. (9 : 1) with 87% ee (Table 1, entries 6-9). The result obtained employing other chiral N-Me-Xiang-Phos ligands indicated that the introduction of steric hindrance on the phenyl backbone and enhancement of the electron-donating character were benefcial for the catalytic enantio-selectivity and regioselectivity (Table 1, entries 10-14). Employing the newly modifed N-Me-X5 as the chiral ligand, a series of Pd precursors were then screened, showing that a fvemembered cyclic palladium precatalyst was competent for the carboamination cyclization (Table 1, entries 15-19). Comparably better outcomes were obtained under mild conditions by lowering the temperature to 20 C (Table 1, entries 20-23). Inspired by previous fndings that the addition of a trace amount of water may help to increase the reactivity and the stereoselectivity, 17 the water effect was studied, and indeed, we found that the addition of 2 equivalents of water to the system led to a signifcantly improved reactivity and reproducible enantioselectivity. In terms of the reactivity and enantio-and regio-selectivity, the reaction conditions illustrated in entry 22 were utilized in the following substrate scope investigations (please see Table S1 in the ESI for details †). Various substituted N-(2-bromophenyl)-p-tolylsulfonamide derivatives 1a-r were subsequently employed as coupling partners in the enantioselective intermolecular carboamination of 2,3-dhf (Scheme 2). Remarkably, a wide range of 2-Br-anilines bearing electronically diverse substituents at C4 and C5 such as halogens, -Me, -OMe, -CF 3 , -OCF 3 , and -CO 2 Me reacted smoothly and furnished the corresponding furoindolines 3aa-3la in good yields (up to 97%) and ee's (up to 96%). Further substrate scope investigations demonstrated that the electronic properties of substituents at C3 and C6 did affect the yields and enantioselectivities, and produced products 3na and 3oa in lower yields comparatively. Notably, the disubstituents on phenyl rings were also applicable in this cyclization reaction, affording 3pa in 84% yield with 89% ee, as well as 3qa containing a heterocycle in 87% yield and 95% ee. When N-(2bromophenyl)-benzenesulfonamide was explored as an alternative to 1a, to our delight, the furoindoline product 3ra was formed in a nearly quantitative yield (92%) with high enantioselectivity (95%). We also replaced the protective groups on the nitrogen atom with Ms and Ns, but only trace products could be detected by NMR. Other nitrogen protecting groups on aniline, such as Boc, Cbz and Bz, were not tolerated, and in these cases the desired product was not observed. To our delight, a gram-scale reaction was conducted to further demonstrate the potential synthetic utility of this methodology, delivering 1.2 g of 3aa in 77% yield and 94% ee with 2.5 mol% palladium catalyst at 20 C for 6 days. The absolute confguration of this series of products was confrmed by the X-ray diffraction analysis of 3aa. 18 After a quick survey of the construction of tetrahydrofurobenzofuran (Table 2, please see Table S2 in the ESI for details †), the optimal reaction conditions were identifed (Table 2, entry 7). A series of substituted 2-bromophenol derivatives 5a-h were subsequently employed as coupling partners in the enantio-selective intermolecular carboetherifcation of 2,3-dhf with the use of N-Me-X4 under mild conditions (Scheme 3). Of note, an exciting enantioselective cyclization was realized when substrates containing diverse substituents at C4, C5 and C6 with different electronic properties (-F, -Me, and -OMe) participated smoothly, delivering products in 90% to 99% ee's. It is a pity that only a trace amount of product was observed when substituents were present at C3 of the phenyl ring. To delve into the construction of poly-substituted fused furoindolines and tetrahydrofurobenzofurans, a variety of dihydrofuran derivatives 2b-2e, which could be readily prepared by a classic Heck reaction, were subjected to carbofunctionalization. In these two cases, variation of the electronic parameters of the phenyl groups on dhf rings had a slight influence on the yields and enantiocontrol of the carbohetero-functionalizations (Scheme 4, 3ab/6ab, 3ac/6ac). Considering the O-and Ncontaining heterocycle substituents on dhf rings, the phenyl groups could be swapped for the benzofuran and quinoline substituents (3ad/6ad, 6ae), still maintaining the efficiency of the transformations. 5-Methyl-2,3-dhf was next examined to investigate the formation of an all-carbon quaternary stereocenter. The carboetherifcation reaction took place smoothly a Unless otherwise specifed, all reactions were carried out with 1a (0.2 mmol), 2a (0.8 mmol, 4 eq.), a [Pd] source (0.01 mmol, 5 mol%), N-Me-Xiangphos (0.024 mmol, 12 mol%), base (0.8 mmol, 4 eq.), and H 2 O (7.2 mL, 2 eq.) in a solvent (1 mL, 0.2 M). when increasing the loading of the palladium precatalyst and chiral ligand at a higher temperature (6af). However, only the corresponding debromination product was detected in the carboamination reaction system. The absolute confguration of the poly-substituted carboamination products was confrmed by X-ray diffraction analysis of 3ac, 18 while the absolute confguration of poly-substituted carbo-etherifcation products was assigned by comparing the rotational value and 1 H, 1 H-NOESY-NMR spectrum (please see the ESI for details †) of 6ab between our work and Mazet's work. 11 The substituted aromatic ring on the tetrahydrofuran ring was in the (S)-confguration, and is a diastereomer of the corresponding product in Mazet's work. Based on Mazet's studies, as well as our observations on Pd/ Sadphos catalytic systems, the chirality-induction models of carbo-amination and -etherifcation were proposed according to the absolute confguration of products 3aa and 6aa, as shown in Scheme 5. We supposed that the reaction was initiated by a classic oxidative addition, which would be followed by ligand exchange, deprotonation and coordination of 2,3-dhf. The key step of asymmetric hetero-palladation was hypothesized to occur to ultimately construct optically active benzofused heterocycles with high regio-and enantio-selectivities. a Unless otherwise specifed, all reactions were carried out with 5a (0.2 mmol), 2a (1 mmol, 5 eq.), a [Pd] source (0.005 mmol, 2.5 mol%), N-Me-Xiang-Phos (0.01 mmol, 5 mol%), base (0.4 mmol, 2 eq.), and H 2 O (3.6 mL, 1 eq.) in a solvent (1 mL, 0.2 M). b Yield of isolated product. c Determined by chiral HPLC. d Pd 2 (dba) 3 was added to 5 mol%, and L3 was added to 10 mol%. Scheme 2 Scope of carboamination of 2,3-dhf with 2-bromoanilines. ## Conclusions In summary, we have demonstrated an efficient Pd-catalyzed enantioselective intermolecular carboheterofunctionalization of 2,3-dihydrofurans for the synthesis of poly-substituted benzofused heterocycles. The new N-Me-Xiang-Phos X4/X5 ligands are responsible for the high reactivity and enantioselectivity. This strategy could be conducted under mild conditions and easily extended to a wide range of chiral fused furoindolines and tetrahydrofurobenzofurans with high chemo-, regio-, and enantio-selectivities, which made the method extremely attractive. In addition, a gram-scale reaction of the representative product 3aa was investigated to further demonstrate the potential synthetic applications of this method. Further applications of Sadphos in other transition-metal-catalyzed reactions are underway in our group and will be reported in due course. ## Conflicts of interest There are no conflicts to declare.

chemsum

{"title": "Pd/Xiang-Phos-catalyzed enantioselective intermolecular carboheterofunctionalization under mild conditions", "journal": "Royal Society of Chemistry (RSC)"}

the_role_of_hypervalent_iodine(<scp>iii</scp>)_reagents_in_promoting_alkoxylation_of_unactivated_c(s

## Abstract: Although Pd(OAc) 2 -catalysed alkoxylation of the C(sp 3 )-H bonds mediated by hypervalent iodine(III) reagents (ArIX 2 ) has been developed by several prominent researchers, there is no clear mechanism yet for such crucial transformations. In this study, we shed light on this important issue with the aid of the density functional theory (DFT) calculations for alkoxylation of butyramide derivatives. We found that the previously proposed mechanism in the literature is not consistent with the experimental observations and thus cannot be operating. The calculations allowed us to discover an unprecedented mechanism composed of four main steps as follows: (i) activation of the C(sp 3 )-H bond, (ii) oxidative addition, (iii) reductive elimination and (iv) regeneration of the active catalyst. After completion of step (i) via the CMD mechanism, the oxidative addition commences with an X ligand transfer from the iodine(III) reagent (ArIX 2 ) to Pd(II) to form a square pyramidal complex in which an iodonium occupies the apical position.Interestingly, a simple isomerization of the resultant five-coordinate complex triggers the Pd(II) oxidation. Accordingly, the movement of the ligand trans to the Pd-C(sp 3 ) bond to the apical position promotes the electron transfer from Pd(II) to iodine(III), resulting in the reduction of iodine(III) concomitant with the ejection of the second X ligand as a free anion. The ensuing Pd(IV) complex then undergoes the C-O reductive elimination by nucleophilic attack of the solvent (alcohol) on the sp 3 carbon via an outersphere S N 2 mechanism assisted by the X À anion. Noteworthy, starting from the five coordinate complex, the oxidative addition and reductive elimination processes occur with a very low activation barrier (DG ‡ 0-6 kcal mol À1 ). The strong coordination of the alkoxylated product to the Pd(II) centre causes the regeneration of the active catalyst, i.e. step (iv), to be considerably endergonic, leading to subsequent catalytic cycles to proceed with a much higher activation barrier than the first cycle. We also found that although, in most cases, the alkoxylation reactions proceed via a Pd(II)-Pd(IV)-Pd(II) catalytic cycle, the other alternative in which the oxidation state of the Pd(II) centre remains unchanged during the catalysis could be operative, depending on the nature of the organic substrate.Scheme 1 Palladium-catalyzed C(sp 3 )-H alkoxylation using the cyclic iodine(III) reagent 2 (BI-OMe) developed by Rao et al. ## Introduction Hypervalent iodine compounds have been widely used in organic synthesis as oxidants and reaction promoters over recent decades. 1 Although these compounds themselves have the capacity to mediate many organic reactions, the addition of a catalyst is usually a prerequisite for certain transformations to occur. 2 Among these catalysts, palladium complexes have been demonstrated to have great potential to promote numerous iodine(III)-mediated processes including the C-H functionalization and coupling reactions. 3 In this context, Rao et al. developed a method for alkoxylation of unactivated methylene and methyl C(sp 3 )-H bonds to prepare alkyl ethers using hypervalent iodine(III) reagents in the presence of a Pd(II) catalyst. 4 According to the developed method, they showed that butyramide derivative 1 is alkoxylated in the presence of an alcohol as the solvent and the alkoxylation reagent, methoxybenziodoxole (BI-OMe, 2) as the oxidant, and Pd(OAc) 2 as the catalyst (Scheme 1). This strategy for alkoxylation might receive specifc attention due to its signifcance in modifcation of anti-inflammatory drugs such as ibuprofen and naproxen. It is worth mentioning that the importance of such a transformation has also prompted others to use similar methodologies for installing alkoxy groups onto other organic molecules using hypervalent iodine(III) reagents under Pd(II) catalysis even, in some cases, before that developed by Rao et al. 5,6 In an attempt to explore the reaction mechanism, Rao et al. conducted an isotope labeling experiment in deuterated methanol (Scheme 2). On that basis, the only observed product was 4, implying that the methoxy group installed on the product must originate from the solvent and not from the oxidant. Interestingly, when different alcohols were used as the solvent, only product 5 was practically observed and when the reaction was run in 1,2dichloroethane (DCE), the carboxylated product 6 was attained. All these fndings suggest that the methoxy group of the oxidant must act as a spectator ligand and thus is never involved in the reductive elimination step in the catalytic cycle (vide infra). Based on the preliminary results briefly discussed above, Rao et al. proposed the catalytic cycle outlined in Scheme 3. Accordingly, the reaction was surmised to commence with the coordination of substrate 1 to Pd(OAc) 2 , followed by OAc ligands-assisted N-H and C(sp 3 )-H activation processes via the concerted metallation-deprotonation (CMD) mechanism 7 to yield cyclopalladated intermediate 7. Subsequently, the oxidation of Pd(II) to Pd(IV) by the cyclic iodine(III) reagent 2 generates intermediate 8 from which a ligand exchange with the alcohol occurs to afford 9. The resultant intermediate fnally forms product 3 by undergoing the C-OR reductive elimination followed by a substitution reaction. It is worth noting that this plausible mechanism has also been summarized in several recent reviews, 3e,8 and proposed by several other researchers for interpreting alkoxylation reactions mediated by iodine(III) and catalyzed by palladium(II). 4,5,6e,f,9a,b,f,h-j,m,n The above-simplifed description prompted us to investigate the detailed mechanism of the title reaction by applying density functional theory (DFT) with the aim of addressing the following questions: (i) how the Pd(II) is oxidized to the Pd(IV) by BI-OMe? (ii) Are 8 and 9 key intermediates on the catalytic cycle? If so, why does not the reductive elimination occur from 9 to give 11? If not, what are the key intermediates? Why is the carboxylate group, and not the methoxy, installed on the organic molecule when the solvent is not the alcohol (Scheme 2)? Is the formation of a Pd(IV) intermediate inevitable in the catalytic process? Does the C-O reductive elimination take place via an inner-sphere mechanism? Why does the reaction require a high temperature to occur? Through answering these questions, we hope to enhance understanding of the fundamental processes involved in numerous Pd(II)-catalyzed I(III)-mediated alkoxylation reactions. 4,5,6e-g,9 Results and discussion ## CMD mechanism As described in the Introduction, the reaction is proposed to be initiated by activation of the N-H and C-H bonds of substrate 1 by Pd(OAc) 2 via the CMD mechanism. In analogy with the previous DFT studies, 10 trimeric Pd 3 (OAc) 6 is considered as the precatalyst for the palladium acetate; the trimeric form is computed to be 15.6 kcal mol 1 more stable than the monomeric Pd(OAc) 2 (Fig. 1a). The breakdown of Pd 3 (OAc) 6 into square planar complex 13 is computed to be exergonic by about 1.8 kcal mol 1 (Fig. 1b). The N-H deprotonation of the coordinated substrate by one of the acetate ligands gives complex 14 in which acetic acid occupies a vacant coordination site of Pd(II). This process is predicted to be extremely fast with DG ‡ ¼ 5.0 kcal mol 1 . The acetic acid in this complex (14) binds weakly to the metal center and thus it can be easily replaced to give complex 15 with a Pd-H-C agostic interaction (Fig. 1c). The resultant complex then involves a second deprotonation process by the other acetate ligand to yield 16. We found that this key step is nearly thermoneutral and proceeds with an overall activation barrier of 19.7 kcal mol 1 . The replacement of the weakly bonded acetic acid in 16 by BI-OMe gives 17 as an active species on the catalytic cycle (vide infra). process (Scheme 3). On that basis, the reductive elimination from 9 is expected to produce both products 3 and 11, which disagrees with the experimental fndings. Similar key intermediates were also suggested by some other researchers for analogous alkoxylation reactions catalyzed by Pd(II) complexes. 4,5,6e,f,9a,b,f,h-j,m,n To confrm the proposed mechanism is inoperative, we calculated the energy of intermediates 8 and 9a, and then evaluated the C-O reductive elimination from these two intermediates (Fig. 2). Several trends are apparent from this DFT investigation. First, based on the proposed mechanism (Scheme 3), the ligand exchange between 8 and the alcohol (solvent) is a prerequisite for the reductive elimination to produce the desired product. Our calculations show that the ligand exchange process is endergonic by 6.4 kcal mol 1 where the alcohol is MeOH, implying that the Pd center prefers to remain coordinated with the carboxylate ligand. Second, transition structures TS 1 and TS 2 are signifcantly lower in energy than TS 3 . It follows that the reductive elimination should preferentially occur from 8 and not 9a. Third, transition structure TS 2 is by 4.5 kcal mol 1 lower in energy than TS 1 indicating that 8 is more prone to be involved in the carbon-carboxylate coupling and not the carbon-alkoxy coupling. It is inferred from the above results that the mechanism proposed in the literature is not capable of accounting for the experimental data. This inconsistency spurred us to seek for some other alternatives. The following discloses a novel mechanism by which one can easily interpret the results of the Pd(II)catalyzed I(III)-mediated alkoxylation reactions developed by Rao et al. and others. 4,5,9 Oxidative addition Our calculations show that the oxidative addition step starts with the formation of adduct 17 in which reagent BI-OMe binds to the palladium center through its methoxy group (Fig. 3a). Next, the OMe ligand is transferred from the iodine to the palladium via transition structure TS 17-18 to give square pyramidal complex 18. The resultant fve coordinate complex is formally an iodonium salt in which the Pd d z 2 orbital interacts with an empty orbital on the iodine(III) having three-center-fourelectron (3c-4e) character (Fig. 3b); the spatial distribution for the LUMO of 18 (Fig. 3b) indicates the antibonding orbital relating to such an interaction. Interestingly, we found that this complex is extremely reactive toward a redox process through undergoing a simple isomerization. The strong trans-influencing feature of the alkyl group causes the quinoline moiety of the tridentate ligand in 17 to coordinate relatively weakly to the Pd center and thus be highly susceptible to rearrangement. In this situation, the quinoline readily moves from the basal to the apical position via trigonal bipyramidal transition structure TS 18-19 lying only 3.1 kcal mol 1 above 18 to give another square pyramidal structure (19) in that the alkyl group occupies the apical position. This ligand movement turns on a repulsive interaction between the nitrogen lone pair and the flled d z 2 orbital, destabilizing the d z 2 orbital, resulting in two electrons being transferred from the Pd(II) to the iodine(III) center, furnishing 19. 11 As depicted in Fig. 3b, such an electron transfer turns off the Pd-N repulsive interaction and formally changes the oxidation state of the palladium center from +2 to +4. The spatial distribution for the LUMO of 19 (Fig. 3b reduced to the iodine(I) in 19. Due to an increase in population of the iodine p z orbital, the I-O a bond distance becomes longer upon going from 17 to 19 (Fig. 3a) and fnally is cleaved in 19 supported by Wiberg bond index (WBI) analysis showing an almost zero-bond order (0.016) between the I and O a atoms. We found an excellent correlation between the I-O a bond distance and the population of the iodine p z orbital with R 2 ¼ 0.92 (Fig. 3d). 12 These computational results are clearly consistent with the oxidation of Pd(II) by the iodine(III) reagent through the unprecedented mechanism outlined in Fig. 3a. ## Reductive elimination Once the oxidative addition has taken place, the fve-coordinate Pd(IV) complex 19 with a pendant carboxylate group is formed. The sixth coordination site of this Pd(IV) complex can be flled by this pendant carboxylate group to furnish complex 20 (Fig. 4). However, owing to the strong trans-influencing character of the alkyl moiety, the carboxylate binds weakly to the Pd(IV) center in 20, indicating a comparable stability for these two intermediates (19 and 20). Now, we turn our attention to investigation of the C-O reductive elimination from these two Pd(IV) complexes. Based on our calculations, this step can proceed via at least four different variants, as shown in Fig. 4. Accordingly, pathways A and B involve the inner-sphere reductive elimination from complexes 19 and 20 by crossing concerted transition structures TS 4 and TS 5 , respectively. These two pathways install the methoxy group of the oxidant on the fnal product. In pathway C, the C-O reductive elimination occurs through an outersphere S N 2 mechanism involving the nucleophilic addition of the pendant carboxylate to the sp 3 carbon bonded to the palladium by passing through transition structure TS 6 . This pathway leads to formation of the carboxylated product. In pathway D, a methanol pre-activated by the pendant carboxylate nucleophilically attacks the sp 3 carbon to yield 22 via transition structure TS 21-22 . In this pathway, the solvent serves as the alkoxylation reagent, and addition of the methanol to the sp 3 carbon is accompanied by its deprotonation by the pendant carboxylate group. As can be seen from Fig. 4, TS 21-22 is much lower in energy than other transition structures, implying that pathway D is kinetically favored over pathways A, B, and C. However, when the solvent is not the alcohol, pathway D is turned off and thus, in this case, the reaction should preferentially proceed via pathway C, confrmed by the fact that TS 6 energetically lies considerably below TS 4 and TS 5 . The energy profle given in Fig. 4 can answer several questions asked in the Introduction. For example, on that basis, one can explain why the OMe group on the oxidant does not appear in the product, why the solvent (alcohol) serves as the alkoxylation reagent and why when the solvent is not an alcohol, the carboxylate group is installed on the product. ## Regeneration of active catalyst 13 After completion of the C-O reductive elimination via pathway D, intermediate 22 is formed. The dissociation of the carboxylic acid from the resulting intermediate via trigonal bipyramidal transition structure TS 22-23 yields 23 in which the methoxylated organic molecule acts as a tridentate ligand (Fig. 5). This complex subsequently participates in ligand exchange with one of the acetic acids produced from the CMD mechanism (Fig. 1) and generates the more stable intermediate 26 by passing through two transition structures TS 24-25 and TS 25-26 . The higher stability of 26 than 23 implies that the Pd(II) center prefers to coordinate to the acetate rather than the methoxy ligand. The addition of the second acetic acid to 26 via transition structure TS 26-27 gives 27 from which 28 is formed by a proton transfer from the coordinated acetic acid to the nitrogen atom of the methoxylated molecule. The overall activation barrier for formation of 28 from 22 is computed to be less than 15 kcal mol 1 . Finally, displacement of the coordinated product by substrate 1 leads to regeneration of active catalyst 13. From this point on, the second catalytic cycle begins. Since the active catalyst 13 lies 10.3 kcal mol 1 higher in energy than 26, the overall activation barrier to the C(sp 3 )-H activation in the second catalytic cycle increases to 30.0 kcal mol 1 (Fig. 5). Indeed, the ability of the N-H deprotonated product to form a tridentate complex causes this species to bind more strongly to the palladium(II) center than substrate 1. This feature retards the alkoxylation reaction by decreasing the activity of the catalyst, resulting in the process requiring a high temperature for completion. ## Catalytic cycle proposed by the DFT calculations The catalytic cycle shown in Scheme 4 summarizes our calculation results related to the mechanism of the title reaction. The reaction is initiated by coordination of substrate 1 to the Pd complex followed by deprotonation of the N-H and C(sp 3 )-H bonds by the OAc ligands to give cyclopalladated intermediate 16. 13 Subsequently, the substitution of BI-OMe for HOAc affords 17. The OMe ligand in the resultant intermediate ( 17) is then transferred from iodine(III) to palladium(II) to give iodonium 18 stabilized by an anionic palladium(II) complex. Afterward, intermediate 18 undergoes isomerization by moving the quinoline moiety from the basal to the apical position, triggering a redox process by promoting two electrons from Pd(II) to iodine(III), leading to formation of Pd(IV) complex 19 with a pendant carboxylate group. This isomerization not only promotes the Pd(II) oxidation but also sets the stage ready for the reductive elimination by formation of a fve-coordinate complex in which the reacting alkyl moiety occupies a position trans to the empty site. 14 The addition of a methanol (solvent) to the ensuing complex then affords 21. In this intermediate, the methanol is stabilized by a hydrogen bonding interaction with the pendant carboxylate group. Next, the C-O reductive elimination takes place by nucleophilic attack of the methanol on the sp 3 carbon via an outer-sphere S N 2 mechanism to yield 22. Our calculations predict that starting from iodonium salt 18, the oxidative addition and the reductive elimination steps are extremely fast with DG ‡ < 3.5 kcal mol 1 (Fig. 3 and 4). Later, the more stable complex 26 is furnished by following a series of chemical steps. The high stability of this species causes the regeneration of the active catalyst 13 to be considerably endergonic ($10 kcal mol 1 ), resulting in the overall activation barrier to the subsequent catalytic cycle increasing to 30.0 kcal mol 1 (Fig. 5). This fnding clearly explains why the alkoxylation reaction developed by Rao et al. requires a high temperature for completion (Scheme 1). ## Impact of the substituent on the reaction mechanism In a separate study, Rao et al. used a similar method for preparation of acetals through double alkoxylation of the C(sp 3 )-H bonds of butyramide derivative 29 using BI-OMe as the oxidant and Pd(OAc) 2 as the catalyst (Scheme 5). 9b Based on the preliminary results, the authors proposed that the product of the frst alkoxylation is the substrate for the second one. What interest us here is to explore how the electronic feature of the R 0 substituent on a substrate affects the alkoxylation mechanism. Fig. 6 compares the energy profles for alkoxylation of the substrates with R 0 ¼ H, Me, and OMe. Several points emerge from this comparison. First, the overall activation energy of the C-H activation step increases in the order R 0 ¼ H < Scheme 4 Catalytic cycle proposed by the DFT calculations for palladium-catalyzed C(sp 3 )-H alkoxylation using BI-OMe. Me < OMe. It is inferred from this result that the greater the electron donor property of the R group, the higher the activation barrier of the C-H activation. Indeed, a substituent with strong electron donating ability decreases the acidity of the hydrogen being abstracted, leading to the C-H activation to become more energy demanding. Second, the stability of three coordinate Fig. 6 Calculated energy profiles for palladium-catalyzed alkoxylation of the C(sp 3 )-H bond of butyramide derivatives with different R 0 substituent. Free energies (potential energies) are given in kcal mol 1 . complex 7 (7_R 0 ) is determined by the nature of the R 0 group; an R 0 group with strong electron donating ability increases the intrinsic stability of this coordinatively unsaturated species. It also fnds that the relative energy of the transition structure TS 17-18 (TS 17-18_R 0 ) depends on the intrinsic stability of this three coordinate species and decreases in the order R ¼ H > Me > OMe. Third, although the formation of a Pd(IV) complex is unavoidable for the substrates with R 0 ¼ H and Me, this is not the case for the substrate with R 0 ¼ OMe. The IRC calculation shows that transition structure TS 17-18_OMe directly connects 17 -OMe to 19 0 _OMe. It follows that intermediates 18_OMe and 19_OMe are not local minimum and thus no Pd(IV) intermediate is formed in this case. Indeed, if we assume that 19_OMe is generated during the reaction, it is highly unstable and rapidly undergoes a redox process. This is because of the strong pdonor property of the OMe substituent, forcing the Pd IV -C(sp 3 ) s-bond in 19_OMe to be completely polarized toward the palladium center, giving zwitterion complex 19 0 _OMe in which the Pd center bears an oxidation state of +2. In this case, the C-O coupling process from this zwitterion complex does not involve the reductive elimination step, and instead, it takes place via nucleophilic addition of the carboxylate-activated MeOH to the oxonium ion. The DFT calculations show that the addition of the methanol to the oxonium ion is extremely fast and proceeds without involvement of an intermediate. As a result, it is evident from our calculations that a change in the R substituent leads to an abrupt alteration in the reaction mechanism. Scheme 6 shows the modifed catalytic cycle for alkoxylation of the substrate with R ¼ OMe. ## Assessing the generality of our proposed mechanism To evaluate whether the mechanism proposed in this study is applicable to interpret other alkoxylation reactions catalysed by Pd(II) complexes, we investigated the mechanism of the reaction shown in Scheme 7 developed by Shi. 5 In this reaction, iodobenzenediacetate (PIDA) is used as the oxidant. The energy profle given in Fig. 7 ## Conclusion In this work, we performed DFT calculations to elucidate the mechanism of the alkoxylation of the C(sp 3 )-H bonds using hypervalent iodine(III) reagents (ArIX 2 ) catalysed by Pd(OAc) 2 . An unprecedented mechanism for this transformation is revealed by clarifying the oxidative addition step. The calculations indicate that this key step begins with the transfer of an X ligand from ArIX 2 to a Pd-alkyl intermediate. The ligand transfer forms a square pyramidal Pd(II) complex in which iodonium [ArIX] + occupies the apical position and is stabilized by interaction with the Pd d z 2 orbital. The resultant complex then undergoes an isomerization by moving the ligand trans to the Pd-alkyl bond to the apical position. This isomerization considerably destabilizes the d z 2 orbital, promoting the electron transfer from Pd(II) to I(III), resulting in the reduction of I(III) concomitant with the extrusion of the second X ligand as a free anion. The released anion then assists the alcohol (solvent) to nucleophilically attack the Pd(IV)-alkyl bond via an S N 2 mechanism to form a new C-O bond. The information reported in this study is important in enhancing our understanding of the fundamental processes that underpin many catalytic reactions mediated by iodine(III) reagents and catalyzed by Pd(II) complexes and could assist scientists to design new catalytic reactions. ## Computational details Gaussian 16 (ref. 15) was used to fully optimize all the structures reported in this paper at the M06 level of theory. 16 For all the calculations, solvent effects were considered using the SMD solvation model 17 with methanol as the solvent. The SDD basis set 18 with effective core potential (ECP) was chosen to describe iodine and palladium. The 6-31G(d) basis set 19 was used for other atoms. This basis set combination will be referred to as BS1. Frequency calculations were carried out at the same level of theory as those for the structural optimization. Transition structures were located using the Berny algorithm. Intrinsic reaction coordinate (IRC) calculations were used to confrm the connectivity between transition structures and minima. 20 To further refne the energies obtained from the SMD/M06/BS1 calculations, we carried out single-point energy calculations in methanol using the M06 functional method for all of the structures with a larger basis set (BS2). BS2 utilizes the def2-TZVP basis set 21 on all atoms with an effective core potential including scalar relativistic effect for palladium and iodine. Tight convergence criterion was also employed to increase the accuracy of the calculations. In this work, the free energy for each species in solution was calculated using the following formula: where DG 1 atm/1 M ¼ 1.89 kcal mol 1 is the free-energy change for compression of 1 mol of an ideal gas from 1 atm to the 1 M solution phase standard state. An additional correction to Gibbs free energies was made to consider methanol concentration where a MeOH is directly involved in transformations. In such a case, the free energy of MeOH is described as follows: G(MeOH) ¼ E(BS2) + G(BS1) E(BS1) + DG 1 atm/1 M + RT ln(24.72) where the last term corresponds to the free energy required to change the standard state of MeOH from 24.72 M to 1 M. 22 The orbital population analysis and determination of the WBI bond orders were carried out by the NBO7 program. 23

chemsum

{"title": "The role of hypervalent iodine(<scp>iii</scp>) reagents in promoting alkoxylation of unactivated C(sp³)\u2013H bonds catalyzed by palladium(<scp>ii</scp>) complexes", "journal": "Royal Society of Chemistry (RSC)"}

visible_light_switching_of_metallosupramolecular_assemblies

## Abstract: A photoswitchable ligand and palladium(II) ions form a dynamic mixture of self-assembled metallosupramolecular structures. The photoswitching ligand is an ortho-fluoroazobenzene with appended pyridyl groups. The E-isomer is combined with palladium(II) salts affords a double-walled triangle with composition [Pd3L6] 6+ and a distorted tetrahedron [Pd4L8] 8+ (1:2 ratio at 298 K). Irradiation with 410 nm light generates a photostationary state with ~80% of the E-isomer of the ligand which results in the selective disassembly of the tetrahedron, the more thermodynamically stable structure, and the formation of the triangle, the kinetic product. The triangle is then slowly transformed back into the tetrahedron over 2 days at 333 K. The Z-isomer of the ligand does not form any well-defined structures and has a thermal half-life of 25 days at 298K. This approach shows how a thermodynamically preferred self-assembled structure can be reversibly pumped to a kinetic trap by small perturbations of the isomer distribution using non-destructive visible light. ## INTRODUCTION The structure and function of self-assembled species, such as molecular cages, can be controlled using stimuliresponsive components. Different stimuli have been used to perturb metal-template supramolecular assemblies 1 including light, 2 guest molecules, 3 pH changes, 4 competing ligands 5 and changes to solvent. 6 Light, especially the visible spectrum, 7 is appealing due to its easy use, potential for highly specific targeting, and the high resolution of spatial and temporal application. 8 Molecular photoswitches, 9 can be isomerized reversibly by light, with each isomer having different geometries and electronic properties. These differences in properties have been used to control the properties of gels, 10 polymers assemblies, 11 or liquid crystals, 12 and to perform functions including acting as light-activated receptors 13 or pharmaceticals, 14 or pumping systems away from thermodynamic equilibrium. 15 The most studied photoswitches are those based on azobenzene, 16 which can be isomerized between a stable E-isomer and a metastable Z-isomer. However, unsubstituted azobenzene requires potentially destructive UV light to form the meta-stable Z-isomer that has a thermal half-life at room temperature of only 2 days. Significant advances have been made in developing azobenzene-type molecules that operate effectively with visible light, 9c,17 with one of the most successful modifications being the introduction of ortho-fluoro substituents (Figure 1a). 18 These ortho-fluoroazobenzenes allow bidirectional visible-light switching with thermal half-lives that can exceed 2 years and have been incorporated into MOFs 18g and discrete self-assembled structures, 18h and have been used to control molecular folding 18f or the function of enzymes. 18j Conceptually there are two approaches for combining photoswitches with discrete self-assembled structures: encapsulation or direct incorporation as part of the struc-ture. The first strategy involves binding the photoactive unit inside a cavity, such as encapsulating azobenzene type derivatives. 19 Encapsulation a photoswitch can also restrict switching or perturb the balance of isomers. 19b,20 Using photoswitches as structural components of selfassembled structures has proven more difficult. Although there are many examples of large photoswitchable assemblies, 21 such as micelles, vesicles or liquid crystals 12 formed with polymers, 11 there are relatively few examples of photoswitches being self-assembled into well-defined, discrete structures. In a key example, pyridine-based ligands and palladium(II) were self-assembled into a [Pd12L24] 24+ molecular sphere with endohedral azobenzene groups 2a which could be switched with UV to increase the hydrophilicity of the sphere's cavity. Some other examples of pyridyl-functionalized switches include [M2L4] 4+ cages formed with stiff-stilbenes and palladium(II), 22 chiral [M6L3] 6+ metallocycles formed from dithienylethene (DTEs) 23 and platinum(II), 24 and related ligands reacted with iron(II/III) to form [Fe2L3] n+ helicates. 25 The first example of a molecular cage with functioning azobenzene-type photoswitches as linkers used cyclotriguaiacylene units with three appended pyridyl-azo-phenyl photoswitches and iridium(III) complexes to form [Ir3L2] 3+ cages. 26 The flexible linkers allowed photoswitching to occur without disrupting the cage topology. The most wellstudied photoswitchable cages are based on pyridylfunctionalized DTE photoswitches assembled with palladium(II) ions reported by the Clever group. 27 The difference in geometries has been exploited for selective guest uptake, 27a,27d,27e and control over macromolecular properties when incorporated into gels. 28 Photoswitching units can also modulate the topology of metallosupramolecular structures; however, this usually leads to the assembly of new non-discrete structures. 30 There are few reported examples of modulation between discrete metallosupramolecular structures, with some key examples represented in Figure 1b,c. 27b, 29 One example used azobenzene or stilbene based ligands to form [M2M'2L4] 8+ (M = Pd, M' = Pd or Re) macrocycles where UV irradiation isomerize the azo unit to contract the macrocycle to the smaller [M2L2] 4+ species. 29,31 Other examples use DTE-based ligands. 27b,27e For one system, the open and closed isomers give rise to a double-walled triangle (as the major component) and a cuboctahedral sphere, respectively. 27b These species can be interconverted using UV and green light, giving reversible control over the structure by external stimuli although the conversion was relatively slow, with a full cycle taking over 3 days. A more recent example was able to eject one ligand from a Pd2L4 cage upon irradiation. 27e Despite these examples, there are no reports of metallosupramolecular structures which can be reversibly rearranged using visible light only. Herein we report a system of two discrete metallosupramolecular assemblies, formed from an orthofluoroazobenzene ligand (Figure 1d). The system can be driven out-of-equilibrium with visible light due to the different kinetic labilities of the structures. To the best of our knowledge this is one of the only examples of light-induced topology changes and the first example of all-visible light switching between discrete structures. ## RESULTS AND DISCUSSION We synthesized substituted ortho-fluoroazobenzenes in moderate yield over three steps from commercially available 4-bromo-2,6-difluoroaniline (see SI-1,2 for details). 32 Compound 1 was obtained in 65% yield using a methodology previously used to generate unsymmetrical azobenzene derivatives, 18a,18b,33a,33b Boronic ester substituted ortho-fluoroazobenzene 2 has been previously reported, 18c but use of microwave heating allowed us to reduce the reaction time to 15 minutes with a trivial work-up that excluded chromatography. Suzuki coupling gave the photoswitchable ligand 3 (53% yield) and the control compound, phenyl derivative 4 (20% yield). The second coupling reaction did not always reach completion despite the arylhalide being in excess, with the mono-substituted product being identified and characterized (see SI-2.4, SI-16 for details). This suggests the second coupling reaction is considerably more difficult than the first. The compounds were isolated as mixtures of the thermodynamically favored E-isomer and the metastable Z-isomer. Heating a solution of 3 in DMSO-d6 generated the pure E-3 isomer as observed by 1 H and 19 F NMR spectra (Figure 2b). 34 The UV-vis absorption of photoswitchable ligand E-3 (Figure 2c) extends into the visible, with a visible absorption maximum at 466 nm assigned as the n-π* band and a band at 356 nm assigned to the π-π* transition (in DMSO at 298 K). Both transitions are red-shifted relative to the parent ortho-fluoroazobenzene, which has an n-π* transition at 460 nm and a π-π* transition at 314 nm (in DMSO at 298 K). 18a The larger red-shift of the n-π* band compared to the π-π* was also reported for 2,2',6,6'tetrafluoro-4,4'-diacetamidoazobenzene, 18a suggesting this effect is due to substitution with electron donating groups. Photoswitchable ligand 3 undergoes reversible photoswitching with visible light (Figure 2c). Irradiation of 3 with an LED centered at 530 nm generated a photostationary state comprising 80% Z-3 (calculated from 19 F NMR signal integrations, Figure 2b and SI-4.1). Subsequent irradiation at 410 nm generated a new photostationary state comprising 85% E-3. The calculated absorption spectrum 35 of Z-3 shows an n-π* transition with an absorption maximum at 432 nm, slightly red-shifted compared to unsubstituted or ester substituted orthofluoroazobenzenes (λmax = 417-421 nm). 18a The separation between the n-π* bands for the two isomers of 3 (Δλn-π* = 33 nm) is similar to that found for other orthofluoroazobenzenes with electron-donating groups in the para position, 18a but less than that for the parent orthofluoroazobenzenes or examples with electron withdrawing groups (Δλn-π* = 30 to 50 nm). 18b Nonetheless, selective photoswitching is still achieved between the isomers. Photoswitchable ligand Z-3 has a thermal half-life of ≈ 25 days at 298 K (thermal barrier of 110 kJmol -1 , measured at 333 K in DMSO, see SI-4.3). Photoswitch 4 has similar properties to photoswitchable ligand 3. For example, photoswitch 4 has an n-π* absorption band at 462 nm and a π-π* band at 360 nm, and Z-4 has a thermal half-life of ≈37 days at 298 K (see SI-5 for details). Photoswitches 3 and 4 both have shorter half-lives compared to the parent ortho-fluoroazobenzene which has a half-life of 700 days (thermal barrier of 117 kJmol -1 , measured at 333-373 K in DMSO) 18b Having characterized photoswitchable ligand 3, we next investigated its self-assembly with palladium(II) ions. When [Pd(CH3CN)4](BF4)2 was added to E-3 in DMSO-d6 a new species was immediately formed as observed by 1 H and 19 F NMR spectroscopy (Figure 3b, SI-6). Equilibration in the dark at room temperature over 10 days led to the formation of a new, lower symmetry, assembly comprising 69% of the mixture (Figure 3c). Using 1 H NMR diffusion (Figure 4a) and ROESY NMR (Figure 4b) data we identified two separate species, with the higher symmetry species having a smaller hydrodynamic diameters (27 vs 31 ). Similar self-assembly using a more soluble palladium salt 36 gave the same two structures albeit with a slightly different relative abundance, see SI-7. Characteristic downfield shifts of the 1 H NMR signals for pyridyl protons indicate coordination to the metal ion (see SI-8.1 for full details). 37 In the initially formed species the ligand retains its original symmetry and the 1 H NMR signal for H a (see Fig. 3 for atom labels) shifts upfield by ≈ 0.2 ppm, consistent with shielding effects commonly seen for related structures. 38,37b The lower symmetry species has a doubling of all ligand signals (Figure 3c), with a significant upfield shift (≈ 0.4 ppm) of the H e ' proton compared to the symmetric species. The 19 F NMR spectrum confirms the reduced symmetry with two peaks observed for the lower symmetry species. The significant peak shifts observed in the NMR spectra did not allow unambiguous assignment of the E/Z-isomerization state. The UV-visible absorption spectrum of the mixture was also unhelpful for assigning the E or Z isomer composition (SI-15.1). Therefore, a degradation experiment was performed. 4-Dimethylaminopyridine (DMAP) was added to the equilibrated mixture in the dark which rapidly disassembled the structures to form exclusively E-3 and [Pd(DMAP)4](BF4)2 as seen by 1 H NMR spectroscopy (See SI-12). Due to the long thermal half-life of Z-3, this degradation experiment indicates that the observed 1 H NMR peak shifts and changes in the UV-visible absorption spectra are due to the constrained local environment or distortions of the E-3 ligand imposed by the structure, rather than isomerization of the ligand. High resolution electrospray ionization mass spectrometry (ESI-MS) identified two major species, a [Pd3(3)6] 6+ and a [Pd4(3)8] 8+ assembly (Figure 5, SI-9) with a range of charge states corresponding to sequential loss of BF4anions from these structures. The combination of NMR and MS data, together with preliminary molecular modelling, was used to propose the topologies of the self-assembled structures for [Pd3(3)6] 6+ and a [Pd4(3)8] 8+ (Figure 6b and 6c). For the [Pd3(3)6] 6+ species, the NMR spectra indicates a highly symmetrical structure, which is assigned as a double-walled triangle. 39 For the species with composition [Pd4(3)8] 8+ , several possibilities can be considered (Figure 6a): a doublewalled square, 39b,39e an interpenetrated double cage 40 or a distorted tetrahedron. 39a,39h,41 The double-walled square would nominally have D4h symmetry with all pyridyl rings being equivalent. This is inconsistent with the observed number of signals in the NMR spectra. The interpenetrated double cage structure would show a doubling of the 1 H and 19 F NMR signals as observed. However, in previous reports of such topologies the transient formation of a [Pd2L4] 4+ cage was observed in the 1 H NMR spectrum and by ESI-MS. 40b Such species were not observed for the current system and molecular modelling also suggests significant strain would be required in the [Pd2(3)4] 4+ subunit. The structure is therefore proposed as a distorted tetrahedron with C2v symmetry. Ligands with 3-pyridyl groups bridged by phenyl 39a,41b or BINOL linkers 41a have been previously assembled into analogous distorted tetrahedra with palladium(II), but the topology remains rare. 39a,41a,41b,41d,39h For [Pd4(3)8] 8+ the groups of signals from the non-equivalent ligands were assigned using 2D NMR techniques and by comparing to previously reported examples. 39a The local environment for the double-bridged ligands is similar to that observed for the double-walled triangle [Pd3(3)6] 6+ species. The single-bridged ligands are more similar to free E-3, especially the phenyl proton (H a' ) which is distal to coordinating pyridines. The molecular model suggests a longest axis (28 ) in agreement with the calculated hydrodynamic diameter (31 ) from the diffusion NMR data. Variable temperature 1 H NMR spectra (SI-8.3) confirmed the two species were in equilibrium. Increasing the temperature to 333 K gave a mixture containing 63% of the smaller [Pd3(3)3] 6+ species. This is ascribed to entropic considerations, as proposed in other systems. 42 The system initially remained out of equilibrium upon cooling to 298 K, reaching the original distribution after 18 hours in the dark. This indicates that the double-walled triangle acts as a kinetic trap for the system, consistent with the initial observations upon combination of E-3 and [Pd(CH3CN)4](BF4)2. Having investigated the self-assembly properties of E-3, we next investigated the behavior of the Z-3 isomer. A sample of 3 was enriched to 80% Z-3 by irradiation with 530 nm light, then combined with [Pd(CH3CN)4](BF4)2 in DMSO-d6. The resulting poorly resolved 1 H NMR spectrum suggests the formation of non-distinct or polymeric products, which do not significantly resolve over time (see SI-13). To understand the self-assembly behavior, we next considered the binding affinity of the ligand for palladium(II) centers. To the best of our knowledge, and despite their widespread use in supramolecular self-assembly, quantitative binding constants for simple pyridine derivatives to palladium(II) ions do not appear to be reported. To study a single 1:1 binding event, we used a palladium(II) complex with a tridentate terpyridine ligand (ttpy = 4'-(para-tolyl)-2,2':6',2''-terpyridine), [Pd(ttpy)(DMSO)](BF4)2, which has a weakly bound solvent molecule that can be readily exchanged for the other ligands. We used 3-methylpyridine as a simple monodentate ligand (SI-3 for synthetic details). Isothermal titration calorimetry (ITC) was used to measure the 1:1 binding constant (see SI-14.1). The relative binding constant is 1.73 ×10 4 mol -1 in DMSO, equivalent to a binding energy of just 24 kJ•mol -1 at 298 K. Similar ITC measurements with 3 and [Pd(ttpy)(DMSO)](BF4)2 indicated only weaker binding (Ka <1000), although solubility difficulties prevented quantitative measurements. Competitive binding experiments monitored by 1 H NMR spectroscopy confirmed that 3 is nearly completely displaced from [Pd(ttpy)(DMSO)](BF4)2 when one equivalent of 3methylpyridine is added (see SI-14.2), consistent with 3 being a surprisingly poor ligand for palladium(II). We also investigated the influence of palladium(II) ions on the photoswitching behavior of ligand 3. When 100 equivalents of [Pd(ttpy)(DMSO)](BF4)2 was added to ligand 3 and the sample was irradiated with 530 nm light, the same thermal Z→E half-life in the dark was measured by UV-vis absorption, (see SI-4.3, SI-4.4). As ligand 3 has only weak affinity for palladium(II), its ability to assemble into discrete structures suggests that cooperativity is responsible for stabilizing the resulting self-assembled structures. The distribution between [Pd3L6] 6+ and [Pd4L8] 8+ can be pumped away from equilibrium using light, even though Z-3 did not self-assemble into well-defined structures with palladium(II) ions. After irradiating a mixture of [Pd3(3)6] 6+ and [Pd4(3)8] 8+ in DMSO-d6 with 410 nm light for 10 minutes, 1 H and 19 F NMR spectroscopy reveals a significant increase in the population of [Pd3(3)6] 6+ , while also showing the concomitant decrease of [Pd4(3)8] 8+ (Figure 7b, ii). No new signals were observed, suggesting no new well-defined self-assembled species were formed. This observation was reaffirmed by high-resolution ESI-MS, the relative population of [Pd3(3)6] 6+ increased after irradiation with 410 nm light (see SI-15). Irradiating the sample with 530 nm light for 10 minutes resulted in the deformation of the two species as seen in the poorly defined 1 H and 19 F NMR spectra, suggesting the formation of polymeric species, or other low symmetry species (Figure 7b, iii). The large population of [Pd3(3)6] 6+ could be recovered by irradiating the system again with 410 nm light for 10 minutes (Figure 7b, iv), demonstrating selective and reversible assembly and disassembly of the triangle species. After heating the sample at 60 °C for 2 days the original distribution was largely recovered ([Pd3(3)6] 6+ : [Pd4(3)8] 8+ = 3:4, Figure 7b, v)), although some chemical shift changes and peak broadening had occurred. The broad peaks observed in the 1 H and 19 F NMR spectra after irradiation are consistent with the involvement of Z-3 within the self-assembled species (Figure 7b, iii), either as structural components or as guest molecules. This effect is far more pronounced within [Pd4(3)8] 8+ , supporting the notion that [Pd4(3)8] 8+ is more flexible and able to accommodate the mismatched ligand whereas [Pd3(3)6] 6+ is more rigid and well-defined. 3)6] 6+ ); ii) the same sample after irradiation with 410 nm light for 10 minutes (82% [Pd3(3)6] 6+ ); iii) the same sample after irradiation with 530 nm light for 10 minutes; iv) the same sample after irradiation with 410 nm light for 10 minutes again; and v) the same sample after 2 days of being heated at 60 °C followed by 6 h of equilibrating at room temperature. The selective disassembly of [Pd4(3)8] 8+ can be rationalized by considering the composition of ligands, the rate of ligand exchange for each species, and the constraints imposed on the photoswitching of ligand 3 while assembled. Variable temperature NMR experiments confirm the struc-tures are dynamic with exchange of ligands and solvent molecules, as is common for palladium(II)-pyridyl assemblies (see SI-8.3). 43 If photoisomerization is suppressed within the self-assembled structures, as observed for a DTE-based cage, 27b ligand 3 can only isomerize after dissociating from palladium. For the tetrahedron [Pd4(3)8] 8+ , a E-3 ligand can dissociate and photoisomerize, but the newly generated Z-3 ligand cannot reassemble into the same original structure. We propose that a metastable [Pd4(3)7] 8+ structure is formed and the ligands rapidly rearrange to form the double-walled triangle, [Pd3(3)6] 6+ . As [Pd3(3)6] 6+ is more inert, any free E-3 in solution will be kinetically trapped as [Pd3(3)6] 6+ . As such, irradiation with 410 nm light continuously pumps the system out-ofequilibrium to favour the formation of the less thermodynamically preferred [Pd3(3)6] 6+ . The PSS generated when irradiated with 530 nm light comprises only 20% E-3, which appears too low to form a significant amount of [Pd3(3)6] 6+ . This finding is consistent with our experiments using a sample of enriched Z-3 and palladium(II) which also resulted in the same ill-defined mixtures. The observed behavior is surprising as it results from a relatively small change (~20%) in the isomer distribution caused by irradiating with 410 nm light. Typically, stimuli responsive architectures are designed to maximize the proportion of components that are switched. This work offers a different approach, where small changes in isomer distribution can be amplified to significant changes within the system, similar to the sergeants-and-soldiers concept 44 in self-sorting. To the best of our knowledge, this is the first example of a self-assembled system where the configuration can be controlled using only visible light and the resultant distribution contains the same sub-components as the equilibrium distribution. ## CONCLUSION We have shown that building visible-light switchable ofluoroazobenzenes into palladium(II)-pyridyl selfassemblies leads to visible-light responsive systems. Irradiating with visible light reversibly redistributes the subcomponents, driving the system out-of-equilibrium to form the higher energy, but less labile, structure. Unlike previous examples, the distinct assemblies contain the same photoisomer of the ligand. This approach of pumping systems to metastable states exploits kinetic effects to amplify small changes in photoisomer distributions to generate large changes in structural distributions.

chemsum

{"title": "Visible light switching of metallosupramolecular assemblies", "journal": "ChemRxiv"}

“broken-hearted”_carbon_bowl_<i>via</i>_electron_shuttle_reaction:_energetics_and_electron_coupling

## Abstract: Unprecedented one-step C]C bond cleavage leading to opening of the buckybowl (p-bowl), that could provide access to carbon-rich structures with previously inaccessible topologies, is reported; highlighting the possibility to implement drastically different synthetic routes to p-bowls in contrast to conventional ones applied for polycyclic aromatic hydrocarbons. Through theoretical modeling, we evaluated the mechanistic pathways feasible for p-bowl planarization and factors that could affect such a transformation including strain and released energies. Through employment of Marcus theory, optical spectroscopy, and crystallographic analysis, we estimated the possibility of charge transfer and electron coupling between "open" corannulene and a strong electron acceptor such as 7,7,8,8tetracyanoquinodimethane. Alternative to a one-pot solid-state corannulene "unzipping" route, we reported a nine-step solution-based approach for preparation of novel planar "open" corannulene-based derivatives in which electronic structures and photophysical profiles were estimated through the energies and isosurfaces of the frontier natural transition orbitals. ## Introduction Unzipping nanotubes, 1-5 nanosheets, 6,7 buckyballs, or annulenes is driven by the renewed interest in fundamental understanding and practical access to novel structural transformations, 17,18 leading to materials with unique optical and electronic profles. For instance, cutting and unravelling of nanotubes resulted in nanoribbons having electronic properties that can be varied as a function of their width, and therefore, applied in a variety of electronic devices including feld-effect transistors, light-emitting diodes, and transparent conductive electrodes. 19,20 In addition, hydrogenation of graphene nanoribbons led to enhanced photoluminescent properties that could pave the way for the development of optically active graphene nanoribbon-based devices. 21 However, promotion of selective C]C bond cleavage in graphitic materials is challenging, and although there have been examples of structural changes due to periphery modifcations of buckybowls (p-bowls), ring expansion, 22 or opening of the strained pbowl, 29,30 these accounts are overall very limited. Pursuing the route of C-C bond activation in curved p-bowl-containing systems is advantageous as this could lead to addressing challenges such as selective sphere opening for preparation of endohedral fullerenes, shortening carbon nanotubes (CNTs), guest integration within the CNT body, as well as access to a class of materials that has not been prepared through "wetchemistry, conventional" routes. One strategy to facilitate C-C bond activation is to employ, for instance, strain energy as a variable, release of which could energetically promote such transformations. 31 Indeed, as presented in this report, release of strain energy can be the driving force for planarization of the naturally curved buckybowl surface (e.g., C 20 H 10 , corannulene), since there is no direct route to cleave a C]C bond, except through uncontrollable flash vacuum pyrolysis 22,27,29 or addition of a directing group (and a catalyst). 32 Although there are numerous reports of catalytic hydrocracking of planar polycyclic aromatic hydrocarbons (PAHs) i.e., increasing the ratio of hydrogen-to-carbon, there are very few accounts on C-C bond cleavage following the hydrogenation step. 37 The literature precedent for C-C bond scission primarily relies on the assistance of transition metal catalysts, high hydrogen pressure, elevated temperatures, or a combination of all three parameters. Therefore, unexpected C-C bond cleavage (discovered from photophysical studies of alignment of electron donor (corannulene) and acceptor (7,7,8,8-tetracyanoquinodimethane, TCNQ) in the solid state) reported herein led us to probe mechanistic pathways to determine the feasibility for p-bowl planarization and factors that could affect such a transformation including strain energy (E s ) and released energy (E 0 , Scheme 1, see more details in the ESI †). The electron coupling and charge transfer (CT) rates between "open" corannulene (or parent corannulene) and TCNQ were evaluated by applying Marcus theory. In addition to the solid-state reaction, we also offer more "conventional" solution-based nine-step synthetic routes for the preparation of novel "open" corannulene analogs. In the reported fndings, we also discuss the electronic structure and photophysical profles of the synthesized "open" analogs through estimation of their energies and isosurfaces of the frontier highest occupied and lowest unoccupied natural transition orbitals (HONTO and LUNTO). ## Results and discussion The reductive C]C bond cleavage and consecutive corannulene planarization to form 5,6-dimethyl-benzo[ghi]fluoranthene (planar corannulene analog (P-C 20 H 14 ), Scheme 1) was achieved through a one-pot solid-state reaction, in which corannulene (15 mg, 0.060 mmol), TCNQ (an electron shuttle; 14 mg, 0.068 mmol), and zinc powder (a reducing agent; 50 mg, 0.76 mmol) were ground together (further experimental details can be found in the ESI †). After that, the reaction mixture was placed in a glass tube, a drop of hydrochloric acid (proton source) was added, and the glass tube was flame-sealed under dynamic vacuum (4 10 5 mbar). Heating the reaction mixture at 200 C for six days resulted in the formation of dark brown needle-shaped crystals suitable for single-crystal X-ray diffraction analysis (Scheme 1). As shown in Scheme 1, such treatment resulted in planarization of the corannulene bowl through partial hydrogenation and formation of P-C 20 H 14 . X-ray crystallographic studies of (C 20 H 10 )$(P-C 20 H 14 )$(TCNQ) (1) cocrystals revealed that the packing consists of alternating columns of TCNQ and P-C 20 H 14 along the c-axis direction (Fig. S1 and S2 †). Furthermore, neither mass spectrometry nor spectroscopic studies identifed the presence of any other partially hydrogenated products (Fig. S1-S3 †). To gain insight into a plausible mechanism of such p-bowl opening during the one-pot solid-state synthesis (Scheme 1), we initially tested the hypothesis of whether all components of the reaction mixture were essential to perform the solid-state C]C bond cleavage. Our results illustrated that the absence of one of the components of the reaction mixture resulted in either no transformation or formation of (corannulene) 2 $(TCNQ) cocrystals, previously reported in the literature (CCDC 1037414) 45 and also detected in our studies (Fig. S4 †). Utilization of a different redox mediator rather than TCNQ (e.g., methyl viologen) did not lead to corannulene opening despite previous reports in which TCNQ and methyl viologen have both been used as electron shuttles in various biological applications. Variation of synthetic conditions, for instance, replacement of the zinc powder with sodium dithionite 50 as a reducing agent did not lead to hydrogenated products (see ESI † for more details). Utilization of more conventional solution-based routes through heating the same reagents (C 20 H 10 /TCNQ/Zn/HCl) in a series of organic solvents was also attempted. We varied the reaction media starting with the solvents possessing low boiling points (e.g., dichloromethane or methanol), transitioning to dichloromethane/water or methanol/water mixtures, and fnally attempting heating in the higher boiling glycerol (b.p. ¼ 290 C) or ethylene glycol (b.p. ¼ 197 C) to more closely match the reaction temperature (200 C) of the solid-state synthesis. In all reactions, no evidence of P-C 20 H 14 was detected according to the 1 H nuclear magnetic resonance (NMR) spectroscopic or mass spectrometry analysis. Notably, the reported hydrogenation reactions of corannulene typically occurred under relatively harsh conditions (e.g., electron bombardment, alkyllithium reagents, or alkali metals), and even despite them, reactions typically led to hydrogenation of one or two rim C]C bonds without carbon-carbon bond cleavage. Since the developed conditions (C 20 H 10 /TCNQ/Zn/HCl) required the presence of zinc, we also probed the Clemmensen reduction that uses zinc amalgam and concentrated hydrochloric acid. 58 Mass spectrometry and 1 H NMR spectroscopy studies of reaction products detected the presence of only pristine corannulene and did not detect any traces of corannulene hydrogenation. As a logical progression, we surveyed an electrochemical method suitable for arene reductive transformations, 59 but proved unsuccessful. Finally, attempts to electrochemically cleave the C]C bond by bulk electrolysis were performed in anhydrous N,N-dimethylformamide or acetonitrile for up to two days, but were also not successful. In line with these studies, we probed the reaction conditions previously utilized for the ring-opening of other nonplanar structures such as o-carborane. For that, we used a triosmium carbonyl complex, Os 3 (CO) 10 (NCMe) 2 ; however, no ring-opening of corannulene was observed, while successful ocarborane opening occurred at 150 C in a nonane reflux. 62 Further experimental investigations were pursued to rule out aromaticity stabilization as a main factor by performing reactions with signifcantly less strained PAHs including pyrene (0.0 kJ mol 1 ) 31 or phenanthrene (0.0 kJ mol 1 ) 63 under experimental conditions similar to those used for the reaction with highly-strained corannulene (101 kJ mol 1 ; 31 see ESI † for more details). As a result, no bond cleavage was detected in any of these systems, and formation of only PAH$TCNQ complexes was observed (e.g., phenanthrene$TCNQ complex (similar to the structure reported in the literature 64 ) or pyrene$TCNQ complex, Fig. S17; see more details in the ESI †). Thus, this experimental evidence suggests that one of the driving forces for the observed solid-state reaction could potentially be a release of energy through buckybowl planarization (Fig. 2a). To prove this hypothesis, we estimated released energy, E 0 , for PAHs and carbon p-bowls as shown in Scheme 1. For instance, E 0 for corannulene was calculated to be 202.0 kJ mol 1 (Fig. S9, B3LYP/6-31+G*, see the ESI for more details †). In contrast, E 0 calculated for the PAHs in Scheme 1 was found to be less than 135.7 kJ mol 1 (Table S3 †). Therefore, corannulene opening is much more energetically favorable in comparison with the PAHs shown in Scheme 1. A similar statement is also valid for a family of extended p-bowls for which the estimated E 0 was even higher than that of corannulene (Scheme S3 †). The estimated enthalpy of the reaction (eqn (1)), was found to be 179.5 kJ mol 1 (239 and 59.5 kJ mol 1 for only the electronic and the ZPE-corrected electronic energies, respectively, using density functional theory (DFT, Table S2, see the ESI for more details †)). Thus, a combination of two parameters, strain energy (E s ) and released energy (E 0 ), highlights the unique nature of buckybowls in comparison with the considered PAHs (Scheme 1). As a next step, we took a closer look at a possible mechanism for p-bowl hydrogenation and C-C bond cleavage. On the basis of our theoretical calculations, experimental results, and literature reports, we hypothesized that the transformation of C 20 H 10 to P-C 20 H 14 occurs in a series of reactions that is frst initiated by a sequence of electron and proton transfers in which hydrochloric acid acts as the proton source (Fig. 1). Moreover, probing the strength of the C-C bond revealed a signifcantly weaker bond (115 kJ mol 1 for C 20 H 12 and 9 kJ mol 1 for C 20 H 12 c, Fig. 1) than a typical C-C bond in RCH 2 -CH 2 R systems, allowing for bond cleavage to occur. 65,66 For instance, if R is a substituent on a pyrene or coronene core then the electronic energy of the C-C bond would be approximately 350 kJ mol 1 (Fig. S14 †) and 302 kJ mol 1 , respectively (Fig. S15 †). Comprehensive analysis of photophysical data for the obtained crystals of 1 revealed properties that are uncharacteristic of the individual components i.e., corannulene and TCNQ themselves. Based on photoluminescence and epifluorescence microscopy studies, the obtained crystals of 1 exhibited red emission (l max ¼ 705 nm, l ex ¼ 365 nm) in contrast to their constituents (l max (TCNQ crystal) ¼ undetectable emission and l max (corannulene crystal) ¼ 490 nm, l ex ¼ 365 nm, Fig. S5 †). Furthermore, in contrast to diffuse reflectance (DR) profles of pristine corannulene and TCNQ (Fig. S6 †), the appearance of a new red-shifted band (550 nm) in the DR profle of 1 was detected. Based on our theoretical calculations using timedependent density functional theory (TDDFT), the new band (550 nm) is characteristic of CT complex formation (see the ESI for more details †) that is in line with a previous report on PAHs and TCNQ co-crystals. 24 In particular, according to our studies using the B3LYP-D3/6-311+G** level of theory (Fig. 2b), both a bathochromic shift and new band appearance could be attributed to CT 66,67 between the HOMO2 and LUMO of a TCNQ/P-C 20 H 14 "stack" (Fig. 2c). To further shed light on the experimental changes of the emission profle, we examined optical excitations of isolated corannulene, P-C 20 H 14 , TCNQ, and the relevant dimers through the TDDFT calculations based on the B3LYP-D3/6-311+G** method. The considered TCNQ/P-C 20 H 14 "stack" is the only species with excitation energies of appreciable strength around 690 nm (1.8 eV; Fig. S7 †), which is in agreement with the experimentally observed red emission at l max ¼ 705 nm (Fig. S5 †). The lowest excitations for TCNQ, pbowl, and P-C 20 H 14 are 413, 288, and 344 nm (3.0, 4.3, and 3.6 eV), respectively (Fig. S7 †). To further probe the idea that CT is more effective in an exclusively planar TCNQ/P-C 20 H 14 "stack" rather than in a TCNQ/C 20 H 10 "stack", that encounters steric hindrance from the curved surface of the p-bowl, 45 we employed Marcus theory 68 to compare the electron coupling (that is proportional to CT rate) between TCNQ/C 20 H 10 and TCNQ/P-C 20 H 14 using eqn (2): where k ¼ charge transfer rate, V c ¼ electron coupling, l ¼ reorganization energy of the system, and DG 0 ¼ energy difference between the initial and fnal states (see the ESI for more details †). According to the Marcus theory model, TCNQ/P-C 20 H 14 could result in ca. 128-fold increase in electron coupling compared to TCNQ/C 20 H 10 (Fig. S8 and Table S5, see the ESI for more details †). Since electron coupling is related to the electron transfer rate, we can surmise that there is likely an increased electron transfer rate as well. 66 The charge on the TCNQ molecule was evaluated by applying the Kistenmacher relationship (i.e., correlation between TCNQ intramolecular bond distances and charge on TCNQ) 69 using the crystallographic data of 1 and (corannulene) 2 $TCNQ co-crystals. 45 In the case of 1, the charge on TCNQ was estimated to be 0.84 and for (corannulene) 2 -$TCNQ co-crystals was found to be 0.20, suggesting more effective CT can occur in 1 between "open" corannulene (P-C 20 H 14 ) and TCNQ. We calculated electronic transitions corresponding to the ground state, frst and second excited singlet states of P-C 20 H 14 in THF (Fig. 3 and S29 †). Delocalization of excited energy levels in P-C 20 H 14 was slightly enhanced, leading to optical transitions of the frst and second excited states with values of 399 and 409 nm, respectively (Fig. S29 †). As an alternative scalable approach to access a family of "open" corannulene-containing derivatives, we report a solution-phase route (see ESI †). 70 Despite a number of required steps (Schemes S1 and S2 †), in comparison with the one-step solid-state synthesis, the "solution" approach has some advantages since it does not rely on selection of the specifc substrate/electron shuttle/reducing agent system and also provides a scalable route for the synthesis of a library of new planar corannulene-type analogs. The synthetic details for preparation of 5-methylbenzo[ghi]fluoranthene (C 19 H 12 , X, Fig. 4a) and 5-ethyl-6-methylbenzo[ghi]fluoranthene (C 21 H 16 , X 0 , Fig. 4b), using this approach, are provided in ESI. † Sublimation of X (Scheme S1 †) allowed for the formation of single crystals of X suitable for X-ray diffraction (Fig. S24 †). The structure of X 0 was confrmed using 1 H and 13 C NMR spectroscopy and mass spectrometry (Fig. S21 †). As in the case of solid-state "open" P-C 20 H 14 , both X and X 0 structures possess a planar geometry (Fig. S24 and S25 †). The emission studies of the prepared X and X 0 compounds revealed a red-shifted emission (l max ¼ 548 nm and 573 nm, l ex ¼ 365 nm, respectively, Fig. S26 †) in contrast to pristine corannulene (l max ¼ 490 nm, l ex ¼ 365 nm, Fig. S5 †). The emission maxima of X and X 0 in THF was found to be 479 nm and 502 nm, respectively, (l ex ¼ 365 nm) and is hypsochromically shifted compared to the solid-state 548 and 573 nm-centered emission, respectively (l ex ¼ 365 nm, Fig. 4a, b, S27, and S28 †). In a similar vein to TDDFT calculations of P-C 20 H 14 , we determined the optical transitions corresponding to the ground state, frst and second excited singlet states of X and X 0 (Fig. S27 and S28 †). While the electronic transition of the frst excited singlet state for both X and X 0 did not differ from the optical transitions corresponding to the ground state (351 nm for X and 361 nm for X 0 ), optical transitions for the second excited singlet states were determined to be 395 nm and 412 nm for X and X 0 , respectively. Electronic transitions corresponding to the second excited singlet state can be associated with the emission profles that are similar to the experimental data (see the ESI for more details). ## Conclusions To summarize, we report the frst example of a unique one-step C]C bond cleavage in the traditionally very robust p-bowl occurring via an electron shuttle reaction. Such ring opening is unprecedented in the literature and has not been observed for pristine p-bowls (e.g., corannulene) to date (with the exception of uncontrollable brute force vacuum pyrolysis 22 ). PAH hydrogenation has been previously observed under harsh experimental conditions (e.g., high hydrogen pressure or extreme temperatures above 1000 C), therefore the formation of P-C 20 H 14 in a one-pot synthesis under relatively mild conditions is an unexpected and remarkable result. Through employment of Marcus theory, optical spectroscopy, and crystallographic analysis, we estimated the electron coupling between "open" corannulene and a strong electron acceptor, TCNQ. A solutionphase route was employed for preparation of two novel "open" corannulene-based derivatives with the corresponding spectroscopic analysis of their properties experimentally and theoretically. Furthermore, through a combination of theoretical modeling with experimental results, mechanistic studies were undertaken to shed light on possible factors (such as strain energy) that could act as a driving force for the observed p-bowl opening. Our studies highlight the possibility to implement novel synthetic routes for p-bowl transformations, that are drastically different from the conventional approaches toward derivatization of traditional PAHs. Thus, the presented solidstate, solution, and theoretical methodology are the frst steps Fig. 4 (a) (Top left) Single-crystal X-ray structure of X. (Bottom left) Optical transition strengths computed at the ground state optimal geometry for X in THF (blue) and at the second singlet excited state optimal geometry for X in THF (red). (Right) Energies and isosurfaces of the HONTO and LUNTO of X in the ground and the second singlet excited states. S 0 and S 2 are the ground and excited states for X of the ground state. S 0 and S 2 0 are the ground and excited state intermediates for the minimum energy geometry of the second excited singlet state. The black solid and wavy arrows indicate absorption (S 0 / S 2 ) or emission (S 2 0 / S 0 0 ) and vibrational relaxation (S 2 / S 2 0 and S 2 0 / S 2 ), respectively. The theory level is TDDFT/RPA based on the B3LYP-D3/ 6-31+G* method. (b) (Top left) Geometrically optimized structure of X 0 based on B3LYP-D3/6-31+G* level of theory. (Bottom left) Optical transition strengths computed at the ground state optimal geometry for X 0 in THF (blue) and at the second singlet excited state optimal geometry for X 0 in THF (red). (Right) Energies and isosurfaces of HONTO and LUNTO of X 0 in the ground and the second singlet excited states. S 0 and S 2 are the ground and excited states for X 0 of the ground state. S 0 ' and S 2 0 are the ground and excited state intermediates for the minimum energy geometry of the second excited singlet state. The black solid and wavy arrows indicate absorption (S 0 / S 2 ) or emission (S 2 0 / S 0 0 ) and vibrational relaxation (S 2 / S 2 0 and S 2 0 / S 2 ), respectively. The theory level is TDDFT/RPA based on the B3LYP-D3/ 6-31+G* method. toward understanding possible avenues to prepare barely accessible structures by "unlocking" the corannulene core and application of the latter for molecular electronic development.

chemsum

{"title": "\u201cBroken-hearted\u201d carbon bowl <i>via</i> electron shuttle reaction: energetics and electron coupling", "journal": "Royal Society of Chemistry (RSC)"}

virtual_screening_and_free_energy_estimation_for_identifying_mycobacterium_tuberculosis_flavoenzyme_

## Abstract: In Mycobacterium tuberculosis (MTB), the cell wall synthesis flavoenzyme decaprenylphosphoryl-β-dribose 2'-epimerase (DprE1) plays a crucial role in host pathogenesis, virulence, lethality and survival under stress. The emergence of different variants of drug resistant MTB are one of the major threats worldwide which essentially requires more effective new drug molecules with no major side effects.Here, we used structure based virtual screening of bioactive molecules from ChEMBL database targeting DprE1, having bioactive 78,713 molecules known for anti-tuberculosis activity. An extensive molecular docking, binding affinity and pharmacokinetics profile filtering results in the selection four compounds, C5 (ChEMBL2441313), C6 (ChEMBL2338605), C8 (ChEMBL441373) and C10 (ChEMBL1607606) which may explore as potential drug candidates. The obtained results were validated with thirteen known DprE1 inhibitors. We further estimated the free-binding energy, solvation and entropy terms underlying the binding properties of DprE1-ligand interactions with the implication of MD simulation, MM-GBSA, MM-PBSA and MM-3D-RISM. Interestingly, we find that C6 shows highest ΔG values (-41.28±3.51, -22.36±3.17, -10.33±5.70 kcal mol -1 ) in MM-GBSA, MM-PBSA and MM-3D-RISM assay, respectively. Whereas, the minimum ΔG scores (-35.31±3.44, -13.67±2.65, -3.40±4.06 kcal mol -1 ) observed for CT319, the inhibitor co-crystallized with DprE1. Collectively, the results demonstrated that hit-molecules C5, C6, C8 and C10 having better free binding energy and molecular affinity as compared to CT319. Thus, we proposed that selected compounds may be explored as lead molecules in MTB therapy. ## Introduction Mycobacterium tuberculosis (MTB) is a slow growing and widely spread pathogen, survive in both, intra-cellular and extracellular systems of patients, and infection may result in chronic and complex disease state. During the treatment, it can go to latency which revert to exponential growth on the immune defiance conditions of hosts . In recent years, WHO reports suggested that around 10.0 million (range, 9.0-11.1 million) individuals infected and 1.3 million (range, 1.2-1.4 million) people died from tuberculosis (TB) . Moreover, the infection of MTB is one of the major causes of death worldwide, possessing the global health crisis, especially for the immunocompromised and HIV patients . Although, the specific treatment may cure MTB, however, it requires multiple drug therapy for a longer period . Furthermore, the development of multi-and extensively-drug-resistant (MDR-TB and XDR-TB) MTB strains are the big challenges to control TB infections . In several conditions, it may turn into totally drug-resistant (TDR) tuberculosis which may worsen the condition of patients and therapy . Thus, the potential drug candidates, having minimal or no side effects are highly sought in MTB therapy . In recent years, several proteins involved in MTB survival and metabolism have been explored as potential drug targets and are progress in the drug development. During the evolution, mycobacteria have developed well-orchestrated and complex biosynthetic pathways to sustain a unique and thick cell wall which helps in maintaining the cellular integrity, survival under stress and dormancy, and eluding the host's immune systems. In MTB, the cell wall consists of the polymers of mycolyl-arabinogalactanpeptidoglycan, covalently connected with peptidoglycan and trehalose dimycolate that protects from stress, antibiotics and the hots immune systems . The flavoenzyme decaprenylphosphoryl-β-d-ribose 2'-epimerase (DprE1) involve in the biosynthesis of cell wall, plays critical role in formation of peptidoglycan-arbinogalactan-mycolic acid complex (PAM) and arabinogalactan and lipoarabinomannan (LAM) which are the essential building blocks and play crucial role in survival and host pathogenesis, virulence, and lethality. DprE1 catalyses the first stage of epimerization reaction especially in the presence of FAD, it oxidizes C2' hydroxyl site of DPR to produce the keto intermediary decaprenyl-2'-keto-D-arabinose(DPX) and then DPA is formed by using decaprenyl-phosphoryl-D-2keto-erythro-pentose reductase (DprE2) and reduced form of nicotinamide adenine dinucleotide (NADH) as a cofactor . Thus, the catalytic activity of DprE1 is one of the potential drug targets in the development of tuberculosis therapy . Recently, the benzothiazinones (BTZs) derivatives have shown higher potency for inhibition of DprE1, and efficacy against XDR and MDR mycobacterium clinical isolates. To improve the pharmacological properties of the compounds, chemical scaffold piperazine was added to BTZ. Further, the lead optimization of PBTZ derivatives results in the discovery of more potent compounds which are currently in clinical trials . In this view, several structurally distinct chemical scaffolds are in drug screening as DprE1 inhibitors. Broadly, these inhibitors can be categorized as covalent or noncovalent, distinctly involved in interaction at the catalytic domain of DprE1 . To elucidate the action and interaction of BTZs compounds, Batt el al., solved the X-ray crystal structure of DPrE1 in both, ligand free and bound form. He found that the structure of DprE1 consists of two functional domains, FAD binding domain and substrate binding domain. The co-factor was buried deeply in highly conserved FAD domain. The substrate binding extended for FAD, decorated largely with antiparallel β-strands (β10-16) and included disordered loops at surface which govern the wide and open active site. The nitroaromatic inhibitors (e.g., BTZ, VI-9376, nitroimidazole 377790) possesses nitro moiety which involved in covalent interaction at C387, whereas, the noncovalent inhibitors (e.g., TCA1, 1,4-azaindoles, pyrazolopyridones, 4-aminoquinolone piperidine amides, Ty38c) potentially inhibit the enzymatic function of DprE1 showed that hydrophobic, electrostatic, and van der Waals interactions are critical for the spatial stability of inhibitors at the active site of DprE1 . Thus, the exploration of crystal structure of DprE1 has been largely facilitated the drug discovery efforts to tend the molecules effective against MDR and XDR strains . Recent studies on the development of DprE1 inhibitors suggested a major contribution of molecular modelling, high throughput screening, docking, functional genomics and proteomics in paradigm of identifying novel chemical scaffolds as potential molecules for TB chemotherapy . Although, molecular docking programs provide the description of protein-ligand interactions. However, a better understanding of protein-ligands interactions requires an accurate description of the spatial orientation of ligands at the active site of protein, conformational dynamics of protein and active sites residues, interaction energy and molecular stability . In this view, MD simulation is an efficient and wellestablished computational method which mimics the flexible nature of bio-molecules, protein conformational changes, protein-ligand interactions, structural perturbation and provide more realistic picture with atomic details in reference to time . Moreover, the free binding energy estimation, effect of solvation and thermodynamic integration is the central focus to understand the molecular interactions which can be well achieved by the implication of MM-GBSA, MM-PBSA and MM-3D-RISM using the trajectories obtained from MD simulation [12, . In this context, we employed the structure based virtual screening for identification of promising chemical entities as DprE1 inhibitors from the ChEMBL database. We find that 78,713 small molecules at ChEMBL database suggested for the anti-mycobacterial activity. The three steps molecular docking and binding affinity estimation process lead to the selection of 10 hit-molecules. Similar procedures were applied on the selected 13 DprE1 inhibitors for the comparison of results with hit-molecules. Multiple MD simulations were performed on the DprE1 complex with hit-molecules and inhibitor (CT319) and the spatial stability of ligand molecules at active site of protein was estimated in terms of binding free energy using MM/PBSA/GBSA, and MM-3D-RISM . The extensive evaluation of pharmacokinetic profile and drug-likeness properties analyses suggested that four chemical entities, compounds C5 (ChEMBL2441313), C6 (ChEMBL2338605), C8 (ChEMBL441373) and C10 (ChEMBL1607606) may be explored as potential lead molecules for the development of promising DprE1 inhibitors in MTB therapy. ## Protein preparation The X-ray structure of DprE1 with inhibitor CT319 and cofactor FAD (PDB ID: 4FDO) was taken from the protein data bank (www.rcsb.org) . The structure of DprE1 consist of two domains, the FAD binding domain comprised with α/β folds (residues 7-196, 413-461) and another domain, substrate binding includes extended conformation and antiparallel β-sheets (residues 197-412). In the crystal structure, the spatial orientation of FAD-binding domain and residues involved in interactions were highly conserved and critical, the cofactor is deeply buried, with the isoalloxazine at the interface to the substrate-binding domain . And, the substrate-binding domain orientated towards the interface of flavin binding at centre. Thus, to prepare the protein files for molecular docking studies, the bound complex of protein with cofactor FAD was used. The other heteroatoms, co-crystallised inhibitor (CT319) and water molecules were removed. The structural regions lacking for low electron density were prepared, using Chimera tools . Protein preparation wizard of Glide was used to assign hydrogen atoms and examine the structural correctness . Finally, the optimized coordinates of DprE1 with FAD was used to carry out molecular docking and virtual screening against the selected antimycobacterial compounds from ChEMBL database. ## Ligand preparation Here, we used the anti-tuberculosis compounds taken from the publicly available chemical compounds database, ChEMBL. It consists of 2,101,843 compounds, out of these 78,713 compounds were bioactive molecules, having anti-tuberculosis activity downloaded from the ChEMBL database . After sorting of compounds based on the repetitive entries, 30,789 ChEMBL compounds were found unique which were used for the ligand preparation. The SMILES (simplified molecular-input line-entry system) strings formats of compounds were converted to 3D SDF format, missing hydrogen atoms were added, and the structures were optimized using CORINA v2.64 software package . The module, Ligprep of Schrodinger suite 2017-3 used to generate compounds with low energy 3D structures . The ionization/ tautomeric states of the selected compounds were taken care of by Epik parameters. The compounds chirality was taken from the original state. All the conformations were minimized and produced at a maximum of 32 conformations per ligand using the OPLS-2005 force field at a pH 7+ 2 . ## Virtual Screening and Scoring The structure based virtual screening of compounds against DprE1 was performed using Glide, Schrodinger, LLC . The grid box define over the active site of DprE1, having outer box size X= 30, Y= 30, Z= 30 with grid center, X=40.1971, Y= 16.829, and Z= 9.172 . The high throughput virtual screening involved three step filtering processes (i) selection of top ten percent ligand molecules using standard precision, (ii) then, docking of molecules by the extra precision mode of Glide (Glide XP) which allow the flexibility of ligands and (iii) the best-docked compounds were chosen using a Glide Emodel energy, Glide energy and Glide score function. The Glide Emodel includes a combination of Coulombic and van der Waals interaction energy, Glide score and strain energy of ligands which were used to selected lowest energy docked complexes on which the post-docking analyses were performed. Further analysis involved the re-scoring of selected docked complex using X-score v1.2.1 and the protein-ligand molecular interactions were examined using Ligplot and Discovery Studio Visualizer (Accelrys, San Diego, CA, USA). ## ADME studies The bioavailability of selected lead molecules, ADME (adsorption, distribution, metabolism and excretion) properties were calculated using module Qikprop v5.7 of Schrodinger 2018-3. ADME descriptors includes, central nervous system (CNS), molecular weight (MW), prediction of octanol/water partition coefficient (QPlogPo/w), aqueous solubility (QPlogS), IC50 value for blockage of HERG K+ channels (QPlogHERG), gut blood barrier (QPPCaco), brain/blood partition coefficient (QPlogBB), binding to human serum albumin (QPlogKhsa), Lipinski's rule of five (RO5) and percentage of human oral absorption (% of Human oral absorption). ## In-silico drug-likeness and toxicity prediction The molecular properties predictor tool, OSIRIS was used for the prediction of side effect risks of the hit-molecules, such as mutagenicity, tumorigenicity, irritant and reproductive effects. It also calculates the drug-relevant properties: cLogP, solubility (LogS), molecular weight (MW) and based on overall drug-score suggested the drug-likeness properties of molecules . ## Molecular dynamics (MD) simulation MD simulation was performed on the coordinates of DprE1 and DprE1-ligand complexes, using Amber16 biosimulation package. The force field ff14SB with TIP3P water model was used for the solvation of prepared systems. The charges, parameters and force field for cofactor (FAD) and ligands were defined by AM1-BCC charges and force field GAFF, using Antechamber tool. Here, six independent MD runs were carried out for the prepared systems, DprE1-FAD, DprE1-FAD-CT319, DprE1-FAD-ChEMBL1607606, DprE1-FAD-ChEMBL2338605, DprE1-FAD-ChEMBL2441313 and DprE1-FAD-ChEMBL441373 complexes. And, the systems were prepared using tleap tool of Amber16 with buffer distance (12 ) in the octahedral simulation box. To neutralize the system 0.15 M counter ions (Na + and Cl -) were added . Bonds involving hydrogens were treated with SHAKE algorithm and the long-range electrostatic forces were handled using Particle mesh Ewald summation. During the simulation, Berendson's barostat and Langevin thermostat were used to maintain the Pressure and temperature, respectively. The energy minimization processes involved two phases. First phase included 3000 minimization steps, which involved 2500 steps of steepest gradient and remaining 500 steps of conjugate gradient algorithm. The solute atoms were restrained (100 kcal mol -1 -2 ) and the only movements of water molecules and counter ions were allowed. The second phase of minimization included 5000 minimization steps (steepest gradient: 4500 steps and conjugate gradient: 500 used) without restraints on any atom. Minimization step followed by heating equilibration of system from 0 to 298 K with a time step of 1fs for 30 ps and consecutive equilibration run of 100 ps using time step of 2 fs with NPT ensemble. Using pmemd.cuda, the production run was performed on NPT ensemble for period of 100 ns and the time step was set to 2 fs. All files, trajectories, velocity and energy were saved at a gap of every 10 ps. The simulation trajectories were analysed using cpptraj tool available in Amber16. ## Binding free energy calculation Free energy change of a protein-ligand binding can be represented as follows: ΔG=⟨GRL⟩-⟨GR⟩-⟨GL⟩ eq. ( 2) where, ΔG, GRL, GR and GL represent the binding free energy of protein-ligand system, free energy of protein complexed with ligand, free energy of protein and free energy of ligand, respectively. The angular brackets represent ensemble average. Neglecting the entropy change of protein and ligand as a result of binding, equation (eq.) 2 can approximately be written as: where, ΔE is the interaction energy change (gas-phase) upon ligand binding. ΔGSOLV is the solvation free energy change on ligand binding. Here, ΔE can be computed using molecular-mechanics force field and the second term, ΔGSOL can be estimated with the help of a proper solvation model. Solvation models can be categorized in one of the two classes, implicit solvation model and explicit solvation model. Implicit solvation models consider the solvent molecules and dissolved salt ions as a mean field dielectric continuum. In contrary to implicit solvation models, explicit solvation models define solvent species at atomistic detail. Generalized Born Surface Area (GBSA), Poisson-Boltzmann Surface area (PBSA) are the two most common implicit solvation models used for solvation free energy calculation. Whereas, using the first principles 3D-RISM-KH (three-dimensional reference interaction site model with Kovalenko-Hirata) provides a 3D maps of solvation structure, thermodynamics and, more accurately predicting the parameters accounts for the ligands binding interactions and affinities. ## Generalized Born Surface Area (GBSA) In GBSA approach, the solvation free energy of a solute molecule, G SOL is calculated in two parts: polar or electrostatics (G SOL−GB ) and nonpolar or non-electrostatics G SOL−GB is estimated using the following generalized Born expression: where qi, qj represent charges on solute atoms i, j respectively and r ij is the distance between them, Ri represents effective Born radii (estimated using van der Waals radius and burial of atom), and the summation runs over all pairs of atoms in the solute molecule. The screening effect produced by the monovalent salt ions is incorporated in eq. 4 through the Debye-Huckel screening parameter κ. The nonpolar or non-electrostatics contribution to solvation free energy (also known as cavitation term),G SURF is determined on the basis of solvent accessible surface area (SASA) of molecules. ## Poisson Boltzmann Surface Area (PBSA) In PBSA model, the solute is represented in atomic detail with molecular mechanics force field and the solvent molecules along with dissolved electrolytes is represented as a dielectric continuum. This approach considers the solute molecule as a dielectric object whose shape is determined by the atomic coordinates and their radii. Electric charges present on atoms of solute molecule produce electric field and in response the solvent also produce a reaction field. The electrostatic potentialϕ(r)at a point satisfies the Poisson-Boltzmann (PB) equation and can be computed by solving it: Where, ε(r) is the dielectric constant, ϕ(r) is the electrostatic potential, ρ(r) is the solute charge density at position r, zi is the charge on ion i, ci is the bulk number density of ion i, k is the Boltzmann constant, and T is the absolute temperature; the summation is over all different ion types. when the solute does not carry a high charge, the second term in eq. 5 can be linearized and it results into linearized PB equation. In Amber PB equation is solved numerically for solutes of arbitrary shape which gives electrostatic potentialϕ(r)at each point of the system. Once we know the electrostatic potential ϕ(r) at each point, we can calculate the polar solvation free energy (G SOL−PB ) of the solute by multiplying each solute charge q i by electrostatic potential ϕ(r i ) at that point. Here, the nonpolar part of solvation free energy G NPOL is also calculated through SASA of solute, as described in GBSA analyses. ## Three-dimensional reference interaction site model with KH closure (3D-RISM-KH) 3DRISM is a semi analytical theory based on statistical mechanical Ornstein-Zernike (OZ) equation which is a contrary to MM/GBSA and MM/PBSA, considers molecular structure of solvent and salt ions. OZ equation splits the total correlation between two molecules into direct correlation between them and indirect correlation, which comes from other particles present in the system. In this theory, the molecular interactions are converted to sum of site-site interactions where atomic centres are taken as sites. The three-dimensional solvent distribution functions are obtained from the solution of following 3DRISM integral equation: Where, h γ (𝐫) = g γ (r − 1) and c γ (𝐫) are the total and direct correlations of solvent in 3D and summation is taken over all interaction sites of all solvent species. The susceptibility function χ αγ (r′)for solvent was calculated using dielectrically consistent reference interaction site model (DRISM) theory and used as input to 3DRISM calculation. As eq. 6 involves two variables; hand c, therefore, we need another closure relation to solve it. Here, we have used Kovalenko-Hirata (KH) closure which is as following: and g(𝐫) = 1 − βu γ (r) + h γ (r) − c γ (r) for g γ (r) > 1 (eq. 7) The one-dimensional site-site solvent susceptibility of solvent is defined in two parts as follows: Where, the intramolecular correlation function ω αγ (r) incorporates the molecular geometry of solvent and h αγ (r) is the total correlation between solvent sitesαand γ. In Amber, eqs. ( 6) and ( 7) are solved numerically to obtain three-dimensional solvent distribution functions around a fixed solute geometry and solvation free energy of the solute is calculated using the following equation which is an extension of Singer-Chandler formula: Where, Θ is the heaviside function and the summation runs over all the solvent sites i. The nonpolar part of GSOL-NPOL is calculated by assigning all solute charges zero. Further, the solvation free energy GSOL can also be decomposed into energetic and entropic components (∆GSOL-E and -∆GSOL-TS) using temperature derivatives. ## Results and Discussion The drug development process involves several expansive steps and complex strategies. Recent advancement in the computational modelling techniques, molecular docking, high-throughput virtual screening, pharmacokinetic profile (ADME), toxicity and bioavailability analyses of the molecules have been perceived as well-established techniques to accelerate the drug development processes [13,31, . Further, the integration of MD simulation and estimation of free-binding energy provide an accuracy on the spatial fitting, interaction stability and binding affinity of ligands at the active site of protein [15, 38]. Herein, we systematically utilized the structure based virtual screening of compounds from ChEMBL database, having bioactive 78,713 molecules known for anti-tuberculosis activity (Figure 1). The initial sorting of molecules leads to selection of 30,789 molecules which were subjected for molecular docking against the protein DprE1, the oxidoreductase enzyme involved in cell wall synthesis of MTB. The extensive ADME, toxicity and pharmacokinetic profile analyses were performed on hitmolecules which results in the selection of four ChEMBL compounds as potential lead molecules for DprE1 inhibitors. To improvise the molecular docking results, multiple MD simulations carried out on DprE1, DprE1-CT319 and DprE1--hit molecules (C5, C6, C8, C10) complexes. Typically, MD simulation deciphers the structural stability of protein-ligand interaction, conformational orientation, stability and molecular interactions of ligands at active site . Moreover, the obtained MD trajectories utilized to calculate MM/PBSA, MM/GBSA and 3D-RISM-KH, which provide a robust estimation of free-binding energy, contacts and effect of solvent underlying the binding affinity of ligand molecules . ## Virtual screening and docking analysis against DprE1 The structure based virtual screening was performed on 30,789 small molecules taken from ChEMBL database, having anti-tuberculosis biological activity. Glide based molecular docking involves various filtering steps for the high throughput virtual screening (HTVS) . During the initial step, docking leads to the selection of 3,078 compounds, top scored 10 % compounds. These top scored 10 % compounds are subjected for standard precision (SP) docking which results in the selection of 307 compounds. The extra precision (XP) filtering is applied on another top scored 10 % (307 molecules) hit-molecules. Finally, the best scored top 10 hit-molecules are selected for the comparative studies with known DprE1 inhibitors and drug molecules taken from the recent literatures . The 2D interactions of top 10 hit-molecules and 13 inhibitors are shown in Supplementary information S1A and S1B. The same procedures, SP followed by XP were applied for molecular docking of DprE1 inhibitors and X-score re-scoring method was applied to measure the binding affinity of molecules with DprE1 . Glide docking scores and binding affinity (X-score) of top 10 hit-molecules and DprE1 inhibitors are summarized in Table 1 and 2, respectively. The DprE1 inhibitor, CT319 shows highest docking score -5.48 kcal mol -1 , whereas, BTZ-N3 shows the lowest docking score -1.82 kcal mol -1 . However, the X-score results show highest binding affinity of BTz043 (-9.62 kcal mol -1 ) with DprE1, whereas, the lowest binding affinity found with TBA-7371 (-7.87 kcal mol -1 ). Among the compounds taken from ChEMBL database, compound ChEMBL2323138 (C1) shows highest docking score of (-10.198 kcal mol -1 ) with DprE1and the minimum docking score -8.795 kcal mol -1 is obtained for ChEMBL1607606 (C10). Whereas, the X-score results show highest binding affinity -10.74 kcal mol -1 with ChEMBL2338605 (C6) and lowest affinity -8.60 kcal mol -1 for ChEMBL1607606 (C10). ## ADME property analysis against DprE1 receptor Another filtering method involves the pharmacokinetic properties (ADME) analysis of hit-molecules . Predicting the bioavailability, toxicity and safety of compounds is an important and integral component of drug designing process . We employed QikProp v5.7 available with Schrodinger 2018-3 to analyse the ADME properties of compounds and compared with DprE1 inhibitors. Results show that compounds, C5, C6, C8 and C10 having the CNS activity with the potential range of drug molecules -1 to 0 (Table 3). All hit-molecules having molecular weight <500. The optimal range value recommended for the lipophilicity (QPlogPo/w) of compound is between -2.0 -6.5. Result shows that all 10 molecules having QPlogPo/w <6.5, however, the lowest value (0.199) is observed for C4. Whereas, the higher QPlogPo/w value 5.108 is obtained for C5. The QPlogS (potential range -6.5 -0.5) defines the aqueous solubility of compounds which are observed within the favourable range for all 10 compounds. The recommended range for predicting IC50 value for blockage of hERG K + channel is QPlogHERG <-5 which is well satisfied by all compounds. The compounds having the predicted apparent Caco-2 cell permeability test (QPlogCaco) > 500 is recommended. Out of 10 molecules, only four compounds, C5, C6, C8 and C10 having a value range > 500. The recommended range for QPlogBB is -3.0 -1.2 which is observed favourable for all molecules. Similarly, it is observed that all hit-molecules obeyed the drug likeness properties RO5 (Lipinski's rule of five) and found within the recommended range for human oral absorption (PHOA). The ADME analysis of DprE1 inhibitors shows that all 13 compounds are lying within the recommended ranges for predicted ADME descriptors (Table 4). ## In silico drug-likeness and toxicity predictions The physiochemical properties, toxicity, tumorigenicity and mutagenesis risk of the compounds are investigated by OSIRIS Property Explorer and compared with the DprE1 inhibitors (Table 5 and 6). Results show that out of 10 hit-molecules, 9 compounds are estimated as no risk for mutagenicity (MUT and tumorigenicity (TUMO), whereas, C2 shows higher risk for both MUT and TUMO test (Table 5). All 10 compounds show no risk for irritation (IR), however, reproductive development (REP) toxicity result shows high risk for C6, whereas, C9 shows medium risk for REP. The drug score (DS) is representing the combined score value of compounds solubility, polar surface area, toxicity, druglikeness and CLogP which define the overall sensitivity of drug molecules. Result shows highest DS value 0.79 for the hit-molecule, C4 and the least observed DS score (0.14) for C2. The four compounds, C5, C6, C8 and C10 which successfully cross the ADME test, show moderate DS score ranges 0.26 -0.42. The compound C8 shows higher DS score 0.42, C5, C8 and C10 predicted as no risk for toxicity parameters. Whereas, C6 predicted as high toxic risk for reproductive effect (REP), as the chemical scaffold contains ketone moiety. Table 6 shows the toxicity and drug-likeness parameters index of DprE1 inhibitors. The inhibitors, Ty38c and 4AQs predicted as higher risk for mutagenesis, whereas, VI-9376 shows medium risk and other inhibitors observed as no risk. All selected inhibitors having no risk for TUMO. Only, two inhibitors show medium risk for IR and all are predicted as no risk for REP. The predicted DS score of inhibitors ranges 0.22 -0.88. The inhibitor CT319, co-crystallized with DprE1 X-ray structure shows moderate range of DS score 0.37, however, no risk is observed for toxicity parameters. ## Molecular Interactions The crystal structure of DprE1 consists of well separated, conversed FAD-domain and the substrate binding domain. The deeply buried FAD-domain is composed of an α/β fold, formed by the residues belonging to N-terminus (residues 7-196) and C-terminus (413-461). The substrate binding domain is extended from flavin at centre to surface, orchestrated by anti-parallel β-sheets (β10-16) and helices (α5, 9 and 10). The wide-open active site of DprE1 is governed by two loops at surface which facilitate the accessibility and flexibility of ligand binding. We find that the top 10 hit-molecules (C1-C10) taken from ChEMBL shared a common interaction with active site residues which is summarized in Table 1. The 2D molecular interaction of hit-molecules C1-C10 and inhibitors at the active site of DprE1 are shown in Supplementary Figure S2 and S3. The active site amino acid residues involved in interactions with DprE1 inhibitors are tabulated in Table 2. The co-crystallized structure of DprE1-CT319 shows that the inhibitor at the active site is largely stabilized by the hydrophobic interactions. Trifluoromethyl moiety of CT319 involved in hydrophobic interaction with Lys134 and Tyr314 and the nitro-benzene is stabilized with Lys317 and Val365. The nitro (NO2)-group shows H-bond with His312 and Lys418, and the phenylethyl moiety forms hydrophobic interaction with Tyr60, Trp230, Phe320, Leu363, Val365 and FAD present in the vicinity of catalytic domain. Furthermore, the structural studies of BTZs derivative inhibitors demonstrated that the hydrophobic amino acid residues at the active site, Trp60, Gly117, His132, Gly133, Lys134, Ser228, Phe231, Tyr314, Leu317, Phe320, Gln321, Trp323, Asn324, Gln334, Q336, Leu363, Val365, Phe366, Lys367, Phe369, Asn385, Ile386, Cys387, Asp389 and Lys418 are critical for the ligand recognition. And, some inhibitors are covalently linked with C387. Molecular docking result shows that at the active site of DprE1, CT319 forms H-bind with Gln336, Asn385, Lys418, alkyl interactions with Tyr314, 𝜋alkyl interaction with Leu317, Leu363, Cys387 and the 𝜋-Sigma bond with Val365 (Figure 2). The trifluoromethyl moiety of CT319 interacting with Pro116, Gly133, Lys134 and Tyr314, whereas, the both benzene rings are stabilized with hydrophobic and van der Waals interaction interacted with residues: Tyr60, Gly117, His132, Lys134, Ser228, Phe320, Gly321, Lys367, Asp389 which is observed consistent with the co-crystal structure . show 𝜋-alkyl interaction, Cys387 engaged in 𝜋-sulfur interaction and Lys418 and Val365 are involved in 𝜋-cation and 𝜋-Sigma interactions, respectively. And, the hydrophobic and van der Waals interactions of Gly117, His132, Gly133, Lys134, Ser228, Phe320, Gly321, Gln334, Gln336, Phe369, Asn385, Asp389 provided the additional stability to C5 at the active site of DprE1 (Supplementary Figure 4). The molecular binding of C8 (1-cyclohexyl-5-oxo-N-(3-phenylphenyl)pyrrolidine-3-carboxamide) at the active site of shows H-bond interaction with Lys418, 𝜋-donor H-bonding with Tyr60, the alkyl and 𝜋-alkyl interactions with Lys134, Lys367 and Leu317, Leu363, Val365, respectively. And, the hydrophobic and van der Waals interactions with Pro116, Gly117, His132, Gly133, Ser228, Tyr314, Phe231, Phe320, Gly321, Trp323, Asn324, Gln334, Gln336, Phe366, Phe369, Asn385, Ile386, Asp389 at the active of DprE1 (Supplementary Figure 5). ## The spatial orientation of C10 (3-(3-hydroxypropyl)-7-(2-thiophen-3-ylethynyl)isochromen-1-one) shows H-bond interaction with Asn324, Cys387, the carbon hydrogen bond with Phe230 and 𝜋-alkyl interaction with Leu317, Leu363, Lys367. The thiophen moiety of C10 involve in 𝜋-sulfur interactions with His132, Phe369, whereas, Val365 and Lys418 show 𝜋-sigma interaction and 𝜋-cation interaction. And, the ligand is stabilized by the hydrophobic and van der Waals interactions of residues: Tyr60, Trp66, Gly117, Gly133, Lys134, Gly321, Glu322, Trp323, Arg325, Gln336, Asn385, Val388, Asn389 which is shown in Supplementary Figure 6. These molecular docking results suggested that apart from the conventional H-bonding at the active site of DprE1, several other interactions, 𝜋-alkyl (Leu317, Lys367), 𝜋-sulfur (Cys387), 𝜋-Sigma and the van der Waals interactions of amino residues His132, Gly133, Lys134 and Asn385 are critically involved in interactions with lead-molecules which observed consist with the co-crystalized structure of DprE1-CT319. Thus, the molecular docking results, binding affinity scores and pharmacokinetic analysis of hit-molecules suggested that compounds, C5 (docking score -9.248 kcal/mol, X-score -9.66 kcal/mol), C6 (-9.211 kcal/mol, X-score -10.74 kcal/mol), C8 (-9.106 kcal/mol, X-score -10.64 kcal/mol) and C10 (docking score -8.795 kcal/mol, X-score -8.60 kcal/mol) may be explored as promising candidates for further lead optimization as DprE1 inhibitors. ## Conformational dynamics and stability of protein-ligand complex The solvent environment around the protein influences the molecular interaction. Thus, the various interactions observed during the molecular docking may or may not exist during the simulation . To examine the conformational stability, dynamics and structural integrity of DprE1 complex with novel hit-molecules, multiple MD simulations were performed in aqueous environment for the period of 100 ns, at 300 K. The conformational dynamics of DprE1 and DprE1-CT319 during the MD simulation used as a control to elucidate the structural stability of DprE1 complexed with ChEMBL compounds. Trajectories obtained from the simulation were further used for the binding free energy estimation of molecules, using MM/GBSA, MM/PBSA and MM/3DRISM . To determine the conformational stability of DprE1 with hit-molecules, we measured all atom Cα-RMSD of protein-ligand complexes and compared the results in reference of DprE1-CT319 complex (Figure 4). Results show that the structure of DprE1 remains stable with an average change in RMSD value 3.57±0.49 . The trajectory of DprE1 archives equilibrium at ~25 ns and a continuous stable equilibrium can be seen up to 100 ns of simulation. RMSD plot of DprE1 complex with inhibitor CT319 shows that trajectory achieves equilibrium at ~15 ns and the complex structure remains stable for the remaining period of simulation with change in RMSD 2.80±0.26 . The RMSD plot of DprE1-C5 shows a continuous increase in trajectory during initial 0-35 ns. The complex structure remains stable for the period of 35-55 ns and a small drift of 0.5 is observed at 60 ns. The RMSD trajectory during 60-100 ns suggested a stabilized structure of DprE1-C5 complex for the last 40 ns of simulation. The trajectory of DprE1-C6 achieves equilibrium in 0-15 ns and remains stable till the simulation finished at 100 ns. We observed a consistent and overlapped RMSD plot of DprE1-C6 with DprE1-CT319 complex. The RMSD plot of DprE1-C8 shows initial perturbation in structure during 0-25 ns, however, reaches to equilibrium at ~30 ns, after that the conformational dynamics remains stable around RMSD 3.66±0.37 . Although the structure of DprE1-C10 achieves equilibrium earlier at ~10 ns and remains stable up to 70 ns, however, we observed structural adjustment with the drift of ~1 at 75 ns and the structural dynamics remains stable with an average change of RMSD value 3.18±0.47 . To understand the spatial stability of ligand molecules at active site of DprE1, we also calculated the time evolution plot of distance of hit-molecules and inhibitor CT319 from the centre of binding pocket, as shown in Figure 5. We observed that the average distance of CT319, C6 and C8 remains quite stable suggesting that ligand is spatially well occupied at active site and stabilized with molecular interaction, during the simulation. The compound C5 shows continuous drop down in distance during 0-40 ns and it stabilized at distance ~4 which is seen up to 100 ns. The distance plot of C10 shows fluctuating behaviour which suggested the spatial adjustment at the binding pocket, thus, we observed a small drift of 1 in RMSD plot of DprE1-C10 complex. The conformational order parameter, radius of gyration (Rg) represents structural compactness and integrity of a protein structure . The Rg plot shows that all five complexes are stabilized around average Rg value ~21-23 , suggesting that all the ligand molecules were well occupied at the binding pocket of DprE1 during the simulation (Figure 6). Similar to RMSD results, we observed slightly higher Rg value 22.02±0.13 for the ligand unbound structure of DprE1. The complex of DprE1-C6 shows lowest Rg value 21.75±0.06 and highest 22.28±0.17 for DprE1-C5, whereas, the structure of DprE1-CT319 is stabilized around Rg score 22.10±0.07 . The marginal differences in Rg value of DprE, ligand bound, and unbound structures suggested the stable interaction of novel hits molecules at the active site of protein. These results provide a clear evidence that the selected ligands are well accommodated in binding pocket, having consistent interactions with active site residues as observed during the molecular docking. We further investigated the conformational fluctuations and local dynamics through the calculation of average fluctuation of each amino acid residue of DprE1. The RMSF plot of all C α -atoms of DprE1 and docked complexes with ligands are shown in Figure 7. In this Figure 7, we can see the comparative results of DprE1, and each ligand bind complex with DprE1. Results show that the average fluctuation of residues are reduced on binding of ligands at the binding pocket of DprE1 which is suggesting a favourable molecular interaction. We observed that average fluctuation of C-terminal residues is increased on binding of CT319 and C5. The RMSF plots show higher mobility for residues 150-200, belonging to β8-10, which can be seen in CT319, C5 and C6. Whereas, DprE1-C8 and DprE1-C10 show lower average fluctuation in compared to FAD bound DprE1. However, the secondary structure analysis plots using DSSP suggested that no significant conformational changes are observed in the secondary structure, upon the binding of ligands during simulation which provide an elegance evidence of stable molecular interactions of ligands with DprE1 (Supplementary Figure 7). ## Hydrogen bond analysis The structure of protein is largely stabilized with the network of H-bond which plays critical role in the conformational adaptability, mobility and interaction with biomolecules. Apart from the structural stability, H-bond interactions are crucial in molecular recognition and protein-ligand interactions. Thus, we analysed H-bond interactions between protein and ligands, using cpptraj module of Amber with distance cut off 3.5 and angle cut off 135°. Results show that average two to four H-bond interactions are involved in DprE1 interactions with ligands (Figure 8). (Figure 8C). In C5, the H-bond formed between N atom and Asn324 is broken and new H-bonds are formed between this N-atom and H atoms at zeta position of Lys418 during the simulation (Figure 8B). The O atom present in C5 was also found to form H-bonds with H atoms at epsilon position of His132, ## Essential Dynamics We further performed essential dynamics (ED) analysis to understand the dynamics of protein-ligand complexes. ED analysis involves representation of collective motion of the most variable region of protein in terms of two principal components PC1 and PC2. The projection of each protein-ligand complex trajectory along with native protein onto two principal components PC1 and PC2 is shown in These observations thus support the idea of decrease in flexibility in the presence of DprE1 inhibitor, CT319 and compounds, C6 and C10. In the presence of CT319, DprE1 is restricted to small excursions slightly away from its initial conformation. The finding of a strong restriction in the size of the explored conformational space with only a minor reduction in RMSF which can be seen in Figure 7, indicates that local fluctuations take place but that collective motions have been compromised or more likely slowed down in the presence of CT319, C6 and C10 (Figure 9A, 9C and 9E) as compared to DprE1 complex with C5 and C8 (Figure 9B and 9D). Thus, the ED results along with H-bond interactions suggested that protein-ligands interactions remain consistent during the simulation, however, the most stable conformational dynamics is observed for CT319, C6 and C10. ## Binding free energy analysis The quantitative assessment of molecular binding interaction of DprE1 inhibitor CT319 and lead molecules C5, C6, C8 and C10 are estimated using three different methods for the molecular theory of solvation, MM/GBSA, MM/PBSA and MM/3DRISM-KH. MM/GBSA and MM/PBSA calculations are performed on the 5000 frames taken from the last 50 ns of the MD simulation . Considering the large computational cost, 100 equally spaced frames taken from the last 50 ns of simulation are used for MM/3DRISM-KH analysis. The result of these three calculations are given in Furthermore, MM/3DRISM method also gives the energetic and entropic component of the solvation free energy (Table 9) which shows that it is the entropic part of solvation free energy not energetic, which favours the protein-ligand binding for all the ligands. Thus, the all three methods do not give the same order of binding free energy for the protein-ligands, but the comparison of predicted binding free energies provide an important clue to evaluate the relative stabilities and flexibilities of compounds at the active site of DprE1. In the table 7, results of MM-GBSA analyses show the higher value of combined ΔG (-41.28±3.51 kcal mol -1 ) for compound C6, whereas, lowest estimated ΔG for CT319 (-35.31±3.44 kcal mol -1 ). The compound C5 and C10 shared the almost similar binding free energies (ΔG ~36 kcal mol -1 ) and the slightly better binding energy predicted for C8 (-40.75±3.86 kcal mol -1 ). Moreover, the binding energies estimated by MM-PBSA (Table 8) also suggested the major contribution of non-polar solvation energies in the molecular interaction of compounds at the active site of DprE1. We find that the ΔG values for lead compounds ranges -16.08 --22.36 kcal mol -1 , whereas, the estimated ΔG -13.67±2.65 kcal mol -1 for CT319. In another analysis, the molecular theory solvation, MM-RISM-KH which yields the broader picture of molecular interactions on solvation structure and implication of thermodynamics from the first principles, accounts for solvent and biomolecules to describe the relative binding affinities . Table 9 shows that binding interaction of compounds stayed stable as perceived from the molecular docking, however, results again reveal the highest ΔG value -10.33±5.70 kcal mol -1 for C6 and lowest for CT319 (ΔG value -3.40±4.06 kcal mol -1 ). Furthermore, the binding free energy approximation by all three methods suggested the larger contribution of van der Waals energies for ligands interactions and stability at the active site of DprE1. Collectively, the results demonstrated that all four hit-molecules C5, C6, C8 and C10 have better binding affinity with DprE1 as compared to inhibitor CT319 (Figure 10). Thus, the lead optimization of selected four compounds from ChEMBL chemical database may provide a new endeavour for the development of DprE1 inhibitors in MTB therapy. ## Conclusion In conclusion, we have explored structure based virtual screening for the identification of promising chemical entities as DprE1 inhibitors from ChEMBL database. Initial sorting of compounds results in the selection of 30,789 small molecules which are suggested for the anti-mycobacterial activity. The three steps molecular docking and binding affinity estimation processes lead to the selection of bioactive 10 hit-molecules. Similar procedures were applied on the selected 13 DprE1-inhibitors which were used to compare the results with hit-molecules. The extensive evaluation of pharmacokinetic profile and druglikeness properties analyses using ADME, toxicity and OSIRIS properties explorer suggested that four chemical entities, C5 (ChEMBL2441313), C6 (ChEMBL2338605), C8 (ChEMBL441373) and C10 (ChEMBL1607606) may be explored as potential candidates for the lead optimization as DprE1 inhibitors. To determine the conformational stability of hit-molecules at the active site of DprE1 in aqueous environment, multiple MD simulation were performed on the complex of DprE1 with lead molecules and inhibitor CT319. The binding free energy estimation using MM/PBSA, MM/GBSA and 3D-RISM-KH revealed that compounds C5, C6, C8 and C10 show better binding affinity as compared to DprE1 inhibitors. Thus, our comparative studies suggested that the selected compounds (C5, C6, C8 and C10) could be further investigated as novel lead molecules for the rational drug designing of DprE1inhibitors in MTB therapy.

chemsum

{"title": "Virtual Screening and Free Energy Estimation for Identifying Mycobacterium Tuberculosis Flavoenzyme DprE1 Inhibitors", "journal": "ChemRxiv"}

a_fast_approximation_for_adaptive_wavelength_selection_for_infrared_chemical_sensors_†

## Abstract: Active mid-infrared spectroscopy with tunable lasers is a leading technology for standoff detection and identification of trace chemicals. Information-theoretic optimal selection of the laser wavelength offers the promise of increased detection confidence at lower abundances and with fewer wavelengths. Reducing the number of wavelengths required enables faster detections and lowers sensor power consumption while keeping the optical power under eye safety limits. This paper presents an approximation to the mutual information which operates ∼40 000× faster than traditional techniques, thereby making near-optimal real-time sensor control computationally feasible. Application of this technique to synthetic data suggests it can reduce the number of wavelengths needed by a factor of two relative to an evenly-spaced grid, with even higher gains for chemicals with weak signatures. Active mid-infrared (MIR) spectroscopy 1 is a popular technique for detection and identification of trace chemicals (both surface residues and vapors/aerosols) at distances of up to tens of meters. The MIR reflectance spectrum can be obtained either by illuminating the target with a broadband source and dispersing the received light at the detector, or by scanning the wavelength 14 of a tunable, narrowband source such as a quantum cascade laser. 10,11 Use of a tunable source offers many advantages, including higher optical throughput (for a given level of illumination, such as dictated by eye safety), simpler detectors (no need for a Fourier-transform infrared or other complex spectrometer), and the ability to select only wavelengths which help discriminate between target chemicals (i.e., no photons need be emitted at uninformative wavelengths). A block diagram of a laser-based chemical sensor is shown in Figure 1. The active IR Figure 1: Block diagram of an active MIR chemical sensor. An active IR sensor interrogates the target with a tunable laser and reports the reflectance spectrum to the detection and control algorithm. The detection algorithm compares the measured spectrum to a library of spectral signatures, and the control algorithm selects additional wavelengths to measure until the desired level of confidence is reached. sensor comprises a tunable laser (which can be raster-scanned over the target to build up an image) and a broadband imager. The active sensor reports the reflectance spectrum, in the form of a hyperspectral image (HSI), to the detection and control algorithm. The detection algorithm compares the measured spectrum to a library of spectral signatures. If the results of the detection algorithm are inconclusive, the control algorithm selects further wavelengths to measure. This process repeats until the desired detection confidence is obtained. Information theory is often used for feature selection in the machine learning community, 19 and has been applied to the optimization of various chemical sensors, the characterization of spectral variability in hyperspectral images, 23 and the characterization of limits of detection in chemical sensors. 24 In particular, previous work has presented techniques for information-theoretic optimal wavelength selection (OWLS). 25 Specifically, given candidate wavelengths Ω, OWLS seeks where R(Λ) is the reflectance at the wavelengths Λ, Y is the identity of the chemical, and is the mutual information between R and Y . 26 The wavelengths selected with this scheme are optimal in the sense that maximizing I(R; Y ) minimizes the mis-classification rate. 27 The simplest way to use OWLS is a priori: the sequence of wavelengths which is expected to deliver the best accuracy is selected offline, before any data have been measured. It is desirable, however, to select the next wavelength(s) to measure adaptively based on the data which have been measured so far, as described in Algorithm 1. For offline use, I(R; Y ) is estimated at each step (line 6) using the Kozachenko-Leonenko (KL) estimator. 28 Computing the first 10 wavelengths (from 500 candidates) to optimally discriminate between 67 chemicals takes six minutes on a typical laptop, and is therefore infeasible for real-time, adaptive wavelength selection. Instead, this paper presents a geometric approximation which exploits the structure of the signature model to deliver comparable results in under 10 ms: a 40 000× speedup. ## Description of the Algorithm In order to enable real-time adaptive OWLS, we need an algorithm which has the following properties: Algorithm 1 Adaptive OWLS (Algorithm 2 of Ref. 25) 1: function AdaptiveOWLSDetection(Λ, Ω, α, q) Λ is the initial set of wavelengths Ω is the set of candidate wavelengths (Λ ∩ Ω = ∅) α is the required level of detection confidence q is the quantile of chemicals to retain at each step Ŷ ← {y | s y ≥ q(s)} Ŷ are the most likely chemicals 6: return arg max s Return chemical with highest detection score • Selects wavelengths which contain comparable mutual information to those selected by more rigorous means (such as the KL estimator). • Can incrementally add new wavelengths to the existing set Λ. • Can select wavelengths in less time than it would take to simply measure all of the candidate wavelengths. (Current systems can measure as fast as 1 ms per wavelength. 13 ) The geometric approximation is inspired by the structure of the signature model illustrated in Figure 2: the cluster corresponding to each chemical forms a line emanating from a single point (the reflectance of the bare substrate). For double pass absorption (such as from a thin film on a metallic surface), the linear relationship is exact in the absorbance domain (i.e., the logarithm of reflectance). For more general cases, this linear behavior is a reasonable approximation for low abundances (where the signal-to-noise ratio (SNR) is low and the most benefit can be gleaned from OWLS). Therefore, each chemical can be thought of as corresponding to a vector which begins at the point corresponding to bare substrate and ends at the point corresponding to a typical abundance. A possible way to pick informative wavelengths is then to find the set Λ which maximizes the squared Euclidean distance between the endpoints of all pairs of chemicals a and b: where R a (λ) is the reflectance at wavelength λ for a typical concentration of chemical a. Simply picking wavelengths which maximize the sum of this quantity over all pairs of chemicals will not provide a good approximation to the wavelengths selected using an actual calculation of mutual information, however: given an initial set of wavelengths, the next wavelength should be the one which contributes the most to separating the chemical pairs which are not already well-separated. Therefore, when assessing the potential gain from adding a wavelength, the objective function should include some form of discounting to reduce the weight of gains which come from chemical pairs which are already well-separated. This suggests a function of the form which has two parameters: • δ ≥ 0 sets how strongly gains for chemical pairs which already have large separation are discounted. Setting δ = 0 ignores how well the current set Λ already separates any given pair of chemicals. • p selects how the discounted gains for each chemical pair are aggregated. The extremes are p = 1 (sum of discounted gains) and p = ∞ (maximum of discounted gains), with intermediate values interpolating between the two behaviors. The full procedure for geometric real-time optimal wavelength selection (GROWLS) is shown in Algorithm 2. In practice, this function replaces the computation of I(R; Y ) on line 6 of Algorithm 1. ## Parameter Tuning In order for GROWLS to be effective, the parameters δ and p must be tuned so that it best approximates the behavior of the KL estimator. One way of selecting the parameters a priori is to minimize the sum of the difference in mutual information captured by the KL estimator and GROWLS: where k = |Λ| is the number of wavelengths selected. Figure 3(a) shows ∆ as a function of δ and p for k = 50. For very low δ, too many wavelengths which only help easily-separable Algorithm 2 Geometric Real-Time Optimal Wavelength Selection (GROWLS) 1: function GROWLS(Ω, δ, p, k) Ω is the set of candidate wavelengths δ is the discount exponent p is the norm order k ≤ |Ω| is the number of wavelengths to select 2: Λ ← {λ new } 4: while |Λ| < k do 6: return Λ chemical pairs are selected, which results in performance which is worse than using evenlyspaced wavenumbers. For all three values of p there is a broad minimum near δ = 1.6. There is only a weak dependence on p, but the minimum for p = 1 is slightly lower than for p = 2 and p = ∞. Therefore, the parameters p = 1, δ = 1.6 were used for the remainder of this work. Figure 3(b) shows the trajectories of mutual information accumulation for evenly-spaced wavenumbers, the optimal wavelengths computed using the KL estimator, and the approximate wavelengths computed using GROWLS. The geometric approximation nearly matches the performance of the KL estimator, indicating that this scheme should deliver comparable detection performance to the full mutual information calculation. Figure 3(c) shows the spectral library of 67 chemicals together with the wavelengths selected by the three techniques. Both the KL estimator and GROWLS pick wavelengths which are focused on informative parts of the spectrum, ignoring the uninformative regions that the evenly-spaced scheme samples. In order to illustrate the utility of adaptive OWLS using GROWLS ("adaptive GROWLS"), we generated noisy synthetic data corresponding to liquid films on metallic substrates and assessed the chemical identification accuracy. As a simplification to the industry-standard adaptive cosine estimator (ACE), 15 we used cosine similarity to match the simulated absorbance data a to the library signatures a y : Figure 4(a) shows the contours corresponding to 90% chemical identification accuracy for evenly-spaced wavenumbers, a priori OWLS (i.e., wavelengths selected a priori using the KL estimator), and adaptive GROWLS. For this example, six wavelengths were added at each step and GROWLS used q = 0.8 (i.e., the top 20% of chemicals were retained at each iteration of Algorithm 1), p = 1, δ = 1.6. The two OWLS schemes deliver comparable performance to each other, and require roughly half the number of wavelengths as the evenly-spaced scheme for moderate abundances. The time to reach a detection of a given confidence is given by where t d is the time-to-detection, N b is the number of batches of N λ wavelengths which are measured, t m is the time to measure a single wavelength (∼1 ms), t c is the time to run the detection/identification algorithm to determine the detection confidence (and, for adaptive GROWLS, to rank the chemicals) after each batch (∼10 µs for cosine similarity with 67 candidates), and t s is the time to select the next N λ wavelengths (∼2 ms for GROWLS with Despite the additional computational overhead, adaptive GROWLS delivers comparable per-Figure 4: (a) Contours corresponding to 90% accuracy, averaged over all 67 chemicals in the library. Adaptive GROWLS delivers comparable average performance to a priori OWLS, and both OWLS schemes out-perform the evenly-spaced wavenumbers, thereby permitting successful detection at lower concentrations while using fewer wavelengths. (b) Time-todetection when 10 µg/cm 2 of the target is present, averaged over all 67 chemicals. Despite the additional computational overhead, adaptive GROWLS delivers accuracy above 90% within about 20 ms, comparable to the performance of a priori OWLS. formance to a priori OWLS: both OWLS schemes reach 90% accuracy in less than half the time it takes evenly-spaced wavenumbers to reach this point. The comparable performance of a priori OWLS and adaptive GROWLS is expected at moderate concentrations: we are adding six wavelengths at each step, but often a few dozen wavelengths are sufficient at moderate concentrations, so adaptive GROWLS does not have much of a chance to make a difference. As noted previously, however, adaptive GROWLS makes a substantial difference for chemicals with low absorbance. 25 This is illustrated by Figure 5, which shows the detection performance when the target is 10 µg/cm 2 of benzene. Adaptive GROWLS obtains an accuracy of 90% after just 15 ms, compared to 30 ms for a priori OWLS and 85 ms for evenly-spaced wavenumbers: adaptive GROWLS is able to substantially reduce the time-to-detection in this case. This paper has presented a geometric approximation to mutual information which runs approximately 40 000× faster than the standard KL estimator, thereby enabling real-time adaptive selection of wavelengths to optimize detection accuracy. It was shown that, for moderate abundances, this approach halves the number wavelengths needed to reach an identification accuracy of 90%. For weak absorbers, the gains are even greater: benzene can be identified 6× faster, even after accounting for the overhead to compute the next batch of wavelengths. These simulated results indicate that adaptive GROWLS can enable sensors with tunable sources to obtain more rapid detections. Furthermore, the approximation presented here is applicable to any situation where the classes have the general structure illustrated in Figure 2, enabling efficient feature selection in a wide variety of machine learning contexts.

chemsum

{"title": "A Fast Approximation for Adaptive Wavelength Selection for Infrared Chemical Sensors \u2020", "journal": "ChemRxiv"}

understanding_how_lewis_acids_dope_organic_semiconductors:_a_“complex”_story

## Abstract: We report on computational studies of the potential of three borane Lewis acids (LAs) (B(C 6 F 5 ) 3 (BCF), BF 3 , and BBr 3 ) to form stable adducts and/or to generate positive polarons with three different semiconducting p-conjugated polymers (PFPT, PCPDTPT and PCPDTBT). Density functional theory (DFT) and timedependent DFT (TD-DFT) calculations based on range-separated hybrid (RSH) functionals provide insight into changes in the electronic structure and optical properties upon adduct formation between LAs and the two polymers containing pyridine moieties, PFPT and PCPDTPT, unravelling the complex interplay between partial hybridization, charge transfer and changes in the polymer backbone conformation. We then assess the potential of BCF to induce p-doping in PCPDTBT, which does not contain pyridine groups, by computing the energetics of various reaction mechanisms proposed in the literature. We find that reaction of BCF(OH 2 ) to form protonated PCPDTBT and [BCF(OH)] À , followed by electron transfer from a pristine to a protonated PCPDTBT chain is highly endergonic, and thus unlikely at low doping concentration. The theoretical and experimental data can, however, be reconciled if one considers the formation of [BCF(OH)BCF] À or [BCF(OH)(OH 2 )BCF] À counterions rather than [BCF(OH)] À and invokes subsequent reactions resulting in the elimination of H 2 . ## Introduction Molecular doping is a paramount topic in the organic semiconductor community, where it can enhance charge-carrier density and therefore electrical conductivity, improve charge injection and lower contact resistance, or increase charge mobility thanks by flling traps. The most straightforward approach to p-or n-doping is to use simple one-electron oxidants or reductants that react with the semiconductor to generate radical cations or anions (positive or negative polarons). A less intuitive approach to doping involves Lewis acids (LAs), notably tris(pentafluorophenyl)borane (BCF). Depending on the nature of the semiconducting polymers, LAs either effectively act as p-dopants or form Lewis Acid-Base (LAB) adducts. 7 The aim of this computational study is to give insight into these two types of reactivity. A decade ago, it was demonstrated that LAs can form physical complexes with semiconducting p-conjugated polymers, 8 a process driven by the interaction between the empty p-orbitals of the centrally electrophilic boron atom in the LA and the electron lone pair of a Lewis base (LB) site on the polymer, such as a pyridyl nitrogen. The formation of a new stable covalent bond yields a LAB adduct with a specifc fngerprint in optical absorption 9 and increased charge carrier density with respect to the unbound polymer, representing a means of postsynthetic engineering. 13 More specifcally, alternating donoracceptor conjugated copolymers, where the acceptor moiety is pyridylthiadiazole (PT), are able to strongly coordinate LAs, such as BCF, likely resulting in partial ground-state charge transfer (CT). The interaction with BCF has been shown to translate into a red-shifted onset in optical absorption of the organic semiconductor by $0.3 eV, a shift primarily due to the effect of the electron-withdrawing LA moiety on the electron affinity in presence of the LA itself. 13 Rather unexpectedly, BCF can also act as an apparent oxidant. Indeed, in the late 1990s Doerrer and Green 14 demonstrated that BCFeither when used intentionally as its 1 : 1 water complex BCF(OH 2 ), which is a strong Brønsted acid, or in the presence of adventitious watercan behave as a strong oxidant, converting metallocenes (MCp 2 , M ¼ Fe, Cr, Co) to the corresponding MCp 2 + . They considered that oxidation likely proceeded by protonation of MCp 2 by BCF(OH 2 ), followed by elimination of H 2 from two MCp 2 H + ions. Interestingly, the products they obtained did not contain the simple [BCF(OH)] anion (which is known and crystallographically characterized in other contexts 15 ), but rather either [BCF(OH)BCF] or [BCF(OH)(OH 2 )BCF] anions. More recently, the oxidizing characteristics of BCF have been rediscovered in the context of the p-doping of organic semiconductors. 16 BCF behaves as a strong oxidant, consistent with the fndings of Doerrer and Green, but inconsistent with a simple one-electron transfer from polymer to BCF. It has been observed that BCF is reduced to the unstable radical anion at ca. 1.7 to 1.8 V versus ferrocene, 17 whereas polymers that have been doped by BCF are oxidized at potentials comparable to, or more positive than, ferrocene, indicating that such an electron transfer would be highly endergonic. Thus, BCF(OH 2 ), or other BCF(OH 2 ) n adducts, which are strong Brønsted acids and are formed by the hygroscopic BCF (unless water is scrupulously excluded), are thought to be the likely oxidant, if not by a direct one-electron transfer manner. In some cases, the use of BCF may be desirable relative to the very widely used 2,3,5,6-tetrafluoro-7,7,8,8tetracyanoquinodimethane (F 4 TCNQ), due to its solubility in organic solvents, its lower volatility, and its ability to dope molecular materials with a relative high ionization potential ($5.8 eV). 11,12,18 On the other hand, other p-dopants that act as clean one-electron-oxidants may be more predictable in their behaviour as a consequence of their more straightforward chemistry. 19,20 In any case, Yan et al. have successfully used BCF as molecular dopant in a donor:acceptor planar heterojunction device structure and found that LA doping plays a synergistic role in changing the opto-electronic properties and nanomorphology of the blends leading to improved device performances, even at low doping concentration. Consistent with the work of Doerrer and Green, 14 it has been suggested that some particular polymers like poly-cyclopentadithiophene-benzothiadiazole (PCPDTBT) can be also oxidized by BCF(OH 2 ) via an initial protonation step of the cyclopentadithiophene (CPDT) unit in the polymer backbone. In ref. 16 it was proposed that the resulting protonated, positively charged, polymer chain would undergo an increase in electron affinity (compared to the pristine polymer) large enough to prompt an electron transfer from another, pristine, polymer chain (or chain section), resulting in the presence of two radical species, i.e., a neutral "protonated radical" and a radical cation (positive polaron). Continuous-wave electronnuclear double resonance (ENDOR) spectroscopy affords a spectrum that is consistent with the presence of both radicals; specifcally, a structureless spectrum is observed similar to what is expected for the "protonated radical", while the polaron is expected to contribute a much less intense structured pattern. However, in a later work on p-doping of poly(3-hexylthiophene) (P3HT), Arvind et al. could observe only the radical cation using high-resolution electron paramagnetic resonance (EPR) spectroscopy, suggesting either the "protonated radical" does not form or that it is unstable against further chemical reactions. 24 In particular, H 2 elimination, as previously invoked in the contexts of both metallocene oxidation by BCF(OH 2 ) and spiro-OMeTAD p-doping by HN(SO 2 CF 3 ) 2 (another strong Brønsted acid), 25 has been suggested to play a paramount role, but to our knowledge formation of H2 has yet to be observed directly. A comprehensive description of how LAs interact with semiconducting p-conjugated polymers is currently lacking. Here, we report on state-of-the-art calculations investigating the potential of three boron-based LAs to either form physical complexes or undergo chemical reactions involving oneelectron oxidation of the semiconductor with three different p-conjugated polymers (Fig. 1). Using density functional theory (DFT) and time-dependent DFT (TD-DFT) calculations based on optimally tuned (OT) range-separated hybrid (RSH) functionals, 26,27 we frst analyse the structural, energetics, and optical signature of ground-state complexes formed between three LAs and poly-fluorene-pyridylthiadiazole (PFPT) and poly-cyclopentadithiophene-pyridylthiadiazole (PCPDTPT) tetramers, fnding good agreement with experiment and highlighting the factors affecting the changes in optical absorption. Though there is clear experimental evidence that LAs are able to dope some polymer semiconductors, the mechanistic aspects of the doping have not been elaborated yet. We thus move on in investigating the doping mechanisms of a PCPDTBT tetramer by BCF(OH 2 ) from frst-principles. This involves identifying the most likely protonation sites and assessing the energetics of previously proposed reactions. Our results show that those are highly endergonic, mostly due to the thermodynamically unfavourable protonation to form [BCF(OH)] , thus ruling out all proposed mechanistic scenarios proposed in the literature. Capitalizing on the seminal work by Doerrer and Green, we instead consider reactions leading to the formation of larger complex anions, as observed in the context of metallocene oxidation. 14 Remarkably, we then fnd that the resulting protonated PCPDTBT chains can undergo moderately endergonic reactions when eliminating H 2 to produce a single spincarrying charged species. ## Methods Gas-phase ground-state equilibrium geometries of two representative tetramers, PFPT and PCPDTPT, were obtained by performing DFT optimization at the RSH functional level of theory, using the exchange-correlation uB97X-D functional 28 and the 6-31G(d,p) split-valence Pople's basis set for all the atomic species. The tetramers containing the PT moiety were optimized as an alternating copolymer of formula H-(-A-B-) 4 -H considering the regiochemical alternation between successive PT groups. For the sake of simplicity and to speed up the calculations, the alkyl chains were substituted with methyl groups in all investigated tetramers, a licit procedure as recently shown in the literature. 29 The same level of theory was used for all the structural optimizations in gas-phase when we introduced the three different LAs to form the LAB adducts with the tetramer PFPT and PCPDTPT. We also checked the influence of the OT range separation parameter u on the resulting optimized structures. 30 Using a RSH functional often comes along with a non-empirical tuning of u. In fact, for each specifc N-electron system, an optimal value of u can be found by enforcing the exchange-correlation functional to obey the DFT version of Koopman's theorem by aligning the negative energy of the HOMO with the gas-phase vertical IP. In practice, one computes the total energy difference between the N-electron and the (N-1)electron system and tries to minimize the overall error by minimizing the following target function: In addition, for a better description of the fundamental gap, the gas-phase vertical EA of the N-electron system can be represented by the vertical IP of the (N+1)-electron system, barring relaxation effects. In order to perform a gap tuning procedure, the modifed target function to minimize is the following: By doing that, the difference between the HOMO and LUMO energies of the N-electron systems in OT-RSH functionals provides a good approximation to the fundamental gap, that is the difference between IP and EA. In tuning the u value, we resorted to a polarizable continuum model 37 (PCM) using a screening dielectric constant of 3 ¼ 5.0, with the role of solvent polarity being addressed elsewhere. 29 With this caveat, from now on, we will refer to the highest occupied molecular orbital (HOMO) negative energy as the vertical ionization potential (IP) of the molecule and to the lowest unoccupied molecular orbital (LUMO) negative energy as its vertical electron affinity (EA). For the neat PFPT and PCPDTPT tetramer and their relatives LAB adducts, the absorption spectra were computed with full TD-DFT calculations and a ground-state population analysis was performed by means of the Charge Model 5 (CM5), 38 at the OT-RSH + PCM level of theory. In order to identify the most likely protonation site by mimicking the protonation mediated by a Brønsted acid of the PCPDTBT tetramer, we modelled in a frst place a CPDT-BT-CPDT unit (see sketch in Table 3). The pristine and protonated model moieties were tightly optimized in gas-phase at the uB97X-D/6-31G(d,p) level of theory. Proton affinity (P(A)) is defned as the negative of the protonation reaction enthalpy at room temperature (T ¼ 298 K): where DZPE is the corrected zero-point vibrational energy (ZPE) of the normal modes, DH 0 elec is the variation in the electronic enthalpy going from the pristine to the protonated model moiety and R is the ideal gas constant. Then, in order to evaluate the thermodynamic properties of all the reactions presented below, each molecule was tightly optimized at the uB97X-D/6-31G(d,p) level of theory in conjunction with PCM and 3 ¼ 5.0. The 3N 6 frequencies of the vibrational normal modes (all checked to be positive) were computed and scaled by 0.949 in order to correct for anharmonicity effects. 39 In a given reaction, the Gibbs free energy difference DG 0 reads: where DH 0 is the enthalpy and DS 0 is the entropy, both Tdependent. Moreover, each contribution can be decomposed in an electronic and a vibrational term (neglecting the rotational and translational ones, as they are not expected to contribute signifcantly), so that: DS 0 (T) Within the harmonic approximation, the vibrational enthalpy H 0 vib (T) and the vibrational entropy S 0 vib (T) can be computed as: where n i is the frequency of the i-th normal mode, h is the Planck constant, k B is the Boltzmann constant and both the sums run over the 3N 6 normal modes. The electronic enthalpy H 0 elec is directly computed at the DFT level, while the electronic entropy S 0 elec can be estimated as: where S is the spin multiplicity. Here we present reactions at room temperature that involve neutral (S ¼ 0) and radical (S ¼ 1/2) species: thus, only the latter have an electronic entropic contribution. In each investigated reaction, its DG 0 was computed as an energy difference between the products and the reactants, by calculating the enthalpic and entropic contribution of each species separately. DFT and TD-DFT calculations were performed using the GAUSSIAN16 package, 40 ## Results and discussion The optimized pristine PFPT oligomer shows a rather twisted structure. Due to the steric repulsion experienced by the nearest hydrogen atoms in the fluorene group and the -CH side of the PT moiety (see Fig. S1 and Table S1 in ESI ‡), the dihedral angles between these two groups are 39 , while the lower steric bulk on the N-bearing side of the PT results in a smaller PT/fluorene dihedral angles of 17-19 . Irrespective of its nature, the addition of one LA borane molecule with the boron atom in front of the pyridyl nitrogen in the PT group increases the dihedral angle up to 49-52 , while the other dihedrals further away from the LA remain unaltered. Gas-phase LAB adduct binding energies were estimated for the three LAs as total energy differences between the adduct coordinated with a LA and the sum of the isolated neat oligomer and LA molecule. The calculated binding energies prove the higher affinity of BBr 3 (29.5 kcal mol 1 ), followed by BCF and BF 3 (22.7 kcal mol 1 and 21.3 kcal mol 1 , respectively), in line with previous theoretical and experimental works. 43,44 The vertical IP and EA values of the neat PFPT oligomer and the corresponding adducts are reported in Table 1 (see also Fig. 2a). A clear stabilization of the charge-transport energy levels is observed in presence of LAs, i.e., both the IP and EA of the LAB adducts are increased. These changes are asymmetric, with a larger impact on EA than IP, resulting in a lowering of the transport gap, E gap . In the case of BCF, the IP increases by 0.14 eV and the EA by 0.39 eV, for an overall reduction in E gap of 0.25 eV. The changes in IP and EA are mostly driven by the partial ground-state CT taking place from the PT group to the LA, with changes across the series BBr 3 , BCF and BF 3 also reflecting various degrees of hybridization of the unoccupied electronic levels (see ESI and Fig. S2 ‡). The predicted $0.1 eV change in IP upon complexation with BCF agrees with ultraviolet photoemission data. 13 TD-DFT calculations (Fig. 2b, Table 1) indicate the emergence of a new, red-shifted, optical absorption band upon complexation. 45 As detailed below, the additional optical feature at wavelengths above 600 nm directly reflects the section of the polymer backbone interacting with the LA, with regions spatially away from the contact points contributing to the feature that is seen at $520-550 nm, slightly blue-shifted from that of the neat oligomer. We observe the largest red-shift of the lowest electronic excitation for BBr 3 (0.30 eV), followed by BCF (0.23 eV) and BF 3 (0.15 eV). The predicted red-shift (by 0.23 eV) of the lowest electronic transition is in excellent agreement with experimental optical absorption at 1 molar equivalent and above of BCF, showing a $0.3 eV red-shift of the maximum absorption peak in both flm and solution. 13 Natural transition orbitals (NTOs) pertaining to the lowest electronic excitation of the neat oligomer and the adduct with BCF are reported in Fig. 2c. In the neat PFPT oligomer, the hole density is delocalized over the entire molecular backbone, but the electron density has larger weights on the PT electron-accepting units (with dominant contributions on the two inner rings), consistent with the lowest excited state having signifcant intramolecular CT character. When BCF binds a pyridyl nitrogen on the PT group, the hole density distribution remains essentially unaltered (despite the slight increase in IP relative to the pristine oligomer), and the electron density is now fully confned to the PT moiety that is in direct interaction with the LA (as this PT unit is now electron poorer and has higher EA). The lowest electronic excitation NTOs of the adduct with BF 3 and BBr 3 are shown in Fig. S3 in ESI. ‡ In order to assess the influence of polymer chain length and its potential impact on the nature of the optical excitations, 46 we also modelled a neat PFPT octamer and its LAB adduct with BCF (see Table S2 and Fig. S4 in ESI ‡). By doubling the molecular length, we note that E gap is only slightly reduced (by $0.1 eV), mainly due to a destabilization of the IP. Irrespective of the conjugation length, the lowest electronic transition of the LAB adduct is red-shifted by 0.20 eV compared to the neat polymer chain. We performed the same analysis for another donor-acceptor oligomer, PCPDTPT, differing from PFPT by the nature of the electron-donating units (see Fig. S5 and Table S3 in ESI ‡). In contrast to PFPT, the PCPDTPT oligomer has a perfectly planar backbone with all dihedral angles equal to 0 in the pristine form. However, the addition of a LA molecule dramatically distorts the structure of the oligomer because of steric effects: the bulkier the LA, the higher the degree of distortion. In particular, the dihedral angle between the LA-bound side of the PT and the CPDT moiety reaches 112 (almost orthogonal orientation) in the adduct formed with BCF, 46 with BBr 3 and 39 with BF 3 . We stress that these substantial changes in the conformation of the molecular backbone are expected to strongly perturb the optical properties of the LAB adduct, as a result of the reduced p-conjugation. A similar effect was also observed by Schier et al. 47 for a quarterthiophene (4T) doped by BCF, with the presence of the LA interacting with the oligomer inducing substantial structural distortions. The calculated IP and EA values of the neat PCPDTPT and its respective LAB adducts, reported in Table 2 and Fig. 3a, show that, upon binding, there is an effective decrease in E gap . However, this effect is far less pronounced than for the PFPT oligomer, with the largest lowering of E gap being 0.14 eV in the case of BBr 3 (versus 0.35 eV for PFPT:BBr 3 ). As in the PFPT case, the IP, EA and E gap values are dictated by a partial ground-state CT and orbital hybridization in the LUMO of the adduct (see ESI and Fig. S6 ‡). We attribute the reduced spectral change to a competition to the opposing effects exerted by electronic CT and hybridization (which tend to reduce the gap) and conformational distortions away from planarity (which tend to increase the gap). TD-DFT optical absorption spectra in Fig. 3b (see also Table 2) show that the formation of the LAB adduct is accompanied by the appearance of a new, red-shifted, optical transition fngerprint, as in the PFPT case. The largest red-shift is predicted for BBr 3 (0.11 eV), followed by BCF (0.07 eV) and BF 3 (0.04 eV), following the trend of the calculated E gap values and similar to what reported above for PFPT. We also note that optical absorption measurements on PCPDTPT:BCF thin flms point to a larger spectral shift (reaching almost 0.4 eV) 16 than predicted, a discrepancy that could arise from conformational restraints in the solid-state (see Fig. In contrast to the previous two tetramers that were investigated, PCPDTBT does not undergo any binding reaction with LAs, 16 as the benzothiadiazole (BT) moiety lacks a pyridyl nitrogen able to share an electron lone pair with the empty boron p-orbital of the LA. Instead, adding BCF to a PCPDTBT based flm leads to an increase in electrical conductivity and to the formation of positive polarons, i.e., molecular pdoping. 16,24,48 As in the mechanism proposed by Doerrer and Green for oxidation of metallocenes, 14 Yurash et al. suggested that the frst step of this p-doping was the protonation by the highly Brønsted acidic complex BCF(OH 2 ) of the CPDT moiety of the polymer backbone. 16 They further proposed that protonation would increase the EA sufficiently that a nearby neutral chain segment would be able to transfer an electron to the (positively charged) protonated segment (with the segments belonging either to the same or different physical polymer chains, if the process is intrachain or interchain, respectively). This mechanism results in the formation of two radical species: a neutral, "protonated radical" and a radical cation, as shown in Scheme 1: The optimized PCPDTBT structure in PCM yields a slightly twisted backbone, with all the dihedral angles of about 20 (see Fig. S9 and Table S5 in ESI ‡). In an attempt to identify the most likely protonation site along the polymer backbone, we performed P(A) calculations. The results reported in Table 3 show that (in contrast to ref. 16 in which position 3 was assumed to be protonated) position 1 (an a-carbon atom) in the CPDT moiety is the most favorable site to be protonated, followed by position 3 (a b-carbon atom) and 2. As a result, the DH 0 elec penalty for the protonation step is signifcantly overestimated in the modeling work by Yurash et al. compared to the value reported here (+40.4 kcal mol 1 in ref. 16 versus +22.9 kcal mol 1 here). The addition of one proton (or hydrogen atom) to position 1 on the CPDT group dramatically affects the polymer backbone planarity since it breaks the p-conjugation by introducing sp 3 carbon atoms and the oligomer becomes quite twisted. By computing the thermodynamic properties of all the species (i.e., proposed reactants, intermediates and products) involved in the above reactions, our calculations show that both the protonation and the one electron-transfer processes are substantially endergonic, with DG 0 values of +23.0 and +13.1 kcal mol 1 , respectively (see Scheme 1), implying a total DG 0 of +36.1 kcal mol 1 (or +1.57 eV), thus suggesting the overall reaction to be very unlikely. In a recent study by Arvind et al. 24 on P3HT, EPR measurements performed on BCF-doped samples revealed the formation of free radical cations on the polymer backbone, yet showing no indication for the presence of another radical species (i.e., associated with the "protonated radical"). If BCF doping of PCPDTBT proceeds in analogous fashion to that proposed for the BCF-induced doping of P3HT by Arvind et al. the overall reaction would be that shown in Scheme 2: The computed DG 0 value for the overall reaction is +31.5 kcal mol 1 , smaller than that for Scheme 1, but still highly endergonic. As shown in Scheme 3, several possible pathways might lead to the same overall reaction as that shown in Scheme 2: In scenario A, the protonation step is followed by electron transfer (as in Scheme 1), but here two neutral "protonated radicals" subsequently react to eliminate H 2 to regenerate two neutral closed-shell polymers (shown for one such radical affording half a molecule of H 2 ), contributing a negative (exergonic) DG 0 ¼ 4.6 kcal mol 1 (or 0.20 eV). Scenario B is a variant of scenario A where H 2 is eliminated from two protonated cationic polymers, contributing with a DG 0 ¼ +8.5 (13.1-4.6) kcal mol 1 (or +0.37 eV). Finally, scenario C is a combination of scenarios A and B, leading, as expected, to a twofold increase in the total DG 0 ¼ +63.0 (2 31.5) kcal mol 1 . We note that reactions of the type shown in Schemes 2 and 3 (and the similar overall reactions involving larger counter-ions that are discussed in the following section) are apparently at odds with the CW ENDOR results of ref. 16. However, although the structureless feature is consistent with that expected for the "protonated radical", it could also in principle arise from dynamic effects leading to loss of the structure expected for the polaron signal, or even from other radicals formed through side reactions. We also reckon that, as observed elsewhere in the literature, 24, the polymer conjugation length plays a paramount role in the context of molecular doping, since different mechanisms might occur depending on the extension of the polymer backbone. To address this point, Table S6 in ESI ‡ reports the computed DH 0 elec values pertaining to Scheme 1 and Scheme 2, using either a PCPDTBT tetramer or an octamer as representative model. The computed Scheme 1 Reaction mechanism similar to that proposed by Yurash et al., involving a protonation followed by an electron-transfer reaction (this mechanism differs from that in ref. 16 in the position of the protonated site, see below). Calculations reported here yield DG 0 ¼ +23.0 kcal mol 1 (or +1.00 eV) for the protonation and DG 0 ¼ +13.1 kcal mol 1 (or +0.57 eV) for the electron transfer. For the sake of simplicity, the distinct structures are shown for single tetramer repeat unit, while we acknowledge that both spin and charge will be delocalized over multiple repeat units to varying extents. DH 0 elec values are found to be comparable, which comforts our choice of tetramer models as providing a good trade-off between accuracy and computational cost. Neither the overall reactions of Scheme 1 nor Scheme 2 appear likely to represent the mechanism responsible for the formation of excess charge carriers in PCPDTBT upon LA doping, since the overall reactions are highly endergonic, with a particularly high energy penalty being associated with the protonation of the pristine polymer chains by BCF(OH 2 ) complex with concomitant formation of [BCF(OH)] . However, in the previous work on metallocene oxidation by BCF(OH 2 ), 14 [BCF(OH)] was not observed, but rather [BCF(OH)(OH 2 )BCF] (Scheme 4), in which [BCF(OH)] is hydrogen bonded to another BCF(OH 2 ) complex, and (Scheme 5), where [BCF(OH)] coordinates a BCF molecule. We reconsidered, therefore, Scheme 2 based on that proposed by Arvind et al. for P3HT, but now forming anions containing two BCF units of the two types observed by Doerrer and Green: If we consider the protonation reaction as forming the "fourbody" [BCF(OH)(OH 2 )BCF] anion of Scheme 4, the highly endergonic (DG 0 ¼ +23.0 kcal mol 1 ) protonation reaction found when [BCF(OH)] is formed now becomes highly exergonic (DG 0 ¼ 22.2 kcal mol 1 ). Consequently, the overall DG 0 for the reaction presented in Scheme 1 and that for a H 2forming reaction in Scheme 2 is now negative: 9.0 kcal mol 1 (or 0.39 eV) for the former and 13.7 kcal mol 1 (or 0.59 eV) for the latter. Moreover, none of the proposed steps after protonation is prohibitively endergonic, and thus may be kinetically feasible, while irreversible loss of gaseous H 2 can drive the doping reaction to the right. The greater exergonicity ## Conclusions We modelled the interactions between three boron-based LAs and different semiconducting p-conjugated polymers, performing detailed quantum-chemical calculations of the structural, energetics and optical signatures for ground-state LAB adducts between LAs and either PFPT or PCPDTPT. Our calculations demonstrate that the observed red-shifted optical absorption in the adducts results from a complex interplay between hybridization, partial CT and changes in the polymer conformation. In assessing the potential of BCF to induce molecular doping in PCPDTBT based on calculated Gibbs free energies of different proposed reactions, we came to the conclusions that both the overall processes proposed by Yurash et al. 16 and by Arvind et al. 24 are highly endergonic, mostly because of the thermodynamically unfavourable protonation by BCF(OH 2 ). Reconciling theory with experiment requires considering complexation of the [BCF(OH)] with another BCF or BCF(OH 2 ) moiety to form more stable anions of the stoichiometry and structure observed crystallographically by Doerrer and Green; 14 these offer a dramatic reduction in the DG 0 penalty for forming the protonated intermediates. We propose that this is followed by moderately endergonic reactions resulting in the elimination of H 2 (as also suggested for the case of metallocene oxidation), either directly from two protonated cationic segments of polymer chains, from "protonated radicals" formed by electron transfer between neutral and protonated cationic segments, or from a protonated cation and a protonated radical (Scheme 3), hence explaining why a single spin-carrying species is observed in EPR measurements. Overall, our calculations highlight the necessity of H 2 loss for the overall feasibility of the reaction, and most importantly, the key role played by the formation of diboron-containing bridged anions in the doping mechanism. Those bridged anions were known, as was the monomeric [BCF(OH)] , but the energetic benefts of bridged anion formation, and therefore its effect on overall reaction feasibility, had not been recognized and certainly not quanti-fed, neither in ref. 16 nor in other works dealing with the doping of p-conjugated polymers with LAs. This is the likely mechanism prevailing at dopant concentrations large enough that BCF dopants can encounter and form complex anions derived from two BCF moieties. In addition, at low dopant concentration and if the dopant is rigorously waterfree, it is also possible that highly hygroscopic BCF molecules could free hole carriers from trapping sites associated with water and/or water-oxygen complexes, 52,53 rather than create excess charges through a conventional doping mechanism. Additional experimental and theoretical work is needed to confrm or reject this hypothesis, as well as to unravel the exact nature of the BCF(OH 2 ) adducts present in doping solutions and the anions present in doped solids. However, this is likely to be very challenging as, even in solution, 1 H and 19 F NMR spectroscopies are unable to reliably distinguish between BCF(OH 2 ) n complexes with different n, 54 while neither the 11 B nor 19 F NMR spectra of [BCF(OH)BCF] differ signifcantly from that of [BCF(OH)(OH 2 )BCF] in solution. 22 Finally we note that the non-straightforward doping nature of the BCF-induced doping process potentially complicates predictions regarding its applicability to other semiconductors. Although variations of the thermodynamic feasibility of the proposed overall p-doping reaction (Scheme 2, but with a complex counterion) for different semiconductors will depend only on the IP of the semiconductor, the kinetic feasibility is expected to depend critically on the ability to protonate the semiconductor. Moreover, different mechanisms may be operative for different semiconductors, for example, if they form substantially more stable "protonated radicals" than PCPDTBT. Finally, the use of BCF as a p-dopant relies on adventitious water and to obtain reproducible doping levels it is likely desirable to use a well-defned and intentionally synthesized BCF(OH 2 ) complex. However, in the presence of additional adventitious water the Brønsted acidity (and thus oxidant strength) of BCF(OH 2 ) is likely decreased. In addition, BCF(OH 2 ) decomposes to (C 6 F 5 ) 2 BOH and C 6 F 5 H on heating, 55 potentially leading to an ill-defned mixture of species in doping solutions or doped flms. It will be useful to carry out further work to identify other Brønsted acids that may be used as effective dopants and that avoid some of these drawbacks.

chemsum

{"title": "Understanding how Lewis acids dope organic semiconductors: a \u201ccomplex\u201d story", "journal": "Royal Society of Chemistry (RSC)"}

concentration_dependence_of_the_sol-gel_phase_behavior_of_agarose-water_system_observed_by_the_optic

## Abstract: concentration dependence of the sol-gel phase behavior of agarose-water system observed by the optical bubble pressure tensiometry nobuyuki ichinose * & Hodaka UraWe have studied an expansion behavior of pressurized bubbles at the orifice of a capillary inserted in gelator-solvent (agarose-water) mixtures as a function of the gelator concentration in which the phase transition points are included. the pressure (P) -dependence of the radius of the curvature (R) of the bubbles monitored by laser beam has shown a discontinuous decrease in the exponent (m) of the experimental power law R = KΔP −m (K: constant) from 1 to 1/2 and a discontinuous increase in the average surface tension (γ ave ) obtained from the work-area plots of the mixtures exceeding that of pure water (72.6 mN/m) at 0.02 < [agarose] < 0.03 wt%, which is attributed to the disappearance of the fluidity. The apparent surface tension (γ app = ΔP/2 R) of the system in the concentration range of 0.03-0.20 wt% has been analyzed by a modified Shuttleworth equation γ app = σ 0 + τln(A/A 0 ), where σ 0 is an isotropic constant component and the second term is a surface area (A) -dependent elastic component, in which τ is the coefficient and A 0 is the area of the orifice. The analysis has indicated that σ 0 coincides with the γ app value of the mixture of 0.02 wt% and the second term at >0.02 wt% is the dominant component. from the appearance of the elastic component and concentration dependence of τ, the plateau of τ for the agarose-water mixtures at 0.03-0.10 wt% (Region II) has been explained to the phase separation giving two-phase mixtures of 0.02 wt% sol and 0.10 wt% gel and the upward inflection of τ at 0.10 wt% has been assigned to an increase in the elasticity of the gel with the increase of the agarose concentration in the range of >0.10 wt% (Region III). On considering the concentration dependence of the surface tension of agarose-water mixtures, the discontinuous and inflection points were assigned to the 1st-and 2nd-order phase transition concentrations of the agarose gel, respectively. Given the results with our tensiometry based on the optical bubble pressure method, distinct gelation points for other systems could be determined both mechanically and thermodynamically which will provide a diagnostic criterion of sol-gel transitions.Gel is a versatile state of substances widely seen in nature, industrial products, and foods due to its solid-fluid dualism, which is owing to the high holding content of solvent in the 3D network which reveals its viscoelastic nature 1,2 . Natural and synthetic water-soluble polymers often form hydrogels where more than 90% of water is contained in weight. Polymer gels are formed by the introduction of crosslinking bonds (chemical gels) or by the mutual aggregation through an increase in the concentration or cooling of the sol (physical gel) 2 . Agarose (Fig. 1) is a polysaccharide taken from a seaweed family (Geridiaceae) [3][4][5][6] , whose molecular weight has been reported to be M w = 0.8-1.4 × 10 5 g/mol 7 and is a typical substance showing physical gel formation, which has been reported to be above the concentration of 0.13 wt% at 20.0 °C8 . However, there are several reports on the minimum gelation concentration (MGC) and sol-gel transition temperature inconsistent each other.Sol-gel transition of physical gels has been studied intensively from mechanistic viewpoints, and several approaches by mechanical, thermal and spectroscopic measurement techniques have been conducted through the dynamic methods 1,2, . However, it remains difficult to determine the transition temperature or concentration through the dynamic rheological 8 , spectroscopic 9 , or thermal 10 measurements owing to the variation in the definition of gel, which depends on the measurement technique employed 2 . Similarly, it is not feasible to determine MGC for physical gels as a critical phase transition concentration because of the difficulty in observing precisely the point of disappearance of fluidity and the point of the appearance of a gel state. The mechanistic study on the gelation of polymers also has a long history of experimental and theoretical chemistry and physics 1,2,12 . Although the main mechanism of the gelation has focused on cross-linking and fibril formation, their order is not clear unless the polymer chains are not intended to be cross-linked in the preparation process of a gel. Although fibrils of linear polymers and even small molecules have been observed by transmission electron microscopy (TEM) and atomic force microscopy (AFM) techniques for flash-freeze-dried agarose gels 2 , evaporation or freezing of the solvent from the systems may induce the fibril formation, and it is not clear whether the fibril formation is essential for gelation. Although the fibril formation affords the "elasticity" or the "plasticity" to gels, which would ensure measurement of mechanical properties of the soft materials, it is not easy to assign the fibril formation as a phase transition phenomenon in a "quasi-static" thermodynamic sense. Thermal hysteresis observed during gelation also makes the analysis complicated. Although agarose gel, for example, melts at 80-90 °C upon heating, the gel is formed at 35-40 °C upon cooling 13 . Since the hysteresis could arise from the existence of several species such as aggregates with single-and double-helices and free polymer in the mixture owing to the extremely slow diffusion and relaxation processes, we planned the surface tension measurement at a constant temperature by means of the bubble pressure method, a tensiometry (surface tension measurement) used for liquids 13 , to detect the deformation of the surface of soft materials including liquids. Since the surface tension (γ) is defined as a ratio of a mechanical work (dW) needed to create a new surface area of an infinitesimal amount (dA), γ = dW/dA and this work is identical to the change in the Helmholtz energy (F) at constant volume and temperature or to the change in the Gibbs energy (G) at constant pressure (p) and temperature (T). Furthermore, since change in the surface tension is also related to the sum of the chemical potentials (μ J ) of the chemical components (J) at the surface area (the Gibbs isotherm) at an equilibrium with the bulk, the mechanical measurement of the surface of a sol or a gel containing J's as a solvent and a gelator can be linked directly with the thermodynamic functions of the system. According to the Ehrenfest classification of phase transition 14,15 , continuity of 1st order derivatives of the chemical potential (μ) as intensive variables are supposed to be the key diagnostic criteria (see also Supporting Information). For one-component systems, μ is a function of T and p, μ = μ(p,T). Since μ requires another variable X J , the mole fraction of component J (J = 1, 2) for two-component systems, phase transition also occurs as a function of X J at a finite condition of temperature and pressure. Since the surface tension is a 1st order derivative of the Gibbs energy, dG/dA for unit mole could be considered as the 1st order derivative of the chemical potential, dμ/dA which is equal to γ for unit mole and unit area. Therefore, some thermodynamic functions such as molar entropy can be derived from the surface tension in a quantity per unit area. For this reason, the surface tension can be a diagnostic parameter for phase transition phenomena on extending the Ehrenfest classification. For example, reported surface tension values of metals below and above the melting point are quite different each other 16 . However, the surface tension is less employed to study phase transition phenomena 17 because of the difficulty in the measurement of the surface tension with the same tensiometry for two different phases, except for liquid-like monolayers at the liquid-air or liquid-liquid interfaces 18,19 where the same tensiometry can be applied and changes in the surface tension can be treated with the Gibbs isotherm. The purpose of our study is to elucidate the sol-gel phase transition through a tensiometry. In other words, changes in an intensive variable γ will indicate the changes in other intensive variables, i.e. the chemical potentials upon the phase transition. However, the surface tension of gels has not been known and the Young-Laplace relationship is not assured in the gel phase. We now report the volume expansion behavior of pressurized bubbles at the orifice of a capillary inserted in gelator-solvent (agarose-water) mixtures as well as the surface tension of the mixtures as a function of the gelator concentration in which the transition point can be determined. We chose agarose as a sample of gelator because of the plentiful reference data in the literature . In this article, the pressure-dependence of the radius of the curvature of the bubbles monitored by laser beam has been studied as a function of gelator concentration to examine the Young-Laplace relationship in the sol and gel states, and to establish a measurement method of the surface tension of gels. We also have studied the discontinuity and the inflection in the surface tension induced by the increase of the concentration in the mixture to examine the 1st-and 2nd-order phase transitions, respectively 14,15 . www.nature.com/scientificreports www.nature.com/scientificreports/ a manometer to the bubble and the static pressure at the depth of the orifice (ΔP). The bubbles in the aqueous solutions of agarose with the concentrations of ≤0.02 wt% (Region I) showed a decrease of the radius upon the increase of ΔP obeying the Young-Laplace relationship with a constant surface tension. The surface tension was obtained by a nonlinear least square curve fitting showing a gradual decrease from that of pure water (72.6 mN/m) to 63.4 mN/m at 0.02 wt% with the agarose concentration as observed for amphiphilic polymers such as polyethylene oxide. The decrease of the radius of the meniscus for the mixtures of ≥0.03 wt% by the applied pressure became smaller compared to that for the dilute solutions, although the appearance of the mixture was solution-like. The double logarithmic plot of R versus ΔP indicated a slope of −1 for the solutions and a slope of −1/2 for the mixture of ≥0.05 wt%, and mixtures of 0.03 and 0.04 wt% showed an intermediate value of the slope of ≈−2/3. This abrupt change in the R-ΔP power law behavior of the bubble strongly suggests that the bubble surface of the mixtures of ≥0.03 wt% is no longer liquid-like (Fig. 4). To compare the surface property of the mixture of ≥0.03 wt% with that of solutions of ≤0.02 wt%, we estimated the surface tension of the mixture without using the Young-Laplace relationship throughout the R-ΔP sets at the agarose concentrations. Assuming the spherical surface of the meniscus, the radius of the curvature (R) can been converted to the surface area (A) and the volume (V) of the meniscus, and the R-ΔP curve has been translated into a ΔP-V curve, then accumulation of the ΔP-V curve gives a course of to the Gibbs energy change (ΔG) as the meniscus has done the work (ΔW) for the volume expansion against the surface tension of the agarose-water mixture. The plot of ΔG vs A shows straight lines for all the mixtures (Fig. 5). Then, the slopes of the lines indicates the average surface tension of the system (γ ave ) as defined thermodynamically, Applied pressure / Pa www.nature.com/scientificreports www.nature.com/scientificreports/ γ = (∂G/∂A) T,p . The error in the surface tension of the solutions with that obtained by the nonlinear least square method is less than 1% (±0.1 mN/m). The surface tension estimated from the Gibbs energy change has a slightly larger error (±0.2 mN/m) than that obtained by the non-linear curve fitting due to the trapezium approximation of 5 sets of the data points in the integration of the ΔP-V curve. Nevertheless, the surface tension values obtained by the two calculations for the solution with concentrations of ≤0.02 wt% agreed within 0.2%. The average surface tension plotted vs. agarose concentration clearly indicates a jump from 63.4 to 75.0 mN/m between 0.02 and 0.03 wt% (Fig. 6). The surface tension of the 0.03 wt% mixture exceeds that of pure water (72.6 mN/m measured with our system) ruling out the exuding of pure water from the 3D network. The surface tension of the mixture is almost constant up to 0.10 wt%. However, it increases for the mixtures of >0.10 wt% until 154.7 mN/m for the 0.20 wt% mixture, but our measurement system was not able to measure the surface tension of mixtures of >0.20 wt%. These results clearly indicate that the mixtures undergo the sol-gel phase transition and the expansion behavior of the bubbles in the gels is network-limited. ## Discussion As the Young-Laplace relationship has been derived from the equation concerning the work for volume expansion and that for the increase of the surface area of a bubble with a small change in the radius (dR), it is necessary to consider the work for the volume expansion for the agarose-water mixtures of ≥0.03 wt%. On considering the gel as a solid, we cannot apply the equation with a constant surface tension, dW = dAγ because the presence of the surface strain due to the elasticity. However, the Young-Laplace equation must hold at a given ΔP. We have calculated the surface tension for each R-ΔP data set as apparent surface tension (γ app ) and reconsider the surface tension with the equation dW = d(Aσ) = σdA + Adσ, where σ is the thermodynamic surface tension. Differentiation of this equation with respect to A gives an equation γ mec = σ + dσ/dlnA according to the treatment of the surface tension of solids 16,17 . As shown in Fig. 7, γ mec increases linearly with the increase of ln(A/A 0 ) and tends to converge to the γ mec value for the 0.02 wt% solution (γ mec = 63.4 mN/m) upon extrapolation of γ mec values for various mixtures at zero expansion (A = A 0 ). The intercepts are almost independent of the agarose concentration, while the slope is dependent on the agarose concentration. Therefore, we can rewrite the equation for γ mec as γ mec = σ 0 + dσ/dln(A/A 0 ), where σ 0 = 63.4 mN/m. The average surface tension γ ave can be considered as a representative value of γ mec at the average surface area (A = Ᾱ), which is dependent on the applied pressure within the fracture limit of the bubble at an agarose concentration. We now return to the R-ΔP relationship. The experimental power law observed for the radius of the curvature to the applied pressure showed an exponent of −1/2, i.e. R = K(ΔP) −1/2 , where K is the constant. This relationship could be obtained from the Young-Laplace equation R = ΔP/2γ mec and the equation for γ mec at the corresponding surface area A = π(R 0 2 + z c 2 ). Although we did not obtain the analytical solution for R = R(ΔP), the second term (compressibility modulus) in γ mec originating from the surface strain appears and dominates with the increase of the agarose concentration above the minimum gel concentration and this will cause the change in the exponent. We also obtained a linear relationship between K and γ mec in their double logarithmic plot (Fig. S5). As a result of the introduction of the surface strain term for the analysis of the R-ΔP data, we could conclude that the observed change in the exponent m is phenomenal although detailed numerical analysis might reveal the R-ΔP relationship. We have introduced the surface strain term dσ/dln(A/A 0 ) in the mechanical surface tension of agarose-water mixture with the concentration of >0.02 wt% and now show a plot of γ mec -concentration (Fig. 8). The introduction of the surface strain term indicates the abrupt increase of γ mec at 0.02 < [agarose] < 0.03 wt% and an inflection at 0.10 wt% more clearly as compared to the γ ave -concentration plot (Fig. 6). As shown in Fig. 8 (and Fig. S6), a plateau is seen in the concentration range of 0.03-0.10 wt% (Region I). According to Herring, γ mec contains a scalar www.nature.com/scientificreports www.nature.com/scientificreports/ and a tensoric terms, the latter is corresponding to the surface strain term. Rusanov 23,24 explained nonequivalence of the mechanical (γ mec ) and thermo-dynamic (σ) surface tension for wetting of an isotropic solid surface co-existing mobile components by relating σ to γ mec and the sum of the chemical potentials of the mobile (I) and immobile (J) components dσ = − s (J) dT + (γ mec − σ)dlnA − ΣΓ I(J) μ I , where s (J) is the entropy surface density and Γ I(J) is the surface excess of the mobile component at the surface of the immobile component. For the present case, I = water and J = agarose, ΣΓ I(J) μ I can be considered to be independent of A and the experiment has conducted at the constant temperature. Therefore, we obtain γ mec = σ + dσ/dlnA = σ 0 + dσ/dln(A/A 0 ) again on considering γ mec = σ 0 at A = A 0 . This means the mechanical surface tension is consisted of scalar term as the surface tension of σ 0 value for the 0.02 wt% solution and the strain term dσ/dlnA due to the strain of the agarose gel as a solid. The strain term is also related to the chemical potentials of the immobile components. However, there is no fast diffusion and no equilibrium between the surface and the bulk phase meaning the inequality of the chemical potentials and there is an area-dependent strain due to the measurement. For this reason, we cannot treat the strain term directly as a parameter for the phase behavior of the solid. Fortunately, the elasticity of polymer gels originates from the entropy of the polymer chain, i.e. ΔF = −TΔS 25 . Now we can treat the strain term as the entropy surface density of the immobile component, i.e. the gel network. With a small isotropic deformation of the surface area, www.nature.com/scientificreports www.nature.com/scientificreports/ experimental value of the strain term is given by (dσ/dlnA)ln(A/A 0 ) ≈ 2ετ, where ε is the isotropic tensor and τ = dσ/dlnA. Therefore, the mechanical surface tension γ mec can be a thermodynamic quantity of gels and inflection of the strain term is attributable to a change in the elasticity due to a structural change of the gel. We have concluded that the surface tension can be a criterion for sol-gel transition: the observed discontinuity explained by the appearance of the strain term is assigned to the sol-gel phase transition as the 1st-order phase transition is, and the inflection of τ reflects the change in the elasticity of the solid gel phase as the 2nd-order phase transition. Finally, we will enter the phase behavior of the agarose-water mixture. The surface tension of the mixture of ≤0.02 wt% (Region I) is the liquid with no doubt because of the Young-Laplace behavior and γ mec = σ throughout the applied ΔP range. The observation of the plateau in γ mec in the range of 0.03-0.10 wt% (Region II) indicates that the strain term of the gel is almost constant. This strongly suggests that the chemical composition of the gel is almost identical despite of the change in the net concentration. As judged from the above discussion on σ 0 , the chemical composition of the sol is also constant. Therefore, only quantities of both phases seem to vary with the concentration. This has invoked the idea of phase separation and the lever rule to explain the slight increase of γ mec in Region II. Spinodal decomposition of the mixture followed by gelation and concomitant phase diagrams have been reported for the mixture of atactic polystyrene in cyclohexane as judged by the test tube tilting and ball-drop method 26 . Indei has reported network formation of poly(vinyl alcohol)-borax by percolation as judged by the micro-rheology 27 . Similar mechanisms for gelation of the agarose-water system at lower concentrations have been proposed on the basis of the rheological 8 and dynamic light scattering measurements 28 . Spinodal decomposition of agarose-water at 0.02 < [agarose] < 0.03 wt% gives 0.02 wt% solution and 0.10 wt% gel whose amounts are determined by the lever rule as an origin of σ 0 and weak concentration-dependence of γ mec in Region II. This leads us to a conclusion that the gel in Region II is two-phase gel with a percolated or a fractal-like structure because of the freedom of the system for intensive variables (f) is 0 for the sake of the phase rule (f = c -p + 2, where c and p are the numbers of components and phases, respectively. Note that two of f are occupied by the temperature and pressure of the experimental condition.). The slight increase in γ mec can be explained by the number of the cross-linking points which increases with the increase of the amount of the gel domain in the mixture. As Rees predicted that agarose molecules form double helix aggregates 5 , which have been observed by X-ray diffraction and other micro-imaging techniques with ≥0.1 wt% mixtures, Liu et al. proposed a "primary fiber" as an intermediate in the gelation of agarose 6 . The primary fiber could be an aggregate formed by twisting of two agarose chains which act as a substrate for self-epitaxial nucleation to form the 3D network. Our observation with the mixtures in Region II and the >0.10 wt% concentration region (Region III) strongly suggests that the expansion of the network requires some energy to unwind the primary double helix, i.e. the energy to break the hydrogen and the hydrophobic bindings and the energy to unwind the multiple helix requires increases by the increase of the concentration. Tieleman et al. have measured the area-dependence of the surface tension of a lipid monolayer, which undergoes phase transition by area expansion 29 . They have reported the compressibility moduli (corresponding to τ = dσ/dlnA) of dipalmitoyl-phosphatidylcholine (DPPC) to be 1400 ± 20 mN/m for liquid condensed phase at a molecular area (A L ) of A L = 0.475 nm 2 , 200 ± 12 mN/m for liquid condensed and liquid expanded coexisting phase at A L = 0.58 nm 2 , and 100 ± 20 mN/m for liquid expanded phase at A L = 0.620 nm 2 , respectively 29 . They have also observed inflection points in the γ-A plot 29 . On the other hand, our observation indicated the values of dσ/dlnA = 300-500 mN/m in Region II and 500-3500 mN/m in Region III together with the inflection point at 0.10 wt% (Fig. S6). These data clearly show that the two-phase gel in Region II has a weak binding nature as compared to the monolayer of alkyl chains. Upon increasing the agarose concentration, dσ/dlnA values of the gel increase drastically by the increase of the density of cross-linking points 12 . This means that the structure of the gel changes with the concentration and the freedom f = 1 corresponding one-phase gel (p = 1) in Region III. Although melting of the agarose gels of >0.10 wt% (Region III) is well known to occur at 35-50 °C5,6 , the systems below this concentration (Region II) have been reported to be a suspension of micro-gel 8,28 , and its phase transition behavior upon the concentration change has been mentioned less frequently. Although intensive study including temperature dependence is needed to draw an exact phase diagram, we show a schematic phase diagram which explain the observed behavior of the surface tension as a function of agarose concentration (Fig. 9). ## Material Agarose (Agarose-S tablets, Nippon Gene, for electrophoresis use, sulfate (−SO 4 ) content ≤0.1%) was used as received. Its molecular weight of Agarose-S as given by the manufacturer was 2-3 × 10 5 g/mol determined by liquid chromatography. Mother solution of agarose (1.0 wt%) was prepared by heating a mixture of an agarose tablet (≈0.50 g) and distilled deionized water (500 mL) was heated to 90-95 °C in a beaker with a microwave oven. The solution was diluted at ≈40 °C (above the gelation point of 37-39 °C for 1.5 wt% solution). ## Method We have modified the bubble pressure method where the experimental apparatus possesses a Teflon cell holding a capillary and an optical window which enable the pressurization of the sample and monitoring of the radius of the curvature (R) of the meniscus (bubble) optically with a laser beam passing through the capillary (pressurizing optical probe) 30,31 . The meniscus formed in the solution or gel acts as a concave lens for the laser beam. The focal length of the meniscus is determined by an external optics and a photodetector system. The radius of the curvature was translated into the surface tension of the mixture, as a first derivative of Gibbs energy (G) with respect to the surface area (A), through the translation into the surface area and volume (V) of the meniscus as a function of the pressure applied (P). The gelation has been judged by the apparent surface tension exceeding that of pure water. The details are described in the Information. www.nature.com/scientificreports www.nature.com/scientificreports/ conclusion In conclusion, we have established a tensiometry for weak gels using the optical bubble pressure method through the demonstration of the sol-gel transition behavior of the agarose-water mixtures as consecutive occurrences of the loss of the fluidity and the increase of the surface tension upon the increase of the gelator concentration, which can be attributed to the 1st-and 2nd-order phase transitions through the analyses of the Gibbs energy change of the meniscus for its expansion and the mechanical surface tension for solids as introduced by Shuttleworth 32 and Herring 33 , and with the aid of the concept of entropic elasticity. Therefore, the increase in the surface tension upon the gelation can be attributed to the changes in the mechanical properties of the polymer network as a solid with entropic elasticity and can be interpreted into thermodynamic phase behaviors. The present results has indicated that the surface tension measurement will provide a reliable criterion for the sol-gel transition of other polymeric systems and also for that of low molecular-mass organic gelator systems, which are being increasingly studied 34,35 .

chemsum

{"title": "Concentration dependence of the sol-gel phase behavior of agarose-water system observed by the optical bubble pressure tensiometry", "journal": "Scientific Reports - Nature"}

chloromethyl-modified_ru(<scp>ii</scp>)_complexes_enabling_large_ph_jumps_at_low_concentrations_thro

## Abstract: Photoacid generators (PAGs) are finding increasing applications in spatial and temporal modulation of biological events in vitro and in vivo. In these applications, large pH jumps at low PAG concentrations are of great importance to achieve maximal expected manipulation but minimal unwanted interference. To this end, both high photoacid quantum yield and capacity are essential, where the capacity refers to the proton number that a PAG molecule can release. Up to now, most PAGs only produce one proton for each molecule. In this work, the hydrolysis reaction of benzyl chlorides was successfully leveraged to develop a novel type of PAG. Upon visible light irradiation, Ru(II) polypyridyl complexes modified with chloromethyl groups can undergo full hydrolysis with photoacid quantum yield as high as 0.6.Depending on the number of the chloromethyl groups, the examined Ru(II) complexes can release multiple protons per molecule, leading to large pH jumps at very low PAG concentrations, a feature particularly favorable for bio-related applications. ## Introduction Photoacid generators (PAGs) are molecules which can generate acids under light irradiation, 1 and have been widely used as photo-initiators of cationic polymerization in photolithography, photocuring, and three-dimensional (3D) printing. 2 The great success that PAGs have achieved in the felds of the microelectronic industry and microfabrication relies on their spatial and temporal control of proton concentrations. Protons also play pivotal roles in a variety of biological events. As a result, PAGs have found ever-increasing applications in biochemical, biological, and biomedical areas in recent years, such as photodynamic therapy, 3 drug delivery, 4 adenosine triphosphate (ATP) biosynthesis, 5 photocontrol of enzyme activity, 6 and protein conformations, 7 as well as proton transfer in biomolecules. 8 For these bio-related applications, PAGs may have some specifc characters, 9 including proper solubility in aqueous solutions, light response in the visible or near-infrared region for deep tissue penetration and minimal damage to biomolecules, 10 and large pH jumps to get remarkable bio-effects. To reach instant large pH jumps, high photoacid quantum yields were vigorously pursued. However, mM levels of PAGs, concentrations not easy to realize under in vitro and in vivo conditions, were usually needed because each PAG molecule can theoretically generate one proton only. 4a,5-7 High concentrations of PAGs and their corresponding photoproducts may also interfere with the examined biological processes, hampering their application severely. Though efforts have been made to address this issue by integrating two PAG moieties into a single molecule, the resultant PAGs showed poor photoacid quantum yields. 1c,3 New strategies that can endow PAGs with both high photoacid capacity (i.e. the proton number a PAG molecule can release) and high photoacid quantum yield are in urgent demand to boost their bio-related utilization. Thermal hydrolysis of haloalkanes may release HX (X ¼ Cl, Br, I) and the acid capacity depends strictly on their halogenation levels. 11 To the best of our knowledge, such a classic reaction has not been capitalized on in PAGs. What attracts our attention is the hydrolysis of benzyl chloride, the kinetics of which can be effectively modulated by electron donating/ withdrawing groups. 12 Electron donating groups will accelerate the process while an opposite effect may be observed for electron withdrawing groups. As reported by A. Fry and S. Yamabe, para-methoxy benzyl chloride has a hydrolysis rate of 4 orders of magnitude higher than that of the para-NO 2 counterpart. 12 Inspired by these facts, we herein synthesized a series of 4,4 0 -bis(chloromethyl)-2,2 0 -bipyridine (bcm-bpy) coordinated Ru(II) complexes (1-3, Scheme 1) to explore their PAG capability. The rationale behind our design is as follows. (1) The bcm-bpy ligand is expected to be inactive in thermal hydrolysis due to the electron-defcient feature of the pyridine ring, which may be consolidated further upon coordination to the Ru(II) center. 13 A good stability in the dark is a prerequisite for a desired PAG. (2) The highest occupied molecular orbital (HOMO) of Ru(II) polypyridyl complexes is generally Ru(II) centered, 14 while the lowest unoccupied molecular orbital (LUMO) of complexes 1-3 should localize on the bcm-bpy ligand due to the electronegativity of Cl atoms. 15 Thus, the bcm-bpy related metal-to-ligand charge transfer (MLCT) state will be accessed preferentially upon light irradiation, from which an efficient hydrolysis of the chloromethyl groups is also anticipated due to the greatly enhanced electron density on the bcm-bpy ligand. This is indeed what we observed in our experiments. Complexes 1-3 can undergo efficient hydrolysis upon visible light irradiation in aqueous solutions, and release 2, 4, and 6 equivalents of HCl, respectively, with photoacid quantum yields as high as 0.6. Both high photoacid capacity and high photoacid quantum yield of complex 3 make large pH jumps feasible at low concentrations. Complex 3 is not only the frst PAG which can release six protons per molecule, but also the frst type of PAG that makes use of hydrolysis reaction of benzyl chloride groups. As a conceptual demonstration, complex 3 was successfully used to switch on the activity of acid phosphatase upon visible light irradiation at a concentration as low as 10 mM. 6 ## Theoretical calculations Theoretical calculations based on the Gaussian 09 program package 16 (see ESI †) were carried out before experiments to examine our ideas. As expected, the calculated HOMO and LUMO of 1-3 are mainly Ru(II) and bcm-bpy based, respectively (Fig. 1 and S1-S2 †). Thus MLCT Ru/bcm-bpy excitation means pumping one electron from the Ru(II) center to the bcm-bpy ligand, which will greatly enhance the electron density of bcmbpy. Such an expectation is further convinced by comparison of Mulliken charges of the ground state (GS) and T1 state. Taking complex 1 as an example (Table S1 †), the selected Mulliken charges of the atoms on the bcm-bpy ligand in the T1 state are all negatively shifted compared with that in the GS. In addition, the length of C-Cl bonds stretches from 1.83 (GS) to 1.85 (T1) (Fig. S3-S5 †). All these results are favorable for photoinduced hydrolysis of the chloromethyl groups. ## Synthesis and characterization The synthesis and characterization of complexes 1-3 were reported in our previous work. 13 The aqueous solutions of all the complexes are quite stable in the dark, as evidenced by the negligible changes in absorption spectra, emission spectra, and 1 H NMR (nuclear magnetic resonance) spectra, as well as pH values of the solutions (Fig. S7 †). Upon irradiation at 520 nm, rapid changes in these spectra were observed. Taking complex 3 as an example, the absorbance in the regions of 315-375 nm and 425-500 nm decreased gradually, along with a slight blue-shift of the ligand-based transition peak centered at 291 nm (Fig. 2). The emission intensity also increased quickly upon irradiation (Fig. 2). The absorption and emission spectra did not change any more after irradiation for only 3 min. At that time, the solution exhibited nearly the same absorption and emission as that of [Ru(bhm-bpy) 3 ] 2+ (20 mM, bhm-bpy ¼ 4,4 0bis(hydroxymethyl)-2,2 0 -bipyridine), suggesting full hydrolysis of the six chloromethyl groups. The excited [Ru(bhm-bpy) 3 ] 2+ without competing photolysis decay should have a longer excitation lifetime and also a higher luminescence quantum yield compared with complex 3, and thus a turn-on luminescence was observed. The known PAGs usually show absorption changes upon light irradiation. Our complexes represent the frst class of PAGs that exhibit remarkable luminescence turn-on along with proton release. By virtue of the higher sensitivity of fluorescence detection as well as the wide application of fluorescence confocal microscopy, this feature may facilitate biological studies. In addition, the transformation of 3 into [Ru(bhmbpy) 3 ] 2+ is also confrmed by high resolution electron spray ionization mass spectra (HR ESI-MS) (Fig. S16 †) and 1 H NMR spectra (Fig. 3). Only an m/z peak of 375.0861 which can be ascribed to the product with six hydroxymethyl groups was observed after irradiation in H 2 O. A similar result was obtained Scheme 1 Chemical structures of complexes 1-3. ## pH changes The pH changes of the aqueous solutions of complexes 1-3 (10 mM) were monitored with a pH meter (Fig. 4). The initial pH values of the solutions were about 6.6. The weak acidity may be attributable to the dissolved CO 2 absorbed from ambient air. 17 The pH values of the solutions kept unchanged in the dark for 24 h (Fig. S7 †). Upon LED (light-emitting diode) irradiation at 520 nm, the pH values decreased quickly to 4.9 for 1, 4.6 for 2 and 4.3 for 3 in 3 min (the theoretically calculated results are 4.7, 4.4 and 4.2 for complexes 1-3, respectively). A large pH jump of 2.3 units was obtained for complex 3 at a concentration as low as 10 mM, which obviously profts from its ability to generate 6 protons for each molecule. To reach a similar pH jump, much larger concentrations, usually several hundreds of micromolar, were generally needed for the reported systems, 6,17,18 unfavorable for biological applications. In bio-related applications, bioactive molecules, such as glutathione (GSH), may serve as strong nucleophilic agents to impair the dark stability of these chloromethyl-modifed Ru complexes. 19 Thus, the effect of GSH was also studied. The initial pH value of a GSH (1 mM) aqueous solution was measured to be 3.4. Addition of 1-3 (10 mM) did not cause any pH changes in 24 h without irradiation (Fig. S20 †). The negligible pH changes may also be the result of the strong buffering ability of GSH. To rule out this possibility, ethanethiol (1 mM) was used instead of GSH. Similar results were obtained again at either room temperature or 37 C (Fig. S21 and S24 †), con-frming the good dark stability of complexes 1-3 even in the presence of strong nucleophilic agents. Upon 520 nm irradiation, the pH values of 1-3 (10 mM) solutions containing ethanethiol decreased quickly (Fig. S22-S24 †), showing their desirable anti-interference ability. ## Photoacid quantum yields The photoacid quantum yields of 1-3 (50 mM) in H 2 O were measured using potassium ferrioxalate as a chemical actinometer (ESI). 20 The light of 520 nm was not suitable for potassium ferrioxalate due to its quite small extinction coefficient at 520 nm, and thus a 470 nm LED was selected. The obtained quantum yields at 470 nm are 0.62 AE 0.01 for complex 1, 0.65 AE 0.02 for complex 2 and 0.61 AE 0.01 for complex 3, which are among the highest reported for PAGs. 1c,d,2a,b ## Photoacid mechanism The high photoacid quantum yields and photoacid capacities of 1-3, along with their good dark stability, make the hydrolysis mechanism of the chloromethyl groups anchored on these complexes particularly interesting. Generally, a thermal hydrolysis of a haloalkane may undergo through either an S N 1 (unimolecular) or an S N 2 (bimolecular) pathway. 12 For benzyl chlorides, the mechanism may change from S N 1 to S N 2 depending on the electron donating/withdrawing abilities of substituents. According to the reports of A. Fry and S. Yamabe, the hydrolysis of para-methoxy benzyl chloride proceeds through an S N 1 way, while S N 2 is found for the para-NO 2 compound. For complexes 1-3, MLCT excitation enhances the electron density of the bcm-bpy ligand, which is favorable for S N 1 rather than S N 2 Based on the theoretical and experimental results, a possible photoacid mechanism is schematically illustrated in Scheme 2, taking complex 1 as an example. Furthermore, we also examined the photolysis of the bcm-bpy ligand in H 2 O. The absorption of the free bcm-bpy ligand appears below 310 nm. Under direct UV light (254 nm) irradiation, HCl was also generated, most probably through a different mechanism involving homolysis of a carbon-chorine bond as reported by R. Sinta et al. in the studying of photoacid properties of 4,6-bis(trichloromethyl)-1,3,5-triazines. 21 Compared with the clean transformation of 1-3 into their corresponding hydroxymethyl products, the photolysis of bcm-bpy To the best of our knowledge, this is the frst time that the hydrolysis reaction was applied to develop a novel type of PAG. Since their discovery four decades ago, PAGs have been dominated by iodonium and sulfonium salts. Only over the past decade, new structures along with their unique properties have emerged in the PAG arsenal, such as photochromic triangle terarylenes, 1e,2a the open form of spiropyrans, 17,18 and N-oxyimidesulfonates. 22 Bearing in mind that the benzyl chloride groups are easy to integrate into a variety of photosensitizers, such as porphyrins, phthalocyanines, and boron dipyrromethene (BODIPY) dyes, more diverse PAG structures and their broader applications are expected. ## Biological applications As a conceptual demonstration of its possible biological applications, complex 3 at 10 mM was used to modulate the activity of acid phosphatase (ACP). 6 ACPs are widely distributed in the human body. An abnormal activity of ACPs is related to many diseases, including prostate cancer, kidney disease, multiple myeloma, Gaucher disease, etc. 23 Therefore, controlling the ACP activity may fnd potential applications in disease treatment and drug development. S. Kohse and co-workers successfully utilized a photoacid (2-nitrobenzaldehyde) to realize efficient tuning of the ACP activity. 6 Similarly, 4-methylumbelliferyl phosphate (MUP) was selected as the substrate of ACP. The transformation of MUP into 4-methylumbelliferone (MU) was monitored by absorption spectra. The initial pH of the solution of MUP and acid phosphatase was kept at 8.0, where the activity of the enzyme was faint as negligible spectral changes were observed. Upon addition of complex 3 (10 mM) and irradiation for 90 s with an LED at 520 nm, the activity of the enzyme was successfully switched on, and elevated to a level of 0.05 mmol (min mg enzyme ) 1 (Fig. 5 and S27 †), which is consistent with the reported results. 6 However, a PAG concentration as high as 500 mM was implemented in their experiments. Light irradiation in the control experiment without the enzyme caused no effects on MUP (Fig. S28 †). ## Conclusions In conclusion, we designed and synthesized three bcm-bpy based Ru(II) complexes as novel PAGs, which can work in aqueous solution and be excited by green light with photoacid quantum yields of about 0.6. Complex 3 can release six protons per molecule, leading to large pH jumps at low concentrations. Theoretically, more protons may be released provided more chloromethyl groups are anchored on the ligand. Our results may open new avenues for developing novel PAGs with both high photoacid quantum yield and capacity to meet more challenging demands particularly in bio-related areas. ## Conflicts of interest There are no conflicts to declare.

chemsum

{"title": "Chloromethyl-modified Ru(<scp>ii</scp>) complexes enabling large pH jumps at low concentrations through photoinduced hydrolysis", "journal": "Royal Society of Chemistry (RSC)"}

dispersion_of_tio2_nanoparticles_improves_burn_wound_healing_and_tissue_regeneration_through_specifi

## Abstract: Burn wounds are one of the most important causes of mortality and especially morbidity around the world. Burn wound healing and skin tissue regeneration remain thus one of the most important challenges facing the mankind. In the present study we have addressed this challenge, applying a solution-stabilized dispersion TiO 2 nanoparticles, hypothesizing that their ability to adsorb proteins will render them a strong capacity in inducing body fluid coagulation and create a protective hybrid material coating. The in vitro study of interaction between human blood and titania resulted at enhanced TiO 2 concentrations in formation of rather dense gel composite materials and even at lower content revealed specific adsorption pattern initiating the cascade response, promising to facilitate the regrowth of the skin. The subsequent in vivo study of the healing of burn wounds in rats demonstrated formation of a strongly adherent crust of a nanocomposite, preventing infection and inflammation with quicker reduction of wound area compared to untreated control. The most important result in applying the TiO 2 dispersion was the apparently improved regeneration of damaged tissues with appreciable decrease in scar formation and skin color anomalies.Accelerated and less painful healing of wounds caused by burn or mechanical injuries and skin and muscle tissue engineering for decreased scar formation and minimization of permanent damage belong to most prominent challenges in modern surgery 1 . Application of nanostructured materials for improved tissue regeneration has become a well-developed and accepted practice in the application of metal bone implants, where a thin layer of nanostructured titanium dioxide is deposited on the top of the implant surface 2,3 . Nano titania is rapidly getting coated with proteins when immersed into the biological fluids due to its well-recognized ability to adsorb and coordinate proteins 4 and phospholipids 5 on its surface. Adsorption of phospholipids can be considered as one of the factors guiding the attachment of cells and grafting on the growing tissue on an implant 6 . In the domain of skin regeneration a strong effort so far has been set on application of stem cells. They have been applied in different approaches, in particular, including sprays 7 . Use of nanomaterials for wound treatment and skin repair has also been intensively investigated ranging from silicone based artificial skin layers 8 , to the development of new materials for wound dressing with delayed and prolonged release of medicines 9,10 and even to direct application of nanoparticle dispersions either possessing themselves antibacterial effects, such as silver and gold nanoparticles [11][12][13] , or loaded with painkillers and antibiotics. The latter have been realized with metal oxide nanoparticles which the FDA approved for intravenous application, namely with Al 2 O 3 14 or with Fe 3 O 4 15 as carriers. The use of medicine-loaded sol-gel alumina and iron oxide resulted in appreciable reduction of the scar tissue sizes with wound healing times not principally different, however, from those when only a solution of the selected medicines was applied directly to the wound 14,15 . Considerable improvement in the size and structure of the scar could be observed when anti-inflammatory natural medicine curcumin loaded siloxane gels 16 , while using silver nanoparticle-graphene-polymer nanocomposites an appreciable acceleration of wound healing was achieved 17 . Nano TiO 2 has also drawn a considerable attention in the recent years. In the view of broad application of nano titania a lot of effort were set on, in the first hand, investigation of its potential toxicity in different forms 18 . It has been demonstrated in numerous studies reviewed, analyzed and reproduced in 19 that in the dark no or negligible toxicity could be associated with essentially any form of nano titania. The apparent toxicity, in particular DNA damage, could be observed for both human cells 19 and for microorganisms, such as, for example, micro algae 20 in the presence of highly crystalline larger (over 25 nm and in most apparent cases -about 100 nm 20 ) particles, obtained by spray pyrolysis, subjected to irradiation by UV and visible 20 and in some cases even IR light 21 . This effect was unequivocally related to the photocatalytic properties of the applied titania and was more pronounced for the more catalytically active anatase phase 20 . The application of nano TiO 2 in wound healing was also very much focused on the exploitation of the photocatalytic effect for production of reactive oxygen species to target bacteria in wound infections 21,22 . To enhance the bactericidal action of titania via photochemical effects, quite a number of applications have been developed for grafting titania onto either fabrics 23,24 or porous (hydrophilic) polymer nano composites 25,26 . In some cases such composites were used, however, primarily for controlled drug delivery to the wounds 9,10 . In the present work we have chosen a completely different approach in use of nano titania for wound healing. The material applied here was a dispersion of small (less than 10 nm) anatase particles produced by sol-gel method in solution. The colloid was stabilized by grafting of (protonated) antioxidant ligand triethanolamine on the surface of the particles, making them positively charged in the originally applied media and photochemically inactive 27 . The small anatase particles stabilized by antioxidant ligands have been demonstrated to be biocompatible for both human cells 28 and for bacteria 27 and micro algae 29 . The applied dispersion did not contain any additional bioactive substances or medicines. The aim was set to investigate possible effects originating solely via surface interactions (adsorption) of the nano titania, possessing large active surface area, with body fluids. ## Results and Discussion Nanoparticles of titania, independently of their phase composition and in spite of their broadly recognized adsorbent properties towards biomolecules, are commonly considered as inert in relation to living systems 18 . Titania as anatase powder is broadly used in food and hygiene industry, referred to as E171 food colorant. Commercially available Degussa P25 titania nano powder is considered as broadly accepted negative standard in the in vitro acute toxicity studies 18 . The dispersion of sol-gel produced anatase nanoparticles stabilized by charging via surface complexation with triethanolammonium ligands applied in this work has been characterized in several earlier publications and was proved to be biocompatible in contact with both human and plant cells up to rather high concentrations reaching 100 μg/mL 30,31 . Recent investigation of the Degussa P25 nanoparticles at very low concentration of below 50 ng/mL has demonstrated them to be capable to induce activation of the contact system eliciting thromboinflammation 32 . This latter observation was considered to be associated with potential health risk from titania nanoparticles if they emerge in the body fluids. On the contrary, we hypothesized that a dispersion of TiO 2 nanoparticles should be applied on the skin to cause enhanced blood coagulation, which is an important first step in initialization of the wound healing processes. The nanoparticles applied in this work were produced by hydrolytic route from titanium ethoxide modified by triethanolamine ligands, following the procedure adopted from 27 with some minor adjustments (Please, see the experimental part and Supplementary). They belong to the anatase phase as is clearly indicated by the distances between fringes for the aligned {101} planes of 0.354 nm in the high resolution TEM images (see Fig. 1A and Figs S1, S3 and S4). The size of the particles is rather uniform and is well in agreement with the observed hydrodynamic size in both the initial alcohol-based dispersion and in the dispersion obtained by its 10 times dilution by de-ionized water (Figs 1B and S2). In contrast, the dilution by 0.9 wt% NaCl (physiological solution), while not leading to precipitation is associated with extended aggregation and apparent increase in the hydrodynamic size of the produced aggregates with distribution between about 100 nm and several micrometers (Fig. 1C). Spectrophotometric measurements of the clot formation at quite enhanced final TiO 2 concentrations in human blood plasma (1 mg/mL and 10 mg/mL respectively) showed that it was strongly accelerated compared to normal clotting in air. The process was completed in just over 30 s at 1 mg/mL and in about 10 s for 10 mg/mL compared to over 1 min with untreated blood serum (see Fig. 2). In order to bring insight into the interaction of blood with the applied titania dispersion, we produced clots by addition of a droplet (0.02 mL) of the dispersion (or PBS solution for reference) to 0.10 mL of whole blood drop placed onto an optical glass slide. Quick clotting in case of TiO 2 dispersion led to formation of a composite in a form of dark brown, almost black, brittle solid. The SEM analysis revealed formation of a dense solid TiO 2 film on the outer surface of solidified droplet with thickness of 30-40 μm, covering a complex structure of iron-rich protein composite with only 3-4 wt% content of titania as calculated from EDS analysis (see Fig. 3B,C and Supplementary Fig. S3, and Table S1). The structure of the protein clot is apparently different for pure blood or the blood with added PBS solution on one hand, and the material produced on interaction with titania dispersion on the other hand: the microstructure in the latter case is larger and more smooth, indicating stronger interactions within the clot. It indicates that the addition of titania is resulting in stronger interactions within the forming solid. We have applied rather high concentration of titania aiming to produce a dense nanocomposite material with perspective to form such composite coatings on the wounds as protective patches instead of using polymer patches as recommended earlier 9, . Interaction of the dispersion with freshly removed human epidermis (outer skin layer) was also investigated, showing that the complex skin structure becomes coated with a uniform dense layer of solid titania film (Fig. 3D-F) with thickness dependent of the concentration and amount of added dispersion, but typically thicker than 10 μm. The thickness of a film obtained by single deposition is apparently quite high, which leads to formation of a uniform system of surface cracks originating from gel densification on the evaporation of solvent. The coating has apparently good adhesion to the skin surface and is not removed mechanically when dry. Washing off with mechanical brushing in a water flow removes major part of it with residues remaining persistently in the skin micro folds. These features in combination indicated that treatment of the wounds with the applied concentrated titania dispersion were going to create on their surface quite dense mechanically tough and strongly adhering hybrid coatings. Such coatings could be potentially capable to serve as protective patches on their own, eliminating the need to cover a wound with some additional protective material/bandage. Platelet activation was measured as the reduction in platelet numbers in the blood after incubation with the TiO 2 nanoparticle-coated surface (TiO 2 ), polystyrene surface (PS) and Corline heparin surfaces (CHS) compared to the initial blood samples (i.e., not exposed to the chambers). The TiO 2 surface induced a clear reduction in platelets as only 25.8 ± 6.4% (mean ± SEM) of the platelets remained in the blood after the incubation, while the blood incubated with the plain PS surface still had 66 ± 9.4% of the platelets and the in the blood from the CHS surface more than 90% (93.2 ± 2.1%) of the platelets were remaining after the same incubation time (See Fig. 4A). Activation of the coagulation system was further analyzed by measuring the generation of TAT complexes after blood exposure to the TiO 2 nanoparticle containing surface, the PS surface and the CHS surface. As expected the TiO 2 -nanoparticle coated surface resulted in a large increase in TAT levels compared to the PS surface that just showed a small increase in TAT levels, and the CHS control surface that gave no significant rise in TAT concentration compared to the initial blood samples (See Fig. 4B). The intrinsic pathway of coagulation is triggered when FXII come in contact with foreign materials, which initiates the clot formation and thereby also the wound healing process. Hence, it would be of importance to see if the TiO 2 nanoparticle-coated surfaces induce this type of contact activation. In a previous publication we investigated protein adsorption and contact system activation induced by TiO 2 nanoparticles incubated in Figure 1. TEM image of the dried dispersion of the applied TiO 2 nanoparticles produced using modified methodology from ref. 18 (A). Hydrodynamic size of the particles in water (B) and in isotonic salt solution ((C), 0.5 ml of dispersion diluted by 10 ml isotonic NaCl), both by DLS. Natural blood clot and its EDS analysis (A), enlarged structure of the inner part of the blood clot forming on interaction with the TiO 2 dispersion and its EDS analysis (B), TiO2 crust on the surface of a treated blood clot and its EDS analysis (C). The integrations of EDS spectra are presented in Table S1 (Supplementary). Untreated skin sample (D) coating on the surface of skin sample (E), enlarged view of the titania film on a skin sample (F). human EDTA-plasma and whole blood without anticoagulantia, respectively 32 . The formed protein corona was abundant in most contact activation proteins; five out of the ten protein identified with highest score identified by MALDI-TOF belonged to the contact system. High amounts of contact system activation complexes were generated reflecting this binding 32 . In the present study, the generation of FXIIa-AT and FXIIa-C1INH complexes was measured in the blood after incubation with the TiO 2 coated surface, with and without the FXII-specific inhibitor CTI. The result showed that both FXIIa-AT and FXIIa-C1INH complexes were formed in the plasma after contact with the TiO 2 nanoparticle coated surface, but the addition of CTI inhibited the formation of these complexes with ca. 80%, thus confirming an FXIIa-dependent complex formation (data not shown). In the view of the observed strong effects on blood coagulation potentially attractive for wound healing, it was decided to evaluate the use of titania sol in a spray application on burn wounds in vivo in rats that were treated with a pre-heated copper disc, causing burns of second (Groups 1 and 2, untreated and treated respectively) and fourth (Groups 3 and 4, untreated and treated respectively) degree. The rats with untreated wounds were used as controls for both types of incurred damages. Following the healing processes it was possible to note that while the duration of the healing processes in total did not differ appreciably, the dynamics of wound surface reduction was clearly and for more severe wounds even dramatically different. The treatment with a titania sol was apparently resulting in quicker decrease of the exposed wound area, reaching for 4 th degree burns as much as 30% reduction in the middle of the healing period (see Fig. 5A,B). Application of the titania colloid onto the wounds led also to an apparent reduction of the forming scar tissue (see Fig. 6) and its less abnormal appearance. Histological analysis of the healed wound tissues turned to be fully in line with the visual observations. In case of Group 1, i.e. the healing of untreated wounds, the epidermis was not changed, with normal keratinization. In the papillary layer the fibers were thickened and lied more tightly than normal. The number of glands was greatly reduced, in particular, the sebaceous were reduced in size, the sweat epithelium was flattened (with a reduced height of the cells), and hair follicles were isolated (no more than three in sight). The reticular layer was showing more pronounced fibrosis with the activation of fibroblasts, overproduction of the basic substance, and hypervascular focal perivascular leukocyte infiltration (see Fig. 6A,B)). ## Figure 5. (A) Wound surface reduction: Group 1 -healing of untreated 2 nd degree burns, Group 2 -second degree burns treated with nano titania, Group 3 -untreated 4 th degree burns, Group 4-4 th degree burns treated with nano titania. Each value is an average of 3 rats/group (for details, please, see Supplementary Table Histological analysis of the results from Group 2, i.e. healing of second degree wounds treated daily with a titania colloid, gave very exciting and encouraging results. The healed area demonstrated unchanged normal skin structure without any skin structure alterations (see Fig. 6C,D). For the Group 3, where 4 th degree burns were healing without treatment, the epithelium was thinned, sometimes missing, exposing the dense connective tissue that replaced rarely thinned dermis through the entire thickness. Dermal papilla also were smoothed and skin appendages absent. Vascularization of all layers as compared with the sample of Group 4 was reduced. In the deep were observed poorly developed networks of thin-walled vessels with weak perivascular inflammatory infiltration. More fibrosis hypodermis with complete replacement of the connective tissue layer of fat and muscle fibers could be seen (see Fig. 6E,F). Tripe corresponded to 4 th degree burns. For the Group 4, where the 4 th degree burns were treated daily with titania, the epithelium was a thickened prickly layer (acanthosis). Dermal papillae were completely smoothed out (unavailable). Dermal thickness was reduced, fibrosis was more pronounced with increasing abundance of fibers and increase in the number of fibroblasts and the base material. Skin appendages -all glands and hair follicles -were completely absent. The reticular layer was a more developed network of young newly formed blood vessels, arranged vertically. Among them -a moderate diffuse leukocyte infiltration, focal hemosiderin deposits and accumulation of hemosiderophages occurred. Severe fibrosis of hypodermis with almost complete replacement of adipose tissue and muscle fiber atrophy took place. Tripe corresponded to the 3rd degree of the burn (Fig. 6G,H). The main message of the work presented here is the ability of stabilized anatase TiO 2 NPs (applied as a solution) to promote burn wound healing tested in an in vivo rat model. This was evident by the formation of a firm crust of hybrid protein-titania nanocomposite with significantly higher anti-bacterial and anti-inflammatory properties compared to that of untreated controls. The rats did not reveal any abnormal behavior or apparent pathologies after completion of the healing process. To get an insight into possible side effects in treatment with nano titania we have carried out a thorough investigation of the tissues (liver, kidney, spleen, and brain) of treated animals (and untreated ones as reference) with respect to possible retention of titanium (for details, please, see the description below in Methods). It was clearly demonstrated that the content of titanium did not increase in any of the vital organs of the treated rats, staying at the same level as in the control animals (see Fig. 7 and Supplementary S5 and S6). This result appears quite logical in the view that the applied small TiO 2 particles possess, as revealed, very strong affinity to proteins. They are apparently either fixed on the surface inside the blood clot or adsorbed directly on the walls if they come into contact with damaged body fluid vessels. It has to be mentioned that the starting level of titanium content in vital organs was in all the studied cases quite low and did not show any statistically appreciable difference between the test and the control samples. The only case, where the difference on the first glance appeared considerable was for the starting spleen samples. It was in this case actually the control that displayed higher titanium content. The reason of the latter might be that the 4 studied rats in this selection ( 1 / 3 of the starting 12 animals sacrificed on the first day of the experiment) have by accident eaten some titanium-containing material (paper, straw or sand). The difference would most probably not be statistically significant if a bigger group of test animals could be investigated, but this would be not ethically acceptable. Multiple reports are found in the literature describing activities of TiO 2 immobilized to different matrices, e.g., antibacterial activity in vitro , as well as anti-inflammatory and accelerated wound healing activity 36 but to our knowledge, this is the first report of these activities induced by topical application of a solution of TiO 2 NPs. Previously, we have studied innate immunity activation by low concentrations of TiO 2 NPs in whole human blood. We found dose dependent platelet activation, monitored as TAT complex formation, release of thrombospondin-1 (a platelet β-granule protein), and platelet loss. There was substantial activation of the contact/kallikrein system, reflected as generation of FXIIa-AT and FXIIa-C1-INH complexes, and concomitant production and release of the pro-inflammatory chemokines Interleukin (IL-8), Monocyte chemoattractant protein (MCP)-1, Macrophage inflammatory protein (MIP)-1α and MIP-1β (detected using a multiplex analytical panel). All these parameters, except production of MIP-1α and MIP-1β were inhibited by the specific contact system inhibitor CTI 32 . That study confirmed earlier results from our group where we found substantial platelet activation (TAT, platelet loss, release of beta-thromboglobulin [another platelet β-granule protein], generation of FXIIa-AT/ C1-INH, and release of the platelet derived growth factor [PDGF]) induced by planar Ti and TiN surfaces 37 . It should be noted that PDGF was not included in multiplex panel used in 22 and therefore was not detected in that study. The link between contact system activation and the release of the same chemokines (IL-8, MCP-1, MIP-1-α, MIP-1β, and PDGF) in addition to vascular endothelial growth factor (VEGF) was also evident in a previous study where we utilized a number of polymers as a tool to investigate these interactions 38 . In the present work we observe that TiO 2 NPs greatly accelerates blood clotting in vitro in two different models, first by turbidimetry when added to human citrate-plasma, and secondly when evaporated onto polystyrene surfaces which were then incubated with whole human blood. In the latter case, the readouts were TAT generation, platelet loss, and generation of FXIIa-AT and FXIIa-C1-INH complexes, both of which decreased in the presence of CTI. Cutaneous wound healing is a multistep process where coagulation-induced inflammation is a critical first event 39 . During this initial phase, a protective fibrin clot is formed, platelets are activated to contribute to clot formation, but also to release chemokines and growth factors, which recruit and activate neutrophiles and monocytes. Chemokines, which are essential to promote wound healing include PDGF (chemoattractant for neutrophils, monocytes and fibroblast), IL-8 (the major attract and activator for neutrophils), MCP-1 and MIP-1-α, (which, in conjunction promote macrophage response), MIP-1-β (mixed leukocyte recruiter), and VEGF (which promotes angiogenesis at a late stage in the healing process). Since the production and secretion of PDGF, IL-8, MCP-1 and MIP-1-α, MIP-1-β, and VEGF all have been shown to be induced by TiO 2 , in conjunction to contact system (FXII) activation, we conclude that this, at least to a certain extent, explains why the administration of TiO 2 nanopaticles accelerates wound healing. ## Conclusions Colloidal solution of pH-neutral stabilized titania displayed clear trends to enhanced and accelerated blood clotting. This process was not hindered by addition of common anti-coagulants such as heparin. Interaction of titania dispersion with both blood and skin samples resulted in formation of dense films on the surface with uniform micro cracks caused apparently by contraction of the gel on evaporation of the solvent. The biochemical analysis indicated clearly that this was associated, on one hand, with apparent strong blood clotting ability, and, on the other hand, with activation of the contact system resulting in enhanced wound healing effect. Using the colloidal titania for treatment of burn wounds in vivo resulted in apparently quicker reduction of the exposed wound area, while the duration until skin total recovery was comparable with untreated wounds. The most striking effect in application of titania was its logicrent ability to promote restoring of the normal skin structure resulting in the absence of the scar tissue after healing of the 2 nd degree burns and improvement of the scar tissue to the appearance typical of a 3 rd degree burns in the cases of the 4 th degree burn damage. ## Methods Preparation of sol-gel titania. The synthesis of the stable size-uniform titania colloids used in this work was made following the earlier described technique 27 . For producing the initial precursor solution Ti(OEt) 4 (5 mL) was dissolved in anhydrous ethanol (5 mL) and then 1.5 mL of triethanolamine were added on continuous stirring. Hydrolyzing solution (1 mL) was produced by mixing 0.5 M nitric acid, HNO 3 (0.5 mL), with ethanol, EtOH (2.0 mL). The resulting clear transparent yellowish solution contained 120 mg/mL TiO 2 according to TGA measurements. The details of particle characterization are provided in the Supplementary. ## Particle characterization. The size of the initial particles in the aqueous sols was measured by dynamic light scattering (Microtrac instrument). FTIR spectra of sols and gels were recorded with a Perkin-Elmer Spectrum 100 instrument without dilution in a cell fitted with CaF 2 windows. The morphology of the xerogels was studied with a Hitachi TM-1000-μ-DeX 15 kV scanning electron microscope (SEM), and the agglomerate size and crystallinity were studied with a Topcon EM-002 B ultrahigh-resolution analytical electron microscope (TEM). UV/Vis spectra were recorded using a Hitachi U-2001 spectrophotometer. ## Thrombin time test. Lyophilized citrate human plasma and human thrombin (150 NIH units/mg) were obtained from «Kvik» LTD Company, Russia. Thrombin time was measured as a period for clot formation from human citrate plasma with known concentration of plasminogen and fibrinogen. With this aim, 10 mg of lyophilized human plasma was disssolved in 1 mL of triple distilled water (giving final plasminogen concentration-102 μg/mL, fibrinogen concentration -2.8 mg/mL) and then 0.1 mL of the plasma solution was mixed with 1 mL of 0.9% NaCl solution (isotonic). Thrombin solution was prepared by solving 1 mg of thrombin in 1.5 mL of NaCl solution (isotonic). Clotting mixture was prepared in 1 × 0.5 mm plastic cuvette by mixing 1.1 mL plasma solution and 0.1 mL thrombin solution respectively. Turbidity at 315 nm was immediately monitored during 175 sec. For the tests with titania sol, before addition of thrombin solution 10 or 100 µl titania sol has been added (corresponding to final TiO 2 content in the mixture of 1 mg/mL and ca. 10 mg/mL respectively) and compared with the samples diluted with the same volumes of isotonic NaCl. ## Blood sampling. Fresh human blood samples were obtained from healthy volunteers who had not received any medication for at least 10 days prior to donation. Blood samples were collected in an open system with no soluble anticoagulant. In this system, any material that comes into contact with blood is furnished with the Corline heparin surface (Corline Systems AB, Uppsala, Sweden) to prevent material-induced contact activation. Preparation followed the manufacturer´s recommendations. Ethical approval was obtained from the regional ethics committee (Uppsala University Hospital). All methods were carried out in accordance with relevant guidelines and regulations, in particular, complying with the rules summarized by the Swedish Research Council for treatment of human tissue samples summarized at http://www. codex.vr.se/en/manniska4.shtml. ## Skin sampling. The skin samples 1.5-2 mm in diameter were donated by the corresponding author (VGK) and cut by a scalpel from the finger tips. Written informed consent was obtained from all patients involved in the study. The whole blood model. To investigate the influence of the TiO 2 -particles on the blood coagulation cascade in human whole blood a slide chamber model was used, which has been described previously (by Hong et al. 40 ), containing two circular wells with an inner diameter of 17 mm. The test surfaces with TiO 2 -nanoparticles were prepared by adding 0.5 mL of TiO 2 particle suspension (120 mg/mL in ethanol) to polystyrene (PS) microscope slides followed by evaporation overnight. As a reference PS slides were treated the same way, but without the TiO 2 -nanoparticles. The chambers, the control surface (PVC) and the tubes, tips and tubing to be used in contact with the blood were pre-coated with heparin (Corline Systems AB). Blood was drawn from healthy volunteers, who not had received any medication at least 10 days prior to blood donation. The wells were filled with 1.5 mL freshly drawn blood containing 0.5 IU/mL heparin (Leo Pharma) and the test surface was attached with two clips, thereby constituting a lid over the two chamber wells. These devices were then incubated under constant rotation at 30 rpm for 60 min. at 37 °C. After incubation the blood was mixed with EDTA at a final concentration of 10 mM to inhibit further activation of the blood cascade systems. Before centrifugation platelet counts were performed. The blood samples were then centrifuged at 2500 g for 15 min., the plasma was collected and stored at −70 °C for further analysis of coagulation markers. The experiment was repeated four times (different blood donors each time) in duplicates. To one series of experiments 3.5 μM Corn Trypsin Inhibitor (CTI; Enzyme Research Laboratories), which is a specific FXIIa inhibitor, was added to the blood prior to incubation with the surfaces. Platelet count. The number of platelets was analyzed in the blood samples before and after incubation with the test surfaces using a Sysmex XP-300 Hematology Analyzer (Sysmex Corp.). Platelet count was calculated as the remaining amount as compared to the initial sample (before incubation in the chambers) and was expressed as mean percent of initial ± SEM. Thrombin-Antithrombin complexes (TAT) ELISA. Plasma levels of TAT were analyzed by a conventional sandwich ELISA. The plasma samples were diluted in normal citrate-phosphate-dextrose plasma. The TAT complexes were captured by an anti-human thrombin antibody (Enzyme Research Laboratories) and detected with an HRP-conjugated anti-human AT antibody (Enzyme Research Laboratories). As standard pooled human serum diluted in in normal citrate-phosphate-dextrose plasma was used. All values were given in μg/L. ## Contact activation complexes. For the detection of FXIIa-antithrombin (AT) and FXIIa-C1-inhibitor (C1INH) complexes in the plasma samples a standard sandwich ELISA described by Sanchez et al. 41 was used. Microtiter plates were coated with anti-human FXIIa antibodies (Enzyme Research Laboratories) and captured complexes were subsequently detected with either biotinylated anti-human AT (Dako) or biotinylated anti-human C1INH (Enzyme Research Laboratories) followed by HRP-conjugated streptavidin (GE Healthcare). Standard solutions were diluted in normal plasma. All measured values are given in nmol/L. In vivo Investigation of burn wound healing properties. Male Hooded rats (body weight range 200-250 g) were used for the study. Animals were acclimatized under standard animal laboratory condition for 7 days prior to the experiment. All experiments were approved by institutional animal ethical committee (Ivanovo State Medical Academy, Russia, Protocol No. 2 from 06.04.2015) and are in agreement with the guidelines for the proper use of animals for biomedical research 42 . Animals were divided into 4 groups, each consisting of 3 rats: I group -rats with 10 sec treatment with heated disc; II group -rats with 10 sec treatment with heated disc and healing titania; III group -rats with 20 sec treatment with heated disc; IV group -rats with 20 sec treatment with heated disc and healing titania. All animals survived and did not suffer weight loss within standard deviation until the last day of the experiment (day 19). Animals were anesthetized with ketamine (dose 60 mg/kg), acting as both sedative and long-term pain-killer agent 43 , by intraperitoneal injection, the dorsal hair was shaved and disinfected. Burns were made 1 cm diameter copper disc preliminary heated up to 300 °C. For groups II and IV the materials were applied on excised burns. The burns were treated daily with 0.1 mL of prepared titania solution. Wound sizes were measured daily until the healing is complete. The wound outline was transferred to transparent films and scanned with an Epson Perfection 2480 scanner. The wound area was calculated with ImageJ 1.30 v. software. The percentage wound reduction was calculated according to the following formula: where C n is the percentage of wound size reduction, S o is initial wound size, S n is wound size on respective day. The rats were kept in individual cages 20 × 30 cm 2 area and had free ability to motion and access to both food and water. As the wounds were located in the dorsal area, there was no risk that the animals should lick or bite their own wounds. ## Histological analysis. Fragments of skin with scar excised and completely fixed in 10% formalin solution during 24 hours. After routine gynecological wiring samples were poured into paraffin. 20 slices with 5 μm thickness were prepared and stained with hematoxylin and eosin from each paraffin block. Chemical analysis of tissue samples. Analysis of the content of titania in the organs was carried out according the following procedure. Animals (in the experiment, 12 rats were used) on 1, 14 and 28 days respectively after wound healing with titania were anesthetized with isoflurane and were killed by cervical dislocation and organs (liver, kidney, spleen and lung) were collected and weighed immediately after killing of the animals. Dissolution of organs was carried out with a mixture of concentrated sulfuric and nitric acids. Completeness of dissolution was achieved by organs heating in heat-resistant glasses. 1 ml of concentrated HNO 3 and 3 ml of concentrated HCl were added to the resulting syrup-like transparent solutions with following transferring to 25 ml volumetric flasks using distilled water. The titanium concentration was determined by atomic-absorption spectroscopy with inductively coupled plasma. The studies were carried out on HORIBA Jobin Yvon ULTIMA 2. Three rats without titania treatment were used as a control sample for organs. The in vivo experiment was repeated for another group of animals and in this case liver, and brain tissues were removed from one representative animal per group, cut into pieces about 0,2 g that were weighed and then dissolved in 3 ml of aqua regia. The pH after dissolution was adjusted to 3.0 by addition of 1.0 M NH 3 solution. The produced liquids were analyzed with ICP-AES Spectro Cirros CCD Instrument, Kleve, Germany. Two of the authors, GAS and VGK declare their involvement in the activities of the CaptiGel AB company, Sweden, developing metal oxide colloids for environmental and biomedical applications.

chemsum

{"title": "Dispersion of TiO2 nanoparticles improves burn wound healing and tissue regeneration through specific interaction with blood serum proteins", "journal": "Scientific Reports - Nature"}

controllable_stereoinversion_in_dna-catalyzed_olefin_cyclopropanation_<i>via</i>_cofactor_modificati

## Abstract: The assembly of DNA with metal-complex cofactors can form promising biocatalysts for asymmetric reactions, although catalytic performance is typically limited by low enantioselectivities and stereocontrol remains a challenge. Here, we engineer G-quadruplex-based DNA biocatalysts for an asymmetric cyclopropanation reaction, achieving enantiomeric excess (ee trans ) values of up to +91% with controllable stereoinversion, where the enantioselectivity switches to À72% ee trans through modification of the Fe-porphyrin cofactor. Complementary circular dichroism, nuclear magnetic resonance, and fluorescence titration experiments show that the porphyrin ligand of the cofactor participates in the regulation of the catalytic enantioselectivity via a synergetic effect with DNA residues at the active site. These findings underline the important role of cofactor modification in DNA catalysis and thus pave the way for the rational engineering of DNA-based biocatalysts. ## Introduction Cyclopropane motifs feature in many natural products and medicinal agents 1 which constitute versatile intermediates for the total synthesis of therapeutic compounds. 2 As these compounds are in high demand, signifcant effort has been devoted to the development of cyclopropane synthesis, in particular via the use of hemoprotein enzymes engineered via directed evolution. Cytochrome P450 (ref. 3 and 4 ) and myoglobin enzymes have been evolved to catalyze asymmetric olefn cyclopropanations with excellent performance. This method has been applied to the synthesis of drug molecules 15,16 and natural product scaffolds. 17,18 Despite this progress, the developed biocatalytic protocols are generally restricted to protein enzyme engineering. 19 The discovery of the catalytic functions of nucleic acids 20,21 has expanded the breadth of biocatalytic protocols to include RNA and DNA catalysts, initiating the pursuit of nucleic acidbased enzymes. DNA possesses inherent advantages as a catalyst. A catalytic sequence can be entirely identifed from a random sequence population and folds into its practical tertiary structure spontaneously. A number of asymmetric reactions, especially Lewis acid-catalyzed reactions, have been successfully realized using DNA-based biocatalysts resulting in remarkable performances. Recently, the Roelfes 32 and Sen 33 groups expanded the scope of reactions catalyzed by DNA-based biocatalysts to include olefn cyclopropanation. Although the enantioselectivities achieved were moderate, this has paved the way for DNA-catalyzed carbene transfer reactions. Therefore, designing a DNA-based biocatalyst that catalyzes cyclopropanation in high enantiomeric excess (ee) remains a challenge, especially to achieve an ee greater than 90%. 34 DNA catalysis is opening a promising avenue for biosynthesis, but the catalytic scope and performance of DNA catalysts still need improvement in comparison with catalysts based on protein enzymes. Considering the great role played by the cofactor ligand in the frst-coordination-sphere in catalytic reactions, cofactor modifcation, which is as powerful as directed evolution but seriously disregarded in the feld of DNA catalysis, is introduced for the development of DNA-based biocatalysts. Here, we report a cyclopropanation reaction catalyzed by a G-quadruplex (G4)-Fe-porphyrin biocatalyst that results in enantioselectivity as high as 91%. By tuning the N-methyl position of Fe-porphyrin from the para-to the ortho-position, the catalytic enantioselectivity of the reconstituted G4-Fe-porphyrin biocatalyst reverses to 72% ee trans . Complementary spectral, nuclear magnetic and isothermal titration characterization studies reveal that the stereo-divergence of the product mainly arises from the participation of the porphyrin ligand in the regulation of the enantioselectivity. This work succeeds in diversifying the functionality of DNA-based biocatalysts via cofactor modifcation, highlighting the great potential for cofactor modifcation in the feld of DNA-based biocatalyst engineering. ## Results and discussion Fig. 1a illustrates the design of the G4-based biocatalysts. Three Fe-meso-tetra-(N-methylpyridyl)porphyrins with para-(FeTM-PyP4), meta-(FeTMPyP3), and ortho-(FeTMPyP2) N-methyl substituents were chosen as parallel cofactors to investigate the effect of the frst-coordination-sphere on the catalytic performance. The non-covalent binding of the cofactors with mA9A G4 (d[G 2 T 2 G 2 TGAG 2 T 2 G 2 A]), a thrombin binding aptamer (TBA) variant, formed the G4-based biocatalysts. The assembled biocatalysts were then tested in a reaction between styrene and ethyl diazoacetate (EDA), which results in a chiral cyclopropane product (Fig. 1b). Table 1 lists the results obtained using the Fe-porphyrins and their corresponding G4 biocatalysts. The use of the free FeTMPyPn (n ¼ 4, 3, 2) cofactor as the catalyst led to relatively low activities and no chiral induction (Table 1, entries 1-3). The biocatalysts (mA9A-FeTMPyPn, n ¼ 4, 3, 2) assembled from FeTMPyPn and mA9A G4 signifcantly improved the catalytic activities with turnover frequencies (TOF) increased about 10fold when compared to those of the free FeTMPyPns. More surprisingly, mA9A-FeTMPyP2 induces the inversion of the enantioselectivity relative to that of mA9A-FeTMPyP4 from +74% to 46% (Table 1, entries 4-6). To further understand the complementary chiral induction mechanism, circular dichroism (CD), nuclear magnetic resonance (NMR), fluorescence titration and other characterization methods were then performed. The CD spectra in Fig. 2a show that the antiparallel G4 signatures of mA9A, featuring two positive peaks at 245 nm and 295 nm, and one negative peak at 265 nm, are still present in the mA9A-FeTMPyPn (n ¼ 4, 3, 2) catalysts. But of note is that the CD spectrum of mA9A-FeTMPyP2 shows a specifcally induced CD signal (ICD) at 420 nm. Only when an asymmetric conformation is formed can an ICD signal be induced. DNA does not absorb at 420 nm, but FeTMPyP2 does (Fig. 2b). Therefore, the ICD signal can be attributed to FeTMPyP2, whose planar symmetry is broken due to the interaction with mA9A. The ultraviolet-visible (UV) absorption spectra (Fig. 2b) of mA9A-FeTMPyPn further support the CD results. The spectra of FeTMPyPn (n ¼ 4, 3, 2) exhibit broad Soret absorption bands at around 420 nm. As the N-methyl group varies from the para-to the ortho-position, the Soret band shows a blue shift, indicating that FeTMPyP2 has decreased electronic conjugation relative to the other porphyrins. Due to the steric hindrance of the 2-Nmethyl group, the bond between the pyridine ring and the porphyrin ring of FeTMPyP2 has a tendency to rotate, causing the pyridine ring and the porphyrin ring to be non-co-planar (Fig. 2b, right inset), which explains the generation of an ICD signal in mA9A-FeTMPyP2. When well-folded mA9A is added to FeTMPyPn, the Soret bands frst fall to a minimum and then slightly rise for all cofactors. This suggests that FeTMPyPn (n ¼ 4, 3, 2) have a similar external stacking mode on mA9A, rather than intercalating between two G-quartets. Moreover, binding with FeTMPyP4 and FeTMPyP3 makes mA9A more stable according to UV melting experiments (Fig. S1 †), while binding with FeTMPyP2 does not. To obtain further structural information on the mA9A-FeTMPyPn catalysts, we conducted an NMR characterization of the systems. The NMR spectrum of mA9A shows seven discrete peaks over the range of 11.5-12.5 ppm (Fig. 2c), which can be assigned to the eight guanine imino protons (H1) of the Gquartet. These signals indicate the formation of a two-layer antiparallel G4 structure in the potassium phosphate buffer (Fig. 2d), like that of TBA. 48,49 NMR titration experiments (Fig. 2c) where mA9A is added to the three FeTMPyPn (n ¼ 4, 3, 2) cofactors show similar line broadening and reductions in peak intensity. However, there are two adjacent peaks around 12.0 ppm (peaks labelled with *) with different relative declines, indicating that the coordination modes between mA9A and the different Fe-porphyrin cofactors vary slightly. Moreover, the addition of FeTMPyP2 causes an obvious upfeld peak-shift, which is attributed to the strong electronic shielding effect arising from the porphyrin ligand stacking upon the Gquartet. 50 Therefore, in combination with the CD and UV-vis results, we fnd that FeTMPyP2 adjusts to an optimal conformation through bond rotation and distortion, thereby forming a tight p-p stacking mode with mA9A and creating a strong electronic shielding effect on the imino protons of the Gquartet. A fluorescence-based binding assay was then implemented to locate the catalytic sites. 51 Since UV titration had determined the external stacking mode of the FeTMPyPn (n ¼ 4, 3, 2) cofactors on mA9A, the 5 0 end (5 0 FAM-mA9A) and loop 1 (int 0 FAM-mA9A) of the two G-quartets of mA9A were labelled separately with the fluorophore 5-carboxyfluorescein (FAM). The fluorescence of FAM can be suppressed by the presence of FeTMPyPn (Fig. 3a and S2 †). By adding FeTMPyPn dropwise to the FAM-labelled mA9A, we obtained the titration quenching curves (Fig. S3 †). The ftted apparent equilibrium dissociation constants (Kd app ) at the different FAM-labelled sites, shown in Fig. 3b, indicate that all the FeTMPyPn (n ¼ 4, 3, 2) cofactors prefer binding at the 5 , 3 0 end of the G-quartet. Considering that the ratio of mA9A and FeTMPyPn is greater than 1 : 1 when assembling the DNA-based biocatalyst and that the ee trans values for the catalytic cyclopropanation can be maintained at the highest level, the preferential binding site of FeTMPyPn with mA9A is regarded as the active center for chiral regulation. Therefore, according to the fluorescence-based binding assay, FeTMPyPn binding at the 5 0 , 3 0 -end of the G-quartet of mA9A constructs the active catalytic site of mA9A-FeTMPyPn. The Kd app of FeTMPyP2 is the lowest at 38 nM, compared to 70 nM for FeTMPyP3 and 52 nM for FeTMPyP4, indicating that FeTMPyP2 binds the strongest with mA9A. Isothermal titration calorimetry (ITC) provides information on intermolecular interactions by recording the heat discharged or consumed during a bimolecular reaction. Fig. 3c shows that there are several distinct exothermic processes during the titration of mA9A with FeTMPyP4 and FeTMPyP3, indicating complicated multiple binding behaviours between FeTMPyP4, FeTMPyP3 and mA9A. As explained in the above section, although mA9A enables the binding of multiple iron porphyrins, the strongest binding sites were the catalytic sites for the chiral regulation of the cyclopropanation reaction. The titration curves were ftted using a sequential binding sites model to calculate the binding parameters. The highest affinities between FeTMPyPn and mA9A were quantifed to be 33 nM (n ¼ 4), 50 nM (n ¼ 3) and 14 nM (n ¼ 2), coinciding with the trend of Kd app as measured using fluorescence titration. This further supports the conclusion that the 5 0 , 3 0 -end of the G-quartet of mA9A constructs the active catalytic sites. Given that FeTMPyP2 has flexible peripheral groups (according to UV spectra), the tight assembly with mA9A can be attributed to the conformational transition of the TMPyP2 ligand, which is also confrmed by the CD spectra. NMR, UV and fluorescence titration studies indicate that fne-tuning the N-methyl position of the cofactor does not change the preference of FeTMPyPn for binding at the 5 0 , 3 0 -end of the G-quartet of mA9A. This suggests that a similar active DNA pocket for cyclopropanation catalysis is provided by all the mA9A-FeTMPyPn (n ¼ 4, 3, 2) catalysts (Fig. 4a). The ligands TMPyP4 and TMPyP3 are planar symmetric molecules that are unable to induce chirality in the catalytic process of the mA9A-FeTMPyP4 and mA9A-FeTMPyP3 catalysts. It is the deoxynucleotide residues at the 5 0 , 3 0 -end of the G-quartet of mA9A that hold the iron porphyrin carbene (IPC) intermediate in a certain orientation and defne the confguration of the product (Fig. 4b). Nevertheless, mA9A-FeTMPyP3 shows a similar but lower enantioselectivity than mA9A-FeTMPyP4. Considering the similar binding strengths of FeTMPyP3 with the two G-quartets of mA9A (Fig. 3b), the reduction in enantioselectivity can be reasonably attributed to the multi-site binding behaviour of FeTMPyP3. In contrast to mA9A-FeTMPyP4 and mA9A-FeTMPyP3, the symmetry breaking of the porphyrin ring in TMPyP2 allows it to participate in chiral regulation. This, in synergy with the deoxynucleotide residues, induces the formation of the cyclopropane product with the opposite confguration to that catalyzed by mA9A-FeTMPyP4 (Fig. 4b). To extend the substrate scope of the mA9A-FeTMPyPn catalyzed cyclopropanation reaction, a series of olefns and diazoesters were investigated (Table 2). All three mA9A-FeTMPyPn (n ¼ 4, 3, 2) catalysts show obvious substrate specifcities. The catalytic enantioselectivities of mA9A-FeTMPyP4 and mA9A-FeTMPyP3 toward olefn substrates with substituents on the phenyl group are reduced (entries 1 vs. 2-6). The enlargement of the diazoester functional group from ethyl (Et) to -CCH 3 (i-Pr) 2 (i-Pr is the abbreviation for isopropyl) enables a signifcant enhancement in the ee trans values to 91% (entries 1 vs. 8-10). However, the use of a -CH(Cy) 2 (Cy is the abbreviation for cyclohexyl, entry 11) group does not result in an enhancement. Several studies have reported that the modifcation of diazoester substituents causes a signifcant impact on the activities and selectivities of carbene transfer reactions. 6,52 The G4-Feporphyrin-catalyzed cyclopropanation has been characterized to proceed through a catalytic IPC intermediate. 34 Diazoester reagents attack the active [Fe] center to form an IPC intermediate and release one molecule of N 2 . An appropriate substituent on the diazoester reagent can coordinate with the deoxynucleotide residues (especially dA9) through hydrophobic interactions to directly determine the conformation and properties of the IPC intermediate. Therefore, a -CCH 3 (i-Pr) 2 substituent promotes IPC convergence to a single well-defned orientation, and the steric hindrance of the deoxynucleotide residues allows one face of the IPC to be exposed to the olefn, while keeping the other inaccessible, resulting in high enantioselectivity. For mA9A-FeTMPyP2, variation in the substituents on the phenyl group of the olefn increases the enantioselectivity to 72% ee trans (entry 6), but the catalytic cyclopropanation activities are generally lower than those of the other two biocatalysts. Although the three mA9A-FeTMPyPn (n ¼ 4, 3, 2) catalysts show different responses to substrate variation, they are all trans product-selective with trans/cis ratios of more than 86 : 14, and almost nonselective toward the R1substituted olefn substrate (entry 7). ## Conclusions In conclusion, stereo-divergence of G4 biocatalyst-catalyzed olefn cyclopropanation was achieved via cofactor modifcation. By tuning the N-methyl substituent of the porphyrin ligand in the cofactor from the para-to the ortho-position, the selfassembled G4-Fe-porphyrin biocatalysts are able to switch the enantioselectivity of the reaction from +91% to 72% ee trans . CD, NMR, ITC, and other characterization studies reveal that the porphyrin ligand cooperating with the deoxynucleotide residues gives the IPC intermediate a single well-defned confguration and results in a specifc enantiopreference. This fnding is down to the rational design of DNA-based biocatalysts through cofactor modifcation, a method which serves as an effective way to regulate the catalytic performance of DNA-based biocatalysts.

chemsum

{"title": "Controllable stereoinversion in DNA-catalyzed olefin cyclopropanation <i>via</i> cofactor modification", "journal": "Royal Society of Chemistry (RSC)"}

comparison_of_rhenium–porphyrin_dyads_for_co₂_photoreduction:_photocatalytic_studies_and_

## Abstract: We report a study of the photocatalytic reduction of CO 2 to CO by zinc porphyrins covalently linked to [Re I (2,2 0 -bipyridine)(CO) 3 L] +/0 moieties with visible light of wavelength >520 nm. Dyad 1 contains an amide C 6 H 4 NHC(O) link from porphyrin to bipyridine (Bpy), Dyad 2 contains an additional methoxybenzamide within the bridge C 6 H 4 NHC(O)C 6 H 3 (OMe)NHC(O), while Dyad 3 has a saturated bridge C 6 H 4 NHC(O)CH 2 ; each dyad is studied with either L ¼ Br or 3-picoline. The syntheses, spectroscopic characterisation and cyclic voltammetry of Dyad 3 Br and [Dyad 3 pic]OTf are described. The photocatalytic performance of [Dyad 3 pic]OTf in DMF/triethanolamine (5 : 1) is approximately an order of magnitude better than [Dyad 1 pic]PF 6 or [Dyad 2 pic]OTf in turnover frequency and turnover number, reaching a turnover number of 360. The performance of the dyads with Re-Br units is very similar to that of the dyads with [Re-pic] + units in spite of the adverse free energy of electron transfer. The dyads undergo reactions during photocatalysis: hydrogenation of the porphyrin to form chlorin and isobacteriochlorin units is detected by visible absorption spectroscopy, while IR spectroscopy reveals replacement of the axial ligand by a triethanolaminato group and insertion of CO 2 into the latter to form a carbonate. Time-resolved IR spectra of [Dyad 2 pic]OTf and [Dyad 3 pic]OTf (560 nm excitation in CH 2 Cl 2 ) demonstrated electron transfer from porphyrin to Re(Bpy) units resulting in a shift of n(CO) bands to low wavenumbers. The rise time of the charge-separated species for [Dyad 3 pic]OTf is longest at 8 (AE1) ps and its lifetime is also the longest at 320 (AE15) ps. The TRIR spectra of Dyad 1 Br and Dyad 2 Br are quite different showing a mixture of 3 MLCT, IL and charge-separated excited states. In the case of Dyad 3 Br, the charge-separated state is absent altogether. The TRIR spectra emphasize the very different excited states of the bromide complexes and the picoline complexes. Thus, the similarity of the photocatalytic data for bromide and picoline dyads suggests that they share common intermediates. Most likely, these involve hydrogenation of the porphyrin and substitution of the axial ligand at rhenium. ## Introduction Much of the world's energy need is satisfed by the combustion of fossil fuels. The processes associated with generating energy in this way release gigatons of CO 2 into the atmosphere every year, contributing to climate change. 1 In addition to environmental unsustainability, the fossil fuels are essentially fnite as they require geological timescales to form. The sun provides a clean source of energy that can satisfy our energy demands now and in the future. 2 It is critical therefore, that we develop systems that can harvest visible light and store the energy as chemical fuel: systems that perform artifcial photosynthesis. Supramolecular assemblies containing components capable of light harvesting and catalysis can in principle perform arti-fcial photosynthesis. There are several examples of this type of system for water oxidation, 3,4 proton reduction, and CO 2 reduction. For supramolecular assemblies to be active for photocatalytic redox reactions, they must be designed such that photoinduced electron transfer is favourable and such that charge separation lifetimes are sufficiently long for the catalytic reaction to occur prior to recombination. Photocatalytic CO 2 reduction to CO is an attractive choice because CO 2 is consumed and CO can subsequently be converted into energy-dense hydrocarbon fuels. CO is also an industrial feedstock and a fuel in its own right. 17 Diimine complexes of rhenium have received much attention since the discovery, reported in 1983, that they are active and selective photo-and electro-catalysts for CO 2 reduction to CO. 18 In the context of solar fuels, the rhenium complexes are limited because they cannot utilize much of the solar spectrum and turnover numbers of CO (TON CO ) are low due to catalyst instability. 19 Introduction of a sensitizer molecule can improve visible light absorption. The use of lower energy radiation and transferring the role of light absorption to another molecular unit will remove pathways of photo-degradation for the rhenium complex and increase stability. Indeed high TON CO have been reported for dyads consisting of rhenium catalysts covalently linked to ruthenium bipyridyl units. 10, Sensitizing dyes have also been used in association with Re catalysts supported on TiO 2 . 32 Zinc porphyrins are good candidates for sensitization for several reasons. 33 They show intense absorption in the visible spectrum, in particular the Q bands centred around 560 nm. 11 The excited state redox potential of zinc porphyrin can be tuned to be negative with respect to the ground state of rhenium diimine complexes. 34 The porphyrin motif is closely related to chlorophylls 35 that are utilized in natural photosynthesis for light harvesting and charge separation. 36 The visible light absorption and photoinduced electron-transfer ability of zinc porphyrins has led to high efficiencies in dye-sensitized solar cells. 37 We and others recently demonstrated that zinc porphyrins can sensitize rhenium diimine complexes for CO 2 reduction to CO with long-wavelength visible light. 11,38,39 Rhenium bipyridine tricarbonyl complexes have been used extensively for photocatalytic and electrocatalytic CO 2 reduction. 18, There have been important recent developments in understanding the mechanism of such reactions. Kubiak has tracked reduced intermediates and their reactivity toward CO 2 . 45,46,51,52 Ishitani has shown that the usual sacrifcial reducing agent, triethanolamine (TEOA), coordinates to rhenium by deprotonation to form a rhenium alkoxide of the type ReOCH 2 CH 2 N(CH 2 CH 2 OH) 2 which can insert CO 2 to form a rhenium carbonate derivative. 53 Inoue et al. have used mass spectrometry to study reduction of ReCl(4,4 0 dimethyl-2,2 0bipyridine)(CO) 3 with triethylamine. 54 They demonstrate that CO 2 displaces a solvent molecule in the one-electron reduced complex to form a Re-CO 2 radical which is then protonated to form a Re-COOH radical cation. Thus there is good evidence of direct CO 2 coordination in the absence of TEOA and a complete cycle has been postulated for the electrochemical reaction. 52 For the photochemical reaction with TEOA, the new evidence indicates CO 2 insertion into the alkoxide complex, but the subsequent steps remain undefned. Closely related zinc porphyrins bound to rhenium carbonyls have been investigated for photo-induced charge separation. 55,56 Iron porphyrins have also been used successfully as electrocatalysts for CO 2 reduction. There are several photophysical investigations into porphyrins linked to metal carbonyl complexes, 38, but investigations connecting photophysical data and photocatalytic activity across a range of catalyst structures are scarce. 10,20 Pump-probe time resolved infrared spectroscopy (TRIR) is an invaluable technique for measuring excited state dynamics in this kind of assembly. 34, Metal carbonyl n(CO) stretches can be observed with high intensity in a region of the infrared where few other vibrational bands are present. Crucially, they are very sensitive to the electron density on the metal centre and can be used to monitor charge transfer. In our previous investigations of long-wavelength (l > 520 nm) photocatalytic CO 2 reduction with Re complexes covalently linked to zinc porphyrins, we investigated [Dyad 1 pic]PF 6 (Fig. 1) with a C 6 H 4 NHCO bridge. 11 To increase catalytic activity we sought to reduce the rate of charge recombination by increasing the separation between donor and acceptor 74,75 by inclusion of a methoxybenzamide molecular spacer ([Dyad 2 pic]OTf), and this dyad indeed displayed higher catalytic activity. 11 We now report the synthesis and catalytic activity of a new dyad with a C 6 H 4 NHCOCH 2 saturated molecular spacer [Dyad 3 pic]OTf (Fig. 1). We also compare the catalytic performance of these three cationic dyads to those of the corresponding neutral bromide complexes Dyad 1 Br, Dyad 2 Br and Dyad 3 Br. To our surprise the catalytic performance of each of the bromide complexes is very similar to that of the corresponding cationic dyads. This is intriguing as the bromide dyads do not undergo photoinduced reaction with intermolecular electron donors and their reduction potentials are signifcantly more negative than those of the cationic complexes. 81,82 In our previous investigations of [Dyad 1 pic]OTf we showed by TRIR spectroscopy that charge separation occurs within a few ps and the lifetime of the charge-separated state is of the order of tens of ps. We now report on the TRIR spectroscopy of [Dyad 2 pic]OTf, [Dyad 3 pic]OTF and that of all three bromide complexes. We also show by TRIR spectroscopy that the excited state behaviour of the neutral bromide complexes is very different from that of the cationic picoline complexes. We propose mechanisms that can reconcile the different excited state and electrochemical behaviour with the similar photocatalysis. ## General procedures Chemicals were obtained from the following suppliers: diisopropylamine, 2.5 M n-butyl lithium in hexanes (Acros); EDTA, AgOTf, 4,4 0 -dimethyl-2,2 0 -bipyridine, sodium sulfde, celite 512 medium, Et 3 N, 2-chloro-methylpyridinium iodide, triethanolamine, anhydrous DMF, methyl chloroformate, copper(II) acetate (Aldrich); 3-picoline (BDH Chemicals); CO 2 -CP-grade with 5% CH 4 (or 1% CH 4 ) (BOC); Na 2 SO 4 , Na 2 CO 3 , NaHCO 3 , NaOH, HCl, KOH, ammonium hydroxide (Fisher); Zn(OAc) 2 $H 2 O, CH 3 CO 2 Na (Fisons). Solvents for general use were obtained from Fisher. Solvents were dried by refluxing over sodium wire (C 6 H 6 , THF, toluene) or over CaH 2 (CH 2 Cl 2 ). DMF was dried using a Pure Solv 400-3-MD (Innovative Technology). For TRIR experiments, CH 2 Cl 2 (99.9%, Merck) was distilled under an inert atmosphere of Ar from calcium hydride and anhydrous THF ($99.9%, inhibitorfree, Sigma Aldrich) was used as supplied and stored in a glove box. CD 2 Cl 2 , CD 3 OD, DMSO-d 6 and CDCl 3 were used as obtained (Aldrich) and THF-d 8 was dried over potassium. Diisopropylamine was distilled from sodium hydroxide. Methyl chloroformate was distilled prior to use. n-BuLi was titrated against nbenzylbenzamide prior to use. Routine separation of porphyrins by flash chromatography was performed on a CombiFlash Rf system using 24 g RediSep Rf silica columns (Teledyne Isco), and dry-loading the samples on silica (Fluka). NMR spectroscopy. NMR spectra were run on a Bruker AV500 ( 1 H at 500 MHz) spectrometer or Bruker ECS400 (400 MHz). 1 83 IR and UV/vis absorption and emission. IR spectra were recorded on a Mattson RS FTIR instrument, averaging 64 scans at resolution 2 cm 1 . ATR-IR spectra were an average of 32 scans. UV/visible absorption spectra were measured using an Agilent 8453 spectrometer. Steady state emission spectra were measured using a Hitachi F-4500 fluorimeter. The fluorescence was taken against a ZnTPP reference for the bromide complexes and against the individual dyad ligand ZnTPP-link-Bpy for the picoline complexes. Time-resolved emission was measured with an Edinburgh Instruments FLS980 equipped with a 560 nm pulsed LED (EPLED 560, pulsewidth 1.5 ns) and a red PMT detector. All samples were either degassed by three freezepump-thaw cycles or de-aerated by purging the sample with Ar. Correction was applied for instrument response. All absorption and emission measurements were made in 10 10 mm quartz cuvettes. Mass spectrometry. ESI mass spectra were recorded on a Bruker micrOTOF instrument with a sample flow rate of 0.2 mL min 1 , nebuliser gas pressure of 1.5 bar, dry gas flow of 8 L min 1 and a dry gas temperature of 180 C. EI mass spectra were run on a Waters GCT premier with a source temperature of 180 C, electron energy of 70 eV and a trap current of 200 mA. Some compounds and the reaction mixture ESI mass spectra were run on a Bruker Esquire 6000 via direct infusion using a syringe pump at 240 mL min 1 . Nebuliser gas and dry gas flows and temperatures were optimised for each individual sample along with the spray voltage. m/z values are quoted for 64 Zn, 185 Re and 79 Br. Electrochemistry. Cyclic voltammetry was performed in CH 2 Cl 2 with 0.1 M [Bu 4 N][PF 6 ] (TBAP) electrolyte. The setup comprised reference electrode (Ag/AgCl, 3 M NaCl), working electrode (platinum disc) and counter electrode (platinum wire). Ferrocene was used as internal standard. All scans were made at 100 mV s 1 . Cyclic voltammetric experiments used a BASi Epsilon potentiostat with C3 cell stand. X-ray diffraction. X-ray diffraction data for [Dyad 1 pic]PF 6 and 5-[4-[(2-methoxy-4-nitro-phenylcarbonyl)-amino]phenyl]-10,15,20-triphenyl porphyrin were collected at 110 K on an Agilent SuperNova diffractometer with MoKa radiation (l ¼ 0.71073 ). Data collection, unit cell determination and frame integration were carried out with "CrysalisPro". Absorption corrections were applied using crystal face-indexing and the ABSPACK absorption correction software within CrysalisPro. Structures were solved and refned using Olex2 implementing SHELX algorithms. [Dyad 1 pic]PF 6 was solved using SUPER-FLIP 84 whereas 5-[4-[(2-methoxy-4-nitro-phenylcarbonyl)-amino] phenyl]-10,15,20-triphenyl porphyrin was solved using direct methods within the SHELXS algorithm. Structures were refned by full-matrix least squares using SHELXL-97. All non-hydrogen atoms were refned anisotropically. Carbon-bound hydrogen atoms were placed at calculated positions and refned using a "riding model". For [Dyad 1 pic]PF 6 , one of the phenyl groups on the porphyrin ring was disordered and modelled in two positions with refned occupancies of 0.817 : 0.183 (12). The ADP of equivalent carbons in the disordered phenyl were constrained to be equal, e.g. C51 & C51A. The hexafluorophosphate was disordered over two sites. For one of these, the phosphorus was centred on a special position and for the other, the occupancy was 50% with a dichloromethane of crystallisation occupying the site at other times. In addition to the ordered dichloromethanes of crystallisation, the crystal also contained some disordered solvent, believed to be a mix of hexane and dichloromethane for which a suitable discrete model could not be obtained. This was accounted for using a solvent mask; this space had a volume of 213 3 and predicted to contain ca. 17 electrons. The large residual density peaks are believed to provide evidence for twinning but a suitable method for modelling this was not found. For 5-[4-[(2-methoxy-4-nitro-phenylcarbonyl)-amino]phenyl]-10,15,20-triphenyl porphyrin, the NH hydrogen was located by difference map. The crystal also contained dichloromethanes of crystallisation. One was partially occupied and was modelled with an occupancy of 0.1875; the carbon of this CH 2 Cl 2 was restrained to be approximately isotropic. The other was fully occupied but disordered and modelled with the carbon in two different positions with relative occupancies of 0.814 : 0.186 (12). Crystallographic parameters are listed in the ESI. † Ultrafast infrared experiments. Picosecond time-resolved infrared (TRIR) spectra were obtained using purpose-built equipment based on a pump-probe approach. Details of the equipment and methods used for the TRIR studies have been described previously, 85,86 a brief description of which is given here. The pump beam (560 nm, ca. 150 fs) and tunable probe beam (180 cm 1 spectral band width, ca. 150 fs) were generated from a commercial Ti:sapphire oscillator (MaiTai)/regenerative amplifer system (Spitfre Pro, Spectra Physics). The mid-IR probe was detected using a 128-element HgCdTe array detector (Infrared Associates) typically with a resolution of ca. 4 cm 1 . All the solutions for analysis were prepared under an inert atmosphere of Ar, degassed by three freeze-pump-thaw cycles and put under Ar. [Dyad 2 pic]OTf and [Dyad 3 pic]OTf were run in CH 2 Cl 2 at 1.5 mM and 1.0 mM respectively, with a path length of 0.5 mm. Dyad 1 Br, Dyad 2 Br and Dyad 3 Br were run in THF at 1 mM with a path length of 0.25 mm. A Harrick solution cell with CaF 2 windows was used and 20 mL of solution was continuously circulated during the measurements. Photocatalysis. Photocatalysis was performed in a custommade cell 67 comprised of a 10 10 mm quartz cuvette with a headspace of a minimum volume of 10 mL. Above the headspace was a ground glass joint, which was sealed with a size 21 septum. Samples were taken through this septum for GC analysis. The headspace had a sidearm, isolated by a Young's tap, joining it to a gas phase IR cell with CaF 2 windows. The IR cell was connected to a vacuum joint via a second Young's tap. The IR cell was put under vacuum. At the end of a catalytic run the headspace was opened to the IR cell and the gas produced from the reaction would be drawn through and could be monitored by IR spectroscopy. The concentration of catalytic solution was typically 0.05 mM, making the absorbance of the porphyrin Q band at 560 nm, Q(1, 0), ca. 1 by UV/vis spectroscopy. A 10 mL stock solution of 0.25 mM catalyst in DMF would typically be made. These stock solutions allowed the catalysts to be weighed out in amounts greater than 1 mg. They were stored in a freezer at 25 C and could be used up to a month later without noticeable degradation in their catalytic performance, UV/vis spectrum or mass spectrometric analysis. The 0.05 mM catalytic solution was made from the stock by diluting 2 mL into 10 mL. To make a 10 mL solution in DMF : TEOA 5 : 1, 1.87 g TEOA was weighed into a 10 mL volumetric flask, approximately 2 mL of DMF was added so the catalytic stock was not being added to neat TEOA. Then 2 mL of stock was added, followed by DMF up to the 10 mL mark. The catalytic solutions were protected from light as much as possible and stored in the freezer. A sample (3 mL) of catalytic solution was added to the photoreaction cuvette and was bubbled with CO 2 /CH 4 95/5 for 10 min, protected from light throughout this time. Irradiation of all samples was performed with an ILC 302 Xe arc lamp. Light from the lamp was directed through a water flter (10 cm) and a 660 nm short pass flter (<660 nm, Knight Optical) to remove heat, such that any sample directly in the beam was at a temperature of 33 C. A l > 520 nm optical flter was added (Schott). The amount of CO produced was determined by GC analysis using a UnicamProGC+ (ThermoONIX) with a thermal conductivity detector. Air, CO, CH 4 and CO 2 were separated on a Restek ShinCarbonST 100/120 micropacked column (2 m, 1/16 00 OD, 1.0 mm ID) ftted with "pigtails" of Restek intermediate-polarity deactivated guard column on either end (fused silica, 0.53 mm ID, 0.69 AE 0.05 mm OD). The carrier gas was ultra high purity He (N6.0, BOC gases) passed through a GC triple flter (Focus Technical) to remove trace impurities prior to the column. The GC method began with 1 min at 40 C followed by a 5 C min 1 gradient up to 120 C (16 min). Injections (200 mL) were made manually with a Hamilton gastight locking syringe (500 mL) at 220 C with a 30 mL min 1 split flow. The carrier gas was kept at constant pressure (165 kPa). The detector block and transfer temperatures were 200 and 190 C respectively, at a constant voltage of 10 V with makeup and reference flows of 29 and 30 mL min 1 respectively. The amount of CO was determined using a calibration plot. Known volumes of CO were mixed with a mimic experimental solution (3 mL DMF : TEOA 5 : 1 (v/v)), headspace and solution were purged with CO 2 : CH 4 (99 : 1 or 95 : 5) and sampled to GC. Quantifcation was by comparison of integrations of the CO peak against the CH 4 internal standard. Corrections were made for temperature and the change in headspace pressure at each injection. ## Synthesis Re(CO) 5 Br, 87 4 0 -Methyl-2,2 0 -bipyridine-4-acetic acid (AABpy). Procedure 1: a modifcation of that by Ciana. 89 A 250 cm 3 round-bottomed flask was flame dried and flushed with Ar. THF (3 cm 3 ) and diisopropylamine (2.1 cm 3 ) were added and the mixture cooled to 78 C. 2.5 M butyl lithium in hexanes (6 cm 3 ) was added via syringe and the mixture was stirred for 0.75 h. A solution of 4,4 0dimethyl bipyridine (3 g) in THF (72 cm 3 ) was added, the solution turned black and was stirred for 2 h at 78 C. Dry CO 2 (g) was set bubbling through a flame dried round-bottomed flask charged with Et 2 O (30 cm 3 ) and cooled to 78 C. The black lithiated bipyridine solution was added to the Et 2 O/CO 2 mixture via cannula and a yellow precipitate soon appeared. The reaction was left under an atmosphere of CO 2 overnight and allowed to warm to RT. Et 2 O was added (30 cm 3 ) and the product extracted with 3 M NaOH (3 30 cm 3 ). The alkaline layer was then acidifed to pH 1 with concentrated HCl and cooling. The product was then extracted with Et 2 O (30 cm 3 ) and buffered to pH 5 with solid CH 3 CO 2 Na. A saturated aqueous solution of Cu(CH 3 CO 2 ) 2 was added causing precipitation of a blue Cu complex. The solid was fltered off with a microfber flter paper and washed with water, ethanol and ether and then air-dried. The product was suspended in water (60 cm 3 ) and H 2 S bubbled through for 20 min resulting in a dark brown colour. The product was fltered through celite, concentrated to 9 cm 3 and fltered again. The solution was evaporated to dryness under reduced pressure to yield a yellow oil. Recrystallisation twice from ethanol/hexane yielded pure AABpy (569 mg, 2.682 mmol, 16%). Analysis was in agreement with the literature. 89 Procedure 2: a modifcation of that by Tomioka. 90 To a flame dried 100 mL round-bottomed flask was added THF (5 mL) and freshly distilled diisopropylamine. The mixture was cooled to 78 C and freshly titrated n-butyl lithium (1.1 eq.) was added. Dimethylbipyridine (1 g, 5.43 mmol) was dissolved in THF (20 mL) and added by cannula. The mixture was stirred at 78 C for 2 h and then freshly distilled methyl chloroformate (0.6 mL) in THF (2 mL) was added by syringe. The reaction was stirred at 78 C for 1 h and then at RT for 2 h. The mixture was then washed with saturated NaHCO 3 solution and extracted into ethyl acetate. The extracts were washed with brine and dried over Na 2 SO 4 . The product was purifed on Si-60 eluting with 2% Et 3 N in pentane and 0-10% EtOAc. The second fraction was collected and the solvent removed (257 mg, 0.858 mmol, 22%). The methyl ester was hydrolysed to produce the free acid. A 50 mL round-bottomed flask was charged with the methyl ester (284 mg), which was dissolved in the minimum amount of methanol. KOH (131 mg) was added. The reaction was stirred at 35 C for 2 h. The solvent was removed and the solid taken up in H 2 O and titrated to pH 7 with a 10% solution of HCl. The H 2 O was removed and the product used without purifcation. (15 cm 3 ) was added, followed by Et 3 N dropwise. The mixture was stirred at 0 C for 5 min and then warmed to RT. After stirring at RT for 0.5 h TLC showed negligible quantities of starting porphyrin and so the reaction was stopped. The reaction was quenched with 10% HCl (50 cm 3 ) and the porphyrin extracted with CH 2 Cl 2 . The extract was washed with saturated NaHCO 3 followed by brine and then dried over MgSO 4 . The product was purifed with column chromatography on Si-60 eluting with CH 2 Cl 2 and CH 3 OH (0% to 3%). The second fraction was collected and the solvent removed to yield the desired product (156 mg, 0.186 mmol, 94%). A 100 cm 3 round-bottomed flask was charged with CH 2 Bpy-H 2 TPP (152 mg, 181 mmol), Zn(OAc) 2 (179 mg, 815 mmol), CH 3 OH (5 cm 3 ) and CHCl 3 (25 cm 3 ). The mixture was heated to reflux for 1 h and the reaction was followed by UV/vis spectroscopy. The reaction mixture was allowed to cool, pumped to dryness and redissolved in 100 cm 3 CHCl 2 and 20 cm 3 CHCl 3. This was washed with EDTA solution (2 g in 200 cm 3 10% Na 2 CO 3 solution), water (3 200 cm 3 ), dried (MgSO 4 ) and the solvent removed to yield the desired compound (160 mg, 178 mmol, 98%). 1 5-{4-[Rhenium(I)tricarbonyl(bromide)-4-methyl-2,2 0 -bipyridine-4 0 -methylene carboxyamidyl]phenyl}-10,15,20-triphenylporphyrinatozinc(II) (Dyad 3 Br). A two-neck 50 cm 3 round-bottomed flask was ftted with a reflux condenser and gas valve. The setup was flame dried. Under Ar CH 2 Bpy-ZnTPP (200 mg, 221 mmol) was added, followed by ReBr(CO) 5 (90 mg, 221 mmol). Dry benzene was added (30 cm 3 ) by syringe. The mixture was heated to 65 C and the reaction was followed by IR spectroscopy and judged to be complete after 22 h. The reaction mixture was fltered to leave a solid product and used without further purifcation (263 mg, 210 mmol, 95%). 5-{4-[Rhenium(I)tricarbonyl(3-picoline)-4-methyl-2,2 0 -bipyridine-4 0 -methylene carboxyamidyl]phenyl}-10,15,20-triphenylporphyrinatozinc(II) trifluoromethanesulfonate ([Dyad 3 pic] OTf). A two-neck 50 cm 3 round-bottomed flask was ftted with a gas valve and flame dried. It was taken into a glovebox and AgOTf was added (206 mg, 800 mmol). A condenser ftted with a single-neck round-bottomed flask and gas valve was flame dried, then the round-bottomed flask was removed under Ar and the condenser and reaction flask were brought together. THF and 3-picoline (1.09 mL, 11.2 mmol) were added and fnally Dyad 3 Br (200 mg, 160 mmol). The mixture was heated to reflux for 2 h and checked for completion by IR spectroscopy. The mixture was allowed to cool, fltered to remove AgBr and dried under vacuum for 72 h. The oil was re-dissolved in THF and applied to Sephadex LH20 eluting with THF. The THF was removed and the solid washed with an ethanol/petrol 20/ 80 mixture. The solid was dried to yield [Dyad 3 pic]OTf (95 mg, 67.10 mmol, 42%). 1 ## Synthetic methodology The synthetic methods for the preparation of [Dyad 1 pic]OTf, Dyad 1 Br, [Dyad 2 pic]OTf and Dyad 2 Br have been reported previously. 11,34 The synthetic strategy for [Dyad 3 pic]OTf is shown in Fig. 2. AABpy 89 and NH 2 -H 2 TPP 88 were prepared by literature procedures. The two were coupled using 2-chloromethylpyridinium iodide in excellent yield (94%). 91 Zinc was inserted into the porphyrin and Re(CO) 3 Br was complexed to the Bpy as reported previously for [Dyad 1 pic]OTf. 34 Bromide was substituted for 3-picoline using AgOTf in THF. The product was purifed using size exclusion chromatography (Sephadex LH20) eluting with THF. (Fig. S13 †). Two reversible oxidation waves were observed on scanning to anodic potentials, corresponding to the frst and second oxidation of the porphyrin. 92 In the cathodic direction a quasi-reversible reduction wave was observed, corresponding to the frst reduction of the rhenium unit. The frst oxidation of the porphyrin is at similar potential to that of [Dyad 2 pic]OTf and the Re-free porphyrin CH 2 -Bpy-ZnTPP, but the reduction of the rhenium is at a more negative potential than those of either [Dyad 1 pic]PF 6 or [Dyad 2 pic]OTf (Table 1). 11 [(2-methoxy-4-nitro-phenylcarbonyl)-amino]phenyl]-10,15,20triphenyl porphyrin (Fig. S16 †). The structure confrms that the hydrogen atom on the porphyrin amide forms a hydrogen bond with the oxygen of the methoxy group, consistent with the low feld at which the amide proton resonates in the 1 H NMR spectrum. 11 ## Emission spectroscopy Comparison of the singlet pp* fluorescence from the porphyrin moiety of a dyad, with that of a suitable rhenium-free analogue provides valuable information on the quenching ability of the rhenium unit. We have previously reported emission quenching determined in this way for the zinc porphyrin unit in [Dyad 1 pic]OTf (>95% in PrCN) and for Dyad 1 Br (50% in THF) compared with the emission of the rhenium-free ZnTPP-link-Bpy analogue. 34,81 Similar emission measurements were performed with [Dyad 2 pic]OTf and [Dyad 3 pic]OTf (Table 2) showing 55% and 23% emission quenching, respectively (Fig. S17 and S18 †). The emission lifetimes (Table 2) show a similar trend to the steady state measurements. [Dyad 1 pic]PF 6 has the shortest lifetime, followed by [Dyad 2 pic]OTf, which also shows a longer component. [Dyad 3 pic]OTf shows the least shortening of the fluorescence lifetime (Fig. S19 †). The emission quenching is attributed principally to electron transfer from excited state zinc porphyrin to the rhenium with only minor heavy atom effects. The large variation in quenching is suggestive of corresponding variations in electron transfer rates. Emission quenching in Dyad 1 Br, Dyad 2 Br and Dyad 3 Br was measured relative to zinc tetraphenylporphyrin in THF (Fig. S20 †). In agreement with previous reports, 81 Dyad 1 Br displays 41% emission quenching relative to a simple zinc porphyrin while Dyad 2 Br and Dyad 3 Br show 11% and 0% emission quenching, respectively. A very similar trend is observed in the emission lifetimes (Table 2, Fig. S21 †). We also checked the emission yield of CH 2 -Bpy-ZnTPP (Fig. 2) relative to unsubstituted ZnTPP and found that the emission of CH 2 -Bpy-ZnTPP is 15% more intense than that of ZnTPP for samples of equal absorbance at the exciting wavelength. The minor quenching of the bromide complexes demonstrates that heavy atom effects are unimportant and that electron transfer plays a less signifcant role than in the corresponding picoline complexes. We note also that ZnTPP and zinc tetraphenyl chlorin fluorescence is not quenched by TEOA. 11 ## Photocatalysis All six dyads were tested for CO 2 photoreduction to CO under irradiation with l > 520 nm in a solution of DMF : TEOA 5 : 1 at 0.05 mM (Fig. 4 and Table 3). Overall turnover frequencies (overall TOF) are calculated over the full period of irradiation, whereas maximum turnover frequencies ## Photocatalytic intermediates UV/vis spectra were taken at regular intervals during photocatalysis. We previously reported that signifcant changes occur in the Q-bands of the porphyrins during CO 2 photoreduction catalysis by the porphyrin-rhenium dyads. 11 The Q-bands of porphyrins and their derivatives provide excellent spectroscopic handles in the visible region, providing a clear indication of structural changes. These changes were assigned to formation of chlorin, a reduction product of the porphyrin in which one CC bond of a pyrrole group is saturated. The UV/vis spectra of [Dyad 3 pic]OTf during catalysis are shown in Fig. 5. At early photolysis times, the Q-bands of the porphyrin decrease in intensity and a product band grows at 625 nm with a shoulder at 610 nm, seen most clearly in the difference spectra (Fig. 5(b)). The relative intensities of the 610 and 625 nm bands change with time. The band at 625 nm may be assigned to zinc chlorin product, while the 610 nm band is assigned to the zinc isobacteriochlorin, the derivative in which two adjacent pyrrole groups are saturated. 94 This second hydrogenation product is formed in greater amounts for [Dyad 3 pic]OTf and persists longer. For all dyads the photocatalytic conditions eventually lead to complete bleaching of the Q-band region of the spectrum. The exact chemical structure of the chlorin cannot be determined from UV/vis spectroscopy alone. It has been shown previously that triethylamine can add to the pyrrole to form both the simple hydrogenation product and a product in which a C-H bond has been formally added across the C]C bond. 95 A large-scale (50 mg) photolysis (l > 520 nm) was performed on ZnTPP in DMF : TEOA 5 : 1 under Ar and the product was exhaustively extracted into ether after addition of water. The ether was separated and the product dried under vacuum. The 1 H NMR spectrum of the product dissolved in CDCl 3 matches the spectrum of an authentic sample of zinc tetraphenylchlorin (Fig. S23 †), demonstrating that the major product is formed by simple hydrogenation (Fig. 6). ESI-mass spectrometry measurements were made on samples from CO 2 photoreduction by [Re(Bpy)(CO) 3 (pic)][PF 6 ] and zinc tetraphenyl porphyrin (ZnTPP). Zinc possesses several isotopes of signifcant abundance producing a pattern that spans several m/z units. As a result, the signals for the various hydrogenation products of zinc porphyrin overlap closely. The signals obtained centre around m/z ¼ 680 and match well with the calculated isotope pattern for a mixture of ZnTPP and the dihydrogenated (chlorin) and tetra-hydrogenated (isobacteriochlorin) products (Fig. S24 † and 6). (Fig. S27 †). 96 A similar experiment with Dyad 2 Br produced no change thermally, but a product with bands at the same wavenumbers appeared on photolysis with l > 520 nm under N 2 (Fig. S28 †). In an attempt to increase conversion and the signal of the substitution product, CO 2 was bubbled through Dyad 2 Br in DMF : TEOA 5 : 1 under l > 520 nm irradiation. This time, different signals were observed at 2015, 1911 and 1885 cm 1 that correspond closely to those reported 53 for the carbonato complex (Fig. 7 and S28 †). Considering the excellent ft to Ishitani's data for Re(OCH 2 -CH 2 NR 2 )(Bpy)(CO) 3 , the product from [Dyad 2 pic]OTf may be assigned as Dyad 2 OCH 2 CH 2 NR 2 . We are not able to show defnitively whether the product from Dyad 2 Br is the same or the analogue where the porphyrin has been reduced to chlorin, since the timescale for hydrogenation is similar the timescale for reaction with TEOA. Nevertheless, the IR evidence supports CO 2 insertion into the metal-oxygen bond to form species containing the Re{OC(O)OCH 2 CH 2 NR 2 }(Bpy)(CO) 3 unit. 53 Picosecond time-resolved infrared spectroscopypicoline complexes Extensive previous TRIR investigations of the excited states of Re(Bpy)(CO) 3 derivatives 34,68,80 show that formation of 3 MLCT excited states result in high frequency shifts of the carbonyl vibrations, whereas charge transfer to the Re(CO) 3 results in substantial low frequency shifts. 34 The photophysics and photochemistry of [Dyad 1 pic]OTf have previously been investigated using TRIR spectroscopy in PrCN. 34 Excitation of [Dyad 1 pic]OTf at 600 nm resulted in the initial formation of an excited state localised on the porphyrin moiety of the dyad, followed by subsequent electron transfer to the Re(diimine) ligand generating a charge-separated (CS) state. The CS state reached a maximum within 10 ps and decayed over 40 ps. Charge recombination back to the porphyrin moiety via a hot ground (HG) state regenerated the parent complex within 200 ps. In addition, a sharp peak in the TRIR spectra at 2026 cm 1 could be observed during the frst 5 ps, which was tentatively assigned to the formation of an intraligand (IL) pp* excited state. 97 The complexity of the transient spectroscopy was reconciled with a model which postulates that the dyad molecules adopt a range of conformations each with their own kinetics. We have performed 4. ## Picosecond time-resolved infrared spectroscopybromide complexes The photophysics and photochemistry of Dyad 1 Br, Dyad 2 Br and Dyad 3 Br were monitored using TRIR spectroscopy following excitation at 560 nm in THF. All TRIR spectra were obtained in THF since the dyads are not sufficiently soluble in CH 2 Cl 2 . In general, the TRIR spectra of the bromide dyads are more complex than those of the picoline dyads, with multiple transient species observable in the spectra. The TRIR spectra obtained following excitation of Dyad 1 Br are shown in Fig. 10. Three negative bands are observed corresponding to the parent complex at 2022, 1922 and 1900 cm 1 . At early time delays (<50 ps) a band at 2055 cm 1 and a broad band at ca. 1960 cm 1 can be observed, characteristic of the high frequency shift associated with the formation of a 3 MLCT state on the Re moiety of the dyad. 68,77,98,99 This 3 MLCT excited state is formed initially <10 ps after excitation from vibrationally hot excited states and decays over the subsequent 250 ps (Fig. 10(c), blue squares). The formation of peaks at 1998 cm 1 and 2015 cm 1 can also be observed on a similar timescale to the decay of the 3 MLCT state. The peak at 1998 cm 1 is analogous to observations made on the picoline dyads (see above) and is assigned to the formation of a CS state. The corresponding lower energy bands associated with the CS species can be observed at ca. 1880 cm 1 , but due to their weak intensity, the exact band positions could not be determined. The band at 2015 cm 1 suggests the simultaneous formation of an IL pp* excited state, 77,97 similar to that observed following the photolysis of [Dyad 1 pic]OTf in PrCN. 34 The associated low energy bands of the IL pp* excited state cannot be observed as they are low intensity and fall in a similar region of the spectrum to the ground state bleach. Bleaching of the ground state does not reach a maximum negative signal until 15 ps, and it recovers over the subsequent 1000 ps (Fig. 10(c), red dots). The recovery of the signal at 2022 cm 1 occurs over two distinct timescales. The frst (0-100 ps) is mainly associated with deactivation of the 3 MLCT and the second (100-1000 ps) is principally due to the decay of the CS and pp* excited state. The kinetics of the CS state and the IL pp* excited state were not fully determined as the bands are weak and overlap with other bands in this region of the spectrum. The TRIR spectra obtained following excitation of Dyad 2 Br are shown in Fig. 11. Parent bleaches at 2020, 1922 and 1900 cm 1 can be observed as well as the formation of two transient species (Fig. 11(b)). At all time delays, bands 2057 cm 1 and ca. 1975 cm 1 (broad) are visible, associated with the formation of a 3 MLCT excited state on the Re moiety. This 3 MLCT state is initially formed from vibrationally hot excited states at time delays <10 ps. In addition, bands at 1997, 1887 and 1871 cm 1 can be observed <500 ps after excitation, which are assigned to the formation of a CS state. The CS species grows in on a timescale faster than 2 ps and decays over the subsequent 1000 ps (Fig. 11(c), black squares) as the parent bleach partially recovers (65%, Fig. 11(c), red dots). An IL pp* excited state was not observed at any time delay in this experiment. At 500 ps after photolysis, the only bands visible in the TRIR spectrum are those originating from the 3 MLCT and these bands along with the parent bleaches do not change intensity signifcantly on the timescale of this experiment (up to 1000 ps). The TRIR spectra recorded after flash photolysis of Dyad 3 Br are shown in Fig. 12. Bands associated with the formation of a 3 MLCT excited state at 2055 cm 1 and ca. 1975 cm 1 (broad) grow in over the frst 100 ps and do not deplete signifcantly up to 1000 ps after excitation. In addition, an IL pp* excited state band at 2014 cm 1 can be observed that grows in over the frst 30 ps and completely decays by 100 ps. The low energy bands of the IL pp* excited state cannot be observed as they are weak in intensity and overlap with the ground state bleaches. The 3 MLCT state is probably formed via energy transfer from the porphyrin pp* excited state. 97 This is energetically feasible as the higher energy emission maximum of Dyad 1 Br is at 606 nm, compared to the emission maximum for the 3 MLCT state of ReBr(Bpy)(CO) 3 at 620 nm. 81 In contrast to Dyad 1 Br and Dyad 2 Br, a CS state was not observed following the photolysis of Dyad 3 Br. The ground state bleach reaches a maximum at 30 ps and has recovered by 65% at 1000 ps after excitation. Through a separate ns-TRIR experiment we determined that the 3 MLCT state decays with a lifetime of ca. 2 ns as the parent complex reforms. However, this experiment had to utilise a 532 nm excitation pulse which is not ideal as it falls at the edge of the porphyrin Q band absorption and led to relatively weak TRIR signals. We examined the possible quenching of the 3 MLCT Br. quenching of the 3 MLCT state is expected to be a small component of the decay because of the short excited state lifetime. However, no reductive quenching was observed. Given the low signal-to-noise of these measurements due to the unfavourable excitation wavelength, we can only state that if quenching occurs then it represents less than 1% of the 3 MLCT decay. ## Energetics of electron transfer The change in free energy for the dyad picoline cations on intramolecular electron transfer from the excited state of the sensitizer to rhenium can be estimated using eqn (1) where E ox and E red are taken as the potentials for the frst oxidation of the sensitizer and frst reduction of the rhenium, respectively. The potentials were estimated from cyclic voltammograms measured in CH 2 Cl 2 . For E 00 , we used the highest energy emission maximum of the sensitizer, measured at room temperature. The potentials, emission maxima, driving forces and maximum TON CO are given in Table 5. We do not report quantum yields for CO production because there is signifcant photoreaction at the porphyrin during catalysis. For the bromide dyads, we expect an additional electrostatic contribution to the free energy of electron transfer, since the electron transfer generates a pair of charges. The edge-to-edge distance from porphyrin to Bpy may be regarded as the minimum distance for electron transfer and is measured at 8.0 in the crystal structure of [Dyad 1 pic]PF 6 . The electrostatic contribution in CH 2 Cl 2 is calculated as 0.20 eV and may be taken as an upper limiting value for Dyad 1 Br. The corresponding value for Dyad 2 Br would be signifcantly less negative, while that for Dyad 3 Br may be more negative at ca. 0.27 eV because of its ability to fold about the CH 2 group. However, in DMF, the solvent used for CO 2 reduction, these values become of little importance because of the high dielectric constant of the solvent: 0.04 eV for Dyad 1 Br and 0.05 eV for Dyad 3 Br. 5 for [Dyad 3 pic]OTf and Dyad 1 Br are close to zero, while that for Dyad 3 Br is positive. The bromide dyads have completely different potentials from the picoline dyads yet their photocatalytic behaviour is very similar and sometimes superimposable. Furthermore, Dyad 3 Br is very active, yet the driving force in DMF is not favourable for electron transfer. Considering just the picoline complexes, the greater is the driving force for electron transfer, the lower is the observed maximum turnover number. We can deduce from these points that the porphyrin dyad bromides are not the active species in photocatalysis. However, we have previously shown that the driving force for electron transfer from the excited state of zinc tetraphenylchlorin to Re complexes is 150 meV more negative than that for the excited state of zinc tetraphenylporphyrin. 11 Thus, it is possible that bromide dyads become more active on reduction to the chlorin derivative (see Mechanism section of Discussion, below). The corresponding values of the reduction potentials of Re(OCH 2 CH 2 NR 2 )(Bpy)(CO) 3 (R ¼ CH 2 CH 2 OH) and related dyads are not known, but we would expect them to be close to those of the bromide complexes. The reduction potential of Re(OCOOCH 2 CH 2 NR 2 )(Bpy)(CO) 3 (R ¼ CH 2 CH 2 OH) is reported to be very similar to that of the simple bromide complex. 53 ## Emission spectroscopy Emission data allow us to derive yields for fluorescence from the singlet pp* state and estimates of the intramolecular quenching rate (Table 2). The quantum yields are based on the quantum yield of fluorescence for ZnTPP in toluene of 3%. 93 Across both the picoline and bromide dyads, steady state emission quantum yields and emission lifetimes increase signifcantly as photocatalytic activity increases. This lack of correlation may be resolved if some of the dyads are emissive but inactive with respect to charge separation, while others are non-emissive but undergo charge separation. Thus the fluorescence data show that there are signifcant excited state populations that do not directly lead to photocatalysis. This diversity of behaviour is attributed to conformers which are not predisposed to the required electron transfer process. If there was no issue of multiple conformers, all the molecules would end up in the state with the shortest rise-time following formation of the porphyrin p-p* S 1 state. ## Time resolved infrared spectroscopy There are some striking differences in the TRIR spectra and kinetics obtained between the different dyads. are not ideal for electron transfer and represent one conformer out of many that may be present in solution (Fig. 3). Comparison shows that [Dyad 3 pic]OTf exhibits the longest risetime for charge separation and the longest lived charge-separated state (Table 4). Only [Dyad 1 pic]OTf undergoes charge recombination via a hot ground state. 34,100 Although the lifetimes of the CS states correlate with photoactivity, they are extremely short if bimolecular reaction is to occur with any species other than either triethanolamine or DMF which are components of the solvent, even for [Dyad 3 pic]OTf. The TRIR spectra of the bromide complexes are very different from those of the picoline complexes. In the bromide complexes, we observe a 3 MLCT state in all three dyads and the CS state can only be clearly observed in Dyad 1 Br and Dyad 2 Br. These observations are consistent with the driving force calculations above. The 3 MLCT states of Dyad 2 Br and Dyad 3 Br have lifetimes on the ns timescale. The risetimes of the 3 MLCT states are in the range of tens of picoseconds which is again incompatible with the rate of quenching of the pp* states. We suggest that these differences reflect the presence of multiple conformers. The absence of the CS state of Dyad 3 Br appears remarkable considering its strong photocatalytic activity. ## Mechanism On photo-excitation, [Dyad 2 pic]OTf and [Dyad 3 pic]OTf form charge-separated states, as seen for [Dyad 1 pic]OTf. 34 The close match in the photocatalysis curves for [Dyad 2 pic]OTf and Dyad 2 Br (Fig. S22 †) suggests that these compounds share a catalytically active species. The other two pairs also exhibit strong similarity in the curves. The parallel photocatalytic behaviour contrasts with the very different excited states observed by TRIR spectroscopy (see above). We have presented evidence from steady-state spectroscopy that the photocatalysts undergo reaction with triethanolamine both at the porphyrin centre and the rhenium centre. Photoreaction at the porphyrin causes initial 2-electron hydrogenation to the chlorin and subsequently a further 2-electron hydrogenation to the isobacteriochlorin. Thermal reaction of [Dyad 2 pic]OTf and photoreaction of Dyad 1 Br yield evidence for formation of Re(OCH 2 CH 2 NR 2 )-(Bpy)(CO) 3 derivatives. In addition, triethanolamine and DMF are capable of coordinating to zinc, probably forming equilibrium mixtures. An electron can be supplied to the oxidised porphyrin by the ffth ligand on zinc. Taken together with the arguments presented above on energetics and TRIR spectra, these considerations indicate that the dyads act as pre-catalysts. Within the frst 30 min of irradiation, signifcant reduction to chlorin and reaction with triethanolamine occurs, probably generating the true photocatalysts. ## Conclusions We have shown that a new design of zinc porphyrin-Re bipyridine tricarbonyl dyad with a methylene spacer is active for the photocatalytic reduction of CO 2 to CO with l > 520 nm irradiation. The TON is higher than those for previously reported dyads by a factor of ten and there is a major increase in TOF. These fgures exceed those for the two component system ZnTPP + [ReBpy(CO) 3 (pic)][PF 6 ] by a factor of three. 11 The beneft of using a saturated bridge between Bpy and the porphyrin is in accord with Ishitani's binuclear Ru-Re complexes which were also most effective with a saturated bridge. 10 The most likely reason for the enhanced activity is the flexibility for the Re(Bpy) unit to adopt the best orientation and closest approach to the porphyrin unit. Furthermore, in these dyads the groups at the 4 and 4 0 positions of the Bpy are CH species, followed by formation of an isobacteriochlorin and eventually complete bleaching. Previous results indicate the chlorin intermediates are active catalysts. 11 It is likely that the isobacteriochlorin intermediates are active also. Complete bleaching renders the dyads unable to absorb visible light and is one route for deactivation. The dyads react with triethanolamine in DMF to form alkoxide complexes containing a Re(OCH 2 CH 2 NR 2 )Bpy(CO) 3 moiety which undergoes CO 2 insertion. The picoline complexes undergo this transformation thermally while the bromide complexes require irradiation. [Dyad 3 pic]OTf undergoes charge separation in 8 ps and the charge-separated state has a lifetime of 320 ps. For comparison, [Dyad 2 pic]OTf undergoes faster charge separation but the majority of the CS photoproduct decays much faster (with time constant of 42 ps). The bridge in Dyad 3 has slowed down charge separation and charge recombination. The chargeseparated state is one order of magnitude longer-lived in Dyad 3 than in Dyad 2. The bromide complexes show very different photochemical behaviour on the ps timescale with combinations of either 3 MLCT and CS, or 3 MLCT and IL excited state products. Dyad 1 Br and Dyad 2 Br form some CS product, whereas Dyad 3 Br does not. The CS state is unlikely to be responsible for the activity of the bromide complexes since Dyad 3 Br is very active. All three bromide dyads display formation of a 3 MLCT state, the lifetimes of which decrease in the order Dyad 3 Br > Dyad 2 Br > Dyad 1 Br, in line with their photocatalytic activities. The 3 MLCT state may be responsible for the activity of the bromide dyads. As expected for such short lifetimes, ns-TRIR experiments on Dyad 3 Br showed little or no bimolecular reaction with TEOA but this cannot be ruled out. Thus the bromide complexes display very similar photocatalytic behaviour to the picoline complexes but totally different excited states. Taken together, the data strongly suggest that the active photocatalyst is formed by a combination of reaction of triethanolamine at rhenium and photoreduction of the porphyrin. We have previously shown that zinc chlorin is more reducing than zinc porphyrin 11 and may allow for the formation of signifcant amounts of charge-separated state in the bromide complexes as well as the picoline complexes. This hydrogenation can also explain why the picoline dyads are not de-activated on thermal substitution of picoline for the anionic alkoxide/ carbonato complexes, which would be expected to have similar reduction potentials to the bromides. TRIR experiments demonstrate that bimolecular reaction of the 3 MLCT state of Dyad 3 Br is minimal and thus support the chlorin theory.

chemsum

{"title": "Comparison of rhenium\u2013porphyrin dyads for CO₂ photoreduction: photocatalytic studies and charge separation dynamics studied by time-resolved IR spectroscopy", "journal": "Royal Society of Chemistry (RSC)"}

drug_repurposing_of_approved_drugs_elbasvir,_ledipasvir,_paritaprevir,_velpatasvir,_antrafenine_and_

## Abstract: Aims: Pneumonia of unknown cause detected in Wuhan, China was first reported to the WHO Country Office in China on 31 December 2019. The outbreak was declared a Public Health Emergency of International Concern on 30 January 2020. Currently, there is no Vaccine against COVID-19 pandemic and infection is spreading worldwide very rapidly. The present study aimed to meet the exigent requirement of practicable COVID19 drug treatment with a computational multitarget drug repurposing approach.Main methods: Many reports are available with in silico drug repurposing. However, the majority of them engrossed on a single target. In the present study, 1735 FDA approved drugs screened with molecular docking approach against Covid19 protein and extracts the drug combination targeting COVID19 proteins comprehensively. ## Introduction The World Health Organization announced in February 2020 that COVID-19 is the official name of the disease. World Health Organization chief Tedros Adhanom Ghebreyesus explained that CO stands for corona, VI for virus and D for disease, while 19 is for the year that the outbreak was first identified; 31 December 2019 . Coronavirus disease 2019 (COVID-19) is a contagious disease caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) previously referred to as the 2019 novel coronavirus (2019-nCoV) . In 2019 at Wuhan, the capital of Hubei, China, the disease was first reported and then it spread worldwide, resulting in the 2019-20 coronavirus pandemic . At present, there are no clinically proven vaccines and medicines for COVID-19 prevention and treatment as per U.S. Food and Drug Administration (FDA), the Centers for Disease Control and Prevention (CDC) and World Health Organization (WHO) . On an interactive web-based dashboard to track COVID-19 in real-time as on April 19, 2020, in the entire world more than 200 countries/territories are having 2.3 Million confirmed cases and 1,61,283 deaths worldwide and more than 50000 increase daily reported since March 26, 2020 . Recovery observed in patients of Covid19 treated with the mixture of Anti-HIV drugs like Libonavir & Ritonavir, Anti-SARS drugs Oseltamivir and Anti-malarial drug Chloroquine in India (Wadhawan, 2020). In South Korea, human MERS-CoV successfully revokes the viral clearance using a combination of Lopinavir/Ritonavir (LPV/RTV) (Anti-HIV drugs) pegylated interferon and ribavirin . Still, the treatment with anti-HIV drugs is a mystery for the Patients and Researchers as well . As per data available on https://www.excelra.com/covid-19-drug-repurposing-database/data currently, 125 different drug molecules are reported under trial. Recently, diverse computational approaches explored for screening the drug molecules for COVID19 treatment. Molecular docking of Lopinavir, Darunavir and Ritonavir reported against the modelled structure of COVID19 proteases, coronavirus endopeptidase C30 (CEP_C30) and papain-like viral protease (PLVP) . MM-PBSA-WSAS (Molecular dynamics simulations followed by binding free energy calculations using an endpoint method) employed for Fast Identification of Possible Drug against COVID-19 protease . Anti-HCV drugs, Sofosbuvir, IDX-184, Ribavirin, and Remidisvir reported promising drug candidates with a docking approach against modelled COVID-19 RNA dependent RNA polymerase (RdRp) . Hirokawa et al., (2020) identified one hundred and several dozen potentially candidate drugs for 3CL protease inhibitors, which are already approved as antiviral, HIV protease inhibitors, antibacterial or antineoplastic agents with in silico docking-based screening approach, which combines molecular docking with a protein-ligand interaction fingerprint (PLIF) scoring method. Many other reports are also available with in silico drug repurposing but the majority of them engrossed on a single target. The present study designed for docking FDA approved Drugbank compounds with molecular weight less than 700 against all COVID-19 experimentally and computationally generated protein structure. ## Target preparation Crystal structure of COVID-19 main protease in complex with an inhibitor N3 (PDB ID: 6LU7) , SARS-Coronavirus NSP12 bound to NSP7 and NSP8 co-factors(6NUR) and Pre-fusion 2019-nCoV spike glycoprotein with a single receptor-binding domain (6VSB) retrieved from the Protein Data Bank available at https://www.rcsb.org/. Additionally, computationally predicted 24 modelled structures based on protein sequences translated from the complete genome of Severe acute respiratory syndrome coronavirus 2 isolate Wuhan-Hu1 (Genbank Accession: MN908947.3) available on I-TASSAR online server also included in the present study (Table 1) . Total 27 receptors threedimensional structures were subjected addition of Hydrogen for pH 7.0 and Gasteiger charges were using Open Babel . The resulting structures converted to PDBQT format using python script -prepare_receptor.py‖ from AutoDock Tools . ## Data Analysis The Binding energy for each drug ligand exported for analysis from the Autodock Vina v4.2 software and need-based analyses performed manually using python script and excel software. Drug target interaction analysed with Ligplot + . ## Results Drug repurposing study for FDA approved 1735 drug molecules < 700 MW included in the present study. The molecular docking accomplished against a total of 24 modeled proteins and 3 PDB structures followed by seven-stage screening to determine the best combination for COVID19 treatment. In the first stage, a total of 4,68,650 docking solutions obtained from Molecular docking of approved 1735 drug molecules against 27 COVID19 proteins. From Autodock Vina output, one best pose was extracted from out of 10 poses in the second stage and 46,865 pose data utilized for further analysis (Table 2). In stage three screening step, the top 20 drug molecules based on binding energy were separated for each COVID19 Protein. Which lead to 540 drugs with a potential range of docking scores (Table 1 & 2, Figure 1). The redundant drugs removed with the merger of the top 20 dock score for all COVID19 protein together yields 133 unique drugs from the 540 drugs (Supplementary material). All the 133 drugs were manually inspected for pharmacological activity reported in Drugbank. Consequently, 18 anticancer drugs, 6 anti-inflammatory, 7 anti-HIV, 8 anti-HCV, 6 drugs for lung disease, 3 drugs for anti-parasitic activity, 2 anti-migraine and few other activities reported. The anti-cancer drugs and drugs with clinically major side effects were omitted. The remaining 35 drugs were selected for further comparative analysis. This included anti-HIV, anti-HCV, anti-Inflammatory, Lung disease, anti-migraine activity, and anti-parasitic activity. Along with nine clinically reported promising anti-COVD19 drugs also included for further analysis. The binding energy of selected 35 compounds graded in five grades displaying dark green, light green, yellow, orange and red for easy comparison. Concerning anti-COVID19 activity against single and multiple target analysis, the best results obtained for 8 drugs from anti-HCV followed by 7 anti-HIV and 3 currently clinically used drugs. In addition to that, 1 drug from an anti-inflammatory and anti-migraine group also had a good docking score. The overall analysis presented the comparatively better scoring of the selected drug in this study over currently clinically applied drugs for COVID19 treatment (Figure 1) To narrow down the number of drug combinations and removing drug acting on the same target binding energy replaced with docking rank. Based on dock rank 10 drugs selected for further analysis (Figure 2). To reduce the drug for treatment and avoid duplication, the drug acting on the same target was removed. Consequently, the combination of Elbasvir, Ledipasvir, Paritaprevir, Velpatasvir, Antrafenine Ergotamine drug figured out as a potential cocktail for COVID19 treatment (Figure 3). ## Comparative binding interaction of drugs against HCV and COVID19 To analyse the comparative binding interaction with HCV targets for anti-COVID19 receptors, three targets of HCV NS3 Helicase, NS5B RNA-dependent RNA polymerase and NS3/4A S168A protease docked against 26 potential drugs along with known inhibitors. The result indicated, Chloroquine antiplasmodial drug computationally significant for three HCV receptors but the same is less significant for COVID19 (Figure 4). Similarly, anti-HIV drugs Dolutegravir and Maraviroc had considerable binding against HCV targets but less potential with COVID19 receptors. Conversely, Elbasvir, Ledipasvir reacted with related potential against Helicase, RdRp and Protease from both virus HCV and COVID19. This projected the potential of anti-HCV drugs for inhibition of the COVID19 virus. While ergotamine used in the present study proposes significant binding against all three COVID19 proteins but no promising binding evident with HCV. This suggests the presence of a specific binding site present in COVID19 which might be absent in HCV. ## Discussion The earlier report of drug repurposing mainly focused on a single target option we represented the holistic approach of targeting all the viral proteins for effective anti-COVID19 activity. The combination suggested will effective on the nonstructural proteins, structural proteins of COVID19 and also included drug the anti-inflammatory with synergic anti-COVID19 activity. Comparative computational analysis of the proposed drug is superior to currently clinically used drug combinations. Which, reflects the possibility of the promising effect of the proposed drug combination in further clinical applications. Ledipasvir is an inhibitor of the Hepatitis C Virus (HCV) Non-Structural Protein 5A (NS5A) . This protein is crucial for viral RNA replication and assembly of HCV virions. Ledipasvir and Velpatasvir were reported promising in one of the recent studies in virtual screening . Clinical trials against anti-HCV suggest the Ledipasvir treatment significantly improves the patients within one to twelve weeks . Similarly, Velpatasvir acts as a defective substrate for NS5A (Non-Structural Protein 5A) sharing a similar function as Ledipasvir. The preclinical study shown the patient having an infection of HCV genotype 1 to 6 can be treated with Velpatasvir . Elbasvir was primarily used for the treatment of HCV. However, some trials are also carried out for using Elbasvir in the treatment of COVID19 . Paritaprevir reported as an inhibitor of COVID19 Protease . No reports are available for anti-COVD19 effectively of Antrafenine and Ergotamine. ## Conclusion There is an urgent need for anti-COVID-19 drugs to address the global medical emergency. Several drugs are currently employed with empirical clinical knowledge along with some contradictions. Many drug repurposing reports are available but those are mainly concentrated on a single target. The present study directed with a holistic approach of targeting multiple COVID19 proteins with a mixture of FDA approved drugs. The proposed blend drugs include Elbasvir, Ledipasvir, Paritaprevir, and Velpatasvir which currently used for HCV treatment. Inclusion anti-inflammatory Antrafenine and anti-migraine Ergotamine drugs can be effective for dual action of inflammation reduction and COVID19 inhibition. The anticipated combination of drugs acting on both non-structural and structural proteins therefore, can be able to reduce the COVID19 infection process and also reduce viral multiplication. The present study can be immediacy explore further by medical, pharm and research experts effectively to find out the best strategy for anti-COVID19 treatment. The Green colored cell contains the lowest binding energy for each row followed by the greenish-yellow, yellow, orange and red color. The lowest binding energy represents the firm binding of ligand and protein molecules. The Comparative analysis establishes the drugs used for the treatment of Hepatitis C represented a more number of greenish cells. This indicates these drugs stay on top throughout the library used in the present study. The Elbasvir, Ledipasvir, Paritaprevir, Velpatasvir from Anti-HCV followed by Anti-HIV drugs Saquinavir, Ritonavir, Dolutegravir, Maraviroc, Tipranavir, Etravirine and Rilpivirine having good binding energy one or more targets. Only one drug, Antrafenine used for anti-inflammatory and Ergotamine used for anti-migraine activity shown significant activity against the different targets. However, the drugs used worldwide for clinical purposes except for Danoprevir and Lopinavir have shown poor binding with receptor molecules used in the study. The drugs from the list removed having overlapping target applicability. The removal identified 6 unique drugs acting on the multiple targets. The combination of these drugs overcomes the lower affinity of the drug molecules to the specific receptors targeting all the proteins available in the present study. Ergotamine shown good binding against 18 receptors, followed by Lepidasvir, Elbasivir and Pariptapevir respectively against 16, 12 and 9 receptors. While velpatasvir and anterafenine showed binding against 5 which might be very useful in a multi-drug combinatorial approach. The results graded on five color green for lowest binding energy followed by greenish-yellow, yellow, orange and red for the highest binding energy for each receptor. Figure 4: Comparison of HCV targets as a reference against COVID19.

chemsum

{"title": "Drug repurposing of approved drugs Elbasvir, Ledipasvir, Paritaprevir, Velpatasvir, Antrafenine and Ergotamine for combating COVID19", "journal": "ChemRxiv"}

merging_of_the_photocatalysis_and_copper_catalysis_in_metal–organic_frameworks_for_oxidative_c–c_bon

## Abstract: The direct formation of new C-C bonds through photocatalytic oxidative coupling from low reactive sp 3 C-H bonds using environmentally benign and cheap oxygen as oxidant is an important area in sustainable chemistry. By incorporating the photoredox catalyst [SiW 11 O 39 Ru(H 2 O)] 5À into the pores of Cu-based metal-organic frameworks, a new approach for merging Cu-catalysis/Ru-photocatalysis within one single MOF was achieved. The direct Cu II -O-W(Ru) bridges made the two metal catalyses being synergetic, enabling the application on the catalysis of the oxidative coupling C-C bond formation from acetophenones and N-phenyl-tetrahydroisoquinoline with excellent conversion and sizeselectivity. The method takes advantage of visible light photoredox catalysis to generate iminium ion intermediate from N-phenyl-tetrahydroisoquinoline under mild conditions and the easy combination with Cu-catalyzed activation of nucleophiles. Control catalytic experiments using similar Cu-based sheets but with the photoredox catalytic anions embedded was also investigated for comparison. ## Introduction The direct formation of new C-C bonds through oxidative coupling reactions from the lower active sp 3 C-H bonds using oxygen as oxidant is an important area in sustainable chemistry. 1 Among the reported promising examples, the oxidative activation of the C-H bonds adjacent to nitrogen atom in tertiary amines represents a powerful strategy, giving valuable, highly reactive iminium ion intermediates for further functionalization. 2 Recent investigations also revealed that the visible light photoredox catalysis was a promising approach to such reaction sequences 3 with respect to the development of new sustainable and green synthetic methods. It was also postulated that the combination of the photocatalysis and the metal catalysis within a dual catalytic transformation is attractive to circumvent the potential side reactions relative to the highly active intermediates that exist in the photocatalysis. 4 The hurdles that need to be overcome include the careful adaptation and the fne tuning of the reaction rates of the two catalytic cycles, 5 beside the appropriate choice of the metal catalysis and photocatalysis. 6 Metal-organic frameworks are hybrid solids with infnite network structures built from organic bridging ligands and inorganic connecting nodes. Besides the potential applications in many diverse areas, 7 MOFs are ideally suited for catalytic conversions, since they can impose size and shape selective restriction through readily fne-tuned channels or pores, 8 providing precise knowledge about the pore structure, the nature and distribution of catalytically active sites. 9 In comparison to the heterogeneous catalytic systems that have been examined earlier, the design flexibility and framework tunability resulting from the huge variations of metal nodes and organic linkers allow the introduction of more than two independent catalyses in one single MOF. 10 The combination of photocatalysis with the metal ions or organocatalysis was expected to be a promising approach to create synergistic catalysts. 11 By incorporating a ruthenium(III) substituted polyoxometalate [SiW 11 O 39 Ru(H 2 O)] 5 within the pores of copper(II)-bipyridine MOFs, herein, we reported a new approach to merge the visible light photocatalytic aerobic oxidation and copper(II) catalytic coupling reaction within one MOF (Scheme 1). We envisioned that the ruthenium-containing fragments possibly worked as oxidative photocatalyst to generate the iminium ion from N-phenyl-tetrahydroisoquinolines, 12 whereas the Cu-based MOF potentially activated the nucleophiles, as it was shown in the oxidative C-C bond coupling. 13 ## Synthesis and characterizations of CR-BPY1 Solvothermal reaction of 4,4 0 -bipyridine (BPY), Cu(NO 3 ) 2 $3H 2 O and K 5 [SiW 11 O 39 Ru(H 2 O)]$10H 2 O gave CR-BPY1 in a yield of 52%. Elemental analyses and powder X-ray analysis indicated the pure phase of its bulk sample. Single-crystal structural analysis revealed that CR-BPY1 crystallized in a space group P42 1 m. Two crystallographically independent copper(II) ions are connected by BPY ligands and m 2 -water bridges alternatively to produce 2D wavy-like Cu-BPY sheets (Fig. S5, ESI †). The Cu(2) atom adopted a six-coordinate octahedral geometry with four nitrogen atoms from four BPY ligands positioned in the equatorial plane and two water molecules occupied the axial positions. The Cu(1) atom displayed a fve-coordinate square pyramidal geometry with two m 2 -water groups and two nitrogen atoms of BPY ligands positioned in the basal plane, and a terminal oxygen atom of the depronated [SiW 11 O 39 Ru(H 2 O)] 5 polyoxoanion occupied the vertex position. The ruthenium atom disordered in the twelve equivalent positions within a depronated [SiW 11 O 39 Ru(H 2 O)] 5 . 14 The availability of vacant dorbitals on the metal atoms adjacent to the heteroatom allows the polyoxometalate matrix to function as a p-acceptor ligand. 15 While these copper atoms were connected by the BPY ligands to form two-dimension square grid at frst, adjacent sheets were connected together using the deprotonated [SiW 11 O 39 Ru(H 2 O)] 5 polyoxoanion by Cu II -O-W(Ru) bridges to generate a 3D framework. Two symmetric-related frameworks further interpenetrated each other perpendicularly to consolidate the robust structure (Fig. 1), in which the opening of the pores was reduced to 10.0 5.3 . To the best of our knowledge, CR-BPY1 represents the frst example of MOFs which are comprised of ruthenium substituted polyoxometalate [SiW 11 O 39 Ru(H 2 O)] 5 . As the noble metal substituted polyoxometalates exhibited excellent photoreactivity in various catalytic oxidation processes of organic substrates, 16 such kinds of MOFs potentially allow the combination of photocatalysis and MOF-based heterogeneous catalysis to achieve synthetically useful organic transformations. Moreover, the directly bridging of the copper and ruthenium by Cu II -O-W(Ru) provided a promising way to achieve the synergistic catalysis between photocatalyst and metal catalyst. Confocal fluorescence microscopy has attracted much attention in biological imaging. It may provide a way to analyse relatively thick porous materials, because it offers the advantage of increased penetration depth (>500 mm). 17 The assessment of guest-accessible volume in MOFs can be reliably done by using confocal fluorescence microscopy with a tool-box of dyes with a wide range of sizes. It would be applicable to any porous materials, whose single-crystal structures are not available, or non-crystalline materials. 18 Dye uptake investigation was carried out by soaking CR-BPY1 in a methanol solution of 2 0 ,7 0 -dichlorofluorescein. It gave the quantum uptake equivalent to 5% of the MOF weight (Fig. S11, ESI †). 19 The confocal laser scanning microscopy exhibited strong green fluorescence (l ex ¼ 488 nm) assignable to the emission of the fluorescein dye (Fig. 2), confrming the successful uptake of the dye molecules inside the crystals of the MOF. 20 Furthermore, the rather uniform distribution of the dye molecules throughout the crystal suggested that the dyes penetrated deeply into the crystal rather than staying on the external surface. Without guest water molecules, the effective free volume of CR-BPY1 was estimated to be 29.0% by PLATON software. 21 CR-BPY1 exhibited an absorption band centered at 398 nm in the solid state UV-vis absorption spectrum (Fig. S1, ESI †), assignable to the transitions of [SiW 11 O 39 Ru(H 2 O)] 5 . 22 Upon excitation at this band, CR-BPY1 did not exhibit any obvious emission, however, progressive addition of the N-phenyl-tetrahydroisoquinoline into the dichloromethane suspension of CR-BPY1 up to 0.50 mM caused the appearance of the Ru IIrelative emission band at about 422 nm (Fig. 3a). 23 The results suggested that CR-BPY1 oxidized N-phenyl-tetrahydroiso-quinoline to form the Ru II species and the iminium intermediate. 24 Electrospray ionization mass spectrometry of the CH 2 Cl 2 suspension containing N-phenyl-tetrahydroisoquinoline and CR-BPY1 after 3 hours light irradiation exhibited an intense peak at m/z ¼ 208. This peak was assignable to the relative imine ion, confrming that CR-BPY1 oxidized N-phenyl-tetrahydroisoquinoline to form the Ru II species and the iminium intermediate (Fig. S13, ESI †). The electron paramagnetic resonance (EPR) of CR-BPY1 exhibited the characteristic signal of Cu II with g ¼ 2.14 (Fig. 3c). Solid state electrochemical measurements (Fig. 3d) exhibited a broad redox band centred at 186 mV (vs. SCE) relative to the overlap of the Cu II /Cu I and Ru III /Ru II redox couples. The potentials were comparable to these Cu II and Ru III -containing catalysts, 25 and enabled CR-BPY1 to prompt the oxidative coupling of N-phenyl-tetrahydroisoquinoline with nucleophiles under light. 26 It seems that CR-BPY1 adsorbed the N-phenyl-tetrahydroisoquinoline in its pores and activated the substrate to form the iminium intermediate. ## Catalysis details of CR-BPY1 The catalysis was examined initially using N-phenyl-tetrahydroisoquinoline and nitromethane as the coupling partners, along with a common fluorescent lamp (18 W) as the light source. The resulting reaction gave a yield of 90% after 24 hours irradiation. The removal of CR-BPY1 by fltration after 18 hours shut down the reaction, and the fltrate afforded only 12% additional conversion for another 18 hours at the same reaction conditions. The observation suggested that CR-BPY1 was a true heterogeneous catalyst. 27 Solids of CR-BPY1 could be isolated from the reaction suspension by simple fltration alone and reused at least three times with moderate loss of activity (from 90% to 82% of yield after three cycles). The index of XRD patterns of CR-BPY1 fltrated off from the reaction mixture suggested the maintenance of the crystallinity (Fig. S14, ESI †). With the size of the microcrystals reduced to 2 mm by grinding CR-BPY1 crystals for 20 min, the time of the reaction giving the same conversion to that of the as-synthesized materials was reduced by about 10% (Fig. S15, ESI †). It seems that the MOFbased particles having well-defned size were really helpful for the catalytic reactions, but the size of the crystals did not dominate the catalysis directly. Control experiments for the C-C coupling reaction of N-phenyl-tetrahydroisoquinoline and nitromethane were carried out and summarized in Table 1. Almost no conversion was observed when the reaction was conducted in the dark (entry 7), while a very slow background reaction was observed in the absence of catalyst (entry 6), which demonstrated that both the light and the photocatalyst are required for efficient conversion to the coupling products. In addition, using the same equiv. of copper(II) salts or/and K 5 [SiW 11 O 39 Ru(H 2 O)] as catalysts, respectively gives conversions of 39%, 25% and 42% in homogeneous fashion (entry 3-5). These results suggested that the direct connection of copper(II) ions to [SiW 11 O 40 Ru] 7 anions not only enabled the dual catalysts to individually activate N-phenyl-tetrahydroisoquinoline and nitromethane, but also enforced the proximity between the potential intermediates i.e. the iminium ion and nucleophile, avoiding the unwanted side reactions or reverse reactions. 28 Although several examples of photocatalysts and metal copper catalysts have been reported to prompt the oxidative coupling C-C bond formation, CR-BPY1 represents a new example of a heterogeneous bimetal catalyst that merges the copper catalyst and the ruthenium(III) substituted polyoxometalate catalyst within one single material. The high The reactions were carried out in the presence of a common used secondary amine, L-proline, as an organic co-catalyst to activate the ketones. 30 In the case of the acetophenone as reactant with a fluorescent lamp (18 W) as the light source; the catalytic reaction gave a yield of 72%. Control experiments demonstrated that the use of K 5 [SiW 11 O 39 Ru(H 2 O)] or copper(II) salts as catalysts, only gave less than 25% of the conversions, respectively. The results indicated the signifcant contribution of cooperative effects of the individual parts within one single MOF. From the mechanistic point of view, the ruthenium(III) of the polyoxometalate [SiW 11 O 39 Ru(H 2 O)] 5 interacted with Nphenyl-tetrahydroisoquinoline to form iminium ions, whereas the copper atoms coordinated to the acetophenones weakly to form the enol intermediate that worked as active nucleophile for the oxidative coupling C-C bond formation. At the same time, the presence of copper ions could enhance the activation of N-phenyl-tetrahydroisoquinoline, benefting the synergistic catalysis between photocatalyst and metal catalyst. Importantly, in contrast to the smooth reactions of substrates 1-3, the C-C coupling reaction in the presence of bulky ketone (1-(3 0 ,5 0 -ditert-butyl [1,1 0 -biphenyl]-4-yl)-ethanone) 4, gave less than 10% conversion under the same reaction conditions (Table 2, entry 4). The negligible adsorption by immersing CR-BPY1 into a dioxane solution of substrate 4, coupled with the fact that the size of substrate 4 was larger than that of the channels, 31 revealed that 4 was too large to be adsorbed in the channels. Furthermore, it is suggested that the synergistic catalytic coupling reaction indeed occurred in the channels of the MOF, not on the external surface. ## Synthesis and catalytic characterizations of CR-BPY2 To further investigate the synergistic interactions between the inorganic copper and [SiW 11 O 39 Ru(H 2 O)] 5 anion, a reference compound CR-BPY2 was assembled using the same starting components but different synthetic conditions (hierarchical diffusion). CR-BPY2 was synthesized by a diffusion method in a test tube by laying a solution of 4,4 0 -bipyridine in acetonitrile onto the solution of K 5 [SiW 11 O 39 Ru(H 2 O)]$10H 2 O and Cu(NO 3 ) 2 $3H 2 O in water for several days in a yield of 59%. Elemental analyses and powder X-ray analysis indicated the pure phase of its bulk sample. Single-crystal structural analysis revealed that CR-BPY2 crystallized in the orthorhombic lattice with a space group Pccn. Two crystallographically independent copper(II) ions connected four BPY bridges alternatively to produce a 2D sheet (Fig. 5), which were further stacked paralleled along the crystallographic a axis to form the 3D structure with embedded [SiW 11 O 39 Ru(H 2 O)] 5 (Fig. S8, ESI †). The copper(II) ions resided in an octahedral geometries with the equatorial plane which was defned by four nitrogen atoms of BPY ligands, and the axial positions were occupied by two water molecules (Fig. S7, ESI †). Without guest water molecules, the effective free volume of CR-BPY2 was also estimated to be 33.9% by PLATON software, which is quite larger than that of CR-BPY1. These results suggested that the pore of CR-BPY2 is larger enough to adsorb the substrates. Since [SiW 11 O 39 Ru(H 2 O)] 5 polyoxoanions were embedded in the channels, it is thus an excellent reference for investigating the catalytic activity on the same coupling reaction. a Reaction conditions: N-phenyl-tetrahydroisoquinoline (0.25 mmol), 1 mol% catalyst, 2.0 mL nitromethane, 18 W fluorescent lamp at room temperature. b The conversions after 24 hour irradiation were determined by 1 H NMR of crude products. CR-BPY2 also exhibited an absorption band centered at 398 nm in the solid state UV-vis absorption spectrum. Upon excitation at this band, CR-BPY2 did not exhibit obvious emission, however, progressive addition of the N-phenyl-tetrahydroisoquinoline into the dichloromethane suspension of CR-BPY2 up to 0.50 mM caused the appearance of the Ru II -relative emission band at about 422 nm, suggesting that CR-BPY2 oxidized N-phenyl-tetrahydroisoquinoline to form the Ru II species. The EPR of CR-BPY2 exhibited the characteristic signal of Cu II (g ¼ 2.13). 32 The sharper peak shape compared to that of CR-BPY1 might be one of the indicator of isolated Cu II ions in CR-BPY2. No metal-metal interactions were found corresponding to the Cu II ions in CR-BPY2. Solid state electrochemical measurements exhibited two redox peaks corresponding to the Cu II /Cu I and Ru III /Ru II redox couples, with the redox potential calculated at 75 mV and 84 mV (vs. SCE). The potentials enabled CR-BPY2 to prompt the oxidative coupling of N-phenyl-tetrahydroisoquinoline with nucleophiles under light. However, the redox peaks also suggested that these Cu I and Ru III ions did not interacted directly. It seems that CR-BPY2 adsorbed the N-phenyl-tetrahydroisoquinoline in its pores and was a convincing reference to investigate the synergistic action between Cu II -O-W(Ru) bimetal of CR-BPY1. The catalytic activities of CR-BPY2 in the C-C coupling reactions were examined under the same conditions using nitromethane and N-phenyl-tetrahydroisoquinoline as the reactants. About 1 mol% loading amount of the catalyst gave rise to a 46% conversion, which was superior to the case when copper(II) salts and the K 5 [SiW 11 O 39 Ru(H 2 O)] were employed as catalysts, indicating the signifcance of the two constitute parts for CR-BPY2 as a photocatalyst. However, the catalytic activities of CR-BPY2 were signifcantly weaker than that of CR-BPY1 (Fig. 3b). It should be concluded that the direct bridging of the copper and ruthenium by Cu II -O-W(Ru) provided a promising way to achieve the synergistic catalysis between photocatalyst and metal catalyst, and the high reaction efficiency in the reactions was dominated by the spacious environment of the channels, like those of CR-BPY1. ## Conclusions In a summary, we reported the new example of copper MOFs containing the ruthenium substituted polyoxometalate with the aim of merging the synergistic Cu-catalysis/Ru-photocatalysis in a single MOF. CR-BPY1 exhibited perpendicularly inter-penetrated structure and the catalytic sites positioned in the robust pores of MOFs. Luminescence titration and IR spectra of the MOF-based material revealed the adsorbance and activation of N-phenyl-tetrahydroisoquinoline and acetophenone, by the ruthenium center and copper ions, respectively. The direct connection of copper(II) ions to [SiW 11 O 40 Ru(H 2 O)] 5 not only provided the possibility of the dual catalysts to individually activate the substrates, but also enforced the proximity between the intermediates, avoiding the unwanted side reactions or reverse reactions. CR-BPY1 exhibited high activity for the photocatalytic oxidative coupling C-C bond formation with excellent size-selectivity, suggesting the catalytic reactions occurred in the channels of the MOF, and not on the external surface. ## General methods and materials All chemicals were of reagent grade quality obtained from commercial sources and used without further purifcation. 1 H NMR was measured on a Varian INOVA-400 spectrometer with ## Synthesis of CR-BPY2 The CR-BPY2 was synthesized by a diffusion method in a test tube. A mixture of acetonitrile and water (1 : 1, 10. ## X-ray crystallography Data of CR-BPY1 and CR-BPY2 were collected on a Bruker Smart APEX CCD diffractometer with graphite-monochromated Mo-Ka (l ¼ 0.71073 ) using the SMART and SAINT programs. 34 Their structures were determined and the heavy atoms were found by direct methods using the SHELXTL-97 program package. 35 Crystallographic data for them are summarized in Table 3. Except some partly occupied solvent water molecules, the other non-hydrogen atoms were refned anisotropically. Hydrogen atoms within the ligand backbones were fxed geometrically at their positions and allowed to ride on the parent atoms. In both of the two structures, the ruthenium atoms were disordered in the equivalent positions of tungsten atoms. For CR-BPY2, several bond distances constraints were used to help the refnement on the BPY moiety, and thermal parameters on adjacent oxygen atoms of the polyoxometalate anion were restrained to be similar.

chemsum

{"title": "Merging of the photocatalysis and copper catalysis in metal\u2013organic frameworks for oxidative C\u2013C bond formation", "journal": "Royal Society of Chemistry (RSC)"}

nucleophilic_vinylic_substitution_in_bicyclic_methyleneaziridines:_s_nv_π_or_s<s

## Abstract: A stereodefined monodeuterated methyleneaziridine is shown to be prepared via coordinated reductive ring-opening of an alkynyl epoxide and diastereoselective tethered allene aziridination. Ring-opening of this aziridine with copper-based organometallics follows a pathway that results in stereoretentive substitution, replacing the exo-C-N bond with a corresponding C-C bond; this stereochemical outcome supports either an overall S N V p mechanism or a C-N insertion/reductive coupling process.Scheme 1 The formation and S N V ring-opening of fused 1,3-oxazolidin-2-one methyleneaziridines. ## Introduction In 2010 both we 1 and Blakey 2 reported the first examples of intramolecular allene aziridination with sulfamate substrates, with the major products being derived in most cases via 2-amidoallylcation intermediates. 3 Our group followed this up with the first report 4 of analogous reactions of carbamate substrates 1 (Scheme 1) and, in that work, somewhat unstable bicyclic 1,3-oxazolidin-2-one methyleneaziridines 2 were obtained following Lebel's modification 5 of the Du Bois protocol 6 for Rh(II)-nitrenoid generation. Soon afterwards, Schomaker's group took on the area and developed it extensively, optimising the conditions for generating the methyleneaziridines, engineering the substrates for synthetic tractability (non-terminal allenes, formation of 1,3-oxazinan-2-ones), and elaborating the products into a variety of hydroxy/amino stereotriads and -tetrads and rearranged heterocycles. 7 In our original publication we noted that the methyleneaziridines were constrained by the ring-fusion such that only the exocyclic aziridine C-N bond is electronically activated in the ground state through hyperconjugation with the carbamate carbonyl p-system. This suggested the possibility of effecting direct substitution/ ring-opening at the sp 2 -carbon, in contrast to the prevailing reactivity of unconstrained methyleneaziridines in which ring-opening occurs preferentially at the sp 3 -carbon. 8 At the time, the only sp 2 -C-N bondcleaving processes involved either transition metal-mediated processes 9 or stepwise radical addition/b-scission. 10 In the event, treatment of methyleneaziridine 2 (R = i-Pr) with lithium diphenylcuprate, or various Grignard reagents in the presence of CuI, led to moderate to good yields of the products 3 of nucleophilic vinylic substitution (S N V). 11 That publication concluded with an intention to clarify the stereochemical details of the S N V reaction; the current paper describes studies to that end. ## Results and discussion A stereochemically defined monodeuterated analogue 4 (Scheme 2) of methyleneaziridine 2 (R = i-Pr) was targeted that would allow the stereochemistry of the S N V process to be probed without presenting any steric or electronic bias compared with the original methyleneaziridine. At the outset of this study, a dissociative mechanism for the substitution reaction was ruled out on the basis of the aprotic, low-temperature conditions for the process and the relative instability of a vinylic cation. An out-of-plane (relative to the cleaving C-N bond) stepwise p-addition/elimination process, proceeding via a short-lived formal carbanion located on the terminal methylene carbon, or an equivalent concerted mechanism, would proceed with retention of configuration (S N V p pathway, -5). An in-plane concerted process, akin to an S N 2 reaction in aliphatic substrates, would lead to inversion of configuration (S N V s pathway, -6). The operation of either of these reaction modes would then be revealed in the relative disposition of the newlyformed C-C bond and the adjacent H/D atoms, as shown. In the absence of any literature precedent for the synthesis of a stereodefined terminally monodeuterated buta-2,3dienol, 12 a synthesis of methyleneaziridine 4 was proposed based upon diastereoselective coordinated delivery of hydride 13 to deuterated alkynyl epoxide 7 (Scheme 3) and the known stereochemical course of the intramolecular aziridination. Following this proposal, trans-2-ethynyl-3-isopropyloxirane 14 was stirred with an excess of D 2 O under basic conditions 15 to yield the deuterated alkyne 7 (94% deuterium incorporation). Alkyne 7 was treated with DIBAL in dichloromethane as a non-coordinating solvent that would support epoxide chelation with the aluminium centre, and allene 8 was isolated apparently as one predominant stereoisomer, 16 depicted as that expected, and confirmed retrospectively from the NMR data for methyleneaziridine 4. A slightly modified variant of Lebel's protocol for nitrenoid formation afforded consistent yields (B25%) of methyleneaziridine 4 from N-tosyloxy carbamate 9; lower yields were obtained from carbamate 10 with a range of Rh(II) catalysts including Rh 2 (OAc) 4 , Rh 2 (esp) 2 , 17 and Rh 2 (TPA) 4 . 18 The stereochemistry in methyleneaziridine 4 was confirmed by comparisons with the NMR data for non-deuterated methyleneaziridine 2, 4 and the NOE correlations shown in Fig. 1. In the 1 H NMR spectrum of methyleneaziridine 4, the adjacent methine protons at d 1.81 and 4.35 show 3 J HH = 9.5 Hz, indicating a dominating trans-antiperiplanar disposition that places one of the diastereotopic methyls more regularly in close proximity to the CHN and =CHD protons, as seen in the NOE spectra. A simple dihedral drive calculation supports this view (ESI †). 19 Two variants of the S N V reaction were carried out, both of which converted methyleneaziridine 4 into products with reasonable overall efficiency (Scheme 4). In the first, addition of lithium dimethylcuprate gave a 77% isolated yield of 4-isopropenyl oxazolidinone 11, in which the methyl group was found (see below) to be cis-to the deuterium atom. In the second, a copper-catalysed Grignard reaction with vinylmagnesium bromide gave 4-(buta-1,3-dien-2-yl) oxazolidine 12 as the major product, again with the new C-C bond formed cis-to the deuterium atom. The azirine 13 was also isolated in this work; its formation may be explained by competing addition at the carbonyl followed by 1,4-vinylation of the so-formed a,b-unsaturated ester. 20 A combination of NMR experiments, including NOE (Fig. 2) provided support for the stereochemical assignments in S N V products 11 and 12. Notably, in 11 no NOE correlation was observed between the vinyl methyl protons andQCHD; similarly, in compound 12, there were no significant correlations between the vinyl protons and QCHD. An invertive S N V s reaction appears to be stereoelectronically accessible in methyleneaziridines 2 and 4, and the microscopic reverse of such a process is supported in the NaNH 2 -mediated formation of simple methyleneaziridines from 2-bromoallylic amines. 21 Despite this, our results clearly rule out the S N V s mode of ring-opening, the stereochemical outcome being consistent with a (retentive) S N V p mode of reaction. Setting aside the extent of the involvement of the metal counterions in this process, at one simplistic mechanistic extreme, as the delivery of the methyl or vinyl ligand to the methylene group initiates and charge begins to build on the terminal carbon, the sp 2 -C-N bond weakens, with progression along this pathway ## Conclusions To the best of our knowledge, the direct nucleophilic sp 2 C-N bond cleavage reactions that we reported in 2010 remain the only examples in methyleneaziridine chemistry. In this work, we have demonstrated that the substitution is stereoretentive, ruling out an S N V s pathway, but the detailed mechanism of these reactions remains open to speculation and further work is intended to close this particular chapter of methyleneaziridine reactivity. 23 ## General information All solvents for anhydrous reactions were obtained dry from Grubbs solvent dispenser units after being passed through an activated alumina column under argon. THF was additionally distilled from sodium/benzophenone ketyl under argon. Commercially available reagents were used as supplied unless otherwise specified. Triethylamine was distilled from CaH 2 and stored over KOH pellets under argon. 'Petrol' refers to the fraction of light petroleum ether boiling between 30 and 40 1C; 'ether' refers to diethyl ether. All reactions were carried out in oven-dried glassware and under an atmosphere of argon unless otherwise specified. Thin layer chromatography (TLC) was carried out using Merck aluminium backed DC60 F254 0.2 mm precoated plates. Spots were then visualised by the quenching of ultraviolet light fluorescence (l max 254 nm) and then stained and heated with either anisaldehyde or KMnO 4 solutions as appropriate. Retention factors (R f ) are reported along with the solvent system used in parentheses. Flash column chromatography was performed using Merck 60 silica gel (particle size 40-63 mm) and the solvent system used is reported in parentheses. Infrared spectra were recorded using a Bruker Tensor 27 FT-IR fitted with a diamond ATR module. Absorption maxima (n max ) are reported in wavenumbers (cm 1 ) and are described as strong (s), medium (m), weak (w) or broad (br). Proton ( 1 H) and carbon-13 ( 13 C) spectra were recorded on Bruker AVIII HD 500, AVII 500, or AVIII HD 400 spectrometers. Chemical shifts (d H or d C ) are reported in parts per million (p.p.m.) downfield of tetramethylsilane, internally referenced (in MestReNova) to the appropriate solvent peak: CDCl 3 , 7.26/77.16; acetone-d 6 , 2.05/29.84. Peak multiplicities are described as singlet (s), doublet (d), triplet (t), quartet (q), septet (sept), octet (oct), multiplet (m), and broad (br) or a combination thereof. Coupling constants (J) are rounded to the nearest 0.5 Hz. Assignments are made on the basis of chemical shifts, integrations, and coupling constants, using COSY, HSQC and nOe experiments where appropriate. High Resolution Mass Spectra (HRMS) were recorded by the staff at the Chemistry Research Laboratory (University of Oxford) using a Waters GC-TOF spectrometer (EI/FI). Melting points were recorded on a Griffin melting point apparatus and are uncorrected. ## Trans-2-(deuterioethynyl)-3-isopropyloxirane (7) Trans-2-ethynyl-3-isopropyloxirane (2.03 g, 18.4 mmol) was added to a stirring solution of K 2 CO 3 (3.76 g, 27.2 mmol) in acetonitrile (42 mL). After 30 min, D 2 O (20 mL) was added and stirring was continued for 5 h. The product was extracted from the reaction mixture into petrol (5 100 mL). The combined Scheme 4 Reagents and conditions: (i) Me 2 CuLi (1.0 eq.), THF, 20 1C -RT, 30 min (11, 77%); (ii) vinyl-MgBr (2.0 eq), CuI (5 mol%), THF, 50 1C -0 1C, 1 h (12, 31%; 13, 23%). extracts were dried (MgSO 4 ) and the solvent was removed in vacuo [CARE: the product is volatile] to afford the title compound as a pale yellow oil (1.50 g, 73%, 94% deuterium incorporation). R f 0.58 (petrol/ether, 3 : 1); n max /cm 1 (thin film): 2966m, 2589m, 1980w, 1469m; d H (400 MHz, CDCl 3 ) 0.97 (3H, d, J = 7.0 Hz), 0.99 (3H, d, J = 7.0 Hz), 1.52 (1H, oct, J = 7.0 Hz), 2.29 (0.1H, d, J = 1.5 Hz, residual RCH), 2.89 (1H, dd, J = 7.0, 2.0 Hz), 3.11 (1H, d, J = 2.0 Hz); d C (100 MHz, CDCl 3 ) 18.1, 18.7, 30.4, 43.9, 65.4, 71.5 (t, J = 38.5 Hz), 80.3 (t, J = 7.5 Hz). (3R*,5R*)-6-Deuterio-2-methylhexa-4,5-dien-3-ol (8) A solution of epoxyalkyne 7 (1.50 g, 13.5 mmol) in dichloromethane (100 mL) was added dropwise to a stirred solution of DIBAL (33.7 mL, 1.0 M in hexane, 33.7 mmol) in dichloromethane (100 mL) at 0 1C. The reaction mixture was stirred for 1 h then quenched by careful addition of water. A satd. aq. solution of Rochelle's salt (200 mL) was added dropwise and the mixture was stirred overnight to allow the solvent layers to separate completely. The mixture was extracted with dichloromethane (5 50 mL), the organic layers were combined and washed with brine (50 mL), then dried (MgSO 4 ) and the solvent removed in vacuo. Purification by flash chromatography (petrol/ ether, 8 : 1) afforded the title compound as a pale yellow oil (810 mg, 53%). R f 0.23 (petrol/ether 3 : 1); n max /cm 1 (thin film) 3409br, 2957m, 2924s, 2854m, 1953w, 1464w, 1379w, 1261w, 1024w; DNNaO 3 , 195.0850; found, 195.0846. Recrystallised p-TsCl (619 mg, 3.25 mmol) was added to a stirred solution of the hydroxycarbamate (554 mg, 3.22 mmol) in dry ether (30 mL) at 0 1C. Triethylamine (0.45 mL, 3.23 mmol) was then added dropwise and stirring was continued for 18 h. The mixture was then diluted with ether (30 mL), washed with brine (2 20 mL), dried (Na 2 SO 4 ), and the solvent removed in vacuo. The crude product was purified by flash chromatography (petrol/ether, 5 : 1 pure ether) to afford the title compound as a pale yellow oil (893 mg, 85%). R f 0.50 (petrol/ether, 1 : 1); n max /cm 1 (thin film) 3284br, 2967m, 1770s, 1737s, 1598m, 1467m, 1379s, 1192s, 1179s, 1019m, 742m; d H (400 MHz, CDCl 3 ) 0.82 (3H, d, J = 7.0 Hz), 0.84 (3H, d, J = 7.0 Hz), 1.

chemsum

{"title": "Nucleophilic vinylic substitution in bicyclic methyleneaziridines: S_NV_\u03c0 or S_NV_\u03c3?", "journal": "Royal Society of Chemistry (RSC)"}

non-iterative_method_for_constructing_valence_antibonding_molecular_orbitals_and_a_molecule-adapted_

## Abstract: While bonding molecular orbitals exhibit constructive interference relative to atomic orbitals, antibonding orbitals show destructive interference. When full localization of occupied orbitals into bonds is possible, bonding and antibonding orbitals exist in 1:1 correspondence with each other. Antibonding orbitals play an important role in chemistry because they are frontier orbitals that determine orbital interactions, as well as much of the response of the bonding orbital to perturbations. In this work, we present an efficient method to construct antibonding orbitals by finding the orbital that yields the maximum opposite spin pair correlation amplitude in second order perturbation theory (AB2) and compare it with other techniques with increasing the size of the basis set. We conclude the AB2 antibonding orbitals are a more robust alternative to the Sano orbitals as initial guesses for valence bond calculations, due to having a useful basis set limit. The AB2 orbitals are also useful for efficiently constructing an active space, and work as good initial guesses for valence excited states. In addition, when combined with the localized occupied orbitals, and relocalized, the result is a set of 1 molecule-adapted minimal basis functions that is built without any reference to atomic orbitals of the free atom. As examples, they are applied to population analysis of halogenated methane derivatives, H-Be-Cl, and SF 6 where they show some advantages relative to good alternative methods. ## Introduction Virtual orbitals are important in chemistry as they play a central role in molecular orbital theory. From a computational standpoint, orbital mixing between occupieds and virtuals determines the optimal occupied orbitals in mean-field Hartree-Fock theory and Kohn-Sham density functional theory. In wavefunction theory, electron correlation is typically described by amplitudes such as the pair correlations describing the simultaneous promotion of two electrons from occupied to virtual orbitals. The virtual orbitals span the unoccupied space, and the choice of representation is important. Canonical virtual orbitals are delocalized levels that are appropriate for electron attachment. Localized virtuals, such as the redundant non-orthogonal basis of atomic orbitals projected into the virtual space, 8,9 permit development of efficient local correlation methods, because the amplitude tensors describing correlation become sparse. 10 Other prescriptions for localized orthogonal virtuals exist, as well as proposals to form sets of virtuals that are specifically optimized for correlations that involve a given occupied level, as will be discussed below. The virtual orbitals span the entire unoccupied space, which can be contrasted with the intuitive notion of antibonding orbitals that exist in 1:1 correspondence with bonding orbitals. The 1:1 correspondence is evident from constructive and destructive interference of a pair of 1s-type functions on two hydrogen atoms in H 2 : Antibonding orbitals themselves play a central role in describing chemical reactivity of one molecule with another through donor-acceptor interactions between a high-lying occupied of one species with a low-lying antibonding orbital of the other. Frontier orbital theory is constructed on these ideas. Antibonding orbitals also play an important role in describing strong electron correlations. A simple example is the stretching of the H-H bond which leads, in a minimal basis, to a strong increases in the amplitude for σσ → σ σ excitation which breaks the bond. While the antibonding orbitals are intuitive, 20 it is nonetheless not routine to extract them from modern quantum chemistry calculations performed in extended basis sets, which return canonical orbitals. By contrast, in a minimal basis description of hydrocarbons, the space of antibonding orbitals is naturally spanned by the canonical virtual orbitals. In larger basis sets however, different methods have been developed to extract the antibonding orbitals, often by relying on projection back onto some chosen minimal basis, 13, typically a tabulated one for a specific free-atom Hartree-Fock energy eigenstate. For example, Schmidt et. al. found antibonding orbitals by performing an SVD of the overlap between the virtual orbitals and a minimal basis to produce valence virtual orbitals. 13 Some methods have been developed to produce a minimal basis specifically adapted to a molecular environment, 26,27 but those are non-linear optimization procedures that are often iterative and costly. One famous method that does not rely on a reference minimal basis is the Natural Bond Orbital (NBO) procedure, 14,15 where the density matrix coupling between multiple atom-tagged orbitals is utilized to produce bonding and anti-bonding orbitals. However, atom tagging of basis functions plays a critical role in the NBO procedure -in fact, the standard NBO method is specific to atom-centered orbital (AO) basis calculations. Few methods cut the umbilical cord to the minimal basis in producing antibonding orbitals. Aside from the Sano antibonding orbitals 28 (discussed below), Foster and Boys 29 suggested oscillator orbitals which are virtual orbitals with the maximum dipole from localized occupied orbitals. Local correlation has been intensively studied, 8,9,25, leading to the conclusion that dynamic correlation can be well approximated using domains of localized virtual orbitals that are in the same spatial region as a localized occupied orbital. 8,9 This reduces the 4th rank tensor of pair correlation amplitudes to an asymptotically linear number of significant elements. Nevertheless, all virtual orbitals are required for post-SCF methods such as coupled cluster theory that recover dynamic correlation, rather than just the much smaller set of valence virtual orbitals. By contrast, static or strong correlation, resides mostly in the valence virtuals (i.e. the antibonding orbitals). Thus complete active space (CAS) methods that seek to describe strong correlation require only a description of the valence virtuals. Methods in this class include CASSCF, 2,39-42 spin-coupled valence bond (VB), and approximations such as generalized valence bond (GVB), 47 coupled cluster valence bond (CCVB), etc. CAS, GVB and CCVB methods thus need an initial guess for the antibonding orbitals. We do note that the orbitals associated with key amplitudes for strong correlation are not necessarily spatially localized. One method used to obtain initial guess antibonding orbitals is the so-called Sano procedure. 28 In brief, after localizing a set of occupied orbitals using standard methods, 12, the Sano procedure finds the virtual orbital that has maximum exchange interaction with each given localized occupied orbital. The idea of maximizing exchange is very old 58,59 and comes from its predecessor, the modified virtual orbitals 60,61 (note that modified virtual orbitals have been since used to refer to any non-canonical set of virtual orbitals 62 ). The resulting orbitals are symmetrically orthogonalized to yield a set of valence antibonding orbitals. This method has worked quite well for GVB-PP and CCVB calculations in moderately sized basis sets. 63,64 In this work we will show that the Sano procedure shows undesirable behavior with increasing the size of the AO basis set. This motivates the need for a better behaved alternative. We suggest that finding the antibonding orbital which gives the largest first order perturbation amplitude for exciting an electron pair from a given bonding orbital is a suitable alternative. A range of numerical results confirm this to be the case. These antibonding orbitals can be viewed as a specific instance of orbital specific virtuals. 2 Theory ## Defining the set of antibonding orbitals Solving the mean field Hartree-Fock (HF) equation self consistently gives the lowest energy single Slater determinant electronic wave function. To solve the many-body problem, one needs to include the missing correlation energy. 65 Second order Møller-Plesset (MP2) perturbation theory 66,67 offers a useful and computationally inexpensive approximation to treat the correlation yielding the following expression in the case of restricted HF orbitals: where This expression folds together contributions from the correlation of two electrons of opposite spin (OS), with amplitudes: together with the contribution of correlations of electrons with the same spin. The twoelectron repulsion integrals (ERIs) over spatial orbitals describing the interaction of each occupied with each virtual are: Let us collect the ERIs associated with occupied orbital i into the symmetric matrix K i , where: K i is positive semi-definite, and thus the eigenvector belonging to its largest eigenvalue will correspond to the virtual level with the strongest exchange interaction with occupied level i. That is the Sano prescription 28 for finding the antibonding orbital associated with i. We can likewise define a matrix of second order pair correlation amplitudes, T i , associated with a given occupied orbital: This matrix is negative semi-definite since the denominators are negative for the ground state determinant. We can therefore find the largest OS pair-correlation amplitude as the lowest eigenvalue, t i max of T i , and the associated virtual orbital, |i * = a |a c ai * is the eigenvector, with expansion coefficients c ai * in the original virtual basis: Upon repeating for each occupied level, most naturally in a localized representation, and using similar arguments to Kapuy's zeroth in the Fock and 2nd order in correlation approximation, 68,69 we suggest that this is an appropriate non-iterative way to find a set of antibonding orbitals in 1:1 correspondence with the bonding orbitals. This approach may be contrasted with Sano's suggestion to obtain the virtual orbital with maximum repulsion from the bonding orbital by solving the eigenvalue problem for each orbital using K i rather than T i . Inclusion of orbital denominators in Eq. 9 provides a clear physical meaning of the antibonding orbital as having strongest pair correlation amplitude with its parent bonding orbital. As will be demonstrated numerically later, this property also dramatically improves basis set convergence relative to the Sano definition. We will refer to these virtual orbitals as "second order antibonding" (AB2) MOs to emphasize their second order origins, and their 1:1 correspondence with bonding MOs. In terms of existing literature, the AB2s are directly related to the "orbital-specific virtual" (OSV) orbitals 30,31 that are sometimes used to evaluate the correlation energy. Each AB2 orbital is the most important OSV for a given localized bonding orbital. Of course the reason for selecting the amplitudes associated with MP2 is computational efficiency. The exact limit of this procedure would be to diagonalize the corresponding exact (i.e. from Full CI) doubles amplitudes; T i ab , via Eq. 9. A closely related alternative that has some advantages over Eq. 9 above is to define the space of valence antibonding orbitals from the virtual-virtual block of the MP2 one-particle density matrix: 70,71 Upon diagonalizing, the (M − O) eigenvectors with largest occupation numbers span the valence antibonding orbital space, and, together with the occupied space, complete the span of a molecule-adapted minimal basis. Localization of these valence virtual orbitals will then yield an alternative to the localized virtuals above. The advantage of this approach is for cases where there is no simple 1:1 mapping between bonding and antibonding orbitals, as discussed more later. The virtual orbitals obtained this way are the valence subset of the "frozen natural orbitals" (FNO), 70,71 and we emphasize that they are not generally localized in contrast to the AB2 MOs. They are close to the virtual natural orbitals associated with P MP2 as defined by the gradient of the MP2 energy, with the caveat that only the virtual-virtual block is diagonalized. ## Population analysis using the effective minimal basis Finding a suitable set of antibonding orbitals provides the missing part of the valence space not spanned by the occupied orbitals. Thus the union of the occupied space and the space of antibonding orbitals spans the space of an effective minimal basis. It is well accepted that full valence CASSCF wavefunction is spanned by an effective minimal basis within the molecule for this reason. 39,40 Accordingly, localizing the union of the occupied orbitals with the antibonding orbitals reveals a set of molecule-adapted atomic orbitals (MAOs): 75,76 For a given pair of well-localized bonding and antibonding orbitals (say σ and σ * ), this procedure amounts to inverting Eqs. 2 to discover the corresponding MAOs even though we may be using a very extended basis, or even a non-atom-centered basis, such as plane waves or a real-space grid, to perform the calculations. The resulting MAOs, χ are thus expressed in terms of the AO's, ω, as χ = ωC MAO . The MAOs are orthogonal, and typically localize onto atoms. The MAOs exactly span the space of the occupied orbitals, and can be used for population analysis among other things. 26,40, Let us denote p as an MAO label for χ p , which is centered at r p = χ p |r|χ p . Using A, B as atom labels, and given that the density matrix in the MAO basis is one can make a population analysis as follows: where Q A and Z A are the atomic charge and the nuclear charge, respectively. Such a population analysis has no dependence on atom-tagging of the underlying basis, and does not rely upon a reference minimal basis. Therefore it generalizes nicely to plane wave basis and real space methods. If the orbitals are unrestricted, we construct antibonding pairs and the MAOs for the alpha and beta spin spaces independently. This approach to generating an MAO representation does have some limitations. First, it assumes that there is a 1:1 mapping between bonding and antibonding orbitals. One class of exceptions can be found in electron deficient molecules (e.g. LiH will not recover 2p-like orbitals on Li, and BH 3 will not recover a 2p z orbital on B). Such species can be said to have "virtual lone pairs", whose identification is a problem that we shall not address here. A second class of exceptions lie in species such as cyclopentadiene anion, where there are 3 semi-localized π occupied orbitals, but the valence space only admits 2 antibonding orbitals. Thirdly, in symmetric systems with multiple Lewis structures (e.g. C 6 H 6 ), the MAOs will derive from localized bond and antibonding orbitals corresponding to a single Lewis structure and may not reflect the indistinguishability of the atoms. Broadly, we can say that this MAO approach is readily applicable to neutral molecules with a single dominant Lewis structure. ## Implementation details Computational efficiency is very important for quantum chemistry in order treat molecules that are as large as possible for given computational resources (computer speed, memory size, etc). Our AB2 implementation uses exact 4-center integrals in a basis of Gaussian-type atomic orbitals (other alternatives such as using auxiliary basis expansions can also be readily implemented). Each step with its computational complexity is shown in Fig. 1. Note that for the figure and the discussion here we use O, V , and N for the number of bonding orbitals, virtual orbitals, and AO basis functions, respectively. We start by making a pseudo-density To generate the two-electron integrals (µν|λσ), Q-Chem 84 only generates significant µν (i.e. AO basis) pairs to some target numerical cutoff, yielding a total that we term as for small systems but approaches linear scaling (i.e. (N N ) cut ∝ N ) in the limit of large system size. The integrals are made and contracted on-the-fly with the bonding orbitals' pseudo-densities to make bonding-specific exchange integrals K i µν with compute effort scaling as O(O(N N ) 2 cut ). The K i µν matrices are then transformed into the virtual space as K i ab in Eq. 7 with compute cost scaling as O(OV N 2 + OV 2 N ). Asymptotically this is the dominant step in this method unless more careful thresholding is considered. 85 Then, we divide by the appropriate denominator to get T i ab in Eq. 8 (with O(OV 2 ) effort). Lastly, we diagonalize T i for each bonding orbital to get the AB2 antibonding orbitals as in Eq. 9 with O(OV 3 ) effort. Note that the last step can in principle be made O(OV 2 ) since we are only solving for the eigenvector with the largest amplitude in each matrix. We can contrast this procedure with the modified FNO approach which has a dominant computational step that scales as the 5 th power of molecule size: constructing P ab in Eq. 10 with complexity of O(O 2 V 3 ). Figure 1: A chart illustrating the mathematical steps needed to construct AB2 orbitals with the appropriate computational complexity for each step indicated. Here, O, V , N , and (N N ) cut refer to the number of occupied orbitals, virtual orbitals, AO basis functions, and significant AO pairs, respectively. One reason for the efficiency of the AB2 approach compared to FNO comes from focusing on the bonding orbitals one at a time rather than the whole occupied space at once. It is then important to start by localizing the occupied space, which is known to be a cubic scaling iterative procedure for e.g. the Boys and Pipek-Mezey localization measures. 86,87 Then, one must also distinguish between localized orbitals with different character: specifically core, bonding, and non-bonding, e.g. lone pairs. Our implementation uses an automatic bonding detection option that runs before AB2. The detection process is simply determined by Pipek's delocalization measure 88 on Mulliken charges, where measures amounting to 1 indicate an orbital localized on an atom (core or non-bonding) and measures around 2 correspond to orbitals split between two atoms. ## Computational details All methods discussed here were implemented in a developer version of Q-Chem 5. 84 The geometries used for molecular calculations were optimized at the ωB97X-D/def2-TZVPD level of theory. All geometries are included in the Supplementary Material (SI). ## Results and discussion We will compare different approaches to generating effective antibonding orbitals: in particular we are interested in whether the second order antibonding (AB2) MOs significantly improve upon the Sano antibonding orbitals, as measured by usage-relevant metrics obtained from a set of numerical experiments. We will first examine orbital plots, orbital energies, and orbital variances. We then test the applicability of Sano and AB2 MOs to several valence correlation methods: coupled cluster valence bond (CCVB), 49,50 complete active space configuration interaction (CASCI), and complete active space self-consistent field (CASSCF). Next, we look into their uses for describing valence excited states. For basis set, we are using the Dunning basis set family 93 and Ahlrichs. 94 These are available in Q-Chem 5.3 with an automated detection of bonding orbitals. ## Orbitals, orbital Energy, and orbital variance We start by looking at the σ * orbital of H 2 , as shown in Fig. 2, evaluated by the Sano procedure, the AB2 approach, and CAS(2,2) (performed as 1-pair perfect pairing). It is visually clear that the Sano σ * orbital is contracting as the basis set is improved. Fig. 3 displays the orbital energy (diagonal matrix element of the Fock operator) and the variance ( r 2 − r 2 ) of the σ bonding orbital, and the Sano and AB2 models of the antibonding orbital. The variance confirms that the size of the Sano σ * -orbital contracts with basis size, while its orbital energy increases (reflecting increasing electron confinement) unsatisfactorily. By contrast the behavior of the AB2 orbital is very close to the bonding orbital, with pleasing stability in both energy and variance as the basis set is converged towards completeness. The stark difference is due to Sano orbitals including high energy orbitals to maximize the exchange interaction whereas AB2 biases against those higher energy orbitals with the denominator penalty in Eq. 8. orbitals. We will therefore use Pipek-Mezey orbitals whenever we encounter π orbitals. Inspecting the σ C-C bond in C 2 H 4 in Fig. 4 shows that the shape of the occupied Pipek-Mezey and converged CCVB bonding orbitals both do not change much upon increasing the size of the basis set. By contrast, when looking at σ * in Fig. 4 we see even poorer behaviour of the Sano C-C antibonding orbital as a function of basis set size than we did for H 2 . This is confirmed in Fig. 5 where we compare the orbital energy and the orbital variance of the bonding and the antibonding C-C σ orbital in C 2 H 4 . The Sano σ * orbital does not converge with the size of the basis set, with the variance decreasing, and the energy increasing. By contrast, the AB2 σ * orbital converges rapidly both in terms of energy and variance for similar reasons to before. In Fig. 5 we compare the orbital energy and the orbital variance of the bonding and the antibonding orbitals for the C-C π in C 2 H 4 . The shortcomings of Sano seem to be much less severe in π * orbitals. We believe this is due to the diffuse nature of the π orbitals making the maximum exchange, thus spatial locality, sufficient to describe the π * . However, we can still see that the orbital energy and variance do not converge for Sano while they do for AB2, and converge to drastically different orbital energy and orbital variance. The quantitative advantage of the AB2 antibonding orbitals relative to the Sano orbitals seen so far can also become qualitative advantages in systems with more complex electronic structure. One such example is Cu 2 , which, considering that the valence state of Cu can be taken as 3d 10 4s 1 , is isoelectronic to H 2 . The σ orbital (HOMO) of Cu 2 is shown in the upper panel of Fig. 6, along with the optimized correlating orbital from CCVB, as well as the Sano and AB2 antibonding orbitals. Maximizing exchange results in a Sano antibonding orbital that resembles an empty π-bond between the two metals. By contrast, the AB2 and CCVB orbitals look qualitatively identical. Figure 6: Comparison of the shape of the orbitals in Cu 2 where the σ bond is used to produce Sano and AB2 antibonding orbitals. While the AB2 method produces very similar orbitals to CCVB, the Sano approach fails to give a qualitatively correct antibonding orbital. ## CCVB iterations The CCVB method is a simple low-scaling approximation to exponentially scaling spincoupled valence bond theory that can separate a system of 2n electrons into fragments with spin purity, provided that UHF can also reach the dissociation limit. One price to be paid for these advantages is a challenging orbital optimization problem: the CCVB orbitals have no invariances to rotations within the active space, in contrast to CASSCF. Hence a good initial guess is very important. Sano orbitals 28 have been commonly as a starting guess for valence bond methods 51,95 such as CCVB due to their resemblance to antibonding orbitals. For simple alkanes, we examine how many iterations are needed to converge a CCVB calculation with Sano and compare with AB2 shown in Fig. 7 with increasing the molecule size and the basis set size (using the Dunning cc-pVXZ sequence of basis sets 93 ). Since the doublezeta basis set does not involve many high energy orbitals, both methods converge almost at the same speed. Upon increasing the size of the basis set, overly-contracted Sano orbitals deviate more from the optimal antibonding orbitals, and therefore require far more iterations to converge. For this reason, we recommend using AB2 orbitals as a starting guess for valence bond methods instead of the Sano orbitals. Figure 7: Number of iterations needed to converge CCVB calculations on alkanes of increasing size, with increasing ζ of the basis set. This shows a relatively constant number of iterations needed for AB2 regardless of system size, while the number of iterations rise unfavorably for the Sano guess in large basis sets. Geometric direct minimization (GDM) 95,96 is used to determine the steps. ## CAS methods The relative fraction of correlation energy recovered using AB2, Sano, FNO or other choices for antibonding orbitals to complete an active space can help us discern which ones are most appropriate to use for configuration interaction with fixed orbitals, as well as for a CASSCF initial guess. As a simple example, we stretch the C-C bond in C 2 H 4 while keeping the geometry of the methylene groups fixed at those of the equilibrium ground state geometry of ethene. Looking at Fig. 8, we see that canonical virtual orbitals capture less and less correlation as the def2 basis set is improved from SVP to TZVPP to QZVPP. We also observe that the gap between Sano and AB2 orbitals increases with increasing the size of the basis set. Finally, we can see that FNO and AB2 orbitals perform almost identically and are the Since the AB2 and FNO orbitals seem to capture quite a lot of the static correlation, we sought to compare them to CASSCF orbitals. In Fig. 9 we are comparing the smallest singular value of the overlap matrix between the CASSCF orbitals and those of canonical, Sano, AB2, and FNO, at the optimized geometry of C 2 H 4 . Once again, the canonical orbitals become dramatically worse with increasing the basis set size. Sano and FNO both become very slightly worse with increasing the size of the basis set, namely by increasing zeta, while AB2 seems to be nearly basis set-independent. Figure 9: The smallest singular value from the overlap of CASSCF (12e,12o) orbitals with those from Sano, AB2, FNO and canonical orbitals. Canonical orbitals with the lowest energy and FNOs with the highest occupancy were selected. Canonical orbitals are differ strongly from optimized CASSCF orbitals while AB2 orbitals have the highest agreement. ## Excited States Since the AB2 orbitals seem to be good guesses for GVB methods, and yield orbitals close to converged CASSCF orbitals, this led us to believe that they could also provide a good description of valence excited states. State-specific methods, such as orbital-optimized DFT (OO-DFT) 97 need a suitable starting guess, as convergence is typically to the nearest stationary point., 98 so we used Sano and AB2 guesses for the π → π * excitation in methanal (H 2 CO). For our purposes we employed the square gradient minimization method 98 which looks for saddle points in the orbital Hilbert space to converge restricted open-shell Kohn-Sham (ROKS). 97,99,100 In Fig. 10 we compare the overlap of the π * orbital from converged singlet open shell HF calculations with Sano and AB2 orbitals. For this excitation, AB2 orbitals overlap the optimized orbital by at least 0.9, and vary minimally with the size of the basis set. We note here that aside from the double-zeta case, converging the excited state starting from the Sano orbital sometimes lands on a Rydberg excited state, while AB2 landed on the correct π * state in all cases. Figure 10: The overlap of the converged ROKS-HF antibonding orbital with the Sano and AB2 initial guesses in H 2 CO for the π → π * excitation. The π * orbital is well described by AB2 regardless of basis set size. ## Population Analysis Antibonding orbitals belong to the valence space, and contribute to making a minimal basis that can be used to gain insight into chemistry, for instance via population analysis to assign effective charges on each atom. The population analysis we present here is constructed from the union of the occupied space and the antibonding orbitals without dependence on the basis set used. To study our atomic charge predictions and compare it to some other methods in the literature, we look into fluoro-and chloro-substituted methanes which have been studied theoretically and experimentally. 104,105 These simple systems are nonetheless interesting because they manifest the effect of substituting electron withdrawing halogen atoms of different sizes and electronegativities for hydrogen in methane. How consistent or inconsistent are different atomic population analysis schemes as descriptors of these chemical substitutions? In Fig. 11 we examine the effect of progressive substitution of hydrogen by chlorine and fluorine in the methane molecule on the computed net charge at the C atom. We consider some commonly used methods, specifically charges on electrostatic potential grid (ChElPG), 106 iterative Hirshfeld (Iter-Hirsh), 107,108 intrinsic atomic orbitals (IAO), 109,110 and the method presented in this work, molecular atomic orbitals (MAO). Most obviously, the charge transferred upon halogen substitution will depend strongly on the electronegativity difference between X and H. Furthermore, while halogens are more electronegative than hydrogen (or carbon), the electron donating capacity of C is not unlimited, and so we expect the first halogen substituted to pull away a greater fraction of an electron from C compared to the next, and so forth. Such a change will also have some dependence on the X vs H electronegativity difference. With these preambles aside, atomic charges are not observables and therefore no single answer should be viewed as strictly correct. Nevertheless, we can examine the results of each population analysis for signs of incorrectness relative to physical intuition. Figure 11: The charge on the carbon atom for successive chlorination and fluorination of methane predicted using four different population analysis methods (see text for the names). The triangle, square, hexagon, and octagon correspond to charges using def2-SV(P), def2-SVPD, def2-TZVPD, and def2-QZVPD, respectively. For instance, while all methods agree that the C -H bonds of CH relative to the slight positive charge predicted by IAO. Finally we examine an unusual linear molecule, which is the result of insertion of Be into HCl, yielding H -Be -Cl. 111,112 While H -Cl is polarized as H δ+ Cl δ− , Be has lower electronegativity than H, and so there will be substantial charge transfer. Indeed the ionic limit would be H -Be 2+ Cl -. What then, is the actual charge distribution when we consider the covalent character of the molecular orbitals? The calculated populations are shown in Fig. 12, and it is immediately evident that predicted charges on Be vary widely. The least polar picture comes from ChElPG and MAO, with q(Be)≈+0.5, while the IAO scheme suggests q(Be)≈+1. 35. How should we understand this dramatic difference and suggest which might be more correct? Figure 12: The charges on each atom in the BeHCl molecule predicted by the four methods mentioned in the text. The triangle, square, hexagon, and octagon correspond to charges using def2-SV(P), def2-SVPD, def2-TZVPD, and def2-QZVPD, respectively. From the MAO perspective, there are two σ bonds involving Be, one with H and one with Cl. Each is made from sp hybrid orbitals on Be, meaning that the p orbitals of Be are at play in this σ bonding, as shown in Fig. 13. These bonds are both polarized away from Be, as expected. The origin of the much larger IAO charge can now be understood. The IAO reference minimal basis set, known as 'MINAO', 109 does not include 2p orbitals for Be, and therefore we are instead seeing essentially only the Be(2s) charge via the IAO approach! Iterative Hirshfeld evidently struggles with a similar issue, leading to similar overestimation of Be charge. Overall, this case nicely illustrates the advantages of the MAO population scheme that is based entirely on the system at hand, rather than some reference atomic orbitals or states. Next we study the hypervalent molecule, SF 6 , which has O h symmetry, and whose chemical bonding has long been of interest. 113 While empty 3d functions on sulfur are needed to form 6 equivalent sp 3 d 2 hybrids, the energetic cost of promoting electrons to the 3d shell is too high for d-orbital participation in the bonding to be chemically important. 46, Rather, the bonding may be thought of as resonance between Lewis structures with 4 covalent S-F bonds, and 2 F − anions, with a formal charge of +2 on S. 118 There are some limitations associated with reducing the union of the localized occupied orbitals and the AB2 antibonding orbitals to a set of MAOs that should be mentioned. First, some conjugated π systems, such as benzene and C 5 H 5 -, present a multiple minimum solution problem for orbital localization methods. Since our method relies on the localization procedure heavily, we expect there will be inconsistencies in these systems. For example, in benzene, there are different sets of solutions for the localized π orbitals, nominally corresponding to the two different Kekule structures. Using the Boys localized orbitals yields populations that reflect D 6H symmetry, while the Pipek-Mezey scheme gives alternating charges on successive carbons going around the ring. There is a second class of molecules that are inaccessible in our method. These are anions where the natural valence minimal basis is too small to provide an antibonding orbital for each bonding orbital. One such example is C 5 H 5 -, the cyclopentadienyl anion. Forming the set of AB2 valence orbitals and taking the union with the occupied space leads to a set of orbitals that cannot be localized to atoms. Broadly, we can say that neutral species with a single Lewis structure are wellhandled by the approach described here; as well as some more complex bonding situations like SF 6 discussed above. ## Conclusion We presented a relatively cheap, non-iterative procedure to produce a set of antibonding orbitals that vary minimally with the size of the atomic orbital basis set from which they are constructed. Specifically, antibonding second order (AB2) orbitals show far less variation with basis than the Sano orbitals which are sometimes used as valence antibonding orbitals. We showed that use of AB2 rather than Sano orbitals as initial guesses provides improved convergence for valence bond methods (specifically CCVB), as well as for CASSCF. The AB2 orbitals were successfully used as guesses for state-specific ROKS calculations of excited states, where they better resemble the converged orbitals than does the corresponding Sano orbital guess. We have shown how these AB2 orbitals can be used with the localized occupied orbitals to construct an effective minimal basis that can be used for population analysis among other things. Population analysis on the substituted fluormethane and chloromethane sequence shows the method is stable and consistent with other common methods that accord with chemical intuition. For the insertion of Be into HCl, the resulting charges show some advantages. Overall, the AB2 antibonding orbitals are relatively efficient to compute and quite useful for a variety of applications in quantum chemistry.

chemsum

{"title": "Non-iterative Method for Constructing Valence Antibonding Molecular Orbitals and a Molecule-adapted Minimum Basis", "journal": "ChemRxiv"}

layered_double_hydroxide_of_cd-al/c_for_the_mineralization_and_de-coloration_of_dyes_in_solar_and_vi

## Abstract: Cd-Al/C layered double hydroxide (Cd-Al/C-LDH) and Cd-Sb/C nanocatalyst are reported here for the decoloration and mineralization of organic dyes. These catalysts were largely characterized by FESEM, EDS, XRD, FTIR, XPS, PL and DRS. The diffuse reflectance data showed a band gap at 2.92 and 2.983 eV for Cd-Al/C-LDH and Cd-Sb/C respectively. The band gap suggested that both catalysts work well in visible range. The photoluminescence spectra indicated a peak at 623 nm for both the catalysts which further support the effectiveness of the respective catalyst in visible range. Both catalysts also showed good recyclability and durability till 4 th cycle. Five dyes, acridine orange (AO), malachite green (MG), crystal violet (CV), congo red (CR) and methyl orange (MO) were used in this experiment. Various parameters of different light intensity such as visible, ultraviolet, sunlight and dark condition are observed for the de-coloration of these dyes. The de-coloration phenomenon was proceeded through adsorption assisted phot-degradation. The low cost, abundant nature, good recyclability and better dye removal efficiency make these catalysts suitable candidates for the de-coloration and mineralization of organic dyes.Dyes industries play an important role in the progress and development of a country and make the human life beautiful. In pre-historical time people used various dyes to make their environment gorgeous. Most of the dyes stuff are categorized on the basis of its coloring properties, solubility, and chemical nature 1 . These dyes are used to color our clothes, food materials and beverages and even make some medicine colored. Literature survey revealed that there are approximately 10,000 commercially available dyes and 7, 00,000 tons are manufactured per annum worldwide [2][3][4] . During coloring practices most of the dyes were not intact and therefore, a large percentage of these remaining dyes were dumped into the stream. Approximately 10-15% of the dyes stuff are discharge into the environment which are esthetically not favorable 2 . Recently, dyes stuff are of pronounced environmental distress because of their carcinogenicity and mutagenicity 5 . More than 90% of approximately 4000 dyes were experienced in an ETAD survey (Ecological and Toxicological Association of the Dye stuffs) indicating more than 2 × 10 3 mg/ kg LD 50 values. The highest toxicity were found in diazo and basic dyes 6 . UK make an environmental policies in Sept. 1997 according to which zero synthetic chemical substances are to be released in the marine environment and ensure that the textile industries should treat their effluent before discharging into the water resources 6 . Developed countries and European community become more rigorous to control the dyes stuff from the industrial effluents 7 . Dyes industries contributed to the development of a country, but unfortunately most of the dyes stuff are discharge into our water resources without any treatment 8 . This led to the contamination of underground water resources by passing from soil to water beds and by owing its carcinogenic nature exposing human health and other organisms to high risk. Dyes stuff in minute quantity colorized the water and make a foam like layers on the surface of water 9 . These layers halted the penetration of sunlight and oxygen into water which finally led to the death of aquatic flora and fauna 10 . Therefore, it is very important to purify these effluent before discharging into the water resources.Dyes removal from water is one of main environmental problem due to its carcinogenic and mutagenic nature to aquatic life [11][12][13] . Previous methods such as physical, chemical and biochemical are not useful for the removal of dyes from the effluent 6 . Several methods have been reported in the literature for the de-coloration techniques such as activated carbon 7,14-18 however, the advanced oxidation processes (AOPs) effectively used for the degradation and mineralization of dyes into CO 2 and some inorganic ions. During the removal of dyes the AOPs generated highly reactive species known as ROS (reactive oxygen species i.e. O •2− , • OH, or HO •2 ), which possibly invade almost all contaminants. In all known AOPs, the heterogeneous photocatalytic process is the most effective method because of its high availability, low toxicity, inexpensive and diverse nature that might attacked and mineralize a large number of contaminants 13 . A class of new materials with hydrotalcite like structure known as layered double hydroxide recently attract researchers attention due to their varied application and high anion exchange capacity which make them suitable candidates for catalysis 19 . Hydrotalcite is comparable to brucite like structure Mg(OH) 2 , in which divalent Mg +2 is octahedraly sandwich between hydroxyl groups 6,20 . LDH has positively charge cations on the surface intercalated negatively charge anion and water molecules 21 . The general formula for LDH is [M II 1−x M III x (OH) 2 ] z+ (A n− ) z/n •y H 2 O M II and M III are di-and trivalent cation, (Zn +2 , Cu +2 , Al +3 and Fe +3 etc.) while A n− is the counter anion for cation present in the interlayer of brucite like sheets while x is the (M II /M II + M III ) ratio. A number of authors have been reported various metal-metal combinations for the synthesis of layered double hydroxide through various techniques for various application. For instance, Gaini et al. reported Mg-Al-CO 3 −2 for the de-coloration of indigo carmine 6 . Panda et al. reported Mg/Al-LDH and study its various factor on its growth such as various metal cation concentration, pH and aging time 22 . Similarly, Shao et al. investigated ZnTi-LDH by co-precipitation methods for the de-coloration of methylene blue dyes under the influence of visible light irradiation 23 . The ZnAl-LDH was synthesized by Beak et al. through hydrothermal methods 22 . Beside metal-metal combination the LDH was supported on various support to increase its catalytic performance. For instance, carbon nanotubes was used by Li et al. as a support for NiAl-LDH through solution method 24 . Similarly, reduced graphene oxide were used as a support for CoAl-LDH for the asymmetric electrochemical capacitors 25 . An efficient MgAl-LDH grown on multi-walled carbon nano tubes MWCNT for CO 2 adsorption 26 . Composite of NiCoAl-LDH coupled with Ni-Co-carbonate hydroxide supported on graphite paper were used for the asymmetric supercapacitors 27 . In the present work we synthesized Cd-Al and Cd-Sb grafted on activated carbon in order to increase the electronic conductivity through co-precipitation method in which Cd-Al/C were grown in layered double hydroxide morphologies. The synthesized materials were employed for the removal of five dyes AO, MG, CV, CR and MO through adsorption and adsorption assisted photo-degradation. ## Experimental Chemicals and reagents. Salt of Al and cadmium nitrate and chloride of cadmium and antimony along with other reagents and dyes mentioned in this manuscript were obtained from Sigma-Aldrich, Ireland. Millipore-Q machine was used for double distilled water, present in chemistry department, King Abdulaziz University Saudi Arabia. ## Synthesis of Cd-Al/C-LDH. Salt of Cd and Al nitrate were well mixed in double distilled water and then mixed with activated carbon through co-precipitation method 19,28,29 . Briefly salt of Al(NO 3 ) 3 and Cd(NO 3 ) 2 were dissolved thoroughly in double distilled water in 1:3 molar ratio. To this reaction mixture, 10 wt% of activated carbon was added and well dispersed by continuous stirring with the help of magnetic stirrer. To this mixture freshly prepared 0.1 M NaOH solution are added and continuously monitored till pH 9. After this the reaction mixture were placed on a hot plate for 6 h at 60 °C with homogenous stirring. After completion of the reaction the surplus solution is removed and the precipitate was washed thrice with C 2 H 5 OH:H 2 O mixture (8:2). The resultant product was dried in an oven for overnight at 50 °C and store in clean tube for further characterization. Instrumental analysis and Characterization. FTIR (Thermo Scientific) for functional group analysis and powder X-ray diffractometer (PXRD) with a Kα radiations (λ = 0.154 nm) source were used for the purity and crystallinity of the catalyst. Field emission-scanning electron microscope (FESEM), JEOL (JSM-7600F, Japan) for surface morphology and average size of the particles, while, energy dispersive X-rays spectrometry (EDS) of oxford-EDS system was employed for the elemental composition of the catalyst. X-ray photoelectron spectroscopy (XPS) Thermo Scientific K-Alpha KA1066 spectrometer (Germany) in the range of 0 to 1350 eV was investigated for the elemental analysis as well as for the determination of binding energy in the respective catalyst. The photocatalytic reaction was monitored through Evolution 300 UV-visible spectrophotometer (Thermo scientific). The effect of visible and ultraviolet light on adsorption-degradation process of dyes were observed under visible lamp (OSRAM, 400 watt) and ultraviolet lamp (Smiec Shanghai China, 230 V, 11 watt) respectively. The solar light effect was studied under normal day sunlight in open atmosphere and dark effect in complete absence of light. Photoluminescence emission spectra were confirmed at 320 nm excitation wavelength (fluorescence spectrofluorophotometer), RF-5301 PC, Shimadzu, Japan. The UV-vis diffuse reflectance spectroscopy was recorded by PerkinElmer UV-vis diffuse reflectance spectrophotometer. ## Synthesis of Procedure for dye removal. For this study 0.025% mmol effluent solution of acridine orange (AO), methyl orange (MO), congo red (CR), malachite green (MG) and crystal violet (CV) were prepared. The catalytic activity of Cd-Al/C-LDH and Cd-Sb/C were evaluated against the respective dyes under visible, solar, ultraviolet light, and dark condition. The adjusted dose 10 mg of the respective catalyst were added in 100 mL of 0.025 mmol concentration of dye solution and the gradual decrease in concentration of all the dyes were explored through UV-vis spectrophotometer. The % removal efficiency A.E. (%) of each catalyst was evaluated by using the following equation. C 0 represents the original concentration of each dye solution at time = 0, C t is the concentration of dye solution by adding the catalyst after some time = t as indicated in equation (1). Similarly, A 0 designated the absorbance of the original concentration of the dye solution at time = 0 and A t is the absorbance of dye solution during reaction progress after passing some time = t. Each time 3 mL of the aliquot were taken after specified time and checked the reaction progress by using UV-vis. spectrophotometer. ## Results and Discussion Structural characterization of catalyst. The average size and morphology was scrutinize through FESEM. The FESEM images shows the sheet morphology of Cd-Al/C-LDH (Fig. 1a,b) and Cd-Sb/C (Fig. 1c,d) and the sheets are composed of nanoparticles with average particle size of 50 and 40 nm respectively. In both catalyst small particles are aggregated to form the sheet morphology, however the particle is in photoluminescence spectra of both catalyst further support the efficiency of the respective catalyst in visible range as presented in Fig. 6. ## Photocatalytic activity. The photocatalytic activity of the Cd-Al/C-LDH and Cd-Sb/C was carried out for the de-coloration of five dyes AO, CR, MO, MG and CV under solar, visible, ultraviolet light and dark condition. Visible light exposure. Prior to the effect of solar, ultraviolet and dark condition the effect of visible light was study on the de-coloration of cationic and anionic dyes. Both Cd-Al/C-LDH and Cd-Sb/C were evaluated against three different dyes cationic acridine orange (AO) and anionic congo red (CR) and methyl orange (MO) under visible of 400 watt. Adjustment of catalytic dose. The catalytic dose was adjusted under visible light by selecting AO dye. Initially, the catalytic doses were adjusted with AO by starting from 100 mg of the respective catalyst in 100 mL of 0.025 mmol of AO. The Cd-Al/C-LDH de-colorize 70% AO while Cd-Sb/C 53% in 1 h. The amount of both the catalyst were decreased to 40 mg in 100 mL of 0.025 mmol of AO in which 63% of AO is removed with Cd-Al/C-LDH and approximately 45% with Cd-Sb/C in 1 h. However, the removal efficiency of dyes is further increased as we increased the contact time. For instance, after 2.5 h under the same condition the % removal of AO with Cd-Al/C-LDH and Cd-Sb/C was approximately 80 and 63% respectively as shown in Fig. 7a,b. Further the amount of the catalysts were decreased to 10 mg in 100 mL of AO. This time the removal efficiency was dropped to 43 and 36% respectively with Cd-Al/C-LDH and Cd-Sb/C in 1 h as shown in Fig. 8a,b. However, by increasing the exposure time of reaction mixture with light the rate of dye removal is also increases and vice versa. During the optimization of the catalyst to dye solution the Cd-Al/C-LDH showed superior activity than Cd-Sb/C (Fig. 9a,b). Using small amount of the catalyst to dye solution is eco-friendly and therefore, we selected 10 mg of the respective catalyst as an optimized amount for the further detail study. 10 mg of the respective catalyst were further used for the de-colorization of AO, CR and MO. After the first 15 min the % removal efficiency of Cd-Al/C-LDH against AO, CR and MO was 27.0, 13.7 and 4.8% respectively. However, at the same condition the Cd-Sb/C showed 30.7, 3.9 and 3.7% removal efficiency respectively. During the start of the reaction Cd-Sb/C showed slightly good response then Cd-Al/C-LDH. However, onward its activity was much lower than Cd-Al/C-LDH and this might be due to the LDH nature of Cd-Al/C-LDH. By increasing the contact time the % de-coloration of all dyes is increased and it was found that after 200 min, the Cd-Al/C-LDH showed strong response then Cd-Sb/C. For instance, after 200 min the % de-coloration of AO with Cd-Al/C-LDH is 69.4% as compared to Cd-Sb/C which was only 44.0%. It was found that both catalyst showed superior response against cationic dye (AO) as compared to anionic dye (CR and MO) and this is probably due to the large structure of these mentioned anionic dyes. It was also found that for all dyes Cd-Al/C-LDH showed superior performance than Cd-Sb/C. The reaction was also monitored without catalyst with almost no change in dye concentration under visible light, which shows that visible light itself has no role in the de-coloration phenomena. The decrease in the concentration and % removal efficiency of CR and MO are illustrated in Fig. 10a,b (CR), Fig. 11a,b (MO). ## Selectivity of Dye. Figure 12 showing the selective removal of AO under visible light exposure as compared to CR and MO. Therefore, AO was selected for the further detail study with both catalyst under solar, ultraviolet light and dark condition. Sunlight exposure. Under the same condition both the catalyst were evaluated for the de-coloration of AO in normal day sunlight exposure. The adjusted dose 10 mg of both the catalyst were added in 100 mL of 0.025 mmol of AO solution. The Cd-Al/C-LDH showed faster and better response then Cd-Sb/C in sunlight exposure. After the first 15 min of experiment the Cd-Al/C-LDH showed 13.5 while Cd-Sb/C showed 18.5% de-coloration of AO. Similarly, after 180 min the % de-coloration of AO with Cd-Al/C-LDH increased from 13.5 to 82.2% and Cd-Sb/C from 18.5 to 76.0%. This inferred the better performance of both catalyst with the passage of time as shown in the inset of Fig. 13a-d. Under visible and solar light irradiation it was concluded that cationic dye AO is selectively removed with the corresponding catalyst. It was also confirmed that cationic dye AO is predominantly removed in solar light as The catalytic activity of the respective catalysts were excellent, better and good for AO, MG and CV respectively. Cd-Al/C-LDH showed stronger catalytic activity over Cd-Sb/C catalyst and selectively de-colorized AO over MG and CV as presented in Fig. 16. Therefore, we select AO for further study under ultraviolet light and dark condition. Ultraviolet light exposure. UV lamp (230 volt, 11 watt) was used for the de-coloration of 0.025 mmol AO solution. Keeping the same condition as adjusted for visible and solar light, 10 mg of both the catalyst were used for the de-coloration of 100 mL of 0.025 mmol AO solution. After the first 15 min of exposure time the % removal efficiency of Cd-Al/C-LDH and Cd-Sb/C against AO was 8.5 and 7.5% respectively. However, after 200 min, the removal efficiency of AO (%) was increased to 36.3% with Cd-Al/C-LDH and 29.1% with Cd-Sb/C. Like visible and solar light the Cd-Al/C-LDH showing superior removal for AO as compared to Cd-Sb/C. The decrease in concentration and % removal efficiency of AO under ultraviolet light exposure are presented in the inset of Fig. 17a,b. ## Dark condition exposure. Prior to solar and ultraviolet light the effect of dark was studied for both catalyst to know the adsorption or degradation phenomena. The adjusted dose 10 mg of the respective catalyst were put in a beaker containing 100 mL of 0.025 mmol AO solution by providing complete dark condition. After the first 15 min of the reaction progress the Cd-Al/C-LDH displayed 11.0% and Cd-Sb/C 10.9% removal efficiency. However, after 200 min, the Cd-Al/C-LDH was showing 31.0% and Cd-Sb/C 26.0% removal of AO as indicated in Fig. 18a,b. After the detailed study for the de-coloration of dyes (AO, CR, MO, MG, CV) under solar, visible, ultraviolet light and absence of light, it was concluded that cationic dyes removed preferentially then anionic dyes. Among the cationic dyes (AO, MG and CV) AO was predominantly removed by both catalyst in solar light. However, the removal efficiency of MG and CV are also good. It was inferred that AO adsorbed in the absences of light and adsorbed plus degraded in ultraviolet, visible and solar light. The adsorption process is triggered by the presence of activated carbon in the respective catalyst. The dyes is adsorbed on the catalyst and then degraded as the reaction proceeded. Kinetic study of the reaction. The kinetics of the reaction was determined by applying pseudo first order kinetics (lnC t /C o ). This model was applied to compare the rate of Cd-Al/C-LDH and Cd-Sb/C in visible light by using different amount of the catalyst against the removal of AO. This equation also applied to investigate the rate of reaction in different parameters like dark, ultraviolet, visible and solar light. The rate of the reaction was determined by plotting lnC t /C o should be subscript verses time. The model showed the highest rate of Cd-Al/C-LDH then Cd-Sb/C in all conditions. The rate of reaction is directly related to the amount of catalyst. For instance, at 40 mg of Cd-Al/C-LDH the rate of reaction was 5.88 × 10 −3 mol L −1 min −1 as compared to 4.92 × 10 −3 mol L −1 min −1 obtained when 10 mg of the same catalyst was used under solar light. The same trends was observed in visible light. However, the rate of degradation is slow from solar light as inspected in Fig. 19a. In all other parameters Cd-Al/C-LDH showed higher rate of reaction then Cd-Sb/C as indicated in the inset of Fig. 19b,c. Catalytic recyclability. The catalytic recyclability is critical issue while carrying catalysis. Most of the catalyst become de-activated after first or second cycle. During the recyclability of the catalyst 100 mg of the respective catalyst was added in 100 mL (0.025 mmol) of AO under visible light of 400 watt. After one hour, the reaction was stopped and 3 mL of the aliquot was taken through a clean syringe and examined in UV-vis. spectrophotometer. After this the catalyst was separated through filtration process and the filtrate was washed thrice with acetone. The recovered washed catalyst (obtained from first run) was used in the next cycle for 100 mL of 0.025 mmol AO solution without drying or heating. The reaction mixture was again placed for 1 h under visible light exposure. Similarly, after one hour the reaction was stopped for monitoring the decrease in concentration and catalytic activity through UV-vis spectrophotometer. The catalyst is likewise separated carefully and washed thrice with Structural feature of dyes and photocatalytic activity. The structural features of dyes play a significant role in dyes light. However, it is necessary to select specific materials according to the functional groups in dyes. Dyes degradation generally undergoes through breakage of various functional group i.e. cleavage of the aromatic, C-S bond breaking occurred between an aromatic ring and the sulphur of a sulphonyl group. Similarly, other functional group such as C-C, C-N and azo bond breakage will also happened during dyes degradation 31 . Both catalyst showed better performance for cationic dyes AO, MG and CV then anionic dyes CR and MO. Among the cationic dyes AO selectively removed as compared to MG and CV. The higher adsorption assisted photo-degradation of AO was probably due to the cleavage of C-C and C-N bond of aromatic ring and their non-bulky nature. While, lesser removal of CR and MO were due to the presence of azo group and large bulky groups which interfere with the charge transfer during de-coloration process. The LDH has layered double structure with upper cationic and inner anionic layers. The schematic phenomena of adsorption assisted photo-degradation of dyes is shown in Fig. 22. ## Conclusion The inexpensive and diverse morphology make layered double hydroxide a suitable candidate in the field of catalysis. Cd-Al/C-LDH and Cd-Sb/C were synthesized through co-precipitation method. Both the catalyst showed a narrow band gap which indicated its effectiveness in visible region. The PL also showed a peak at 623 nm for both catalyst which showed its efficiency in visible range. Both catalyst grown in nanosheet morphologies. The Cd-Al/C has LDH nature as confirmed from FTIR and XRD. The respective catalyst were used for the de-coloration and mineralization of organic dyes from the effluent under solar, ultraviolet, visible light and dark condition. The Cd-Al/C-LDH shown better catalytic activity in all conditions and this is probably due to its layered double hydroxide nature. The rate of reaction was determined from Langmuir isotherm indicating high rate of Cd-Al/C-LDH as compared to Cd-Sb/C. A predominate degradation and negligible adsorption was found in solar light and an equal percentage of degradation and adsorption were found in visible light. Similarly, adsorption was observed in dark condition and ultraviolet light. The Cd-Al/C-LDH and Cd-Sb/C showed good recyclability, durability and easy separation. These results showed the high activity and the ease in separation of the catalyst which are routinely encounter in nanocatalysis.

chemsum

{"title": "Layered double hydroxide of Cd-Al/C for the Mineralization and De-coloration of Dyes in Solar and Visible Light Exposure", "journal": "Scientific Reports - Nature"}

towards_a_computational_ecotoxicity_assay

## Abstract: Thousands of anthropogenic chemicals are released into the environment each year, posing potential hazards to human and environmental health. Toxic chemicals may cause a variety of adverse health effects, triggering immediate symptoms or delayed effects over longer periods of time. It is thus crucial to develop methods that can rapidly screen and predict the toxicity of chemicals, to limit the potential harmful impacts of chemical pollutants. Computational methods are being increasingly used in toxicity predictions. Here, the method of molecular docking is assessed for screening potential toxicity of a variety of xenobiotic compounds, including pesticides, pharmaceuticals, pollutants and toxins deriving from the chemical industry. The method predicts the binding energy of the pollutants to a set of carefully selected receptors, under the assumption that toxicity in many cases is related to interference with biochemical pathways. The strength of the applied method lies in its rapid generation of interaction maps between potential toxins and the targeted enzymes, which could quickly yield molecularlevel information and insight into potential perturbation pathways, aiding in the prioritisation 1 of chemicals for further tests. Two scoring functions are compared, Autodock Vina and the machine-learning scoring function RF-Score-VS. The results are promising, though hampered by the accuracy of the scoring functions. The strengths and weaknesses of the docking protocol are discussed, as well as future directions for improving the accuracy for the purpose of toxicity predictions. ## Introduction Environmental pollution and ecotoxic stress through habitat-destruction, further exacerbated by climate change, have led to the onset of the sixth great mass-extinction event. 1 Pollution affects all organisms in the environment including humans, through a range of mechanisms, from chronic toxicity via the respiratory function and the gastrointestinal channel to dermatological uptake. This may trigger the detoxification pathways of the body in an attempt to eliminate toxic compounds from the system. 2,3 At continuous low-dose exposure, pollutants may also incur long-termed effects. 4,5 There is, however, limited or no available toxicity data for tens of thousands of compounds to which organisms in the environment, including humans, are exposed, due to the high cost and laborious nature of traditional toxicity testing. 6,7 This underscores the need for approaches that can rapidly screen and predict the toxicity of chemicals and prevent them from being released in the environment. In 2007, the National Research Council, Committee on Toxicity Testing and Assessment of Environmental Agents (U.S.A.), proposed a new framework, emphasising the importance of integrating in vitro-based and computational methods for evaluating toxicity. 8 In response, a series of programs have been initiated in the last decade, aimed at identifying assays relevant to toxicity and screening the biological activity of large numbers of candidate pollutants in a cost-efficient and timely manner. In particular, the Toxicity Forecasting (ToxCast) project of the Environmental Protection Agency (U.S.A.) has been instrumental in setting the stage for in vitro high-throughput assays 9,10 In a significant effort, over 900 chemicals were employed in over 300 concentrationresponse assays to produce a large matrix with estimates of the chemical potency, in terms of the AC 50 parameter. 10 The launch of Phase III further broadened the chemical and assay space of ToxCast. Although the measured AC 50 values have no direct toxicological interpretation and estimated values may have large standard errors, 11 high-throughput screening (HTS) nevertheless offers an important benchmark and a means of comparing chemical activities. Other ongoing programs for predicting the hazardous character of pollutants rely on computational measurements and properties of the chemical structures under consideration. The most common approach is the Quantitative Structure Activity Relationship (QSAR) method, which is based on the chemical similarity principle. Although these algorithms present good predictions for various compounds, they give no information on the mechanism of action of pollutants, which often requires structural and molecular biology methods of study. Several factors affect the toxicity of a given chemical, including its dose and route, frequency and duration of exposure. Many modes of molecular toxicity may be generalised as the binding between a chemical and a biomolecular target. The ability of a chemical to interact with a protein has, for a long time, been modelled within the pharmaceutical industry using the molecular docking formalism. By yielding a prediction of the binding affinity of chemicals to targets, putatively related to toxicity, docking could provide molecular-level information and insight into relevant interactions. Since the targets culpable for the underlying adverse effects remain unknown for most environmental chemicals, docking could provide information on the most likely targets. This could, for example, aid in designing in vitro HTS bioassays. A multiple-target approach which considers all possible combinations of receptors and ligands could also yield a more complete picture of potential toxicity and the potential perturbation pathways. Furthermore, the docking output could also be combined with other computational approaches, for example to derive QSAR models to predict toxicity. The main result of this paper is a computational ecotoxicity assay (CETOXA), containing 65 prepared protein targets belonging to various protein families. The methods for developing this computational study have been investigated in a previous paper published by our group on an ecotoxicological case study on vitamin B deficiency in moose. 24 Here, we explore the feasibility of applying a popular classical scoring function, Autodock Vina, and one of the most promising of the newer generation of machine-learning scoring functions, the Random Forest-based scoring function RF-Score-VS, for toxicity screening. The output from docking is a matrix with docking scores, representing an estimate of the binding energies, for all possible combinations of receptors and ligands. Of the two scoring functions considered, Vina shows promising results as it on average predicts higher affinities for active chemicals than for inactive ones. Furthermore, we also show that compound similarity is a strong predictor of binding affinity and that similar chemicals tend to show highest affinity for the same binding site of a given protein. Ultimately, the results may be integrated with existing toxicity and bioassay data to compare chemical profiles, to extract binding patterns and associations that may be key steps in triggering an adverse biological effect. ## Methods A flow-chart for the entire method of deriving the computational ecotoxicity assay is given in Fig. 1. The different steps in the flowchart are detailed below. ## Receptors Based on the assays present in Phases I & II of the ToxCast program, 10 65 receptors were selected for which experimental structures are available in the Protein Data Bank (PDB). A complete list of the proteins can be found in Table S1 in the Supporting Information (SI), and the list includes kinases (22), phosphatases (5), proteases (6), G-protein-coupled receptors (9), nuclear receptors (11), and cytochromes P450 (4). These proteins represent a broad range of cellular functions that are critical for the survival and proliferation of cells. PDB structures were chosen based on the availability of high-quality crystal structures. Only deposits with a resolution better than 2.85 were included. The protein structure quality was assessed using MolProbity. 25 Structures with an overall MolProbity score (representing a log- weighted combination of the clashscore, rotamer, and Ramachandran evaluations) lower than the crystallographic resolution were kept. In many cases, better scores were found for the corresponding structure in the PDB-REDO databank, which contains re-refined and rebuilt PDB entries. Chains were removed as necessary, to ensure that each structure corresponds to the reported biologically-relevant assembly. Ligands, cofactors, and co-crystallised lysozyme molecules were also removed as needed (see Table S1). Zinc ions were kept, given their functional role in proteins, although they are not taken into account during docking. Common structural problems were corrected using Dock Prep 29 through the Chimera software. Missing non-terminal backbone residues were modelled using Modeller 30 and a local energy minimization was performed with the Amber ff14SB force field 31 through Chimera to relieve potential atomic clashes. ## Pollutants The ligands in the ToxCast chemical library can be broadly separated into four chemical use groups: phthalates and alternative plasticizers, pesticides, pharmaceuticals (both marketed and failed), and other consumer use chemicals (such as food additives, soaps, and shampoos), all together comprising a structurally diverse chemical space. Ligands in reference 10 are denoted by PubChem codes and were downloaded using a script. About 1/3 of the compounds were available only as 2D chemical coordinates and a semi-automated procedure to make 3D coordinates was performed, using the obgen tool in OpenBabel 32 followed by manual curation of all the structures. Compounds with unsupported atoms, including metals and, e.g., boron were removed, leaving 957 environmentally relevant compounds. Both ligands and receptors were treated by scripts from AutodockTools 33 to prepare them for docking. ## Identification of Binding Sites Blind docking was employed to explore the protein surface for prospective binding sites. 34,35 For each potential chemical-target complex, a blind docking calculation was performed using QVina-W (an extension of QVina2, see below, specifically designed for blind docking), 36 with 64 independent runs per docking. Nine results were stored per complex, producing 9 x 957 bound ligand poses per target. Common binding sites were identified by clustering the 9 x 957 centres of mass of the bound ligands using the OPTICS algorithm, 37 a density-based clustering algorithm similar to DBSCAN, available in the Python machine learning library scikit-learn. 38 To reduce errors associated with inappropriate choice of binding site, multiple binding sites were considered. For each receptor, the centres of mass of up to four of the most populated clusters were stored. In most cases, fewer than four clusters were found by the algorithm (see SI, Table S1). An example of the procedure is The receptors, with 1-4 binding sites each, define the computational ecotoxicity assay that can be used for in silico scoring of new and existing compounds. ## Molecular Docking AutoDock Vina 40 (or Vina) has been a popular choice for high-throughput screening, due to its efficiency and relatively high accuracy. Here, we use a revised version, Quick-Vina2 (or QVina2), which has improved on the local search algorithm, achieving significant speed-up in computation time without compromising accuracy. 41 ## Re-docking co-crystallised ligands In order to evaluate the docking protocol, preliminary docking simulations were performed in which each of the co-crystallised ligands was re-docked into the active site of its cognate protein target. The coordinates for the ligands were extracted from the PDB files, randomised using obconformer in OpenBabel, and prepared for docking using AutodockTools. The search box dimension was set as in ref. 40. Re-docking was performed with QVina2, with an exhaustiveness level of 8, producing up to nine possible poses for each run. The symmetry-corrected RMSD was calculated using DockRMSD. 42 ## Targeted Docking and Scoring Using the computational ecotoxicity assay, targeted docking of the pollutants was performed on the identified binding sites for each receptor. A total of 160,776 targeted dockings and scorings were performed, using the same settings as during re-docking. The lowest score, corresponding to the strongest binding, of the 1-4 binding sites for each complex was kept, producing a cross-docking matrix of binding free energy predictions for 62,205 complexes, for all possible combinations of receptors and ligands. ## Binding affinity prediction For our purposes, the ability to accurately predict the binding pose is not as important as the ability to accurately rank chemicals based on the binding affinity. Enhancing the accuracy of scoring functions for predicting binding affinities or biological activity remains a challenge. It has been noted that most classical scoring functions suffer from limitations and are unable to accurately predict biological activities. 43 In order to move beyond the limitations of docking codes solvent effects may have to be included, at a substantial additional computational cost. Among the classical scoring functions currently available, Vina has one of the best scoring powers. 47 In recent years, however, a new class of scoring functions have emerged that use a nonparametric machine learning approach instead of imposing a predetermined functional form. Such machine-learning (ML) scoring functions have been found to show improvements over classical scoring functions, in terms of ranking compounds by binding affinity. It should be noted that while there is at least one ML scoring function specifically designed to perform well at experimental pose prediction, 52 other ML scoring functions do not necessarily improve success rates. 48,53 For the application of binding affinity predictions, however, the Random Forest-based scoring function RF-Score developed by Ballester et al. 54 has shown promising results, outperforming classical scoring functions such as the one used by Vina. 55 In particular, RF-Score-VS was adapted to virtual screening by training also on negative data (i.e., known inactive ligands), although the improvement in performance appears to be less substantial when applied to new targets not included in the training set. 51 Having generated an ensemble of viable docking poses with Vina, the top scored poses were rescored using the standalone version of RF-Score-VS. The Vina predicted binding affinities will be reported in terms of ∆G bind and the result of RF-Score-VS rescoring in terms of pKd. ## Re-docking co-crystallised ligands In order to assess the quality and reliability of the docking protocol, the co-crystallised ligands were redocked and the resulting pose compared to the experimental one. Pose generation error is commonly assessed by measuring the root mean square deviation (RMSD) of the predicted pose from the experimental binding orientation. The cumulative frequency curve of the RMSD between the native and predicted conformation is shown in Fig. 3. With 2 as a commonly accepted RMSD cutoff value, the success rate (i.e., fraction of predicted poses with an RMSD ≤ 2.0 ) for the top pose of Vina is 50.0%. For comparison, a benchmark study by Wang et al. 47 found the average success rate for the top scored pose among academic docking programs to be 47.4%, with second best performance achieved by Vina (with a success rate of 49.0%). Our results are thus within the expected accuracy of the docking method. Considering instead the best of the nine output poses (i.e., the pose with the lowest RMSD), the success rate increases to 67.2 % (with 80% falling below 3.0 , see Fig. 3). In other words, Vina manages to find a pose close to the experimental binding orientation, but may fail to rank it as the best-scoring pose, consistent with previous studies. 39,47,56 It should be noted that a large pose generation error does not by default imply inaccurate binding affinity predictions. A study by Li et al. found that there is a low correlation between pose generation error and binding affinity prediction error. 50 For our purposes, the latter is the more relevant quantity. ## Rescoring with RF-Score-VS The nine output poses of Vina were rescored using RF-Score-VS, giving a new top pose. This leads to a drop in success rate to 30.1% (Fig. 3). Note that it is known that machine-learning scoring functions do not necessarily outperform classical scoring functions in regards to pose prediction. 48,53 Considering instead the reliability of the scoring functions in predicting binding affinities, a comparison of different docking programs using the PDBbind benchmark dataset 57,58 found that Vina had the best Pearson correlation coefficient and Spearman ranking coefficient between the docked scores and experimental binding affinities, with values of 0.564 and 0.580, respectively. 47 In comparison, using the DUD-E database, 59 Vina had an effective Pearson correlation of 0.18 while RF-Score-VS had a correlation of 0.56 when both training and test sets contained data from all targets, and a more modest correlation of 0.2 when the training and test data were created independently. 51 The performance of the scoring functions may also be judged by their ability to predict high scores for known binders. The crystal structure of the complexes were scored (i.e., without docking) using Vina and RF-Score-VS. For a direct comparison between the two functions, the Vina score was converted to pKd units by using pKd= − log(e)/RT • ∆G (where R is the gas constant and T = 298.15 K), 60 see Fig. 4. Values of pKd < 4 would suggest weakly bound compounds, while values above 10 would indicate tightly bound compounds. 61 As the activity cutoff, the value of pKd = 6 has previously been used to distinguish between active and inactive chemicals. 51 Vina predicts the majority (75%) of ligands to have a predicted binding affinity pKd > 5.2 (alt. < -7.1 kcal/mol), with a mean binding affinity of pKd = 6.2 (alt. -8.4 kcal/mol ). A number of the cocrystallised ligands are predicted to be weakly bound and some outliers even fall below pKd 4. Among the latter, relatively poor binding affinities were predicted for a ligand involving halogenbonding interactions (3W2S), which is not accounted for in most classical scoring functions, and three ligands involving metal-coordination (1HFC, 4G9L, and 4H3X). In comparison, none of the known binders are predicted to be weakly binding by RF-Score-VS. Yet, none are predicted to be particularly potent either, generally falling below pKd 6.6. Evaluating the linear and monotonic relationship of the two scoring functions, the Pearson's correlation coefficient and Spearman's ranking coefficient are 0.23 and 0.26, respectively. It can be noted that chemicals with high scores from RF-Score-VS also tend to have high Vina scores while no correlation is found for more weakly binding chemicals. ## Identification of active chemicals The pollutants from the ToxCast chemical library were docked to the 1-4 binding sites of each target, identified as shown in Fig. 2, and the Vina score for each complex was stored. The top poses were further rescored with RF-Score-VS. For toxicity screening, the goal is to minimise the number of false negatives, meaning that no harmful chemicals should be classified as inactive. This poses a particular challenge since environmental chemicals can also elicit adverse effects from weak interactions. To be able to make predictions on potential activity based solely on the scoring function, it would be necessary to identify a target-specific threshold score which could successfully delineate active chemicals from inactive ones while minimising the number of false negatives. Ideally, the active and inactive sets should be separated to such a degree to enable predictions by simply choosing an appropriate threshold value. Some promising results were found in one toxicity screening of environmentally pertinent chemicals, where two different molecular docking software (eHiTS and FRED) were compared and found to have the capacity to identify weakly active chemicals from inactive chemicals binding to rat estrogen receptors. 62 For the dataset investigated here, however, the two sets show significant overlap. As an example, Fig. 5 shows the sets of active and inactive chemicals binding to CYP 2C9, as determined by the recorded activities in the ToxCast HTS, as a function of the Vina docking score (Fig. 5 The application of molecular docking for toxicity screening is an intriguing method, however, the results indicate the limitations of using current scoring functions to predict chemical activity. Possible routes for improving the performance could be to investigate complexation in a more detailed fashion through, e.g., molecular dynamics methods. Although implicit-solvent models remain popular, there is quite some evidence that they do not give sufficiently accurate results for predicting binding free energies. Therefore, free energy calculations using explicit solvent will likely be required to get accurate estimates of binding energies. 67 In some cases, quantum-chemical estimates of binding-strength for toxin-receptor models 68,69 may yield additional insights as well. Finally, whereas the purpose of drug discovery is to increase the enrichment factor, focusing only on the best ranking compounds, toxicity screening using molecular docking should also consider weakly binding chemicals, as these may still elicit adverse biological effects. This requires additional considerations and places new requirements on scoring functions. 70 Although a critical score cannot be set, Vina on average predicts better scores for active chem-icals, illustrated as follows. The average binding affinity for active and inactive chemicals to each target which had at least 5 active chemicals was calculated and the set values subtracted, ∆G bind,act − ∆G bind,inact . The difference is plotted in Fig. 6 and can be seen to be mostly negative. These are encouraging results as Vina would appear to somewhat consistently predict stronger binding affinities for active chemicals than for inactive ones. A similar trend can be seen for RF-Score-VS, although showing a poorer performance, at best producing a difference of -0.27 kcal/mol. ## Chemical similarity of ligands It is generally assumed that structurally similar molecules exhibit similar biological activities. 71 This notion is, for example, used in drug discovery to search chemical libraries for matching compounds 72,73 and to generate information about common receptors. 74 Toxicological data gaps may also be filled by read-across, in which chemicals with known toxicity are used to predict the toxicity of untested chemicals, based on their chemical similarity. It should be noted, however, that structurally similar compounds may in some cases show large differences in potency, due to so-called activity cliffs. 78,79 In this context, reproducibility of docking results may be evaluated by testing whether the docking program can reproduce the same binding pattern for structurally similar chemicals (see SI for details). The Tanimoto coefficient for pairs of chemicals was computed and the generated similarity matrix hierarchically clustered, see Fig. 7. A value of zero implies no shared chemical fragments and a value of one represents identity. The dendrogram in Fig. 7 is labeled by the four chemical categories used in ToxCast. Some degree of clustering may be seen within the chemical class groups, notably among pesticides and pharmaceuticals. Overall, however, the chemical library is structurally diverse with only few chemicals classified as highly similar. Binding patterns were assessed for each pair of chemicals by computing the mean absolute deviation (MAD) in binding affinity to the various targets. A low MAD value would suggest similar binding behaviour to the different targets and vice versa. These were correlated to the Tanimoto coefficient by colour-encoding the clustered similarity matrix with the MAD values (Fig. 8). A qualitative comparison of the two matrices reveals that clusters along the diagonal in Fig. MAD 7, corresponding to blocks of chemicals with shared fragments, also interact with similar binding affinities with the different protein targets, as indicated by their lower MAD values in Fig. 8. Low MAD values are also observed for pairs of highly dissimilar chemicals, notably for a group of consumer use chemicals and pesticides. These would appear to be chemicals which bind weakly to all targets, typically with an average Vina score > -6.0 kcal/mol. A more rigorous approach was employed for a quantitative comparison, by looking at the spread in binding affinities among similar compounds. For each query compound in the pollutant library, a subset of N sim similar chemicals was stored, selected based on the corresponding Tan-imoto threshold (see SI). If N sim > 3, the standard deviation in binding affinity for this subset, std(∆G bind ), was calculated (Fig. 9). For comparison, a subset of least similar chemicals as well as randomly chosen chemicals was also stored. As reference, the target-based standard deviation of all chemicals and the overall standard deviation among all docked complexes (vertical line in Fig. 9) are shown. It can be seen that compound similarity is generally a strong predictor of binding affinity, with Vina predicting more tightly packed scores for the most similar compounds. Conversely, the chemicals least similar to the query molecule also have a wider spread in affinities compared to the query. It should be noted that the latter subset may still contain chemicals that are similar to one another. Having compared patterns in binding affinity, it is also of interest to investigate whether similar chemicals show highest affinity for the same binding site of a given protein. At the onset, each protein was assigned up to 4 docking sites. Using the same subsets as above, for each protein with at least 2 docking sites and each query molecule binding with the lowest Vina score to site A, the fraction pA = N sim,A /N sim of similar chemicals also binding best to A was calculated. The value was normalized by the number of hits on each binding site, i.e., pAtot = N tot,A /N tot , to remove any potential bias for a more promiscuous binding site. The histogram of pA/pAtot is shown in Fig. 10, where a value greater than 1 indicates that the compounds are more likely to bind to the same site. It can be seen that Vina predicts structurally similar chemicals to preferentially bind to the same site compared to randomly chosen chemicals, with a distribution skewed towards larger pA/pAtot values, whereas the least similar ones, which may be significantly different in terms of size and functional groups present, have a distribution distinctly skewed towards zero, i.e., dissimilar chemicals prefer to bind to an alternative site. The fact that binding is in general largely correlated with structural similarity motivates the use of chemical similarity in predicting toxicity, as mentioned earlier. Based on available toxicity data, this would allow us to identify protein-ligand associations and predict a set of chemical fragments likely to contribute to the toxicity of a given molecule, which could be highly relevant in drug design. An example of the implementation is shown in SI. ## Hierarchical clustering of binding affinities Hierarchical clustering was performed using the Seaborn package in Python, with Euclidean distance as the similarity metric and Ward variance minimization algorithm as the linkage method. Some prominent protein family clusters can be seen, particularly homogeneous clusters of NR, Kinase, and GPCR. Among the various protein families, CYP and GPCR are predicted to be the most affected, with over half the chemicals binding with an affinity < -8 kcal/mol to CYP 1A1, CYP 1A2, and the muscarinic acetylcholine receptors. Comparing chemicals in the different chemical use categories, consumer use chemicals and phthalates and alternative plasticisers tend to have low predicted binding affinities while pesticides and pharmaceuticals show a broader distribution of Vina scores. This can be seen more clearly in Fig. 12, which shows the binding free energy, averaged over all receptors, for pollutants within each chemical use group. A wide band of pesticides, pharmaceuticals, and consumer use products in Fig. 11, corresponding to 10% of all chemicals considered, have low binding affinities (∆G bind > -6 kcal/mol) to all protein targets. Among the chemicals that show specificity by binding to only a few targets with higher affinities are Mirex, a persistent organic pollutant, binding to the farnesoid X receptor α (∆G bind = -8.4 kcal/mol) and PI3K α (∆G bind = -7.7 kcal/mol); Pentachlorophenol, a pesticide and disinfectant, binding to CYP 1A1 and 1A2 (∆G bind = -7.6 and -7.7 kcal/mol, respectively); Lindane, an insecticide, pediculicide, and scabicide, binding to CYP 1A2 (∆G bind = -7.5 kcal/mol). Interestingly, Mirex has been found to inhibit particularly Adenosine triphosphatase, a protein which is central for liver function. 80 This enzyme shares structural motifs with Phosphoinositide 3-kinase, which binds Mirex as calculated by CETOXA with a free energy of -7.7 kcal/mol. The common features between these two proteins suggest that the docking procedure detects structural similarities of particular enzyme motifs, even though these two proteins pertain quite different 3D structures. The second important finding is the docking result associated with the Farnesoid receptor. Indeed, this receptor is critical for liver function and has a central role in the detoxification mechanism in hepatocytes. This protein is also related to carcinogenesis and its inhibition causes considerable liver and gallbladder complications. The result that Mirex binds to the Farnesoid receptor is thus relevant for the biochemistry of this receptor, as it is for the case of Adenosine triphosphatase. Another interesting finding is that pentachlorophenol binds strongly to CYP1A1 and -1A2. Both these enzymes use O-deethylation to oxidize phenolic substances and both are expressed in the liver and perform detoxification in the liver microsomes of pyrimidinelike and phenolic xenobiotics such as caffeine, alpha-napthoflavone and 7-ethoxyocourmarine and phenacetin. 83 The results also indicate that lindane may be destined for detoxification by CYP1A2. Interestingly, lindane is a hexa-chlorinated compound that binds particularly well to CYP1A2, as CYP1A2 has a 30% higher affinity towards halogenated compounds compared to CYP1A1. This indicates that the docking procedure in this case has the potential to distinguish enzymes with very similar specificity with however a 30% preference difference towards halogenated aromatics. 83 Among the plasticizers, one phthalate (butyl benzyl phthalate) and three alternative plasticizers (dipropylene glycol dibenzoate, pentane-1,5-diyl dibenzoate, and hexane-1,6-diyl dibenzoate) showed promiscuous behaviour, targeting over 1/3 of the receptors with ∆G bind < -8 kcal/mol. Among the consumer use chemicals identified as potentially highly promiscuous are the surfactants perfluorodecanoic acid and perfluoroundecanoic acid, as well as the widely used synthetic food dyes allura red and FD&C Blue no. 1. Among the pesticides targetting multiple targets with high binding affinities are Famoxadone, Novaluron, and Prosulfon. The interactions of Famoxadone with multiple protein targets has been confirmed in the literature. In a study by , 84 Famoxadone was found to inhibit both mitochondrial enzymes as well as the cytochrome system (particularly CYPBc1 85 ). The heatmap notably contains a cluster of about 28 pharmaceuticals with high Vina scores, binding to multiple targets. These include failed or terminated drugs, such as the thiazolidinedione based and non-thiazolidinedione based antihyperglycemic agents Troglitazone and Farglitazar, which are specific ligands for peroxisome proliferator-activated receptors (PPAR). Troglitazone has been found to have hepatotoxic effects, 86 while Farglitazar did not reach past phase III clinical trials. Zamifenacin in this list was also identified to have highly promiscuous biological activity in the ToxCast HTS. 10 Although drugs are typically designed to interact with a specific target, unintended drug-target interactions are one of the major challenges in drug design, associated with side-effects and high failure rates. 87,88 Assessment of Vina high scores The highest binding affinity per chemical in the heat map was extracted and the top 20 chemicaltarget pairs are shown in Table 1. the triple arrangement of their aromatic rings. They expose their lateral electrons for oxidation by detoxification enzymes and are then converted to their diol-epoxide forms which react with DNA. 92,93 These compounds are ranked as a top compound to target CYP1A2 (Table 1), which reflects well with empirical studies 94 showing a high preference of CYP1A2 towards ethoxyresorufin. Ethoxyresorufin is geometrically similar to these compounds, suggesting the docking system recognizes the conserved molecular volume and geometry. There are no studies on fluroanthene and benzo [b]fluoranthene which could confirm their interaction with CYP1A2, however, by their elongated geometries, it is highly feasible that CYP1A2 is the detoxification enzyme for these two compounds, as both are known to be genotoxic. 95,96 A further study showed that CYP1A2 is activated by benzo[b]fluroanthene, however the expression of CYP1A2 was not confirmed. 96 In the prediction, the docking procedure indicated that CYP1A2 was the preferred target for benzo[b]fluroanthene, which indicates that physiological responses may be different than cellular responses when considered on a cohort of individuals exposed to the compound via airparticulate matter. The potentially genotoxic and carcinogenic 97 food dye C.I. Solvent yellow 14 is predicted to bind strongly to both P450 1A1 and 1A2, with an affinity of -11.7 kcal/mol and -12.2 kcal/mol, respectively. Although 1A1 and 1A2 are highly homologous, CYP1A1 is considered to be most efficient in metabolizing C.I. Solvent yellow 14, whereas other CYPs, including CYP1A2, were previously found to be almost ineffective. 98 The discontinued PPAR agonist Farglitazar is predicted to bind strongly to the beta-2 adrenergic receptor, although it was developed to treat type 2 diabetes. 99 It has a predicted binding affinity to PPAR α and PPAR γ of -9.4 kcal/mol and -11.3 kcal/mol, respectively. In other words, many of the Vina high scores can be correlated with known toxic compounds. ## Assessment of known interactions Based on literature data and as considered in the ToxCast publications, 9,10 we assessed whether Vina succeeds to predict high scores for chemicals that are known to interact with certain proteins. Some of these known complexes and their predicted binding affinities are listed in Table 2. Bisphenol A 100 and 2,2-bis(4-hydroxyphenyl)-1,1,1-trichloro ethane, 101 a metabolite of the pesticide methoxychlor, are known estrogen receptor agonists. Perfluorooctanoic acid (PFOA) and perfluorooctane sulfonic acid (PFOS) are chemicals with widespread use and with multiple reported toxicities, generally thought to be triggered by activating PPAR α. 102,103 In the HTS screening in ref. 9, both were active for PPAR γ, but only PFOA was active for PPAR α. In our case, the predicted binding affinity is relatively high also for PPAR γ. It should be noted that one in vivo study found that neither compound activates PPAR γ, 104 illustrating potential risks in extrapolat-ing in vitro or in silico results to in vivo. The pesticide lactofen 105 is also an expected PPAR activator found to be positive for PPAR assays in ToxCast. 9 The pharmaceuticals CP-471358 and CP-544439 are known inhibitors of matrix metalloproteinases MMP2, MMP9, and MMP13, 106,107 and these complexes all have high Vina scores. However, relatively low predicted binding affinities are found for the complexes involving the androgen receptor (AR). The pesticides linuron, prochloraz, and vinclozolin are known to cause toxicity by interacting with AR as antagonists. 108 Binding to the ligand binding domain of nuclear receptors is known to cause structural and dynamic changes in the protein. 109 For example, one study used computer simulations to follow the long-time scale conformational fluctuations of AR interacting with different ligands and found that agonists and antagonists induce distinct conformational changes. 110 Since the androgen receptor used for docking here corresponds to an agonist complex, 111 with an associated redocking score of -8.9 kcal/mol, the low binding affinities of the (antagonist) pesticides may be due to the shortcomings of neglecting protein flexibility. This issue can be addressed in the future by expanding the computational ecotoxicity assay to include multiple protein structures for the same receptor, with both agonist and antagonist co-crystallised conformations to account for multiple binding modes. The HTS screening in ref. 10 in some cases missed known active compounds. Zamifenacin is a muscarinic antagonist 112 that did not show activity for the M1 assay. 5,5-diphenylhydantoin is a known substrate and inhibitor for many CYPs, 113 but was not found to inhibit any of the CYPs in the HTS screening. These chemicals show high predicted binding affinities to their respective targets. Additional chemical-target interactions that were considered, including estrogenic and nonestrogenic compounds within the EPA's Endocrine Disruptor Program, can be found in SI Tables S2 and S3. ## Assessment of Vina low scores Considering potentially missed interactions and weakly binding active chemicals, the 20 least promiscuous chemicals in the docking simulations were checked for measured activity in the HTS assays. Of these 1300 complexes with low predicted binding affinities, only 8 were reported as active in ToxCast. These assays include a number of nuclear receptors which, as noted above, may undergo conformational changes upon ligand binding. In one case, sodium nitrite in complex with monoamine oxidase A (MAO), the low predicted binding affinity may be due to the mechanism of inhibition, as nitrates would appear to inhibit MAO through oxidation of the SH-groups, 114 a mechanism not captured in current docking methods. ## Potential refinement of the protocol In this work, we assessed the strengths and weaknesses in employing molecular docking for screening potential toxicity of xenobiotic compounds. This approach allows the generation of interaction maps between potential toxins and targets linked to perturbation pathways in a a cost-efficient and timely manner, yielding molecular-level information. While neither the classical Vina scoring function nor the machine-learning scoring function RF-Score-VS considered here were capable of distinguishing active chemicals, Vina on average predicted higher affinities for active chemicals and succeeded in identifying several interactions that could be confirmed in the literature. To be able to achieve higher docking accuracy with fewer false negatives, the method can be further developed by considering, for example, protein side-chain conformational changes, covalent interactions, charge redistribution, and bio-transformation products. A further refinement could be to consider multiple structures per target to take into account multiple binding modes. More exact methods may be required to distinguish binders from non-binders and the obvious challenge there remains to balance computational cost and accuracy. Finally, current scoring functions are limited in their ability to predict biological activity, underscoring the need to consider other properties

chemsum

{"title": "Towards a Computational Ecotoxicity Assay", "journal": "ChemRxiv"}

facile_synthesis_of_aiegens_with_wide_color_tunability_for_cellular_imaging_and_therapy

## Abstract: Luminogens with aggregation-induced emission (AIE) characteristics are nowadays undergoing explosive development in the fields of imaging, process visualization, diagnosis and therapy. However, exploration of an AIE luminogen (AIEgen) system allowing for extremely wide color tunability remains challenging. In this contribution, the facile synthesis of triphenylamine (TPA)-thiophene building block-based AIEgens having tunable maximum emission wavelengths covering violet, blue, green, yellow, orange, red, deep red and NIR regions is reported. The obtained AIEgens can be utilized as extraordinary fluorescent probes for lipid droplet (LD)-specific cell imaging and cell fusion assessment, showing excellent image contrast to the cell background and high photostability, as well as satisfactory visualization outcomes.Interestingly, quantitative evaluation of the phototherapy effect demonstrates that one of these presented AIEgens, namely TTNIR, performs well as a photosensitizer for photodynamic ablation of cancer cells upon white light irradiation. This study thus provides useful insights into rational design of fluorescence systems for widely tuning emission colors with high brightness, and remarkably extends the applications of AIEgens. ## Introduction The exploration of fluorescent materials and technologies has opened new avenues to scientifc advancement, societal development and public health, 1 which is exemplifed by the Nobel Prize successively awarded to fluorescence-related research. As one of the most important branches of fluorescent materials, fluorescent bio-materials that offer researchers a powerful platform for analytical sensing and optical imaging have been proven to be extremely useful for biological visualization, clinical diagnosis and disease treatment by virtue of their noninvasiveness, in situ workability, excellent accuracy, superb sensitivity and simple operation. 2 Although many types of fluorophores have been commercialized for biological applications, the current situation is still far from ideal, mainly due to some limitations: (1) inherent fluorescence quenching upon aggregate formation due to intermolecular p-p stacking and other nonradiative pathways, which is notoriously known as aggregation-caused quenching (ACQ); 3 (2) the difficulty of widely tuning emission colors by simple modifcation of molecular structures; and (3) complicated and laborious syntheses of fluorophores. 4 As an anti-ACQ phenomenon, aggregation-induced emission (AIE) was coined in 2001 by Tang's group, 5 which refers to a unique phenomenon that a novel class of fluorophores are non-emissive or weakly emissive in the molecularly dissolved state but they emit intensively in aggregated states owing to the restriction of the intramolecular motions (RIMs). 6 Remarkably, the AIE principle has triggered state-of-the-art developments in an array of biological felds, ranging from bioimaging, biosensing, stimuli-responsive systems, and therapeutics to theranostics, mainly resulting from various impressive advantages of AIE luminogens (AIEgens), such as high photobleaching threshold, high signal-to-noise ratio for imaging, excellent tolerance for any concentrations, large Stokes shift, turn-on feature for detecting analytes, and efficient photosensitizing ability. 7 Although numerous AIEgens have been constructed on the basis of different structural motifs including tetraphenylethene, 8 hexaphenylsilole, 9 tetraphenylpyrazine 10 and distyrylanthracene, 11 to the best of our knowledge, there has been no single AIE system which allows arbitrarily tuning emissions ranging from each color of visible light to the nearinfrared (NIR) region. Considering the great signifcance of tunable fluorescent systems in the applications of multi-target sensing, optoelectronic devices and full-color bio-imaging, 12 the development of an AIE system exhibiting wide color tunability is highly desired and remains a challenging task. Compared with inorganic complexes and quantum dots, organic fluorophores are advantageous for bio-imaging, diagnosis and therapy, benefting from their good bio-compatibility, tunable molecular structures and chemical compositions, and scalable synthesis. 13 Evidently, the exploration of an organic fluorophore system with both the AIE attribute and emission color tunability across a wide wavelength range would captivate much interest. Herein, we report for the frst time the design and synthesis of a series of AIEgens having widely tunable emissions covering violet, blue, green, yellow, orange, red, deep red and NIR regions (Fig. 1). Each AIEgen comprising the triphenylamine (TPA)-thiophene building block is facilely obtained through oneor two-step reaction, and the emission colors are tuned by simple alteration of the HOMO-LUMO energy level by the introduction of electron donor (D)-acceptor (A) substituents. 14 Moreover, these AIEgens can be successfully utilized as extraordinary lipid droplet (LD)-specifc bioprobes in cell imaging, determination of cell fusion, and photodynamic cancer cell ablation. ## Synthesis and single crystal analysis As depicted in Scheme 1, the desired compounds were facilely prepared through one or two steps. TTV was synthesized through the Suzuki-Miyaura coupling reaction of 4-bromo-N,N-diphenylaniline with thiophen-2-ylboronic acid in the presence of the palladium catalyst using mixed THF/H 2 O as the solvent at 75 C. The same synthetic procedure was successfully conducted by employing substituted 4-bromo-N,N-diphenylaniline and modi-fed thiophen-2-ylboronic acid as starting materials, producing compounds TTG, TTY, TTO and TTR. The reactions between TTG/TTO and malononitrile proceeded smoothly, giving the corresponding products TTDR and TTNIR with moderate yields. In addition, TTB was obtained by the Suzuki-Miyaura coupling reaction of (4-(1,2,2-triphenylvinyl)phenyl)boronic acid with intermediate product 1, which was isolated through the Suzuki-Miyaura coupling reaction between (4-(diphenylamino)phenyl) boronic acid and 2,5-dibromothiophene. All compounds are composed of sufficient moieties that can freely rotate in the single-molecule state leading to energy consumption of the excited state through non-radiative pathways, thus ensuring that these compounds are weakly emissive in solution. Aiming to further study and deciphering their optical properties in the aggregation state, single crystals of TTG, TTY and TTDR were grown in DCM-MeOH mixtures by slow solvent evaporation. As illustrated in Fig. 2, S1 and S2, † the twisted conformation of the TPA segment extends the intermolecular distance (>3.2 ) between two parallel planes, remarkably reducing or avoiding the intermolecular p-p interactions, and essentially preventing emission quenching in its aggregation state. Moreover, the molecular conformation can be strongly rigidifed by abundant intermolecular interactions (such as C-H/O, C-H/C, and S/C) resulting in the restriction of molecular motions, which is benefcial for enhancing the solid state emission efficiency. On the basis of the above-mentioned XRD results, it is believed that these synthesized compounds are potentially AIE-active. ## Photophysical properties The UV-vis absorption spectra of TTV, TTB, TTG, TTY, TTO, TTR, TTDR and TTNIR were measured in acetonitrile (ACN). As shown in Fig. 3A and Table S1, † the solution of building block TTV displays a maximum absorption band at 348 nm, and the maximum absorption peaks of these modifed compounds are located ranging from 383 nm to 512 nm. The gradually red-shifted absorption wavelengths can be attributed to the orderly enhanced D-A effect from TTV to TTNIR. To investigate their AIE features, an ACN/H 2 O mixture with different H 2 O fractions was utilized as the solvent system. It was observed that compounds TTB, TTG, TTY, TTO, TTR, TTDR and TTNIR exhibit typical AIE features (Fig. 3C). Taking TTY as an example, there is almost no fluorescence emission when the H 2 O fraction is below 60%. Afterwards, the PL intensity increases dramatically along with raising the fraction of water because of activation of RIM by molecular aggregation and reaches its maximum at 90% water fraction, which is 185-fold higher than that in ACN solution (Fig. 3B). Although the fluorescence intensity of TTV is inversely proportional to the water fraction, its quantum yield in the solid state (27.5%) is higher than that in the solution state (18.6%), defnitely demonstrating an aggregation-induced emission enhancement (AIEE) attribute. The gradually decreased fluorescence intensity of TTV along with the increased water fraction could be attributed to its twisted intramolecular charge transfer (TICT) feature, 15 which was determined by both the red-shifted emission wavelength and the declined emission efficiency accompanying the raised solvent polarity (Fig. S3 †). As one of the nonradiative pathways for the excited state to relax and 2019, 10, 3494-3501 | 3495 deactivate, the TICT effect is competitive with AIE properties in determining the PL intensity and efficiency using the ACN/ H 2 O solution system. In the case of TTV, the AIE feature is strongly depressed by the TICT effect in the nanoaggregation state. As illustrated in Fig. 3D and E and Table S1, † these TPAthiophene building block-based AIEgens emit efficiently in both nanoaggregation and solid states exhibiting relatively high quantum yields ranging from 3.11% to 40.79%. Each maximum emission wavelength accurately peaks in violet (402 nm), blue (482 nm), green (531 nm), yellow (580 nm), orange (612 nm), red (649 nm), deep red (667 nm) and NIR (724 nm) regions, respectively, suggesting the extremely wide emission color tunability, which is ascribed to both of their varied pconjugation and D-A effect. Additionally, the fluorescence decay curves in the solid state show that their lifetimes range from 0.64 to 3.69 ns (Fig. 3F and Table S1 †). ## Theoretical calculations To better understand the optical properties of these AIEgens, density functional theory (DFT) calculations were carried out at the B3LYP/6-31+G(d) level with molecular geometries optimized at the TD-B3LYP/6-31+G(d) level (Fig. 4). It was observed that, from TTV to TTNIR, the calculated HOMO-LUMO energy gaps generally decrease, and the results are in good accordance with the experimental data of emission maximums. The orderly declined energy gaps are realized through ingenious modifcation of the TPA-thiophene building block with diverse electron-donating (thienyl or methoxyl groups) and electron-accepting (aldehyde or cyano groups) units or the p-bridge. Except for TTB, the HOMOs of the remaining AIEgens are delocalized at the TPA moiety, whereas their LUMOs are distributed on the other side of the structures, demonstrating typical D-A structural features. It has been demonstrated that the separation of HOMO and LUMO distributions is essential to effectively reduce the singlet-triplet energy gap, which facilitates the generation of reactive oxygen species (ROS), 16 further endowing these AIEgens with prominent potential for photodynamic therapy (PDT) applications. 17 In contrast, TTB possesses an evenly distributed HOMO and LUMO, resulting from its both imperceptible D-A effect and long p-conjugation bridges. ## Bio-imaging, visualization of cell fusion and photodynamic therapy In the preliminary bioimaging experiment, the cell imaging study was conducted by using HeLa cells as a cell model. Cells were incubated with 1 mM of TTNIR for 20 min; as illustrated in Fig. 5B, bright fluorescence within cells can be observed showing excellent image contrast to the cell background. The co-localization study further proceeded upon incubating HeLa cells with TTNIR and BODIPY493/503 Green. The latter dye is a commercially available bioprobe for the LDs, which are ubiquitous lipid-rich spherical organelles and actively involved in various biofunctions, such as signal transduction, lipid , 2019, 10, 3494-3501 This journal is © The Royal Society of Chemistry 2019 metabolism, and protein degradation. The perfect overlap between TTNIR and BODIPY493/503 Green in cell imaging output indicates the excellent LD-specifc targeting capability of TTNIR (Fig. 5B-D). Photostability is a key criterion for evaluating the overall stability of photosensitive substances. The continuous scanning method was then utilized to quantitatively study and compare the photostability of TTNIR and BODIPY493/503 Green. As shown in Fig. 5E-I, after 15 minutes of laser irradiation, the fluorescence intensity of BODIPY493/ 503 Green encounters an obvious decline, whereas TTNIR shows negligible photobleaching. Moreover, the photostability assessment was also conducted towards Nile Red, which is another commercially available dye for LD-staining (Fig. S13 †). It was observed that Nile Red suffered an obvious signal loss with remaining signal intensity around 60% after 15 minutes of laser irradiation, strongly suggesting that the photostability of TTNIR is superior to that of commercially available bioprobes. To further prove its applicability, this staining and imaging strategy using TTNIR is exploited for other cell lines, including NCM460, DLD1, SW480, SW620 and COS-7 (Fig. S4 †). In each case after incubation with TTNIR for 20 min, it shows strong and specifc internalization into the LDs. Moreover, other AIEgens including TTV, TTB, TTG, TTY, TTO, TTR and TTDR were also investigated for cell imaging. It was observed that LDs can be clearly visualized with excellent image contrast to the cell background through respective incubation of cells with these presented AIEgens (Fig. 6). Pearson's correlation coefficients between AIEgens and commercially available LD-bioprobes were determined to be 90-95%, solidly demonstrating the high specifcity of these AIEgens for staining LDs (Fig. S5-S12 †). Their excellent LD-staining specifcity reasonably results from the lipophilic properties, which bring about efficient accumulation of them in the hydrophobic spherical LDs due to the "like-like" interactions. Evidently, these AIEgens possess various impressive features, such as high brightness, excellent targeting specifcities to LDs, extraordinary photostabilities and widely tunable emission colors, making them remarkably important in visualization of biological structures and processes. As a common phenomenon in nature, cell fusion is highly associated with many cellular processes, including fertilization, development of placental, regeneration of skeletal muscle, oncogenesis, aneuploidy, chromosomal instability and DNA damage. 18,19 In addition, a recent study shows that cell fusion could play a vital role in alternative therapies for restoring organ function through repairing cellular dysfunction. 19 Therefore, the development of effective methods for visualizing cell fusion is of great importance. Encouraged by the excellent cell imaging results and homology of the presented AIEgens, a straightforward method for visualization of the cell fusion outcome was conducted by using the combination of TTG and TTNIR as cell imaging agents, due to their minimal overlap of the emission range. In this experiment, two sets of cells were respectively stained with TTG and TTNIR, which were then mingled and treated with polyethylene glycol (PEG) to induce cell fusion. 20 As illustrated in Fig. 7, after treatment with PEG, both green and red fluorescence of lipid droplets were observed within one single cell, suggesting that cell fusion between TTG-and TTNIRstaining cells successfully proceeded. In addition, the cell fusion outcome was also solidly verifed through a commercially available nuclei-staining agent Hoechst 33258. The appearance of two stained nuclei within one single cell (Fig. 7D) indicated that the visualization strategy of the cell fusion outcome by using two AIEgens with different emission ranges is defnitely reliable. Evidently, the developed AIEgens having widely tunable emissions and high emission efficiencies are potentially useful in the fundamental study of cell fusion. Intense fluorescence in the near-infrared (NIR) region is highly desirable for many clinical processes, due to the salient advantages of deep tissue penetration, minimal photodamage to biological structures, and high image contrast to the physiological background. 21 Moreover, NIR emission is generally realized by intensifying the D-A effect of the structure, resulting For TTG, l ex : 405 nm (1% laser power), l em : 425-540 nm. For TTNIR, l ex : 560 nm (6.5% laser power), l em : 600-740 nm. For Hoechst 33258, l ex : 405 nm (3.5% laser power), l em : 425-540 nm. Concentrations: TTG (500 nM), TTNIR (2 mM), Hoechst 33258 (2.5 mM). Scale bar ¼ 20 mm. in the separation of HOMO and LUMO distribution, as well as the decrease of the singlet-triplet energy gap, thus facilitating the generation efficiency of ROS. Therefore, the AIEgen TTNIR with both bright NIR emission and the strong D-A effect is potentially efficient for PDT, which is an extraordinary therapeutic modality, and has captivated much interest for treating various malignant and non-malignant diseases with minimal invasiveness and precise controllability. In the preliminary test, the ROS generation efficiency of TTNIR was investigated using H2DCF-DA as the indicator, which can emit fluorescence at around 534 nm triggered by ROS. As shown in Fig. 8A, in the presence of TTNIR, the emission of H2DCF-DA was rapidly intensifed with the increase of irradiation time using white light as the irradiation source, reaching 36-fold enhancement in 6 min compared with the original emission intensity. In contrast, the fluorescence intensities of AIEgens or H2DCF-DA alone were very low and remained almost constant under the same irradiation conditions. These results reveal good photo-sensitizing properties for ROS generation. Quantitative evaluation of the phototherapy effect of TTNIR on HeLa cells was then explored through the standard 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. The dose-dependent toxicity study shows that there is no obvious cytotoxicity observed for the HeLa cells treated with TTNIR in the dark, even with the TTNIR concentration reaching as high as 20 mM (Fig. 8B). Upon white light exposure, cell viability dropped gradually with raising the concentration of TTNIR. Only 7% of cell viability remained with utilizing 20 mM of TTNIR, demonstrating almost complete cell apoptosis. Apparently, TTNIR holds high effectiveness for cancer cell ablation by means of PDT. ## Conclusions To sum up, we report the frst series of AIEgens with widely tunable emissions covering the whole visible region extending to the NIR area. These TPA-thiophene building block-based AIEgens can be facilely prepared by extremely simple synthetic protocols, and show high fluorescence quantum yields up to 40.79% in the solid state benefting from their intrinsic aggregation-induced emission nature. They have been successfully utilized for LD-specifc cell imaging, showing excellent image contrast to the cell background and higher photostability than the commercial LD-staining fluorophore. Additionally, the high brightness and homology of these AIEgens endow them with excellent performance for visualizing cell fusion. To the best of our knowledge, this would be the frst report on using AIEgens as fluorescent probes for assessing cell fusion. Notably, upon exposure to white light irradiation, one of these presented AIEgens, namely TTNIR, displays high ROS generation efficiency, enabling its effective application for photodynamic ablation of cancer cells. Our fndings in this study provide an ideal fluorescence system for widely tuning emission colors with high brightness at will. This successful example would further facilitate the exploration of organic fluorophores with AIE features for preclinical research and clinical applications. ## Materials and methods Chemicals for synthesis were purchased from Sigma-Aldrich, MERYER or J&K, and used without further purifcation. All solvents were purifed and dried following standard procedures. 1 H spectra were measured on Bruker ARX 400 NMR spectrometers using CD 2 Cl 2 or CDCl 3 as the deuterated solvent. Mass spectrometric measurements (HRMS) were performed on a Finnigan MAT TSQ 7000 mass spectrometer system operating in matrix-assisted laser desorption/ ionization time of flight mass spectrometry (MALDI-TOF) mode. UV-vis spectra were measured on a Milton Ray Spectronic 3000 array spectrophotometer. Steady-state photoluminescence (PL) spectra were recorded on a PerkinElmer LS 55 spectrophotometer. Fluorescence images of AIEgens in the solid state and aggregation state were collected on an Olympus BX 41 fluorescence microscope. The cellular fluorescence images were taken using a Zeiss laser scanning confocal microscope (LSM7 DUO) and analyzed using ZEN 2009 software (Carl Zeiss). ## Synthesis of compound TTV 7c,22 A mixture of the bromide substituted triphenylamine moiety (1.2 mmol), thiophen-2-ylboronic acid moiety (1 mmol), THF (20 mL), K 2 CO 3 aqueous solution (2 M, 1.6 mL), and Pd(PPh 3 ) 4 (58 mg, 0.05 mmol) was degassed and charged with N 2 . The mixture was refluxed overnight. The reaction was quenched by the addition of water (30 mL) and extracted with CH 2 Cl 2 (3 30 mL). The combined organic layer was dried over anhydrous Na 2 SO 4 and evaporated. The residue was purifed by column chromatography over silica gel using petroleum ether to afford the desired product TTV with a yield of 78%. 1 H NMR (400 MHz, CD 2 Cl 2 ): 7.60 (d, J ¼ 6.8 Hz, 2H), 7.41 (d, J ¼ 8 Hz, 2H), 7.37-7.33 (m, 4H), 7.13-7.06 (m, 9H). 13 C NMR (100 MHz, CDCl 3 ): 147. 49, 147.20, 144.26, 129.27, 128.54, 127.95, 126.71, 124.42, 123.98, 123.75, 123.02, 122.21 ## Cell imaging and confocal co-localization In 35 mm glass-bottomed dishes, the cells (NCM460, DLD1, SW480, and SW620) were seeded and cultured at 37 C. After incubation with TTNIR (1 mM) for 20 min, the cells were washed with PBS three times and subjected to imaging analysis using a laser scanning confocal microscope (Zeiss Laser Scanning Confocal Microscope; LSM7 DUO). The excitation flter was 488 nm and the emission flter was 570-740 nm. For the costaining assay, the AIEgen loaded COS-7 cells were subjected to incubation with BODIPY 493/503 Green or Nile red for 20 min. Afterwards, the cells were washed with PBS and then observed with CLSM. The cells were imaged using appropriate excitation and emission flters for each dye. The co-localization efficiency was analyzed with Olympus FV10-ASW software, in which the calculated Pearson's coefficient was above 0.90. ## Photostability For the photostability test, the cells were imaged using a confocal microscope (Zeiss Laser Scanning Confocal Microscope; LSM7 DUO) and analyzed using ZEN 2009 software (Carl Zeiss). Both TTNIR and BODIPY493/503 Green were excited at 488 nm for one-photon imaging (1% laser power). The scanning speed was 22.4 s per scan, and the repeated image scans were taken 40 times. The frst scan of both TTNIR and BODIPY493/ 503 Green was set to 100%, followed by which the pixel intensity values were averaged and plotted against the scan number. The resulting curve represents the bleaching rate. ## ROS generation and PDT study H2DCF-DA was used as the ROS generation indicator. In the experiments, 10 mL of H2DCF-DA stock solution (1.0 mM) was added to 2 mL of TTNIR suspension, and white light (18 mW cm 2 ) was employed as the irradiation source. The emission of H2DCF-DA at 534 nm was recorded at various irradiation periods. HeLa cells were seeded in 96-well plates (Costar, IL, USA) at a density of 6000-8000 cells per well. After overnight cell culture, the medium in each well was replaced with 100 mL fresh medium containing different concentrations of TTNIR. Following 30 min incubation, the plates containing HeLa cells were exposed to white light (around 18 mW cm 2 ) for 30 min, and another array of plates with cells were kept in the dark as the control. ## Cell fusion Two dishes of COS-7 cells were incubated with TTG and TTNIR for half an hour separately. After that the cells were washed with PBS 3 times, collected by adding trypsin, and centrifuged respectively. Then the cells were mixed together and incubated for 2 hours in another Petri dish with a cover glass. 10 g of polyethylene glycol 3400 was dissolved in 10 mL of Dulbecco's modifed Eagle's medium (DMEM) without FBS. The mixed culture was overlaid for 5 min at 37 C with 2 mL PEG solution. Then the PEG solution was gradually diluted with DMEM in four steps at the interval of 2 min, by adding 0.5, 1, 2, and 4 mL DMEM, respectively, after which the liquid was removed and replaced with DMEM. ## Conflicts of interest There are no conflicts to declare.

chemsum

{"title": "Facile synthesis of AIEgens with wide color tunability for cellular imaging and therapy", "journal": "Royal Society of Chemistry (RSC)"}

fluorofoldamer-based_salt-and_proton-rejecting_artificial_water_channels_for_ultrafast_water_transpo

## Abstract: We report here the best artificial water channel ever reported in terms of structural robustness, facile synthesis and water transport property.Here, we report on a novel class of fluorofoldamer-based artificial water channels (AWCs) that combines excellent water over ion selectivity with extraordinarily high water transport efficiency and structural simplicity and robustness. These AWCs were produced by a facile one-pot copolymerization reaction under mild conditions. Among these channels, the best-performing channel (AWC 1) is a n-C8H17-decorated foldamer nanotube with an average channel length of 2.8 nm and a pore diameter of 5.2 Å. AWC 1 demonstrates an ultrafast water conduction rate of 1.4 × 10 10 H2O/s per channel, outperforming the archetypal biological water channel, aquaporin 1, by 27%, while excluding salts (i.e., NaCl and KCl) and protons. Unique to this class of channels, the inwardly facing C(sp2)-F moieties are proposed as being critical to enabling the ultrafast and superselective water transport properties observed. ## INTRODUCTION Scarcity of clean water is one of the critical grand challenges facing humanity that currently affects over 4 billion people worldwide (1,2). An important state-of-the-art technology for clean water production and wastewater reuse is reverse-osmosis (RO) membrane desalination (3). The key to RO desalination is precise control over transient or fixed sub-nanometer scale passages across the membrane that only allow water molecules to pass through while excluding other solutes like salt ions (4). In Nature, living organisms regulate transmembrane water flow by membrane-embedded water channels, viz. aquaporins (AQPs). These proteinic channels facilitate superfast water translocation and at the same time completely reject salts and even protons (5,6). For instance, AqpZ, isolated from E. coli, features a water transport rate of ~ 6 × 10 9 H2O/s (7,8). The other type of AQPs, AQP1 that is present in specific human cells can transport ~ 1.1 × 10 10 H2O/s ( 9), yet with remarkably high water to monovalent ion selectivity over 10 9 . Integration of such water-permeating and salt-rejecting AQPs into polymer-based membranes represents an emerging approach for developing the next generation of water desalination and purification technology (10)(11)(12). Nevertheless, membrane proteins like AQPs usually suffer from high production costs, challenges with scalability, and questions about structural stability in abiotic environments (13), making them less ideal for large-scale industrial applications. Motivated by the superior performance of natural AQPs, researchers have expanded extensive effort in developing artificial water channels (AWCs) with simpler structures yet comparable or even exceeding water transport capabilities (14)(15)(16). In 2007, Percec and co-workers reported the pioneering work in this field, wherein dendritic dipeptides were employed for the construction of AWCs in lipid membrane (17). Thereafter, various types of unimolecular or self-assembled AWCs have been designed and characterized, including imidazole-quartet (18)(19)(20)(21)(22)(23), pillar [n]arenes (24)(25)(26)(27), aromatic macrocycles (28), carbon nanotube porins (CNTPs) (29,30), porous organic cages (31), helically folded polymeric nanotubes (32)(33)(34), and hydrophilic hydroxyl assemblies (35). The collective conclusion states that water transport efficiency and selectivity highly depend on the geometry and surface chemistry of the channel interior lumen, in which channel-water and water-water interactions occurs, primarily via H-bonds (9,33,34). However, concurrently achieving high single-channel water permeability and high transport selectivity (e.g., rejection of salts and protons) in a single AWC still remains a daunting task to date that has been addressed in only a few studies (26,34). Here we report on such a high-performance salt-and proton-rejecting AWC system that has a 5.2 -diameter cavity and transports water at a remarkable rate of ~1.3 times that of AQP1, outperforming by at least a factor of 4 all other hitherto known salt-rejecting AWCs , except for one very recent example (34). ## Molecular design Although the lone pair donation from fluorine, being the most electronegative element in the periodic table, is significantly suppressed, making it a poor H-bond acceptor (36), early studies have established the ability of C(sp2)-F to form weak intramolecular H-bonds in foldamer structures (37)(38)(39). Further, fluorine atoms may differ considerably from other H-bond-forming groups in determining the foldamer channel construct and guest binding behaviors (40) by altering the interior pore size and wall smoothness, channel backbone distortion, intermolecular host-guest H-bond interactions, etc. With these concepts in mind, we decided to explore fluorofoldamer-based polymeric hollow channels, having inward-facing fluorine atoms decorating the channel lumen, as possible AWCs. ## AWC 1 exhibits the best water transport performance Screening a matrix of such AWCs, combinatorically derived from different reaction conditions and monomer structures, culminated in a discovery of the best-performing AWC, water channel 1. AWC 1 is found to conduct water at an ultrafast rate of 1.4 × 10 10 H2O/s across lipid bilayer membrane, a value that is about 1.3 times higher than that of AQP1 and two times higher than its methoxy-containing analogous channel 1-OMe (see later discussions). Further, the readily synthesized 1 also demonstrates near-perfect salt (NaCl and KCl) and proton rejection, making it an excellent replacement of natural AQPs for possible industrial uses in fabricating next-generation of AWC-based RO membrane for seawater desalination or for use in therapeutics (26). Synthesis of 1 was carried out by following a previously reported protocol (41). Briefly, a facile one-pot copolymerization reaction between diamine monomer A1 (with n-C8H17 side chains) and fluoro-containing diacid B using HBTU as the coupling reagent readily produced an off-white powdery product 1 with ~80% isolated yield (Fig. 1a). Apart from extensive π-π stacking, intramolecular H-bonds are also expected to stabilize the polymeric product in a helically folded configuration (42). Molecular dynamics (MD) simulations of the pore scaffold shows that the optimized structure exhibits expected helical tubular shape with The inner pore diameter is ~5.2 after subtracting the van der Waals radii of the interior atoms. This pore diameter is larger than a water molecule (2.8 ), but smaller than first-shell hydrated Na + or K + ions. The average molecular weight of 1 was measured to be 13.9 kDa using gel permeation chromatography (GPC). A NMR-based method was also applied to determine the molecular weight, in which a chiral group was introduced at the amine end of 1 as an internal NMR standard (Scheme S3). Based on the area integration ratio of specific 1 H signals (Fig. S1), molecular weight of 1 was determined as 15.4 kDa (Table S2), agreeing well with the GPC-derived value. Using the simulated pore structure as the guide (three AB units per helical turn, MW of AB unit = 616.7 Da; see Fig. 1b and Supplementary Table S1), 1 contains 25 AB units in average, measuring at 2.8 nm in average nanotubular length that is dimensionally comparable to the thickness of typical lipid bilayer membranes (e.g., 2.7 nm for DOPC) (43). In addition, 1 also displays a characteristic mass pattern with a repeating unit of 617 Da in the MALDI-TOF spectrum (Fig. S2). The unique structural features of 1, including (i) appropriate inner pore diameter intermediate between a water molecule and hydrated ions (e.g., Na + ), (ii) membrane-spanning channel length of 2.8 nm, and (iii) special lumen surface chemistry with H-bond donors/acceptors and dipolar C(sp2)-F moieties, lays the structural basis for its water transport property. Stopped flow light-scattering method was employed to quantify the water transport efficiency, using large unilamellar vesicle (LUVs, 120 nm diameter, Fig. 1c) with channel 1 pre-inserted in the LUV wall (24,25). Under the shrinkage mode, LUVs were exposed to hypertonic buffer solution containing 200 mM sucrose, which induces water efflux and vesicle shrinking. The time-dependent variation of the light-scattering intensity was then captured and analyzed (Fig. 1d), from which water transport rate can be reliably determined. As shown in Fig. 1e, water permeability of 1 was largely independent of the lipid to channel molar ratio (mLCR), and the profile peaks at 12000:1, giving water permeability PW of (41.2 ± 2.1) × 10 -14 cm 3 /s. With a channel insertion efficiency of 87.9% at this mLCR (Table S3), water permeability translates into a single-channel water transport rate of (1.4 ± 0.07) × 10 10 H2O/s, which is 133% and 27% faster than the biological AqpZ and AQP1 water channels, respectively (44). The value of 1.4 × 10 10 H2O/s becomes 0.78 × 10 10 H2O/s using the new equation for Pf correction (44). The water permeability of 1-reconstituted LUVs at different temperatures (6-25 °C) were measured, from which its activation energy Ea is calculated as 7.1 ± 1.2 Kcal mol -1 using the Arrhenius Equation (Fig. 1f). It is higher than that of the AQPs (~ 5 Kcal mol -1 ), but much lower than that from the blank DOPC LUV (12.3 ± 0.2 Kcal mol -1 ). In view of the superior water conduction rate of 1 compared to AQPs, we assume that low activation energy might not be a necessary feature for highly permeable AWCs, likely because the transport mechanisms differ from that of AQPs in Nature, as proposed before (33,34). ## Impact of channel length on water permeability Following identical synthetic protocols (41), other amide coupling reagents (HATU, BOP and TBTU) produce the same A1B type channels with NMR-derived molecular weights of 20.1, 19.9 and 13.1 kDa (Table S1) that correspond to channel lengths of 4.0, 4.0 and 2.6 nm, respectively. As summarized in Fig. S3, their water transport rates were all found to be lower that of channel 1 (MW = 13.9 kDa, 2.8 nm). More specifically, at the mLCR of 12000:1 and compared to 1 of 2.8 nm (Pw = 41.2 × 10 -14 cm 3 /s), A1B type channels produced using HATU (4.0 nm), BOP (4.0 nm) and TBTU (2.6 nm) show much lower Pw values of 21.9 × 10 -14 , 22.0 × 10 -14 and 32.4 × 10 -14 cm 3 /s, respectively. ## Impact of side chain type on water permeability To examine the impact of channel side chains on water transport property, diamine monomers A2 and A3 (carrying n-(CH2CH2O)3CH3 and i C4H9 side chains respectively) were also employed in the HBTU-facilitated copolymerization reaction. Their corresponding products (A2B)n and (A3B)n were named channels 2 and 3, respectively. Monomers A1 and A2 were further pre-mixed in 1:1 ratio, and then stoichiometrically reacted with B to produce mixed copolymers (A1BA2B)n (e.g., channel 4). From their NMR-derived molecular weights (Supplementary Table 2), the channel tubular lengths can be estimated to be 3.1, 2.9 and 4.1 nm for 2, 3 and 4, respectively. At the mLCR of 12000:1, 2 -4 show much lower Pw values of 2.8 × 10 -14 , 20.2 × 10 -14 and 29.4 × 10 -14 cm 3 /s, respectively. The comparative data among 1-3 indicates the importance of channel side chain lipophilicity on water transport efficiency, and clearly the linear n-C8H17 represents the best performer for the fluorofoldamer-based AWC scaffold. ## High salt and proton rejection capacity of AWC 1 Besides ultrafast water conduction, the other major challenges for AWCs in mimicking AQP performance are to achieve complete rejection of salts and protons. To this end, we firstly compared the osmotic water permeability (Pf in cm/s) values of 1 under three hypertonic conditions (300 mM sucrose, 150 mM NaCl, or 150 mM KCl, Fig. 2a). Since large sucrose molecules are not able to permeate through channel 1, the reflection coefficient, defined as Pf(MCl)/Pf(sucrose) where M + = Na + or K + , was used to approximately gauge the transport of salt ions. The well-established dimeric cation-transporting channel gramicidin A (gA) was employed as the positive control, which shows expected reflection coefficients of 0.53 ± 0.02 and 0.07 ± 0.001 for NaCl and KCl, arising from its high permeability to both Na + and K + ions. In contrast, the reflection coefficients of 1 were calculated to be 1.02 ± 0.01 for Na + and 1.05 ± 0.01 for K + at 12000:1 mLCR, confirming the inability of 1 to transport either cation across the membrane and its near-perfect salt rejection property (19). The rejection of Na + and K + cations was further validated by the fluorescence-based HPTS assay, with pH-sensitive HPTS dye molecules entrapped in the LUVs (Fig. 2b). The intravesicular region is set pH 7, whereas the extravesicular environment is maintained at the same pH but with 200 mM M2SO4 (M = Na or K). Under this high salt gradient, H + /M + antiport will increase the pH of the intravesicular region and hence enhance the HPTS fluorescence intensity. As shown in Fig. 2b, 1 at 1 µM was found non-responsive towards Na + or K + gradient, affirming the impermeability of neither cation through the inner pore of 1. On the contrary, gA at the identical channel concentration (1 µM) could efficiently transport Na + (291%) and K + (343%) cations. Such observation is in excellent agreement with the reflection coefficient results described earlier, both confirming the inability of 1 to transport cations. The anion transport ability of 1 was further examined by using Clˉ-sensitive SPQ dye molecules entrapped in LUVs (Fig. 2d). As expected, 1 at 1 µM displayed similar SPQ quenching as background (9%), whereas the well-established chloride transporter L8 (45) at 6 µM (corresponding to 1 µM channel concentration) displayed significant decrease (45%) in the SPQ fluorescence intensity. In another set of LUV-based experiments where intravesicular region has 100 mM NaCl at pH 7 and extravesicular region has 67 mM Na2SO4 at pH 8 (Fig. S5), gA (cation channel, 1 µM), FCCP (proton carrier, 1 µM) and L8 (anion channel, 1.3 µM) induce HPTS fluorescence increase of 56%, 22% and 138% respectively. In sharp contrast, 1 at 1 µM does not cause any fluorescence change. Proton translocation was probed using the pH gradient set across the membrane (Fig. 2c). Serving such a purpose, the intra-LUV region contains 100 mM NaCl at pH 7, whereas the extravesicular region was set pH 8 with 100 mM NaCl or KCl. If 1 is able to transport protons, the proton efflux (coupled with passive diffusion of cations or anions for charge neutralization) will induce significant pH increase in the intravesicular region, and dramatic change in HPTS fluorescence intensity will emerge. Experimentally, no fluorescence change was observed at all after addition of 1 (1 µM), suggesting negligible transport of protons. Using a conservative approach (for details, see the Supplementary Section of "Estimation of Proton Transport Rate"), the proton transport rate of 1 is estimated to be less than 0.25 proton/s. Stopped-flow fluorescence analysis was further applied to quantitatively measure chloride permeability through DOPC membrane in the absence and presence of 1 (Fig. S6) (34,46). Based on the determined single-channel Clpermeability PCl of (1.7 ± 07) × 10 -20 cm 3 /s, the water-to-Clpermselectivity (e.g., Pw/PCl) for 1 was calculated to be (2.5 ± 1.2) × 10 7 . Since NaCl permeability is limited by the Na + ions in actual desalination processes (46), 2.5 × 10 7 represents a conservative estimate of the water-to-NaCl permselectivity for 1. As compiled in Fig. 3a, this value exceeds the permeability-selective trade-off trendline of current desalination membranes (11,47) by a factor of ~10 2 , signifying good potential for developing novel AWC-based desalination membrane that incorporates or is made of 1. ## Comparison with two high-performance AWCs As summarized in Fig. 3b, currently there are only two water-transporting systems having higher water conduction rates than both AQP1 (1.1 × 10 10 H2O/s) (9) and AWC 1 (1.4 × 10 10 H2O/s), i.e., the relatively low selectivity CNT porin (2.3 × 10 10 H2O/s) (30) and the highly selective AWC 4-LA (2.7 × 10 10 H2O/s) (34). But it is worth emphasizing that while the water-transporting CNT porin also conducts ions and protons (30), channel 4-LA requires additional lipid anchors (LA) installed at the helical ends to orient the channel's alignment to achieve the ultrafast water conduction (34). Without such LA modifications, it's water transport rate drastically drops by 75% to ~ 0.6 × 10 10 H2O/s (34), a value that is ~43% capacity of 1. Further, it is possible that the LA-enhanced water transport property might deteriorate over time or be altered by the complex environment of a water purification membrane. All these make both CNTP and 4-LA potentially less competitive for fabricating practical AWC-based biomimetic water purification membranes than 1 developed in this work (12,48). ## Critical roles played by fluorine atoms To demonstrate the crucial role of C(sp2)-F moieties in determining the water transport property of channel 1, we compared it with the recently reported analogous channel denoted as 1-OMe of 3.0 nm in height, which differs from 1 in that 1-OMe contains methoxy groups in the positions of F-atoms of 1 (34). As a result of bulky hydrophobic methyl groups helically arranged around the pore interior of 1-OMe, its helical backbone is slightly less curved than that of 1 having its pore surface decorated by F-atoms. Consequently, the pore diameter of 1-OMe is enlarged to 6.5 across, which is larger than 1 (5.2 across). Under the identical conditions, the water transport rate of 1-OMe was determined at ~ 5 × 10 9 H2O/s (34), which is 36% that of 1. Furthermore, unlike 1 with excellent ion-rejection capability, 1-OMe was permeable to anions (41). Therefore, we speculate that the superior water transport properties of 1 should arise from a collection of influencing factors induced by the inward-facing C(sp2)-F moieties, including the smaller atomic size, weak H-bond acceptor ability, dipolar bond characteristics and good hydrophobicity. Further investigative efforts to decipher these factors are currently underway in our laboratory. ## Molecular dynamics simulation To provide a molecular level explanation of transmembrane water transport through the pore of 1 embedded in POPC lipid bilayer membrane, we performed 800 ns long all-atom molecular dynamics (MD) simulations, (Fig. 3c and Supplementary Video 1). To maintain QM-derived diameter of 1 at the MD level, we used the RMSD colvar module of NAMD during the course of MD simulation. Fig. S8 shows the RMSD of 1 as a function simulation time. As the simulation begins, water molecules rapidly start permeating across the lipid bilayer through 1 (Fig. 3d). A linear fit to the water permeation vs simulation time (excluding first 200 ns) yields a permeation rate of ~3 water molecules/ns, which is higher than 1.2 water molecules/ns for AQP1. At any given instant of time, a water cluster typically having 30-50 water molecules resides inside the channel, with a mean of 40.5 water molecules (Fig. 3e). Among them, 40.7% or 16.5 water molecules are considered as proton wire breakers (Fig. 3f), which were described and defined in our recent study (34). Interacting with the neighboring water molecules via zero or just one H-bond, or two H-bonds solely via only O-atoms or only H-atoms (Fig. 3g), these breakers prevent forming a continuously H-bonded channelspanning water chain through which protons hop via the Grotthuss mechanism. Interestingly, the breaker type involving the formation of two H-bonds with the adjacent water molecules using only H-atoms is also observed in the NPA motif of AQPs (49). The existence of these proton wire breakers accounts for low proton permeability of 1. Due to the narrow pore, each water molecule forms 1.94 H-bonds with other water molecules inside the channel (Fig. S9a,b) and 0.79 H-bonds with the channel wall (Fig. S9c), leading to a total of 2.73 H-bonds per water molecule. Taking 4 H-bonds per water molecule (EH-bond = 5.1 Kcal/mol) in bulk water (50,51), the activation energy for water entry into 1 can be estimated to be 6.5 Kcal/mol, which is consistent with the experimentally determined value of 7.1 Kcal/mol (Fig. 1f). The fact that 1 has a higher activation energy but transports water faster than AQP1 can be largely attributed to its larger pore diameter of 5.2 vs ~ 2.8 opening in the central channel of AQP1 as well as the more than one water wire molecule occupying the pore lumen that differs from the single file transport seen in AQP1 (34). ## CONCLUSIONS In summary, we have demonstrated ultrahigh water transport efficiency and excellent selectivity of a novel class of fluorofoldamer-based artificial water channels. Produced by facile one-pot copolymerization reaction with good yields, the best-performing water channel 1 of 2.8 nm in average channel length shows a remarkable water conduction rate of 1.4 × 10 10 H2O/s and near-perfect rejection of salt ions (Na + , K + , Cl -) and protons. This work uncovers the positive effects of introducing C(sp2)-F moieties on the inner rim of foldamer-based water channel pores, providing a new dimension of channel design principles. This, we believe, will stimulate further development towards the next-generation of membrane technologies for water desalination, nano-filtration and medical dialysis applications. ## Materials All reagents were obtained from commercial suppliers and used as received unless otherwise noted. Aqueous solutions were prepared from MilliQ water. Egg yolk L-α-phosphatidylcholine (EYPC) and 1,2-dioleoyl-sn-glycero-3-phosphocholine lipid (DOPC) were obtained from Avanti Polar Lipids. HEPES, HPTS, SPQ and FCCP refer to 4-(2-hydroxyethyl)-1-piperazine-ethane sulfonic acid, 8-hydroxypyrene-1,3,6-trisulfonic acid, 6-methoxy-N-(3-sulfopropyl)quinolinium, carbonyl cyanide-p-trifluoromethoxyphenylhydrazone, respectively. ## Water transport study In a 2 mL microcentrifuge tube, 6 mg DOPC (0.24 mL, 25 mg/mL in CHCl3, Avanti Polar Lipids, USA) and water channel compound (dissolved in CHCl3) were mixed at different molar ratios (4000:1 to 15000:1). The solvent was slowly removed by N2 flow and the resulting thin film was dried under high vacuum overnight. 1 mL HEPES buffer (10 mM HEPES, 100 mM NaCl, pH = 7.0) was then added into each tube for lipid hydration. In order to maximize incorporation of channel molecules into the lipid bilayer, each microtube was vortexed for 30 s and sonicated for 150 s (37 kHz, power 100, 70 °C) for 10 cycles. A glass spatula was used to scratch down all the lipid residues from the microtube wall to minimize lipid loss and maximize channel incorporation whenever necessary. The lipid/channel mixture was then subjected to 10 freeze-thaw cycles (freezing in liquid N2 for 1 min and heating at 55 °C in water bath for 2.5 min). The mixture was then extruded through polycarbonate membrane (0.1 μm) at 80 °C for 15 times to give LUVs at 6 mg mL -1 lipid concentration. For stopped-flow experiments, this LUV solution was diluted to 1 mg mL -1 with buffer (10 mM HEPES, 100 mM NaCl, pH = 7.0). The LUVs were then exposed to a hypertonic solution (200 mM sucrose, 10 mM HEPES, 100 mM NaCl, pH = 7.0). During stopped-flow experiment, the abrupt decrease in vesicle size was expected due to transport of water to the extravesicular pool and this event leads to increase in the light scattering intensity of 90° angle according to the Rayleigh-Gans theory. The changes of light scattering intensity caused by vesicle shrinkage were recorded at a wavelength of 577 nm and all these plots were fitted in the following form of single exponential function to give rate constant (k) value using the equation shown below: y = Aexp(-kt) + y0 where y is change in the light scattering, k is the exponential coefficient of the change in the light scattering and t is time. With the assumption that change in the light scattering intensity is proportional to the change in the vesicle volume (ΔV/V0) based on the Boyle-van't Hoff law, the osmotic permeability (Pf) in the unit of cm/s was commonly calculated as follow: where k is the exponential coefficient of the change in the light scattering ; S and V0 are the initial surface area and volume of the vesicles, respectively; Vw is the molar volume of water, and Δosm is the osmolarity difference. The size of LUVs was determined by dynamic light scattering after 10 times dilution of the aforementioned LUVs solution (i.e., 1 mg mL -1 ) with buffer (10 mM HEPES, 100 mM NaCl, pH = 7.0). Following the new approach proposed by Horner and co-workers (44), the water permeability can be alternatively calculated using the new equation shown below: Pf = k x (Cin,o + Cout)/ ((2Cout 2 ) x ((S/V0) x VW)) Where Cin,0 and Cout are the osmolytes concentration inside at t = 0 s and outside of vesicles, respectively. To calculate the true water permeability (PW in the unit of cm 3 /s) of water channels, the Pf(blank) value of the blank vesicle without water channels needs to be deducted from Pf(channel), which was multiplied by the vesicle surface area (S) and divided by the number of water channels (N) incorporated in the liposome as shown below. Further taking into consideration of channel incorporation efficiency (CIE, Supplementary Table 3), the Pw values can be calculated by the following equation: PW = (Pf(blank) -Pf(channel)) x (S/(N x CIE)) The HPTS assay for cation transport EYPC (1 mL, 25 mg/mL in CHCl3, Avanti Polar Lipids, USA) was placed in a 10 mL round bottomed flask and solvent was evaporated by slowly purging N2. After drying the resulting thin film under high vacuum overnight at room temperature, the film was hydrated with a HEPES buffer solution (1 mL, 10 mM HEPES, pH = 7.0) containing pH-sensitive HPTS dye molecules (0.5 mM) at room temperature for 1 hour (with occasional vortexing after every 15 minutes) to give a milky suspension. The mixture was then subjected to 10 freeze-thaw cycles (freezing in liquid N2 for 1 minute and heating at 55 o C in water bath for 2 minutes). The vesicle suspension was extruded through polycarbonate membrane (0.1 μm) to produce mostly monodispersed LUVs of about 120 nm in diameter with HPTS dyes encapsulated inside. The extravesicular HPTS dye was removed by using size exclusion chromatography (stationary phase: Sephadex G-50, GE Healthcare, USA; mobile phase: 10 mM HEPES buffer, pH = 7.0) and diluted with the mobile phase to yield 3. ## The HPTS assay for anion transport The SPQ-containing LUV suspension (30 μL, 10 mM lipid, 200 mM NaNO3) was added to a NaCl solution (1.97 mL, 200 mM NaCl) to create a chloride concentration gradient for ion transport observation. A solution of 1 at 120 µM in DMSO was then injected into the suspension under gentle stirring. Upon channel addition, the SPQ dye emission was immediately monitored at 430 nm with excitation at 360 nm for 300 seconds using fluorescence spectrophotometer (Hitachi, Model F-7100, Japan), after which time an aqueous solution of Triton X-100 (20 μL, 20% v/v) was added to completely eliminate the chloride gradient. The final transport trace was obtained by normalizing the fluorescence intensity using equation shown below. ## IF = [(Ft -F0)/(F1 -F0)] x 100 Where F0 = fluorescence intensity just before the channel addition (at t = 0 s), Ft = Fluorescence intensity at time t, and F1 = fluorescence intensity after addition of Triton-X100. ## Activation energy measurements To determine activation energies for water transport, we conducted water permeability measurements at different temperatures at intervals of 5 °C between 20 and 40 °C. For these experiments, the solution reservoir and the measurement cell of the stopped-flow instrument were maintained at a set temperature by a recirculating heater/chiller (Polystat, Cole Parmer). Permeability rates through channels at varying temperatures were used to construct an Arrhenius plot. ## Ln(k) = Ln(A) -Ea/(RT) where k is the exponential coefficient of the change in the light scattering; A is pre-exponential factor; Ea is activation energy; R is gas constant; T is absolute temperature in Kelvin. ## Molecular dynamics simulations All MD simulations were performed using the MD program NAMD2 with periodic boundary conditions and using the particle mesh Ewald (PME) method to calculate the long-range electrostatics. The Nose-Hoover Langevin piston and Langevin thermostat were used to maintain the constant pressure and temperature in the system. CHARMM36 force field parameters describe the bonded and non-bonded interactions of among, lipid bilayer membranes, water and ions. An 8-10-12 cutoff scheme was used to calculate van der Waals and short range electrostatics forces. All simulations were performed using a 2 fs time step to integrate the equation of motion. SETTLE algorithm was applied to keep water molecules rigid whereas RATTLE algorithm constrained all other covalent bonds involving hydrogen atoms. The coordinates of the system were saved at an interval of 19.2 ps. The analysis and post processing the simulation trajectories were performed using VMD and CPPTRAJ. The initial structure of channel 1 having 25 AB repeating units was built using a fragment-assembly strategy. Specifically, a helical fragment containing 8 AB repeating units was built using Gaussview and optimized at the HF/6-31G(d) level. Based on the optimized structural parameters (bond angle/length, dihedral angle, etc) of this helical fragment, we then built longer channel 1. The topology and force field parameters for the monomeric unit of 1 were created using the CHARMM general force fields (CGenFF) webserver. Subsequently, the channel was embedded into a 10.5 x 10.5 nm 2 patch of pre-equilibrated POPC lipid bilayer membrane. The lipid patch was generated using the CHARMM-GUI membrane builder and pre-equilibrated for approximately 400 ns. Lipid molecules that overlapped with the channel were removed. The system was then solvated with water using the Solvate plugin of VMD. Sodium and chloride ions were added to 0.6 M concentration using the Autoionize plugin of VMD. The final assembled system measured 10.5 x 10.5 x 9.0 nm 3 and contained 100,482 atoms. Following the assembly, the system underwent 1200 steps of energy minimization using the conjugate gradient method to remove steric clashes. After energy minimization, the system was subjected to a 48 ns equilibration at a constant number of atoms (N), pressure (P = 1 bar) and temperature (T = 300 K), the NPT ensemble, with harmonic restraints applied to all non-hydrogen atoms of channels that surrounded the transmembrane pore. The restraints were applied relative to the initial coordinates of the atoms, with spring constants at 1 kcal mol -1 -2 . After 48 ns, the harmonic restraints were removed, and the system was equilibrated while restraining the RMSD of the channel with respect to its QM optimized initial conformation using the colvar module of NAMD. For the corresponding references, see the supplementary information.

chemsum

{"title": "Fluorofoldamer-Based Salt-and Proton-Rejecting Artificial Water Channels for Ultrafast Water Transport", "journal": "ChemRxiv"}

a_new,_<i>substituted</i>_palladacycle_for_ppm_level_pd-catalyzed_suzuki–miyaura_cross_couplings_in_

## Abstract: A newly engineered palladacycle that contains substituents on the biphenyl rings along with the ligand HandaPhos is especially well-matched to an aqueous micellar medium, enabling valued Suzuki-Miyaura couplings to be run not only in water under mild conditions, but at 300 ppm of Pd catalyst. This general methodology has been applied to several targets in the pharmaceutical area. Multiple recyclings of the aqueous reaction mixture involving both the same as well as different coupling partners is demonstrated.Low temperature microscopy (cryo-TEM) indicates the nature and size of the particles acting as nanoreactors. Importantly, given the low loadings of Pd invested per reaction, ICP-MS analyses of residual palladium in the products shows levels to be expected that are well within FDA allowable limits. ## Introduction In a recent review by Alami and Messaoudi, 1 palladacycles were characterized as among the "most powerful" pre-catalysts to highly reactive, mono-ligated forms of Pd(0). 2 While they are stable, quite convenient, and broadly applicable, these features may not be sufficiently attractive for long term usage typically in the 1-5 mol% range. That is, not only is awareness of the endangered status of Pd gaining in appreciation, such reagents are also almost invariably used in environmentally egregious organic solvents and with limited levels of solvent and precious metal recycling. 3 Put another way, and notwithstanding Nobel Prize-level recognition bestowed in 2010 on Pd-catalyzed Suzuki-Miyaura (SM) cross-coupling chemistry developed several decades earlier, 4 such an approach to modern Pd-based catalysis as practiced today is both costly and not sustainable. Access to worldwide supplies, even under the best of geopolitical circumstances, is determined by current limits of technology that prevent access to metals that lie too deep within the Earth's surface. 5 One solution to this inevitable shortage calls for switching, in large measure, from a petroleum-to a water-based discipline, 6 akin to the role of water as the reaction medium in nature, in general, and biocatalysis in particular. 7 Along with this gradual transition comes myriad opportunities for developing new catalysts engineered to function both within an altered reaction medium and under newly unfolding rules for which analogies in organic solvent are non-existent. 8 Palladacycles present one such opportunity where, under traditional conditions (i.e., use in organic solvents), the focus has been on modifcations that include, e.g., the nature of the leaving group. Such design changes that feature steric and stereoelectronic effects, conformational biases, etc., have little to do with palladacycle solubility. In an aqueous medium containing micellarbased nanoreactors, 9 however, solubilization becomes a crucial parameter. Thus, one key to successful couplings in water involves influencing the binding constant of a reagent to the micellar inner core: the greater the incentive to enter the site of reaction, the more catalytic activity is to be expected and the lower the catalyst loadings. Hence, the question arises: could the appropriate substitution pattern on the biaryl skeleton within a palladacycle pre-catalyst enhance its micellar entry, and thereby reduce the required level of otherwise precious metal, and associated (oftentimes equally precious) ligand, in a cross-coupling reaction? Substituted biarylamine-based palladacycles are currently unknown, 10 since in organic solvent there is no reason to pursue such derivatization. In water, however, where new rules are operating, 8b the prospects for not only providing the convenience of palladacycles that lead to especially reactive catalysts as well as the potential for addressing the endangered nature of platinoids provides more than ample justifcation for investigating this nontraditional approach. In this report we describe such a newly adorned palladacycle pre-catalyst that, indeed, allows for a general and environmentally responsible process for SM couplings to be run, in most cases, in water under mild conditions and at the 300 ppm (0.03 mol%) level of Pd (Scheme 1). ## Results and discussion Several newly substituted palladacycles were prepared initially containing the ligand rac-HandaPhos 11 (P1, P4, P5, P6 and P7), the structures of which are shown in Scheme 2. These included either one or two lipophilic t-butyl residues (as in P4 and P5, respectively), along with those bearing one (P6) or two isopropyl moieties (P7), as compared to the parent array (P1). Several additional ligands commonly used within palladacycles were also prepared and tested under identical conditions in the model reaction between bromide 1 and boronic acid 2 to arrive at biaryl 3. Clearly, the most effective catalyst, by far, is the diisopropyl-substituted HandaPhos palladacycle, P7. Surprisingly, even the di-t-butyl analog, P5, was not as effective. What may also be found striking at frst, but is perfectly in harmony with the "new rules" associated with this chemistry in water, 8 are the results observed for catalysts P9 and P10, where HandaPhos has been replaced by XPhos and SPhos, respectively. While both are excellent choices for SM couplings in traditional organic solvents (e.g., toluene or dioxane), 12 they are non-functional at 300 ppm in water. Likewise, other well-known ligands that make up catalysts P11 and P12 were not competitive under these conditions. Additional optimization studies regarding the choice of surfactant, base, as well as results using organic solvents are described in the associated ESI. † Several examples of SM couplings were carried out using 300 ppm of catalyst P7, as summarized in Table 1. While most cases were amenable to this very low loading of Pd, some could be conducted at levels even down to 25 (4) to 100 (5) ppm of P7. On the other hand, some cases (10, 12, 13, 18-21, 13 and 25) required up to 500 ppm, possibly reflecting competition by the product for palladium. Both electron-donating and -releasing groups in the educt are tolerated. Heterocycles present in either the halide or boronic acid partner could be used. Partners with protecting groups are easily coupled (20 and 21). Polyaromatics such as 27 and 29 are easily fashioned. Alternatives to boronic acids, including Bpin, Molanderate BF 3 K salts, 14 and MIDA boronates 15 (see 22 and 23) appear to be compatible partners as well. Double SM couplings using the corresponding precursor dibromides proceeded smoothly to give the anticipated products (30, 31, and 32), using what is, formally, only 150 ppm of this Pd catalyst per bond formed. The importance of organic co-solvents has also been addressed, as these additives can play a dramatic role in scaling up reactions under micellar conditions. 16 The co-solvent effect is responsible not only for increasing solubility of highly crystalline educts, but also enlarges micellar size, thereby expanding the interior volume available for reaction. The observed impact of three organic solvents used as 10 vol% in this aqueous surfactant system is shown in Scheme 3 involving coupling partners 33 and 34. Each solvent (THF, toluene, and acetone) was found to increase the rate of conversion. Analysis of the medium for 2 wt% TPGS-750-M/H 2 O vs. that with 10% THF by cryo-TEM (Scheme 3, bottom) revealed the enlargement of the former (ca. 50 nm) to ca. 200 nm due to the presence of THF, suggesting that larger nanoreactors may be responsible for enhancing the overall rates of these cross-couplings. The potential use of less reactive aryl chlorides was briefly examined at the 500 ppm level of Pd catalysis (0.05 mol%). As the examples in Table 2 show, a variety of aromatic and heteroaromatic chlorides and boronic acids could be employed, arriving at the targeted biaryls in good isolated yields. Included in this study is the late stage derivatization of aryl chloride fenofbrate 17 to analog 38. Table 1 Substrate scope for couplings with ppm Pd pre-catalyst P7 in water a a Reaction conditions unless otherwise noted: 0.5 mmol aryl halide, 0.6 mmol aryl boronic acid, 1.0 mmol Et 3 N, 25-500 ppm P7 stirred at 45 C in TPGS-750-M/H 2 O (0.5 M); isolated yields are shown. Double SM couplings were carried out using 1.2 mmol of aryl boronic acid, 2.0 mmol of Et 3 N, and 10% THF as a co-solvent. As an illustration of the opportunities to carry out multi-step processes given the commonality of reaction conditions (i.e., in aqueous nanoreactors at rt-45 C), the commonly used fungicide boscalid 18 could be prepared in three steps using a 1pot protocol (Scheme 4). Initially, biaryl 37 was constructed that, without isolation, was subjected to nitro group reduction using our previously described carbonyl iron powder. 19 The Scheme 4 Boscalid: 3-step, 1-pot synthesis in water. resulting aniline was then treated directly with 2-chloronicotynyl chloride. The fnal product, boscalid, was ultimately isolated in 80% overall yield. The seemingly incompatible addition of this acid chloride to this aqueous medium, while counter-intuitive at frst, is yet another example of the "new rules" associated with chemistry in water. Reagents and/or reaction partners that are sensitive albeit insoluble in water simply do not hydrolyze or quench; rather, upon stirring they enter the hydrophobic inner micellar core where they react, usually as desired. A multi-gram scale reaction between educts 45 and 46 was run in water using Pd catalyst P7 to document the prospects for scaling up these SM couplings (Scheme 5). Use of 24 mmol of 45 and 20 mmol of 46 in the presence of 40 mmol of Et 3 N were exposed to 300 ppm of P7. Stirring this heterogeneous mixture for 15 hours yielded 94% of the desired coupled product 47. In this case, the reaction was quite efficient in the absence of a cosolvent. That is, stirring was not an issue throughout the 15 h reaction period (see images (a-c)) The product 47 could be isolated as a white solid (image (d)), purifed by simple fltration through silica gel. Biologically active targets, such as precursors to (a) Merck's anacetrapib (48), 20a (b) sonidegib (49), 20b and (c) Novartis' valsartan (50), 20c could also be prepared efficiently under mild conditions using 300-500 ppm of catalyst P7 (Table 3). Additional representative examples of biaryls (51-53) en route to anticancer drugs are also to be found in this table. 21 Facile recycling of the aqueous TPGS-750-M solution is an important aspect to this environmentally responsible technology, leading to very low levels of aqueous waste streams. 22 By contrast, recycling of organic media typically requires fractional distillation to separate reaction and workup solvents for re-use. Scheme 6 illustrates just how effective a 2 wt% aqueous solution of TPGS-750-M can be, thereby dramatically minimizing aqueous waste streams. Recycling could be carried out using a different SM reaction with each of four recycles, following an initial coupling. Products were either separated via fltration or by decantation of the aqueous mixture; hence, individual extractions were not required prior to purifcation. cryo-TEM analysis of the aqueous mixture after fve uses revealed that while the nanomicelles were of the same shape, they were unexpectedly larger (ca. 75 nm; Scheme 6). From the perspective of the pharmaceutical industry, it is commonly assumed that under traditional SM cross-coupling conditions the amount of residual Pd in the product is going to be outside of the acceptable 10 ppm limit imposed by the US FDA. 23 Hence, additional processing is usually anticipated, potentially adding time and expense to the eventual API. But use of such low levels of Pd catalysts rarely exceed this limit. With catalyst P7 at the 300 and even 500 ppm loadings it was not surprising that, for the three cases randomly selected and examined by ICP-MS, no more than 6 ppm Pd was found for biaryls 12, 24, and 51 (Fig. 1). On the other hand, following traditional literature conditions used to make each of these biaryl products (e.g., 2 mol%, or 20 000 ppm Pd), residual levels of Pd were found to be orders of magnitude greater. ## Conclusions In summary, a new palladacycle has been uncovered that mediates Suzuki-Miyaura couplings in water at the 300 ppm level of precious metal. Key to this methodology is placement of an isopropyl group on each aromatic ring of the biaryl skeleton making up the palladacycle, a substitution pattern that could not have been predicted given the lack of precedent for such pre-catalysts. Likewise, screening of several monophosphines, including some of the most commonly used for such Pd-catalyzed cross-couplings, ultimately identifying HandaPhos as the preferred ligand (i.e., P7), requires further study to rationalize the effectiveness of this novel ligand/ palladacycle precursor combination. Applications to various targets within the pharma, agro, and materials domains have been demonstrated, along with the potential for large scale use, recycling of the aqueous reaction medium, and tandem 1pot processes. The nature of the nanomicelles involved has been determined via cryo-TEM measurements, both initially as well as after use in the presence of added co-solvent. Residual levels of Pd in the products formed have been shown to be well within governmental limits for safety, further enhancing the attractiveness of this technology. The prognosis for use of the same pre-catalyst for other types of Pd-catalyzed crosscouplings (e.g., Stille, Sonogashira, and Heck couplings) looks encouraging, with the results from these ongoing studies to be reported in due course.

chemsum

{"title": "A new, <i>substituted</i> palladacycle for ppm level Pd-catalyzed Suzuki\u2013Miyaura cross couplings in water", "journal": "Royal Society of Chemistry (RSC)"}

characterization_of_three_tetrabromobisphenol-s_derivatives_in_mollusks_from_chinese_bohai_sea:_a_st

## Abstract: Identification of novel brominated contaminants in the environment, especially the derivatives and byproducts of brominated flame retardants (BFRs), has become a wide concern because of their adverse effects on human health. Herein, we qualitatively and quantitatively identified three byproducts of tetrabromobisphenol-S bis(2,3-dibromopropyl ether) (TBBPS-BDBPE), including TBBPS mono(allyl ether) (TBBPS-MAE), TBBPS mono(2-bromoallyl ether) (TBBPS-MBAE) and TBBPS mono(2,3-dibromopropyl ether) (TBBPS-MDBPE) as novel brominated contaminants. Meanwhile, the mass spectra and analytical method for determination of TBBPS-BDBPE byproducts were presented for the first time. The detectable concentrations (dry weight) of TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE were in the ranges 28-394 μg/g in technical TBBPS-BDBPE and 0.1-4.1 ng/g in mollusks collected from the Chinese Bohai Sea. The detection frequencies in mollusk samples were 5%, 39%, 95% for TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE, respectively, indicating their prevailing in the environment. The results showed that they could be co-produced and leaked into the environment with production process, and might be more bioaccumulative and toxic than TBBPS-BDBPE. Therefore, the production and use of TBBPS derivatives lead to unexpected contamination to the surrounding environment. This study also provided an effective approach for identification of novel contaminants in the environment with synthesized standards and Orbitrap high resolution mass spectrometry.Recently, increasing studies have been carried out to identify novel brominated contaminants in the environment, especially the derivatives, byproducts and degradation products of brominated flame retardants (BFRs) 1-3 . For example, the mono-modified byproducts or degradation products of tetrabromobisphenol-A (TBBPA) derivatives, including TBBPA mono(allyl ether) (TBBPA-MAE), TBBPA mono(2,3-dibromopropyl ether) (TBBPA-MDBPE), have been found in various environment samples, such as soil, sediment, earthworm and mollusks 2,3 . More importantly, these byproducts and degradation products showed higher bioaccumulation and toxicity than main technical TBBPA products 2,4,5 . Due to the similar structures and production process 6,7 , there should be more mono-modified byproducts of TBBPA derivatives co-produced and leaked into environment, which could cause widespread contamination and deserve our more attention. As important alternatives of TBBPA, the most widely used BFR, tetrabromobisphenol-S (TBBPS) and TBBPS bis(2,3-dibromopropyl ether) (TBBPS-BDBPE) are extensively produced and applied in electronic devices, plastics, rubber and textiles 8 . As a result, TBBPS and TBBPS-BDBPE have been detected in waste water at a concentration up to 10 μ g/L 9 . TBBPS-BDBPE was also found in herring gull eggs collected from colonies in the Laurentian Great Lakes 10 . Because TBBPS-BDBPE is synthesized by modification of the two phenol groups of TBBPS 7 , the mono-modified byproducts of TBBPS-BDBPE might also be co-produced with technical products and leaked into environment as potential contaminants, such as TBBPS mono(allyl ether) (TBBPS-MAE), TBBPS mono(2-bromoallyl ether) (TBBPS-MBAE) and TBBPS mono(2,3-dibromopropyl ether) (TBBPS-MDBPE). However, the byproducts of TBBPS derivatives were largely ignored in most studies, and there are even no pure standards available. Much still remains unknown about their environmental distribution and risks. The lack of analytical methods is another main obstacle for identifying novel contaminants. Because of the thermolability of TBBPA, TBBPS and their derivatives, gas chromatography mass spectrometry (GC-MS) is not applicable for direct analysis of these compounds 11,12 . The high performance liquid chromatography coupled with tandem MS (HPLC-MS/MS) has been developed for analysis of TBBPA and TBBPS derivatives 3,8,10,13 . However, electrospray ionization (ESI) source was reported with poor sensitivity because of the weak polarity of TBBPA and TBBPS derivatives 8 . Although atmospheric pressure chemical ionization (APCI) and atmospheric pressure photoionization (APPI) mass spectrometry methods have been developed, they were not sufficient for the trace level determination of these derivatives in the environment matrices 10,13 . With the rapid development of high resolution mass spectrometry (HRMS), such as time of flight (TOF) HRMS and Orbitrap HRMS, the novel contaminants could be identified and quantified through the full scan acquisition spectrum . An attractive advantage of full scan HRMS is that there is no number limitation of analytes in one single injection, which is enormously beneficial to the retrospective analysis of untargeted contaminants 18 . Furthermore, the exact mass information is helpful for identification of compounds without standards, which largely extends its application 19 . In this view, the combination of ultra HPLC (UHPLC) with Orbitrap Fusion HRMS technique would provide a high accuracy as well as a low detection limit for the mono-modified byproducts of TBBPS-BDBPE. The aim of this study was to identify three potential byproducts of TBBPS-BDBPE as novel brominated contaminants. The standards of TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE were synthesized with high purity. A sensitive and accurate method for simultaneous determination of these novel TBBPS derivatives was developed with UHPLC-Orbitrap HRMS. Their distribution in mollusk samples collected from the Chinese Bohai Sea and potential risks were discussed in detail. The strategy used in this work could also be an effective approach for identifying other novel brominated pollutants related to BFRs. ## Structure Confirmation of Synthesized TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE. The standards of TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE were self-synthesized with purities of 99%, 98% and 96%, respectively. The synthesis schemes and 1 HNMR spectra of these compounds are provided in Figure S1 (Supporting Information). Orbitrap Fusion HRMS was employed to further identify the target compounds in the full scan mode (Table 1, The recoveries for TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE were all higher than 76% with ultrasound method and accelerated solvent extraction (ASE) method at the spiking amounts of 10 ng in 0.5 g Neverita didyma (Nev) samples (n = 7, Table 2). The recoveries of ultrasound method were slightly higher than ASE method and the standard deviations (SDs) were lower than ASE method. Finally, the samples were extracted by ultrasound method and cleaned by ENVI-carb cartridges. The mean recoveries for TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE were all higher than 70% and the SDs were all less than 10% at three different spiking amounts, 100 ng (n = 5), 10 ng (n = 7) and 1 ng (n = 5) (Table 2). The method DLs (MDLs) of TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE were 0.04 ng/g dry weight (dw), 0.08 ng/g dw and 0.06 ng/g dw, respectively. The mean recovery of the internal standard, 13 C labeled 3,5-dibromophenol (ISDBP), was 101% ± 7% at the spiking amount of 10 ng (n = 7). The recoveries of ISDBP from the real samples ranged from 81% to 104% with a mean recovery of 91% and SD of 6% (n = 38). The matrix effects ranged from 0.86 to 1.05 (Table 2) at three different spiking concentrations, 1 ng/mL, 5 ng/mL and 10 ng/mL. The pretreatment method of TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE from the real samples was reliable and repeatable. Details for the development and optimization of extraction and solid phase extraction (SPE) cleanup procedure were provided in Supporting Information. ## TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE in Technical Products and Mollusk Samples. The technical product of TBBPS-BDBPE purchased from a BFRs factory (purity > 90%) was dissolved in methanol at a concentration of 100 μ g/mL and determined by Orbitrap Fusion HRMS. The concentrations of TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE in this technical TBBPS-BDBPE were 28 μ g/g, 87 μ g/g and 394 μ g/g, respectively. The concentrations of TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE in total 38 mollusk samples collected from Bohai Sea during 2009 to 2013 were also analyzed with an external standard method (Table S1). TBBPS-MAE was only found in two mollusk samples with concentrations of 0.1 ng/g dw and 0.2 ng/g dw. TBBPS-MBAE was detectable in 15 samples with the concentrations ranging from 0.1 to 1.6 ng/g dw. In these 15 samples, thirteen ones had the concentrations between 0.1 and 0.3 ng/g dw. TBBPS-MDBPE was detectable in 36 samples with the concentrations ranging from 0.3 to 4.1 ng/g dw, among which 20 ones contained the compound higher than 1.0 ng/g dw. The detection frequencies of these three compounds were in the order of TBBPS-MDBPE (95%) > TBBPS-MBAE (39%) > TBBPS-MAE (5%). As shown in Fig. 3, the mean concentration of TBBPS-MDBPE was higher than that of TBBPS-MBAE, and the concentration of TBBPS-MAE was the lowest. A typical mass chromatogram for the three compounds detected in mollusk sample is shown in Fig. 2(C,D). ## Environmental Risk Prediction of TBBPS Derivatives. The physical-chemical properties of TBBPS, TBBPS derivatives and other well concerned contaminants were calculated by US EPA EPI Suite V4.1, which has been widely employed for screening of potentially persistent and bioaccumulative contaminants . As shown in Table 3, the log K ow values of TBBPS and the derivatives ranged from 5.21 to 9.52, log K oa values ranged from 16.83 to 21.83, log K oc values ranged from 4.16 to 6.33, and log K aw values were all lower than − 8.81. The bioconcentration factor (BCF) values of TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE were 10730, 13200 and 8829, respectively, which were significantly higher than those of TBBPS (1266), TBBPS bis(allyl ether) (TBBPS-BAE) (4207) and TBBPS-BDBPE (775). In addition, the potential toxicity of TBBPS-BDBPE, TBBPS-BAE, TBBPS, TBBPS-MAE, TBBPA-MBAE and TBBPS-MDBPE were also estimated with the primary cerebellum granule cells (CGCs) as the model, which were usually used for neurotoxicity studies 13,27,28 . The IC 50 of TBBPS, TBBPS-MAE, TBBPA-MBAE and TBBPS-MDBPE were 0.45, 0.19, 0.20 and 0.17 μ M, respectively. The IC 50 of TBBPS-BAE and TBBPS-BDBPE were 13.1 and 11.2 μ M. TBBPS and the three derivatives with phenol groups, TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE, inhabited 50% of the cell viability at a much lower concentration than TBBPS-BAE and TBBPS-BDBPE. ## Discussion Since the standards of TBBPS-BDBPE byproducts were not commercially available, TBBPS-MAE, TBBPS-MDBPE and TBBPS-MBAE were self-synthesized and further characterized by 1 HNMR, UHPLC-Orbitrap Fusion HRMS (Thermo Fisher scientific, USA) and HPLC-UV. The results indicated the successful synthesis with high purity (> 96%) of TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE. These compounds could be used as the standards for the further analysis. In order to identify the target compounds in samples, an accurate and sensitive method was developed. By using the highly sensitive Orbitrap Fusion HRMS, the IDLs for TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE were in the range 0.06-0.1 pg which were lower than those acquired with HPLC-ESI-MS/ MS. The target compounds could be identified according to the accurate m/z values of precursor ions within a mass tolerance of 5 ppm. Meanwhile, in the full scan mode, the isotope information was positively observed with the quantification process. As for optimizing the extraction method, the ultrasound method showed slightly higher recoveries and lower SDs which meant it was a stable and reliable extraction method. Meanwhile, the matrix effects were all close to 1.0 which indicated the interference from the matrix could be ignored. ENVI-Carb SPE cartridges could effectively eliminate the interference and concentrate the target compounds. The pretreatment method was reliable and repeatable for the identification of TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE in mollusk samples. Meanwhile, ISDBP was selected as an appropriate internal standard for the recovery monitor of TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE from the mollusk samples. The recoveries of ISDBP were all higher than 80% in all the samples, indicated the recoveries of target compounds from real samples were reliable. With the proposed method, the existence of TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE in technical TBBPS-BDBPE and mollusk samples from the Chinese Bohai Sea was studied in detail. In the sampling area of this work, several BFRs factories produce TBBPS and TBBPS-BDBPE on a large scale. TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE were detectable in technical TBBPS-BDBPE of the BFRs factory with the concentrations ranged from 28 to 394 μ g/g. Consequently, in mollusk samples, TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE were detectable at a level of ng/g dw. Therefore, the BFRs factories might be the point sources of TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE. They were probably produced with the manufacture process, and leaked into the environment through the production and application process. The production process of TBBPS-BDBPE might influence the concentration and detection frequency of the byproducts. TBBPS-BDBPE is synthesized from TBBPS-BAE, and TBBPS-BAE is synthesized from TBBPS. Therefore, TBBPS-MDBPE, which has the most similar structure with TBBPS-BDBPE, is the main byproduct of TBBPS-BDBPE. As a result, TBBPS-MDBPE was detected at the highest concentration and detection frequency. While the technical TBBPS-BAE is produced as intermediate of TBBPS-BDBPE, its structure related byproduct, TBBPS-MAE, showed the lowest concentration and detection frequency. TBBPS-MAE and TBBPS-MBAE were detected in the mollusk samples at the similar concentration level with previously reported for TBBPA derivatives, TBBPA-MAE and TBBPA-MDBPE 3 . TBBPS-MDBPE showed higher concentration level (> 1.0 ng/g dw) than TBBPA-MDBPE (< 1.0 ng/g dw) in the mollusk samples. 3 The detected concentrations of TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE were lower than Tris-(2,3-dibromopropyl) isocyanurate (TBC) 29 , hexabromocyclododecane (HBCD) 29 and polybrominated diphenyl ether (PBDE) 29,30 which were also detected in the mollusk samples with the concentration ranges of below detection limit (nd) to 12.1 ng/g dw, nd to 28.8 ng/g dw and 0.01 to 59 ng/g dw, respectively. The detection frequency of TBBPS-MDBPE (95%) was comparable with that reported for HBCD (99%) and PBDE (100%) 29,30 . The difference of the concentrations of TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE between years was not significant. Interestingly, the detection frequency of TBBPS-MDBPE was 95% which indicated it was probably one widely dispersed brominated compound. Significant difference was not observed among the concentrations of TBBPS-MAE, TBBPS-MBAE and TBBP-MDBPE in different mollusk species. The concentrations of TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE remained at a similar level which indicated these compounds could persistently accumulate in the mollusks. The property of persistent accumulation in the biota system may result in their potential health risks posing on the aquatic ecosystem. Furthermore, the environmental risks of TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE were evaluated. Their physical-chemical properties were calculated by US EPA EPI suite. The log K ow values of TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE were close to TBBPA (7.20) 20 , TBC (7.37) 22 and HBCD (7.74) 29 and higher than 5. These results indicated the accumulation of TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE in organic materials such as fat-rich organisms. The high K oa values of TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE indicated low respiratory elimination rate and high bioaccumulation ability in respiratory organisms. Meanwhile, the low K aw values implied that large amount of TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE would participate in water rather than the air at the boundary exchange process. The comparable K oc values with TBBPA (5.24) and TBC (4.92) indicated TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE could be absorbed by the sediment and soil with a considerable amount. Usually, the chemicals with log K ow > 5 and BCF > 5000 are considered to be bioaccumulative 31 . In addition, the BCF values of TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE were higher than TBBPS-BDBPE, TBBPA derivatives and other environmental contaminants in Table 3. They were more bioaccumulative than TBBPS-BDBPE and TBBPA derivatives. As novel brominated contaminants, the toxicity of TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE were essential for their environment risks assessment. By using the CGCs as a model, the IC 50 of TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE were all lower than 0.2 μ M, suggesting that they were more toxic than TBBPS-BDBPE and TBBPS-BAE with IC 50 higher than 11 μ M (Figure S3). This is probably accused by the phenol group in the structure which probably increased the toxicity of the BFRs 32 . The production of technical products would bring some novel compounds with more severe toxicity into the environment. These byproducts in the environment showed potential health risk to human. Further studies about the toxicity of these byproducts are urgently needed. By using Orbitrap HRMS, we also found several unknown brominated compounds. With the accurate results determined by Orbitrap HRMS, the compound contained Br is easier to be identified because of the special properties: 1) m/z value of the decimal part decreases with the number increase of 81 Br; 2) the isotope ratio is different for compounds contained different number of Br. In this study, some untargeted brominated organics also showed up in the samples. Their spectra are shown in Fig. 4 and the molecular formulas were calculated by Xcalibur software. The three compounds detected in mollusk samples showed the properties of containing bromine atoms. Within delta ppm < 5, the m/z detected at RTs 2.8 min (isotope ratio, 1:3:3:1), 2.6 min (isotope ratio, 4:6:4) and 1.7 min (isotope ratio, 1:2:1) were calculated to be [C 6 H 2 OBr 3 ] − , [C 12 H 5 O 4 Br 4 S] − and [C 6 H 2 O 3 NBr 2 ] − , respectively. They might be three kinds of bromophenols. These three untargeted compounds were further analyzed with Orbitrap Fusion HRMS with reference to the standards, including TBBPS, 2,6-dibromo-4-nitrophenol (DBNP) and four kinds of tribromophenol (Figure S4). The untargeted peak detected at RT 2.64 min (Fig. 4B2) showed similar RT and mass spectra with TBBPS (RT 2.62 min, isotope ratio 1:4:6:4:1). The untargeted peak detected at RT 1.76/1.73 min (Fig. 4A1,B1) had the similar RT and mass spectra with DBNP (RT 1.72 min, isotope ratio 1:3:1). DBNP was also identified as novel bromophenol compounds showing toxicity and potential risk to human . DBNP could formed in the chlorination of drinking water and saline sewage effluent 36 . The untargeted peak detected at RT 2.82 min (Fig. 4A2) presented the similar mass spectra with all the four kinds of tribromophenol and the similar RTs with 2,3,4-tribromophenol, 2,4,6-tribromophenol and 2,4,5-tribromophenol. It might be 2,4,6-tribromophenol as it is one kind of mass-produced BFRs in the sampling area. We did not quantify these untargeted compounds because of the lack of reliable pretreatment method. However, their presence in the mollusk samples could be determined with the quantification of our target compounds by Orbitrap Fusion HRMS. In all the 38 mollusk samples, 7 samples were detected to contain tribromophenol, 8 samples for TBBPS and 17 samples for DBNP. The detection frequencies of these three compounds were all higher than TBBPS-MAE. The anthropogenic activities might result in the emergence of 2,4,6-tribromophenol, DBNP and TBBPS in the environment as they were not reported as natural compounds 37 . The untargeted compounds might also become novel brominated contaminants. In this view, further investigation is needed to be conducted on the identification and environmental fate of these compounds. It is worth mentioning that the Orbitrap Fusion HRMS is a powerful tool for the quantification of novel contaminants and qualitative analysis of unknown contaminants with one injection. Most BFRs, such as poly brominated diphenyl ether (PBDE), TBBPA and TBBPS derivatives, usually share the ether bond linked structure 6 . For the production of ether bond derived organic aromatic chemicals with several bromine atoms, the left-over starting reagents, co-produced phenol and less brominated byproducts could be potential environmental contaminants together with the desired BFRs products. The byproducts generated from manufacture production or degradation draw great attention because they were found in various environment compartments as novel or emerging BFRs . For example, the byproducts of TBBPA and its derivatives, TBBPA-MAE, TBBPA-MDBPE, TBBPA mono(2-hydroxylethyl ether), TBBPA mono(glycidyl ether), dibromobisphenol A and tribromobisphenol A have been determined in water, soil and biota system 2,3,42 . In this work, we found the manufacture process of TBBPS-BDBPE resulted in the occurrence of TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE inevitably. The existence of the phenol byproducts in the aryl-O linked technical products might be a global problem. Although aryl-O bond of organic chemicals is considered very stable in chemical reactions, its cleavage is easy to fulfill under the bacterial biodegradation, the UV irradiation and super-reduced conditions . TBBPA bis(2,3-dibromopropyl ether) (TBBPA-BDBPE) was also found to transform to TBBPA via ether breakage in aquatic mesocosm 46 . The compounds with ether bond are not as stable as suppositional under environmental conditions. TBBPA-MAE and TBBPA-MDBPE were also predicted to be the degradation products of TBBPA bis(allyl ether) (TBBPA-BAE) and TBBPA-BDBPE by the University of Minnesota Pathway Prediction System 2,3 . Through the same microbial transformation, TBBPS derivatives showed the potential ability of ether bond cleavage and form the mono-modified degradation products, TBBPS-MAE and TBBPS-MDBPE. The co-produced byproducts in manufacture process and microbial degradation in the environment contribute to the occurrence of mono-modified byproducts in the environment. The study about the byproducts and degradation products of these ether linked BFRs will supplement the information for novel brominated contaminants. In conclusion, TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE were identified as three novel brominated contaminants, which showed higher bioaccumulation properties and potential severe toxicity compared with TBBPS-BDBPE. They could be co-produced and leaked into the environment along with production process of TBBPS-BDBPE. The occurrence of the mono-modified byproducts or degradation products of the extensively used brominated products might be a widespread problem. This work could promote the further study of the environmental fate and risks of widely used TBBPS and TBBPS-BDBPE. The strategy used in this work, integrating the synthesis of standards and Orbitrap HRMS identification, could also be an effective approach for identifying other novel brominated pollutants related to BFRs. ## Methods Chemicals and Materials. TBBPS (98%) was purchased from Beijing Apisi biotechnology co. ltd., and was used without further purification. Ammonium hydroxide (50%) was purchased from Sigma-Aldrich. Methanol, acetone, hexane and methylene dichloride (DCM) were all HPLC grade. Ultra-pure water was generated by a Milli-Q advantage A10 system. TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE were synthesized and purified in our lab. The synthesis procedures were described in Supporting Information. The purities of these three compounds were 99%, 98% and 96% as determined by HPLC-UV (214 nm). ## Sample Collection. From 2009 to 2013, in August of each year, 11 species of mollusks were collected from one coastal city -Shouguang, Shandong Province. These 11 selected species of mollusks were Rapana venosa (large and small, RapL and RapS), Crassostrea talienwhanensis (Ost), Scapharca subcrenata (Sca), Cyclina sinensis (Cyc), Mya arenaria (Mya), Mactra veneriformis (Mac), Chlamys farreri (Chl), Neverita didyma (Nev) and Meretix meretrix (large and small, MerL and MerS) (Fig. 3, Table S1). After sampling, the mollusks were frozen and transported on ice to the laboratory, and then cleaned by water. The collected samples were disposed according to the previous method 29 . The samples were freeze-dried, grinded, and preserved at − 20 °C until analysis. A total of 38 mollusk samples were obtained and analyzed. Sample Pretreatment. Ultrasound Extraction. Mollusk (0.5 g) samples were mixed with 2 g anhydrous Na 2 SO 4 ; spiked with 10 ng 13 C labeled 3,5-dibromophenol (ISDBP); extracted with 10 mL DCM/ hexane (8/2, V/V) for three times by sonication (30 minutes per time). After centrifugation, the extraction solution was collected and the solvent was removed with rotary evaporator and re-dissolved in 3 mL DCM/hexane (1/1, V/V) before SPE process. SPE Procedures. The SPE cartridges (Supelclean TM ENVI-Carb TM , 0.5 g, 6 mL) were first conditioned by 5 mL acetone, 5 mL DCM and 10 mL hexane and then the samples were loaded. Then the cartridges were cleaned by 5 mL hexane and 5 mL DCM/hexane (1/1, V/V). Finally, the cartridges were eluted with 10 mL acetone (containing 0.5% NH 3 •H 2 O) and the elution were collected and blown to dryness by gentle nitrogen gas flow. The residue was solvent-exchanged to 1 mL methanol and analyzed by UHPLC-Orbitrap Fusion HRMS. Instrument Parameters for UHPLC-Orbitrap Fusion HRMS, HPLC-UV and HPLC-ESI-MS/MS Analysis. The details were described in Supporting Information. ## Analytical Method Validation. TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE in samples were identified by retention time and accurate m/z of the precursor ions comparison with the corresponding standards. Quantification of the target compounds in the environmental samples was performed by peak area of the accurate precursor ions of compounds within 5 ppm mass tolerance. For TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE, precursor ions at m/z 604.69189, 682.60248 and 764.52643 were used as quantification ions, and m/z 602.69415, 684.60022 and 762.52856 were used as the qualitative ions. All calibration standards and spiking solutions were prepared by serial dilution in methanol. A linear calibration curve with seven points ranged from 0.05 to 100 ng/mL was used to quantify the target compounds with a determination coefficient (R 2 ) higher than 0.99 (Table 2). The concentrations of target compounds were determined by an external standard method. Every 9 samples were prepared with one blank sample (only anhydrous Na 2 SO 4 added), and analyzed with methanol as solvent blank to make sure no cross contamination. The DLs were determined by the lowest mass value of the target compounds that Orbitrap Fusion HRMS detected. The IDLs were determined for five times within 20% relative standard deviation for the signals. The MDLs were based on replicate analysis (n = 10) of Nev sample spiked at a mass concentration of 5 times of the IDLs and calculated with the method previously used for HRMS 47 . The recoveries were determined at the spiking amounts of 1 ng (n = 5), 10 ng (n = 7) and 100 ng (n = 5) in 0.5 g Nev samples (not containing target compounds). The internal standard 13 C labeled 3,5-dibromophenol (ISDBP) was used to monitor the pretreatment process and not used for the concentration calculation. The detailed procedures for the SPE optimization and results were described in Supporting Information. The matrix effects were determined according the method reported elsewhere previously 43,48 . Detailed information regarding the synthesis routines and 1 HNMR and MS 2 spectra of TBBPS-MAE, TBBPS-MBAE and TBBPS-MDBPE, the 1 HNMR data, instrumental analysis information of Orbitrap Fusion HRMS, HPLC-UV and HPLC-ESI-MS/MS, the optimization of pretreatment method, the concentration of every mollusk sample, the cell information and cytotoxicity test method, the cytotoxicity of TBBPS and its derivatives, the HRMS chromatograms and spectra of different brominated phenols are provided in the Supporting Information.

chemsum

{"title": "Characterization of Three Tetrabromobisphenol-S Derivatives in Mollusks from Chinese Bohai Sea: A Strategy for Novel Brominated Contaminants Identification", "journal": "Scientific Reports - Nature"}

direct_observation_of_chain_lengths_and_conformations_in_oligofluorene_distributions_from_controlled

## Abstract: Synthetic polymers are mixtures of different length chains, and their chain length and chain conformation is often experimentally characterized by ensemble averages. We demonstrate that Double-Electron-Electron-Resonance (DEER) spectroscopy can reveal the chain length distribution, and chain conformation and flexibility of the individual n-mers in oligo-(9,9-dioctylfluorene) from controlled Suzuki-Miyaura Coupling Polymerization (cSMCP). The required spin-labeled chain ends were introduced efficiently via a TEMPO-substituted initiator and chain terminating agent, respectively, with an in situ catalyst system. Individual precise chain length oligomers as reference materials were obtained by a stepwise approach. Chain length distribution, chain conformation and flexibility can also be accessed within poly(fluorene) nanoparticles. ## Introduction Knowledge of the chain length distribution and of chain conformations are essential to understand and design synthetic procedures, materials properties, and nanoscale structures. 1,2 Consequently numerous methods to access these parameters have been developed to a high level and for practical use. The most prominent and ubiquitious method to access molecular weight distributions is size exclusion chromatography. More recently, mass spectrometry has also been advanced considerably for the lower molecular weight regime, though it is not quantitative. These methods are based on separation of the sample, prior to detection of its individual chain length components. Likewise, an experimental determination of chain conformations on mixtures of different length chains yields an ensemble average only. We now report double-electron-electron-resonance (DEER) 3,4 studies of spin-labeled oligofluorene mixtures that give direct access to the chain length distributions and monitor conformational ensembles and flexibilities 5 of the oligomers. As a probe we chose oligofluorenes from controlled Suzuki-Miyaura coupling polymerization (cSMCP). 6,7 Oligofluorenes are attractive materials due to their photo-and electroluminescence and light-induced charge generation. 8,9 cSMCP proceeds in a chain growth fashion. 10 The controlled character of cSMCP allows for an introduction of functional endgroups at both chain ends. cSMCP has been demonstrated for the synthesis of a scope of polyarylenes, like poly(fluorenes), poly(thiophenes) and poly(phenylenes). 7,13 DEER is a pulsed electron paramagnetic resonance (EPR) method to determine distance distributions between paramagnetic centers. Nitroxide groups, such as 2,2,6,6-tetramethylpiperidinyloxyl (TEMPO) are commonly used as spin labels. DEER is established in biological chemistry where it is commonly used to determine distances between a pair of spins placed in a single defined type of molecule, like a protein. 14,15 In the same way, DEER has also been applied to defined monodisperse synthetic molecules. 16,17 ## Results and Discussion Spin labels were attached at the chain ends of poly(9,9dioctylfluorene) (PF8) directly during polymerization by employing spin-labeled initiators and terminating agents in an otherwise established controlled SMCP protocol. 10,11,18 The polymerization was initiated by an in situ system using chloro[(tri-tert-butylphosphine)-2-(amino-biphenyl)] palladium(II) as Pd(0) source, and TEMPO-labeled 4-bromobenzoic acid which adds oxidatively. Chains were quenched by addition of TEMPO-labeled 4-carboxyphenyl-boronic acid pinacol ester end capping agent, resulting in identical chain termini (Figure 1). The successful incorporation of the TEMPO labels was confirmed by MALDI-TOF MS, identi-fying the doubly labeled polymer chains as major component by the isotopic pattern (cf. Figure S 4.1). As expected, the maximum of the obtained chain length distribution determined by GPC could be adjusted by the ratio of initiator to monomer in the reaction mixture. Thus, the described in situ system offers a robust and versatile method for the synthesis of doubly TEMPO labeled conjugated polymers in a one-step approach. For establishing a chain conformation analysis, doubly labeled monodisperse oligomers with a precise number of repeat units (herein referred to as doubly labeled precise oligomers) were required as a reference. According to reported procedures, the non-labeled oligo(fluorenes) were built up stepwise by repetitive cycles of alternating Suzuki-Miyaura coupling and bromination, starting from the mono-and dibromo-substituted fluorene monomers (Figure S 3.2). Spin labels were attached to the oligomers by a final Suzuki-Miyaura coupling step with a TEMPO-substituted monofunctional arylboronic acid ester. These doubly labeled precise oligomers from stepwise synthesis are abbreviated as Pn, with n representing the chain length in the following. Oligomers up to n = 5 were synthesized. As the synthetic effort of the step-by-step approach considerably increases for each additional monomer unit, reference oligomers with n = 5 and 6, abbreviated as P5´ and P6´, were provided by semi-preparative GPC fractionation (cf. S 3.2). As reference, DEER measurements in Q-band for the precise oligomers were performed upon shock-freezing in toluene-d8 (Figure 2A). The experimental data for all individual oligomers Pn (with n = 1-4), P5´ and P6´ was fitted (Figure 2A) with the worm-like chain model (WLC), 17,22,23 using DEERAnalysis. Assuming that the monomer P1 with just one repeat unit is completely stiff, the width of the respective distance distribution can be accounted to the flexibility of the spin-label end group. It can be described by a Gaussian broadening 16,17 with σ = 0.06 nm. For all further experiments, we used this Gaussian broadening to take the flexibility of the spin-label end groups into account. The persistence length Lp is a global parameter for describing the set of DEER data, the contour length Lc was fitted for each oligomer individually. The obtained distance distributions are shown in Figure 2B, the fit parameters are listed in Table 1. ## Table 1. Oligomers and parameters of model-based fit with WLC. Global parameters: Lp = 14 nm and σ = 0.06 nm. For P4 an additional gauss distance was introduced (r = 3.03 nm, σ = 0.267 nm, 21 % weighting factor). ## Oligomer Lc As an important detail, for P4 we observed occurrence of three-fold bromination (21 %) in the course of the stepwise oligomer synthesis (see SI Figure S4.2) and attachment of a third spin-label in the center of the chain. This results in a broad additional distance contribution below the expected end-to-end distance, which is described by an additional Gaussian in the model (Table 1). The set of doubly labeled oligomers can be described with Lp = 14 nm and Lc is approx. (2.1 + n • 0.8) nm. This is in good agreement, with a monomer length of 0.75 nm, found for poly(9,9-bis(2-ethylhexyl)fluorene-2,7-diyl) (PF2/6). 27 With the WLC parameters for every oligomer in hand, DEER was applied to an oligomer mixture (Mix) obtained from controlled Suzuki-Miyaura polymerization (Figure 3A). A monomer to initiator species of five had been applied in the polymerization experiment. The accessible range in the distance distribution detected by DEER is restricted by the dipolar evolution time, which was limited to 10 μs in our experiments due to spin-spin relaxation. Under these conditions, distance contributions and their width can be detected with an upper limit of 7 nm. The limit for accurate determination of the shape of the distribution is 5 nm. 4 To describe the experimental data, we used a model containing a superposition of different oligomers each described by WLC using the parameters obtained for the respected monodisperse reference oligomers (P1-4, P5´, P6´) according to a Poisson distribution with expectation value λ. As parameters for reference oligomers for n > 6 were not derived individually, we fitted the contour length of an oligomer with n = 7 as an additional parameter in the analysis. Note that the polymerization mixture contains oligomers with n > 7, which are beyond the detection limit in our experiment. This does not disturb the analysis of the data for the oligomers amenable to DEER observation, however. To illustrate the amount of higher oligomers not observed by DEER in the overall distributions, these were calculated according to the found Poisson distribution (Figure 3B, dotted line). In summary, we find that the experimental data is in excellent agreement with i) a Poisson distribution for the chain lengths, with an expectation value λ = 5.4 and ii) a chain conformation described by a WLC with a persistence length of 14 nm. Even when determining the fractions for each chain length individually in a model free approach, we find a reasonable agreement with a Poisson distribution (see SI S9). The agreement of the experimental DEER data with a Poisson distribution also allow for additional mechanistic conclusions on the polymerization reaction. Initiation of chains occurs very efficiently. That is, the initiating Pd-species is formed rapidly and completely in the early stages of the reaction, and it starts growth of a chain efficiently. In many instances, solid polyfluorene materials are of interest rather than solutions. An access to chain conformations and chain flexibility in e.g. films or nanoparticles are desirable as they are instrumental in determining, for example, particle shapes. To demonstrate the principal suitability of the method reported also for nanoparticles, we prepared spherical poly(fluorene) nanoparticles (NP) by emulsification. The bulk polymer was blended with P1, P2 and P3 (4:3:3 doubly labeled species, 0. Prior to DEER measurements, the particles were purified by dialysis for the removal of excessive surfactant and freeze-dried. DEER measurements within the nanoparticle were performed with a dipolar evolution time of 3 µs due to increased spin-spin relaxation in the protonated environment of the nanoparticle. Tm = 1.9 µs compared to 8.0 μs in the deuterated toluene-d8 matrix. The experimental data can be described by using contour length and broadening due to flexibility of the spin-label end group as derived in solution and fitting the weights for P1, P2 and P3 as well as the global persistence length for the oligomers. Figure 4c shows the corresponding distance distribution for spin-labeled oligomers incorporated in nanoparticles. The found fractions of 32 %, 28 % and 40 % are in agreement with the expectations from sample preparation (see above). We derived a persistence length of Lp = 14 nm as found in solution. This suggests a worm-like chain nature with undisturbed flexibility being retained in the nanoparticles. ## Conclusion In conclusion, we have demonstrated the application of DEER distance measurements to a real-life synthetic polymer containing a multitude of different chain length species. The required spin-labeled chain ends could be introduced efficiently by controlled Suzuki-Miyaura coupling Polymerization (sSMCP), cSMCP in general being a state-of-the-art versatile protocol for the synthesis of numerous poly(arylene)s. The DEER data agrees with a Poisson distribution, expected for the case of a living polymerization with fast and efficient initiation, which is given here. The method allows for the quantification of the individual n-mer populations and their respective conformations and flexibilities directly on mixtures. Further, we demonstrated that DEER allows a quantitative analysis of the oligomer fractions as well as characterization of conformation and flexibility inside nanoparticles in principle. The necessity of spin-labeled oligomers is clearly a limitation of the analysis method reported. Thus, it is rather complementary than competitive to established standard methods for chain length distribution and conformation analysis. Its strength is access to properties of the individual n-mers directly on mixtures. ## ASSOCIATED CONTENT Supporting Information. Experimental procedures and characterization data; EPR measurements; Evaluation of DEER experiments.

chemsum

{"title": "Direct Observation of Chain Lengths and Conformations in Oligofluorene Distributions from Controlled Polymerization by Double Electron-Electron Resonance", "journal": "ChemRxiv"}

boronic_acid–dmapo_cooperative_catalysis_for_dehydrative_condensation_between_carboxylic_acids_and_a

## Abstract: Arylboronic acid and 4-(N,N-dimethylamino)pyridine N-oxide (DMAPO) cooperatively catalyse the dehydrative condensation reaction between carboxylic acids and amines to give the corresponding amides under azeotropic reflux conditions. This cooperative use is much more effective than their individual use as catalysts, and chemoselectively promotes the amide condensation of (poly)conjugated carboxylic acids. The present method is practical and scalable, and has been applied to the synthesis of sitagliptin and a drug candidate. ## Introduction The catalytic dehydrative condensation reaction between carboxylic acids and amines is one of the most ideal methods for synthesizing the corresponding amides. 1 In 1996, Yamamoto et al. reported the frst example of the dehydrative amide condensation reaction catalysed by meta-or para-electron-defcient group-substituted phenylboronic acids such as 3,4,5-tri-fluorophenylboronic acid (1) (pK a ¼ 6.8) 2a and 3,5bis(trifluoromethyl)phenylboronic acid (2) (pK a ¼ 7.2) 2b under azeotropic reflux conditions (Scheme 1). 3 These boronic acids are more acidic than phenylboronic acid (pK a ¼ 8.8, 8.9). 2 In 2006, Whiting et al. reported that ortho-Brønsted basesubstituted phenylboronic acids such as 2-(N,N-diisopropylaminomethyl)phenylboronic acid (3) were effective catalysts for the amide condensation of aromatic carboxylic acids under the same conditions as above. 1a,4 In 2008 and 2012, Hall et al. reported that 2-iodophenylboronic acid (4a) and 2-iodo-5methoxyphenylboronic acid (4b) were also effective catalysts for the amide condensation in the presence of drying agents (activated 4 molecular sieves) at lower temperature. 5 The o-iodo group of 4a and 4b assists the catalysis of amide condensation as a weak base. 6 In addition to these boronic acids, boric acid, 7a,c benzo dioxaborol-2-ol, 7b methylboronic acid, 7d and some o-Brønsted base-substituted boronic acids 8 have been reported to be useful as amidation catalysts. However, the substrate scope is still quite limited. For example, harsh conditions (higher temperature, prolonged reaction time, excess amounts of substrates, increased amounts of catalysts, etc.) are required for sterically hindered a-branched carboxylic acids and conjugated carboxylic acids. In 2013, Whiting et al. discovered an interesting synergistic catalytic effect between o-tolylboronic acid (50 mol%) and o-nitrophenylboronic acid (50 mol%) in dipeptide synthesis. 3f To the best of our knowledge, this was the frst example of two cooperative promoters for direct amidation. 3f,9 In the process catalysed by arylboronic acid, a mixed anhydride intermediate 5 is generated from the carboxylic acid and arylboronic acid under azeotropic reflux conditions or in the presence of drying agents in the frst stage (Schemes 1 and 2). This is the frst activation of the carboxylic acid. a tetrahedral intermediate 6, the amide condensation may proceed more rapidly. However, if Nu preferentially coordinates as a Lewis base to the boron atom of 5, a less active species 8 is generated and the amide condensation may be suppressed. Here we report that arylboronic acids and N,N-dimethylaminopyridine N-oxide (DMAPO) cooperatively promote the dehydrative condensation between various carboxylic acids and amines. ## Results and discussion First, the amide condensation reaction between 2-phenylbutyric acid and benzylamine was examined in the presence of 5 mol% each of boronic acid 2 and a nucleophilic additive under azeotropic reflux conditions in fluorobenzene (bp. 85 C) 3f for 17 h (Table 1). Boronic acid 2 did not promote the reaction in the absence of additive under these conditions (entry 1). Tertiary amines such as N,N-diisopropylethylamine and 4-(N,N-dimethylamino)pyridine (DMAP) 11 were not effective as additives (entries 2 and 3). 4-Methoxypyridine N-oxide (MPO) was also less active (entry 4). In contrast, a more nucleophilic but weak base, DMAPO, 12 was quite effective for the amide condensation (entry 5). However, a more nucleophilic additive such as 4-(pyrrolidin-1-yl)pyridine N-oxide (PPYO) was less effective than DMAPO (entry 6), perhaps because the strong nucleophilicity of PPYO might reduce the activity of 7. Next, the cooperative effects of several boronic acids (5 mol%) were compared in the condensation reaction between 2-phenylbutyric acid or benzoic acid and benzylamine in the presence of DMAPO (5 mol%) (Table 2). These less reactive carboxylic acids were not activated by the individual use of boronic acids under the same conditions. As expected, 2-DMAPO and 4b-DMAPO efficiently activated 2-phenylbutyric acid (entries 1 and 3). Phenylboronic acid and 3 were almost inert, even in the presence of DMAPO (entries 2 and 4). Interestingly, 2-DMAPO was more effective than 4b-DMAPO for the amide condensation of benzoic acid (entries 1 and 3). While Whiting's catalyst 3 was quite effective for the amide condensation of benzoic acid, the catalytic activity of 3 was suppressed in the presence of DMAPO (entry 4). 13 To explore the substrate scope using the cooperative catalysts, 2-DMAPO, the amide condensation reactions of several less reactive a-branched carboxylic acids and arenecarboxylic acids were examined under azeotropic reflux conditions in fluorobenzene (bp. 85 C) or toluene (bp. 110 C). As shown in Table 3, in each example, the cooperative catalysts were much more effective than 2 alone, the results for which are shown in brackets. Notably, not only aliphatic primary amines but also sterically hindered aliphatic secondary amines, less nucleophilic anilines and alkoxyamines reacted with these carboxylic acids. In particular, 2-DMAPO was effective in the amidation of arenecarboxylic acids with sterically hindered amines, in comparison with 3 and 4b (entries 9-14). This cooperative method is scalable to practical volumes: the catalytic loading of 2-DMAPO could be reduced to 2.5 mol% for the dehydrative condensation on an 80 mmol scale (entry 4). a A solution of 2-phenylbutyric acid (0.5 mmol) and benzylamine (0.5 mmol) in fluorobenzene was heated in the presence of 2 (5 mol%) and additive (0 or 5 mol%) under azeotropic reflux conditions. b Isolated yield. The boronic acid-catalysed condensation of relatively more reactive a-nonbranched carboxylic acids with sterically hindered secondary amines and less nucleophilic anilines proceeded even in the absence of DMAPO, as shown in brackets in Table 4. Nevertheless, the addition of DMAPO was also quite effective for these reactions. Interestingly, 4b and phenylboronic acid were slightly more reactive than 2 in the presence of DMAPO. In particular, the utility of inexpensive phenylboronic acid is industrially signifcant. This catalytic method is readily scalable. 2.5 g of N-Boc protected sitagliptin, 14 an anti-diabetic drug, was obtained by carrying out the condensation on a 5 mmol scale (entry 4). The results in Tables 1-4 suggest that both the nucleophilicity of the additive and the Lewis acidity and steric effect of the boronic acid are important in the cooperative catalysis with an ArB(OH) 2 -nucleophilic base (Table 5). The reactivity from highest to lowest followed the order arenecarboxylic acids, a-branched carboxylic acids, a-nonbranched carboxylic acids. As a result, 2 was more effective for arenecarboxylic acids and a-branched carboxylic acids. On the other hand, 4b and phenylboronic acid were more effective for a-nonbranched carboxylic acids. The amide condensation reaction should occur through the active intermediate 6 (Scheme 2). However, not only 6 but also a Unless noted otherwise, 0.5 mmol of carboxylic acid and 0.5 mmol of amine were used in the presence of 5 mol% of 2 and 0 or 5 mol% of DMAPO. b The results when both catalysts were used are shown. For comparison, the results without DMAPO are shown in brackets. c 10 mol% of each of the catalysts was used. d 2.5 mol% of each of 2 and DMAPO was used on an 80 mmol scale in 70 mL of toluene. e 15 mol% of each of the catalysts was used. f 99% ee. g 3 was used. h 4b was used. a Unless noted otherwise, 0.55 mmol of carboxylic acid and 0.50 mmol of amine were used in the presence of 5 mol% of ArB(OH) 2 and 0 or 5 mol% of DMAPO. b The results when both catalysts were used are shown. For comparison, the results without DMAPO are shown in brackets. c 10 mol% of each of the catalysts was used. d The reaction was carried out at a 5 mmol scale. Table 5 Relationship between the cooperative effects of boronic acid-DMAPO and the reactivity of carboxylic acids the undesired complex 8 would be generated in an equilibrium mixture. Complex 8 might be converted to the more stable complex 9, which is inert to the amide condensation. In fact, the generation of inert complex 9 was ascertained by 11 B and 1 H NMR analysis in the amidation of less-hindered carboxylic acids. 15 Also, the chemical structure of the cyclic complex prepared from 2, phthalic acid, and DMAPO was determined to be that of 9z by X-ray diffraction analysis (Fig. 1). 16 For sterically hindered carboxylic acids such as arenecarboxylic acids and a-branched carboxylic acids, the desired intermediate 7 was preferentially generated. Thus, o-nonsubstituted and m-or p-electron-defcient group-substituted phenylboronic acids such as 1 and 2 were more suitable. In contrast, for less sterically hindered a-nonbranched carboxylic acids, the undesired complex 9 was generated more easily. In addition, the strong Lewis acidity of 2 helped to stabilize 9 by the tight coordination of DMAPO to the boron centre. This is why 4b and phenylboronic acid were slightly more effective than 2 for the condensation of a-nonbranched carboxylic acids. Not only Lewis acidity, but also the bulkiness of the o-substituent of the boronic acid might suppress the generation and stability of 9. It is noted that the effect of DMAPO was not striking at ambient temperature. Heating was required to accelerate the equilibrium between 6 and 8. The utility of the cooperative catalysts was also demonstrated for the selective amide condensation of b-substituted acrylic acids to give the corresponding amides 10 (Table 6). The production of Michael adducts 11 was fairly minimal. In contrast, when boronic acids were used in the absence of DMAPO, the yield and selectivity of the reaction for 10 were moderate. Control experiments ascertained that 10 (n ¼ 1) was selectively obtained from 13, 17 and 11 (n ¼ 1) was not generated from 10 (n ¼ 1) but 14. Amide 10c is known to be a potential antimitotic agent, especially for brain cancers (entry 6). 18 The cooperative catalysts were effective for the selective amide condensation of not only b-substituted acrylic acids, but also polyconjugated carboxylic acids and but-2-ynoic acid (entries 12-17). ## Conclusions In conclusion, this new cooperative catalytic system is quite effective for the amidation reaction of less reactive carboxylic acids, such as sterically hindered a-branched carboxylic acids and arenecarboxylic acids, and the chemoselective amidation reaction of conjugated carboxylic acids. Based on the NMR spectra and X-ray diffraction analysis of inert species 9, a preliminary mechanism was proposed. Further mechanistic studies are in progress. We believe that these fndings will trigger the further development of high-performance amidation catalysts.

chemsum

{"title": "Boronic acid\u2013DMAPO cooperative catalysis for dehydrative condensation between carboxylic acids and amines", "journal": "Royal Society of Chemistry (RSC)"}

identifying_therapeutic_compounds_targeting_rna-dependent-rna-polymerase_of_sars-cov-2

## Abstract: COVID-19 has emerged as the biggest threat of this century for mankind. This contagious disease was initially transmitted from animals (probably bats or pangolins) to humans and later it spread across the globe through human to human transmission. Scientists rushed to understand the structure and mechanism of the virus so that antiviral drugs or vaccines to control this disease can be developed. A key to stop the progression of the disease is to inhibit the replication mechanism of Sars-Cov-2. RNAdependent-RNA polymerase protein also called RdRp protein is the engine of Sars-Cov-2 that replicates the virus using viral RNA when it gains entry into the human cell. The replication of the virus is the main process that acts as a catalyst in the progression of disease. RdRp is the main target of researchers working to develop antiviral drugs to inhibit the mechanism of the virus. Numerous drugs proposed for the treatment of COVID-19 such as Camostat Mesylate, Remdesivir, Famotidine, Hesperidin, etc. are under trial to analyze the aftermath of their medicinal use. Nature is enriched with compounds that have antiviral activities and can potentially play a pivotal role to inhibit this virus. This study focuses on the phytochemicals that have the potential to exhibit antiviral activities. A large number of compounds were screened and a cohort of most suitable ones are suggested via in-silico evidence that can inhibit the functionality of RdRp and hence the replication of Sars-Cov-2. ## Introduction: Coronavirus is a vast family of viruses. 7 known coronaviruses can enter into human cells. The first case of coronavirus in humans was reported in 1965, which had mild symptoms of flu and fever. Coronaviruses are significant pathogens for both humans and animals. These are medium-sized but can have a very large RNA genome. They can bind with the host cells and mutate when they transfer from one species to another. Subsequent mutation can lead to its transmission into humans. They can bind themselves to the respiratory tract causing an infection. The symptoms of coronavirus infection are: illness, flu, mild fever, diarrhea, and difficulty in breathing. Severe Acute Respiratory Syndrome named as Sars is an infectious disease caused by Sars-CoV that spreads swiftly and causes illness and flu at the initial stage. Sars-Cov-2 is just like Sars-Cov in its working and structure but more dangerous in terms of severity. It spreads from person to person through coughing or sneezing droplets and physical contact. In 2019, Sars-CoV-2 emerged from Wuhan, China, and took the world by storm. The world was not prepared for it and as a result, both humans and the world economy have suffered very adversely. At this point, over 9.5 million infected cases have been reported and the death toll has reached over 480,000. The onset of COVID-19 has led to a drastic reduction in social and economic activities throughout the world. At this point, doctors and researchers from every country are trying hard to devise an effective strategy for controlling the disease. To propose an effective and long-lasting solution, understanding of the structure of the virus and its action is very important. Recent studies have been able to develop an understanding of the mechanism and structure of the virus through 3D modeling. ## Structure of SARS-COV-2: To get an insight into the action of Sars-Cov-2 viruses and discover suitable antiviral compounds, it is very important to elucidate the proteomic buildup of Sars-CoV-2. Its proteomic data encompasses different proteins that form its makeup such as the Spike and RNA dependent RNA polymerase (RdRp) proteins. The entire Genome RNA structure inside the coronavirus is nearly 30000 bases long. As a whole, it contains 4 proteins that form the viral envelope which are the Spike protein, E protein, Hemagglutinin (M) protein, and N protein. It is very important to understand how Sars-Cov-2 gains entry into the human cells. Angiotensinogen is a hormone found inside the liver which is also found in kidneys and different segments of the brain. This hormone is responsible for managing blood pressure. Angiotensinogen is converted into Angiotensin 1 also named AT-I, by an enzyme produced by a kidney called Renin. In the next step, this AT-I is converted into Angiotensin 2 which is named AT-II by an enzyme called ACE which is produced in the lungs. AT-II is a vasoconstrictor that means it narrows the blood vessels, as a result, aldosterone is produced which causes an increase in blood pressure. AT-II creates two states in the body, one is a low state in which ACE2 binds with Angiotensin Receptor I also named ATR-1 on the surface of the membrane. As a result of this ACE2-ATR-I binding, ACE2 creates Angiotensin 17 (AT-17) which is responsible for vasodilation and decreases inflammation which is good for the human body. The second state is called high state in which due to the high level of AT-II, it does not allow to bind ATR-I with the sites of ACE2 resulting in a gap on the site of ACE2. Because of this gap, Spike protein at the surface of Sars-Cov-2 finds sites to attack, it binds with the sites of ACE2 where ATR-I did not bind and Sars-Cov-2 anchors itself to an entry point into the human cell. It is worth mentioning here that Spike binds with Human Ace with an affinity of -21 kcal/mol, if the spike is to be targeted, ligand must have a binding affinity of more than -21 kcal/mol, which makes it almost impossible to find such a ligand which could bind with Spike protein with a higher binding affinity . This leads to the conclusion that finding ligands that target Spike proteins may not prove fruitful. After entry, the virus needs to replicate itself so that it can propagate itself within its host cells. The RdRp protein plays a pivotal role during this replication process. RdRp is the most significant gene in the virus genome which is encoded inside the RNA of the virus, it speeds up the process of RNA replication from the RNA template and provides safe passage to the virus that is just entered into human cells. Endoplasmic reticulum is a system of membrane that performs multiple functions i.e. Modification, folding, and transfer of proteins. After entering into the human cell, the virus contacts this system and persuades the development of a double-membrane vesicle by developing a complex with it. It generates a copy of genomic RNA. Further, it converts this Negative RNA to positive RNA which makes it mRNA. But this mRNA cannot replicate by itself and translate into a protein. The virus exploits the ribosome machinery of the human cell. The ribosome is tricked into working for the virus and translates the mRNA, creating viral proteins in thousands in each replication cycle. These viral proteins are received by the Golgi apparatus which pack them into vesicles and later send to different destinations. In this way, the whole protein creation apparatus of a human cell is used by the virus for its multiplication. Below figure 2 is the illustration of virus attachment and replication mechanism. As depicted in figure 3 below, RdRp which is also Non-Structural Protein 12 (nsp12) illustrated in the complex with two small proteins nsp7 and nsp8 and has right-hand cup structure with palm subdomain, thumb subdomain, and fingers subdomain. Table 1 shows the range of residues that cover palm, fingers, and thumb subdomains in the structure of RNA-Dependent-RNA-Polymerase. ## Subdomain Residue Range Palm T582-P620 and T680-Q815 Fingers L366-A581 and K621 Thumb H816-E920 The Amino acid sequence of the Sars-Cov-2 genome in many respects resembles Sars-Cov that caused the SARS outbreak in 2002-2003. One strategy to inhibit the progression of the disease is to find ligands that target the RdRp protein. Antiviral drugs that can considerably compromise the function of RdRp protein will be able to suppress the viral multiplication and hence disease progression. Researchers are working to discover an antiviral drug that targets its key residues by splitting the strands of RNA that cause replications for the virus. In this way, its replication and connection with the virus could be inhibited compromising its proper functioning. Table 2 shows different Motifs and the residues that cover each motif along with the type of residues. . A drug therapy that targets these residues of RdRp protein will be able to produce an antiviral effect by inhibiting its function. Moreover, clinically proven drugs like Remdesivir binds to THR 680, SER 682, and VAL 557, pp-sofosbuvir binds to ASN 691, ARG 555, and ASP 623 . Binding details of both drugs will give an insight to discover potential compounds that can cover these binding residues as well as remaining residues that are not covered by Remdesivir and pp-sofosbuvir. Phytochemicals are naturally occurring substances that can contain antiviral and antibiotic properties proving effective for the treatment of diseases. Several plants have therapeutic compounds for example compounds of Artemisia can inhibit tumor growth inside the body and can be used as anticancer substances. Compounds of Azadirachta are used as an antibacterial and for the treatment of skin diseases, stomach upsets, diabetes, fever, and different eye diseases . Compounds of aconitum heterophyllum are antibacterial, antiviral, and anti-inflammatory, these are used for fever, flu, cough, upper tract respiratory diseases, and malaria. Many natural plants and herbs contain substances that have been used as antiviral, antioxidant, and antibacterial purposes for centuries. A large set of plants containing flavonoids, alkaloids, Vitamin C, Sennosides etc. have antiviral properties that can be effective for the treatment of disease. There are thousands of phytochemicals and natural substances whose structures are openly accessible in databases like PubChem, RCSB, chEMBL, and ChemSpider. Medicinal trials on these substances for a specific ailment can consume huge effort in terms of time and money and still required results may not be achieved. In-silico simulation techniques can considerably narrow down on the number of relevant substances through very accurate and meticulous modeling. These methods are capable of providing an insight into the compound structure, analyze physical and chemical properties, and predict the suitability of compounds against target receptors. ## Related Work: Different in silico methods have been used recently to simulate interactions and to evaluate the suitability of drugs for a specific disease. ModeBase was used to create the 3D model of Spike Protein to exhibit the binding of Angiotensin-Converting-Enzyme and Spike Protein. Docking was done by using different virtual screening methods through software named Schrodinger. Grid Generator tool was used to create a grid . To analyze correlation, Claudia Cava et al. performed an analysis between Human Ace2 and other proteins by TCGA-LUAD to get all the possible interactions while path enrichment analysis is performed using a Fisher's test . T. Joshi et al. used virtual screening to screen 318 phytochemicals to get a suitable compound to analyze the interaction with Human Ace2. PyMol is used to remove ions and water molecules. Open babel is used to convert the SDF format ligand file to PDB format. Rigid docking method is performed to get different conformations of ligand at different binding sites and in the results observed in Lig-Plot+ software . Ammar D. et al. used Computer-Aided Design (CADD) to show the interactions of ligands and receptors. The study also showed that molecular docking is done to evaluate the interactions between Human enzymes and potential ligands. Molecular docking study and ADMET profiling is used to analyze the inhibitors. Homology modeling is used to develop a structure of a protein by using its sequence and then to perform structure-based virtual screening from a large number of chemical compounds AutoDock Vina is used. For binding residues and pockets, AutoDock Vina 4.2 is used . Manoj Kumar et al. used Molecular Dynamic Simulation to study the structure of protein. Further, the DrugMint server is used to prepare drugs like ligands for screening, CASTp is used to calculate the pockets in the protein. Subsequently, autodock is used for docking ligands with receptors and analyzing binding affinity of every compound to set the threshold. Additionally, the comparative analysis of sequences was performed by Multalin . Several researchers have also applied machine learning and artificial intelligence-based models to study the genomic properties - . In this study, a method is proposed to carefully examine Sars-Cov-2 specific antiviral properties of substances. Irrelevant or ineffective chemicals are screened out. The selected compounds are further scrutinized by different docking and interaction techniques. Based on these results the most suitable compounds are proposed for the treatment of COVID-19 that can be the potential therapeutic candidates for the treatment of COVID-19 and open broad-spectrum treatment for other RNA viruses. ## Phytochemicals Preparation: The selection is performed by analyzing the properties of numerous plants. Then 3D chemical structure of 4596 phytochemicals obtained from natural herbs was extracted from databases like PubChem , ChemSpider , chEMBL and IMMPAT . Compounds converted from SDF format to PDB format using Open Babel. Subsequently, the phytochemical library is prepared for further processing. ## Receptor Protein Preparation: The recent crystal structure of RdRp protein is retrieved from Protein Data Bank (PDB ID: 6M71). Molecule SARS-Cov-2 NSP 12 has one chain with 942 amino acids. Water molecules and hydrogen atoms were removed from the Receptor by using the MGL tools of Autodock Vina. ## Virtual Screening: Virtual screening of phytochemical compounds is performed by the RPBS webserver to narrow down the potential structures that are likely to bind with the receptor. This server uses the AutoDock Vina package which is accurate and yields good screening results . Grid Center Coordinates were set to: X=-2.3, Y=45.7, Z=28.6. The search space was set to: X=55, Y=55, Z=55. Listed compounds were uploaded on the server for virtual screening. Results with a binding energy of a vast number of compounds are analyzed and all ligands which had binding affinity numerically greater than -7 were discarded. ## Molecular Docking: Suitable compounds that were selected from the results of virtual screening were further docked with the target Protein using AutoDock Vina. Grid box parameters were set to: X=-3.27, Y=44.29, Z=-28.65 and Dimensions were set to: X=35, Y=35, Z=35 (Angstrom). Universal Force Field (UFF) method was used for minimization which is more effective in finding the minimized energy than any other method. A webbased tool named admetSAR was used for profiling and finding drug similarity. ## Universal force field (UFF) Optimization: After loading ligands in Autodock Vina, UFF optimization is used to carry out the optimization of molecular geometry with the help of molecular mechanics. Method of energy minimization is used before the process of docking. This ensures that ligand's length, structure, and angles of bonds are precise before performing the docking process. This method provides good results with organic and inorganic compounds. ## Broyden-Fletcher-Goldfarb-Shanno (BFGS) Method: Broyden-Fletcher-Goldfarb-Shanno (BFGS) is used in autodock Vina for local optimization. This method helps to generate different conformers of ligands. Just like other optimization methods, BFGS also uses gradients with scoring function i.e. the derivative of functions with its arguments. In this situation, arguments contain position, orientation of ligands, and torsion values for effective bonds. This gradient is used to decide the direction of local optima. Before calculating the second derivative which may prove costly, BFGS estimates using top-level updates provided by gradient assessment. In the end, an optimized structure is chosen for selection and by using the Metropolis basis, the next iteration will start from this structure and if this structure scores better than the best available solution then this will be again optimized and will be used as the current best solution . This search process continues until the limit of iterations is reached. ## Visualization: After the completion of the above process, the most suitable results and interaction of every Ligand are further analyzed to visualize the ligand-receptor binding sites in PyMol. PyMol is a very efficient visualization software that supports 2D and 3D structure of the complex and binding interactions as well as distance measurement with sequence information of protein and ligand. ## Molecular Dynamic Simulation: Molecular dynamic (MD) simulation is widely used to evaluate the structural behavior and stability of the protein. In this study, Nanoscale Molecular Dynamic (NAMD) software is used to perform simulation on RdRp protein and proposed compounds . The temperature of 310K was set for simulation. Configuration files were generated using the CHARMM website . Parameter files were obtained using the CHARMM General Force Field (CGenFF) tool. Protein with complex was solvated using water molecules. The energy of the system was minimized for 2000 steps and dielectric was set to 1.0. ## Electrostatic Potential Calculation: Electrostatic potential simulation was performed by PyMol which uses the Poisson-Boltzmann method based on cubic spline charge discretization. Solute dielectric and solvent dielectric were set to 2.0 and 78.0 respectively. Temperature was set to 310K whereas the solvent probe radius was 1.400 . Table 3 shows grid values that were used for the calculation of electrostatic charge of RdRp. Figure 4 shows the flow chart of the above-described methodology and a brief overview of the screening and selection process of phytochemicals that are extracted from different databases. ## Results: Table 3 shows the names of selected ligands along with the binding affinity value of each ligand. For each ligand, the residue sites of RdRp with which it binds to are also listed. All screened compounds have at least -7 binding affinity. Mentioned compounds are mostly naturally occurring substances and they have shown promising antiviral activities and very good binding affinity values with RdRp during the docking process. 4 shows a comparison of compounds selected under this criteria with some of those which are currently under trial. Ligands' names along with the binding details, distance of interaction from each binding site, and estimated inhibition constant is also shown. The comparison shows that the proposed compounds have very good binding affinity and bind to key residues to compromise the replication of RdRp. Below compounds (highlighted in Green) have better binding affinity and binding interaction with RdRp key residues than the compounds that are currently under trial i.e. Camostat Mesylate , Hesperidin and Remdesivir . ## Naringin (PubChem ID: 442428): Naringin is a bioflavonoid and it belongs to the family of flavonoids, it is found in citrus fruit and has exhibited antiviral, anti-inflammatory, and possesses antioxidant properties. It is used in the treatment of diabetes, hypertension, and metabolic syndrome. It has also shown anticancer effects as it behaves as suppressing or blocking agents in the treatment of cancer. It induces cell apoptosis and impedes cell proliferation in tumor cells of Bladder cancer, Breast cancer, and cervical cancer . Figure 5 shows the chemical structure of Naringin. ## Desacetylnimbinolide (PubChem ID: 102285346) This naturally occurring substance is extracted from plant Azadirachta Indica which is commonly known as Neem in the Indian subcontinent. It has been used in Chinese and Unani medicines for many years. It is enriched with antioxidants. It plays a vital role in anticancer management . It is very safe for medicinal purposes and used in the treatment of diabetes, fever, and skin disease. Figure 6 depicts the structure of Desacetylnimbinolide. ## Sennaglucosides (PubChem ID: 5199) This is the most effective substance found in this research. This is also a naturally occurring compound that is extracted from a plant called Alexandria Senna. Its leaves are used for medicinal purposes. It is used to treat constipation and has strong laxative effects. It is also used to empty the stomach before surgery and its medication is taken by mouth. It prevents the reabsorption of water and electrolytes which results in increment of fluids in the intestine. It is safe and well-tolerated. Figure 7 is showing the structure of Sennaglucosides. Figure 7 Structure of Sennaglucosides Famotidine (PubChem ID: 5702160) Famotidine is used to decrease acids produced in the stomach and intestine. It is a histamine-2 inhibitor and used to treat ulcers in the stomach. It is also used for ZES (Zollinger-Ellison-Syndrome) in which the stomach produces excessive acids. It also prevents ulcers from coming back. It is available in the market with the name of Pepcid and its medication is taken by mouth. Figure 8 is showing the structure of famotidine. Figure 8 Chemical Structure of Famotidine Famotidone (PubChem ID: 129849878) Famotidone can be used for hayfever, skin allergies, and itchy nose. It can also be used for the treatment of skin rashes for adults and children over 6 years. Figure 9 is showing the structure of famotidone. ## 4-({3,4-dihydroxy-5-[(3,4,5-trihydroxybenzoyl)oxy]benzoyl}oxy)-1-hydroxy-3,5-bis[(3,4,5trihydroxybenzoyl)oxy]cyclohexane-1-carboxylic acid (PubChem ID: 442676) This compound which is also named Quinic acid (multiplied acylated with galloyl moieties) is extracted from Eucalyptus bark. It is antiseptic and anti-inflammatory and used for the treatment of asthma. It contains a substance that kills bacteria. It is also used for skin diseases like skin ulcers and Gout. This compound is extracted from Rutaceae which belongs to the rue family of flowering plants. This is also found in citrus fruits like orange and lemon. It is used in many diseases like asthma, constipation, fever, and diarrhea. Proposed compounds can be used in combinations of 2 or 3 to inhibit the working of RdRp most effectively. Now we shall visualize results in combinations. ## 1) Naringin & Sennaglucosides: Figure 12, shows that a combination of two promising compounds (Sennaglucosides and Naringin) binds to 5 key binding sites of RdRp. Magenta and Green Color represent Sennaglucosides and Naringin respectively. Yellow dots illustrate binding interactions between the combination of compounds and RdRp. ## Figure 12 Combined Interaction of Sennaglucosides and Naringin To further understand the results, table 5 shows that Sennaglucosides and Naringin have inhibition constants of 685 nM and 488 nM respectively, and interact with 5 key sites (both compounds bind with ASP 623 simultaneously, that's why this binding site is neglected for Naringin) of RdRp protein. It is also shown in the table that the combination of Sennaglucosides and Naringin interacts with the key binding residues of RdRp with a good binding affinity of over -8.3. 2) Sennaglucosides and Desacetylnimbinolide: Figure 13 shows the conformations of Sennaglucosides (Magenta) and Desacetylnimbinolide (Green) in the combination which best fit the key residues and cover over 10 binding sites but our main focus is key residues. This combination covers 6 binding residues and binding interactions are shown in yellow dotted lines. ## Figure 13 Combined Interaction of Sennaglucosides and Desacetylnimbinolide Table 6 shows that Sennaglucosides binds to 4 key residues with an inhibition constant of 685 nM and Desacetylnimbinolide binds to 2 key residues with an inhibition constant of 0.146 nM and combination of both compounds can cover 6 key binding sites that are very important to inhibit the function of RdRp. This interaction with the key binding residues can halt the exponential growth of Sars-Cov-2 in human cells by compromising the function of RdRp. ## 3) Desacetylnimbinolide and Naringin: In figure 14, the interaction of Desacetylnimbinolide (represented in Green) and Naringin (represented in Magenta) has been illustrated with key sites of RdRp. Binding interactions are shown in yellow dots. Table 7 shows that the combination of these two compounds covers 3 key binding sites (ARG-555 is common in both compounds' interaction, ARG 555 from Naringin is not included in Figure 13 and neglected in Table 8). The details of binding residues that each compound cover along with the binding affinity and inhibition constant is also shown. These compounds can be very effective for the treatment of RNA-related and antiviral diseases. ## 5) Sennaglucosides, Cyclohexane-1-carboxylic acid and 8-difluoro-7-hydroxy chromen-4one: In figure 16, a combination of Cyclohexane-1-carboxylic acid which is also known as Quinic Acid (represented in Blue), Sennaglucosides (represented in Magenta), and 8-difluoro-7-hydroxy chromen-4one (represented in Orange) yields the best results in terms of binding to key residues. Table 9 shows that Cyclohexane-1-carboxylic acid (Quinic Acid) binds to 3 key residues, Sennaglucosides binds to 4 key residues, and 8-difluoro-7-hydroxy chromen-4-one binds to 1 key residue, making it an effective combination that binds to 8 key residues of RdRp collectively including ASN 691 and THR 680 which are very important and there are very rare compounds that bind to these two (ASN 691, THR 680) key residues. Binding affinity and inhibition constant of each compound is also mentioned. ## Electrostatic Potential Distribution: Electrostatic potential is an effective way to understand the structural properties and characteristics of protein and ligands which bind to it. Electrostatic potential charges are mapped on the surface of the RdRp protein of Sars-Cov-2, to show the distribution of positive and negative charges and the intensity on the surface of the protein. Distribution of Positive potential charges (Blue) covers the inner cavity of binding pockets of RdRp protein and the remaining surface is covered by the negative charges (Red). Above Figure 17 shows the location of combinations of ligands in the inner cavity of binding pockets of protein which is a clear indication of the fact that proposed compounds bind to key binding sites (Cavity of Binding pockets is shown in the blue) which are the main cause of replication and progression of Virus in the host cells. Thus covering these sites will inhibit the working of RdRp protein. ## Molecular Dynamic Simulation Analysis: In this study, Root Mean Square Deviation (RMSD) is measured to evaluate the distance between backbone atoms of superimposed molecules. As shown in Figure 18, RMSD of RdRp protein remained stable between 16ns to 25 ns timescale at 1.581 , then showed a slight upward deviation until 34ns and at 35ns it persisted at 1.581 till the end. The RMSD of RdRp_ Quinic Acid, Sennaglucosides, and 8difluoro-7-hydroxy chromen-4-one showed rise until 20ns at 1.7 and after slight fluctuation it gained stability at 25ns at 1.76 . The RMSD of RdRp_Sennaglucosides and Desacetylnimbinolide showed stability at 15ns timescale at 1.55 and after slight upward fluctuation it system was balanced at 28na timescale at 1.77 . The RMSD of RdRp_Naringin and Sennaglucosides increased up to 15ns timescale at 1.62 and then fluctuated downward on timescale at 1.57 and the system was balanced at 34ns timescale at 1.66 . The RMSD of RdRp_Desacetylnimbinolide and Naringin gained stability at 21ns timescale at 1.61 . The RMSD of Famotidine and Famotidone ascended until 11ns and then the system was stable until 22ns timescale at 1.53 . Figure 18 shows the RMSD plots of protein with all suggested compounds. A brief analysis has shown the Root Mean Square Fluctuation (RMSF) of residues of RdRp protein with its complexes. In Figure 19, RdRp and its binding compounds have shown the fluctuations between 1.2 and 1.8 . This depicts that proposed compounds have maintained close binding contact with the binding residues during Molecular Dynamic simulation. ## Suggested Combination: By analyzing the above results, we can predict the most suitable combination according to the number of key residues that it covers. In table 11, combinations of the proposed compounds are listed according to the most suitable first. The table also shows the names of binding residues that these compounds cover along with the number of compounds in each combination. Combinations of compounds are selected according to the ability to cover the maximum key residues as well as binding affinity with RdRp. ## Conclusion: COVID-19 is a viral disease that has caused a pandemic in the modern era. Not only has it affected social life but it has imparted an impeding effect on world economies. People having an underlying health condition are at great risk. The only way to undo this threat is either by finding a vaccine or a potent antiviral therapy against the virus. Researchers all over the world have proposed numerous drug therapies for the disease. This study covers in-silico identification of phytochemicals that can prove effective in inhibiting the function of RdRp proteins of Sars-Cov-2. The study proposes 7 compounds that can prove effective as per in-silico evidence when used in combinations or individually. These compounds have shown promising signs towards the development of antiviral medications for the COVID-19. Most of them are naturally occurring substances with low toxicity, very few side effects, proven anti-pathogenic effects, and most importantly are easily available. They bind to the key sites of RdRp protein to inhibit its functioning and stop the replication of coronavirus. All the results have been carefully analyzed through the use of in silico methods and machine learning models. Their binding affinities and binding sites are thoroughly observed for result compilation. The most promising observation from the simulation is that a therapy based on the combination of Cyclohexane-1-carboxylic acid (Quinic Acid), Sennaglucosides, and 8-difluoro-7-hydroxy chromen-4-one can bind to eight out of nine key residue sites of RdRp protein of Sars-Cov-2. This is a strong indication that the combination of these compounds can significantly compromise the replication cycle of Sars-Cov-2 and hence alleviate the severity of the disease. ## Future Work: All the results shown in the study are obtained from in silico methods. The proper clinical trial and medical observation will reveal more crucial information about their effectiveness. If the proposed compounds make an impact in the development of the vaccine of COVID-19 then these compounds can also be used in further research of RNA-related viral and other contagious diseases.

chemsum

{"title": "Identifying Therapeutic Compounds Targeting RNA-Dependent-RNA-Polymerase of Sars-Cov-2", "journal": "ChemRxiv"}

fully_oxygen-tolerant_atom_transfer_radical_polymerization_triggered_by_sodium_pyruvate

## Abstract: ATRP (atom transfer radical polymerization) is one of the most robust reversible deactivation radical polymerization (RDRP) systems. However, the limited oxygen tolerance of conventional ATRP impedes its practical use in an ambient atmosphere. In this work, we developed a fully oxygen-tolerant PICAR (photoinduced initiators for continuous activator regeneration) ATRP process occurring in both water and organic solvents in an open reaction vessel. Continuous regeneration of the oxidized form of the copper catalyst with sodium pyruvate through UV excitation allowed the chemical removal of oxygen from the reaction mixture while maintaining a well-controlled polymerization of N-isopropylacrylamide (NIPAM) or methyl acrylate (MA) monomers. The polymerizations of NIPAM were conducted with 250 ppm (with respect to the monomer) or lower concentrations of CuBr 2 and a tris [2-(dimethylamino) ethyl]amine ligand. The polymers were synthesized to nearly quantitative monomer conversions (>99%), high molecular weights (M n > 270 000), and low dispersities (1.16 < Đ < 1.44) in less than 30 min under biologically relevant conditions. The reported method provided a well-controlled ATRP (Đ ¼ 1.16) of MA in dimethyl sulfoxide despite oxygen diffusion from the atmosphere into the reaction system. ## Introduction According to the IUPAC report, reversible deactivation radical polymerization (RDRP) is one of the top ten emerging technologies in chemistry that could change the world. 1 Atom transfer radical polymerization (ATRP) is one of the most widely used RDRP methods, providing access to well-defned, complex polymer architectures. ATRP is catalyzed by transition metal complexes in their lower oxidation state. It is exceptionally tolerant to a wide variety of functional groups, solvents, and impurities. However, like any radical polymerization, ATRP is inhibited by oxygen. Recently, several avenues to design oxygen tolerant RDRP systems have been reviewed. 8 The most active copper catalysts with highly negative redox potentials allow a well-controlled polymerization at a loading of only 10 ppm relative to the monomer. Even trace amounts of oxygen can inhibit polymerization by rapidly oxidizing the activator form of the catalyst Cu I /L to the inactive Cu II /L complex. 13 Furthermore, oxygen molecules can react with the propagating carbon-based radicals, thus terminating the polymerization process. 14 The sensitivity of ATRP to oxygen necessitates the use of specialized equipment or deoxygenation by inert gas sparging before the polymerization (Scheme 1). As a result, ATRP techniques can be cumbersome to non-experts. On top of that, inert gas sparging or freeze-pump-thaw degassing are often incompatible with the synthesis of hybrid biomacromolecules, 15,16 as they may cause protein denaturation or a loss of enzymatic activity. 17 Scheme 1 Approaches for oxygen scrubbing and achieving oxygen tolerance in ATRP. The Cu I /L ATRP catalyst activates the dormant C(sp 3 )-X polymer chain end, resulting in the formation of the X-Cu II /L complex and a carbon-centered radical. Both carbon-based radicals and Cu I /L species react with molecular oxygen with diffusion control to form peroxy radicals or hydroperoxides and Cu II /L complexes, respectively (Scheme 1). However, since Cu I /L is at a concentration thousands to millions times higher than the concentration of propagating radicals, oxidation of the Cu I / L activator to Cu II /L is predominant. Thus, continuous regeneration of the oxidized form of the catalyst Cu II /L with a reducing agent allows the chemical removal of oxygen from the reaction system (Scheme 1). In 1998, we demonstrated that a well-controlled ATRP could occur in the presence of a limited amount of oxygen using a zero-valent copper powder as a reducing agent. 18 This concept was later extended to ATRP with copper wire or copper plate and other reducing agents, such as ascorbic acid, tin(II) 2-ethylhexanoate, 30 tertiary amine, 31 nitrogen-based ligands, 32 phenols, 33 alcohols, 34 sodium dithionite, 35 and zerovalent iron. 36 Another area where signifcant progress has been made towards oxygen-tolerant ATRP is photoinduced polymerization. In photoinduced ATRP, catalyst regeneration occurs by excitation of the Cu II /L complex, followed by a single electron donation from the amine-based ligand. Photoirradiation of a copper catalyst in the presence of an electron donor in excess enables removal of dissolved oxygen. Despite these great developments, the vast majority of reported methods are successful only when polymerization is performed in sealed vessels with a limited amount of oxygen in the reaction mixture. So far, only a few ATRP systems, mainly based on enzymatic degassing, can be carried out in a completely open reaction vessel, where oxygen continuously diffuses into the system from the atmosphere. In 2018, inspired by the works of Yagci 59 and Stevens, 60,61 we developed a "breathing ATRP" of oligo(ethylene oxide)methyl ether methacrylate that used glucose oxidase (GOx) as a highly efficient scavenger for oxygen. 55 GOx catalyzes the oxidation of b-D-glucose to D-glucono-1,5-lactone and hydrogen peroxide. However, hydrogen peroxide reacts with Cu I /L in a Fenton-type reaction to form a hydroxyl radical and the Cu II /L complex. Hydroxyl radicals can initiate new polymer chains, decreasing average molecular weights (M n ) as compared to the theoretical values. To suppress this undesirable process, we developed a bio-inspired ATRP system in which GOx removed oxygen, while sodium pyruvate (SP) acted as a hydrogen peroxide scavenger and prevented the formation of new polymer chains. This study was later extended to "oxygen-fueled" ATRP by employing Horseradish peroxidase (HRP) as a catalyst for the generation of radicals from acetylacetone in the presence of hydrogen peroxide produced by GOx. This enzymatic cascade enabled a well-controlled ATRP in a reaction vessel open to the air. 56 However, these high-performance biocatalytic systems created a new challenge: the synthesized polymers or polymer bioconjugates were contaminated with enzymes, which are particularly difficult to separate from biohybrids. Also, the methods were limited to aqueous media. Recently, Keitz et al. harnessed an even more complex biological system, microbial metabolism, to develop an aerobic ATRP in water. 62 As with the enzymatic degassing, the use of cellular respiration machinery in bioconjugates synthesis can complicate the purifcation process. The development of efficient small molecule-based ATRP methods tolerant to oxygen that are compatible with water and organic solvents is therefore highly desirable. Poly(N-isopropylacrylamide) (PNIPAM) is a temperatureresponsive, biocompatible polymer that has a lower critical solution temperature in water of $32 C. 63 This feature is widely used in the design of controlled drug delivery systems, 64 tissue engineering 65 and biosensing. 66,67 Low dispersity PNIPAM with varying molecular weights can be synthesized using a variety of ATRP techniques. However, the methods reported so far exhibit at least one critical flaw, such as the use of high loadings of copper catalysts, a relatively long reaction time, or oxygen intolerance. Recently, the disproportionation of Cu I /Me 6 TREN in water was shown to enable the ATRP of N-isopropylacrylamide (NIPAM) in open-air conditions. However, the use of high copper concentration (2000-8000 ppm relative to NIPAM) was necessary to attain high monomer conversions and low dispersity values. 73 Herein, we demonstrate the frst fully oxygen tolerant, photoinduced ICAR ATRP of NIPAM with ppm level of Cu catalyst in water, enabling a quantitative conversion of the monomer in less than 30 min. This simplifed, non-enzymatic ATRP system uses sodium pyruvate as both a hydrogen peroxide scavenger and a "fuel" for the continuous regeneration of the catalyst and can be easily transferred to organic solvents. ## Results and discussion Initial studies began by polymerizing NIPAM in water (targeting a degree of polymerization 200) under UV LED irradiation (l ¼ 394 nm, 2.6 mW cm 2 ), using 2-hydroxyethyl 2-bromoisobutyrate (HOBiB) as the initiator, CuBr 2 as the precatalyst, and tris [2-(dimethylamino)ethyl]amine (Me 6 TREN) as the ligand (Table 1). The reactions were carried out in sealed vials with a septum (see Fig. S1 in the ESI †) at 6 C and in the presence of limited amounts of oxygen (without degassing the reaction mixture). Table 1 shows the results of the polymerization of NIPAM and the effect of different components involved in the PICAR ATRP system. A set of control experiments was performed to evaluate the influence of SP on the ATRP process (Table 1, entries 1-3). The initial conditions used 250 ppm of CuBr 2 (with respect to the monomer) with a six-fold excess of Me 6 TREN ligand to Cu II and no SP. After 12 h of UV irradiation, the conversion of NIPAM measured by 1 H NMR was only 16%. Furthermore, size exclusion chromatography (SEC) analysis showed that the polymer had a high dispersity (Đ) of 1.86 (Table 1, entry 1). The use of Cubased ATRP catalysts in water typically results in a signifcant dissociation of the [X-Cu II /L] + deactivator to the "naked" [Cu II / L] 2+ dication and a free halide anion. The [Cu II /L] 2+ complex cannot act as a true deactivator, leading to poorly controlled polymerizations. To counteract this problem, aqueous ATRP is performed in the presence of halide anions to suppress the deactivator dissociation. 74 The use of modifed phosphate-buffered saline (PBS) solution containing bromide anions gave slightly better results (Table 1, entry 2; Đ ¼ 1.66). To our delight, when both the Br-based PBS and SP were used (Table 1, entry 4), a quantitative conversion was achieved within 30 min, and the polymerization was well-controlled (M n ¼ 31 700, Đ ¼ 1.16). These experiments showed the critical role of SP. The reaction without the addition of buffer components (Na 2 HPO 4 and KH 2 PO 4 salts) reached quantitative monomer conversion with a similar rate of polymerization. However, the dispersity of the resulting polymer was 1.42 (Table 1, entry 3). For aqueous ATRP, the optimal pH is 7.5. 75 UV irradiation induces the homolytic cleavage of SP (see further section on proposed mechanism), which leads to the protonation of the ligand, decreasing its ability to coordinate the metal center, resulting in a loss of control over the polymerization. Maintaining a constant pH $ 7.4 during the polymerization prevents this process. Next, the performance of sodium pyruvate-based ATRP system was evaluated in the presence of varying amounts of CuBr 2 (Table 1, entries 4-7). Despite decreasing the amount of CuBr 2 to just 50 ppm relative to NIPAM, the reaction still proceeded to high monomer conversion (>97%) and yielding a polymer with a dispersity of 1.37 (Table 1, entry 6). Increasing the amount of CuBr 2 to 1000 ppm did not improve the outcome, yielding similar control as with 250 ppm of CuBr 2 (Table 1, entry 7). In conventional photoinduced ATRP, a Cu II complex in the excited state reacts with an amine-based ligand, which acts as an electron donor, resulting in the formation of the activator Cu I /L and a radical cation from the donor. 39,76 Since this process consumes the ligand, it must be present in excess. In our PICAR ATRP, SP is the dominant electron donor. However, in the presence of dissolved oxygen, the ligand oxidation may still occur during photoirradiation. This explains why the ratio [CuBr 2 ]/[Me 6 TREN] ¼ 1/1 was not sufficient to achieve wellcontrolled polymerization while maintaining high conversion (Table 1, entry 9). The use of a 1/3 or 1/6 ratio allowed much better control over polymerization of NIPAM (Table 1, entry 8 and 4). Increasing the concentration of SP from 100 mM to 200 mM caused a slight increase in PNIPAM dispersity (Table 1, entry 10; Đ ¼ 1.20). This could be attributed to a higher concentration of radicals resulting from the homolytic cleavage of SP under UV irradiation. The radicals thus formed could initiate new polymer chains or terminate polymerization by radical-radical coupling. Decreasing the SP concentration to 50 mM caused only a slight decrease in monomer conversion (98%), while maintaining the low Đ ¼ 1.16 (Table 1, entry 11). However, further tests were carried out with a concentration of SP of 100 mM to make the polymerization more tolerant to oxygen. Several hypotheses have been proposed to explain the difficulty of obtaining good control in the polymerization of acrylamides by ATRP. One possibility is the intramolecular cyclization reaction leading to the loss of C(sp 3 )-Br chain end. The u-Br chain end functionality was shown to decrease as a function of reaction time and was dependent on the structure of the amide group. 79 Low chain-end fdelity compromises the control over polymerization. The use of the pyridine-based ligands: the less active TPMA or the more active TPMA* 3 ligands (TPMA ¼ tris(2-pyridylmethyl)amine, TPMA* 3 ¼ tris([(4methoxy-2,5-dimethyl)-2-pyridyl]methyl)amine) 81 resulted in a signifcant decrease in control over the polymerization (Table 1, entry 12 and 13). TPMA is the most versatile ligand for aqueous ATRP of acrylates and methacrylates. 75 However, for acrylamides, the [Cu I /TPMA] + catalyst does not provide a sufficiently high ATRP equilibrium constant (K ATRP ). Thus, the rate of the polymerization is slower, then the loss of C(sp 3 )-Br chain-ends via intramolecular cyclization. In turn, poor control provided by very active TPMA* 3 ligand could be explained by the too high value of the ATRP equilibrium constant. Higher K ATRP implies higher radical concentration at equilibrium, which favors bimolecular termination reactions, resulting in diminishing control over the polymerization. Next, the kinetics of the polymerization of NIPAM was investigated in an open reaction vessel (Fig. 1A). The reaction was performed in a Br-based PBS buffer with [NIPAM]/[HOBiB]/ [CuBr 2 ]/[Me 6 TREN] molar ratios of 200/1/0.05/0.30, in the presence of SP (100 mM). A short inhibition period of 15 min was observed, followed by a well-controlled (Đ ¼ 1.15), rapid polymerization with linear semi-logarithmic kinetics, that reached 97% monomer conversion in 15 minutes. We unexpectedly observed that polymerization in an open vessel (Fig. 1B) led to smaller deviation from the theoretical molecular weight value (M n,th ¼ 22 400, M n,GPC ¼ 25 400) than polymerization in a sealed vessel (M n,th ¼ 22 800, M n,GPC ¼ 31 700). Fast activation of initiators, leading to termination of initiating radicals could explain this deviation. 3 The performance of this system was further evaluated in a series of reactions in a closed vessel, with varying target degrees of polymerization (DP) of NIPAM (Table 2). The concentration of CuBr 2 was maintained at 250 ppm relative to the monomer. The results showed a high degree of control for targeted DP ¼ 100, 200, 400, 1000, and 2000 (Table 2, entries 1-5). In all cases, nearly quantitative monomer conversions were reached with Đ in the range 1.16-1.44. However, for higher DP ¼ 4000 and 10 000, a signifcant deviation from the theoretical molecular weights and higher Đ values were observed (Table 2, entries 6 and 7). Moreover, SEC traces of the polymers showed a signifcant tailing (ESI Fig. S5F and G †). The appearance of this tailing could be attributed to the continuous formation of new chains, plausibly generated by radicals formed from the photochemical homolytic cleavage of SP. Then, polymerizations of NIPAM were conducted in the open reaction vessel, targeting DP of 100-2000 (Table 3). A high level of control over polymers was achieved under PICAR ATRP conditions when oxygen continuously diffused into the reaction system from the atmosphere, reaching 65-97% monomer conversions and providing polymers with monomodal, narrow molecular weight distributions (1.15 < Đ < 1.32). In open-air conditions, the Cu I /L activator is constantly oxidized to inactive Cu II /L, which results in lower monomer conversions. In turn, the increased X-Cu II /L deactivator concentration provides better control over polymerization. The promising results of PICAR ATRP in an aqueous medium prompted us to utilize this fully oxygen-tolerant system in an organic solvent, which would signifcantly extend the scope of this method toward hydrophobic monomers. Since the SP-triggered ATRP catalytic system is based on small molecules, it can be transferred to organic solvents much more easily compared to ATRP techniques based on enzymatic degassing. However, sodium pyruvate has limited solubility in organic solvents due to its ionic structure. We used a stoichiometric amount of tetrabutylammonium bromide (TBAB) to increase the solubility of SP in dimethyl sulfoxide (DMSO). This ) in the presence of oxygen (sealed vessel). b Monomer conversion was determined by using 1 H NMR spectroscopy. c See SEC traces in the ESI Fig. S3. All measurements were analyzed using GPC (dimethylformamide as eluent) calibrated to poly(methyl methacrylate) standards. quaternary ammonium salt is commonly used as a phase transfer catalyst in many synthetic transformations. 82 The investigation was started by preparing a reaction mixture that contained all components needed for the polymerization (Fig. 2 ) was performed at room temperature. After 3 h, the conversion of MA measured by 1 H NMR was 84%. SEC analysis showed that the polymer had a low dispersity (Đ ¼ 1.16), and a molecular close to the theoretical value (M n,th ¼ 14 500, M n,GPC ¼ 16 700), indicating a well-controlled polymerization (Fig. 2). This experiment shows that even without a time-consuming, careful optimization, a highly efficient, fully open-air ATRP system could be quickly developed. Further optimization of this method and its applicability to other non-polar monomers will be the subject of a forthcoming publication. To gain insights into the polymerization mechanism, we investigated the reactivity of SP toward Cu complexes. Recently we reported the sono-ATRP of MA in DMSO in the presence of sodium carbonate. 83 In this system, ultrasonication triggered the homolytic cleavage of the in situ formed (CO 3 )-Cu II /TPMA complex, generating Cu I species and a radical carbonate anion. Haddleton et al. observed a similar phenomenon for the photoreduction of (HCO 2 )-Cu II /Me 6 TREN complex. 84,85 In addition, Vaida et al. showed that the UV excitation of pyruvic acid in an aqueous medium causes photodecarboxylation, which forms radicals as intermediates. 86 Furthermore, a-keto acids can undergo decarboxylative acyl radical formation in transition metal-catalyzed radical cross-couplings. 87,88 Based on the above results, we propose that the SP reacts with Cu II species to yield a (CH 3 C(O)CO 2 )-Cu II /L complex by a simple anion dissociation/association process (Scheme 2). Subsequent UV excitation causes the homolytic cleavage of the carbon-carbon bond in the pyruvate moiety. This photolysis induces decarboxylation, which leads to the reduction of Cu II /L to Cu I /L and the formation of the acyl radical. This radical can regenerate the activator or initiate a new polymer chain by addition to the monomer. The role of a buffer medium is to control the pH and, thus, the concentration of the formed acyl radicals. 86 Furthermore, the reaction between the acyl radical and X-Cu II /L deactivator leads to the formation of an acyl halide, which undergoes rapid hydrolysis in a buffer. This prevents the initiation of new polymer chains and the protonation of the ligand. A control experiment without the HOBiB initiator showed that SP could initiate ATRP on its own, but the polymer had a broad molecular weight distribution and a high Đ ¼ 1.92. In order to confrm the role of SP in the catalytic system, UV-vis spectroscopy measurements were performed. Fig. S6A † shows a decrease in the absorbance of Cu II /TPMA complex under UV irradiation in the presence of SP. Fig. S6B † shows that the absorbance of Cu II /TPMA decreased much slower when SP was absent, which indicates that SP is necessary for the efficient reduction of Cu II species. This in turn is critical for ATRP in an ambient atmosphere. The proposed mechanism shown in Scheme 2 can be considered equivalent to PICAR (photoinduced initiators for continuous activator regeneration) ATRP. 89,90 ## Conclusions We have developed the frst example of a photoinduced ATRP system that operates in an open reaction vessel and yields wellcontrolled polymerizations in both aqueous and organic solvents. Sodium pyruvate is the essential component in this novel method, acting as a hydrogen peroxide scavenger and enabling the continuous regeneration of the copper catalyst through UV excitation. This methodology allowed the synthesis of poly(N-isopropylacrylamide) in water with high monomer conversion (97%) and dispersity of 1.15 with 250 ppm of a catalyst in 30 min under an ambient atmosphere. Furthermore, the use of sodium pyruvate with tetrabutylammonium bromide enabled the polymerization of methyl acrylate in DMSO in a fully open vessel without compromising the control over the molecular weight distribution (Đ ¼ 1.16). Non-experts can easily apply this straightforward and robust protocol for the synthesis of well-defned polymers. Expanding the scope of this methodology to more complex polymer architectures and polymer-based biohybrids is currently under investigation. ## Conflicts of interest There are no conflicts to declare.

chemsum

{"title": "Fully oxygen-tolerant atom transfer radical polymerization triggered by sodium pyruvate", "journal": "Royal Society of Chemistry (RSC)"}

thermal_chemical_vapor_deposition_of_epitaxial_rhombohedral_boron_nitride_from_trimethylboron_and_am

## Abstract: Epitaxial rhombohedral boron nitride films were deposited on α-Al2O3(001) substrates by chemical vapor deposition, using trimethylboron, ammonia, and with a low concentration of silane in the growth flux. The depositions were performed at temperatures from 1200 to 1485 °C, pressures from 30 to 90 mbar and N/B ratios from 321 to 1286. The most favorable conditions for epitaxy were: a temperature of 1400 °C, N/B around 964, and pressures below 40 mbar. Analysis by thin film X-ray diffraction showed that most deposited films were polytype-pure epitaxial r-BN with an out-of-plane epitaxial relationship of r-BN[001] ∥ w-AlN[001] ∥ α-Al2O3[001] and with two in-plane relationships of r-BN[110] ∥ w-AlN[110] ∥ α-Al2O3[100] and r-BN[110] ∥ w-AlN[110] ∥ α-Al2O3[1 ̅ 00] due to twinning. ## I. INTRODUCTION The trialkylboron triethylboron (TEB, B(C2H5)3) is commonly used as boron precursor in chemical vapor deposition (CVD) of boron-based thin films as it is less corrosive than the halides BF3 and BCl3, and less poisonous than diborane (B2H6). In a seminal study, Lewis et al. 1 compared the trialylborons TEB, trimethylboron (TMB, B(CH3)3) and tributylboron (TBB, B(C4H9)3) and suggested that TEB was the most suitable for depositing boron carbon films, judged mainly by the high B/C ratio obtained. A recent study on the thermal gas phase chemistry of TEB in CVD 2 confirmed that the molecule is an efficient boron source at temperatures below 1000 °C. TEB has been employed for CVD of boron carbides 2,3 , phosphides and arsenides 4 and, of particular interest for this study, boron nitrides. TEB decomposes primarily by β-hydride elimination, offering a low-temperature synthesis route for boron-rich films 1,2 . On the contrary, a drawback is that the ethyl ligands are suggested to form C2H4 upon β-hydride elimination 2 , which will be reactive as CVD precursors at the high temperatures needed for the growth of boron nitride (around 1500 °C) 5,9 and can therefore lead to carbon impurities in the BN films. In this regard, TMB is a promising alternative to TEB. TMB was recently shown to be an efficient boron precursor for high temperature deposition and suggested to form less reactive CH4 in an α-elimination decomposition. 14 deposited films from 10 B-enriched TMB and ammonia in a nitrogen ambient, 15 but did not report on any characteristics of the process. Here, we investigate CVD of epitaxial rhombohedral boron nitride (r-BN) using TMB and ammonia in hydrogen ambient at 3 temperatures ranging from 1200-1485 °C, pressures from 30 to 90 mbar and N/B ratios ranging from 321 and 1286. ## II. EXPERIMENTAL DETAILS BN films were deposited on α-Al2O3(001) for 120 min at temperatures of 1200, 1300, 1400, and 1485 °C in a hot-wall CVD reactor kept at a base pressure below 10 -7 mbar. The substrates were cut in 10x10 mm 2 pieces and were cleaned according to the following procedure: 3 min in an ultrasonic bath in acetone at 80 °C, 3 min in an ultrasonic bath in ethanol at 80 °C, followed by standard clean 1 (SC1, NH3:H2O2:H2O with relative concentrations 1:1:26 at 80 °C) 16 and standard clean 2 (SC2, HCl:H2O2:H2O with relative concentrations 1:1:22 at 80 °C) 16 . The substrates were placed in the center of a tantalum-carbide-coated elliptical susceptor. Prior to BN deposition, the α-Al2O3 substrates were heated to 1100 °C during 5 min in palladium membrane purified hydrogen gas (H2), after which ammonia (NH3, 99.999 %, further purified with respect to water by a getter filter) was introduced and the temperature ramped up to the selected growth temperature for 10 min to form an insitu aluminum nitride buffer layer as previously reported in 8,9,12 . H2 was used as carrier gas for the boron precursor TMB (99.99 % purity, Voltaix/Air Liquide Advance Materials, FL) as well as the nitrogen source NH3. TMB was flowed in a separate quartz liner to avoid the formation of the NH3:B(CH3)3 adduct 17 . The N/B ratio was varied between 321 and 1286. From previous works 18 , silane (SiH4, 99.999 % purity, 2000 ppm diluted in 99.9996 % H2) was inserted 2 min prior to growth. The process pressure was in the range of 30 to 90 mbar and regulated by a throttle valve. The growth temperature was monitored by a pyrometer (Heitronics KT81R, calibrated by silicon melting). The deposited films were characterized by thin film X-ray scattering, electron microscopy and ion beam analysis. All diffractograms and reflectograms were acquired using Cu Kα radiation. The 2θ/ω diffractograms were acquired with a PANalytical X'Pert PRO, using a Bragg-Brentano HD mirror with 1/2° divergence and anti-scatter slits as primary optics and an X'Celerator detector with a 0.5 mm anti-scatter slit, 0.04 rad Soller slits and nickel Kβ filter as secondary optics. In plane measurements, as azimuthal scans (φ-scans) and Glancing-Incidence Diffraction (GID) were acquired with a Phillips X'Pert MPD, using cross-slits (2x2 mm 2 ) with nickel Kβ filter as primary optics and a proportional detector (PW1711/96) equipped with a parallel plate collimator. The thickness of the film was estimated from scanning electron microscopy (SEM) using an accelerating voltage of 5 kV and an inlens secondary electron detector. The analysis of the composition was performed by time-of-flight energy elastic recoil detection analysis (ToF-E ERDA). The measurements were carried out with a 36 MeV 127 iodine ion beam. The incident angle of primary ions and exit angle of recoils were both 67.5° to the sample surface normal giving a recoil angle of 45°. The measured ToF-E ERDA spectra were converted into relative atomic concentration profiles using the Potku code 19 . and 54.5° originated from the diffraction of sp 2 -BN(00ℓ) and the second order diffraction (002ℓ), suggesting highly-oriented pyrolytic BN 20 or textured h-BN 21 or r-5 BN 22 on the nitridated α-Al2O3(001). We note that the growth temperature of 1400 °C is 100 °C lower than the temperature previously reported for TEB at similar growth conditions 9 . Increasing temperature to 1485 °C or decreasing it to 1300 °C, decreases the intensity of the 00ℓ peak and the second order diffraction peaks are no longer visible. At 1200 °C, no diffraction peak is visible. As for deposition with TEB 9 , the deposition of high-quality sp 2 -BN films seems to be constrained to a narrow temperature window, albeit at 100 °C lower temperature. In addition, to the highintensity 006 peak from the sapphire substrate in all investigated films, the 002 diffraction peak of w-AlN was detected from 1200 °C and with 100 and 110 peaks visible for growth at 1485 °C. SEM measurements of cross sections showed that the average film thickness increased from 896 ± 87 nm at 1200 °C to 1308 ± 194 nm at 1400 °C, corresponding to an average growth rate from 7.5 ± 0.7 nm/min to 10.9 ± The {110} planes of sp 2 -BN cannot be used to determine the BN polytype. By rotating the sample by 30°, it is possible to use GID to investigate the presence of the {100} planes of h-BN, as this family of planes is extinct in the case of r-BN. 22,23 The result is shown in Figure 2.(b), where only diffraction from the aluminum nitride buffer and the sapphire substrate is detected. This shows that it is possible to obtain polytype-pure r-BN films from TMB. In a few films, diffraction of h-BN inclusions could be detected by GID, as previously reported in 24 . ## A. Structural characterization of the BN films The twinning of the r-BN films was also investigated by acquiring azimuthal scans of the {101} of planes of r-BN. These planes have a three-fold symmetry as dictated by rhombohedral crystal system, whereas the φ-scan in Figure 3 shows six peaks for r-BN{101}. This originates from the presence of twin crystals that are rotated by 60°. From this result, two in-plane epitaxial relationship can be determined Twinning of r-BN has been reported in previous works for films deposited on sapphire 9 and SiC 10 and is to be expected due to the 6-fold symmetry of the AlN buffer layer and of the hexagonal polytypes of silicon carbide, respectively. 9 FIG. 3. XRD φ-scans of r-BN{101} (2θ = 42.6835°, ψ = 77.61°) and α-Al2O3{202} (2θ = 46.161°, ψ = 72.20°). Diffraction from crystals oriented -30° with respect to the substrate is indicated by circles, diffraction from crystals oriented +30° is indicated by crosses. ## B. Deposition process In contrast to r-BN deposited from TEB 9 , epitaxial films were obtained in a wider range of N/B ratios and lower pressures using TMB at 1400 °C. At fixed pressure, NH3/TEB ratios below 460 and above 770 were shown to strongly affect r-BN epitaxy 9 , whereas NH3/TMB between 321 and 1286 resulted in epitaxial r-BN, without having any influence on the crystal quality from 2θ/ω scans for N/B ratios above 643, as shown in Figure 4 (a) by the full width at half maximum (FWHM) values from θ/2θ scans of r-BN(003). For comparison, the FWHM of 2θ/ω diffractograms of α-Al2O3(006) was 0.06°. Figure 4 (b) indicates an optimal N/B ratio of 964. This is higher than the value observed for decomposition of TEB (N/B around 10 615-640) and can be explained by the fact that at a lower process temperature the activation of the NH3 molecule is less favorable 25 . In Figure 4 (c), the crystal quality is shown to increase while decreasing the process pressure as illustrated by the proportional decrease in FWHM of θ/2θ scans of r-BN(003) with decreasing process pressure from 90 to 30 mbar. Interestingly, Figure 4 (d) shows that at fixed N/B ratios, the pressure does not affect the total amount of coherently diffracting domains along the c-axis, i.e. the proportion of crystallites is independent of the total pressure at these experimental conditions. i.e. 636 sccm NH3 in 6400 sccm H2 at 40 mbar. The TMB flow was 0.9 sccm and mole fractions were the same as in Figure 1. Lines are guide for the eyes. Similarly as reported for TEB 18 , the deposition process using TMB is dependent on the background silicon concentration. In the absence of silane, the intensity of the (003) diffraction peak of r-BN is significantly reduced and the peak broadens as shown in Figure 5. ToF-ERDA gives 44.5 at% B, 46,1 at% N (B:N ratio of 1:1.04), 4.3 at% C , 3.8 at% O and 1.1 at% H in an r-BN film deposited at 1400 °C, 40 mbar, 0.9 sccm TMB and NH3/TMB ratio of 643. This can be compared to the B:N ratio of 1:0.98; O, H of 0.1 at%, 1 at%, respectively, and C being less than 0.1 at% (below detection threshold) for a film deposited at 1500 °C, 70 mbar, 0.7 sccm TEB and B/N = 643. 8 A reason for the increased carbon content in the film deposited from TMB may be the fact that TMB cannot undergo β-hydride elimination. This may restrict the removal of all ligands from the TMB molecule and result in a less favorable surface chemistry for the removal of the methyl groups, leading to the incorporation of carbon in the films. Similar trends have been observed for the pairs 13 trimethylaluminum/triethylaluminum and trimethylgallium/triethylgallium for AlxGa1-xAs 26 , GaAs 26 , InxGa1-xAs 27 and GaN 28 . ## IV. CONCLUSIONS We demonstrate a deposition process for epitaxial r-BN on α-Al2O3(001) from a reaction between TMB and NH3 in hydrogen. Epitaxial growth was achieved at 1300 °C and with the best conditions at a deposition temperature of 1400 °C. The

chemsum

{"title": "Thermal chemical vapor deposition of epitaxial rhombohedral boron nitride from trimethylboron and ammonia", "journal": "ChemRxiv"}

a_combined_photobiological-photochemical_route_to_c10_cycloalkane_jet_fuels_from_carbon_dioxide_via_

## Abstract: The hemiterpene isoprene is a volatile C5 hydrocarbon with industrial applications.It is generated today from fossil resources, but can also be made in biological processes. We have utilized engineered photosynthetic cyanobacteria for direct, light-driven production of bioisoprene from carbon dioxide, and show that isoprene in a subsequent photochemical step, using either near-UV or simulated or natural solar light, can be dimerized into limonene, paradiprene, and isomeric C10H16 hydrocarbons (monoterpenes) in high yields under photosensitized 2 conditions (above 90% after 44 hours with near-UV and 61% with simulated solar light). The optimal sensitizer in our experiments is di(naphth-1-yl)methanone which we use with a loading of merely 0.1 mol%. It can also easily be recycled for subsequent photodimerization cycles.The isoprene dimers generated are a mixture of [2+2], [4+2] and [4+4] cycloadducts, and after hydrogenation this mixture is nearly ideal as a drop-in jet fuel. Importantly the photodimerization can be carried out at ambient conditions. However, the high content of hydrogenated [2+2] dimers in our isoprene dimer mix lowers the flash point below the threshold (38 °C), yet, these dimers can be converted thermally into [4+2] and [4+4] dimers. When hydrogenated these monoterpenoids fully satisfy the criteria for drop-in jet fuels with regard to energy density, flashpoint, kinematic viscosity, density, and freezing point. Life-cycle assessment results show a potential to produce the fuel in an environmentally sustainable way. TOC graphic:Broader Context: The transportation sector is one of the major contributors to greenhouse gas emissions due to the use of fossil-based fuels. While the automobile industry is slowly shifting 3 towards electric vehicles, the aviation sector is still dependent on fossil fuels. To mitigate the environmental effects, biofuels have been introduced (in early stage of development) in the aviation sector. However, the biofuels production is dependent on biomass as a source of raw material which leads to competition with farmland and to rapid deforestation. Therefore, technology is needed to utilize CO2 as substrate for production of jet fuels, in a true carbon neutral process. Here, we report a combined photobiological and photochemical process for production of jet fuel equivalents, using CO2 as source of carbon and light as source of energy.A small hydrocarbon, isoprene, is produced by engineered photosynthetic cyanobacteria, and subsequently converted to C10 cycloalkanes by a photochemical process followed by catalytic hydrogenation. The C10 cycloalkane blends have all attributes to be used as drop-in jet fuels, ultimately enabling usage of the presently available infrastructure for aviation fuels, and lifecycle assessment (LCA) indicates that it would be possible to reduce climate impacts from jet fuel through this photosynthetic process. In the LCA, production of sodium nitrate dominated the impacts in all environmental categories, and therefore, use of an alternative nitrogen source from waste streams can be a potential solution for further reduction of the overall environmental impacts. ## Introduction In order to mitigate global warming and reach the goals of the Paris agreement, a shift towards carbon neutral fuels is necessary. For year 2050, the International Air Transport Association (IATA) emission reduction roadmap projects a reduction in CO2 emissions from aviation by 50% compared to 2005 levels. 1 This may seem modest, yet, globally air traffic increased by 4.5 -8.7% per year during the period 2009 -2019, 2 and a low annual increase of merely 4% until 2050, resulting from changes in travel patterns due to covid-19 and the installment of alternative transportation infrastructures, 3 still implies more than a three-fold increase in air traffic by 2050 when compared to 2019 and approximately six-fold when compared to 2005. As the increase in air traffic is often considerably steeper in growing economies, fulfilment of the IATA goal requires prompt technological development and introduction of new sustainable aviation fuels far beyond the biofuels currently in use or at the stage to be introduced on the market. Today, there are different technologies and feedstock alternatives to conventional jet fuels. An emerging route to biofuels goes via direct production of hydrocarbons by engineered photosynthetic microorganisms, such as algae or cyanobacteria. Cyanobacteria are photosynthetic bacteria which grow on water, minerals, and CO2 from the atmosphere, using sunlight as their energy source. Many cyanobacterial strains are amenable to genetic engineering, and thus, they are ideal hosts for biotechnological production of sustainable fuels. 11,12 Fossil-based jet fuels consist mostly of C8 -C16 hydrocarbons. More explicitly, they are mixtures of n-, iso-and cyclo-alkanes, small aromatics (< 25%) and alkenes (< 5%). 13,14 The fuel should be a proportional mixture of these compounds in order to follow the strict requirement for jet-fuels in terms of energy density, freezing point, and viscosity. In one typical jet fuel, JP-8, the proportion of C10 hydrocarbons is ~21%. 15 Hydrogenated monoterpenes (C10) and sesquiterpenes (C15) have long been considered as potential jet fuels due to their low viscosity and high energy density. Limonane (hydrogenated limonene) has been in focus among hydrogenated monoterpenes because of its availability from biomass fermentation and the low estimated cost of the resulting fuel (~0.73 USD/L). 16 Sesquiterpenes, e.g. bisabolene, farnesene and epi-isozizaene, are also molecules with potential utility. While biotechnological production of monoterpenes and sesquiterpenes has been demonstrated in various microorganisms, the toxicity of these compounds to the cells is often problematic. 19 Mono-and sesquiterpenoids tend to accumulate in the biological membranes, due to their hydrophobic nature, and interfere with their integrity and function. 20 On the other hand, smaller hydrocarbons, e.g., alkenes such as iso-butene and the 5-carbon-atom hemiterpenoids, are more volatile and tend to easily escape through the cell membranes. 21,22 Their diffusion to the extracellular environment makes them less toxic to the cells and their harvest/capture is less costly since there is no need for cell disruption. We, therefore, suggest a two-step procedure in which these small volatile hydrocarbons (C5 and smaller) are produced photobiologically, followed by their photochemical oligomerization in a second separate step. Isoprene is a volatile five-carbon hydrocarbon and can be an ideal precursor. It contains CC double bonds which are useful as sites for (photo)oligomerization, and its production by photosynthetic engineered cyanobacteria has been demonstrated. 21,23,24 Thus, hydrogenated isoprene oligomers could be ideal as drop-ins into presently used aviation fuels. There are already well-established chemical methods using heterogeneous catalysts common in industry for oligomerization of alkenes and dienes, 25 which require high temperatures and pressures. Recently, Harvey and co-workers reported iron-catalyzed dimerization processes of alkenes and dienes, including isoprene, that run at ambient temperature and pressure and that produce [2+2] and [4+4] cycloadducts (Fig. 1). 26,27 Interestingly, the hydrogenated [4+4] dimers of isoprene have better fuel properties compared to conventional jet fuels (Jet-A), and a life-cycle assessment and technoeconomic analysis showed that the process can be further improved to reduce cost and emission to compete within the sustainable aviation fuel sector. 28 The [2+2] oligomerization of isoprene was not selective for dimers since also trimers and tetramers were formed in significant amounts. Fig. 1 A) The two iron-based catalysts by Harvey and co-workers, 26,27 and B) the catalyzed oligomerization of isoprene. C) Photochemical dimerization of isoprene which resulted in the formation of [2+2], [4+2] and [4+4] photodimers. 29 Bonds formed in the reaction are marked in red. We have explored to what extent isoprene can be dimerized photochemically through triplet sensitizers using as mild conditions as possible, ultimately with solar light and in ambient conditions. The photochemical dimerization of isoprene was reported already in the 60s by Hammond, Turro and Liu using benzophenone (5 mol%) as photosensitizer (Fig. 1C), leading to 65% conversion to isoprene dimers when irradiated for five days in a sealed tube. 29 Interestingly, the composition of the dimer mixtures, i.e., the distribution of [2+2], [4+2] vs. [4+4] cycloadducts, depended on the triplet energies of photosensitizers, 30 yet importantly, trimers and longer oligomers were not formed. Combined with photosynthetic generation of isoprene from CO2, this could provide for sustainable production of hydrocarbons for jet fuels. Here it can be noted that there are only a few earlier studies on the direct production of jet fuels from CO2. An inexpensive heterogenous Fe-Mn-K catalyst prepared by the Organic Combustion Method was utilized for direct conversion of CO2 to jet fuel range hydrocarbons, with a CO2 conversion of 38% when run at 300 °C. 33 Recently, a model of thermochemical solar fuel production has been demonstrated where CO2 and H2O were captured from ambient air in a process that will be suitable for fuel production in desert regions. 34 Yet, we seek a process that requires as modest an energy input as possible. Hence, we now report on the first formation of C10 hydrocarbons, suitable as jet fuel drop-ins after hydrogenation, in a combined two-step photobiological-photochemical approach with CO2 as carbon source and with light, either as (simulated) solar or ambient light, as the energy source. To ensure a sustainable production route, a system analysis perspective is needed as it allows us to understand the different impacts of the product throughout its entire life cycle. 35 Today, life cycle assessment (LCA) is employed as the main decision-support tool for implementing renewable energy technologies using a holistic framework, and several earlier studies have assessed the environmental impacts of biofuel production from microalgae using LCA. Furthermore, it has been shown that algae-derived biodiesel is the most efficient alternative in terms of land use as it avoids competition with food crops. 48,49 The environmental impacts of producing cyanobacteria-based biofuels have also been assessed. 38,40,50 Both Luo et al. and Quiroz-Arita et al. employed LCA to assess the life cycle energy and greenhouse gas (GHG) emissions of ethanol production via cyanobacteria, 40,50 and revealed that the ethanol purification process was the main energy consumer and a significant contributor to the carbon footprint of the process. Nilsson et al. assessed the environmental impacts of photosynthetic butanol production by genetically engineered cyanobacteria, 38 and found that in order to displace fossil fuels using butanol produced by cyanobacteria, significant metabolic engineering based improvements in carbon and energy conversion efficiency per cell are necessary. As the process reported herein is based on a volatile product which spontaneously separates from the cell culture, we can eliminate the energy requiring distillation or processing of biomass, in contrast to ethanol and larger alcohols as well as direct biodiesel production. The process resembles a previously envisioned strategy on catalyzed oligomerization of ethylene produced by cyanobacteria, which was explored in a technoeconomic analysis study and revealed to yield economically viable biofuels in the long term. 51 We used LCA to assess the different environmental impacts of jet fuel production through the combined photobiologicalphotochemical route in order to identify the hot spots and improvement options. Our results should aid the further development of the novel emerging technology presented herein as it pinpoints the hurdles that need to be addressed, and thus, enable a faster realization of the technology at large scale. ## Results and Discussion The photobiological formation and trapping of the isoprene produced by the cyanobacteria are presented first, followed by optimization of the photoinitiated dimerization of isoprene (including bio-isoprene) to yield C10 hydrocarbons (monoterpenoids). The dimerization mechanism is analyzed through density functional theory (DFT) computations, unravelling why isoprene trimers are formed in only trace amounts. To be useful as a fuel, the monoterpenoids formed need to be hydrogenated and we determine various properties and assess the values of our hydrogenated monoterpenoids in relation to what is required for a jet fuel. We also carry out a life cycle assessment in order to pinpoint the different environmental impacts of bio-jet fuel production and to identify the related hot spots and improvement options. The results of the study will facilitate further development of the emerging technology presented. ## Microbial production and trapping of isoprene: Cyanobacteria, like other bacteria, are able to generate terpenoids via the methylerythritol-4-phosphate (MEP) pathway, but do not naturally produce isoprene (Fig. 2A). 52 In previous work, we have established engineered strains of the unicellular cyanobacterium Synechocystis sp. PCC 6803 (hereafter Synechocystis), capable of light-driven isoprene production from CO2, via photosynthesis. This was achieved through the introduction of genes encoding an efficient isoprene synthase (IspS) and two enzymes upstream in the MEP pathway -DXS, 1-deoxy-D-xylulose-5-phosphate synthase, and IDI, isopentenyl-diphosphate isomerase (Fig. 2A). 24 DXS performs the first step of the pathway by combining the two substrates pyruvate and glyceraldehyde-3-phosphate to form 1-deoxy-D-xylulose 5-phosphate (DXP). IDI performs the interconversion of isopentenyl diphosphate (IPP) and dimethylallyl diphosphate (DMAPP), the substrate for the isoprene synthase to form isoprene. 24,52 The reaction catalyzed by DXS includes a decarboxylation step, thereby serving as a gateway for the flux of carbon into the MEP pathway. The expression of IDI is likely necessary to maintain the balance between IPP and DMAPP, and thus enable the synthesis of essential terpenoids downstream in the terpenoid biosynthesis, when IspS expression would otherwise deplete the levels of DMAPP in the cell. 24 Here, we have used the engineered Synechocystis cells for photosynthetic production of isoprene in small-scale cultures. 20 mL of cyanobacterial culture were grown for four days in sealed 60 mL culture tubes under a constant illumination of 50 μmol photons m -2 s -1 , with addition of 50 mM NaHCO3 to the culture medium. Thereafter, the headspace gas was drawn through 20 mL of cold heptane to capture produced bio-isoprene from the cultures (Fig. 2B). Isoprene concentrations in the gas phase of the cultures were determined by gas chromatography comparing to an isoprene standard, before and after capturing of the gas phase. For further experimental details, see Fig. S1, ESI †. We achieved an isoprene titre of 1.60 mg L -1 culture after four days of cultivation under the abovementioned conditions. After capturing the isoprene in heptane in our customized trapping setup, the equivalent of 935 µg L -1 of culture remained in the cultivation tube, which translates into a capture efficiency of 41.4% (Fig. S2, ESI † and Table S1, ESI †). A second cycle of trapping resulted in the capture of ca. 490 µg L -1 culture and a higher efficiency (52.4%), for a combined trapping efficiency of ca. 70%. Additionally, we achieved higher capture efficiencies in a single trapping step for other tests, reaching as high as 89% of the isoprene produced. The bio-isoprene trapped in the heptane of the collector tubes was then used for the photochemical dimerization experiments (see section below on Photodimerization of bio-isoprene). Throughout the experiments, we observed variability in the isoprene production by the engineered strain, likely due to genetic instability of the expression constructs. In order to improve long term isoprene production, we therefore generated another strain of Synechocystis, where the genetic constructs conferring ability to produce isoprene are expressed from the cyanobacterial chromosome rather than from a plasmid. This was achieved by integration into the slr0168 neutral site in the genome (Fig. 3A). 53,54 The resulting strain, ΔNS1::2MEP-EgIspS, was evaluated for isoprene production which was found to be stable for at least several weeks of cultivation (data not shown). Furthermore, since the isoprene production is performed in closed vessels where isoprene accumulates in the headspace, we hypothesized that over time the concentration of isoprene and oxygen in the culture tubes may become inhibitory for cell growth and productivity. We therefore performed a set of experiments where the headspace gas was vented from the cultures at different intervals. In these experiments, closed cultures of ΔNS1::2MEP-EgIspS were grown for 6 days with sampling and removal of the gas phase at 12, 24, 48 or 72 hour intervals, and growth and isoprene production was evaluated (see Fig. S3). In cultures with more frequent venting of the gas phase (12-48h cycles), growth as well as productivity continued for a longer time period, and total cumulative isoprene production and rates of production was higher than in cultures which was vented every 72h (Fig. 3B and C, Table S2, ESI †). Regardless of the periodicity of these cycles, the cumulative amounts of isoprene were always higher than when no cycling was applied. These results are in agreement with previous reports on butanol and isobutanol production in cyanobacteria, where semi-continuous cultivation with frequent dilution resulted in prolonged and enhanced productivity of the cultures. 55,56 The strategy of continuous or fed-batch cultivation with frequent product removal is thus a potential avenue for developing the isoprene production process on larger scale. ## Screening of triplet sensitizers: To establish a photochemical isoprene dimerization process that utilizes solar irradiation (natural or simulated) we started at the triplet sensitized diene dimerization reported by Hammond, Turro and Liu in the 60s. 17 Arylketones are excellent photosensitizers due to their relatively high triplet quantum yields and exceptional photostability. The excitation wavelength of arylketones can be tuned to the visible region by extension of π-conjugation of the aryl groups. Additionally, the triplet quantum yield of ketones can be greatly improved compared to the corresponding arene chromophore. 58 Such modulations push the excitation of the sensitizers toward the visible wavelength region where they can be activated by solar light (see below). In the screening of photosensitizers suitable for photodimerization of isoprene we used benzophenone (9), xanthone (10), thioxanthone (11), di(naphth-1-yl)methanone (12), naphthalen-1-yl(naphth-2-yl)methanone (13), and di(naphth-2yl)methanone (14), see Fig. 4. The synthesis of the photosensitizers is discussed in the ESI †. The triplet energies (E(T1)) of 9 -14 and isoprene, both experimentally determined and calculated using density functional theory (DFT) at the (U)B3LYP-D3/6-311+G(d,p) level, indicate that these ketones are suitable for effective photosensitization because their E(T1)'s are slightly higher than that of isoprene (Fig. 4 and Fig. S4, ESI †). Furthermore, the T1 states of dinapthylketones (12-14) are of * character which prevents the competing H atom abstraction, 63 a photoreaction that many ketones with T1 states of n* character initiate. In a typical photoreaction, a mixture of inhibitor-free isoprene and aryl ketone was contained in a quartz test tube under argon and irradiated with 365 nm light (Fig. S5, ESI †). The solution was stirred during the photoirradiation in order to achieve uniform light exposure. ## Fig. 4 The photosensitizers used in this study as well as isoprene, and in parenthesis, their experimental triplet energies (kJ/mol, in red) and the calculated adiabatic triplet energies (kJ/mol, in blue) at (U)B3LYP-D3/6-311+G(d,p) level. 30, The isoprene dimers formed were characterized by 1 H nuclear magnetic resonance (NMR) and gas chromatography-mass spectrometry (GCMS) analysis (Fig. S6 -S9, ESI †). However, we confirmed the structure of the isomers by 1 H NMR as the GC chromatograms can give erroneous results on the relative product distribution due to thermal rearrangement of the dimers (see below). Seven isomeric isoprene dimers (2 -8) were observed, in line with findings reported by Hammond, Turro and Liu (Fig. 1C and Fig. S9, ESI † ). 29 It was also proposed by Hammond and Liu that cyclooctadienes 7 and 8 might have resulted from thermal rearrangements in the GC, 67 but our 1 H NMR data of the isoprene dimers (purified by silica gel column by using pentane as eluent) reveals that these two dimers originate from photoinitiated dimerization and cyclization. Here it can be noted that the distribution of the various isomers depends on the E(T1)'s of the photosensitizers used. It is also noteworthy that trace amounts of isoprene trimers were formed, but not any longer oligomers (Fig. S7, ESI †). The screening of the photosensitizers was performed by using 2 mol% loading, unless otherwise mentioned in Table 1. Depending on the photosensitizer, with the quartz tube setup (Ø 13 mm, Fig. S5, ESI †) we observed 8 -41% conversion to isoprene dimers with di(naphthalen-1-yl)methanone 12 giving the highest conversion. A control experiment carried out without photosensitizer clarified its crucial role as the conversion dropped to 0.5% after 44 h of irradiation with  = 365 nm (Fig. S10, ESI †). a The actual loading was lower due to poor solubility of the sensitizer in isoprene. Interestingly, the efficiency of the three dinaphthylmethanone isomers (12-14) to convert isoprene to its dimers varied from 21 to 44% due to the positional effect of naphthyl groups. Thus, the isomeric dinaphthylmethanones (12) acts as a better photosensitizer than benzophenone (9), while similar yield of isoprene dimers could be obtained with 13, and comparably the lowest yield could be obtained when 14 was used. If we compare the relative absorbance of the benzophenone (9) and the three dinaphthylmethanone isomers (12-14) at 365 nm, the maximum molar extinction coefficient is observed for 13 and minimum for 9 (Fig. S11-12, ESI †), and from the E(T1)'s of 12 -14 (Fig. 4) it is clear that 12 is the dinaphthylmethanone with a triplet energy closest to that of isoprene. Additionally, the absorption tails of the dinapthylmethanones go beyond 400 nm, which possibly enable solar light photosensitization. As a result, the isoprene photodimerizations using dinaphthylmethanone sensitizers can be run with very low sensitizer loadings and as they absorb solar irradiation, it is apparent that particularly 12 is a suitable photosensitizer. The yields of isoprene dimers when xanthone 10 and thioxanthone 11 were used as photosensitizers were significantly lower as compared to when benzophenone (9) was used, and we initially considered this to arise from their poor solubility in neat isoprene. To improve the solubility, we designed and synthesized 3,6-di(octyloxy)xanthenone (15) with solubilizing alkyl groups (for synthesis see ESI †). Yet, despite an improved solubility, the improvement in the isoprene-to-dimer conversion is minute (from 8 to 11%). Instead, the higher E(T1) of both 10 and 15 compared to 9 may cause less efficient triplet energy transfer to isoprene and, consequently, a less efficient isoprene dimerization. Indeed, the calculated triplet energy of 15 is higher than that of 10 by 10.9 kJ/mol, revealing that substitution allows for further tailoring of xanthone-type sensitizers, similarly as recently reported by Booker-Milburn and coworkers. 68 Optimization of dinaphthylmethanone sensitized dimerization: Having identified the most suitable photosensitizers among those selected, we determined the loading of 12 required for the optimal conversion of isoprene to its dimers. The photosensitizer loadings were screened from 0.5 down to 0.01 mol% with a similar setup as used above (see Table S3, ESI †). We could observe 21% yield of isoprene dimers in 44 h with the loading of 12 as low as 0.01 mol%. It is worth noting that the yield of the isoprene dimers does not correlate linearly with the loading of 12 as the light transmission through the solution, which is a function of the sensitizer concentration, influences the yields. We found that a loading of 12 of 0.1 mol% was adequate to get an optimized yield of the isoprene dimers. Additionally, we re-screened all photosensitizers (9-15) at 0.1 mol% concentration and the results confirmed that 12 was the most efficient photosensitizer at this concentration (Table 1). Further improvement of the photodimerization was carried out in modified reaction setups. We first used a fluorinated ethylene propylene polymer (FEP) tubing (outer diameter: 3.2 mm, ~120 mL loop size) coiled around a water-cooled jacketed beaker (Fig. S13, ESI †). The FEP tubing setup extensively increased the surface area for the incident light, which in turn improved the light absorption. The water-cooled beaker also allows the reaction to run at ~10 ºC which prevents evaporation of the volatile isoprene. With this setup and with 0.1 mol% of loading of 12, we observed 89% yield of isoprene dimers (120 mL scale) when photo-irradiated for 44 h. We attempted to scale up the reaction to 400 mL by using wider FEP tubing (outer diameter: 7.9 mm) coiled around the water-cooled jacketed beaker (Fig. S13, ESI †) and we observed a 48% yield of isoprene dimers when using the reaction conditions described above. Here, the lower yield can be attributed to the increased tube diameter which prohibits an equal light distribution over the width of the tube. The isomer distribution between the isoprene dimers, as quantitatively determined through the 1 H NMR spectrum, were found to be: 30 The lower triplet energy of dinapthylmethanone 12 than of 9 leads to preferential activation of s-cis isoprene, resulting also in high amounts of [4+2] cycloadducts (42.6%). The isoprene dimers and photosensitizer 12 could easily be separated by passing through a short silica gel column by using pentane as eluent or by distillation under reduced pressure (65 ºC at ~0.1 mmHg). The isoprene dimers could be stored at 4 °C for a few months without noticeable decomposition (Fig. S15, ESI †). However, the conversion of kinetically stable [2+2] photodimers to the other thermodynamically more stable dimers was observed after a few months in storage (Fig. S16, ESI †) or upon heating over 100 °C in air. Also noteworthy is that under ambient conditions the photodimers tend to convert slowly over time to the corresponding immiscible epoxides and alcohols (Fig. S17-18, ESI †). Now, can the photochemical formation of isoprene dimers be run under ambient atmosphere? To explore this, we analyzed the photodimerization with the aforementioned setup (120 mL) and photosensitizer content for 44h under ambient conditions and we observed the same yield (86%) as before. The improved photosensitizing efficiency of 12 compared to benzophenone 9 is attributed to the higher absorption at 365 nm (Fig. S11, ESI †), lowest triplet energy difference as well as higher photodimerization quantum yield (ϕ = 0.91 for the dinaphthyl methanone 12 versus ϕ = 0.43 for benzophenone 9, see ESI † for details). It is also noteworthy that 12 is straightforwardly synthesized in a one-pot reaction using readily available and inexpensive reagents, and after the photoirradiation it can easily be recovered (up to 95%), purified, and used for another cycle. Finally, very low amounts of 12 as photosensitizer (0.1 mol% loading) are needed, which together with its recyclability, should significantly reduce the cost for large-scale production of isoprene dimers. ## Dimerization induced by (simulated) solar irradiation: Our ultimate goal is to carry out the photodimerization of isoprene with solar irradiation (Fig. S21, ESI †). Dinaphthylmethanone 12 might be an ideal photosensitizer as its absorption tail stretches until ~400 nm and the solar irradiation has significant light intensity at the surface of Earth at wavelengths longer than 350 nm (Fig. S22, ESI †). For this reason, we first performed the isoprene photodimerization in a solar simulator (1 sun, AM 1.5G) using a newly designed flat spiral coil made of FEP tubing for simulated solar irradiation of isoprene (Fig. 5). Now, we could obtain 61% yield of isoprene dimers (4 mL scale) when irradiated in the solar simulator for 44 h using 0.1 mol% of dinaphthylmethanone 12 as photosensitizer (Fig. S23, ESI †). Thus, the experiments demonstrate that the formation of isoprene dimers under sunlight irradiation is achievable. Furthermore, the higher yield that can be estimated after 20 h in the solar simulator (28%) can be rationalized by the fact that the solar simulator has a higher relative intensity in the 350 -400 nm range when compared to natural solar irradiation (see Fig. S22, ESI †). ## Photodimerization of bio-isoprene: The bio-isoprene produced by the Synechocystis cells and captured in heptane was mixed with dinaphthylmethanone 12 (0.02 M), filled into the flat spiral coil and irradiated in the solar simulator (24 h, 1 sun, AM 1.5 G). Even though the concentration of bio-isoprene was low, the reaction produced bio-isoprene dimers as confirmed by GCMS (Fig. 6), and experiments with commercially available isoprene (0.05 M solution in heptane) gave a similar distribution pattern of dimers (Fig. S26-27, ESI †). This proof-of-principle experiment shows the possibility to turn CO2 used as carbon source into C10 cycloalkanes with our combined photobiological-photochemical approach. Bio-isoprene dimerization was also attempted under natural sunlight, yet, no dimers were detected in GCMS. This might result due to two factors; (i) the weaker intensity of the natural solar light compared to the simulated one in the 350-400 nm range (Fig. S22, ESI †), and (ii) the low concentration of the bio-isoprene in heptane. Thus, one next step is to increase the production of bio-isoprene so that a higher concentration can be achieved. This may be addressed via further metabolic engineering of the cyanobacterial strain to enhance flux of fixed carbon towards the isoprene product combined with more efficient trapping of isoprene from the culture. ## Photodimerization mechanism: The reaction mechanism for light-induced formation of the isoprene dimers involves six steps (steps 1-6, Fig. 7) which we explored through DFT computations at the (U)B3LYP-D3/6-311G(d,p) level (for details on the computations and for additional results at M06-2X/6-311G(d,p) level 62,69 see the ESI †). The first step is the excitation and intersystem crossing (ISC) of the photosensitizer to its triplet state, followed by triplet energy transfer from the sensitizer to isoprene in the ground state, yielding isoprene in its T1 state. The T1 state isoprene can be described as a radical-pair composed of one resonance stabilized allyl radical and one methyl radical, and as such it exists in four different conformers with nearly equal energies, yet, separated by activation barriers of 63 -67 kJ/mol. One molecule of isoprene in its T1 state can add to an S0 state isoprene via a number of reaction paths. Among these, the addition of the methyl radical site of a T1 state isoprene molecule to an S0 state isoprene molecule proceeds over slightly lower activation barriers (step 3, lowest barrier ~56 kJ/mol) than the addition of the allyl radical part of T1 isoprene to an S0 state isoprene (lowest barrier ~61 kJ/mol). The triplet lifetime of isoprene has been determined to 5 s, 30,70 sufficiently long to allow a substantial amount to overcome the activation barrier for dimerization. The additions, which are markedly exergonic (-92 to -71 kJ/mol), lead to intermediate isoprene dimers that can be described as triplet state bis(allyl) radical pairs. Thus, once formed there will be no back reaction. As the two radical sites of the bis(allyl) radical pair are only weakly coupled, the singlet diradical is essentially isoenergetic with the triplet, and a rapid ISC should occur (step 5). Furthermore, the bis(allyl) radical pairs have high conformational flexibilities irrespective of electronic state because the conformer interconversions involve C-C single bond rotations (in the T1 state the rotational barriers are ~16 kJ/mol, step 4). Finally, when a singlet state bis(allyl) radical pair adopts a conformer with the two unpaired electrons at a sufficiently close distance they will combine into a C-C single bond (step 6), leading to the observed isoprene dimers with either cyclobutane, cyclohexene or cyclooctadiene rings (Fig. 1C). ## Fig. 7 The various steps in the reaction mechanism for the formation of the cyclic isoprene dimers (steps 1 to 6) and trimers (steps 7 and 8) with the lowest activation energies at UB3LYP/6-311G(d,p) level. ISC = intersystem crossing. For further details see the Supporting Information, section 6. So why is further oligomerization hampered? As the bis(allyl) radical pairs are composed of two allyl radicals which are internally stabilized through -conjugation, they will be less reactive than triplet state isoprene which can be described as one allyl radical and one reactive methyl radical fragment. Thus, the rate for the addition of the bis(allyl) radical pair to an isoprene in its S0 state, leading to a trimer bis(allyl) radical pair, should be slow (step 7). Indeed, the lowest activation barrier for the addition of the bis(allyl) radical pair to an S0 state isoprene is 76 kJ/mol, significantly higher than the addition of a T1 state isoprene to an S0 state isoprene (56 kJ/mol as seen above). A second potential route to trimers goes via addition of an T1 state isoprene to a C-C double bond of a cycloadduct (step 8), but this process should also be slow as it leads from a single carbon-centered radical to another. For this process we find a lowest calculated activation energy of 71 kJ/mol. Together with the fact that the ring-closure of the dimer bis(allyl) radical pair is a unimolecular reaction in contrast to the bimolecular reaction to trimer bis(allyl) radical pair, this explains why the further oligomerization to trimers, tetramers etc. is not competitive with the closure of the bis(allyl) radical pair to the cyclic dimers observed. Finally, since the combined portions of isoprene dimers that are either [2+2] and [4+4] cycloadducts make up more than half of the dimer mix, we also tested a T1 state concerted mechanism that would involve a transition state with a cyclic array of 4n electrons stabilized by through-space Baird-aromaticity, 71-73 however, we could not locate such a pathway. For further discussions, see ESI †. ## Hydrogenation and fuel performance: The isoprene dimers are unsaturated, which is not ideal if they should function as a jet fuel as soot would form due to incomplete combustion when ignited. The isoprene dimers (here labelled ID-1) were therefore hydrogenated in presence of Pd/C as a catalyst at 10 atm hydrogen pressure, providing hydrogenated isoprene dimers (HID-1) in near quantitative isolated yields (see ESI † for detail procedure). These hydrogenated isoprene dimers appeared as a colorless liquid (Fig. S28, ESI †), and they were further characterized by 1 H NMR and GCMS analysis (Fig. S29, ESI †). The disappearance of the alkene signals of the isoprene dimers in the 1 H NMR spectrum proves a complete reduction of the C-C double bonds, leading us to the cycloalkane-based jet fuel equivalent. For this mixture of hydrogenated isoprene dimers, we determined the key fuel properties, i.e., the net heat of combustion (NHOC), kinematic viscosity, density, and flash point (Table 2). The measured density of HID-1 is 0.77 g/mL at 15 °C (Table S7 and Fig. S40, ESI †) which matches well with the lower required density of Jet-A. The density of the fuel is lower than that of dimethylcyclooctanes (DMCO) due to the presence of high amounts of isomers with cyclobutane rings. Moreover, the hydrogen content of the HID-1 (14.37%) is significantly higher than that of Jet-A due to the absence of aromatic and unsaturated moieties, which eventually gives a higher gravimetric NHOC value and produce clean burn without soot formation. The gravimetric NHOC is an important parameter for a jet fuel, and it should be above 42.8 MJ/kg according to the standard specification for jet fuels. 26 Additionally, the volumetric NHOC value of HID-1 is higher than that of conventional jet fuels (Jet-A). For the two C10 hydrocarbons (18, 19, 25 and 26) in Fig. 8 which have experimentally determined NHOC, 15,26 we find that computed values calculated with a DFT-based procedure by Major and co-workers 74 are in good agreement (for a further description of the method see the caption Fig. 8 and the Supporting Information). Thus, based on the computed NHOC of the C10H20 hydrocarbons contained in HID-1 we can also conclude that their energy contents are in line with expected for an aviation fuel. The NHOC values were computed following a DFT-based procedure by Major and co-workers developed for the M06-2X functional. 74 These values contain two corrections which are needed to achieve accuracy; (i) a correction for the addition of the enthalpy of vaporization of terpenes and water, and (ii) a correction of the enthalpy of O2. The enthalpies of vaporizations were calculated using the SMD solvation model. 75 Additionally, we have measured the kinematic viscosity of HID-1 from -40 °C to 20 °C as it is an important parameter in terms of safety and combustion of the fuel. 76 A higher viscosity leads to a poorer atomization of the fuel which leads to incomplete combustion and formation of soot, eventually reducing fuel efficiency. To achieve proper atomization and combustion of a jet fuel it is strongly recommended to have a kinematic viscosity value below 12.00 mm 2 /s at -40 °C. Rewardingly, the kinematic viscosity of HID-1 (1.71 mm 2 /s at -20 °C) is more than 4.5 times lower than the recommended value for conventional fuel (8.00 mm 2 /s), and it is even 2.4 times lower than that recently reported for DMCO (4.17 mm 2 /s at -20 °C) which is closely related to the structure of the molecule (C10). The kinematic viscosity at -40 °C is 2.60 mm 2 /s (Table S5 and Fig. S39, ESI †), which is 4.6 and 3.1 times lower when compared to Jet-A and DMCO (7.95 mm 2 /s), respectively. The lower kinematic viscosity might result from the higher portion of alkylated cyclobutane isomers over cyclooctane isomers, and it will allow the drop-in to be blended with other conventional jet fuels at any ratio. The freezing point of the jet fuel is also crucial for the safety and the flow of the fuel at high altitudes. We assessed the freezing properties of HID-1 by placing it in a dry ice/acetone bath (-78 °C) for 1.5 h and did not observe any cloudiness or crystallization, indicating that the freezing point of HID-1 is lower than -78 °C, i.e., it is much lower than the recommended value for conventional jet fuel (-40 °C). The low freezing point of HID-1 suggests that it is possible to use as a fuel in high altitude flight. Yet, a drawback of HID-1 is the flashpoint which was found to be 33.5 °C, lower than the specified value for conventional jet fuel (38 °C). The lower flash point may limit the use of HID-1 as jet fuel surrogate due to safety issues, although the commercially available Jet-B and TS-1 have much lower flash points (-18 and 28 °C, respectively) compared to the recommended value. 77 Yet, these fuels have very low freezing points allowing them to be used in extremely cold environments. The low flash point of HID-1 can be attributed to the isomers with cyclobutane rings as these are more volatile. ## Further modification of the C10 fuel: The fact that the flash point is slightly below the recommended value prompted us to consider modifications of the isoprene dimer mix ID-1 before the hydrogenation step. The boiling points of the various isomeric isoprene dimers (2 -8, Fig. 1C) were earlier reported by Hammond, Turro and Liu and it was revealed that the [2+2] isomers have relatively lower boiling points than the others (Fig. S14, ESI †), 29 with 2 having the lowest. This should also contribute to the low flash point of HID-1 as the flash point of a hydrocarbon correlates with its vapor pressure. A further modification of ID-1 could be performed through moderate heating which led to the conversion of cyclobutane-containing isomers to cyclooctadiene-and cyclohexene-containing ones through Cope and other thermal rearrangements. 29 Here we probed two different temperatures, 135 and 160 °C, and subsequent hydrogenation gave the modified hydrogenated isoprene dimers HID-2 and HID-3 (see †ESI for detailed synthetic procedure, Fig. 9). The reaction mixtures were analyzed by 1 H NMR and GCMS measurements (Fig. S30-33, ESI †). When ID-1 is heated at 135 °C for 1.5 h, leading to ID-2, the isomer 2 rearranges to isomers 5 and 8, where isoprene is formed as a byproduct to 5 % (Fig. S34, ESI †). In order to transform all [2+2] isoprene dimers into [4+2] and [4+4] isomers the temperature had to be elevated to 160 °C for 4 h, giving ID-3. Yet, in this case the amount of isoprene formed through a back-reaction increased to 11 %, even though 3 and 4 after prolonged heating remained in the post-modified ID-3 in trace amounts of 1 % and 2 %, respectively (Fig. S35, ESI †). It is worth noting that the post-modification of ID-1 can be justified, as the isoprene formed as a byproduct can be recycled. After removal of isoprene from ID-2 and ID-3, these dimer mixtures were 29 hydrogenated using the conditions described above leading to quantitative formation of HID-2 and HID-3 (Fig. S36-38, ESI †). Here it is noteworthy that the hydrogenation of isoprene dimers (ID-3) could be run at 1 atm H2 pressure to obtain HIDs (HID-3) in quantitative yield. However, the reaction requires longer time (48 h) to complete and 1% p-cymene is formed due to aromatization of limonene (Fig. S69, ESI †). ## Fig. 9 Isomerization of the cyclic [2+2] isoprene dimers to plausible cyclic [4+4] and [4+2] isomers through thermal Cope and other rearrangements. After the heat treatments, the flash points of HID-2 and HID-3 increased to 38.5 °C (Table 2), i.e., above the recommended value. The identical flash point of HID-2 and HID-3 can be rationalized as they are mixtures of hydrogenated cycloalkanes with very similar boiling points. The gravimetric NHOC values of HID-2 and HID-3 decreased to 43.57 and 43.59 MJ/kg, respectively, lower than that of HID-1 which is explained by the reduced amounts of cyclobutane isomers in the modified HID blends. Yet, the modified HID's have higher densities (both 0.809 g/mL at 15 °C) (Table 1, S7 and Fig. S40, ESI †) which leads to higher volumetric NHOC values (35.25 and 35.22 MJ/L, respectively). The volumetric NHOC values for modified fuels are 6.3% greater compared to conventional Jet-A (> 33.17 MJ/L), which should be an added advantage. With regard to the kinematic viscosities (3.16 and 2.92 mm 2 /s at -20 °C for HID-2 and HID-3, respectively) these are higher than that of HID-1 due to their lower contents of cyclobutanes (Table 1, S5 and Fig. S39, ESI †). Still, the values are more than 2.5 times lower than the largest recommended values, facilitating a good atomization of the HID's when used as fuels. Finally, both modified fuels have very low freezing points (<-78 °C), enabling high altitude flight (Table 1). The easy modulation of the ID-1 to ID-2 and ID-3 should be an advantage as they after hydrogenation should be ideal as drop-ins for conventional fuels for high-altitude jet engines. There are also further favorable features of HID-1 -HID-3. Conventional jet fuels contain mixtures of aromatic compounds which have added benefits as they swell the nitrile rubber elastomer valves which helps to protect the integrity of the jet engine. However, modern elastic materials do not require the aromatic content to swell the elastomers, and recent studies have shown that cycloalkane blends have similar properties as aromatics and are able to swell nitrile rubber elastomer valves. 78,79 Additionally, the content of aromatic compounds in jet fuels leads to lower NHOC values as well as formation of carbon soot during the combustion which adversely affects the lifetime of the engine. Finally, aromatic compounds in jet-fuels are major health and environmental hazards. Thus, avoidance of such compounds is favorable for these reasons, and substantial interests have been focused towards development of biocycloalkane based fuels that mitigate the abovementioned problems. 80 The very recent review by Muldoon and Harvey further highlights the potential of bio-cycloalkane based hybrid fuels for future use in military and civilian aviation fuel industries. 80 In this context it can be noted that JP-10 (exo-tetrahydrodicyclopentadiene) is a synthetic C10 cycloalkane-based missile fuel. 81,82 Taken together, our jet fuel mixtures (HID-1 to HID-3), which are C10 cycloalkanes, fulfil all requirements for future, less environmentally hazardous jet fuels, they are devoid of aromatic content and have high NHOC values. ## System, efficiency and scale-up potentials: The emerging technology reported here is at a very early stage of development (approx. at technology readiness level 2 (TRL2)). A technoeconomic analysis is therefore not yet meaningful. To clarify where future research and development need to focus, we instead identify technological challenges by using information from recent analyses of approaches that resemble our combined photobiologicalphotochemical one. We also performed a life-cycle assessment (see below). In order to develop this platform into a commercial production system which is both energetically and economically sustainable, extensive improvements in performance are necessary on several levels. For the photosynthetic production of isoprene, the conversion from solar energy and CO2 to product needs to be more efficient. This will require further engineering of the host organisms, for improved photosynthetic efficiency and increased carbon fixation as well as for increased partitioning of carbon towards product formation. Furthermore, cultivation conditions need to be optimized for cell productivity. Cultivation and harvest systems also need to be further developed. While photobioreactors are commercially produced, albeit still mostly at smaller scale, efficient harvesting of a volatile product from the culture remains a challenge to solve. A technoeconomic analysis of ethylene production by cyanobacteria has earlier been reported, 51 and it was estimated that gasoline-equivalents, produced by oligomerization of the bioethylene, could be sold at a price of $28.66 per gallon in the near-term and at a price of $5.36 in the long-term. The largest cost that determined the gasoline price was the capital investment for the photobiology reactors, followed by the electricity cost for the power intense cryogenic distillation. Isoprene, contrary to ethylene, will not require cryogenic distillation as it condenses at much higher temperatures than ethylene. A further difference is the subsequent oligomerization which in case of ethylene uses a Ziegler catalyst, a mature technology utilized widely within the petrochemical industry. Our photochemical dimerization of isoprene is not an established technology and needs extensive process development, yet, if carried out with natural solar light it will be much less power demanding than the catalytic approach for ethylene oligomerization. The efficiency of the photochemical step is such that we can assume that all isoprene produced photobiologically within one day can be dimerized photochemically within the same amount of time. Thus, the main limiting factor for the photoproduction is the photobiological production step. One drawback of our first strain of cyanobacteria used for isoprene generation was genetic instability of the plasmid-borne DNA construct. We successfully circumvented this by instead inserting the genes required for isoprene production into the genome of the host cyanobacterium (Fig Fig. 3). This enables long-term stable production of isoprene, opening the possibility for continuous cultivation of the production strain for longer time periods. In a fully developed system at large scale, fed-batch or continuous cultivation combined with continuous product removal has the potential to increase productivity of the culture, while further strain engineering to enhance the productivity per cell will also be necessary. As described above, the photochemical dimerization can be performed to very high yields (~90%) in batch setup using thin FEP tubing, yet the yield decreases when the tube diameter increases. Process optimization in which various conditions are varied (flow rate, irradiation intensity, tubing width, laminar vs. tubular flow, and reactor design) is required. One may also search for photosensitizers with smaller S1 -T1 energy gaps than compound 12, yet still with E(T1) above that of isoprene. Such sensitizers could absorb within the visible (blue) wavelength region of the solar spectrum where the intensity is higher and still be able to transfer the triplet energy to isoprene and initiate the dimerization. ## Life cycle assessment: To assess whether large scale production of photosynthetic jet fuel according to our system may become an environmentally sustainable process, we have performed an LCA for the integrated photobiological-photochemical process, using one tonne of fuel as the functional unit (Fig. 10). For the cultivation and production of isoprene from cyanobacteria, we have used as a starting point a scenario described by Nilsson et al., 38 where the authors modelled cyanobacterial butanol production. In this system, cultivation takes place in a 750 m 3 array of serially connected vertical flat panel photobioreactors, covering 1 ha of land. We assume that the cyanobacterial cultivation would be performed in two phases. First, a pre-cultivation in 10% of the whole volume for five days to generate biomass, during which period product formation is inhibited. Second, the biomass is transferred to the reactor volume for a production phase of three weeks where production is induced, and 90% of fixed carbon is directed to isoprene production in the cells. Isoprene product is continuously removed and transferred to the downstream photochemical process. We make the following assumptions: carbon fixation is at 1.2 g L -1 day -1 ; 83 inorganic carbon is supplied from a waste resource such as biomass combustion, thus providing a carbon source at no environmental cost to our system; 80% of the water from the reactor plant is recycled after each production round; electricity needed is supplied in accordance with the Swedish energy mix. In the scenario we modelled, nutrients other than CO2 are supplied based on the composition of BG11 growth medium. 84 With the mentioned carbon fixation rate and 90% of carbon allocated to isoprene formation, 38 the time of operation of the 750 m 3 reactor for generating one tonne of jet fuel at the end of the process is 2.4 days. The bio-isoprene produced will subsequently be dimerized photochemically, and in our modelling, we utilised the input from the lab scale experiment and scaled up to produce 1 tonne of HID-2. Upon solar irradiation of isoprene (60 h) in presence of 12 as a photosensitizer to obtained isoprene dimers in 51% yield. The unreacted isoprene is distilled off to be used in the next cycle, while the isoprene dimers are separated by distillation under reduced pressure (~70 °C, 10 mmHg pressure). The photosensitizer is easily recovered from the residue by washing with pentane and methanol (~95% recovery, see above), and it was therefore excluded from the LCA since merely 0.1 mol% was used in the photoreaction. Further, the isoprene dimers produced will be treated thermally at 135 °C under an inert N2 atmosphere to produce ID-2 in 92% yield. The residual isoprene produced during reaction should be distilled off and used again in the photoreaction cycle. Finally, we assume that heattreated isoprene dimers will be hydrogenated by using 10 wt% Pd/C (0.5 mol%) and H2 to obtain HID-2 in near quantitative yield, utilised as drop-in jet fuel and storable at the production site. The product could be separated by filtration of Pd/C to obtained jet fuel. The excess hydrogen used in this process would be recycled and used in the next hydrogenation cycle. The Pd/C (10 mol%) was not included in the LCA model due to low amount of loading (0.5 mol%) and reusability of the catalyst. The process and system boundaries modelled in the LCA are shown in Figure 10, and all inventory data summarized in Table S8, †ESI. Results from the LCA are presented in Table 3. The climate impact was approximately 0.6 tonne CO2 eq./tonne biofuel (Table S9, †ESI), mainly attributed to emissions from the production of sodium nitrate used in the photobiological processes (Table 3). The climate impact is about 20% of that of fossil jet fuel (approx. 3.8 CO2 eq./tonne for conventional Jet A), 85 and is at the lower end of the range from 0.6 -2.7 tonne CO2 eq./tonne biofuel found in the study by Nilsson et al., 38 which investigated the environmental impacts of cyanobacteria-produced n-butanol using three different reactors. In other studies, some investigated bio-jet fuels had the best result at 0.8 tonne CO2 eq./tonne. 28 Fig. 10. Flow chart for the process and system boundaries for the LCA. From the assessment of the overall environmental impact we see that under the assumptions made, the production of sodium nitrate completely dominates the impacts in all environmental categories (Table 3). This nitrate is used as a nutrient for cultivation of the cyanobacteria. The source of sodium nitrate in our model is the global market, and it is produced using fossil fuels. Use of alternative raw materials from waste streams, such as municipal waste water, as a source for nitrogen instead of sodium nitrate can be a potential solution for reducing the environmental impacts. 86 Increasing the photosynthetic efficiency of the cyanobacteria would also reduce the overall environmental impacts. ## Conclusions and Outlook In this study, we demonstrated that it is possible to generate photosynthetically derived isoprene from CO2 using engineered cyanobacteria, capture the isoprene, and use it for subsequent biofuel generation via a novel photochemical process driven by sunlight. While further optimization of the engineered microorganisms is required for industrial applications, we were able to trap isoprene with high efficiencies relying on a simple, scalable capturing method. We could also show that repeatedly removing the product enhanced productivity from isoprene producing cyanobacterial cultures In a subsequent photochemical step, the isoprene was dimerized into cyclic C10H20 isomers in nearly quantitative yields by usage of dinapthylmethanones as photosensitizers. The photoreaction could be run under ambient conditions, facilitating a fully renewable fuel production. Our current studies reveal that rather simple modifications of the reaction setup can greatly improve the yield of the photoreaction. Combined with a careful choice of photosensitizer this enables photodimerization of isoprene by use of solar light. The isoprene dimer mixture can be further modified by heating at moderately elevated temperatures (130 -160 °C), resulting in C10 hydrocarbon mixtures which after hydrogenation fulfil all criteria to function as drop-ins for conventional jet fuels. Indeed, the modified and hydrogenated isoprene dimers have better fuel properties than the commercially available Jet-A. The very low freezing points and low viscosity should make these fuels ideal for high-altitude flights. It is usually a challenge to compare the results and environmental impacts of an emerging technology with a mature technology due to several uncertainties such as missing data, upscaling assumptions and modelling issues. 87 In case of production of photosynthetic biofuels using microalgae and cyanobacteria, the process is still in its early stages and significant productivity improvements can be expected. The results of the current LCA study will assist in further improving our novel two-step technology for bio-jet fuel production from cyanobacteria. Our LCA showed an overall positive result on the environmental sustainability of our system. It was noted that the production of nutrients, in particular nitrate, dominates the environmental impact categories. Cyanobacteria can also conceivably grow well on municipal wastewater as a source of nutrients, including nitrogen, 86 something we have not included in the above model and which would likely increase sustainability. Hence, our results described are the very first steps toward a completely renewable jet fuel generated from CO2, water and solar light, provided that cultivation is carried out outdoors and that the hydrogenation and thermal rearrangement steps also utilize renewable energy. We report on the first proof-of-principle study of a combined photobiologicalphotochemical approach for jet fuel production. Extensive future research and development along various lines are needed, and several different short alkenes and dienes could be useful for similar processes. In the photochemical dimerization of isoprene presented in this study, we have produced in total ~3L of isoprene dimers. In an estimation, this amount would allow an Airbus A380 to fly (at least) ~174 m based on the fact that it is estimated to burn 13.78 kg/km and that the densities of our fuels are ~0.8 g/mL. 88 Considering this, there is a very long way to go before we have reached fully sustainable jet fuels produced by a combined photobiologicalphotochemical approach. Yet, every journey begins with a single step. ## Conflicts of interest There are no conflicts of interests.

chemsum

{"title": "A combined photobiological-photochemical route to C10 cycloalkane jet fuels from carbon dioxide via isoprene \u2020", "journal": "ChemRxiv"}

rapid_analysis_and_authentication_of_meat_products_using_the_masspec_pen_technology

## Abstract: Food authenticity and safety are major public concerns due to the increasing number of food fraud cases. Meat fraud is an economically motivated practice of covertly replacing one type of meat with a cheaper alternative, raising health, safety, and ethical concerns for consumers.In this study, we implement the MasSpec Pen technology for rapid and direct meat analysis and authentication. The MasSpec Pen is an easyto-use handheld device connected to a mass spectrometer that employs a solvent droplet for gentle chemical analysis of samples. Here, MasSpec Pen analysis was performed directly on several meat types including grain-fed beef, grass-fed beef, venison, cod, halibut, Atlantic salmon, sockeye salmon, and steelhead trout, with a total analysis time of 15 seconds per sample. Statistical models developed with the Lasso method using a training set of samples yielded per-sample accuracies of 95% for the beef model, 100% for the beef versus venison model, and 84% for the multiclass fish model. Metabolic predictors of meat type selected included several metabolites previously described reported in the skeletal muscles of animals, including carnosine, anserine, succinic acid, xanthine and taurine. When testing the models on independent test sets of samples, per-sample accuracies of 100% were achieved for all models, demonstrating the robustness of our method for unadulterated meat authentication. MasSpec Pen feasibility testing for classifying venison and grass-fed beef samples adulterated with grain-fed beef achieved persample prediction accuracies of 100% for both classifiers using test sets of samples. Altogether, the results obtained in this study provide compelling evidence that the MasSpec Pen technology is as a promising alternative analytical method for the investigation of meat fraud. ## Introduction Food fraud is an increasing public health and commercial concern for consumers and vendors. Currently, food fraud costs the global food industry between an estimated $10 billion and $15 billion per year. The 2013 European horse meat scandal, for example, caught international attention of consumers by revealing that meat products labeled as beef contained as high as 80-100% horse meat. Thus occurrence led many countries to examine their own meat and food fraud policies to advance prevention of this crime. 5 The most common method for meat fraud is replacement, which involves the complete or partial substitution of a meat product with a less expensive adulterant, such as replacing beef for horse meat. 1-2, 4, 6-7 The substitution can be made by replacing the whole meat product, such as a steak or fish fillet, or by incorporating the adulterant into a ground meat product at a certain percentage. In addition to the economic and criminal effects of this practice, meat fraud also affects consumers with meat allergies or other dietary, religious and cultural restrictions, while also representing an ethical violation of the trust of the costumers. Physical and molecular evaluation of meat products is routinely performed to verify its authenticity and assure product quality before reaching the costumer. For example, in the United States, meat products from farms are continuously inspected by the United States Department of Agriculture (USDA) Food Safety and Inspection Service (FSIS) for authenticity and quality both before (ante mortem) and after (post-mortem) slaughter. For instance, meat will be inspected ante mortem for disease, illness, injury, identification, etc., and inspected postmortem for incidental or purposeful cross contamination, identification, bacterial growth, etc. Following visual inspection, randomly chosen meat products undergo speciation and hormone tests before reaching the consumer in a FSIS facility to detect the type(s) of meat and hormones present in a product. 1 Polymerase chain reaction (PCR) and liquid chromatography mass spectrometry (LC-MS) are the most commonly used techniques for meat authentication. In PCR assays, accuracies from 96% to 100% are reported for the identification of a variety of meat types. PCR assays require targeted molecular probes for testing of the DNA material extracted from the meat product, and can often take four hours to several days to yield a result. 8,10 Mass spectrometry methods have been increasingly employed as a faster alternative test for meat speciation. 2,8 LC-MS techniques are commonly used to identify meat type based on quantitative analysis of lipids and proteins that are characteristic of each meat type. Detection limits below 3% in terms of partial substitution limits have been reported using LC-MS, in a time frame normally ranging from three hours to two days for completion, depending on the sample preparation methods used. LC-MS is also frequently used to identify the nature of the contaminants present in the meat, or food products. 2,6,15 For example, LC-MS has been used to detect adulterants that are added to a meat product for financial gain, such as soybean protein added to beef or chicken in pork meat. 2,6,14,16 Although LC-MS assays allow faster analysis when compared to PCR, the time required for sample preparation steps and chromatographic separation hinders its use for direct, real-time analysis of food products. 8,17 Moreover, PCR and LC-MS instrumentation is complex and often done off-site at a FSIS facility, thus requiring time for sampling, transportation to the analysis site, and storage. Several ambient ionization MS techniques have been explored as methods for direct and on-site food analysis, including analysis of meat products in an effort to expedite and improve accessibility to fraud testing. Liquid extraction surface analysis MS (LESA-MS), for example, has been explored for meat authentication based on proteomic analysis a sample digest from various raw, cooked, and processed meat products, followed by analysis using nanoESI. Using this approach, M. Montowska and coworkers reported a predictive variation of 94.9% using orthogonal partial least squares discriminant analysis in the identification of cooked horse, pork, turkey, chicken, and beef (n=50) using peptide biomarkers. This study also reported partial substitution limits of detection of 10% for pork, horse, and turkey and 5% for chicken in beef matrices (n=50). 26 Rapid evaporative ionization MS (REIMS) has also been explored for meat authentication. In REMS, a handheld electrocautery device is directly used to thermally ablate the meat sample, leading to the formation of an aerosol that is transported to a mass spectrometer for lipidomic analysis. REIMS was used to analyze a wide range of meat samples, including beef, horse, venison, and various fish species in multiple studies. One of these studies by J. Balog, et al., yielded high prediction accuracies of 100% for horse and beef and 97% prediction accuracy at the breed level (n=20) using leaveone-animal-out-cross validation, as well as 5% detection limit for mixed samples of Wagyu beef and horse meat mixed into venison and grain-fed beef meat, of beef, venison, and horse meat, and of all four meats. 34 Here, we explore the use of the MasSpec Pen technology as an alternative approach for rapid and direct meat authentication of unadulterated and adulterated meat samples. The MasSpec Pen is a handheld device coupled to a mass spectrometer that enables direct sample analysis based on a gentle liquid-extraction process. Upon contacting the device onto a sample surface and pressing a foot pedal, a syringe pump delivers a discrete solvent droplet to a reservoir at the pen tip, where it is held in contact with the sample for 3 seconds. The solvent gently extracts molecules such as small metabolites and lipids from the sample into the droplet, which is then transported through polytetrafluorethylene (PTFE) tubing to the mass spectrometer for analysis. The entire process from contact to analysis is completed in seconds. 35 Notably, the gentle nature of the liquid-extraction process allows molecular analysis without apparent damage to the sample surface. We have previously shown that the molecular data acquired using the MasSpec Pen in conjunction with statistical modeling via the lasso method allows discrimination of normal and cancerous human tissues with overall accuracies over 96%. Here, we show that the MasSpec Pen allows effective molecular analysis of various meat samples, enabling meat authentication in seconds. ## Materials and Method Samples. Meat samples of grain-fed beef (n=13), grass-fed beef (n=13), pork (n=5), chicken (n=5), lamb (n=5), venison (n=13), cod (n=13), halibut (n=13), Atlantic salmon (n=14), sockeye salmon (n=13), and steelhead trout (n=13) were obtained from local grocery stores (Austin, TX) and stored at 4 o C until analysis. Each meat sample (or biological replicate) was from a different animal. Only samples that were verified as being regulated by the USDA were purchased. Mixed meat samples containing either grain-fed beef and venison (n=31) or grain-fed and grassfed beef (n=31) were made in the laboratory using pure meat products that were ground together at different ratios (0, 25, 50, 75, and 100%) using a meat grinder. Samples were processed through the grinder at least three times to yield a uniform distribution of each product and were stored at 4 o C until analysis. Prior to MasSpec Pen analysis, all samples were brought to room temperature and any excess moisture on meat samples was removed with a Kimwipe. MasSpec Pen Analysis. The MasSpec Pen design and experimental setup have been previously described in detail. 35 Polydimethylsiloxane pen tips with a 4 mm reservoir diameter and a solvent droplet volume of 20 µL were used for all experiments. Various solvent systems (water, methanol:water blends, and acetonitrile:dimethylformamide (ACN:DMF) blends), extraction times (3, 5 and 10 seconds), and PTFE tubing lengths (0.5, 0.75, 1.0, and 1.5 m) were tested. Between analyses, a wash step was performed where the pen system was flushed with the solvent system. Tubing and pen tips were changed between meat types. For quality control, a background analysis was performed at the beginning and between analyses to monitor and minimize carryover. The MasSpec Pen was coupled to a Q Exactive Hybrid Quadrupole-Orbitrap and to a Q Exactive HF Hybrid Quadrupole-Orbitrap mass spectrometer (Thermo Scientific, San Jose, CA) for analysis of the meat samples at ambient conditions. All analyses were conducted in the negative ion mode with a mass range of m/z 50-600 and 140,000 resolving power (at m/z 200). Eight or nine analytical replicates were acquired for each biological replicates of each meat type, by performing MasSpec Pen analysis of different regions of the meat piece. For each mixed meat sample, three to five analytical replicates were analyzed depending on the size of the sample. Test set samples were analyzed separately two months apart from the training set of samples and treated as an independent sample set. Ions were tentatively identified based on high mass accuracy measurements (mass error ≤ 5 ppm) and higher energy collision induced dissociation (HCD) tandem MS analyses (Supplemental Figure 3). Statistical Analysis. Following analysis with the MasSpec Pen, three scans were averaged for each analytical replicate. The resulting mass spectrum was extracted from the XCalibur raw data. The relative standard deviation (RSD) of the method was calculated using the average and standard deviation of ratios of biological ions detected in the mass spectra for each meat type in the optimization study. For further statistical analysis, data were imported into the R programming language. All mass spectra were normalized according to total ion count. Following normalization, all background peaks and peaks not present in at least 10% of samples were excluded. Cosine statistical analysis was performed using the las package in R. Principal component analysis (PCA) was performed by centering the preprocessed data to mean zero and computing principal components in R. Lasso was applied to developed classification models using the glmnet package in the R language. Lasso is a logistic regression technique that selects a sparse set of features, specifically m/z values, to create a predictive model, or classifier, capable of discriminating between two or more classes, or in this case different meat types. 37 A training set of MasSpec Pen data was used to build each classifier using leave-one-out cross validation (Supplemental Figure 2). The test set of independent samples was used to evaluate the performance of the classification models. Performance of the classification models was measured by recall for each meat type and by an overall accuracy for each classifier. ## Optimization of the MasSpec Pen for meat analysis We optimized the MasSpec Pen solvent system, extraction time, and tubing length to enable molecular analysis of meat samples. Solvent systems of water, methanol:water blends, and acetonitrile:dimethylformamide (ACN:DMF) were tested by analyzing grain-fed beef, pork, chicken, and lamb and evaluating the diversity of molecular species detected as well as the reproducibility of the mass spectra. Six biological replicates for each meat type and nine analytical replicates per sample were performed. Note that all the solvent systems tested were nondestructive to the meat tissue, thus allowing repeated analyses of the same sampling spot, if desired. Within the solvents tested, pure were observed in the mass spectra obtained with both solvents at high relative abundance. Although complex lipids are commonly observed with the MasSpec Pen in tissue analysis, the m/z range for analysis of meat products was restricted to m/z 600 to focus on the analysis of the detection of small metabolite species. Generally, we observed that ACN:DMF (1:1) allowed detection of a richer diversity of molecular species, with species such as deprotonated galactitol (m/z 181.070), deprotonated taurine (m/z 124.006), and deprotonated inosine (m/z 267.073) detected using ACN:DMF, but undetected with water. Additionally, ACN:DMF yielded higher reproducibility (RSD of 15% ± 4%, n=6 for each meat type) when compared to what achieved with water (RSD of 34% ± 7% for mass spectra of all meat types, n=6 for each meat type) (Supplemental Figure 1). Thus, ACN:DMF (1:1) was used as the solvent system for the remaining experiments performed. Next, we optimized the MasSpec Pen extraction time by varying the time (3, 5 and 10 seconds) that the pen and the solvent droplet was in contact with grain-fed beef samples. To compare the mass spectra between varying extraction times, a cosine similarity analysis was performed. As previously reported in experiments with human tissues, 35 reproducible mass spectral profiles were obtained for the different extraction times explored, yielding an average cosine value of 0.985 ± 0.0.009. An extraction time of 3 seconds was selected for the analyses to expedite total analysis time per sample. Lastly, different PFTE tubing lengths (0.5, 0.75, 1.0. and 1.5 m) were also tested. The mass spectra obtained of grain-fed beef samples using 0.5, 0.75, 1.0. and 1.5 m tubing lengths, resulted in an average cosine value of 0.97 ± 0.02, and shows that the mass spectra profiles are nearly independent of tubing length, which also corroborates previous findings 35 . Thus, the shortest tubing length of 0.5 m was used for all following analyses to conserve PTFE tubing material and expedite analysis time per sample. Using 0.5 m PFTE tubing length, droplet transport from the pen tip to the mass spectrometer was completed in 2 seconds. At the optimized conditions, each analysis was performed with ACN:DMF (1:1) solvent, 3 seconds of extraction time, 0.5 m PFTE tubing, for a total analysis time of ~15 seconds per sample, which is faster than PCR and LC-MS, the current gold standard methods. 2,6,8 ## Pilot study of MasSpec Pen meat speciation We then performed a preliminary study to determine if the molecular profiles obtained with the MasSpec Pen were capable of distinguishing visually distinctive meat types, including grain-fed beef, pork, chicken, and lamb (five biological replicates per meat type, four analytical replicates per sample). As expected, different mass spectral profiles were obtained from the MasSpec Pen analysis of each meat type (Figure 1a). Deprotonated carnosine (m/z 225.099), for example, was detected in all meat types with high abundances in beef and pork, which was previously detected in various meat types using LC-MS. Deprotonated anserine (m/z 239.115) and deprotonated succinic acid (m/z 117.018) were detected in beef, lamb, and chicken. Furthermore, deprotonated xanthine (m/z 151.025) was detected in chicken and beef at a higher abundance when compared to lamb and pork, while deprotonated taurine (m/z 124.006) was detected in relatively high abundance in all meat types. These molecules were previously detected using LC-MS in various meat types. Using PCA, separation between the data obtained from the four meat types evaluated was achieved (Figure 1b), thus confirming that the molecular information obtained using from MasSpec Pen analysis is distinct between meat types. ## Developing beef and venison authentication models using the MasSpec Pen. We next used the MasSpec Pen to analyze commonly substituted meat products. One set of commonly substituted meat products are grain-fed beef and grass-fed beef, which are the same meat species with different feeding habits. Typically, grass-fed beef is the more expensive of the two, thus grain-fed beef can be mislabeled as grass-fed for a larger profit. 42 Another set of commonly substituted meat products are grain-fed beef and venison, which is also performed for financial gain. For each meat type, ten raw samples were analyzed with at least eight analytical replicates per sample, yielding 254 spectra from the 30 meat samples. Representative mass spectra are shown in Figure 2a for a sample of grain-fed beef, grass-fed beef, and venison. Qualitative differences in the relative abundances of several metabolic species previously reported in the skeletal muscles of animals can be seen when comparing the mass spectra obtained. For example, deprotonated carnosine (m/z 225.099), previously reported in skeletal muscles of beef and chicken using LC-MS, was detected at a high relative abundance in both beef profiles when compared to venison. Additionally, deprotonated malic acid (m/z 133.013) and deprotonated anserine (m/z 239.115) are at an increased relative abundance in grain-fed beef when compared to venison and grass-fed beef. Moreover, a higher relative abundances of deprotonated xanthine (m/z 151.025) was detected in venison and grain-fed beef, while a lower relative abundances of deprotonated taurine (m/z 124.006) in grass-fed beef and venison, both of which have been previously observed in beef and fish meat using LC-MS. Interestingly, a qualitative higher relative abundance of the chlorine adduct of hexose (m/z 215.032, [M+Cl] -) was detected in grain-fed beef, which could be associated with the diet of grain-fed cattle containing more sugars than the diet of grass-fed cattle. 43 Using the data collected, we created two two-class classification models, one for grain-fed and grass-fed beef, and one for grain-fed beef and venison using the Lasso method. 35, For the beef two-class model, 176 spectra acquired from 20 samples were used as the training set of data using leave-one-out cross validation (LOOCV). A per-sample accuracy of 95% was achieved (Figure 2b), which is particularly encouraging as the samples are from identical species with the main difference being feeding habits. Notably, within the predictive m/z, deprotonated malic acid (m/z 133.013), deprotonated carnosine (m/z 225.099), and the chlorine adduct of hexose (m/z 215.032) were selected, which corroborate with the trends in ion relative abundances observed in the mass spectra. For the beef versus venison model, 165 spectra acquired from 20 samples were used as the training set, yield a per-sample accuracy of 100% (Figure 2c). The predictive m/z selected included ions deprotonated xanthine (m/z 151.025), deprotonated carnosine (m/z 225.099), and deprotonated anserine (m/z 239.115). To evaluate the performance of the model for meat authentication setting, a test set of samples were used, as discussed later in the manuscript. Developing a fish authentication multiclass model using the MasSpec Pen. We then examined the MasSpec Pen's performance for fish identification for a multiclass fish model. Here, we analyzed five common fish products, steelhead trout, sockeye salmon, Atlantic salmon, cod, and halibut. For each meat type, nine samples were analyzed as a training set of samples, with at least eight analytical replicate analyses performed for each sample, yielding 395 spectra from the 45 samples analyzed. When evaluating the molecular profiles of the raw samples analyzed with the MasSpec Pen, qualitatively district mass spectra profiles were detected for each fish type, as shown in Figure 3a. For example, deprotonated taurine (m/z 124.006), was detected at varying relative abundances in each fish tested. 40 Additionally, m/z 267.074 (unidentified) is seen in a higher relative abundance in cod when compared to the other products. Deprotonated anserine (m/z 239.115) was qualitatively observed in a higher relative abundance for sockeye salmon, 39 while a higher relative abundance of deprotonated xanthine (m/z 151.025) was observed in halibut. Moreover, a qualitatively higher relative abundance of the chlorine adduct of hexose (m/z 215.032) was detected in Atlantic salmon, a farm-raised fish, when compared to wild-caught sockeye salmon, which could be due to their habitats and eating habits. We then used Lasso to create a five-class model for the fish types analyzed (395 mass spectra obtained for the 45 samples) using LOOCV, yielding 84% overall accuracy per sample (Figure 3b). The confusion matrix with recall values for each fish type is provided in Supplemental Table 4. Notably, the highest recall was achieved for cod and halibut, which are the most commonly substituted fish meat products as cod loin is often substituted for halibut. 7,31 Larger confusion was seen when classifying the farm-raised Atlantic salmon and wild-caught sockeye salmon, which is expected as they are the same fish species and commonly substituted with each other and trout. 7,31 The predictive features selected included deprotonated taurine (m/z 124.006), the chlorine adduct of methyluric acid (m/z 217.011), and deprotonated xanthine (m/z 151.025) and reflected the trends in relative ion abundances seen in the molecular profiles. While a five-class classifier was built to include the five fish types analyzed, a twoclass and three-class classifier built using LOOCV were also explored for fish that are more commonly substituted in fraudulent activities. For example, a two-class classifier was built to distinguish halibut and cod, which are visually very similar and often substituted in meat fraud crimes, yielding 100% accuracy with 229 mass spectra acquired from 26 samples (Supplemental Table 5). A three-class classifier was also built to distinguish steelhead trout, sockeye salmon, Atlantic salmon, which are also visually similar and commonly substituted in meat fraud crimes. For this classifier, an overall accuracy of 90% was achieved (352 mass spectra and 40 samples) (Supplemental Table 5). Overall, the classifiers developed here for fish identification demonstrate the potential of the MasSpec Pen for identification of fish type and the investigation of fish fraud. ## Authentication of meat products using the MasSpec Pen. To evaluate the predictive performance of the classifiers built, we analyzed an independent test set of meat and fish samples using the MasSpec Pen following the same experimental approach. For the grain-fed beef, grass-fed beef, and venison, three biological replicates were analyzed with at least eight analytical replicates for each sample, yielding 71 mass spectra for 9 samples. For the five fish types, four samples were analyzed for each meat type with at least eight analytical replicates per sample, yielding 187 spectra for 21 samples. Using the previous statistical models to predict the meat type for each sample, test set accuracies of 100% were achieved for the beef model and the beef versus venison model (Figure 2b and 2c). Similarly, an accuracy of 100% was achieved for the fish multiclass model (Figure 3b). The accuracies are comparable with what previously reported for LESA-MS, 94.9% (n=50) for a training set, and REIMS, 100% (n=20) species identification for a training set. 26,30 Further, the accuracies achieved are comparable with current testing metrics using PCR (96%-100%). The performance achieved with the MasSpec Pen in the test set of samples provides evidence the MasSpec Pen in conjunction with lasso is a robust method for analysis of meat products. ## Extending the MasSpec Pen use to the analysis of mixed meat samples. We next evaluated if the MasSpec Pen could be used to identify adulterated mixed meat samples. To this end, we mixed varying percentages of ground grain-fed beef (0%, 25%, 50%, 75%, and 100%, in weight), as an adulterant into samples of ground grass-fed beef or ground venison, and analyzed the ground samples using the MasSpec Pen. For the grain-fed beef mixed into grass-fed beef, we analyzed nine 0% grain-fed beef samples, four 100% grain-fed beef samples, and six 25%, 50%, and 75% samples. For the grain-fed beef mixed into venison, we analyzed eight 0% grain-fed beef samples, six 100% and 25% grain-fed beef samples, and five 50% and 75% samples. When analyzed, each of the ground meat samples had five analytical replicates for each sample. The mass spectra obtained from the analysis of the venison samples adulterated with grain-fed beef are shown in Figure 4a. Notably, relative abundances of deprotonated taurine (m/z 124.006), deprotonated malic acid (m/z 133.013), and deprotonated carnosine (m/z 225.099) were increasingly higher with increasing percentage of adulterant grain-fed beef in the ground venison samples. On the other hand, a decrease in the relative abundances of deprotonated succinic acid (m/z 117.018) and deprotonated inosine (m/z 267.074) was observed as the percentage of adulterant grain-fed beef increases in the ground venison samples. The changes in the relative abundance of the metabolites agreed with the trends observed in the raw meat samples. We then built statistical classifiers to test if classification of meat samples as adulterated was possible using the metabolic information obtained with the MasSpec Pen. To this end, we trained the Lasso method to identify adulterated and unadulterated meat by including mix samples (amount >25%) into the adulterated class and ground samples (0% mixed) in the unadulterated class. Using the training set data of 76 mass spectra and 20 samples, an overall 93% per spectra accuracy and a 90% per samples accuracy was achieved (Figure 4b). The features were selected for prediction of the two classes and reflected trends in molecular profiles seen in Figure 4a. For example, deprotonated carnosine, which has been previously found to be present in skeletal muscles of beef, was selected as a predictive feature weighted towards the adulterated with grain-fed beef classification, while deprotonated inosine (m/z 267.074) was selected as a predictive feature weighted towards the unadulterated classification. Within the training set, two unadulterated samples were misclassified as adulterated samples, while no adulterated samples were misclassified. We then tested the classifier on 43 mass spectra acquired from 10 samples, including 7 adulterated samples and 3 unadulterated samples. Overall accuracies of 98% per mass spectra and 100% per sample were achieved for the venison mixed samples test set (Figure 4b). The mass spectra obtained from the analysis of the grass-fed beef adulterated with grain-fed beef are shown in Figure 5a. As the amount of grain-fed beef adulterant increased in the samples, the relative abundances of ions such as deprotonated taurine (m/z 124.006), deprotonated malic acid (m/z 133.013), and deprotonated D-erythro-Lgalacto-nonulose (m/z 269.088) also increased, while a decrease in the relative abundances of deprotonated succinic acid (m/z 117.018) and deprotonated xanthine (m/z 151.025) was observed. Similar to the adulterated venison classifier described above, a binary classifier was built for the beef samples in which grain-fed beef was used as an adulterant to grain-fed beef meat. Using a training set of 88 mass spectra and 21 samples, an overall accuracy of 92% per mass spectra and 90% per sample was achieved (Figure 5b). Within the several features selected by the model, including deprotonated malic acid and deprotonated anserine weighted towards the adulterated classification and deprotonated xanthine weighted towards the unadulterated classification. One unadulterated sample was misclassified as adulterated in the training set. When tested on 45 mass spectra acquired from 10 samples in the validation set, 100% accuracy per mass spectra and per sample was achieved (Figure 5b). Multiple LC-MS methods have previously reported detection limits down to 3% for various mixed meat samples. Collectively, the results for both classifiers demonstrate feasibility in developing the MasSpec Pen and lasso as a robust method for the authentication of mixed meat samples, although further studies using lower percentages of substituted meat are needed to validate these findings. In conclusion, in this study we describe the optimization and application of the MasSpec Pen technology for analysis and classification of meat products. We showed that the gentle nature of the MasSpec Pen analysis allows detection of a range of metabolic species directly from fish, beef, and venison without the need of sample digestions or alteration. Several of the metabolic species detected, including deprotonated carnosine, deprotonated xanthine, deprotonated inosine, deprotonated anserine, the chlorine adduct of hexose, and deprotonated taurine have been previously described in meat samples and related to eating habits and other metabolic process in animal skeletal muscle tissues. 43 Using statistical classification with the Lasso method, we demonstrate the robustness and high accuracy of the MasSpec Pen in identifying meat types in training (93% accuracy, n=85) and test sets (100% accuracy, n=33). Further, we showed that common meat replacement could items could be identified using our approach, including discrimination of meats commonly used for replacement fraud such as grain-fed versus grass-fed beef (100% accuracy in test set, n=6), and different fish types (100% accuracy in test set, n=21), with similar performance to what reported with other ambient ionization MS methods and other routine meat testing techniques. 8-12, 26, 30 Lastly, we showed feasibility for identification of adulterated mixed meat samples. Although the MasSpec Pen is limited to qualitative molecular evaluation with substantially lower molecular coverage when compared to LC-MS, the MasSpec pen analysis is completed in less than 15 seconds and does not require any sample pre-processing, which is appealing for routine use in testing of meat products. At a minimum, the 15-second testing time per sample provided by the MasSpec Pen is ~240 times faster than LC-MS (considering a total analysis time of 1 hour/sample), and ~720 times faster than PCR (considering a total analysis time of 3 hours/sample). Further, the ease of use and maneuverability that a handheld device like the MasSpec Pen allows can facilitate implementation and use by users with various levels of expertise. Due to the gentle nature of the analysis, repetitive analysis of different products and regions within the sample of interest can be achieved. Future research would include expanding meat identification classifiers, improving the mixed meat results, and developing methods to quantify the amount of adulterant present in a sample. Furthermore, we will expand our method to other meat products, such as wild fish products and beef products from different countries. Lastly, while a high-performance Orbitrap mass spectrometer was used here for this exploratory study, we are currently exploring integration of the MasSpec Pen with a portable ion trap mass spectrometer for meat analysis to facilitate fieldable use outside of specialized laboratories. Collectively, our study shows compelling evidence that the MasSpec Pen provides a rapid and direct method for investigating meat fraud, allowing for accurate meat identification in less than fifteen seconds, thus providing a powerful alternative technique to traditional meat testing methods.

chemsum

{"title": "Rapid Analysis and Authentication of Meat Products using the MasSpec Pen Technology", "journal": "ChemRxiv"}

complex_coacervates_as_extraction_media

## Abstract: Various solvents such as ionic liquids, deep eutectic solvents, and aqueous two phase systems have been suggested as greener alternatives to existing extraction processes. We propose to add macroscopic complex coacervates to this list. Complex coacervates are liquid-like forms of polyion condensates and consist of a complex of oppositely charged polyions and water. Previous research focussing on the biological significance of these polyion-rich phases has shown that polyion condensates have the ability to extract certain solutes from water and back-extract them by changing parameters such as ionic strength and pH. In this study, we present the distribution coefficients of five commonly used industrial chemicals, namely lactic acid, butanol, and three types of lipase enzymes in poly(ethylenimine)/poly(acrylic acid) complex coacervates. It was found that the distribution coefficients can vary strongly upon variation of tunable parameters such as polyion ratio, ionic strength, polyion and compound concentrations, and temperature. Distribution coefficients ranged from approximately 2 to 50 depending on the tuning of the system parameters. It was also demonstrated that a temperature-swing extraction is possible, with backextraction of butanol from complex coacervates with a recovery of 21.1%, demonstrating their potential as extraction media. ## Introduction Solvent extractions are important processes in many industrial separation processes ranging from the chemical industry, the food industry, to the pharmaceutical industry. An application of liquid extraction that has been receiving increasing attention is in the field of bio-based chemical production. There are many different categories of bio-based chemicals and the feature that they often have in common is that typically large amounts of water are present. Removing water by evaporation is among the costliest operations in industry, and therefore when aqueous solutions are present, liquid-liquid extraction (LLX) may be applied. In LLX an additional liquid phase, typi-cally an organic solvent exhibiting preferential solubility for a specific solute, is used to selectively extract the solute from the initial liquid phase. Unfortunately, organic solvents that have been proven to be effective for extraction can be toxic for individual organisms and/or the environment. 1,2 There is great interest in the design of 'green solvents' that are more environmentally friendly in terms of production, usage, and disposal. For extraction from aqueous solutions, several alternatives to conventional organic solvents have been proposed in the past years such as ionic liquids (ILs), 3,4 deep eutectic solvents (DESs), 5 and aqueous two phase systems (ATPSs). ILs are essentially molten salts with a relatively low melting point ( per definition, ≤100 °C). 9 ILs have shown a broad range of applications in part due to the customization possible as a result of the large variety of composite components. 3,10 They are generally less volatile in nature compared to organic solvents and the negligible vapour pressure eliminates solvent losses through evaporation. 11 Unfortunately, many ILs are potentially toxic and not biodegradable. 12 DESs are mixtures of hydrogen bond donors and hydrogen bond acceptors that form liquids on mixing and exhibit eutectic behaviour by having melting points lower than that those of their constituent components. They have been proposed as new extraction solvents and share many advantageous characteristics with ILs. 5,13 The toxicity of DESs varies, and in some cases the DES is even more toxic than its constituent components, 14,15 which is a factor to be taken into consider-ation when formulating DESs for sustainable extraction. Additionally, due to the fact that DESs are composite solvents, the molar ratio between the hydrogen bond acceptor and donor may change during the extraction. 16 This can result in solidification of the DES components and affect the subsequent extraction steps. ATPSs function via segregative phase separation and consist of two ( partially) immiscible aqueous phases. The most common ATPSs are formed when two constituents (often polymer-polymer or polymer-salt (or even ILs 7,8 )) are mixed in an aqueous solution, resulting in two distinct segregated phases. Each of the segregated phases is rich in one of the two constituents. When used for the separation of molecules, one of these phases will be the preferred phase for the compound of interest, while the remaining impurities hopefully concentrate in the other phase. ATPSs are currently extensively used for the isolation and extraction of various biological compounds ranging from small molecules, hormones, up to the isolation of entire cells. Also, micellar systems have been proposed as the foundation for new greener extraction methods with extraction principles similar to those of ATPSs. 23 Similar to segregative phase separation, two phase systems can also be formed via associative phase separation such as complex coacervation (Fig. 1). This process occurs when oppositely charged polyions (a.k.a. polyelectrolytes) are mixed under conditions that allow them to associate. The formed complex coacervates (CCs) are macroscopic liquid-like aqueous polyion-rich condensates, which are in equilibrium with an aqueous polyion-poor phase, also called the supernatant. Depending on the chemistry of the polyions and the environmental conditions, solid-like condensates can also form, called polyelectrolyte complexes (PECs). In this study, we will make use of complex coacervates. In previous studies, CCs and PECs have been reported with the property of partitioning certain proteins into the complex phase over the supernatant phase. The ability to isolate proteins using single polyions is already well established, but a previous study has shown that in some cases the addition of a mixture of both polycations and polyanions can lead to better partitioning than the addition of only one species of polyions. For example, the addition of the polyanion poly (acrylic acid) (PAA) alone is not enough to extract the positively charged protein lysozyme from an aqueous solution, but with the addition of a polycation (and thus the formation of a PEC), the lysozyme could be extracted completely. 24 CCs therefore show emergent properties that their constituent components do not. A potential advantage of associative phase separation of CCs and PECs over segregative phase separation of ATPSs is that the distribution coefficients of CCs can be dependent on the composition of the CCs, resulting in different partitioning behaviours for the same constituent polyions present in different ratios. 24,25 There are a handful of studies that show that PECs have the ability to partition certain proteins as well as certain small molecule dyes 29,30 from an aqueous solution. In some cases, the distribution coefficients reported were in the order of 10 4 in favour of distribution in the PEC for a specific protein and polyion pair. 24 These studies hint at the potential of CCs and PECs as extraction media, though they are typically concerned with biomedical applications such as intracellular drug delivery. We have previously achieved success in using structurally simple polyions in order to selectively extract lysozyme from an aqueous solution in the presence of another protein. 24 Beyond varying the ratio of the polycation to the polyanion, there are other factors that influence the CC properties such as solution ionic strength, temperature, and varying concentrations of the system's constituents. There are no systematic studies that go into the details of the effect of such system parameters on the partitioning behaviour of the solutes. The inspiration for CCs as extraction media comes from the partitioning behaviour of solutes between cellular fluids and membraneless organelle (MLO) compartments within living cells. MLOs consist of both negatively and positively charged biomacromolecules such as negatively charged RNA and positively charged intrinsically disordered proteins. 31 The MLO phase behaviour strongly resembles the phase behaviour of CCs. Our cells use MLO droplets to perform very specific biological functions, including the partitioning and release of specific targeted compounds in response to changes in the stimuli in the cellular environment. While nature undoubtedly has a head start regarding the design of MLOs, their functionality in cells shows that there is currently untapped potential for CCs as media for extraction processes. Developing CCs with distribution coefficients that are strongly dependent on tunable stimuli and environmental parameters would be of great benefit to the development of extraction processes. In this study, we investigate the extraction of several compounds from aqueous solutions using complex coacervates formed by branched poly(ethylenimine) (PEI) and poly(acrylic acid) (PAA). PEI-based nanocrystals have been used as extraction media for rare earth element recovery and are increasingly used as vehicles for drug delivery. 36,37 Higher molecular weight PEI is typically considered cytotoxic, though this effect can be decreased by using the low molecular weight (1.8 kDa) variant that is used in this study. 38 PAA is commercially used as a thickening agent and water absorber in the hygiene, cosmetic, agricultural, and food industries. In these contexts, PAA is usually known as sodium polyacrylate or waterlock. We consider lactic acid (LA), butanol, and three varieties of industrial lipase enzymes as model compounds for the extraction from the aqueous supernatant into the PEI/PAA CC. These industrially relevant lipases are widely used in food, detergents, and pharmaceuticals 39,40 and represent up to 10% of the total global enzyme market. 41 Lactic acid extraction from an aqueous fermentation broth has received increased attention in the last few years amongst others due to the possibility of poly(lactic acid) being a sustainable alternative to many commonly used plastics. 42 The use of poly(lactic acid) as a competitor to modern plastics is currently restricted to application areas where the higher costs associated with purification and extraction from the fermentation broth can be tolerated. Several techniques have been in development for the recovery of LA from the fermentation broth aiming to reduce the production cost and decrease the impact of by-product formation during lactic acid production on the environment, and CCs may be a new technique to address the LLX of LA. 43,44 Butanol, being a popular solvent and a popular candidate for biofuels, can also be extracted from fermentation broths. 45 In this study, we create macroscopic CCs via associative phase separation of PEI and PAA. We investigate the effect of several parameters such as CC composition, reagent concentrations, and temperature on the partitioning of lipases, lactic acid, and butanol to demonstrate a proof of concept to draw attention to the use of CCs for extraction purposes. ## Materials Poly(acrylic acid) (PAA) sodium salt powder with a molecular weight of 6.0 kDa and branched poly(ethylenimine) (PEI) with a molecular weight of 1.8 kDa were purchased from Polysciences, Inc. Sodium chloride (NaCl, >99%), sodium hydroxide (NaOH, >98%), fuming hydrogen chloride (HCl, 37 ± 1 wt%), n-butanol (>99%), and lipase from porcine pancreas (PPL) were purchased from Sigma-Aldrich/Merck. NovoCor AD L lipase (CALA) and Novozyme CALB lipase (CALB) were donated by Novozymes A/S. Crystalline L-lactic acid was donated by Corbion N.V. Unless otherwise specified, water used for the solutions and dilutions was ultrapure Milli-Q water dispensed from a PURELAB flex system at a resistivity of 18.2 MΩ. ## Experimental methods Complex coacervates were prepared by mixing prepared aqueous polyion solutions (PAA and PEI) for a total polyion concentration of up to 20 g L −1 in the presence of up to 400 mM NaCl. All solutions are set to pH 7 before mixing. In the case of lipases, they are added to the solution with the polyions at a lipase concentration of 67 µM, consistent with earlier studies. 25,26,33 Unless otherwise specified, butanol was added at 400 mM and lactic acid at 100 mM. In the case of butanol and lactic acid, the mixed polyion solution is first left to equilibrate overnight into a CC. Then it is centrifuged at 12 500 g for 30 minutes using the Centrifuge 5425 (Eppendorf ) to expedite the separation of polyion-rich complex coacervates from polyion-poor aqueous supernatant phases. The supernatant is then replaced with a new solution containing either lactic acid or butanol in an aqueous sodium chloride solution with the same NaCl concentration as during the preparation of the CC. Total volumes for each experiment were fixed at 500 µl unless otherwise specified. The composition of the CC is defined via F − ; where n − and n + are the concentrations of PAA and PEI monomers, respectively, which are mixed in solution. For example, at F − = 0.50, there is an equal molar amount of PEI and PAA monomers present, and at F − = 0.75, there are 3 PAA monomers for every 1 PEI monomer. The assumption being that at pH = 7 both polyions are fully charged due to the interaction between the two polyions. 24,25,46,47 Under this assumption, PAA has a mass of 76.7 g mol −1 of negative charge and PEI has a mass of 43.0 g mol −1 of positive charge. ## Analytical methods The total mass of the complex coacervates was determined by comparing the mass of the sample tubes when emptied to that of those containing only the complex coacervates. The volume was determined under the assumption that the density of the complex coacervates is approximately equivalent to that of water. 24 This assumption is based on the densities of PEI (1.03 g ml −1 ) and 50% PAA solution (1.15 g ml −1 ) reported by the manufacturer. Considering that the majority of the CC consists of water, total CC density is within a few percent of water, in the calculated range of 1.02-1.04 g ml −1 . The water content of the PEI/PAA complex coacervates was determined via thermogravimetric analysis (TGA) using a STA 449 F3 Jupiter (Netzsch) thermal analyzer on CCs formed at 10 g L −1 total polyion concentration. The temperature was increased from 30 to 120 °C at a rate of 5 °C min −1 and then kept constant at 120°for 40 min to evaporate the water present in the complex coacervates. The mass of the samples is recorded to obtain the mass loss corresponding with the evaporated water. Prior to the determination of the concentration of the solute present in the experiment, the systems were centrifuged for 30 minutes at 12 500g in an Eppendorf Centrifuge 5425. Enzyme concentration from the supernatant was determined by evaluating the absorbance at 280 nm using a Shimadzu UV-2401PC spectrophotometer. Extinction coefficients for PPL and CALA were calculated to be 68 kM cm −1 and 54 kM cm −1 based on the peptide sequence. The extinction coefficient for CALB has been reported in literature as 41 kM cm −1 . 48 Butanol concentration was determined using a Thermo Scientific Trace 1300 gas chromatograph with two parallel ovens, an auto sampler TriPlus 100 Liquid Samples and an Agilent DB-1MS column (60 m × 0.25 mm × 0.25 μm) with an injection volume of 1 μL diluted in analytical acetone. A ramped temperature profile was used, in which the initial temperature was 30 °C, followed by a ramp of 10 °C min −1 to 140 °C. The second ramp of 50 °C min −1 to 340 °C finished the program, which lasted for 15 min. The flame ionization detector temperature was 440 °C. A column flow of 2 mL min −1 with a split ratio of 25, an airflow of 350 mL min −1 , a helium make-up flow of 40 mL min −1 and a hydrogen flow of 50 mL min −1 was used. Lactic acid concentration was determined using a Grom Resin H + IEX column on a Metrohm 850 Professional ion chromatograph. The mobile phase was 1 mM H 2 SO 4 solution with a flow rate of 0.6 mL min −1 . The column temperature was 45 °C. As the total amount of the added compound is known and the concentration of the compound in the supernatant is measured, the compound concentration in the complex coacervate can be calculated. The distribution coefficient is then determined via where [X] CC and [X] SN are the concentrations of the compound in the complex coacervate and supernatant, respectively. The distribution coefficient changes depending on the varied parameter, resulting in a distribution profile. ## Butanol extraction and back-extraction PEI/PAA CC systems were prepared with a total polyion concentration of 50 g L −1 in 1 ml with a composition of F − = 0.26. The increased polyion concentration was chosen to produce more CC as a simulation of upscaling compared to the previous experiments. This mixture was centrifuged for 30 minutes at 1000 g. The aqueous supernatant was then replaced with 650 µl of 5.7% butanol and 10 mM NaCl solution. The samples were collected to determine the butanol concentration after 24 h of incubation at room temperature (RT), and again after 24 h of incubation at 70 °C. The supernatant was then decanted, and any excess supernatant drops were removed using pressured nitrogen gas. 600 µl of fresh 10 mM NaCl solution was added to the CC as a back-extraction phase, and the samples were collected from the back-extract after 24 h of incubation at 60 °C. Then, the samples were collected after another 24 h of equilibration at 40 °C, and once more after another 24 h at RT. The butanol concentration of all the samples was determined as described previously and the amount of butanol present in the CC was calculated taking into account the varying volumes of the supernatant due to sample extraction. ## PEI/PAA complex coacervate formation and water content Complex coacervates are formed due to the interactions between the oppositely charged polymer chains, with the driving force being both entropy gain due to the release of counterions and electrostatic interaction. The fraction of the negative and positive charges is important for the total extent of CC formation. To narrow down the region of interest for evaluating the partitioning, we first evaluated the total CC formed as a function of the composition F − and looked at the water content for two CC compositions of interest. In Fig. 2A, it is shown that the largest amount of CC was formed around F − = 0.25 to 0.50, with the highest values found at 0.26 and 0.36 with 23.1 ± 3.3 mg and 22.4 ± 3.3 mg, respectively. Fig. 2B shows the photographs of the relative quantities of CC as a function of F − . As shown in Fig. 2C, we evaluated the water content of the two F − values with the highest CC formation as seen from Fig. 2A and found that for PEI/PAA CCs the water content varies drastically based on CC composition, with the water content for F − = 0.36 being 73.5%, and for F − = 0.26 being 51.9%. Comparing the remaining mass of the polyions in the CC to the total polyions added, it appears that for F − = 0.26 all the polyions form the CC mass, while for F − = 0.36 only approximately 60% of the polyions form the CC, with the rest presumably remaining in solution. Intuitively, it might be expected that the largest volume of CC formation occurs at the composition F − = 0.50, where an equal amount of positive and negative monomers is present. However, this is not necessarily the case as demonstrated by the PEI/PAA CC system. One explanation for this discrepancy is that the interactions between polyions, water, and salts can affect the degree of ionization of the monomers. Water content of CCs and PECs is typically reported to be between 60 and 80%. We found using TGA that for PEI/ PAA CCs at a composition of F − = 0.36 the CCs fall within the reported range, though the water content at F − = 0.26 is approximately 10% lower than expected. The water content of CCs can impact the partitioning behaviour of solutes based on their preferential association with water. For example, lipases in general are known to prefer oil-water interfaces over fully aqueous environments. 52 Both PEI and PAA are not expected to decompose at the given conditions, temperature, and timescale. 53,54 ## Lipase enzyme distribution In this section, the partitioning of several types of lipases in the PEI/PAA complex coacervates is described. In Fig. 3, the distribution coefficients (K D ) of three commonly used lipases PPL, CALB, and CALA as a function of the CC composition, the NaCl concentration, and the total polyion concentration are shown. We found that the K D of all lipase types varies greatly as a result of the adjusted parameters. The charge ratio F − has the most significant consistent effect (Fig. 3A-C), showing distinct K D maxima at a composition of F − = 0.36 for CALB (K D maximum of 11.0 ± 0.9) and F − = 0.26 for CALA (K D maximum of 23.0 ± 0.5) and PPL (K D maximum of 19.2 ± 1.9). These maxima are partially consistent with the maximum values of CC formed (Fig. 2A); however, a small deviation in the composition results in a larger change in K D than that can be solely attributed to a difference in the CC quantity: for the region with the highest constant CC formation (F − ranging from 0.25 to 0.50), there are varieties in the K D for up to a factor 4 for CALA (Fig. 3A). The distribution profiles were found to be dependent on the specific protein investigated. The results for these lipases corroborate the earlier studies that report similar nearly symmetrical distribution profiles (though centred around different F − values) for three proteins with poly(lysine)/poly(glutamate) CCs. 26 Other proteins with different polyion pairs show completely different distribution profiles altogether that are not necessarily symmetrical. 24,25 For now, there are no reliable methods to predict the distribution profile in advance as a result of the parameters. Many studies that look into the partitioning of proteins assume F − = 0.50 is the optimal composition for both PEC formation and partitioning and do not investigate the other charge ratios. 27,28,30 Based on the results presented here, there might be opportunities for working at other compositions that result in more desired K D values. Both CALB and PPL show a similar distribution profile as a function of the NaCl concentration (Fig. 3E and F) with a slight K D increase initially, followed by a decrease. CALA however shows an immediate decrease, followed by a local maximum (Fig. 3D) at a comparatively high salt concentration. By varying the NaCl concentration, the K D varies between approximately 5 and 15-20 for the investigated enzymes. We hypothesize that for CALB and PPL a partial screening of the polyion charges by the salt ions results in the CC being less densely packed, essentially increasing the distance between polyion chains and allowing the proteins (or other solutes) to enter the CC more easily. Polyion condensation in the presence of other ions (such as salt ions) results in ion association with charged monomer subunits of the polyion. This effectively screens the electrostatic interaction between the oppositely charged monomers of each polyion. Indeed, if the ionic strength of the solution becomes too high, the polyion structures dissolve completely as the degree of screening prevents the complex formation between polyions. 46 Between complete complex dissolution and the absence of additional ions beyond the counterions brought in by the polyions, there is a concentration region where the salt ions prevent part of the oppositely charged polyions from associating. Subsequently, this can influence the behaviour of the condensates. All three enzymes showed a similar trend of K D decrease as the total polyion concentration increased. A possible explanation is that as the total mass of CC increases, this does not result in a proportional increase of the CC-water interface, limiting the penetration of the solutes into the CC. It is worth mentioning that there are other advantages of concentrating enzymes in CCs or PECs beyond extraction purposes. It has been reported that the activity of proteins may be enhanced in CCs compared to the same proteins in regular aqueous solutions. 27,47 In addition, the polyions may protect the proteins from degradation, increasing the shelf life of (extracted) proteins. 55 The mechanism for this is unknown, though the ability to both highly concentrate the enzymes and increase their activity is particularly interesting for industrial applications. ## Lactic acid distribution The partitioning of lactic acid into CCs was studied, as lactic acid is an industrially relevant small molecule. The effects of the CC composition, the NaCl concentration, the total polyion concentration, the initial LA concentration, and the temperature on the lactic acid K D were studied, and the results can be observed in Fig. 4. Unlike the distribution profiles for the lipase enzymes, we found only very little effect of the composition on the K D (Fig. 4A), which remained between 2 and 4. In contrast, the effect of NaCl on K D (Fig. 4B) of LA was more pronounced than those of CALB and PPL while following a similar distribution profile. Within our hypothesis of salt ions influencing the distance between polyion chains, the effect of the salt NaCl concentration may be more pronounced for LA, as it is substantially smaller than any of the lipases. By varying the NaCl concentration, we found the highest K D for LA at 7.4 ± 0.5 for 100 mM NaCl. Similar to the trend with lipases, increasing polyion concentration had an adverse effect on the partitioning (Fig. 4C). However, altering the initial concentration of lactic acid only slightly affects the partitioning in the evaluated range (Fig. 4D), suggesting that the saturation point for the CC has not yet been reached as this would result in an expected decrease in K D at higher LA concentrations. 24 Fig. 4E shows that an increase in temperature has a small but consistent positive effect on the K D in the investigated range. This suggests that the extraction process is endothermic and that the driving force behind the partitioning is an increase in total entropy, perhaps similar to how an increase in entropy is the primary driving force for polyion-polyion association in the first place. 23 The optimal K D for LA in our PEI/PAA CC system at approximately 100 mM NaCl is comparable to or greater than many other liquid-liquid extraction systems. 44, A disadvantage of some of these reported systems is their reliance on low pH 59 or the toxicity of the solvents. 58 While some established extraction methods, such as tri-n-octylamine in 1-octanol, 60 outperform CC systems with regard to LA for now, we show that the effects of system parameters for CC systems can substantially alter the K D . Interestingly, where the common method using tri-n-octylamine appears to decrease the distribution coefficient at higher temperatures, the opposite is true for PEI/PAA CCs. 56 There are many additional parameters that can be further fine-tuned, suggesting the ability to achieve much higher K D values. ## Butanol distribution, extraction, and back-extraction We investigated the K D of butanol as a function of the CC composition (Fig. 5A) as well as the temperature (Fig. 5B). As butanol partitioning showed a remarkable temperature sensitivity, we evaluated the possibility of extraction and backextraction of butanol using CCs by alternating between RT and 70 °C (Fig. 5D). Contrary to the lipases, we observe the highest K D as a function of CC composition quite distant from the optimal CC formation, resulting in the highest value of K D of 22.7 ± 0.7 at F − = 0.56. This K D is very similar to that of a reported task-specific IL and substantially higher than the standard of oleyl alcohol, which are K D = 21 and 3.4, respectively. 61,62 Whereas LA demonstrated only a minor temperature dependence of the K D (Fig. 4E), the butanol distribution shows a large difference between RT and 70 °C, roughly at a factor of 4-5. Out of the evaluated parameters, temperature is the most practical to change for the existing systems as it does not require the addition or removal of chemicals and is straightforward to implement. For this reason, we envisioned a PEI/ PAA CC system that was able to partition butanol within the CC to a greater degree at high temperatures and could then be coaxed to release butanol into a separate aqueous environment at lower temperatures such as RT. To evaluate such a system for extraction and back-extraction of butanol, we prepared PEI/ PAA CCs at higher concentrations of polyions (Fig. 5C). The resulting CCs had a mass of 62.2 ± 1.7 mg (average ± standard deviation, n = 4). A supernatant containing butanol was added to the CCs, and the temperature was increased from RT to 70 °C for butanol extraction. For back-extraction, the supernatant was replaced with fresh supernatant containing no butanol, and the temperature was decreased first to 60 °C, then to 40 °C, and finally to RT (Fig. 5D). Consistent with the observations of Fig. 5B, increasing the temperature to 70 °C substantially increases the butanol content in the CC. Fig. 5B shows an approximate quadrupling of the K D , whereas Fig. 5D only shows a CC butanol increase from 8.80 ± 0.03 to 20.39 ± 0.80 mg, corresponding with a decrease of the supernatant butanol concentration from 4.39 ± 0.00% at RT to 2.28 ± 0.14% at 70 °C. A possible explanation for this discrepancy is the difference in the total polyion concentration, as Fig. 3G/H/I and 4C show that increased polyion concentrations do not necessarily lead to an increase in partitioning. By replacing the supernatant and lowering the temperature in the steps from 70 °C back to RT, 21.1 ± 0.6% of the butanol extracted into the CC could be back-extracted into a new aqueous solution. Interestingly, reverting the temperature back to RT did not completely revert the butanol equilibrium and a fraction of butanol remains within the CC. Considering the large number of tunable parameters, it is likely that with alterations a back-extraction higher than 21.1% is achievable. For example, increasing the salt concentration has been used to back-extract proteins from polyion micelles and polyion precipitates by disrupting the polyion complex, 46 while varying the pH has been used to back-extract proteins, keeping the polyion precipitates intact. 24 Other experimental parameters such as increasing the number of temperature steps or increasing the equilibration time may also prove to be beneficial. Further research should find improved recovery methods as well as better understanding of the physicochemical mechanisms allowing for a larger fraction of the CCextracted butanol to be recovered. ## Lactic acid and butanol distribution in the enzyme-filled complex coacervates We hypothesized that the presence of additional components in the CCs can influence the partitioning behaviour of LA and butanol in those CCs. For this reason, we investigated the distribution of LA and butanol in PEI/PAA CCs that already contained PPL, CALB, or CALA enzymes. Similar to the presence of salt ions, the presence of relatively large enzymes in the CCs may change the structure of the polyion complex by altering the distance between polyions and the properties of the CCwater interface. We fixed the compositions of the systems to the F − at which the maximum K D was found; F − = 0.36 for PPL and CALB, and 0.26 for CALA. Then, LA (Fig. 6) and butanol (Fig. 7) partitioning was studied as a function of the ionic strength at 25 and 50 °C. In Fig. 6, we can see a stabilizing effect of the lipases on the LA distribution coefficients as they no longer strongly increase between 10 and 100 mM NaCl compared to the PEI/PAA CCs without the lipases shown in Fig. 4B. In addition, the presence of PPL slightly increases the 'stable' K D to approximately 5 compared to 3 without PPL. CALB increases the K D to approximately 4. For PPL and CALB, a higher temperature resulted in a slightly lower K D , comparable to values where the lipases were not present at all. Similar to Fig. 4E, there is no strongly noticeable difference between the investigated temperatures. A much stronger effect is observed for the distribution of butanol shown in Fig. 7. For PPL and CALA, the K D values are comparable to CCs without the lipases at K D = 10-20, but with increased temperature the K D values increase to 40-50 for PPL and 30-50 for CALA. Interestingly, in the presence of CALB, the K D for butanol increases linearly with the NaCl concentration (Fig. 7B) at RT, but not at higher temperatures. These 'doped' CCs show different distribution profiles than 'empty' CCs. Doped CCs may shield against the effect of increased NaCl concentration or simply increase the distribution coefficient by up to a factor of 3 compared to empty CCs. All in all, the concept of pre-filled CCs gives another parameter to tune and optimise the extraction potential of complex coacervates. ## Conclusion and outlook We present an exploratory study on new applications of complex coacervates. While the partitioning behaviour of CCs has been noted before, the step to develop them as an extraction medium has been absent. From previous studies as well as the results shown in this study, it has become clear that the partitioning behaviour of the compounds in CCs is a complex subject involving many tunable parameters that individually greatly influence the distribution coefficient between the aqueous environment and the CC. In our study, we showed that the distributions of lipase enzymes, lactic acid, and butanol in PEI/PAA complex coacervates are strongly affected by the CC composition, ionic strength as determined by the NaCl concentration, polyion concentration, temperature, and presence of other compounds in the CC. However, the effect of any of these parameters depends on the partitioned compound examined. For example, we found that the CC composition has a great influence on the K D of lipases (Fig. 3A-C), while it has only a minimal effect on the K D of LA (Fig. 4A). Even within the category of lipases, the effect of the NaCl concentration on the K D of CALA is much stronger than on the K D of PPL (Fig. 3D and F, respectively). The only consistent influences of the parameter found were that higher concentrations of polyions above 5 g L −1 or high concentrations of NaCl led to lower K D values, though a small amount of NaCl was often (but not always) beneficial. The highest K D experimentally found and the corresponding parameters for the 5 compounds are presented in Table 1. We demonstrated that several relatively simple and tunable parameters can change the K D by a factor of 4 for lipases and butanol as a result of the changes in the CC composition (Fig. 3A-C) and temperature (Fig. 5B), respectively. It is unfortunate that many studies investigating the partitioning behaviour of solutes in CCs do not investigate different compositions, and instead fix it at F − = 0.50 where they might miss either compositions with greater partitioning or with greater PEC formation. 27,28,30 As is demonstrated with the PEI/PAA system, we have shown that it is far from a safe assumption that the optimal polyion complex formation takes place at F − = 0.50, let alone the assumption that the desired partitioning properties are optimal at this composition. Special emphasis has been laid on temperature as a parameter that is easily physically tunable without adding or removing chemicals to or from the system. Using temperature, we created a PEI/PAA CC temperature-swing extraction system that can extract approximately half the butanol from an aqueous supernatant at 70 °C, and then back-extract 21.1% of the extracted butanol back into a new aqueous phase at RT in a single-step system. In this way, CC extraction media can be considered analogous to, for example, cyclic CO 2 absorption, where typical cyclic capacities are in the order of 5-15%. 63 However, considering the number of tunable parameters, it is almost certain that the cyclic capacity can be made much more efficient, and that extraction/back-extraction of a variety of small molecules as well as proteins is possible. While the results for our butanol extraction were not directly comparable in efficiency to some of the results shown by ATPS systems, 64 where up to 95% of a protein was purified in a single step, such high extraction numbers have been shown with different PECs for different proteins, 24 suggesting that a similar potential for CCs exists. There are several limitations of this study. Some of the experimental protocols in these experiments, such as centrifuging at 12 500 g for 30 minutes, are impractical for industrial applications. These protocols were based on earlier fundamental research 24,25 and it is likely (but not verified) that centrifuging at far lower speeds and durations is sufficient. Indeed, the butanol (back-)extraction was performed without additional centrifugation steps after the addition of butanol to the system. The reasons for the variation in K D values and the mechanisms determining the distributions in CCs or other PECs are not well understood. The partitioning behaviour is currently not well understood and cannot yet be accurately predicted. Currently, this means that extensive testing for the individual compound, polyion pair, and tunable parameters is required in order to learn how the parameters influence partitioning. It would be extremely beneficial for the development of CCs as extraction media if the fundamental mechanisms of partitioning in CCs were better understood. The ability to predict the influence of (combinations of ) parameters on partitioning prevents the necessity of high-throughput testing to optimise the parameters for the extraction of a particular desired compound. With a greater understanding of the underlying mechanisms, complex coacervates show promise as extraction media for a wide variety of compounds. The partitioning of solutes in CCs and PECs is the result of a complex interplay of at least 6 different compounds ( polyanion, polycation, water, two salt ions, and the solute of interest), and the temperature will affect the interactions between all these compounds, making it difficult to predict the partitioning behaviour. For proteins, it is expected that the charge and charge distribution are important, and hydrophobic interactions will also play a role. The temperature-dependent partitioning of butanol is promising, but systematic studies are required to unravel the detailed molecular mechanism.

chemsum

{"title": "Complex coacervates as extraction media", "journal": "Royal Society of Chemistry (RSC)"}

enhancement_of_photochemical_heterogeneous_water_oxidation_by_a_manganese_based_soft_oxometalate_imm

## Abstract: Development of efficient and oxidatively stable molecular catalysts having abundant transition metals at the active site is an immediate challenge to synthetic chemists in order to photochemically split water into clean fuels oxygen and hydrogen to serve the ever-increasing energy demand. Herein we report a soft-oxometalate (SOM)-based heterogeneous photocatalytic system which effectively performs water oxidation giving oxygen. In the present work we placed a double sandwich type manganese-based polyoxometalate (POM), Na 17 [Mn 6 P 3 W 24 O 94 (H 2 O) 2 ]Á43H 2 O, on an electroactive graphene oxide matrix and synthesized a new SOM [Na 17 [Mn 6 P 3 W 24 O 94 (H 2 O) 2 ]Á43H 2 O@graphene oxide] 1 and performed water oxidation with it. The efficiency of photocatalytic water oxidation by SOM 1 is almost double than in the case of Na 17 [Mn 6 P 3 W 24 O 94 (H 2 O) 2 ]Á43H 2 O alone. The rationale behind this lies in the electron accepting nature of the graphene sheets which effectively relay the electrons generated in the water oxidation reaction, thus facilitating the forward reaction and increasing the oxygen yield. Variation of catalyst loading, pH-dependent and time-dependent experiments are performed to study their effects on photocatalytic water-splitting. The reaction kinetics is sigmoidal in nature, suggesting the heterogeneous nature of catalysis.The composite catalyst system is observed to be stable towards the reaction conditions. ## Introduction Water oxidation is one of the most promising routes towards the global goal of alternative energy. Many research groups have developed robust catalysts for efficient water oxidation. Recently chemists have been interested in developing molecular water oxidation catalysts by using cheap and abundant transition metals. 6,8 Different chemical species are used as catalysts for that purpose, e.g. metal organic complexes, nanomaterials, and the recently developed polyoxometalates. 6,8, Polyoxometalates (POM) show higher stability in an oxidizing environment compared to metal-organic complexes where organic ligands tend to get easily oxidized and thus offer better catalyst stability. Different routes of water splitting have been explored, such as chemical, 9,10 electrochemical 2,43-45 and photoelectrochemical methods. However, photochemical water oxidation seems to be the most facile and clean technique. 6,13,53,54 A recent challenge in photochemical water oxidation by polyoxometalates is to enhance the oxygen generation and increase the turnover number (TON) of the reaction. 16 Up to now an iridium based complex has shown the maximum TON reported by Crabtree and Brudvig. 55 We are interested in observing whether the reaction efficiency can be enhanced without changing the active center of the catalyst. It is known that POMs can easily be immobilized on the electroactive surface to form stable composite systems. 56 So, we ask whether it is possible to employ a related composite system to perform water oxidation experiments. 29,33,57 Recently, the Hill group developed a similar method using graphene modified electrodes and ruthenium based POM as active catalyst. 57 The graphene modified electrodes show excellent catalytic activity and high stability toward the electrochemical water oxidation reaction at neutral pH. This work showed enhanced water oxidation reaction electrochemically. 57 Here we ask whether it is possible to make a soft oxometalate 56, based on polyoxometalate-graphene oxide to enhance the efficiency of photochemical water oxidation. In our present work we use a manganese based polyoxometalate Na O@graphene oxide] 1. This catalyst shows a turnover number of 22 at pH 8 for WO reaction. The SOM 1 dispersion is prepared by sonication. Formation of the composite is confirmed by Raman spectra, SEM images and EDX data. Finally we use this SOM as a photocatalyst in water oxidation (Fig. 1). Interestingly, we observe that in the presence of the graphene oxide matrix the water oxidation activity of Mn-POM is almost doubled. A detailed account of synthesis and characterization of the composite catalyst and observations related to photochemical water oxidation studies are provided in the following sections. ## Result and discussion Formation of the SOM 1 composite based on graphene oxide SOM 1 is prepared by following the classical route of immobilization of POM on an electroactive surface. 56 In our present study we initially prepared graphene oxide dispersion in water. To this dispersion Na 17 [Mn 6 P 3 W 24 O 94 (H 2 O) 2 ]Á43H 2 O (Fig. 2) was added and the mixture was sonicated to finally get the composite SOM 1, which forms a stable dispersion. Composite formation was characterized by using Raman spectroscopy, SEM and EDX analysis. ## Scanning electron microscopy (SEM): morphology of SOM 1 GO shows nanosheet type morphology. It is observed from scanning electron microscopy (Fig. 3a). The SOM 1 shows nanospheres embedded on graphene oxide layers (Fig. 3b). The white bright spot indicates the clustering of POM, suggesting that POM units are attached to the surface of GO by electrostatic interactions, as GO has an electron deficient surface (positively charged) and POM are large polyanions (negatively charged). This further indicates the formation of the composite in the reaction system. From the EDX data we can also infer that the molecular integrity of POM is intact in SOM 1. [Note: manganese, phosphorus and tungsten are present in the expected correct ratio of POM in SOM 1 (Fig. 3c).] HATR-IR spectroscopy and stability of POM in SOM 1 HATR-IR spectroscopy of SOM 1 and also HATR-IR spectroscopy of the POM constituent were performed. It was observed that a few broad bands were obtained in each case in the IR spectrum. This broadness is possibly due to the low concentration of the sample in the dispersion. Here we observed common peaks for POM and SOM 1 at 1637, 693, 569, 496 cm 1 respectively (Fig. 4a). Thus from the IR-spectrum we can conclude that the POM constituent remains stable and intact in the SOM 1 after composite formation and no catalyst is degraded at all. We further performed UV-VIS spectroscopy to check changes in the energy gap of the POM constituent after composite formation. ## UV-VIS spectroscopy of the POM constituent and SOM 1 We performed UV-VIS spectroscopy of SOM 1 and the POM constituent in water. For both POM and SOM 1 we got absorbance maxima at 250 nm (Fig. 4b). Thus it may be concluded that the band gap of the POM constituent does not change in the SOM 1 composite, which further proves the stability of the POM constituent in SOM 1 because if it was dissociated to other cluster units then there should have been a clear difference in the UV-VIS spectrum. Raman spectroscopy and the nature of SOM 1 We now want to show the effective formation of SOM. Raman spectra (Fig. 4c) of Na 17 [Mn 6 P 3 W 24 O 94 (H 2 O) 2 ]Á43H 2 O, graphene oxide and the SOM 1 composite were taken; for SOM 1 we observed 4 peaks at 516, 949, 1379, and 1628 cm 1 respectively. We assign these peaks as follows: 516 (n as,Mn-O ), 949 (n WQO ) and the other two peaks at 1379 and 1628 cm 1 for disorder and graphitic nature of graphene oxide respectively. These peaks are blue shifted compared to that of the spectrum of graphene oxide and Na shows peaks at 495 and 927 cm 1 which can be attributed to the following modes: (n as,Mn-O ), (n WQO ). The characteristic peaks at 1371 and 1618 cm 1 are on the other hand due to disorder and graphitic nature of graphene oxide respectively. This shift in the spectrum might indicate that there may be the possible presence of interaction between POM and the graphene oxide layer in SOM 1. In our system electrons are probably transferred from Na 17 [Mn 6 P 3 W 24 O 94 (H 2 O) 2 ]Á43H 2 O to graphene oxide. 62,63 This was further explained by CV. We conclude from this shift that in SOM 1, graphene oxide may act as an electron acceptor and Na 17 [Mn 6 P 3 W 24 O 94 (H 2 O) 2 ]Á43H 2 O may act as an electron donor. The Raman spectrum also reveals that SOM 1 is not a physical mixture of the constituents graphene oxide and POM but an assembly of the two at the molecular level. ## Cyclic voltammogram of the catalyst To further monitor the stability of the POM constituent in SOM 1 we performed CV of SOM 1 and compared it with the CV of the POM constituent of SOM 1 (Fig. 4d) and it is clearly observed that both are identical, which indicates that the redox behavior of the POM constituent in SOM 1 remains unaltered and we can also conclude that the POM constituent is stable after composite formation. Also we observed that the peak current increased for SOM 1 which further indicates facile electron transport from the POM constituent to GO surface. ## Photochemical water splitting Photochemical water splitting experiments were performed under a UV lamp (l max = 373 nm) with the composite catalyst system. The composites were prepared as mentioned in the previous section. The oxygen evolution was monitored using a YSI optical sensor based dissolved oxygen meter and also by cyclic voltammetry. The maximum oxygen yield obtained was 19.2 mmol for 20% SOM 1 loading at pH 8 in phosphate buffer. The graphene oxide acts as an electron acceptor and traps the electrons released in the water oxidation reaction and facilitates electron transport as well. ## Confirmation of water oxidation and the effect of graphene oxide In our present work SOM 1 absorbed light and was elevated to the excited state. This excited SOM 1 generated hole and electron pairs, and the holes oxidized water to oxygen in the presence of light. After photoillumination quantitative determination of evolved oxygen was performed by measuring the evolved oxygen (Fig. 5c) using a YSI optical sensor based dissolved oxygen meter. For further confirmation of evolution of oxygen cyclic voltammetry (Fig. 5a) was performed using samples after photoirradiation, where a sudden rise of current was observed near +1.2 V with respect to the Ag/AgCl reference electrode indicative of oxygen evolution from water. It implies oxidation of water. We observe that the extent of oxygen evolution is almost doubled in the case of the SOM 1 composite catalyst as compared to water oxidation by Na 17 [Mn 6 P 3 W 24 O 94 (H 2 O) 2 ]Á43H 2 O alone. Using POM alone the maximum amount of oxygen liberated is almost 3.2 mmol for 0.071 mmol loading of the catalyst, with a TON of around 46, whereas in the case of the SOM 1 composite catalyst system the amount of O 2 evolved is almost 6.5 mmol for 0.071 mmol loading of the catalyst, with a TON of 92, which is roughly double than that of the POM alone. We thus investigate the role of graphene oxide in water oxidation. In the next set of experiments, Na 17 [Mn 6 P 3 W 24 O 94 (H 2 O) 2 ]Á 43H 2 O loading was kept constant (10 mg/10 ml in all the catalyst dispersions) and the graphene oxide concentration was increased (Fig. 5b). Here we observe a similar type of sigmoidal curves and the maximum O 2 generation is almost 58 mmol for 10% SOM 1 loading. The comparative studies clearly show that the increase of graphene oxide loading has a prominent effect on water oxidation. We observed two different aspects: (i) up to a certain loading of graphene oxide (5 mg) O 2 evolution increases to a maximum of 58 mmol and (ii) thereafter the increase of graphene oxide loading has no further effect on O 2 evolution. We explain this as follows. With the increase in graphene oxide loading the extent of electron relay facilitated by graphene oxide increases, thereby increasing the effective O 2 evolution. However beyond a certain threshold of graphene oxide loading the active POM concentration in SOM 1 gets diluted. Hence the evolution of O 2 does not increase anymore. Needless to say, graphene oxide invariably enhances the water oxidation activity. In SOM 1 we also infer that graphene oxide most likely (1) provides an enhanced active catalytic surface area and (2) facilitates electron transport and thereby enhances water oxidation effectively. Also one of the prominent reasons for the enhancement of water oxidation by using the GO matrix may be due to the increase in the effective surface area of the catalyst. In the case of SOM 1, the hydrodynamic radius is around 300 nm (from dynamic light scattering experiments) as compared to single SOM having a hydrodynamic radius of 130 nm. We have calculated the surface area by assuming the catalyst materials to form nanospheres in dispersion and for a spherical surface we calculated the area by the following equation: surface area = 4pR h 2 . Now we address the problem of how water oxidation is affected with variation in pH and loading of SOM 1 dispersion in the next section. pH dependent study pH dependent water oxidation study reveals some interesting results (Fig. 5d). We observe that with the increase in pH the amount of evolved oxygen increases gradually. At pH 8 we observe the maximum yield. On further increasing the pH, oxygen evolution decreases abruptly and this may be attributed to the degradation of clusters at higher pH. This observation may be explained by the shift of equilibrium involved with oxygen evolution to the right with the increase in pH. ## Catalyst loading variation studies In this set of experiments, the graphene oxide concentration is kept constant (1 mg/10 ml in all the catalyst dispersions) and the SOM 1 loading is increased to observe the change in water oxidation. It is observed that with increasing POM loading oxygen evolution increases for a given pH. The nature of the oxygen evolution curve with catalyst loading reveals that initially with increasing catalyst loading oxygen evolution increases rapidly, but after exceeding a certain loading of POM on SOM 1 catalyst (Fig. 6), the rate of enhancement of oxygen evolution in the reaction decreases to some extent. This may be due to the stability factor of the dispersion. More precisely oxygen evolution decreases when phase separation is observed and when we cross the dispersion stability window. This decrease in oxygen evolution is also due to the decrease in the active surface area of the catalyst. ## Time depended studies of oxygen evolution reaction Time dependent water oxidation experiments show general sigmoidal kinetic patterns typical of heterogeneous catalysis reactions, where up to a certain limit of time oxygen evolution increases and reaches a plateau (Fig. 7a). There is an induction period of reaction which may be due to light absorption limitation. For excitation of SOM 1 it needs to cross a minimum energy barrier, which is attained after some time and therefore initially there is no reaction. When SOM 1 possesses minimum energy for excitation water oxidation starts (Fig. 7b). As water is taken in excess in the reaction, the reaction rate only depends on the intensity of light and not on the amount of water present in the reaction medium. At early times of the reaction, i.e., at low light intensity (up to 9.32 mW cm 2 ) there is no O 2 evolution. However, beyond a threshold light intensity (9.32 mW cm 2 ) O 2 evolution begins. The induction period (before threshold light intensity) probably simply reflects the time for O 2 product equilibration before the analysis of evolved O 2 . It increases in a sigmoidal fashion suggesting co-operative photo-activation of the SOM sites for water oxidation. However with the increase in energy density oxygen evolution reaches saturation. Hence in other words it might be said that water oxidation reaction requires a threshold energy density to begin with, then increases in a sigmoidal fashion implying co-operative photo-activation of the SOM sites, finally reaching saturation with energy density. Thus it implies that the water oxidation reaction is topped off after a certain time. The maximum TOF of the reaction is 0.75 min 1 which is comparatively less than that of the recently developed photochemical water oxidation using cobalt based POM. This difference in TOF may be due to the involvement of different redox couples in the reaction. Here we observe that the maximum amount of O 2 generated is almost 19 mmol for 20% of SOM 1 loading. Now we ask whether SOM 1 is stable in the course of the reaction. To determine its stability we measured the Raman spectrum of SOM 1 before and after the completion of the reaction. A detailed account of this study is provided in the next section. ## Stability of the composite SOM catalyst Raman spectroscopic investigations were performed on the SOM 1 composite catalyst before and after the reaction. The spectra were observed to be identical (Fig. 8a). We also performed the HATR-IR (Fig. 8b) and UV-VIS (Fig. 8c) spectroscopy of the SOM 1 catalyst after the reaction. We also performed cyclic voltammetry (Fig. 8d) with the post reaction dispersion and in all the cases we observed identical spectra compared with the spectra obtained with the dispersion SOM 1 before the reaction. So we can possibly comprehend that the composite catalyst system remains intact during the water splitting reaction. Thus the reported catalyst system is stable under the water oxidation conditions that are used in this study. The POM constituent does not dissociate to form MnO 2 or some other fragment. Therefore during photochemical water oxidation reaction it is reasonable to believe that no MnO 2 is generated under reaction conditions which can possibly oxidize water. Only the POM constituent is clearly responsible for the water oxidation reaction. More experiments and analyses are needed to pin-point the active species, excited species lifetime and other deeper mechanistic details which will be performed by us in the future. We also took SEM images (Fig. 9) of the post reaction composite catalyst and observed almost a similar kind of morphology as in the images taken before the commencement of the reaction. ## Catalytic recyclability of SOM 1 As the catalyst is stable after the reaction, we can effectively reuse this catalyst for further catalytic cycles. For this purpose we checked the recyclability up to 10 catalytic cycles and we observed that each time an equal amount oxygen is evolved in the catalytic cycle (Fig. 10). Therefore the catalyst is completely reusable. ## Mechanism of evolution of O 2 from water Photochemical water oxidation with polyoxometalates generally takes place in the presence of an additional photosensitizer and a sacrificial electron acceptor. In our present work, the photocatalytic heterogeneous reaction possibly follows a completely different pathway. Here we need not add any photosensitizer and sacrificial electron acceptor. SOM 1 itself may be absorbing light and going to the excited state which oxidizes water to oxygen (Fig. 11), but elucidation of the actual photophysical mechanism will require additional studies. To further prove that the excited species oxidized water we added catechol in the reaction and observed that water oxidation ceased under these conditions. This may be due to the fact that catechol oxidation is more favorable compared to water oxidation and therefore water oxidation does not take place in the presence of catechol. The graphene oxide sheets are expected to act as electron acceptor platforms for the electrons generated in the water oxidation process 64,65 and also enhance the surface area of the POM constituent of SOM 1. ## Conclusion To summarize, we have demonstrated the water oxidation by Mnpolyoxometalate (POM) based soft-oxometalate (SOM 1) dispersion and the efficiency is almost doubled by immobilizing Mn-POM on an electroactive graphene oxide matrix. The catalyst system acts as a water oxidizing agent to generate oxygen under photochemical conditions. The graphene oxide layers possibly act as electron acceptors and surface area enhancers and facilitate water oxidation by SOM 1. Thereafter we describe the effect of catalyst loading and pH on photocatalytic water-splitting. From the kinetics of the reaction we show the operation of heterogeneous mode of catalysis. After demonstrating the stability of the catalyst in the course of the water splitting reaction we have proposed the plausible pathway of the catalyst action. Further work is in progress in our laboratory in order to design more SOM based water splitting catalysts. ## Materials and reagents All the materials were purchased from commercially available sources and used without further purification. All the glass apparatus were first boiled in an acid bath, then in water and finally rinsed with acetone. All the glass apparatus were properly dried in a hot air oven overnight. Doubly distilled deionized water was used to carry out all the experiments. ## Synthesis of graphene oxide Graphene oxide was synthesized by the improved Hummers' method. Hummers' method 69 (KMnO 4 , NaNO 3 , H 2 SO 4 ) is the most common method used for preparing graphene oxide. A recent methodology study has modified the process to some extent and improved the efficiency of the oxidation process and this modified Hummers' method 68 was employed here to synthesize graphene oxide for our experiments. Concentrated H 2 SO 4 (69 ml) was added to a mixture of graphite flakes (3.0 g, 1 wt equiv.) and NaNO 3 (1.5 g, 0.5 wt equiv.), and the mixture was cooled using an ice bath to 0 1C. KMnO 4 (9.0 g, 3 wt equiv.) was added slowly to keep the reaction temperature below 20 1C as KMnO 4 addition is exothermic. The reaction was warmed to 35 1C and stirred for 7 h. Additional KMnO 4 (9.0 g, 3 wt equiv.) was added in one portion, and the reaction was stirred for 12 h at 35 1C. The reaction mixture was cooled to room temperature and poured into ice with 30% H 2 O 2 (3 ml). The mixture was then purified following the usual protocol of sifting, filtering, centrifugation, decanting with multiple washes followed by a final vacuum drying to give 4.0 g of solid product. ## Synthesis of SOM 1 1 mg of graphene oxide was added into 10 ml of water and 2 ml of ethylene glycol was added to it for better separation of the graphene oxide sheets. Then it was sonicated for 3 hours at room temperature to prepare graphene oxide dispersion. After that, 10 mg of Na 17 [Mn 6 P 3 W 24 O 94 (H 2 O) 2 ]Á43H 2 O was added and the dispersion was sonicated for 3 more hours. The stability of the dispersion was checked and it was found to be stable. ## Photocatalytic water splitting Photocatalytic water splitting reactions were performed as follows. In the composite dispersion for water oxidation experiment buffer solution of pH 7 was added. The reaction mixture was then sealed and N 2 gas was purged for 3 hours to get rid of the trace amount of oxygen in it. Then the reaction mixture was kept in a photoreactor under UV-light (energy density of the photoreactor is 19.5 mW cm 2 with l max = 373 nm) for 2 hours. After irradiation we measured the amount of evolved oxygen in the reaction by using YSI optical sensor based dissolved oxygen meter standardized by using degassed double distilled water. Evolution of oxygen in the reaction was further investigated by performing cyclic voltammetry using the irradiated samples. In cyclic voltammetry we observed a sharp rise of the current-voltage curve near +1.2 V, which is typically indicative of O 2 generation by water splitting. ## pH dependent water splitting This experiment was performed by following the previous procedure using different buffer solutions in the pH range of 5 to 9. Measurement of oxygen evolution was carried out by a similar method mentioned earlier. ## Characterization techniques SEM-EDX microscopy. SEM measurements were done by drop-casting SOM 1 dispersion on a silicon wafer and drying under vacuum for 2 days. Then SEM imaging was performed and images were taken on a SUPRA 55 VP-41-32 instrument with the Smart SEM version 5.05 Zeiss software. Cyclic voltammetry. A PAR model 273 potentiostat was used for CV experiments. A platinum wire auxiliary electrode, a glassy carbon working electrode with a surface area of 0.002826 cm 2 and an aqueous Ag/Ag + reference electrode which is filled with saturated KCl solution were used in a three electrode configuration. The scan rate was 0.5 V s 1 . The CV spectrum was recorded in the range of 0 to +1.3 V. Blank refers to the amount of oxygen present in distilled water in our mentioned reaction conditions. The pH of the medium was 7. 0.1 M KCl solution was used as a supporting electrolyte in all the experiments. All measurements were done at 298 K in an inert atmosphere. Dynamic light scattering measurements. The average size of the particle was obtained using the dynamic light scattering method in a Malvern Zetasizer instrument. A very dilute solution of SOM 1 was prepared by further dilution of the SOM 1 dispersion and taken in a fluorescence glass cuvette with a square aperture and the instrument was set to take 15 runs before measuring the average hydrodynamic radius of the SOM 1 composite. Raman spectroscopy. A LABRAM HR800 Raman spectrometer was employed using the 633 nm line of a He-Ne ion laser (l = 633 nm) as the excitation source to analyze the sample.

chemsum

{"title": "Enhancement of photochemical heterogeneous water oxidation by a manganese based soft oxometalate immobilized on a graphene oxide matrix", "journal": "Royal Society of Chemistry (RSC)"}

mesoscopic_superstructures_of_flexible_porous_coordination_polymers_synthesized_<i>via</i>_coordinat

## Abstract: The coordination replication technique is employed for the direct conversion of a macro-and mesoporous Cu(OH) 2 -polyacrylamide composite to three-dimensional superstructures consisting of the flexible porous coordination polymers, Cu 2 (bdc) 2 (MeOH) 2 and Cu 2 (bdc) 2 (bpy) (bdc 2À ¼ 1,4-benzenedicarboxylate, bpy ¼ 4,4 0 -bipyridine). Detailed characterization of the replicated systems reveals that the structuralization plays an important role in determining the adsorptive properties of the replicated systems, and that the immobilization of the crystals within a higher-order architecture also affects its structural and dynamic properties. The polyacrylamide polymer is also found to be crucial for maintaining the structuralization of the monolithic systems, and in providing the mechanical robustness required for manual handling. In all, the results discussed here demonstrate a significant expansion in the scope of the coordination replication strategy, and further confirms its utility as a highly versatile platform for the preparation of functional three-dimensional superstructures of porous coordination polymers. ## Introduction The design and synthesis of porous coordination polymers (PCPs) or metal-organic frameworks (MOFs) has experienced an intensive focus in recent years, 1 due to their potential use in applications such as gas storage, molecular separations, and heterogeneous catalysis. 2 These compounds are assembled from metal-containing nodes bridged by organic linkers, which form porous structures that are characterized by high surface areas, as well as tunable pore dimensions and pore surface chemistry. While the ability to conveniently construct new materials from the combination of a metal salt and organic ligand (in the so-called modular approach) has provided researchers with a tremendously large library of compounds, there is an urgent need for versatile synthetic strategies for the convenient fabrication of PCPs in a structuralized form. 3 Here, synthetic routes have begun to emerge for the bottom-up preparation of zero-(e.g. hollow spheres), one-(rods), two-(flms), and three-dimensional (monolithic) superstructures of PCPs. 3a,b A feature common to the preparative methodologies of the systems reported so far is that they provide a precise control of the crystallization interface at which PCP formation occurs, resulting in the precipitation of the PCP with the desired structuralized architecture. An elegant technique that has recently emerged for the preparation of three-dimensional superstructures of PCPs is the so-called coordination replication strategy. In this method, a structuralized metal source (such as a metal oxide) is employed as a template, which undergoes conversion in a ligand solution into a three-dimensional PCP superstructure with retention of the original structure. While this technique has been successfully demonstrated with a small number of PCP systems so far, 4 investigations of the incorporation of molecular-scale flexibility within structuralized systems with sophisticated dynamic properties are yet to emerge. While studies of this type are of high interest from a fundamental perspective due to the prospects of new phenomena emerging from the embedding of such dynamic building blocks in a structuralized form, the identifcation of suitable starting materials and PCP systems is challenging due to the difficulty in preparing metal-based compounds in well-defned structures, as well as the currently limited scope of structuring techniques. In this work, we address these challenges via the structuring of flexible copper-based PCPs, namely Cu 2 (bdc) 2 (MeOH) 2 , which has a two-dimensional interdigitated structure, and Cu 2 (bdc) 2 (bpy), which comprises a three-dimensional interpenetrated structure, into three-dimensional monolithic superstructures (bdc 2 ¼ 1,4-benzenedicarboxylate, bpy ¼ 4,4 0bipyridine). 8 A macro-and mesoporous Cu(OH) 2 -polyacrylamide (PAAm) monolithic material was chosen as a precursor for the coordination replication strategy, which was frstly successfully converted into a Cu 2 (bdc) 2 (MeOH) 2 monolith ("daughter" phase), followed by a PCP-to-PCP replication to fabricate a Cu 2 (bdc) 2 (bpy) monolith ("granddaughter" phase) via the pillar ligand (bpy) insertion process (see Fig. 1). Importantly, unique adsorptive and dynamic properties are observed following immobilization of the PCPs within the three-dimensional superstructures, and the potential origins of these effects are discussed in the context of both the composition and the structures of the monoliths. ## General considerations Unless otherwise noted, all reagents were obtained from commercial vendors and used as received. While all syntheses were carried out in the air, the desolvated forms of each of the compounds were handled and stored in a nitrogen-flled glove box. Solvothermal syntheses were carried out in a DKN302 constant temperature oven (Yamato Scientifc Co., Ltd) using glass vials sealed with Teflon-lined lids. Nitrogen and methanol adsorption measurements were carried out on a BELSORP-max adsorption analyser (BEL Japan, Inc.) equipped with a constant temperature bath. Powder X-ray diffraction patterns were collected using a Smartlab X-ray Diffractometer (Rigaku Corp.) equipped with a Cu Ka source. ## Field-emission scanning electron microscopy (FE-SEM) Scanning electron microscopy (SEM) images were collected using a JEOL JSM-7001F4 electron microscope. Powder and monolith samples were evacuated to remove any residual solvent molecules, and attached to a 13.5 mm substrate using double-sided carbon tape. The samples were then coated with osmium nanoparticles to a thickness of 5 nm, and transferred to the SEM instrument. The images were collected using an emission voltage between 10 and 15 kV. ## Synthetic procedures Cu(OH) 2 -polyacrylamide monolith. The parent phase was synthesized by sol-gel processing as reported recently, 9 using a starting mixture of CuCl 2 $2H 2 O (1.53 g, 8.97 mmol), polyacrylamide (PAAm; 0.60 g, M w $ 10 000), water (1.10 mL), ethanol (0.30 mL), glycerol (2.40 mL), and propylene oxide (1.47 mL, 21.0 mmol). The as-synthesized form of the monolith was stored in 2-propanol, and was rinsed with methanol prior to the coordination replication procedure. Note that after washing, some Cl ions still remain in the composition (ca. 4 wt%), 9 but we refer to the starting structure as "Cu(OH) 2 -polyacrylamide monolith" for simplicity. Fig. 1 A conceptual illustration summarizing the two-step replication procedure employed in this work. In the first step, a macro-and mesoporous Cu(OH) 2 -polyacrylamide (PAAm) composite is subjected to a coordination replication process via treatment with H 2 bdc (bdc 2 ¼ 1,4benzenedicarboxylate), resulting in a monolith consisting of the two-dimensional layered framework, Cu 2 (bdc) 2 (MeOH) 2 . During this step, there is a significant increase in the internal solid volume (versus void volume) due to the Cu 2 (bdc) 2 (MeOH) 2 crystals occupying a much greater volume compared to the precursor. In the actual monolith, this largely eliminates the macroporosity within the structure while keeping the external macroscopic dimensions. In the second step, the obtained monolith is subjected to a PCP-to-PCP replication procedure in the presence of 4,4 0bipyridine (bpy), which leads to the pillaring of the two-dimensional layers and formation of a monolith constructed from the three-dimensional, interpenetrated Cu 2 (bdc) 2 (bpy) framework. Inset: portions of the structures of each of the PCP compounds (one half of the interpenetrated framework of Cu 2 (bdc) 2 (bpy) is shown faded). Green, grey, blue, and red spheres represent Cu, C, N, and O atoms, respectively. H atoms, and solvent molecules (except for the directly coordinated atom) have been omitted for clarity. Bulk Cu 2 (bdc) 2 (bpy). To a 500 mL round-bottom flask, H 2 bdc (210 mg, 1.26 mmol) and methanol (200 mL) were added, and the mixture was refluxed under Ar for 2 h. After this time, a commercially-obtained Cu(OH) 2 powder (121 mg, 1.24 mmol) was added, and the solution was refluxed for a further 3 days. After this time, a sky-blue precipitate was formed, and the reaction solution was cooled to room temperature. Then, a mixture of bpy (100 mg, 0.64 mmol) in methanol (100 mL) was added to the flask, and the solution was stirred vigorously for 3 days at room temperature. This induced a color change of the solid to pale-green. The resulting solid was isolated by vacuum fltration, washed with methanol (3 50 mL), and dried under a reduced pressure. Cu 2 (bdc) 2 (MeOH) 2 -polyacrylamide monolith. An approximately 5 mm 5 mm 5 mm piece of the as-prepared Cu(OH) 2 -polyacrylamide monolith (Fig. 2, left) was inserted into a tapered glass tube (i.d. 8 mm), and placed into a 20 mL glass vial containing H 2 bdc (50 mg, 0.30 mmol) and methanol (20 mL). The vial was sealed and placed in an oven set to a temperature of 60 C for 12 h, after which time the color of the monolith changed from green to sky-blue (see Fig. 2, center). The glass tube (containing the monolith) was then removed and placed in a fresh ligand solution of the same composition, and placed back in the oven for a further 12 h. This procedure was repeated until the total reaction time was 7 days, after which time the fully replicated monolith was washed by immersion in neat methanol (50 mL) for 24 h to remove any unreacted H 2 bdc. The washing procedure was repeated three times, and the material was stored in neat methanol to avoid degradation of the resultant structure. Cu 2 (bdc) 2 (bpy)-polyacrylamide monolith. In a 100 mL glass vial, the Cu 2 (bdc) 2 (MeOH) 2 -polyacrylamide monolith obtained in the previous step was immersed in methanol (45 mL). Then, a solution of bpy (10.0 mg, 64.0 mmol) dissolved in methanol (5 mL) was slowly added, and the contents of the vial were allowed to stand undisturbed at room temperature for 24 h. Then, 5 mL of the solution was removed, replaced with a bpy solution with the same composition added initially, and the mixture was allowed to stand for a further 24 h. This procedure was repeated until the total reaction time was 5 days, after which time the color of the monolith had changed to blue-green (Fig. 2, right). The solid was washed and stored using the same method as described for the Cu 2 (bdc) 2 (MeOH) 2 -polyacrylamide monolith. ## Synthesis and characterization The coordination replication technique is an attractive method for the structuralization of PCP materials, since a potentially wide variety of metal-based precursors can be shaped into a desired form via conventional fabrication techniques, such as sol-gel processing. Here, the main requirement for precursor materials is a slow dissolution rate relative to the crystallization rate of the target PCP crystals, such that crystal growth is spatially constrained at the interface between the solid precursor and the ligand solution. 4 This represents one of the main challenges in expanding the scope of coordination replication synthesis, and precursors that offer the correct balance between solubility and reactivity under the reaction conditions for the PCP formation process have remained limited so far. Among the solid sources containing Cu 2+ considered for the formation of the Cu 2 (bdc) 2 (MeOH) 2 and Cu 2 (bdc) 2 (bpy) frameworks, Cu(OH) 2 was chosen for further study owing to its low solubility in polar organic solvents and high reactivity toward acids. 10 Consequently, synthetic conditions for the synthesis of the two compounds were developed using a commerciallyavailable bulk crystalline powder of Cu(OH) 2 . The screening of various parameters, including the reaction solvent, the metalto-ligand ratio, and the reaction time revealed that the addition of a stoichiometric quantity of Cu(OH) 2 to a refluxing solution of H 2 bdc in methanol afforded Cu 2 (bdc) 2 (MeOH) 2 after a reaction time of 3 days. Next, a suspension of Cu 2 (bdc) 2 (MeOH) 2 was treated with an excess of bpy in methanol, resulting in the installation of bpy pillars between every second square grid layer to produce the interpenetrated Cu 2 (bdc) 2 (bpy) framework (see Fig. S3 †). SEM observation confrmed a plate-like crystal morphology (Fig. S4 †), and N 2 adsorption measurements (Fig. S5 †) at 77 K gave a BET surface area 11 of 1030 m 2 g 1 (Langmuir surface area: 1300 m 2 g 1 ) which is somewhat higher than the corresponding value of 700 m 2 g 1 measured previously for a sample prepared from a conventional method that uses CuSO 4 as the Cu 2+ source. 8a Following the successful demonstration of the synthesis of Cu 2 (bdc) 2 (MeOH) 2 and Cu 2 (bdc) 2 (bpy) from crystalline Cu(OH) 2 powders, a structuralized form of Cu(OH) 2 was required for coordination replication studies. Recently, a method for the preparation of an amorphous macro-and mesoporous Cu(OH) 2 -polyacrylamide composite material via sol-gel processing accompanied by phase separation, and its conversion to the prototypical and rigid PCP, Cu 3 (btc) 2 , was reported. 9,12 The amorphous nature of the Cu(OH) 2 within this parent phase is expected to have a similar (or enhanced) reactivity compared to crystalline Cu(OH) 2 , and was identifed as a suitable candidate for further study. In this case, the synthetic procedure of the Cu(OH) 2 -polyacrylamide parent phase was adapted to prepare a monolithic solid featuring continuous macropores with a diameter of ca. 1 mm (see Fig. 2, left, and Fig. 3A). Analysis of the porosity of the parent monolith used in this work via N 2 adsorption isotherms afforded a type-IV profle typical of a mesoporous solid (Fig. S6 †). The determination of a pore size distribution based on this data revealed a maximum density at ca. 5 nm for the mesopores. Note that the large pores present within this monolith are expected to facilitate the diffusion of the organic linkers throughout the solid, which is required for full conversion during the coordination replication procedure. The Cu(OH) 2 -polyacrylamide monolith was suspended and heated within a solution of H 2 bdc in methanol for an extended period of 7 days (with daily exchange of the mother liquor), which resulted in a color change of the solid from green to skyblue. Importantly, the external dimensions of the monolith and its mechanical integrity were retained despite the long period of treatment (see Fig. 2, center). 13 Observation of the surface of the monolith following replication by SEM revealed the growth of square plate-like crystals approximately 1 mm in width from the walls of the co-continuous structure (see Fig. 3B). SEM observation following slicing of a monolith sample to expose the cross-section (depth direction) of the structure showed crystals of the same morphology had uniformly formed throughout the material (see Fig. S7 †), but the macropores were almost completely eliminated. This is because the conversion from Cu(OH) 2 to Cu 2 (bdc) 2 (MeOH) 2 results in a volume increase of approximately 10 times (based on Cu 2+ density in the bulk, crystalline forms of both compounds). The complete conversion of the Cu(OH) 2 of the parent phase was further confrmed by thermogravimetric analysis (TGA), which did not exhibit a weight loss at the decomposition temperature of Cu(OH) 2 of ca. 80 C (see Fig. S8 †). The TGA data could also be used to estimate a polyacrylamide content of 15.0 wt%, which is close to the composition employed during the preparation of the Cu(OH) 2polyacrylamide monolith of ca. 20.0 wt%. Note that, in the preparation of the Cu 2 (bdc) 2 (MeOH) 2 monolith, the reaction conditions developed for the preparation of bulk powders of the same compound from crystalline Cu(OH) 2 was successfully used for monolith conversion. This agrees with our experience using the coordination replication method for the synthesis of Al-based PCP architectures from Al 2 O 3 phases, 4 which has demonstrated that amorphous or less-dense variants of an inorganic compound tend to dissolve faster or have higher reactivities since they have lower lattice energies. This results in the right balance between precursor dissolution and PCP crystallization, which is required for preservation of the structuring of the parent phase. Next, the Cu 2 (bdc) 2 (MeOH) 2 monolith was immersed in a methanol solution of bpy to induce pillaring of the square grid layers of the two-dimensional framework to afford the three dimensional Cu 2 (bdc) 2 (bpy) compound. After several hours, the color of the monolith changed from sky-blue to blue-green (see Fig. 2, right). SEM data revealed the retention of the structuralization of the monolith following replication accompanied with a slight increase in the thickness of the crystals, which is consistent with the insertion of the bpy pillars between the dinuclear copper paddlewheels of every second square grid layer (Fig. 3C). Estimation of the composition of the monolith via TGA data revealed a polyacrylamide content of 2.0 wt% (Fig. S16 †), the loss of which, as discussed in the following section, has important consequences with respect to the properties of the monoliths. ## Structural features and structural exibility of the replicated frameworks Cu 2 (bdc) 2 (MeOH) 2 monolith. The properties of the replicated monolith were probed using a combination of powder Xray diffraction, SEM, TGA, infrared spectroscopy, and sorption experiments. Diffraction patterns obtained from a solvated fragment of the replicated solid were indicative of a highly crystalline framework phase, with reflections that were wellmatched with those of solvated bulk Cu 2 (bdc) 2 (MeOH) 2 (see Fig. 4). Surprisingly, a signifcant number of peaks were absent in the diffraction pattern of the monolithic phase. Assignment of the diffraction peaks observed for the monolith revealed that the 0k0, 00l, and 0kl reflections were present, while all reflections with a non-zero h component were signifcantly broadened or absent. 14 The structure of the Cu 2 (bdc) 2 (MeOH) 2 compound is such that the crystallographic a-axis (i.e. the h00 reflection) represents the periodicity of the stacking of the twodimensional square grids (see Fig. 1), and the absence of these reflections can be attributed to its disruption (or "amorphization") upon integration into the monolith. This is analogous to a phenomenon observed in carbon-based materials with a turbostratic structure, in which 00l reflections are prominently observed (with a broadened peak width) compared to its crystalline counterpart, graphite. 15 The origins of this unusual feature of the powder X-ray diffraction data were further probed by N 2 adsorption analysis at 77 K after activation of the monolith at 150 C. 16 Fig. 5 displays data collected for the parent Cu(OH) 2 -polyacrylamide monolith, the Cu 2 (bdc) 2 (MeOH) 2 monolith and a bulk Cu 2 (bdc) 2 (MeOH) 2 powder sample. Remarkably, while the bulk material showed a negligible N 2 uptake owing to the inability of N 2 to open and access the interlayer spacing, the structuralized variant exhibited signifcant uptake at low pressures, reminiscent of a type-I isotherm observed for a microporous solid. Indeed, a BET analysis of the sorption data (see Fig. S9 †) afforded a surface area of 520 m 2 g 1 , 17 which is signifcantly greater than can be accounted for by the sorption properties of the parent Cu(OH) 2 -polyacrylamide phase and bulk Cu 2 (bdc) 2 (MeOH) 2 . This suggests that the structural influence of the interactions between the Cu 2 (bdc) 2 (MeOH) 2 crystals and the polyacrylamide chains at the molecular scale in turn impart considerably different sorption properties to the PCP phase compared to its bulk counterpart. The powder diffraction and adsorption data observed here can be reconciled by considering the role of the polyacrylamide polymer in the replicated system. The polyacrylamide content of the Cu 2 (bdc) 2 (MeOH) 2 monolith of approximately 15.0 wt% is a component required for the integrity of the three-dimensional structuralization. Here, it is expected that the anchoring of the crystals to the polymer occurs by way of Cu 2+ -amide interactions, which inherently requires the polymer to become partially incorporated between the layers of the framework (i.e. by coordination to the dinuclear paddlewheels). This is expected to disrupt the periodicity of the PCP in the crystallographic a-direction of the framework (while leaving the crystallinity of the bc plane unaffected), and the creation of uneven spacings between the square grid layers, some of which are sufficiently large for N 2 to be incorporated at low temperatures. This phenomenon is unique to Cu 2 (bdc) 2 (MeOH) 2 in a structuralized state, since such points of anchoring do not exist in the bulk form. Further, it demonstrates the importance of molecular scale interactions between the PCP crystals and the support in determining the adsorptive and dynamic behavior of the system as a whole. The impact of structuralization in the Cu 2 (bdc) 2 (MeOH) 2 system was further investigated by destroying the architecture by mechanical grinding of the monolith into a fne powder. Although the crystallinity of the sample was preserved following this process (see Fig. 4), N 2 adsorption data at 77 K revealed the complete loss of microporosity once in a ground powder form (see Fig. S10 †). This can be ascribed to the pulverization of the crystals as observed by SEM (Fig. S11 †), which leads to most of the crystalline fragments no longer being bound by the polyacrylamide polymer. Indeed, while the microporous region of the N 2 isotherm no longer shows a signifcant uptake, the profle exhibits a monotonic increase up to 190 cm 3 g 1 at a pressure of 1 bar, consistent with surface adsorption of N 2 to the polyacrylamide polymer surface. In addition, preparation of Cu(bdc) 2 (MeOH) 2 from a uniformly ground sample of the Cu(OH) 2 -polyacrylamide parent phase (prepared under the same reaction conditions as bulk Cu(bdc) 2 (MeOH) 2 ) yielded a sample of the same composition as the Cu(bdc) 2 (MeOH) 2 monolith. However, unlike the monolith form, the material displays little microporosity despite the presence of polyacrylamide in the overall composition (see Fig. S12 and S13 †). This further supports the observation that the immobilization of the Cu(bdc) 2 (MeOH) 2 crystals within the three-dimensional architecture provides the additional microporosity observed here. Cu 2 (bdc) 2 (bpy) monolith. The composition, structure, and framework flexibility of the replicated monolith was characterized using a combination of powder X-ray diffraction, SEM, and adsorption experiments. As shown in Fig. 6, powder X-ray diffraction data collected for an as-synthesized sample afforded reflections corresponding to the open pore form of the framework simulated from single-crystal data. In situ activation of the sample under a He flow at 150 C led to a structural change in the framework to the corresponding closed pore form, which is consistent with the removal of the methanol molecules within the pores. Resolvation of the material in methanol resulted in a return to the open form phase with retention of the threedimensional superstructure. Note that this solvation-desolvation process could be repeated several times without loss of the integrity of the monolith, demonstrating the successful preparation of a monolithic structure consisting of reversibly flexible building blocks. A methanol isotherm collected for an activated sample (see Fig. 7) exhibited a stepped isotherm with hysteresis in the desorption branch, which is typical for a gateopening type structural transition of the framework. Comparison of the methanol uptake over several cycles showed no degradation to the adsorption profle (Fig. S14 †), confrming the stability of the monolith with respect to flexing of the framework. Next, the effect of structuralization of the Cu 2 (bdc) 2 (bpy) compound in a monolith form was probed by comparing its methanol adsorption isotherm after mechanical grinding of the framework. Surprisingly, in contrast to the Cu 2 (bdc) 2 (MeOH) 2 monolith, little change was observed after the destruction of the structuralization with regard to both the gate-opening pressure and the quantity of methanol adsorbed (Fig. S15 †). Furthermore, comparison with a bulk powder of Cu 2 (bdc) 2 (bpy) also revealed an almost identical adsorption profle, revealing that both the structural flexibility and the adsorption properties of the monolith are a good match to those of a bulk sample of the same compound. This is a somewhat surprising result given that, based on the unusual properties observed for the Cu 2 (bdc) 2 (MeOH) 2 monolithic system, the immobilization of the Cu 2 (bdc) 2 (bpy) crystals in a monolith form might be expected to influence the adsorptive and dynamic properties of the system. In order to elucidate the origin of this result, IR and TGA data were collected to evaluate the composition of the Cu 2 (bdc) 2 (bpy) replicate. As is clear from the IR data presented in Fig. 8, the spectrum observed for the activated form of the Cu 2 (bdc) 2 (bpy) monolith shows a close match with that of a bulk sample of the same framework. However, in comparison with the parent and Cu 2 (bdc) 2 (MeOH) 2 monolith, the C-N stretch at approximately 1660 cm 1 originating from the amide moiety of the polyacrylamide polymer is greatly diminished, suggesting that the polymer component is excluded from the structure during the insertion of the bpy pillars. This was further confrmed by the TGA data shown in Fig. S16, † which allowed the polyacrylamide content to be calculated as 2.0 wt%, compared with 20.0 wt% and 15.0 wt% in the parent Cu(OH) 2 -polyacrylamide and Cu 2 (bdc) 2 (MeOH) 2 monoliths, respectively. The loss of polyacrylamide from the structure is also consistent with a decrease in the mechanical robustness of the Cu 2 (bdc) 2 (bpy) monolith, emphasizing its key role in providing the effect of structuralization of the monolithic structure following replication. The origin of the loss of polyacrylamide from the structure was probed via a number of control experiments. Immersion of the parent Cu(OH) 2 -polyacrylamide and Cu 2 (bdc) 2 (MeOH) 2 monoliths in methanol resulted in no change to the composition or the structuralization, which provided clear evidence of the stability of the monoliths (and its associated polymer content) under these conditions. Furthermore, immersion of the parent Cu(OH) 2 -polyacrylamide compound in a methanol solution of bpy resulted in no loss in the polyacrylamide component from the structure as evaluated by TGA data (Fig. S20 †). Thus, the polyacrylamide is only lost when the Cu 2 (bdc) 2 (MeOH) 2 undergoes pillaring by the bpy molecules during the second PCP-to-PCP replication step. While an exact mechanism for the loss of polyacrylamide is not yet available, a plausible sequence of events is as follows. In the conversion of the Cu(OH) 2 -polyacrylamide monolith to the Cu 2 (bdc) 2 -(MeOH) 2 replicate, the polyacrylamide directly binds to the Cu 2 (bdc) 2 (MeOH) 2 framework via amide sidechains as described above. This leads to the polymer chains, which are originally buried beneath a colloidal network of Cu(OH) 2 particles, to become exposed after replication. This is due in part to the plate-shaped crystals of Cu 2 (bdc) 2 (MeOH) 2 that are not expected to uniformly protect the polymer chains from access at the molecular scale. Then, upon exposure of the monolith to a solution containing bpy, the amide moieties are displaced from the Cu 2+ centers, leaving the chains unbound and susceptible to dissolution out of the monolith. This dissolution process may additionally be assisted by a partial hydrolysis of the polymer chains, which is known to occur in the presence of basic species. Note that analysis of the reactant solution by IR and 1 H NMR did not reveal the presence of free acrylamide monomers, suggesting a complex decomposition pathway for the PAAm component into a variety of products. As such, after the removal of the polyacrylamide component from the monolith, the limited intergrowth between the Cu 2 (bdc) 2 -(bpy) crystals leads to the structural and sorption properties of the monolith largely reflecting those of a bulk powder, despite the retention of the monolithic structure. ## Conclusions and future outlook The foregoing results have detailed the synthesis and properties of three-dimensional superstructures consisting of the flexible Cu 2 (bdc) 2 (MeOH) 2 and Cu 2 (bdc) 2 (bpy) frameworks via coordination replication from a structuralized macro-and mesoporous Cu(OH) 2 -polyacrylamide composite parent phase. The synthesis of these monolithic systems expands on the scope of the coordination replication technique to include flexible PCPs, but perhaps more importantly, provides monolithic systems that exhibit properties that differ from bulk powders as a result of structuralization. In the case of the Cu 2 (bdc) 2 (MeOH) 2 system, the anchoring of the two-dimensional framework by the polyacrylamide polymer leads to their immobilization within the superstructure, but also results in an amorphization of the interlayer direction of the framework structure. This provides the framework with an ability to adsorb N 2 , which is not observed in the absence of structuralization. For the Cu 2 (bdc) 2 (bpy) system, the framework flexibility is preserved after immobilization, leading to a flexible monolith system. In this case, the sorption and dynamic properties largely reflect the characteristics of the bulk form owing to the dissolution of the polymer phase during the PCP-to-PCP replication step. This emphasizes the importance of the polymer phase in maintaining the connectivity between crystals and in providing the system with the effects of structuralization. The results presented here further demonstrate the versatility of the coordination replication technique, and it is envisaged that a greater library of structuralized PCPs will emerge in the near future for specifc applications in areas such as molecular separations and heterogeneous catalysis. In addition, the new properties observed for the structuralized forms of the compounds suggest that new, rich phenomena could emerge as a result of detailed studies of this type. However, as revealed here, there is an urgent need for preparative routes to new parent materials that are optimized for coordination replication, and care is also needed in the selection of the target PCP system. Specifcally, a high degree of crystal intergrowth is desired in order to achieve cooperative effects stemming from material structuralization. While the polyacrylamide polymer serves as an adhesive between the crystals in this case, greater intergrowth between the PCP crystals themselves would preclude the need for the use of a composite system. For example, optimization of both the crystal size (i.e. smaller crystals) and morphology (i.e. block-shaped crystals) of the PCP phase is expected to facilitate a greater preservation of the original structure of the parent material with a greater degree of intergrowth. Such optimizations of the crystal parameters have already appeared in the case of bulk crystals via the coordination modulation technique, 18 and studies using this strategy for the fabrication of three-dimensionally structuralized systems composed of other functional PCP systems are already underway. mechanical strength measurements of the Cu 2 (bdc) 2 (MeOH) 2 monolith were not successful in this case. While large monoliths (cylindrical tablets with a diameter of 1 cm and a height of 0.5 cm) of the Cu 3 (btc) 2 framework were readily prepared within 30 min from the same starting precursor, 12 the conversion was found to be signifcantly slower in the case of Cu 2 (bdc) 2 (MeOH) 2 . The use of starting monoliths of a sufficient size resulted in samples with unreacted cores even after 14 days, likely due to preferential crystal growth at the exterior of the monolith resulting in macropore blockage, preventing diffusion of the organic linker throughout the solid. The signifcantly different behavior of the two systems highlights potential differences in both the molecular scale replication mechanism and the nature of the crystal growth, which are areas worthy of systematic investigation in order to optimize precursor design for specifc PCP systems. 14 Such effects are often observed in oriented samples or those with highly anisotropic crystal shapes, although this is not expected for the replicated phase studied here due to the random distribution of spatial orientations of the crystals within the monolith. 15 Y. Hishiyama and M. Nakamura, Carbon, 1995, 33, 1399. 16 Note that this slightly lower activation temperature than for bulk powder samples allows the polyacrylamide component to be stably maintained within the framework, while allowing full removal of the methanol within the pores and bound to the Cu 2+ ions of the dinuclear paddlewheel units. 17 The macroporosity is largely eliminated and the mesoporosity signifcantly diminished upon replication, which is due to the Cu 2 (bdc) 2 (MeOH) 2 crystals occupying a (up to 10 times) greater volume compared to the original Cu(OH) 2 component based on the density of Cu 2+ ions in the respective crystal structures. The plate-like morphology of the framework crystals may also provide a less contoured surface providing fewer cavities in the mesopore length scale. 18 (a) T. Tsuruoka, S. Furukawa, Y. Takashima, K. Yoshida,

chemsum

{"title": "Mesoscopic superstructures of flexible porous coordination polymers synthesized <i>via</i> coordination replication", "journal": "Royal Society of Chemistry (RSC)"}

graphite_oxide_improves_adhesion_and_water_resistance_of_canola_protein–graphite_oxide_hybrid_adhesi

## Abstract: Protein derived adhesives are extensively explored as a replacement for synthetic ones, but suffers from weak adhesion and water resistance. Graphite oxide (GO) has been extensively used in nanocomposites, but not in adhesives applications. The objectives of this study were to prepare functionally improved protein adhesive by exfoliating GO with different oxidation levels, and to determine the effect of GO on adhesion mechanism. GO were prepared by oxidizing graphite for 0.5, 2, and 4 h (GO-A, GO-B and GO-C, respectively). Increasing oxidation time decreased C/O ratio; while the relative proportion of C-OH, and C = O groups initially increased up to 2 h of oxidation, but reduced upon further oxidation. Canola protein-GO hybrid adhesive (CPA-GO) was prepared by exfoliating GO at a level of 1% (w/w). GO significantly increased (p < 0.05) adhesion; where GO-B addition showed the highest dry, and wet strength of 11.67 ± 1.00, and 4.85 ± 0.61 MPa, respectively. The improvements in adhesion was due to the improved exfoliation of GO, improved adhesive and cohesive interactions, increased hydrogen bonding, increased hydrophobic interactions and thermal stability of CPA-GO. GO, as we proposed for the first time is easier to process and cost-effective in preparing protein-based adhesives with significantly improved functionalities.Due to increasing concerns over environmental and human health implications of synthetic adhesives, researchers are looking for green materials/biobased adhesives using sustainable and renewable polymers [1][2][3][4] . Proteins are one of the most studied renewable polymers for adhesive applications 5 . Canola is the farm-gate crop in Canada while its meal after oil extraction finds limited value-added applications other than feed and fertilizer uses; thus research on canola protein gains the momentum as an alternative polymer source for adhesive preparation 5,6 . However, similar to other proteins, canola protein derived adhesives also suffered from weak water resistance and adhesion strength, which might limit their widespread applications 5,7 . Therefore, improving water resistance and adhesion strength of canola protein-derived adhesives is essential to succeed as a competitive alternative over synthetic ones. Our previous study found that exfoliating nanomaterials at lower addition levels could significantly increase the adhesion strength and water resistance of canola protein; especially, graphite oxide (GO) and nano crystalline cellulose (NCC) showed superior improvement over other nanomaterials 8 . The dry, wet and soaked adhesion strength of canola protein adhesives was increased from 6.38 ± 0.84 MPa, 1.98 ± 0.22 MPa, and 5.65 ± 0.46 MPa in the pH control samples to 10.37 ± 1.63 MPa, 3.56 ± 0.57 MPa, and 7.66 ± 1.37 MPa, respectively, for the 1% NCC addition (w/w, NCC/protein), and to 8.14 ± 0.45 MPa, 3.25 ± 0.36 MPa, and 7.76 ± 0.53 MPa for the 1% GO (w/w,GO/protein) addition 8 .Although NCC showed greater improvement than GO, NCC is more expensive than that of GO. Furthermore, GO shows excellent exfoliation properties in aqueous and organic solvents, as well as in different polymer matrixes due to hydrophilic nature of GO 9 . Previous studies on composite materials showed that the improvements in mechanical, thermal and electrical properties were directly related to the exfoliation properties of nanomaterials in polymer matrix 2,10 . Therefore, it is essential to use a nanomaterial with better exfoliation properties for adhesive preparation to improve mechanical strength of the adhesive 2 .Carbon based nanomaterials such as carbon nanotubes, graphene, graphite oxide and aerographite have been extensively studied recently in polymer and composite applications, mainly due to their excellent mechanical, thermal and conductive properties 11,12 . First isolated in 2004, graphene consists of two dimensional sheets of carbon molecules bonded via sp 2 -bonds 13 . Pristine graphene has unique material properties such as extremely high Young's modulus (∼1 TPa), fracture strength (∼130 GPa), thermal conductivity (∼5000 Wm −1 K −1 ) and specific surface area (2630 m 2 g −1 ) compared to other carbon based materials 14,15 . Graphite oxide (GO), an intermediary product in mass scale production of graphene, possess similar material properties to graphene 15 . GO represents advantages over graphene, mainly due to their simplicity of production through chemical methods, hydrophilic properties, and potential to convert into graphene or graphene oxide 15,16 either by chemical 17,18 or thermal 19 reduction methods before or after exfoliating in the polymer matrix. In addition, GO can form liquid crystals 20 and microscopic assembly of graphene once incorporated in polymer matrix 21 , which could help develop homogeneous polymer composite with improved mechanical properties 13,22 . The presence of oxygen containing functional groups imparts GO excellent hydrophilic properties, facilitating exfoliation in a polymer matrix 23 . Hydrophilic nature of GO is a vital property in preparing GO exfoliated adhesives using the solution intercalation method. GO has been extensively explored in developing advanced nano-composites in combination with different polymers such as poly (vinyl acetate) 24,25 , chitosan 26 , natural rubber 27 , poly (methyl methacrylate) and epoxy 10,23 . However, there is scanty information available in literature on GO based adhesives, except one study found in literature regarding applicability of graphene in adhesive preparation. Khan et al. (2013) reported that incorporation of 3% graphene (dissolved in tetrahydrofuran) into poly (vinyl acetate) (PVA) adhesives improved both tensile strength (from 0.3 MPa to 0.75 MPa) and shear strength (from 0.5 MPa to 2.2 MPa) at dry conditions 28 . It is well known that the functionality of GO largely depends on the level of its oxidation 9,23,29 ; therefore, there is a need to study the effect of different GO oxidation levels on adhesion strength and water resistance. We hypothesized that adding GO with different oxidation levels will change adhesion strength and water resistance of canola protein derived adhesives. The objectives of this research were to prepare GO with different oxidation levels under various oxidation time, to determine the effect of GO with different oxidation levels on adhesion properties, and to explore the mechanism of GO in adhesion improvement. In this study, GO with different oxidation levels were prepared by oxidizing graphite at different oxidation times. Prepared GO samples were exfoliated in canola protein to produce canola protein-graphite oxide (CPA-GO) hybrid wood adhesive. The effect of oxidation time on C/O ratio, surface functional groups, interlayer spacing, and thermal properties were characterized to identify their effect on GO dispersion in protein matrix, structural and thermal changes, adhesion strength and water resistance of CPA-GO. ## Results and Discussion The functionality of GO depends largely on the methods of preparation and conditions used in the process such as oxidation time and amount of oxidizer 23,29 . In composite research, tailoring conditions of GO preparation have proven to change material properties such as flexural strength and conductivity 29 . However, to best of our knowledge, there were no previous reports in the literature regarding the effect of GO on adhesives derived from biobased polymers/protein-based polymers. Adhesion strength of canola protein-GO hybrid adhesives. Adhesives failure can happen in two occasions, either adhesively at adhesive-wood interface or cohesively within bulk adhesive material 28 . Since amorphous polymer generally has a limited mechanical strength 28 , cohesive failure is more prominent in biobased adhesives. The potential of nanomaterials in improving the bulk adhesion strength of canola protein adhesive was studied. Effects of adding GO on adhesion strength are shown in Fig. 1. All GO samples used in this study significantly increased (p < 0.05) the adhesion strength and water resistance compared to the negative control and the pH control samples. GO prepared at 0.5, 2, and 4 h of oxidation showed a dry adhesion strength of 10.63 ± 0.81, 11.67 ± 1.00, and 11.22 ± 0.82 MPa, respectively. Increasing oxidation time reduced the C/O ratio of GO samples, but showed an increasing trend in dry adhesion strength. Similar trend was also observed in soaked strength, where the highest strength of 10.73 ± 0.45 MPa was observed in GO-B (2 h of oxidation) followed by GO-C and GO-A samples (10.22 ± 0.45, 9.82 ± 0.38 MPa respectively). The wet adhesion was significantly increased from 1.98 ± 0.22 MPa in the pH control sample to 4.85 ± 0.35, 4.85 ± 0.61 and 4.48 ± 0.28 MPa for GO-A, GO-B and GO-C samples respectively, but did not differ among different oxidation times. Protein contains both hydrophilic and hydrophobic residues which makes it an excellent amphiphilic biopolymer with well-known adhesiveness to various solid surfaces 30 . Liu et al. (2010) studied the interactions of GO with bovine serum albumin (BSA) and suggested that conjugated GO-protein complex can act as an adhesive matrix to interact with other solid materials 30 . Studies on PVA polymer composites showed improved interactions and mechanical strength after exfoliating graphene oxide at low concentrations 28,31 . Therefore, GO induced cohesive (protein-protein) and adhesive (protein-wood surface) interactions might be the main contributor to increased adhesion and water resistance observed in this study. Conversion of GO into more hydrophobic and stable reduced graphene oxide (rGO) might be another reason for the improved water resistance. Several authors reported thermal 32 or protein aided reduction 30 of GO into rGO in composite research, which improved the mechanical properties. Adhesive curing at elevated temperature, and the presence of canola protein might trigger the reduction of GO into rGO, thereby improve water resistance of the CPA-GO adhesive. In comparison, canola protein modified with sodium bisulfite showed dry, wet and soaked adhesion strength of 5.28 ± 0.47, 4.07 ± 0.16, and 5.43 ± 0.28 MPa, respectively 6 . In another study, modifying canola protein with 0.5% sodium dodecyl sulphate (SDS) had dry, wet and soaked adhesion of 6.00 ± 0.69, 3.52 ± 0.48, and 6.66 ± 0.07 MPa, respectively. Grafting poly(glycidyl methacrylate) in canola protein was reported to improve the dry, wet and soaked adhesion to 8.25 ± 0.12, 3.80 ± 0.15, and 7.10 ± 0.10 MPa, respectively. Canola protein adhesives prepared with GO as developed in this study substantially improved both adhesion strength and water resistance. Changes in elemental composition, functional groups of GO and their effect on adhesion. GO with variable elemental composition, C/O ratio and functional groups were previously developed via manipulating oxidation conditions 9,23,33 . In this study, we prepared GO with variable properties by changing oxidation time while maintaining other conditions constant. Oxidation conditions used in this study, elemental composition and C/O ratio of prepared GO samples are shown in Table 1. Native graphite mainly consists of carbon and oxygen at percentages of 97.65% and 2.35%, respectively, according to the XPS data (Supplementary information-S1). Graphite showed a C/O ratio of 41.55 while oxidizing graphite for 0.5, 2 and 4 h reduced C/O ratio to 2.06, 1.40 and 1.49, respectively. In addition, GO also contains small amount of sulfur (∼2%) and trace amounts of sodium, and manganese, as the residuals from GO processing. The presence of oxygen containing functional groups was confirmed by analyzing XPS high-resolution C1s spectra of graphite and GO samples. The original high-resolution C1s spectra and fitted peaks are shown in Fig. 2. XPS data processing for C1s spectra of graphite only showed a major peak centered at 284.5 eV which is attributed to sp 2 hybridized carbon, derived from C = C and C-C bonds with delocalized π electrons 29,33 . The other small peak at a binding energy of 285.3 eV resembles to sp 3 carbon hybridization 34 , which attributed to oxidation of graphite in the presence of atmospheric oxygen 35 . GO-A sample shows four new peaks at binding energies around 285.5-288.5 eV, representing oxygen functional groups in addition to the characteristic sp 2 peak at 284.5 eV. Shift of binding energies from 284.5 eV to 285.4 eV, 286.5 eV, 287.2 eV, and 288.5 eV are attributed to the occurrence of carbon sp 3 , C-OH, C-O-C, and C = O functional groups respectively. Previous studies reported similar binding energy shift in GO . Increasing oxidation time to 2 h (GO-B sample) further changed the composition of surface functional groups. Peak corresponding to the carbon sp 3 was disappeared while the relative proportion of C-OH and C = O peaks (286.5 eV and 288.3 eV respectively) increased. Furthermore, C-O-C peak appeared at the binding energy of 287.1 eV. Wang et al. (2012) also reported an increased proportion of C = O and C-OH groups at higher oxidation conditions in graphite oxide 29 . Further oxidation of graphite up to 4 h in GO-C increased the relative proportion of carbon sp 2 , C-O-C, and C = O groups, at the expense of C-OH groups; interestingly, the carbon sp 3 peak re-appeared at 285.4 eV binding energy. Degradation of oxygen functional groups in prolonged oxidation might be the reason for sp 3 hybridization observed in GO-C sample 33 . FTIR spectra of GO samples prepared under different oxidation times are shown in Fig. 3. FTIR peaks were assigned to respective functional groups according to the previous data reported in the literature. In graphite, the peak appearing at 1586 cm −1 generally represents the stretching vibration of C = C bond (vC = C) 35,39,40 . However; after oxidation, the C = C stretching vibrations shifted to 1619 cm −1 , 1623 cm −1 , and 1621 cm −1 wavelengths for GO-A, GO-B and GO-C respectively. Chen et al. (2010), and Stankovich et al. (2006) also reported similar peak shifts in the range of 1618 cm −1 -1626 cm −1 probably due to the oxygen functional groups present in GO 41,42 . The absorption peaks of GO samples at 3424 cm −1 -3436 cm −1 are attributed to the stretching vibration of -OH groups (vO-H) either from -OH groups of absorbed water or -OH groups formed during the oxidation 35,41,43 . Following oxidation, the presence of new peaks at 1729 cm −1 , 1731 cm −1 , and 1725 cm −1 wavelengths respectively in GO-A, GO-B, and GO-C samples was observed; probably due to the formation of oxygen containing functional groups, causing the C = O stretching vibrations (vC = O) 29,41,42 . The intensity of vC = O in GO samples was increasing at increasing oxidation level. Wang et al. (2012) also reported similar trend at increasing oxidation levels 29 . Higher degree of oxidation and oxidation induced cracks in GO edges were reported as the main reasons for increased intensity of vC = O 29,44,45 . C-OH bending vibration (δC-OH) peaks were observed in both GO-A and GO-B samples at 1411 cm −1 and 1423 cm −1 respectively, however the intensity was reduced in GO-C. Vibrations from either alcohols or carboxylic groups of GO were reported as the main contributors to δC-OH 39,40 . The peaks appeared at 1220 cm −1 -1230 cm −1 range were usually assigned to C-O stretching vibrations (vC-O 35, , which attributed to carboxylic acid groups 29 , hydroxyl groups 39,41 , or epoxy groups 40,46 present in GO. The peaks appeared at 1730 cm −1 -1731 cm −1 range were probably attributed to the ester groups that formed during graphite oxidation 47 . The formation of various oxygen containing functional groups in GO might be responsible for the improved adhesion strength. For example, −OH groups in GO might increase −H bonding between adhesive matrix and wood surface; the epoxy groups (C-O-C) in GO can either homopolymerize with another epoxy group in GO, or react with functional groups such as −OH, −COOH on the wood surface, and −NH 2 , −SH in canola protein 48 , thus improving both adhesive and cohesive strength. ## Effect of different GO samples on protein structural changes. Effect of different GO samples on secondary structure of canola protein was studied by creating second derivative of FTIR spectra followed by peak fitting of Amide I peak 49,50 . GO induced protein secondary structural changes are shown in Fig. 4. Exfoliating GO in canola protein has increased the relative proportions of unordered structures (1639-1642 cm −1 wavelength) and turn structures (at wavelength range of 1694-1697 cm −1 ) at the expense of β-sheets in the wavelengths of 1625 cm −1 , 1636 cm −1 and 1673-1675 cm −1 49, 50 . In comparison, GO-B and GO-C samples showed the highest relative proportions of unordered structures and turn structures, compared to the pH control and GO-A samples (Supplementary information-S2). The results observed in protein structural changes were compliment to the changes in adhesion strength of CPA-GO prepared with different GO samples. Increase in unordered structures will exposes more hydrophobic functional groups buried inside protein molecules which increase the hydrophobic interactions with wood surface 51 , thereby increase the water resistance and adhesion strength. Changes in GO crystallinity and their effect on GO dispersion in protein matrix. The effect of oxidation time on glancing angle (2θ) and interlayer spacing (d) of GO samples are shown in Fig. 5. X-ray diffraction of graphite showed one major crystalline peak at a glancing angle of 26.28° with d spacing of 0.338 nm. Shao et al. (2012) also reported a similar peak for graphite at a glancing angle of 26.54° and d spacing of 0.334 nm 23 . After oxidation, the graphite crystalline peak was disappeared in GO-A (0.5 h) but two new peaks appeared at different glancing angles: the first major peak was appeared at glancing angle of 11.28° with d spacing of 0.785 nm while another minor peak was observed at glancing angle of 42.17° with d spacing of 0.214 nm. Shao et al. (2012) also reported the disappearance of the characteristic graphitic peak after oxidation and the formation of a new peak at a glancing angle of 11.3° with increased interlayer spacing of 0.80 nm 23 . Increasing graphite oxidation time from 0.5 h to 2 h significantly changed the crystallinity and d spacing of GO-B sample. Glancing angle of the characteristic GO peak has shifted from 11.28° to 9.40° while d spacing increased from 0.785 nm to 0.939 nm (for GO-A and GO-B respectively). Similar to GO-A, GO-B sample showed another peak at a glancing angle of 42.20° (d = 0.214 nm), and a new crystalline peak at 19.91° (d = 0.495 nm). Further increasing oxidation time to 4 h slightly shifted the glancing angle towards 9.94° while decreased d spacing to 0.889 nm. The reduction in interlayer spacing has been previously reported due to the decomposition of oxygen containing functional groups in GO samples at prolonged oxidation 33,52 . In GO-C, another two peaks were visible at glancing angles of 42.29°, and 17.89° with d spacing of 0.214 nm and 0.495 nm respectively. In addition, the new peak at a glancing angle of 25.33° (d = 0.351 nm) in GO-C showed similarity to the characteristic graphite peak appeared in un-oxidized graphite. The re-appearance of graphite like crystalline peak at higher oxidation level indicate the decomposition of oxygen containing functional groups, re-forming carbon sp 2 bonds and reduction in crystallinity of GO-C samples 33,52 . Proper exfoliation of GO in polymer matrix is one of the major factors affecting the improvement of adhesion strength and water resistance. Aggregation of nanomaterial upon mixing with protein will not improve the adhesion strength 2,53 ; therefore it is important to produce GO with appropriate exfoliation properties. All three GO samples prepared in this study exhibit improved exfoliation in canola protein matrix. X-ray diffraction patterns of GO samples and their dispersion in canola protein are shown in Fig. 6. Two common crystalline peaks were appeared in all three GO samples with diffraction angles (2θ value) around ∼9-11° and ∼42° and one additional crystalline peak was found at ∼25° diffraction angle for GO-C. The disappearance of crystalline peaks after exfoliation of GO in canola protein clearly indicated the uniform exfoliation of GO within protein matrix. As shown in TEM images of exfoliated GO samples (Fig. 7), the appearance of single GO sheets in both CPA GO-A and CPA GO-B adhesive samples further supported the uniform exfoliation of GO in canola protein matrix. However, a slight aggregation of GO was visible in CPA GO-C. Addition of hydrophilic functional groups during graphite oxidation is the major reason for increased interlayer spacing of GO 33 . It was reported that increased interlayer space reduces binding energies of GO, which would facilitate the exfoliation of GO layers in the matrix 54 . Therefore, the uniform exfoliation of GO observed in this study, in particular for GO-B might be due to reduced binding energy resultant from increased interlayer spacing. Ultimately, proper exfoliation of GO will help in improving both adhesion strength and water resistance of the CPA-GO adhesive. Change in thermal properties of graphite oxide and their effect on thermal stability of prepared adhesive. Effect of graphite oxidation time on GO thermal transitions is shown in Fig. 8. An exothermic transition was observed in all GO samples, but with different enthalpy requirement and temperature range. In GO-A (0.5 h) exothermic transition was observed at extrapolated onset and peak temperatures of 159.7 °C 190.0 °C respectively with 1.57 KJ/g ΔH. Increasing oxidation time to 2 h (GO-B) has changed the thermal transition to 145.6 °C, 164.9 °C and 1.16 KJ/g for extrapolated onset, peak temperature and ΔH respectively. Increasing oxidation time to 4 h (GO-C) shifted extrapolated onset and peak temperatures to 146.0 °C and 166.7 C° respectively where ΔH changed to 1.10 KJ/g. The reduction in ΔH and transition temperatures is a result of increased amount of oxygen containing functional groups. Schniepp et al. (2006) also reported similar changes in thermal transitions around ∼200 °C in graphite oxide and attributed them to decomposition of oxygen containing functional groups 36 . They have further analyzed the outlet gas generated from DSC, and showed that major products as CO 2 and H 2 O that were generated during decomposition of oxygen containing functional groups 36 . Effect of different GO samples on thermal transitions of CPA-GO are shown in Table 2. Adding GO into canola protein increased both onset and peak temperatures, as well as the specific heat in transitions. The increased thermal stability is an essential property for adhesive application, as it required to process under higher temperature for adhesive curing 28 . Adding nanomaterials, especially graphene oxide, have been proven to increase thermal stability of protein in previous studies mainly due to improved protein-protein/protein-GO interactions, and 2007) also reported an increased thermal stability and denaturation temperatures of soybean peroxidase enzyme conjugated with graphene oxide nanosheets 55 . Addition of GO into canola protein increased the thermal stability of all CPA-GO samples compared to control samples, which can be related to the increased protein-protein/protein-GO interactions. CPA GO-A showed slightly higher onset and peak temperatures than that of CPA GO-B and CPA GO-C which can be a result of GO induced protein structural changes. Increased unordered structures were observed after adding GO-B and GO-C, at the expense of β-sheets and α-helix which can potentially reduce the thermal stability compared to GO-A. ## Conclusions GO samples with various C/O ratio and surface functional groups were prepared at different oxidation time. Oxidation of graphite for 0.5, 2 and 4 h reduced the C/O ratio of graphite from 41.55 to 2.06, 1.40, and 1.49, respectively. The relative proportion of C-OH and C = O groups as well as interlayer spacing of GO were increased at increasing oxidation time from 0.5 h to 2 h whereas both C-OH content and interlayer spacing were reduced at 4 h of oxidation. GO prepared with different oxidation times improved both adhesion strength and water resistance in all three GO samples; the dry, wet and soaked strength was increased from 6.38 ± 0.84 MPa, 1.98 ± 0.22 MPa, 5.65 ± 0.46 MPa in the pH control sample to 11.67 ± 1.00 MPa, 4.85 ± 0.61 MPa, and 10.73 ± 0.45 MPa, respectively for GO-B exfoliated adhesive. The improved adhesive and water resistance in GO added canola adhesive was due to increased interlayer spacing, improved exfoliation properties, and increased adhesive and cohesive interactions (protein-protein, protein-GO and adhesive-wood surface), hydrophobic interactions and thermal stability. Graphite oxide, instead of graphene, as we proposed for the first time in the study, is easier to process and more cost-effective in preparing protein-based wood adhesives with significantly improved functionalities. ## Materials and Chemicals. Canola meal was provided by Richardson Oilseed Ltd. (Lethbridge, AB, Canada). All chemicals were purchased from Fisher Scientific (Ottawa, ON, Canada) unless otherwise noted. Graphite and cellulose were purchase from Sigma-Aldrich (Sigma Chemical Co, St. Louise, MO, USA). Birch wood veneer with thickness of 0.7 mM was purchased from Windsor Plywood Co (Edmonton, AB, Canada). ## Canola protein extraction. Proteins were extracted from defatted canola meal as described by Manamperi et al. (2010) with slight modifications 56 . Meal was ground to a fine powder using a Hosokawa milling and classifying system (Hosokawa Micron Powder Systems, Summit, NJ, USA) and then passed through a 100-mesh size sieve. Ground canola meal was mixed with mili-Q water in 1:10 (w/v) ratio; pH was adjusted to 12.0 by adding 3 M NaOH and stirred for 30 m (300 RPM, room temperature). The resulting dispersion was centrifuged for 15 m (10000 g, 4 °C). The supernatant was collected, pH was readjusted to 4.0 by adding 3 M HCl, stirred for another 30 m, and centrifuged at the same condition above to collect protein precipitate. The precipitate was washed with deionized water, freeze-dried, and stored at −20 °C for further use. ## Graphite oxide preparation. Graphite oxide nanoparticles (GO) were prepared as described by Hummers and Offeman method 57 with modification for oxidation time to produce GO with different oxidation levels. In brief, 5 g of graphite and 5 g of NaNO 3 were mixed in a glass beaker and 120 mL of concentrated H 2 SO 4 was slowly added while stirring in an ice bath at 200 RPM for 0.5 h, 2 h, and 4 h to prepare GO-A, GO-B and GO-C samples respectively. Then, 15 g of KMnO 4 was slowly added to the reaction mixture while maintaining the temperature at 35 ± 3 °C with stirring for 1 h. At the end of the reaction, 92 mL of deionized water was added and stirred for 15 m. Unreacted KMnO 4 and other leftover chemicals were neutralized by adding 80 mL of hot (60 °C) deionized water containing 3% H 2 O 2 . After cooling to room temperature, samples were centrifuged (10000 g, 15 m, 4 °C) and washed with deionized water to remove any leftover chemicals. Collected GO samples were sonicated for 5 m (at 50% power output); freeze dried, further dried in a vacuum desiccator with P 2 O 5 , and stored in air tight containers at -20 °C for further use. ## Preparation of canola protein-graphite oxide hybrid wood adhesive (CPA-GO). GO with different C/O ratios was exfoliated in canola protein matrix according to our previously reported method. 1% (w/w, GO/protein) GO addition level was selected based on the optimum conditions developed in our previous method 8 . In brief, 3 g of canola protein was mixed with 20 mL of deionized water (15% w/v solution) and stirred for 6 h (300 rpm) at room temperature to disperse canola proteins; and then the pH was adjusted to 5.0 using 1 M HCl solution. GO samples (GO-A, GO-B and GO-C) were separately dispersed in 10 mL of deionized water (equivalent to a final GO/protein ratio of 1%, w/w, GO/protein) by stirring (300 rpm) 5 h at room temperature and another 1 h at 45 ± 3 °C, sonicated for 3 m by providing intermittent pulse dispersion of 5 s (at 3 s intervals and 60% amplitude) using medium size tapered tip attached to a high intensity ultrasonic dismembrator (Model 500, Thermo Fisher Scientific INC, Pittsburg, PA, USA), and then homogenized for 2 m (2000 rpm) using ULTRA TURRAX high shear homogenizer (Model T25 D S1, IKA ® Works, Wilmington, NC, USA). The prepared GO dispersions were slowly added to the protein dispersions dropwise while stirring for 15 m (300 rpm) to have a final protein concentration of 10% (w/v) in the adhesive mixture. The resulting adhesive mixtures were sonicated and homogenized as above and the pH of the adhesive was adjusted to 12.0 by adding 6 M NaOH solution. Negative control was prepared by dispersing canola protein (10% w/v) in deionized water and use as is while pH control was prepared by adjusting the pH of canola protein dispersions (10% w/v) to 12.0 similar to GO dispersed samples, without adding GO. ## Adhesion strength measurement. Hardwood veneer samples (Birch, 1.2 mm thickness) were cut into a dimension of 20 mm × 120 mm (width and length) using a cutting device (Adhesive Evaluation Systems, Corvallis, OR, USA). Veneer samples were conditioned for seven days at 23 °C and 50% humidity in a controlled environment chamber (ETS 5518, Glenside, PA, USA) according to the specifications of ASTM D2339-98 (2011) standard method 58 . CPA-GO hybrid adhesives were spread at an amount of 40 uL/veneer strand in a contact area of 20 mm × 5 mm using a micropipette. Veneer samples were air dried for 5 m and hot pressed for 10 m (at 120 °C and 3.5 MPa) using Carver manual hot press (Model 3851-0, Carver Inc., In, USA). Dry adhesion strength (DAS) was measured according to the ASTM standard method D2339-98 (2011) by measuring tensile loading required to pull bonded veneer using Instron machine (Model 5565, Instron, MA, USA) equipped with 5 kN load cell. Tensile strength data was collected using Bluhill 3.0 software (Instron, MA, USA). Wet adhesion strength (WAS) and soaked adhesion strength (SAS) was measured according to the ASTM standard method D1151-00 (2013) 59 using instron tensile loading. WAS values were measured after submerging bonded veneer samples for 48 h in water (23 °C) where SAS was measured after reconditioning submerged veneer samples for seven days at 23 °C and 50% relative humidity in a controlled environment chamber (ETS 5518, Glenside, PA, USA). Minimum of four bonded veneer samples per formulation were used in testing strength (DAS, WAS, SAS). All samples were clamped to Instron with a 35 mm gauge length and tested at 10 mm/m cross head speed. X-ray Photoelectron spectroscopy (XPS). GO samples were characterized using X-ray photoelectron spectroscopy (XPS) for their elemental composition, carbon/oxygen (C/O) ratio and changes in the functional Binding energy of neutral carbon C1s spectra was adjusted to 284.5 eV as a reference. Oxidation time dependent changes in surface functional groups were characterized by curve fitting of high-resolution C1s spectra assuming a Shirley background and 70%/30% Gaussian/Lorentzian distribution shape. Four peaks were fitted for all other GO samples while five peaks were used in GO-A sample with a lower oxidation time. ## X-ray diffraction (XRD). X-ray diffraction (XRD) of GO and CPA-GO samples were performed using Rigaku Ultima IV powder diffractometer (Rigaku Co. Japan). Cu-Kα radiation (0.154 nm) was used to collect angle data (2ϴ) from 5 to 50 degrees. Interlayer spacing of graphite oxide was calculated using Bragg's equation 60 ; sin θ = nλ/2d where, λ, d and θ represent wavelength of the radiation, spacing between diffraction lattice (interlayer space), and glancing angle (measured diffraction angle) respectively 53,61 . XRD data was analyzed using Origin 2016 software (OriginLab Corporation, MA, USA) to identify effect of oxidation time on exfoliation of GO. ## Differential scanning calorimetry (DSC). Thermal transitions of GO and CPA-GO adhesives were studied using differential scanning calorimeter (Perkin-Elmer, Norwalk, CT, USA). DSC instrument was calibrated for temperature and heat flow using a pure indium reference sample. Sample moisture was first removed by freeze-drying followed by drying with P 2 O 5 for two weeks in a hermetically sealed desiccator. GO and hybrid adhesive samples were accurately weighed into T-Zero hermetic aluminum pans (∼6 mg each), mixed with 60 µL of 0.01 M phosphate buffer, and hermetically sealed with lids. Heat flow differential of samples were recorded against the empty reference pan under continuous nitrogen purging. All samples were equilibrated at 0 °C for 10 m and thermodynamic data was collected while heating from 0 to 250 °C at a ramping rate of 10 °C m −1 . Data was analyzed using Universal Analysis 2000 software for thermal transition changes in adhesives and GO samples (Perkin-Elmer, Norwalk, CT, USA). ## Fourier transform infrared spectroscopy (FTIR). Effect of oxidation time on GO functional groups and GO induced protein secondary structural changes in adhesive samples were characterized using Nicolet 8700 Fourier transform infrared spectrometer (Thermo Eletron Co. WI, USA). Sample moisture was removed prior to FTIR analysis by freeze-drying and further drying with P 2 O 5 in a hermetic desiccator for two weeks. Samples were mixed with potassium bromide (KBr), milled into a powdered pellet prior to FTIR analysis. IR spectra were collected in 400-4000 cm −1 range using 128 scans at a resolution of 4 cm −1 . Collected data was analyzed using Origin 2016 software (OriginLab Corporation, MA, USA) to identify changes in functional groups. Second derivative spectra were generated using Savitzky-Golay smooth function (7 points window) and used for curve fitting to identify GO induced protein structural changes. ## Transmission Electron Microscopy (TEM). Effect of GO samples on exfoliation in canola protein matrix were characterized using transmission electron microscopy (TEM). Images were collected using Philips/FPI transmission electron microscope (Model Morgagni, FEI Co, OR, USA) coupled with Getan digital camera (Getan Inc, CA, USA). Adhesive samples were diluted to 100-fold with ethanol, and a single drop was casted onto 200 mesh holey copper grid covered with carbon film. After 30 seconds of air-drying, the remaining liquid was removed and copper grid was used for collecting TEM images. Statistical Analysis. Adhesion strength data (DAS, WAS, and SAS) was analyzed using analysis of variance (ANOVA) followed by Duncan's Multiple Range (DMR) test to identify the effects of graphite oxidation time on adhesion strength. Collected data was processed using Statistical Analysis System Software (SAS version 9.4, SAS Institute, Cary, NC). Effects of different GO samples on adhesion strength were evaluated at the 95% confidence level.

chemsum

{"title": "Graphite Oxide Improves Adhesion and Water Resistance of Canola Protein\u2013Graphite Oxide Hybrid Adhesive", "journal": "Scientific Reports - Nature"}

revealing_structure-property_relationships_in_polybenzenoid_hydrocarbons_with_interpretable_machine-

## Abstract: The structure-property relationships of polybenzenoid hydrocarbons (PBHs) were investigated with interpretable machine learning, for which two new tools were developed and applied. First, a novel textual molecular representation, based on the annulation sequence of PBHs was defined and developed. This representation can be used either in its textual form or as a basis for a curated feature-vector; both forms show improved interpretability over the standard SMILES representation, and the former also has increased predictive accuracy. Second, the recently-developed model, CUSTODI, was applied for the first time as an interpretable model and identified important structural features that impact various electronic molecular properties. The resulting insights not only validate several well-known "rules of thumb" of organic chemistry but also reveal new behaviors and influential structural motifs, thus providing guiding principles for rational design and fine-tuning of PBHs. ## Introduction In recent years, machine learning (ML) has been increasingly used in chemistry to address a wide variety of challenges, ranging from drug design 1,2 to automatic synthesis, to accelerating traditional computations. Whereas the success of earlier models was measured by efficiency and accuracy in prediction, current models are often aimed towards better "interpretability" -i.e., an ability to provide guiding principles and insight into domain relationships. 9 In other words, scientists wish to understand "what the model has learned", which may serve to validate existing chemical laws and intuitions, 10,11 or, hopefully, even lead to the discovery of new physical and chemical insights. 9,12,13 Recent reports have demonstrated that ML can "rediscover" concepts and conventional wisdom in chemistry and physics. Examples include: the effect of specific functional groups on solubility and HOMO level, 11 the hard and soft acids and bases (HSAB) principle for stability of inorganic complexes, and the identification of important normal modes for molecular dissociation 14 . Alongside these, there is discussion of how ML can lead to entirely new discoveries. 12 It should be mentioned, however, that in all of these cases, domain expertise is required, either to engineer the features given to the model or to place the "understanding" of the model in a domain-appropriate context. In this work, we apply interpretable machine learning to the question of structureproperty relationships in the family of compounds known as cata-condensed polybenzenoid hydrocarbons (PBHs; sometimes also referred to as catafusenes or as polycyclic aromatic hydrocarbons, PAHs). These molecules are impactful in many areas, in particular in human and environmental health 15,16 and in organic electronics. Due to their importance, these compounds have been extensively studied for many decades, both computationally and experimentally. They continue to garner attention for their potential to be used as organic semiconductors 20 and because they are precursors to nanographene sheets. 21 Understanding the properties of PBHs is crucial to both understanding their reactivity and designing new functional materials and new pathways for safe disposal of harmful ones. Thus, obtaining a deeper understanding of structure-property relationships governing the behavior of PBHs is of interest both from the conceptual aspect and from the practical one. Beyond these reasons to study PBHs, there is also a fundamental issue. To paraphrase Randic: 22 in order to understand the behavior of polycyclic aromatic systems (PASs) in a general way, one must first understand the systems comprising the archetypal aromatic unit -benzene. We envision that the current study is the necessary foundation for future investigations of broader swaths of the PAS chemical space. The prevalence of PASs in both natural and man-made materials entails that factors affecting their molecular properties are important to consider in designing new functional molecules and materials. We approach the subject of interpretable ML in the context of aromatic molecules from two directions: a) the introduction of a new type of molecular representation specifically suited to this kind of molecules and b) the application of a novel interpretable ML method, named CUSTODI, 23 which does not require any human-aided feature selection. We show that our new representation is suitable for extracting chemically meaningful insight and has similar performance to state-of-the-art techniques, but with shorter training times and fewer data required for training. The combination of these two new tools allows us both to validate structure-property relationships previously revealed using electronic-structure investigations and also to uncover additional relationships. These can then inform the rational design and/or fine-tuning of properties. ## The LALAS Representation Our group has demonstrated in a series of reports over the past few years that catacondensed PASs can be broken down into their smaller components (monocyclic, bicyclic, and tricyclic), and the magnetic properties of the larger molecules can be predicted by summing the contributions of these smaller subunits using an additivity scheme. For the particular case of the PBHs, molecular properties can be predicted by the type and order of the tricyclic components themselves, where the two tricyclic subunits differ only in their annulation: linear or angular, i.e., anthracene or phenanthrene, respectively. This conclusion allows for a reduction of the molecular structure to the sequence of tricyclic subunits (i.e., the annulation sequence). We have formulated this sequence as a textual representation of the molecule (Figure 1a), containing only the characters "L", "A", "(" and ")" (parentheses are used to denote branching points, where applicable; see Figure 1b for a selection of PBHs and their respective annulation sequences). The resulting names are strings of varying lengths comprising the letters "L" and "A", which we have accordingly named "LALA Strings" or "LALAS" (the terms "LALAS representations", "LALAS", and "annulation sequences" are interchangeable). The annulation sequence, or LALAS, has been clearly demonstrated to be linked to molecular properties: molecules sharing the same annulation sequence are equiaromatic (i.e., the same aromatic behavior) in both the ground state and the lowest excited triplet state. 27 In addition, we have shown that the annulation sequences themselves demonstrate a clear connection to and enable prediction of numerous molecular properties, including relative stability, aromatic character, singlet-triplet energy gaps, and location of spin density in the triplet excited state. 27 The generation of a LALAS for a given molecule proceeds according to the following protocol (similar to IUPAC rules for naming branched alkanes), which we have automated in a modified version of Predi-XY. 28 The modified code for generating the LALAS is freely accessible online. a. For unbranched molecules, each tricyclic subcomponent is denoted as a letter "L" or "A", depending on the type of annulation. The choice of "left-to-right" or "rightto-left" is random, i.e., each molecule has (at least) two valid LALAS. E.g., the molecule LLA (Figure 1B) can also be read as ALL. b. For branched molecules, we search for the longest possible path through the molecule, and denote this the "main branch". E.g., the main branch of molecule LLA()is a chain of 5 rings. c. If there are branching points, they are denoted with "()" (e.g., LLA() in Figure 1B). Note that branching points will always follow an "A", as they are by necessity connected to the middle ring of an angular annulation. Note, also, that the notation "()" implies a branch containing a single ring. d. Branches longer than a single ring will have their own sequence, which will be detailed within the parentheses (e.g., LAA(L)LL in Figure 1B). e. If there are two different paths of similar length, the one with more branching points is chosen as the main branch. We emphasize that, in contrast to SMILES or SELFIES, which describe molecules on an atom-connectivity basis, LALAS describe molecules using ring-based subunits. As such, they reduce the dimensionality of the molecular representation, while retaining important structural information. This trait could allow for significant improvement in efficiency of training ML models -in reducing both the training time and the required training set size. We also note that several graph-theoretical based notations for PBHs have been previously proposed, most notably by Gutman, 29 Balaban, and Cyvin. 33 To the best of our knowledge, these have been used mostly for enumeration of isomers of various types of PBHs. The 3-digit code developed by Balaban in the 1960s, which is the most similar in approach to our own formulation, has been also used to identify correlations to molecular properties (e.g., ionization potential, IP, and electron affinity, EA). 34,35 In this work, LALAS representations were generated using a modified version of the Predi-XY code developed in our group 28 and were used in two ways: a) tokenization directly from the string format (LALAS) and b) as a basis for generating a LALA-based feature vector (LFV) for each molecule (vide infra). ## Data Sources With the advent of more efficient computational techniques, data-driven investigations have become increasingly common; however, it has been difficult to apply such methods to PBHs, as there is a paucity of suitable data. Recently, our group reported on the COMPAS Project: the construction of a novel COMputational database of Polycyclic Aromatic Systems. 36 The first instalment of the database, denoted COMPAS-1D, contains data on ~8,600 cata-condensed PBHs, including their optimized structures and a selection of electronic properties (calculated with DFT at the B3LYP-D3BJ/def2-svp level), as well as their respective SMILES representations and LALAS representations. For the current study, we removed benzene and naphthalene from the dataset, as they are too short to have a LALAS. Both LALAS and SMILES representations were tokenized using two methods: one-hot 37 and CUSTODI. 23 The properties we extracted from the COMPAS-1D database for this study were: a) HOMO energy; b) LUMO energy; c) HOMO-LUMO gap; d) adiabatic ionization potential (AIP); e) adiabatic electron affinity (AEA); f) relative single-point energy. LALA Feature-Vector (LFV) In addition to the LALAS, we generated for each molecule a feature-vector based on curated structural features derived from the LALAS, denoted LFV. This set of chemically intuitive features (detailed in Table 1) was inspired by our collective experience studying PBHs and by structure-property relationships previously found in smaller datasets. 24,27 The purpose of using the LFV as input was threefold: (1) to validate the intuition we developed previously, (2) to check the predictive power of these descriptors, and (3) to compare the conclusions derived from this set of PBH-specific features to those derived from more general chemical representation. The CUSTODI Framework CUSTODI is a recently developed tokenization technique for text-based molecular representations. A full description of the method is beyond the scope of this report. In brief, the approach of CUSTODI is to find, using linear regression, the best fitting tokenization dictionary for a given target property. The resulting dictionary can be used for tokenization (CUSTODI representation) or for prediction (CUSTODI model), as shown in Figure 2. Both the CUSTODI representation and the CUSTODI model were used in this work. For further details on the method, the reader is referred to reference 23 . To perform a methodical comparison, four supervised learning models were used, ranging from standard to state-of-the-art (for further details on each model, please refer to Section S4 in the Supporting Information). Kernel ridge regression (KRR) and random forest (RF) were used in conjunction with CUSTODI tokenization and LALA features as input (denoted as CUSTODI[LALAS]); a recurrent neural network (RNN) was trained on onehot tokenization of the LALAS and SMILES (denoted One-Hot[SMILES] and One-Hot[LALAS], respectively); and a state-of-the-art 38,39 graph convolution (GC) model was used with the MolConv input. 40 The KRR and RF models were implemented using scikitlearn, 41 the RNN model by using tensorflow, 42 and the GC model by using DeepChem 43 with an identical architecture as the MoleculeNet benchmark. 38 The data was split into training and testing sets and hyperparameter optimization was performed using Bayesian optimization algorithm (Gaussian process) as implemented in the scikit-optimize python package. 44 Model selection was done using 5-fold cross validation. Exact details on the hyperparameters of each model are in Section S4.2 in the Supporting Information. The best model was retrained on the whole training set and used to estimate the model's performance. All properties were normalized using z-score normalization (0 mean and 1 standard deviation) and all tokenized strings were padded before insertion into the models. ## Interpretation of CUSTODI The interpretation of CUSTODI is relatively straightforward: each tokenization value corresponds to a substring (e.g., atom or functional group), and these values are used to make the model's prediction (Eq. 1). In this work, each tokenization value corresponds to a tricyclic substructure within the PBH. Where 𝑠 is the string representation of molecule 𝑖 in the database, 𝑐 is the 𝑖th character in 𝑠 , 𝑛 … is the number of occurrences of the substring 𝑐 … 𝑐 in 𝑠 , 𝑥 … is the substring's tokenization value and 𝑝 is the target property. From Eq. 1, the tokenization value is proportional to the significance of the represented substructure for a given property. The tokenization value 𝑥 is not independent of the number of occurrences 𝑛, and there is actually an inverse proportion between them. To account for this proportion, the importance of each substring is given by So that the sum of all the importance terms is 1. We emphasize that the analysis made here can be repeated for many chemical compounds and can produce similar intuition on the effects of various functional groups on properties. Unlike previous reports (a few are detailed in the Introduction, vide supra), CUSTODI does not rely on hand-crafted features. By iterating over all possible substrings, CUSTODI in essence performs its own data-driven feature-engineering. The main advantages are that this does not require any chemical intuition and tests all substructures in the dataset simultaneously. As a result, this reduces possible sources of bias and allows for identification of features that might not be obvious to experts. The disadvantages are that CUSTODI cannot search for varying-length substrings and will likely not identify features that involve non-adjacent structural components. ## Results and Discussion The two main aspects of the work are presented and discussed in the following sections: a) the performance of LALAS as a molecular representation and b) the use of LALAS in conjunction with interpretable ML models (CUSTODI and RF) to gain new chemical insights into PBHs. ## The Performance of LALAS As mentioned above, LALAS are specifically tailored to describe PBH compounds. To test the added value of this dedicated representation versus commonly used general-purpose representations, the performance of several models trained on LALAS was compared to the same models trained on other types of input (see Methods for further details on the selected models for comparison). The models employed are detailed in the Methods section and in Figure 3, which provides an illustration of all input+model combinations. The results obtained with a training set containing 7,674 molecules (90%) are illustrated in Figure 4 (the full fit results on the database are in Section S6 in the Supporting Information). In all other cases, the CUSTODI[LALAS] representation performed markedly better than the CUSTODI[SMILES] representation. What is more, when using CUSTODI[LALAS] as input, the RF model performed similarly to the best-performing RNN and GC models, which are considered much more sophisticated. A possible explanation is that the simpler syntax of LALAS better suits the linear approximation in CUSTODI, thus allowing for better tokenization dictionaries to be generated for LALA, as compared to SMILES. The LFV did not show consistent behavior as an input: the RF model trained on LFVs performed the poorest; however, the KRR model trained on LFVs performed better than both CUSTODI[SMILES] and CUSTODI[LALAS]. This variance in performance is not surprising, as the RF and KRR models work best on significantly different latent spaces. A visual inspection of the individual plots in Figure 4 indicates that the relative energy is the only property with a qualitatively different picture. For this property, it appears that there is a dramatic difference in the performance of the two RNN models, and the performance of RNN model trained on the One-Hot[LALAS] representation appears to be noticeably poorer than the model trained on the One-Hot[SMILES] representation. We must emphasize, however, that such an interpretation is misleading, considering that, in fact, all the models show very satisfactory accuracy: they predict the relative electronic energy with MAE < 0.002 eV, which is smaller than the margin of error of the DFT calculations. Nevertheless, this apparent shift in performance (relative to the other molecular properties) led us to consider possible differences between the representations, which might affect the prediction of relative energy. One important difference is the way LALAS treat angular annulations. Angular annulations can have two types of direction -clockwise and counter-clockwise. Consecutive angular annulations in opposite directions create a zig-zag type of topology, which is planar in the ground state. 45,46 However, consecutive angular annulations in the same direction create cove, fjord, and eventually helix formations (for two, three, and four consecutive A annulations, respectively). These differences do not necessarily affect electronic properties (e.g., molecules with similar annulation sequences are equiaromatic -i.e., have similar aromaticity patternsregardless of the direction of the A annulations), however such substructures can affect relative energy as they have an increasing degree of curvature, which introduces helical strain, i.e., higher relative energy. Whereas SMILES representations include this information, LALAS do not differentiate between the types of angular annulations. Therefore, in principle, there could be performance discrepancies between the two; in practice, we observe that both perform exceedingly well on the given data. Having analyzed the performance of the individual models in terms of prediction accuracy, we now turn to comparing the training time required for each of the models. Table S2 (Supporting Information, Section S2) gives the average time per molecule for each of the models. In general, we find that using the LALAS (or representations derived from LALAS) markedly decreases training time for the RF and RNN models (by factors of ~6 and ~5, respectively), and moderately decreases training time for the KRR and CUSTODI models (by a factor of ~2 for both). Finally, a major advantage of LALAS is revealed when comparing the performance of models trained on smaller training sets. The top four best-performing input+model combinations were identified from Figure 4, and new models were trained on varying training-set sizes. Both LALAS and LFVs indeed show superior performance in small datasets compared to other tested methods. Using only 10% of the data for training, the RNN model trained on One-Hot[LALAS] achieved a normalized test set MAE of 0.12 eV, which is markedly lower than the other three models. At a training-set size of 40%, GC achieved similar results as the RNN, MAE = 0.11 eV, and at 70% all four models showed comparable results. This may be attributed to the concise nature of the LALAS: the LALAS of a given molecule is, on average, shorter by 86.5% (55 characters) than the SMILES of the same molecule. In addition, the complexity of the language is substantially reducedonly four types of characters. Thus, lower variance is expected for models trained on LALAS. ## Interpretation Based on Annulation Sequence As mentioned above, the simplification that is inherent in LALAS affects not only model performance, but also interpretability, which is a main goal of this work. Whereas single atoms or atom-pairs can have meaning as functional groups in many organic molecules, in PBHs individual carbon atoms often do not carry significant chemical meaning. LALAS connect textual characters with chemically meaningful subunits, i.e., specific ring annulation patterns. This makes it amenable to interpretation when used for training textbased models such as CUSTODI (the methodology for interpreting the CUSTODI model dictionary is presented in the Methods section). To extract the most meaningful insights from a given model, one should first ensure that the model shows good and reliable performance. Therefore, we initially performed a benchmarking procedure, to determine the optimal degree of CUSTODI for these data. This procedure is included in the hyperparameter optimization of the CUSTODI model, as the degree of CUSTODI is a hyperparameter of the model (see Methods for details). In other words: CUSTODI-1 was trained on subsequences of a single character (e.g., "L", ")"), CUSTODI-2 was trained subsequences containing either one or two characters (e.g., "L", "LA"), and CUSTODI-3 was trained on subsequences containing either one, two or three characters (e.g., "A", "(L", "ALA"). The best-performing model was found to be CUSTODI-2. The importance terms of the trained CUSTODI-2 model are presented in Figure 6. We emphasize that, while these terms can help assign the importance of the various structural features, they do not tell us in which way the features impact each property, i.e., increasing or decreasing the value of the predicted property. Such an analysis requires different treatment, which is the subject of ongoing work and will be disclosed in due course. Figure 6 shows that the properties HOMO, LUMO, and HOMO-LUMO gap have a similar dependence on particular substring sequences, which is not surprising. In addition, we observe a marked difference between the relative importance of the factors governing these three properties and those determining the relative energy of each molecule (note: the relative energy is calculated with respect to the respective lowest-energy isomer; for further details see 36 ). The adiabatic ionization potential (AIP) and adiabatic electron affinity (AEA) have some similarity to the three aforementioned electronic properties, which is in accordance with Koopman's theorem. 47 Yet, there are also dissimilarities, which demonstrate that the model is capable of distinguishing between the property types. The main factor influencing the HOMO, LUMO, and HOMO-LUMO gap is the presence of linear annulations (L, 𝛽 ̅ 17.7%) and stretches of two consecutive linear annulations (LL, 𝛽 ̅ 9.1%; i.e., four benzene rings annulated linearly, akin to naphthacene). These properties are affected by the presence of angular annulations to a lesser extent (A, 𝛽 ̅ 11.6%), while the existence of branching points does not seem to be important. Our recent analysis of the COMPAS-1D dataset showed that the HOMO, LUMO, and HOMO-LUMO gap all depend on the length of the Longest L subsequence. Because CUSTODI-2 only looks as subsequences up to two letters long, we cannot see here the importance of longer Longest L subsequences. Nevertheless, all of these observations are in line with our previous observations on these compounds. 36 In contrast, the main factors influencing the relative energy are different. We observe the following dependencies: linear annulations (L, 𝛽 16.5%), consecutive series of angular annulations (AA, 𝛽 11.3%), and branching points following an angular annulation (")A", 𝛽 10.8%). The subsequence "AL" also appears, which implies that it is not only the presence of angular annulations that matters, but also what surrounds them, or at what point the A sequence is broken. These results are in line with our previous observations pertaining to prediction of the relative energy, which we attributed to the strain that is incurred by sequences of consecutive A motifs. Specifically, we noted that consecutive sequences of A annulations can be either helical or planar, depending on the direction of the consecutive As. While A annulations in opposing directions lead to "zig-zag" formation that is planar, stretches of A annulations in the same direction lead to the formation of helical structures (known as cove, fjord, and helix). The features entail helical strain which raises the relative energy. Therefore, it is not surprising to find them among the main influencers in the prediction of relative energy. Corroboration for this interpretation can be found in our analysis of the COMPAS-1D dataset, which has shown that the increase in relative electronic energy is correlated to the deviation from planarity. 36 In this context, we note that the relationship between angular annulations and stability has also been investigated with other computational and conceptual tools. For example, the same observations can be interpreted in the context of Clar's rule, 48,49 which states that isomers with a larger number of Clar sextets are more stable than those with fewer Clar sextets. In general, angular annulations and branching points allow for more Clar sextets to be generated, which can therefore influence the relative energy. We are currently investigating the link between Clar structures, aromaticity indices, and the relative energy, to see if this interpretation can be substantiated. Other computational analyses have also rationalized the greater stability of angular isomers in the ground state via graph-theory, 50 additional π-bonding, 51,52 and a greater number of resonance structures. 53 Though the L motifs are predicted by the model to have an importance effect, the direction of this effect is unknown. Hence, it can, in principle, be perceived in two ways: a) following the previous rationalization, the presence of L motifs can be seen as precluding the formation of such non-planar motifs and therefore contributing to stabilization; or b) the L motifs may contribute to destabilization, not via geometric deformation but rather through an electronic effect. Since it is well-established that the most stable isomers are the phenacenes (i.e., the "zig-zag" PBHs), 51,52 one may conclude based on this previous knowledge that the operative case is (b). Nevertheless, we are currently working on implementation of more sophisticated DL models that also reveal the direction of each feature's influence. As mentioned above, the AEA and AIP mostly show similarity to the HOMO, LUMO, and HOMO-LUMO gap analyses, with some exceptions. The main difference is that for both AIP and AEA the angular annulation ("A", 𝛽 ̅ 17.7%) shows slightly greater importance than the linear annulation ("L", 𝛽 ̅ 14.2%). One possible explanation can be found in the work of Khatymov et al., who found that the stabilization of the LUMO is hampered due to specific symmetry features in the angular phenanthrene, which may be generalized to homologous series of angularly annulated PBHs. 54 As a result, within Koopmans' theorem (though just a crude approximation for our DFT-calculated values), the EA is expected to decrease in magnitude. An alternative, or complementary, explanation is that many of the molecules containing multiple A annulations have some degree of helicity, which may affect the charge delocalization. Therefore, the presence of As becomes an important factor for the predictive model. We note that, for all properties, the intercept has a large importance value, i.e., a large influence on the predicted value. As described in the Methods section, the intercept is a constant value that describes the bias of the CUSTODI model. In cases where the bias itself has a large value, relative to the individual tokenization values, the intercept has a strong influence. This can be understood in the following way: the CUSTODI model learns the "average value" of a property and the importance assigned to each of the subsequences represents the effect of the respective subsequence on that relative value. ## Interpretation Based on the LFV The RF model has an inherent way of finding feature importance. 55 Our analysis focuses on the RF model trained on LFVs (Figure 7). The results show very similar patterns to those obtained with CUSTODI [LALAS]. Considering that LFVs are essentially domain expertise-based features which we extracted from LALAS, this implies that the CUSTODI model successfully captures the features directly from the textual representations, without the need for human intervention. The RF model shows that the HOMO, LUMO, and HOMO-LUMO gap are mainly affected by the length of the longest stretch of linear annulations ("Longest L", 87%). Unsurprisingly, the AIP and AEA are also mainly influenced by the linear annulations ("Longest L", 80.2%). However, AEA and AIP are also affected by the number of rings, which is in line with previous reports of a size-dependency for these properties. 56 It is generally considered that the larger a conjugated system is, the better it is expected to stabilize excess charge through delocalization. The relative energy displays a very different set of dependencies, chief among them are the longest stretch of angular annulations ("Longest A", 27.1%), the number of branches in the molecule ("No. Branching points", 21.2%), and the degeneracy of the longest linear sequence ("Longest L Degeneracy", 20.6%). The ratio of L motifs, the longest linear sequence, and the number of LAL sequences also have non-negligible influences (10%, 12.1%, and 6.4%, respectively). As mentioned above, we believe that the impact of the angular annulations can be attributed either to variations in helical strain or to the possible number of Clar sextets that can be formed. Similarly, the number of branches is influential because it is related to the tendency to form helical structures (an increase in branches precludes linear stretches and increases the likelihood of angular stretches in similar directions). As we explained above, we hypothesize that, while the A motifs appear to raise the energy through geometrical deformation, the L motifs raise it via electronic effects. Thus, we observe a dependence also on several features describing the presence of L motifs. As opposed to the other properties, where only the longest linear stretch was important, here also the degeneracy (i.e., the longest stretch that appears more than once) is important. This indicates that the effect of individual linear stretches on the relative energy may be additive, while on other properties it is exclusive. Interestingly, the relative energy also shows a dependence on a specific substructure, "LAL". This particular subsequence was previously noted as behaving in an anomalous manner 24 in the prediction of magnetic behavior in PBHs. A similar analysis using the CUSTODI model trained on SMILES strings yielded no meaningful results, as the substrings used in CUSTODI models are short (this results from the hyperparameter optimization; see Section S4.2 in the Supporting Information for more details). The results of the influence analysis on SMILES strings are also provided in the Supporting Information (Section S5, Figure S1). Similarly, RF trained on CUSTODI[SMILES] did not afford any interpretable results. ## Conclusions In this work, we applied interpretable ML tools to investigate the structure-property relationships in the family of PBHs, which are archetypal polycyclic species. We introduced a new type of textual molecular representation, which is specifically suited for these molecules. This representation can be used either in string form (LALAS) or as the basis for a feature-vector (LFV). In addition, we applied a new type of interpretable ML method, CUSTODI. Comparison to standard models and input types demonstrated the added value of LALAS to both efficiency and interpretability. The application of these two new tools to the newly reported database, COMPAS-1D, [ref: database paper] allowed us to gain chemical insight into the structure-property relationships of PBHs. Four main conclusions were reached: (1) most of the electronic properties of PBHs we studied are primarily influenced by the presence and length of consecutive linear annulations in the molecule; (2) the relative energy of isomeric PBHs is mainly affected by the presence of angular annulations and branching points in the molecule; (3) as expected from Koopmans' theorem, AIP and AEA have similar dependencies as HOMO, LUMO, and HOMO-LUMO gap, however, the former two are also sizedependent while the latter appear not to be; (4) there are "privileged" subsequences, one of which we identified -"LAL". To a certain extent, (some of) these insights may be considered well-known "rules of thumb" or "conventional wisdom" in the chemical community. However, to the best of our knowledge, have never been demonstrated in a data-driven manner. Indeed, the agreement between the ML interpretation and generally accepted chemical behavior indicates that the models performed reliably well, and we have validated these rules of thumb with an unprecedented dataset containing ~8,700 PBHs. Nevertheless, there are also new insights, such as factors influencing relative energy of PBH isomers and the existence of "privileged" subsequences. We also emphasize that the importance analysis presented here indicated that the relationship between linear sequences and the various molecular properties is different. Specifically, for all of the properties except the relative energy, it appears that only the single longest linear stretch is important and how many times such a sequence appears does not matter; in contrast, for the relative energy, the degeneracy of these sequences does matter, which suggests that they might contribute cumulatively to destabilization. Importantly, similar conclusions were obtained using the CUSTODI model, which was trained on LALAS without any preprocessing, and the RF model, which was trained on LFVs -domain-expert curated features. This serves to indicate that the CUSTODI model is capable of extracting the important structural features from this new representation automatically, without expert intervention. We emphasize that CUSTODI can be used in a similar manner on different string representations to derive structure-property relationships. Both the RF and CUSTODI models describe the relative importance of various structural features/subunits, but they could not describe their effect -i.e., increase or decrease in magnitude. Our group is currently exploring the use of additional interpretable algorithms to provide further insight into this, as well as other, aspects. In particular, we are investigating the direct impact of individual structural motifs on different aromaticity indices. Additional emphasis is on the expansion of the LALAS representation concept to include peri-condensed and poly(hetero)cyclic aromatic systems and on generating the relevant data to enable further exploration and analysis of this chemical space. ## Data and Code Availability The full code used in this paper appear in our GitLab repository at https://gitlab.com/porannegroup/lalas. The data was taken from the COMPAS Project repository at https://gitlab.com/porannegroup/compas.

chemsum

{"title": "Revealing Structure-Property Relationships in Polybenzenoid Hydrocarbons with Interpretable Machine-Learning", "journal": "ChemRxiv"}

metal-doped_mesoporous_zro₂_catalyzed_chemoselective_synthesis_of_allylic_alcohols_from_m

## Abstract: Meerwein-Ponndorf-Verley reduction (MPVr) is a sustainable route for the chemoselective transformation of a,b-unsaturated aldehydes. However, tailoring ZrO 2 catalysts for improved surface-active sites and maximum performance in the MPV reaction is still a challenge. Here, we synthesized mesoporous zirconia (ZrO 2 ) and metal-doped zirconia (M_ZrO 2 , M = Cr, Mn, Fe, and Ni). The incorporation of metal dopants into zirconia's crystal framework alters its physico-chemical properties such as surface area and total acidity-basicity. The prepared catalysts were evaluated in the MPVr using 2-propanol as a hydrogen donor under mild reaction conditions. The catalysts' remarkable reactivity depends mainly on their surface mesostructure's intrinsic properties rather than the specific surface area. Cr_ZrO 2 , which is stable and sustainable, presented superior activity and 100% selectivity to unsaturated alcohols. The synergistic effect between Cr and Zr species in the binary oxide facilitated the Lewis acidity-induced performance of the Cr_ZrO 2 catalyst. Our work presents the first innovative application of a welldesigned mesoporous Cr_ZrO 2 in the green synthesis of unsaturated alcohols with exceptional reactivity. ## Introduction The chemoselective reduction of a,b-unsaturated aldehydes to their corresponding allylic alcohols is one of the significant chemical transformations in synthetic organic chemistry. The unsaturated allylic alcohol (UAA) produced through this process is widely utilized as the primary feedstock in food and perfumery industries and intermediates in pharmaceutical industries. 1,2 However, the reaction is classically carried out using gaseous hydrogen in the presence of noble metals as catalysts with significant limitations such as high-pressure requirements and low selectivity to UAA. 3 The high-pressure involved in such a chemical process requires expensive equipment and an elaborate experimental set-up with associated safety risks. 4,5 The problems associated with these classical methods could be avoided in the Meerwein-Ponndorf-Verley (MPV) reduction. The MPV is an alternative means of selective reduction of unsaturated carbonyls at atmospheric pressure using safe and readily available secondary alcohol as a hydrogen donor instead of high-pressure molecular hydrogen or hazardous reduction reagents such as LiAlH 4 and NaBH 4 . 6 The MPV among several reduction routes for functional group conversion has the following prevailing advantages: (i) easy to handle hydrogen donor without the requirement of heavy gas containment, (ii) cheap and environmentally friendly source of hydrogen (iii) enhanced selectivity under mild reaction conditions (iv) safer process (v) minimized waste (vi) reduced cost, e.g., maintenance, separation, and other production logistics (vii) economical and environmentally sustainable process. 7,8 However, to force the equilibrium reaction towards the formation of UAA, the MPV reduction requires excess sacrificial H-donating molecule, 9 which generates by-product. 10,11 The product separation through distillation is a tall task owing to the close boiling points of the a,b-unsaturated aldehydes or ketones, the UAAs, and the sacrificial alcohol. Hence, efforts should be directed towards attaining 100% selectivity and product yield and recycling the by-product generated from the sacrificial hydride donor. 11 The MPV reaction is highly chemoselective to the reduction of CQO bond in the a,b-unsaturated aldehydes but, the hydrogenation of conjugated CQC is preferentially favorable thermodynamically and kinetically over the CQO bond. 3,12 There is an ongoing research effort to develop an efficient catalyst system in MPV reactions. The catalytic performance of metal oxides depends on the surface's structural features or environments, influencing their adsorption strength and activation of the adsorbates. 24,25 Among a wide range of heterogeneous catalytic species utilized in MPV reduction, zirconia showed the most promising performance. 26 The potential catalytic activity of zirconia has increased its application in catalysis. It is known for its acidic properties, 27 high thermal stability, and corrosion resistance. 7,28 To further emphasize this fact, Alvarez-Rodriguez et al. pointed out that the zirconia catalyst is more effective than other conventional catalysts such as alumina or silica to produce unsaturated allylic alcohol from citral and cinnamaldehyde. 29 The surface active sites of zirconia has been verified to improve by addition of dopant. 30 Xie et al. prepared Cr-ZrO 2 using acidbase pair pathway with evaporation-induced self-assembly. 31 They found out that the catalytic performance in dehydrogenation of propane with CO 2 was influenced by the enhanced surface morphology, acidity and redox property of the Cr-ZrO 2 catalysts with different Cr doping percentage. Also, in the report of Wu et al. the catalytic activity of the Cr 2 O 3 -ZrO 2 prepared hydrothermally was attributed to the presence of Cr 6+ species. 32 From the above literature survey coupled with the crucial need for sustainable and environmentally benign catalytic hydrogenation processes, 11 it would be interesting to develop novel mesoporous zirconia-based catalyst systems with increased acidic sites for the transfer hydrogenation process under mild reaction conditions with high selectivity to unsaturated allylic alcohol. Also, a stable catalytic MPV process without the use of additives is imperative for the green and clean synthesis of UAA. The surface properties of zirconia, a major determining factor of its catalytic activity, may depend on the method of synthesis. 33 Several methods of preparing mesoporous zirconia have been reported, but the inverse micelles soft-templated technique is poorly represented. The inverse surfactant micelle approach enables the control of the surfactant-transition metal interactions, hydrolysis, and condensation of inorganic sols. 34,35 Transition metal oxides prepared using the inverse micelles approach are known for their exceptional structural properties such as large surface area (S BET ), porosity, and crystallinity. The large surface area avails the active sites for better catalytic activity than other porous materials with low S BET . Hence, in this study, a triblock copolymer P-123 was employed as a structure-directing agent in the inverse micelles approach 34,35 for the synthesis of mesoporous zirconia and metal-doped zirconia. The mesoporous zirconia-based materials' catalytic potential was investigated as active phases in the MPV reduction of selected aromatic aldehydes to their corresponding alcohols (Scheme 1) without any additives or co-solvent. We optimized the MPV process variables. A deep insight into the effect of cation dopant on the physicochemical properties of ZrO 2 , including crystal phases, structural morphology, surface area, and acidity-basicity, was evaluated. Furthermore, the relativity of these properties to its catalytic performance in the MPV process was also revealed. Elucidation of the activity was based on the observed pseudo-first-order rate constants (k obs ) and calculated conversion Equations S1-3. Also, we show that all the synthesized catalysts exhibit excellent selectivity to the unsaturated allylic alcohol in 2-propanol as H-donor at 80 1C, 450 rpm, and atmospheric pressure. To the best of our knowledge, this is the first work that applied mesoporous Cr_ZrO 2 prepared via inverse micelle for the MPV process. The mesoporous Cr_ZrO 2 showed Scheme 1 MPV reduction of a,b-unsaturated aldehydes to their corresponding unsaturated allylic alcohol over the mesoporous zirconia-based catalyst. considerable conversion with 100% selectivity to the unsaturated alcohols in reducing citral as a model reaction and other aldehydes. The synthesized Cr_ZrO 2 is a sustainable catalyst for the Meerwein-Ponndorf-Verley process. ## Materials Nitric acid (HNO 3 ) (69-70%) was purchased from Rochelle Chemicals (RSA). 1-Butanol (99.8%), ethanol (99.9%), poly(ethylene glycol)block-poly(propylene glycol)-block-poly(ethylene glycol) (PEO20-PPO70-PEO20 or Pluronic P-123), zirconium(IV) butoxide solution (80% in 1-butanol), manganese(II) nitrate tetrahydrate (97%), nickel(II) nitrate hexahydrate (99%), cinnamaldehyde (99%), citral (mixture of cis-and trans-isomers) (96%), benzaldehyde (99%), crotonaldehyde (99%), furfural (99%), 2-propanol (99.5%), decane (99%) were all purchased from Sigma-Aldrich. Ferric nitrate nonahydrate was purchased from SRL chemicals, and chromic(III) nitrate nonahydrate (98%) was purchased from UNIVAR, SAAR CHEM pty. All chemicals were of analytical grade and used as received without further purification. ## Synthesis of catalysts The procedure already reported was followed for the synthesis of ZrO 2 . 34 Briefly, 15.34 g (0.040 mol) of zirconium butoxide was added to a solution containing 25.04 g (0.336 mol) 1-butanol, 4.0 g (6.8 10 4 mol) of P-123, and 4.0 g (0.064 mol) HNO 3 . The mixture was stirred overnight, and the clear gel was dried in an oven at 120 1C for 4 h. The yellow glassy thin flakes obtained were calcined in air at 350 1C for 5 h with a heating rate of 2 1C min 1 . The metal-doped zirconia was synthesized by adding the metal dopant (Cr, Mn, Fe, and Ni) precursor to zirconium butoxide in a molar ratio of 1 : 4. As in the case of pure ZrO 2 , the same thermal treatment was applied (Scheme S1, ESI †). The samples are tagged M_ZrO 2 , M = metal dopant. ## Characterization of catalysts Powder X-ray diffraction (p-XRD) analyses were carried out on a Philips XPERT-PRO diffractometer system operating with Cu Ka1, Ka2, and Ni Kb radiation (l = 1.5406, 1.54443 and 1.39225 , respectively) at 25 1C. Both low and wide 2y range (i.e., 4-901) angle diffraction patterns with a step of 0.1701 were measured. The Debye-Scherrer equation (eqn (S1), ESI †) was used to calculate the mean crystallite size. Micrometric ASAP 2460 sorption system gave the nitrogen sorption measurements. The samples were firstly degassed under flowing nitrogen at 100 1C for 18 h and under vacuum for 10 h at the same temperature before the experiments to remove any physisorbed moisture. The surface areas were calculated using the Brunauer-Emmett-Teller (BET) method. Transmission electron microscopy (TEM) for the confirmation of the mesostructure was achieved on a JEOL Jem-2100F electron microscope with an accelerating voltage of 200 kV. The pore diameter was measured using ImageJ software. A 10 mg of the catalyst was sonicated in 1 ml of methanol for 1 h, and a drop of the suspension was placed on a carbon-coated Cu-grid then allowed to dry before the TEM analysis. The prepared samples' surface morphology was identified on a Tescan Vega 3 LMH scanning electron microscope (SEM) using a scattering electron detector with a high voltage of 20.0 kV. Prior to analysis, the samples were placed on an aluminum stub and carbon-coated in an Agar Turbo carbon coater. The quantity of dopant M n+ was verified with energydisperse X-ray spectroscopy (EDX). The distribution of the metal species was identified by elemental mapping on SEM. Also, the dopant content in the solid samples was measured using a Spectro Acros ICP-OES spectrometer. The Fourier transform infrared spectroscopy (FTIR) spectra of the samples were recorded on a Bruker FTIR Alpha spectrometer in the 4000-400 cm 1 region. The samples were mixed with KBr, and the analysis was performed in the transmission mode under ambient conditions. The NH 3 /CO 2 temperature-programmed desorption (TPD) studies to determine the materials' acidity or basicity were performed on a Micromeritics AutoChem II. About 0.2 g of the sample was loaded in a quartz tube reactor. The loaded sample surface was degassed in a He gas flow at 200 1C for 1 h before the TPD measurement. We used a mixture of NH 3 or CO 2 and helium in the ratio of 10 : 90 as the probe gas at a flow rate of 50 ml min 1 . Measurements were performed in the temperature range of 30-550 1C at a temperature ramp of 10 1C min 1 and 3 1C min 1 for TPD-NH 3 and TPD-CO 2 , respectively. For identifying the Lewis and Brønsted acid sites on the samples, adsorbed pyridine FTIR analysis was carried out. Before the analysis, B0.03 g of the sample was activated by degassing under gaseous nitrogen at 300 1C for an hour and cooled to room temperature. After that, the activated catalyst was contacted with pyridine (200 ml) at 120 1C for 30 min. Subsequently, the physisorbed pyridine was evacuated under vacuum at ambient temperature for an hour, 36 and the sample was analyzed on a Bruker FTIR Alpha spectrometer. The H 2 -TPR (hydrogen-temperature programmed reduction) analysis was conducted on the same Micromeritics Autochem II. Approximately 30 mg of the catalyst was loaded in the quartz tube reactor and pretreated under Argon flow at 200 1C for 1 h to ensure the catalyst surface is clean before each test. After the pretreatment, H 2 /Ar (10 : 90) was passed over the catalyst at a 50 ml min 1 flow rate. The measurements were performed within the ambient temperature to 800 1C with a 10 1C min 1 ramping rate. The prepared samples' thermal stability test was performed on a PerkinElmer STA 6000 thermogravimetric analyzer (TGA). The degradation study temperature was varied from 25-900 1C with a ramping rate of 10 1C min 1 under air at a 20 ml min 1 flow rate. The UV-vis spectra of the samples were obtained on a microplate reader (PowerWave HT.Biotek microplate reader). Before obtaining the UV-vis absorption spectra, about 30 mg of the solid sample was sonicated in 2 ml methanol and decanted. After that, the supernatant was analyzed using a 24-well plate. ## Evaluation of catalytic performance The liquid phase MPV reduction experiments were performed on a carousel reaction station multi-reactor (Radley Discovery Technologies) with twelve 50 ml vials. The 50 ml reactor vial was charged with 0.4 g M-ZrO 2 , 2.50 mmol of aldehyde, 1.00 mmol (200 ml) decane as an internal standard, and 130 mmol (10 ml) 2-propanol. Followed by reflux at 80 1C and stirring at 450 rpm with a 16.5 mm crossbar stirrer. After filtration, the filtrate was analyzed on a Shimadzu GC-2010 with flame ionization detector (FID) using a capillary column (Restek RTX-5; 30 m, 0.25 mm ID, thickness 0.25 mm) in N 2 carrier gas. The injection port and FID temperature were maintained at 200 1C and 350 1C, respectively. The products were further confirmed by a Shimadzu GC-MS QP-2010 using the same capillary column with the injection temperature at 200 1C. The ion source and interface temperatures were 200 1C and 250 1C, respectively. For the GC FID and MS, the column oven temperature program started at 40 1C (hold 2 min), then programmed at 20 1C min 1 to 280 1C (hold 5 min); the total analytical time was 19 min (details in Section S1.3.2, ESI †). The catalysts were screened with the MPV reduction of citral as a model reaction. The catalyst exhibiting the best activity was chosen for the transfer-hydrogenation of selected a,b-unsaturated aldehydes. The substrate conversion, product selectivity, and normalized activity were calculated (eqn (S2)-(S5), ESI †). Furthermore, the observed k obs for each experiment were calculated using Kinetic studio version 2.08 software. For the recyclability study, the catalyst was pretreated by calcining at 350 1C/5 h before reuse. No extra peaks were detected, which confirms the purity of the synthesized t-ZrO 2 . Upon doping t-ZrO 2 with a metal atom, an isomorphous substitution was observed. Suggesting that some surface Zr atoms in the M_ZrO 2 samples are substituted with the dopant atoms and the formation of a homogeneous solid solution of binary M x O y -ZrO 2 . This claim is supported by the gradual shift of the peak at 30.41 (101) of ZrO 2 towards a higher 2y degree. No identifiable peak is associated with the dopants, which is an indication that the dopants are well incorporated into the ZrO 2 matrix and high dispersion of the dopant species. The incorporation of cation into the crystal framework of zirconia significantly influenced its crystallinity. The tetragonal structure with reduced peak intensity remains in the presence of Mn and Fe, while we observed crystal distortion in the case of Ni and Cr dopants. The dopant species weakened the t-ZrO 2 peaks in the case of Mn and Fe and were significantly destroyed in Ni and Cr, forming disordered ZrO 2 . This observation implies the degree of dopant incorporation and distribution and M n+ -Zr 4+ interaction. The observed broader and weaker diffraction peaks in Mn_ZrO 2 and Fe_ZrO 2 explain the occurrence of higher surface area in correlation with the host ZrO 2 . Also, the crystallite size of t-ZrO 2 (5.50 nm) decreased upon doping with Mn_ZrO 2 (1.37 nm) and Fe_ZrO 2 (2.59 nm). The decrease in the crystallite size possibly contributed to expanding the surface area, as depicted in Table 1. ## Surface properties of the M_ZrO 2 catalysts The experiments carried out to confirm the surface structure and porosity of pure and doped ZrO 2 using nitrogen sorption analysis shown in Table 1 revealed that the pure ZrO 2 exhibited pores with an average diameter of 2.53 nm within the meso range with a large BET surface area (S BET ) 206 m 2 g 1 . The BET surface area of the M_ZrO 2 catalysts depends on the final catalyst's crystallinity, which is a function of the dopant's nature. The S BET of pure ZrO 2 (206 m 2 g 1 ) increased when modified with Mn (221 m 2 g 1 ) and Fe (223 m 2 g 1 ) but decreased in the case of Cr (190 m 2 g 1 ) and Ni (193 m 2 g 1 ) dopants. The pore sizes are within the range of 2.5-3.6 nm, a characteristic feature of mesoporous oxides. The metal dopants' addition distinctly enlarged the pore size 2.5-3.6 nm and the pore volume 0.1 to 0.26 cm 3 g 1 . The mesostructuring is associated with the metal-metal grain boundary adhesion/ expansion during the formation of oxo-metal clusters at the stage of sol condensation, diffusion of volatile species, and subsequent removal of the surfactant template. The degree of grain segregation is directly proportional to the porosity of the material. Also, the larger the porosity, the slower the grain growth, as depicted in the correlation of the pore size and the crystallite size (Table 1). The observed physisorption isotherms for all the catalysts in Fig. 2a are typical of Type IV compared with the IUPAC classification reported by Thommes et al. 40,41 This further indicates that all the catalysts are mesoporous materials with thin capillary pores. The ZrO 2 catalysts exhibited hysteresis loops P/P 0 typical of H2 type indicating the occurrence of cavitation controlled evaporation; this depicts that the materials possess a heterogeneous pore network with the neck size (W) distribution much more narrow than the size distribution of the cavities (W c ), that is, W o W c . The pore size distribution of pure zirconia and M_ZrO 2 are shown in Fig. 2b. All the catalysts showed unimodal pore size distribution, with the M_ZrO 2 exhibiting a narrower pore distribution compared to the pure ZrO 2 catalyst. Significantly, the corresponding BET surface area, pore-size distribution, and pore volume is an indication that the catalysts are mesoporous with high surface area, and the surface mesostructure of ZrO 2 could be tailored by adding foreign atomic specie. Moreover, the transmission electron microscopy (TEM) and high transmission (HRTEM) images of the zirconia systems as displayed in Fig. 3a, b, d, e and Fig. S1, S2 (ESI †) reveal that they are made of nanosized particles with intraparticle voids that were preserved upon the addition of different metal ions. The TEM images (Fig. 3a, d and Fig. S1, S2, ESI †) show that the pores are well distributed in the ZrO 2 matrix, supporting the evidence of the presence of pore and the pore enlargement upon doping as presented by the N 2 sorption experiment (Table 1 and Fig. 2b). A similar observation was reported in the work of Xie et al. 31 The surface morphologies of the catalysts are shown in Fig. 3c, f, and Fig. S3 (ESI †). Modification of the t-ZrO 2 surface morphology upon the introduction of dopants is insignificant; this is due to the homogeneity of the M_ZrO 2 solid structures, as observed in the XRD patterns. The EDX mapping (Fig. 3g The thermal stability of the catalysts (Fig. 4a) suggests that the catalysts are thermally stable. The dopant species enhance the thermal stability of the host ZrO 2 (8%), except Mn_ZrO 2 , which shows an approximately similar weight loss of 8.3%. The 5% degradation between 30-257 1C is attributed to the removal of adsorbed surface and bulk water molecules, while above 257 1C could be classified as degradation due to the decomposition of the organic surfactant residue. Above 683 1C, the spectra seemingly flatten out, suggesting minimal or no decomposition and the inorganic material's stability. The samples are also stable in the catalytic characterization and application temperature within this study's scope. The catalysts' hydrogen consumption temperatures (Fig. 4b) and the minimum temperatures (Table 2) suggest the catalysts' reducibility. The pure zirconia showed a poor hydrogen uptake with a small peak around 652 1C, corresponding to the reduction of bulk lattice oxygen of zirconia. We found that doping enhanced the reducibility of ZrO 2 , with a significant shift in its reduction peak to lower temperatures; this depicts the rate of the redox reaction. However, the degree of reducibility is dependent on the kind of doping species. The reduction peaks between 261-360 1C likely represent the reduction of the metal dopant species: Cr 3+ -Cr 2+ , Fe 3+ -Fe 2+ , Mn 2+ -Mn 0 , Ni 2+ -Ni 0 . The reduction peaks demonstrated in the region of 430-486 1C and above 600 1C are typical of surface and bulk reduction of lattice oxygen of zirconia, respectively. The Cr_ZrO 2 demonstrated the superior reduction capacity of surface interaction with hydrogen at the lowest temperature of 261 1C. This is likely due to the strong synergistic interaction between the Cr 3+ and Zr 4+ (Cr x O y -ZrO 2 solid solution). The H 2 -TPR data supported the superior catalytic activity of Cr-Zr active phase species for the H abstraction-release mechanism in the MPV dehydrogenationhydrogenation reaction. We investigated the surface acid-base properties of the prepared catalysts by NH 3 -and CO 2 -TPD analyses. The spectra are depicted in Fig. 5 and 6, respectively. The total acidity and basicity, along with their density, are summarized in Table 2. The total acidity and basicity were obtained from the peak area of NH 3 and CO 2 desorption, respectively. The acid or base sites density was derived by dividing the total acidity or basicity by the surface area (Table 1). The NH 3 /CO 2 desorption peak around 200 1C represents the acid/base sites of a weak strength, from 200 1C to 350 1C depicts medium strength acid/base sites, and above 400 1C corresponds to strong acid/base sites. 43 The surface basic properties of all the catalysts are depicted in Table 2 and Fig. 5. A similar chair-like CO 2 -TPD profile was reported for ZrO 2 44 and Cu/ZrO 2 /CaO. 45 The base concentrations of the prepared catalysts ranged from 0.3-1.1 mmol CO 2 g 1 . The basic sites distribution of the catalysts illustrated in Fig. 5 depicted that all the catalysts exhibited basic sites of both weak and strong strength, although with different peak intensities. The base concentration of the pure ZrO 2 (0.9 mmol CO 2 g 1 ) was approximately similar in the presence of Fe (0.9 mmol CO 2 g 1 ) but slightly increased upon doping with Mn (1.1 mmol CO 2 g 1 ) and Ni (1.0 mmol CO 2 g 1 ). An exception occurred in the case of Cr_ZrO 2 ; the Cr species significantly decreased the base concentration of pure ZrO 2 to 0.3 mmol CO 2 g 1 . The catalysts' basic density showed a similar trend, which was also confirmed by the reduction in the peak intensity representing the weak strength basic sites on the pure ZrO 2 in the case of Cr_ ZrO 2 . Fig. 6 shows the distribution of the acidic sites of weak to strong strength in the meso-ZrO 2 . Upon doping, the acidic sites underwent modulation. In the presence of Mn, Fe, and Ni, only two broad peaks representing weak and medium strength acid sites were observed. Whereas the Cr_ZrO 2 appears unique, which gave a more prominent shoulder desorption peak on the high-temperature side at ca 436 1C, suggesting a stronger surface acidic site. A similar trend of NH 3 desorption over Cr doped ZrO 2 was reported. 31 The Cr_ZrO 2 (0.7 mmol NH3 g 1 ) possessed the highest concentration of acidity among the catalysts (Table 2). The NH 3 desorption results suggest the proton-donating capacity of the surface acid site on the catalysts. The stronger the proton-donor tendency, the more strongly it binds with the base (NH 3 ), and the higher the required NH 3 desorption temperature. Hence, Cr_ZrO 2 possesses stronger electrophilic active sites (acid sites) needed for the selective adsorption of citral via the CQO bond. Comparatively, the NH 3 -and CO 2 -TPD data revealed that all the catalysts exhibit both surface acidic and basic sites. However, these active sites are not equivalent, as observed in the data derived from the acid to base ratio (Table 2); the dominance in terms of strength, total concentration, and density depends on the metal-metal interaction nature. Meanwhile, Cr_ZrO 2 presents more acid sites density and strength than the pure ZrO 2 , which possibly originates from the interaction of the Cr-Zr species at the atomic level. The tuning of the surface acid-base character of ZrO 2 was achieved by incorporating metal ions, which facilitated the understanding of the effect of acid/base sites of the catalysts on the MPV reduction of aldehydes. To further understand the surface components, types, and structures of the acid sites of the catalytic materials, investigation of the surface functionality and nature of acid sites was achieved through FTIR (Fig. 7a and b) and pyridine-adsorbed FTIR (Fig. 7c and d) spectroscopy methods, respectively. In Fig. 7a, the broad absorption band around 500-823 cm 1 is typical of Zr-O-Zr vibration in the tetragonal structure, and 46 This agrees well with the TG analysis (Fig. 4a), indicating residual surfactant is present in the catalysts. The IR peak at 1622 cm 1 suggests the bending hydroxyl group vibrations, while the broad and strong peak at 3435 cm 1 is a reflection of physically adsorbed moisture on the surface, hence showing the O-H stretching of water. 47 Upon doping, the peak intensity at 3435 cm 1 decreased, indicating a decline in hydroxy groups and hydrophilic property of ZrO 2 . 48 Like pXRD patterns (Fig. 1), the peak intensity at 500-823 cm 1 representing the t-ZrO 2 decreased upon the substitution of M n+ into ZrO 2 . Besides, the shifting of the bands towards the lower wavenumber was observed (for instance, 591 to 519 cm 1 ), which is most significant in Cr_ZrO 2 . This shifting is due to variation in the bond length when M n+ ions replace Zr 4+ ions. Hence, it confirmed the successful incorporation of the metal ions into the ZrO 2 lattice. In Cr_ZrO 2 catalyst, small peaks at 1210 and 1740 cm 1 are observed, attributed to the formation of Cr x O y , 49 due to the strong interaction between Cr-Zr. This could be responsible for the generation of more active sites on the surface of Cr_ZrO 2 . It is observed in Fig. 7b that the peak at 1740 cm 1 corresponding to the Cr species disappeared after five catalytic cycles, which suggests that the Cr_ZrO 2 perhaps undergoes a surface structural transformation in the course of further thermal pretreatments during reuse. Pyridine is a sensitive probe molecule for the classification of Lewis acid and Brønsted acid sites. As depicted in Fig. S5.7c (ESI †), the pyridine-IR bands at 957, 1400, and 1615 cm 1 are typical of the Lewis acid site. Two different acidic strengths due to the Lewis acid sites are shown in ZrO 2 , Fe_ZrO 2 , and Mn_ZrO 2 , whereas Lewis acidity of three different strengths was observed for both Ni_ZrO 2 and Cr_ZrO 2 . The higher the assumed frequency of the IR bands, the stronger the acidity of the sites. 50 The IR band at 1538 cm 1 suggests Brønsted acid site while 1485 cm 1 indicates C-C oscillation of pyridine aromatic ring chemisorbed on both Brønsted and Lewis acid sites. 51 The ZrO 2 catalyst showed no peak typical of the Brønsted acid site, but two characteristic bands (1400 and 1615 cm 1 ) associated with pyridinium ions coordinately bonded to Lewis acid sites. 52 These two adsorption bands were retained with increased intensity upon the addition of Mn and Fe species. More broad bands for Lewis acid, Brønsted acid, and a combination of Lewis acid and Brønsted acid sites were found when ZrO 2 was doped with Ni and Cr, accompanied by an increased intensity on Cr_ZrO 2 catalyst. The pyridineadsorbed IR reveals that the total acidity obtained from the NH 3 -TPD data has a larger Lewis to Brønsted acid ratio, which is highest in the Cr_ZrO 2 and the main factor that governs its catalytic activity in this study. The results indicate the possibility of tuning the active sites of ZrO 2 by adding foreign atomic species. The UV-vis absorption spectra of undoped and metal-doped ZrO 2 are shown in Fig. 8. The results indicate that the undoped ZrO 2 and M_ZrO 2 (M = Fe, Mn, and Ni) have no absorption peaks in the visible wavelength region of 300 to 700 nm. However, after Cr doping, a new absorption peak appears at around 360 nm; this is attributed to the band-gap transition of ZrO 2 due to Cr 3+ ions. 53,54 The Cr-3d electronic configuration results in the appearance of some localized states in the host band-gap and makes electron transfer easier than the undoped system. This agrees well with the H 2 -TPR data (Fig. 4b) of Cr_ZrO 2 . All these issues explain the reason for the generation of more Lewis acid sites on Cr_ZrO 2 compared to the undoped system. ## Surface performance of the M_ZrO 2 catalysts in MPV reduction of citral 3.2.1. Dopants vs. catalytic reactivity. The MPV reduction of citral with 2-propanol was selected as a model reaction to examine the activity of the prepared pure and metal-doped mesoporous zirconia (M_ZrO 2 ) catalysts. Table 3 indicates their respective performance. Interestingly, this work's catalytic systems gave 100% selectivity to the unsaturated allylic alcohol (nerol + geraniol). The pure ZrO 2 presented 62.6% citral conversion. Comparison with the pure ZrO 2 , the effect of metal dopant in terms of activity enhancement was only observed when ZrO 2 was doped with Cr species. The Cr_ZrO 2 gave optimal activity of 76.4% conversion of citral. The catalytic activities' observed trend follows the order Cr_ZrO 2 4 ZrO 2 4 Mn_ZrO 2 4 Fe_ZrO 2 4 Ni_ZrO 2 . The correlation of the surface area, acidity, and basicity with the catalytic activity in MPV reduction of citral is displayed in Fig. 9. Generally, the surface area controls the catalyst activity (high surface area, high catalytic activity). This is not the case in our proposed catalytic systems, as catalysts with higher surface area Mn_ZrO 2 and Fe_ZrO 2 gave low conversion of citral. Instead, we found that the catalysts' catalytic activity in the MPV reduction of citral is governed mainly by the kind of metal dopant, acidic site density, and reducibility. As shown in Table 2 and Fig. 9, the catalyst Cr_ZrO 2 with the highest acidity (acid:base ratio) presented the highest activity, whereas catalysts Mn_ZrO 2 , Fe_ZrO 2 , and Ni_ZrO 2 with higher basicity compared to that of Cr_ZrO 2 and the pure ZrO 2 decreased the activity of the host ZrO 2 . It could be deduced from the results that surface-active acid sites, Lewis acid in particular, play an essential role in the MPV reduction of citral. Also, the superior reduction capacity of Cr_ZrO 2 favors its performance. 3.2.2 Recyclability, leaching test, and characterization of Cr_ZrO 2 catalyst after reuse. Developing a recyclable catalyst without decomposing due to long-term reuse has become a critical factor in achieving a sustainable catalytic system. As shown in Fig. 10a, the reduction of citral hardly proceeds after removing the catalyst, indicating no residual active component in the reaction liquid. The recyclability test (Fig. 10b) reveals that the Cr_ZrO 2 is catalytically stable and reusable with excellent selectivity to UAA although, a slight variation in the activity during the 3rd and 4th reaction cycle was observed. The characterization of the spent gave more insights into the possible transformation in the catalyst during reuse. The FTIR (Fig. 7b) shows that the crystal phases remained, the nitrogen sorption results (Table 1 and Fig. S5.7a, b, ESI †) show a significant decrease in the S BET and pore size ascribed to the sintering effect due to repeated thermal treatment during the regeneration process. The adsorbed pyridine experiment (Fig. 7d) reveals a decline in the Lewis acid sites and a significant loss of the Brønsted acid site; this suggests that the Cr_ZrO 2 catalyst undergoes surface restructuring during catalyst pretreatment before reuse. Hence, the effect of Brønsted acidity could be negligible in this study. Nevertheless, the TEM image (Fig. S7c and d, ESI †) shows the stability of the mesostructure after 5 consecutive catalytic cycles. Moreover, the comparison of the Cr content before (8.2 wt%) and after five use (8.1%) as quantified on the ICP/OES showed no leached Cr species. The leaching test (Fig. 10a) and the ICP results (Table 1) confirmed the homogeneity of the Cr_ZrO 2 solid structure. Hence, the mesoporous Cr_ZrO 2 is catalytically stable and reusable with retained activity. ## Substrate scope of mesoporous Cr_ZrO 2 catalyst The chromium doped zirconia exhibited the best catalytic activity in the reduction of citral with 76.4% conversion, Table 3. Hence, the catalytic scope of Cr_ZrO 2 in MPV reduction is extended to some unsaturated aldehydes to form their corresponding unsaturated alcohols Table 4. The Cr_ZrO 2 is most active in the MPV reduction of furfural. The appreciable reactivity suggests that the mesoporous Cr_ZrO 2 acid catalyst is highly active in the MPV process and 100% selective to unsaturated allylic alcohols. ## Discussion Herein, a series of transition metal-doped mesoporous ZrO 2 (M_ZrO 2 , M = Cr, Mn, Fe, and Ni) catalysts were designed for the MPV reduction of aldehydes. Interestingly, the tunability of the surface properties of the resulting catalysts is governed by the metal dopant's nature. The S BET decreases from 206 to 189 and 193 m 2 g 1 upon doping with Cr and Ni, respectively. An improvement in S BET from 206-223 m 2 g 1 was observed when doped with Mn and Fe (221 and 223 m 2 g 1 , respectively). Also, upon doping, there was an enlargement of pore diameter from 2.53-3.63 nm and increased pore volume from 0.10-0.26 cm 3 g 1 (Table 1). The mesostructure properties of the synthesized pure zirconia ZrO 2 and the metal-doped zirconia M_ZrO 2 are typical of type IV hysteresis loops as shown by the BET isotherms (Fig. 2a). This indicates the successful design of a mesopore structure of the materials via a sol-gel approach. This study took advantage of the structure-directing ability of P-123 in the inverse micelles system, which serves as the nanoreactors. Also, a control condensation of the oxo-clusters was achieved by forming NO x species from the nitric acid's thermal decomposition. 34 The increase in the porosity (pore diameter and pore volume) upon doping and S BET in the case of Mn_ZrO 2 and Fe_ZrO 2 could be due to the metal-metal interactions during condensation of the inorganic sols and the condition of the reaction media. A similar phenomenon was explained by Grosso et al., 55 that the mesostructuring occurs during the formation of surfactant-templated inorganic materials by evaporation. The chemical composition of the film governs the meso-organization. Also, it depends on relative vapor pressure in the environment, the evaporation conditions, and the chemical conditions in the initial solution. The changes in the structure and crystallography of ZrO 2 resulting from doping species modified the concentration of active phases involved in the catalyzed reaction on the surface of the ZrO 2 based catalysts. The NH 3 -TPD and CO 2 -TPD experiments showed that the synthesized catalysts possess acid-base properties that could be tuned. The acid/base strength and density are related to the nature of the metal dopant. The acid density decreases in this order Cr_ZrO 2 4 Ni_ZrO 2 4 Mn_ZrO 2 4 Fe_ZrO 2 4 ZrO 2 . On the other hand, the basicity of the M_ZrO 2 gave this trend Mn_ZrO 2 4 Ni_ZrO 2 4 Fe_ZrO 2 4 ZrO 2 4 Cr_ZrO 2 , which could be related to neither the electron density nor ionic charge. These results show that the presence of metal dopant in the ZrO 2 framework possibly tunes the active acid-base sites, resulting mainly from the metal-metal synergy between the metal dopant and the host ZrO 2 . It was revealed that the acid:base in ratio 3 : 1 is required for the chemoselective transfer hydrogenation of citral via the MPV system. To further understand the kinds of acid sites on the solid catalysts, the pyridine-adsorbed experiments indicated that the metal-metal synergy generated both surface Lewis and Brønsted acid sites with a higher concentration of Lewis acid sites having weak, medium, and strong strength. The Lewis acid sites are possibly generated by the concerted metal ions (Cr 3+ and Zr 4+ ) acting as the electron-acceptor. Also, the metal-metal synergy influenced the hydrogen consumption, with the Cr-Zr catalyst showing superior reducibility (H-abstraction capacity). According to Stavale et al., the electronic structure and chemical properties of oxide materials, and their catalytic activities, could be tailored by doping with metal. 56 The approach takes advantage of the metal dopants' tendency to exchange electrons with the host oxide and surface-bound adsorbates. It has also been reported that dopant-modified metal oxides exhibit improved catalytic performance than their pure oxides. 57 In our case, the MPV process is catalytic driven; no activity was observed in the blank reaction. The catalytic activity of the materials in the MPV reduction of a,bunsaturated aldehydes is dependent on the concentration of the surface acidic sites. Among the transition metal dopants incorporated, the catalytic activity of pure ZrO 2 (62.6%) in terms of citral conversion in the model reaction was only improved by Cr_ZrO 2 (76.4%). The observed linear relationship between the surface acidity and activity of the synthesized catalysts suggests that the MPV reduction reaction is perhaps governed by the extent of acidity induced by the electronic interaction between Cr and Zr. The experiment performed with pure chromium oxide showed no activity after 24 h, this suggests that the synergistic interaction might be responsible for the enhanced activity in Cr_ZrO 2 . Also, the acid character of the Cr_ZrO 2 catalyst with the polarity of the citral molecule made it possible for citral to preferably adsorb through the carbonyl group. Hence, the transfer hydrogenation of the carbonyl to produce UAA is favored. A similar scenario in which the adsorption of citral on the Lewis acid site is via the carbonyl was reported. 58 The local structure of the Lewis acid sites on zirconia catalyst was also reported. 59,60 In view of these and the findings of this study, a possible mechanism for MPV reduction of citral on the Lewis acid sites of M_ZrO 2 is proposed (Scheme 2). Distinctively, despite the high acidic properties of the catalytic system, no secondary products were formed. Acidic catalysts frequently favor secondary reactions as either dehydration of alcohol or aldehyde condensation. 61,62 All the prepared catalysts in this work exhibited excellent selectivity to unsaturated allylic alcohol as evidenced in the GC spectra (Fig. S8 and S9, ESI †) compared to their previously reported counterparts in Table 4; this is paramount to a sustainable catalytic process. Moreover, the MPV process in this work was carried out under milder reaction conditions in the absence of additives and gaseous hydrogen. The reactivity retained after five consecutive runs evidence the sustainability of the Cr_ZrO 2 catalyst. Furthermore, the synthesized Cr_ZrO 2 in this work showed considerable reactivity compared to its counterparts in literature Table 5. Our catalyst gave a maximum selectivity of 100% to the UAA under milder reaction conditions in the absence of H 2 gas pressure. For instance, it is more reactive than the ZrSr-PN catalyst in the MPV reduction of cinnamaldehyde; the ZrSr-PN gave 24.0% conversion of cinnamaldehyde in 24 h while our Cr_ZrO 2 gave 60% conversion in 10 h. Also, in the MPV reduction of furfural, our Cr_ZrO 2 showed higher activity of 85% conversion in 4 h and 98.2% in 10 h at 80 1C than P-Zr 200 (55.3%), ME-Zr-200UW (67.6%), and Zr-SBA-15 (54%) after 24, 24 and 6 h, respectively. However, Pt/ZrO 2 synthesized by Wei et al. 4 gave better activity than our catalyst in cinnamaldehyde reduction and Ru/ZrO 2 in citral reduction but, this is due to the H 2 gas pressure used in their catalytic system. ## Conclusion Herein, we have demonstrated that incorporating metal dopants into zirconia's crystal framework alters its physico-chemical properties such as surface area, mesopore structure, crystallinity, basicity, acidity, reducibility, and thermal stability. The reducibility and the strength of the Lewis acid sites govern the activity of ZrO 2 based catalysts in the MPV reduction of citral. Specifically, the Cr dopant weakens the crystallinity of ZrO 2 . However, it improves the reducibility, acidity, and catalytic reactivity for MPV reduction of aldehydes. All the prepared zirconia-based catalysts in this work showed a remarkable selectivity of 100% to UAA. The surfaceinduced performance of the Cr_ZrO 2 is due to the enhanced active centers generated from the synergistic electronic interaction between CrO x and ZrO 2 . Hence, this work unveils that the reactivity of ZrO 2 depends solely on the intrinsic properties of its' surface structure rather than the specific surface expanse. Also, the Cr_ZrO 2 exhibited good stability and recyclability for at least five reaction cycles. We proposed a plausible mechanism of the MPV transformation over the Lewis acidic sites. This work presents the first application of a well-designed mesoporous Cr_ZrO 2 via an inverse micelle approach in the MPV reduction of aldehydes with exceptional reactivity and efficient reusability. The green production of UAA was successfully achieved under mild reaction conditions without pressurized hydrogen gas. The Cr_ZrO 2 is proposed to be a potential sustainable catalyst for industrial applications. ## Conflicts of interest The authors declare no competing interest.

chemsum

{"title": "Metal-doped mesoporous ZrO₂ catalyzed chemoselective synthesis of allylic alcohols from Meerwein\u2013Ponndorf\u2013Verley reduction of \u03b1,\u03b2-unsaturated aldehydes", "journal": "Royal Society of Chemistry (RSC)"}

nanostructured_electrode_enabling_fast_and_fullyreversible_mno_2_-to-mn_2+_conversion_in_mild_buffer

## Abstract: On account of their low-cost, earth abundance, eco-sustainability, and high theoretical charge storage capacity, MnO 2 cathodes have attracted a renewed interest in the development of rechargeable aqueous batteries. However, they currently suffer from limited gravimetric capacities when operating under the preferred mild aqueous conditions, which leads to lower performance as compared to similar devices operating in strongly acidic or basic conditions.Here, we demonstrate how to overcome this limitation by combining a well-defined 3D nanostructured conductive electrode, which ensures an efficient reversible MnO 2 -to-Mn 2+ conversion reaction, with a mild acid buffered electrolyte (pH 5). A reversible gravimetric capacity of 560 mA•h•g -1 (close to the maximal theoretical capacity of 574 mA•h•g -1 estimated from the MnO 2 average oxidation state of 3.86) was obtained over rates ranging from 1 to 10 A•g -1 . The rate capability was also remarkable, demonstrating a capacity retention of 435 mA•h•g -1 at a rate of 110 A•g -1 . These good performances have been attributed to optimal regulation of the mass transport and electronic transfer between the three process actors, i.e. the 3D conductive scaffold, the MnO 2 active material filling it, and the soluble species involved in the reversible conversion reaction. Additionally, the high reversibility and cycling stability of this conversion reaction is demonstrated over 900 cycles with a Coulombic efficiency > 99.4 % at a rate of 44 A•g -1 . Besides these good performances, also demonstrated in a Zn/MnO 2 cell configuration, we discuss the key parameters governing the efficiency of the MnO 2 -to-Mn 2+ conversion. Overall, the present study provides a comprehensive framework for the rational design and optimization of MnO 2 cathodes involved in rechargeable mild aqueous batteries. ## INTRODUCTION The development of rechargeable aqueous batteries for large-scale electrochemical energy storage devices is driven by their ability to achieve appropriate energy densities with safe, sustainable and inexpensive chemicals. 1,2 Among the different aqueous batteries, the rechargeable Zn/MnO 2 battery, able to operate in a mild aqueous electrolyte, is certainly the most attractive, especially since the breakthrough of Pan et al. in 2016, 3 who demonstrated a high reversibility and excellent cycling stability when using a mild acidic electrolyte containing both ZnSO 4 and MnSO 4 . Since then, the concept has been explored in many studies, varying the nature of the MnO 2 cathode and the chemical composition of the Zn 2+ -based aqueous electrolyte. 4 Despite all these efforts, however, the gravimetric capacities of mild aqueous Zn/MnO 2 batteries remain capped to 350-380 mA•h•g -1 , which is far from the maximal theoretical value of 617 mA•h•g -1 associated with the reversible 2-electron reduction of Mn(IV) into Mn(II). Recently, we have revealed that the charge storage mechanism at MnO 2 electrodes is not based on an insertion process but instead on a reversible conversion principle, wherein the weak Brønsted acid AH and conjugated base Apresent in an aqueous buffered electrolyte assist the following proton-coupled electron transfer reaction at the electrode interface: 7 MnO 2(s) + 4 AH + 2 e - Mn 2+ (aq According to this reaction, the 2-electron gravimetric capacity of MnO 2 should be accessible in a mild, non-corrosive buffered aqueous media, which according to the Pourbaix diagram lies within the domain of thermodynamic stability of Mn 2+ . 8 In our previous work, the maximal gravimetric capacity achieved in a buffered electrolyte of pH 5 was 450 mA•h•g -1 (obtained from MnO 2 thin films electrodeposited onto planar electrodes). 7 Although much better than the capacities previously reported for a range of MnO 2 -cathodes in mild unbuffered aqueous electrolytes, this gravimetric capacity remains far from the 570 mA•h•g -1 value recently achieved in a strongly acidic electrolyte (i.e., 0.1 M H 2 SO 4 ). 9 In this case, a similar conversion reaction is at work, but with H 3 O + /H 2 O acting as the proton donor/acceptor couple. Further studies are thus required to better understand the key factors and parameters currently limiting the MnO 2 -to-Mn 2+ conversion in mild aqueous electrolytes. Our previous work leads us to believe that two issues are critical: (i) the efficiency of the electrical wiring between the MnO 2 solid phase and the underlying current collector, and (ii) the appropriate regulation of both the electron transport/transfer at the electrode/MnO 2 film/electrolyte interface and the mass transport of all soluble species involved in reaction 1 (i.e., Mn 2+ , AH, and A -). 10 Indeed, if misbalanced, the Coulombic efficiency (CE) lies below 100 %, inducing a progressive decrease of the electrode gravimetric capacity upon cycling because of the incomplete exploitation of MnO 2 . 7 Concerning the electron transfer between the semi-conductive MnO 2 and the current collector, a few studies have reported on the poor adhesion and mechanical stability of MnO 2 on conductive substrates of various chemical composition and roughness. 11,12 Furthermore, because of the moderate electronic conductivity (10 -3 -10 -4 S•cm -1 ) of MnO 2 , 13,14 the ohmic drop is expected to rise together with the film thickness. All these issues can potentially be solved by using a 3D nanostructured current collector, with a high aspect ratio, well-suited for the conformal electrodeposition of thin films of MnO 2 to provide shorter ion diffusion distances and better electron harvesting paths. 10,15 Such a strategy has previously been successfully exploited to improve the performances of MnO 2 -based supercapacitors, as well as MnO 2 cathodes toward reversible Li + insertion in organic electrolytes. 19 To achieve good conformal electrodeposition of MnO 2 over a 3D porous conductive substrate, the latter must necessarily have a well-opened structure with accessible porosity to avoid any mass transport limitation of the soluble species involved in reaction 1. The approach also requires proper matching of the reactant concentrations and the conversion fluxes, which are linked to both the cycling rate and the specific area developed by the 3D conductive scaffold. In addition, due to the stoichiometry of the conversion process (4 equivalents of proton donor required, see eq 1), strong pH gradients may be generated at the electrode/electrolyte interface, inducing the concomitant precipitation of insoluble phases such as zinc hydroxide sulfates (ZHS), as evidenced in several studies. 3, We believe this point contributes to the limited gravimetric capacities previously observed with unbuffered mild aqueous Zn/MnO 2 batteries. 4 To solve this issue, one strategy is to regulate the pH at the metal oxide/electrolyte interface, using a mild acidic buffered electrolyte at a sufficiently high concentration. 7 In the present study, we demonstrate the decisive advantage of using a nanostructured current collector to fully exploit the reversible MnO 2 -to-Mn 2+ conversion from a buffered aqueous electrolyte containing Mn 2+ . We also show that this conversion reaction remains highly efficient even in a Zn/MnO 2 battery cell configuration, which involves a Zn foil anode paired to the MnO 2 cathode and the presence of Zn 2+ ions, in addition to Mn 2+ , in the buffered electrolyte. For such purpose, we took advantage of 3D nanostructured indium-doped tin oxide (ITO) electrodes (1μm-thick film deposited over a standard flat ITO-coated glass substrate) prepared by glancing angle deposition (GLAD), a technique which allows for the growth of metal oxides nanostructures in different shapes and morphologies. 25 These model mesoporous electrodes are characterized by a reproducible morphology with high aspect ratio and opened porosity, 26 wellsuited for modification by electrodeposition as previously shown with conductive polymers. 27 In addition, their transparency allows for in-situ UV-vis spectroelectrochemical monitoring, providing a real-time quantitative analysis of the amount of MnO 2 that electrodeposits/electrodissolves during the galvanostatic cycles. ## RESULTS AND DISCUSSION The GLAD-ITO electrodes (1 µm-thick film) were prepared according to the published technique 25 using a deposition angle of 80° (see Experimental section) and rapid rotation. This combination leads to growth of vertically-oriented ITO nanopillars on an underlying commercial ITO substrate (Figure 1). The nanopillars are separated from one another with a void spacing in the range of tens of nanometers throughout the entire film thickness, 28 attesting to a well-opened mesoporosity. 29 An electroactive surface area enhancement of 45 was estimated from the capacitive current determined by cyclic voltammetry (CV) at different scan rates (see Supporting Information). The pores of the GLAD-ITO electrodes were filled with MnO 2 by galvanostatic electrodeposition, controlling the amount of MnO 2 through the deposition time (see experimental section). The resulting MnO 2 -modified electrodes were then characterized by SEM, XPS, EDX and UV-vis absorption spectroscopy as detailed in the Supporting Information. Cross-sectional SEM and EDX images reveal that MnO 2 uniformly grows inside the porosity of the GLAD structure and locally forms interwoven nanofibers (Figure 1). The top-view images in Figure 1 show that MnO 2 exhibits a typical carambola-like morphology, similar to that obtained from electrodeposited films on planar ITO electrodes. For the deposited charge of 240 mC•cm -2 , the ITO nanocolumns remains clearly discernible from above (Figure 1), while at higher loads, once the pores of the GLAD ITO are filled, the MnO 2 deposit continues to develop far outside the GLAD-ITO structure, making the ITO nanocolumns no longer discernable. This behavior is well evidenced in Figure 1, where a 0.8 μm-thick MnO 2 film is observable by both SEM and spatially-resolved EDX on top of the GLAD-ITO electrode charged at 400 mC•cm -2 . (An even thicker film, reaching 1.6 µm, is observed for the electrode charged at 800 mC•cm -2 , see Figure S1D). Such a transition between filling of the GLAD-ITO pores with MnO 2 and subsequent growth far beyond the GLAD-ITO structure is also identified in the galvanostatic electrodeposition curves, showing after a certain delay a sudden 20 mV jump in the potential. This potential jump is moreover linearly correlated with the GLAD-ITO thickness as the potential jump shifted from 150 to 90 and then 60 mC•cm -2 when the GLAD-ITO thickness decreased from 1 to 0.6 and then 0.3 µm, respectively (Figure S1A). In order to investigate the benefit of a 3D nanostructured current collector on the reversible conversion of MnO 2 into Mn 2+ , 1 μm-thick GLAD-ITO electrodes were loaded with an intermediate amount of deposited charge of 100 mC•cm -2 (or 48.4 µg•cm -2 ) to ensure the MnO 2 was fully restricted to the interior of the porous structure. The maximal capacity stored in these MnO 2 -modifed electrodes was estimated from the average oxidation state (AOS) of the Mn centers, which in turn was determined from the linear correlation between the amount of electrodeposited Mn centers obtained by ICP quantification and the effective charge passed during the electrodeposition (see Supporting Information and Figure S1B for details). An AOS value of 3.86 was deduced, which is indicative of 1.86 electrons stored per Mn center. This value is in quite good agreement with the AOS of 3.75 determined by XPS analysis of the Mn(3s) peak splitting energy (Figure S1E). We thus conclude a maximal recoverable gravimetric capacity of 574 mAhg -1 for the full conversion of the electrodeposited MnO 2 , and we define the C-rate thanks to this value (1 C corresponding to 0.574 Ag -1 ). MnO 2 -GLAD-ITO electrodes charged at 100 mC•cm -2 were galvanostatically cycled in buffered electrolyte and concomitantly monitored by in-situ UV-vis spectroscopy. The results reported in Figure 2 demonstrate the full exploitation of MnO 2 when working in a 1 M acetate buffer at pH 5, containing 0.1 M MnCl 2 . Indeed, full discharge of the capacity is reported with a near-perfect overlay of the first 20 galvanostatic cycles performed at 11 A•g -1 (19 C-rate), except for the initial discharge which is associated with a slightly lower capacity. Simultaneously, the electrode absorbance oscillates evenly and stably between 0.7 and 0, underlying the excellent reversibility of the MnO 2 electrodeposition/electrodissolution process. The absorbance value recorded at the end of each discharge cycle is close to the absorbance baseline (dashed line in Figure 2B), which demonstrates the complete reductive dissolution of MnO 2 . The small and stable voltage hysteresis (< 0.1 V) observed between the well-defined charge and discharge plateaus at rates lower than 11 A•g -1 is indicative of a rather fast reversible conversion process, occurring under nearly thermodynamic equilibrium (at least at rates < 11 A•g -1 ), as defined by the following Nernst equation: 7 (2) where is the standard potential of the MnO 2 /Mn 2+ redox coupled and the activity of soluble Mn 2+ ions. Hence, based on this equation, the highly stable charge and discharge potentials, leading to well-defined horizontal plateaus, suggest that no significant pH or Mn 2+ gradients develop at the electrode/electrolyte interface within the range of rates (see Supporting Information for details). This mass, equivalent to 145% of the mass electrodeposited during each charging step, indicates that only 40% of the MnO 2 is reversibly exploited. This leads to a significant loss in the gravimetric capacity, reaching only 235 mA•h•g -1 after a few cycles. These observations are in line with what we previously reported at 2D MnO 2 -ITO electrodes using the same electrolyte but with lower amounts of electrodeposited MnO 2 . 7 Another striking difference between the 3D MnO 2 -GLAD-ITO and 2D MnO 2 -ITO electrodes is the shape of the galvanostatic discharge curves, exhibiting one or two plateaus, respectively. At the MnO 2 -GLAD-ITO electrode, the discharge potential centered on a well-defined value of 0.49 V is also remarkably stable until ca. 80% of the MnO 2 is electrodissolved. On contrario, the discharge curves recorded at the 2D MnO 2 -ITO electrode exhibit, after a few cycles, a less welldefined supplementary plateau at a much lower potential of 0.25V, corresponding to roughly 80 % of the total discharge process (Figure 2C). This shift in potential contributes to a huge increase in the voltage hysteresis, thereby severely affecting the charge storage energy efficiency. We previously attributed this second plateau to the formation of a more resistive fraction of MnO 2 at the planar electrode, a fraction which is probably much less electrically connected to the underlying current collector. 7 The lack of a second plateau with the 3D MnO 2 -GLAD-ITO electrode tends to confirm this assumption. It also supports the idea that the 3D nanostructured substrate facilitates the electrical wiring of MnO 2 by shortening the electron transport distances across the semi-conductive MnO 2 , and possibly also by strengthening the interactions between MnO 2 and ITO due to the nanocolumns' roughness visible in SEM images (Figure 1). This result clearly highlights the benefit of the nanostructured electrode to facilitate the reversible electrodissolution/electrodeposition of MnO 2 into/from Mn 2+ , and this benefit should persist as long as the conversion reaction remains within the porosity of the GLAD-ITO electrode. This latter assertion is corroborated by the cycling experiments in Figure S4 carried out with a progressive increase of the deposited charge (i.e., the amount of MnO 2 was increased until it grew far beyond the GLAD-ITO boundary). In the corresponding discharge curves, a second plateau at a lower potential of 0.3 V gradually emerged as the deposited charge increased. The development of this secondary plateau is also accompanied by a progressive and significant decrease in the Coulombic efficiency. The efficiency and rate capability of the conversion process at MnO 2 -GLAD-ITO electrodes was further investigated by cycling electrodes at different rates from 1.4 to 110 A•g -1 (i.e., from (i.e. 38 C). Accordingly, a small fraction of poorly active MnO 2 accumulates at the electrode at the fastest rate, as attested by the nonzero absorbance values A recorded at the end of the discharge steps (Figure 3B). The gravimetric capacities of MnO 2 -GLAD-ITO electrodes cycled at different rates are reported in Figure 3E. The data show that for rates < 11 A•g -1 (i.e., < 19 C), the capacity remains almost constant at 560 ± 10 mA•h•g -1 and close to the maximal value of 574 mA•h•g -1 , while at higher rates, the capacity decreases only slightly, remaining at 435 mA•h•g -1 for 190 C. This remarkable result demonstrates the high rate capability of this conversion process, which is quite attractive for the development of sustainable high-power storage systems. It is worth noting that as the rate increases, the voltage hysteresis also gradually increases from 60 to 340 mV, with a limited ohmic drop contribution (estimated to be 30 ) which only represents 23 % of the total hysteresis at 190 C (Figure 3D). At the higher rates, we also notice an upward and downward drift of potential during the charging and discharging steps, respectively. Based on the Nernst equation (eq. 2), we assume this arises from local pH and/or Mn 2+ gradient changes, driven by mass transport limitations. At the lower rates, while the overall discharge curves overlap, the small increase in the voltage hysteresis results mainly from an increase of the charging potential. This behavior suggests that electrodeposition is intrinsically a slower process than electrodissolution. Overall, these results demonstrate that the combined use of a 3D nanostructured electrode with a mild acidic buffered electrolyte allows for full exploitation of an electrodeposited film of MnO 2 . Such a combination is decisive to provide fast and reversible MnO 2 -to-Mn 2+ conversion, despite the multi-step mechanism, including proton-coupled electron transfer reactions as well as formation/breaking of metal-oxygen bonds. 30 The gravimetric capacity of 560 mA•h•g -1 we report here at pH 5.0 is the highest yet reported in mild acidic electrolytes, 4 and remains competitive with the values recently achieved under much stronger acidic conditions (pH 1). 9 Furthermore, the conversion reaction can be carried out at outstandingly high C-rates, competing with the rate-performances achieved at MnO 2 -based supercapacitors, 16 but with an incomparably greater charge storage capacity. Indeed, as illustrated in the Supporting Information, the present MnO 2 -GLAD-ITO electrode in an unbuffered 1 M KCl electrolyte displayed a capacitance of 240 F•g -1 , which is equivalent to 60 mA•h•g -1 for a potential window of 0.9 V (Figure S2), i.e. 9-times lower than in a buffered electrolyte. The 3D MnO 2 -GLAD-ITO electrode was next assembled in the spectroelectrochemical cell with a Zn foil counter electrode to mimic a Zn/MnO 2 battery. The cell included an Ag/AgCl reference electrode to follow the potential of both MnO 2 and Zn electrodes while simultaneously monitoring the MnO 2 absorbance. Under these conditions, the following overall charge storage conversion reaction is expected: In addition to the acetate buffer (1.5 M, pH 5) and MnCl 2 (0.1 M), 0.25 M ZnCl 2 was added to the electrolyte, as required for the Zn-to-Zn 2+ conversion reaction. The electrochemical performances of the Zn/MnO 2 assembly are given in Figure 4. An initial series of 20 galvanostatic cycles were recorded at an intermediate rate of 7.2 A•g -1 (referred here per gram of MnO 2 ). Well-defined single charge and discharge plateaus were observed at average potential values of 1.65 and 1.50 V vs. Zn 2+ /Zn, respectively, which are very similar to those previously obtained in the Zn 2+ -free electrolyte. The voltage hysteresis slightly increases by 60 mV upon cycling, but without significantly affecting the good energetic efficiency (EE), which remains higher than 86 % over the 20 cycles (blue dots in Figure 4F). Remarkably, after a few preconditioning cycles, the coulombic efficiency remains over 99 %, and from the residual absorbance of the cathode at the end of the discharge steps, we can conclude that only a very small amount of a less active fraction of MnO 2 accumulates, estimated to 6.5 µg•cm -2 ,and mainly arising within the first few cycles. Accordingly, the MnO 2 gravimetric capacity that can be fully exploited here is 500 mAh•g -1 , slightly less than in the absence of Zn 2+ but much higher than has previously been reported for mild aqueous Zn/MnO 2 batteries. 4 Finally, the long-term cyclability of this aqueous Zn/MnO 2 cell configuration was confirmed through the near absence of capacity fading over 400 cycles at 20.6 A•g -1 (i.e., 36 C, Figure 4E), with a Coulombic and an energetic efficiency that rapidly stabilize at 99.7 ± 0.2 % and 82.6 ± 1.3 %, respectively (red dots in Figure 4F). These results clearly demonstrate that the excellent performances of the MnO 2 -GLAD-ITO electrodes are conserved in a Zn/MnO 2 assembly, and that the addition of Zn 2+ in the buffered aqueous electrolyte does not significantly interfere with the reversible MnO 2 -to-Mn 2+ conversion reaction. This thus paves the way to the design of high-performance Zn/MnO 2 batteries in noncorrosive mild aqueous electrolytes. In order to demonstrate the great benefit of using a buffered electrolyte in the aforementioned experiments, the Zn foil/MnO 2 -GLAD-ITO assembly was cycled in an unbuffered aqueous electrolyte (adjusted to pH 5) containing only the inorganic salts required for the conversion reaction (i.e., 0.1 M MnCl 2 and 0.25 M ZnCl 2 along with 0.85 M KCl). The electrochemical performances recorded at 7.2 A•g -1 are reported in Figure 4C. First, it is worth noting that both the Coulombic and the energetic efficiency are significantly deteriorated as compared to those previously obtained in a buffered electrolyte, remaining below 95 % and 79 %, respectively, over the 20 cycles (green dots in Figure 4F). Still, the reversible conversion mechanism is supported by the absorbance change monitored at the cathode, certifying the reversible electrodissolution/electrodeposition of MnO 2 . The absorbance measurements also testify to a significant accumulation of a less active form of MnO 2 , estimated to be 37 µg•cm -2 after 20 cycles (i.e., 76 % of the mass deposited during each charging step), leading to a significant loss in the MnO 2 exploitation and consequently to a lower gravimetric capacity of 310 mA•h•g -1 . It is important to keep in mind that in the absence of acetate buffer, the available proton donors are the hexaaquo complexes [Zn(H 2 O) 6 ] 2+ and [Mn(H 2 O) 6 ] 2+ resulting from the solvation of their parent divalent inorganic salts (here ZnCl 2 and MnCl 2 ). These complexes are characterized by a weak Brønsted acidity with a pK a value of 9.0 and 10.6, respectively. 31 As we have previously demonstrated, 7 both complexes can act as proton donors to assist the electrodissolution of MnO 2 into Mn 2+ according to the following electrochemical reaction: where M is either Zn or Mn. This is typically what we observed in Figure 4C, where, after a few cycles, the galvanostatic discharges lead to a similar areal capacity than in the 1. Mn(OH) 2 ) precipitate, thereby stabilizing the local pH, an effect that is observed through the appearance of a stable plateau at the end of the discharge curves (the beginning of this plateau is indicated by an asterisk on Figure 4C). 24 Such precipitation of insoluble hydroxides over the MnO 2 cathode most likely contributes to a loss in the conversion efficiency, probably by slowing down the ongoing reductive dissolution of MnO 2 . Evidence for the precipitation of such zinc hydroxides has been previously reported in several studies and also related to local pH changes. 3, However, none of these studies explicitly identified the proton source, preventing a clear explanation to the local pH changes and thus the associated precipitation of zinc hydroxides, both governed by the pK a s and local activities of the weak acid and conjugated base acting as proton donor/acceptor. Concerning the charging curves, which involve MnO 2 electrodeposition with the concomitant local release of several equivalents of protons, the first plateau observed at 1.45 V (Figure 4C) suggests a stabilization of the local pH at an intermediate mild acidic value. This effect can be rationalized from the dynamic equilibrium that is expected to persist through the continuous neutralization of the protons released by the Brønsted bases which are generated during the previous discharge step (and which include the precipitated hydroxides). This interpretation is further supported by the absence of a similar plateau when the galvanostatic charge is immediately applied to a MnO 2 -GLAD-ITO electrode without prior discharge (orange curve, Figure 4C), leading to a direct rise of the charging potential up to a plateau at 2 V vs. Zn 2+ /Zn, which is close to the charging potential previously reported for an acidic Zn/MnO 2 battery (i.e., 2.2 V in the presence of a 0.1 M H 2 SO 4 electrolyte). 9 This result is thus indicative of the strong pH gradients that develop during the electrodeposition of MnO 2 from an unbuffered electrolyte. Besides the demonstration of the beneficial effect of buffered electrolytes on the stabilization of the potentials of MnO 2 electrodeposition and electrodissolution, the present work also provides a clear explanation of the good performances recently achieved for a Zn/MnO 2 battery using an unbuffered aqueous electrolyte containing zinc and manganese acetate salts. 32 While the authors focus on the coordinating role of the acetate ions to explain the high performances they obtained and the low MnO 2 electrodeposition potential they observed (1.8 V vs. Zn 2+ /Zn), the present results rather suggest that the increased performance is due to a local buffering effect. Indeed, the protons locally released during the electrodeposition are neutralized by the acetate ions present in the electrolyte, generating acetic acid in dynamic equilibrium with the remaining acetate ions. In the following discharge step, this acetic acid thus behaves as the proton donor to efficiently assist the reductive electrodissolution of MnO 2 . It is worth noting that significant shifts of the charging and discharging potentials were reported in such unbuffered aqueous electrolytes, which most likely arises from significant variations in the local concentrations of acetate and acetic acid, and thus in the local pH. ## CONCLUSION In the present work, we demonstrate the beneficial combination of a nanostructured conductive 3D substrate and a buffered electrolyte to fully exploit the reversible, two-electron MnO 2 -to-Mn 2+ conversion mechanism in mild acidic conditions. The corresponding cathode is characterized by a high gravimetric capacity (560 mA•h•g -1 ), high rate capability (435 mA•h•g -1 at 190 C), low charge/discharge hysteresis (thus high energetic efficiency) and high cyclability (up to 900 cycles with a CE > 98%). Such performances are preserved in a Zn/MnO 2 cell configuration (i.e., 500 mA•h•g -1 at 12 C and a stable Coulombic efficiency of 99.7 % over 400 cycles at 36 C), outperforming all the mild aqueous Zn/MnO 2 assemblies so far described, with the additional advantage of a highly stable potential of 1.5 V over almost the complete discharge. These great performances arise from properly balancing the electrolyte composition (to avoid mass transport limitation) and the MnO 2 mass loading regarding both the surface enhancement and the porosity of the nanostructured conductive 3D substrate (to avoid long-range electron transport throughout the semiconductive MnO 2 film and thus ohmic drop issues). The present demonstration was performed using transparent model GLAD-ITO electrodes, allowing for insitu monitoring of the conversion process. These model electrodes, however, are unsuitable for charge storage applications, and therefore, further developments require scaling-up the conversion process with inexpensive, high surface area conductive substrates to get beyond the low gravimetric capacities of mild aqueous Zn/MnO 2 batteries. ## EXPERIMENTAL SECTION Chemicals. HNO 3 (Suprapur, 65%), acetic acid (Reagent plus, > 99%), KOH, HCl (Normapur, 37%), KCl (GR for analysis), ethanol absolute (EMSURE) and ZnCl 2 were purchased from Sigma-Aldrich/Merck. MnCl 2 tetrahydrate (99%) was purchased from Alfa Aesar. Acetone (Normapur) and chloroform (Normapur) were purchased from VWR Chemicals. ## GLAD-ITO Mesoporous Electrodes. Porous ITO thin films were prepared by the glancing angle deposition (GLAD) method followed by thermal treatment as previously described. 25 Briefly, nanostructured ITO films were deposited from ITO evaporant (Cerac, 91:9 In 2 O 3 /SnO 2 99.99% pure) in an electron-beam physical vapor deposition system (Axxis, Kurt J Lesker) on ITO-coated glass substrates (8-12 Ω/, Delta Technologies Ltd.). Throughout the deposition, substrates were maintained at an 80° angle with respect to impinging evaporant flux, while constantly rotating as a feedback-controlled function of the deposition rate. The film thickness was adjusted between 0.3 and 1 µm. Following deposition, the GLAD-ITO samples were thermally annealed in a two stage process, first under air at 500 °C and subsequently under 5% H 2 /Ar flow at 375 °C, to improve and stabilize the optical and electrical properties. For such deposition conditions, the film porosity was previously estimated to be 0.5 and its density to be 4 g•cm -3 . 25 Prior to the electrochemical experiments, the GLAD-ITO electrodes were cleaned by soaking them successively in chloroform, acetone, and ethanol, each time for 30 min at 50°C. After the electrodes were left to dry, a geometric area of 0.50 ± 0.04 cm 2 (N = 40) was delimited by depositing an insulating layer of nail polish. All gravimetric intensities (A•g -1 ) were calculated from the current density (mA•cm -2 ) applied to the electrode and the expected electrodeposited mass of MnO 2 , deduced from eq. 4. The Coulombic efficiency , energetic efficiency ( ), and gravimetric capacity ( in mA•h•g -1 ) ## Preparation of the were calculated using the following equations: where Q i,ch and Q i,disch are the areal charging and discharging capacities of the i-th cycle in mC•cm -2 , E i,ch and E i,disch the corresponding charging and discharging energy densities in W•h•cm -2 determined from the product of Q and the average charge/discharge voltages, m ch the areal MnO 2 mass deposited during the charging step (i.e., 48.4 µg•cm -2 for a constant charge of 100 mC•cm -2 ), and m acc the total areal inactive MnO 2 mass accumulated at the end of the experiment (the latter being determined either by ICP analysis or from the absorbance of the electrode after discharge). Film characterization. Scanning electron microscopy and energy dispersive x-ray spectroscopy were performed on a Hitachi S5500. XPS spectra were recorded using a K-Alpha+ system (ThermoFisher Scientific, East-Grinsted, UK) fitted with a micro-focused and monochromatic Al Kα X-ray source (1486.6 eV, spot size of 400 µm). The pass energy was set to 150 and 40 eV for the survey and the narrow high resolution regions, respectively. The spectra were calibrated against the (C-C/C-H) C(1s) component set at 285 eV. The chemical composition was determined using the manufacturer's sensitivity factors within the Avantage software (version 5.977). The average oxidation state (AOS) of the Mn centers in the electrodeposited MnO 2 thin-film was estimated on the basis of the XPS Mn(3s) peak splitting energy (ΔBE) and using the following correlation: 33,34 The amount of MnO 2 electrodeposited on ITO was analyzed by inductively coupled plasma atomic emission spectrometry (ICP-AES; Thermo Scientific iCAP 6300 ICP spectrometer) after dissolution of the MnO 2 film in concentrated nitric acid under ultrasonication and then dilution with purified water to have a final 6.5% v/v nitric acid concentration. electrodeposited MnO 2 ( , Figure S1C). The modeling of the experimental data, arising from 20 independent electrodes, gives us the following linear relationship: (S2) The obtained gravimetric extinction coefficient is weaker than that previously determined from thinner MnO 2 films (18.6 × 10 -3 cm 2 µg -1 with < 20 µg.cm -2 ) on 2D ITO, S3 probably because the linear approximation tends to lose its validity for high surface concentrations. Nevertheless, eq. S2allows a rough estimate of the low surface concentrations of MnO 2 on the GLAD-ITO surface, which is typically observed at the end of a galvanostatic discharge. Thus, we are able to calculate the gravimetric capacity of the MnO 2 film at the end of a galvanostatic cycling experiment (see experimental section eq. 5). ## III. Electrochemical characterization of the MnO 2 -GLAD-ITO electrode in a pure KCl electrolyte The double-layer electrical capacitance of a MnO 2 -GLAD-ITO electrode loaded at 100 mC•cm -2 was investigated in a 1 M KCl electrolyte adjusted to pH 5.0. Under these conditions, the electrode exhibits the typical features characterizing the reversible charging/discharging of an electrical double-layer, showing a linear time-dependence of charge and discharge curves with the potential (Figure S2A). The high stability absorbance switching (between 0.76 and 1.35) together with the high Coulombic efficiency of 99.1 ± 0.2 % over 400 cycles also agrees with the charging/discharging behaviour of a true capacitance (Figure S2B). From these data, a specific gravimetric capacitance of C f = 240 ± 5 Fg -1 is obtained (Figure S2C), which lies in the range of capacitances commonly reported for electrodeposited MnO 2 thin-films. S8-S12

chemsum

{"title": "Nanostructured electrode enabling fast and fullyreversible MnO 2 -to-Mn 2+ conversion in mild buffered aqueous electrolytes", "journal": "ChemRxiv"}

discovery_of_benzyl_tetraphosphonate_derivative_as_inhibitor_of_human_factor_xia

## Abstract: The inhibition of factor XIa (FXIa) is a trending paradigm for the development of new generations of anticoagulants without a substantial risk of bleeding. In this report, we present the discovery of a benzyl tetra-phosphonate derivative as a potent and selective inhibitor of human FXIa. Biochemical screening of four phosphonate/phosphate derivatives has led to the identification of the molecule that inhibited human FXIa with an IC 50 value of ~7.4 μM and a submaximal efficacy of ~68 %. The inhibitor was at least 14-fold more selective to FXIa over thrombin, factor IXa, factor Xa, and factor XIIIa. It also inhibited FXIa-mediated activation of factor IX and prolonged the activated partial thromboplastin time of human plasma. In Michaelis-Menten kinetics experiment, inhibitor 1 reduced the V MAX of FXIa hydrolysis of a chromogenic substrate without significantly affecting its K M suggesting an allosteric mechanism of inhibition. The inhibitor also disrupted the formation of FXIa -antithrombin complex and inhibited thrombin-mediated and factor XIIa-mediated formation of FXIa from its zymogen factor XI. Inhibitor 1 has been proposed to bind to or near the heparin/polyphosphate-binding site in the catalytic domain of FXIa. Overall, inhibitor 1 is the first benzyl tetraphosphonate small molecule that allosterically inhibits human FXIa, blocks its physiological function, and prevents its zymogen activation by other clotting factors under in vitro conditions. Thus, we put forward benzyl tetra-phosphonate 1 as a novel lead inhibitor of human FXIa to guide future efforts in the development of allosteric anticoagulants. ## Introduction Thrombosis is a condition in which the blood unnecessarily and/or excessively clots in blood vessels and/or heart chambers leading to life-threatening pathologies. Thrombosis can initiate in a vein-driven manner as it is in deep vein thrombosis and pulmonary embolism. It can also be of arterial origin as in ischemic heart disease and stroke. Thrombosis has also been linked to a host of other chronic and serious diseases including inflammation, cancer, neurodegenerative diseases, and microbial infections. Importantly, micro-and macro-vascular as well as venous and arterial thrombotic conditions have been implicated in the ongoing pandemic of coronavirus disease of 2019 (COVID-19). In fact, COVID-19-associated thrombotic events appear to often lead to poor clinical outcomes of hospitalization, ICU admission and mechanical ventilation, and death. In this arena, important components that variably contribute to thromboembolic diseases are the platelets and the procoagulant factors of the coagulation system. Therefore, drugs that are clinically used to treat thrombosis either target platelet proteins i. e. antiplatelets or inhibit the procoagulant factors of the coagulation system i. e. anticoagulants. Considering the origin of the pathological clots, antiplatelets are generally used in arterial thrombosis whereas anticoagulants are more frequently used in treating and/or preventing venous thromboembolism. Combinations of the two classes of antithrombotics are also used. On the anticoagulants front, clinically available anticoagulants include the indirect anticoagulants of warfarin and heparins as well as the direct anticoagulants of thrombin inhibitors (argatroban, dabigatran, and bivalirudin) and factor Xa inhibitors (rivaroxaban, apixaban, edoxaban, and betrixaban). Despite their structural diversity, all available anticoagulants inhibit thrombin and/or factor Xa (FXa), two serine proteases in the common coagulation pathway (Figure 1). This leads to a very efficient inhibition of the pathological formation of the blood clot, yet it also disrupts hemostasis. Thus, all currently used anticoagulants are associated with serious bleeding events which complicate their effective and safe use in several patient populations such as those with atrial fibrillation or chronic kidney diseases. Therefore, the development of anticoagulants that do not cause bleeding is the main goal of contemporary drug discovery programs in the field. In this direction, several other procoagulant factors including factors VIIa (FVIIa), IXa (FIXa), XIa (FXIa), 33] XIIa (FXIIa), and XIIIa (FXIIIa) have been considered to design and develop new effective anticoagulants with limited-to-none bleeding risk. In particular, FXIa appears to be gaining momentum owing to several epidemiological, animal, and human observations which collectively indicate that FXIa activity contributes to thrombosis but not to hemostasis. 33] In fact, given the promise of FXI(a) as a drug target for safer anticoagulants, several FXI(a)-targeting agents are under development and these include small molecules, monoclonal antibodies, antisense oligonucleotides, and aptamers. Many of those inhibitors are active site inhibitors and few are allosteric inhibitors. Biochemically, FXIa is a serine protease homodimer that belongs to the intrinsic pathway of coagulation (Figure 1). Physiologically, FXIa activates factor IX to FIXa so as to eventually amplify thrombin generation. Importantly, the zymogen form of FXIa i. e. factor XI (FXI) is activated by thrombin or FXIIa in the presence of negatively charged macromolecules such as heparin, inorganic polyphosphates, and dextran sulfate via a template-mediated mechanism. The negatively charged macromolecules binds to anion-binding sites on the activating enzymes as well as on FXI. Interestingly, while the negatively charged heparin facilitates the activation of the zymogen FXI, heparin also directly and allosterically inhibits the active enzyme FXIa. Furthermore, the inorganic polyphosphates have also been found to bind to the same anion binding sites of FXI and acts as cofactors for its autoactivation and for its activation by thrombin and FXIIa. Accordingly, to identify a new line of FXI(a)-targeting anticoagulants, we have considered the aromatic mimetics of inorganic polyphosphates so as to allosterically inhibit the function of human FXIa by targeting its anion-binding sites. In this arena, we tested four phosphonate/phosphate derivatives (1-4) (Figure 2) to evaluate their potential to inhibit human FXIa. As a result, we have identified the benzyl tetraphosphonate derivative (1) as the first aromatic mimetic of inorganic polyphosphates that allosterically inhibits human FXIa, as determined in the corresponding in vitro experiments of chromogenic substrate hydrolysis assay and Michaelis-Menten kinetics. The benzyl tetraphosphonate 1 inhibited human FXIa with an IC 50 value of ~7.4 μM and a submaximal efficacy of ~68 %. The molecule has demonstrated at least 14-fold selectivity toward FXIa over other procoagulant factors of thrombin, FIXa, FXa, and FXIIIa. The inhibitor also selectively prolonged the activated partial thromboplastin time (APTT) of human plasma. Interestingly, inhibitor 1 concentration-dependently inhibited the physiological function of FXIa i. e. FIX activation and inhibited thrombin-mediated and FXIIa-mediated activation of FXI. The inhibitor disrupted the formation of FXIaantithrombin complex in the presence of heparin, suggesting that it may compete with heparin for binding to or near the anion-binding site(s) of FXIa. Overall, we put forward benzyl tetraphosphonate derivative 1, the first potent, selective, and partial allosteric inhibitor of FXIa, to be considered in the development of effective anticoagulants with a limited risk of bleeding complications. Figure 1. The coagulation process is depicted. Among the most important factors are FVIIa of the extrinsic coagulation pathway, FXIIa/FXIa/FIXa of the intrinsic/contact activation pathway, and thrombin and FXa of the common coagulation pathway. Thrombin and FXa are the molecular targets of all currently available anticoagulants. FXIa is the molecular target in this study. Targeting human FXIa is expected to yield effective anticoagulants without the risk of bleeding because FXIa contributes to thrombosis, but not hemostasis. Physiologically, FXIa activates FIX to FIXa which subsequently forms the intrinsic tenase complex that further activates FXa, and then, thrombin. FXIa is produced by inorganic polyphosphate-mediated activation of FXI via the action of thrombin and FXIIa. FXIa can also be produced by autoactivation. ## Figure 2. Chemical structures of phosphonate and phosphate derivatives (1-4) that were screened against human FXIa in this study. ## Rationale for Screening Phosphonate/Phosphate Derivatives (1-4) Against Human FXIa Several approaches have been utilized to discover and/or rationally design inhibitors of FXIa. These approaches include small molecules, polypeptides, aptamers, and monoclonal antibodies. While most of the small molecules are active site inhibitors, sulfated nonsaccharide mimetics of heparin, reported as SPGG and SCI, are allosteric inhibitors of FXIa. They appeared to inhibit FXIa by targeting its anion-binding sites, particularly the site in the catalytic domain. Interestingly, SCI exhibited potent anticoagulant activity with no bleeding complications in rat models of thrombosis. Although their allosteric inhibition mechanism is unique for achieving a high level of functional selectivity, these molecules are highly negatively charged with at least 10 sulfate groups, a structural feature that may compromise their druggability. Thus, we have considered phosphonate and phosphate derivatives 1-4 (Figure 2) to be screened against human FXIa to identify new inhibitors of FXIa with fewer number of negative charges. Given the fact that anion-binding sites of FXI(a) recognize both sulfated heparins as well as inorganic polyphosphates, we have hypothesized that phosphonate and phosphate derivatives will likely exhibit an allosteric inhibition mechanism similar to that exhibited by the sulfated nonsaccharide mimetics of heparin. We have also hypothesized that the phosphonate and phosphate derivatives will likely enjoy a better long-term stability because of the reduced number of negative charges. Not only that, but strategies to develop phosphonate and phosphate prodrugs are also well established, which is necessary to be considered to enhance their overall druggability, especially as it relates to their oral bioavailability. ## FXIa Inhibition Potential of Phosphonate/Phosphate Derivatives (1-4) The four molecules were evaluated for their potential to inhibit FXIa hydrolysis of S-2366, a chromogenic tripeptide substrate, under the physiological conditions of pH 7.4 and 37 °C, as reported earlier. Only the tetraphosphonate derivative 1 and phosphate derivative 4 inhibited FXIa in a dose-dependent fashion. Molecules 2 and 3 did not inhibit FXIa at the highest concentration tested of 100 μM. The inhibition of FXIa by molecules 1 and 4 could be fitted using the logistic equation 1, which resulted in an IC 50 value of 7.4 � 0.9 μM and efficacy of 68.0 � 3.7 % for inhibitor 1 (Figure 3, Table 1) and an IC 50 value of 59.4 � 17.5 μM and efficacy of 111 � 15.6 % for inhibitor 4 (Table 1). The lack of inhibition potential for phosphate derivatives 2 and 3 suggests a rather selective interaction between FXIa and inhibitors 1 and 4. Of note, molecules 2-4 also have two sulfonate groups in addition to the phosphate group, yet only molecule 4 inhibited human FXIa indicating that negative charges are not the only interacting groups and that other structural features are also important. ## Benzyl Tetraphosphonate 1 is a Selective Inhibitor of Human FXIa Over Other Coagulation Proteins The inhibition profiles of benzyl tetraphosphonate 1 against thrombin, FIXa, and FXa were studied using the corresponding chromogenic substrate hydrolysis assays under physiological conditions, as described earlier. In these assays, the inhibition potential was determined by spectrophotometric measurement of the residual protease activity in the presence of varying concentrations of inhibitor 1 (Figure 3). Furthermore, the molecule's activity against human FXIIIa was also studied using the bi-substrate, fluorescence-based trans-glutamination assay, as described earlier. Based on the highest concentration tested of inhibitor 1 against the above enzymes, the calculated IC 50 values are estimated to be > 500 μM for thrombin, > 200 μM for FIXa, > 100 μM for FXa, and > 200 μM for FXIIIa (Table 2), suggesting selectivity indices of > 67.6-fold, 27-fold, > 14-fold, and > 27-fold, respectively. Overall, the above results indicate that benzyl tetraphosphonate 1 is a selective inhibitor for human FXIa, as determined in the corresponding in vitro assays. ## Effect of Inhibitor 1 on Clotting Times of Normal and FXI Deficient Human Plasma Plasma clotting assays of APTT and prothrombin time (PT) are routinely used to investigate the anticoagulation potential of new procoagulant enzyme inhibitors under in vitro conditions. The former time measures the effect of potential anticoagulant on the intrinsic/contact pathway-driven clotting which involves FXIIa, FXIa, and FIXa. The latter time measures the effect of potential anticoagulant on the extrinsic pathway of coagulation which involves FVIIa. The effect of different concentrations of inhibitor 1 on APTT and PT of normal human plasma was measured (Table 3 and Figure 4), as described in earlier studies. Results indicated that inhibitor 1 concentrationdependently prolonged APTT but not PT over the concentration range of 0-470 μM. Figure 4A shows the effect of anti-FXIa monoclonal antibody on the clotting times. The antibody selectively recognizes human FXI, and under our testing conditions, resulted in 1.5-fold increase in APTT at a concentration of 1.4 μg/mL. The antibody did not affect the PT at the highest concentration tested of 2.88 μg/mL. Likewise, Figure 4B shows the variation in APTT and PT in the presence of varying concentrations of inhibitor 1. A 1.5-fold increase in APTT required 311.3 μM of inhibitor 1. However, a 1.5-fold increase in the PT required > 750 μM of inhibitor 1. These results indicate, as expected, that inhibitor 1 is anticoagulant in normal human � 0.9 [b] > 500 [c] > 200 > 100 > 200 [a] The inhibition values were obtained following non-linear regression analysis of direct inhibition of human thrombin, factor Xa, and factor XIIIa in appropriate Tris HCl buffers of pH 7. plasma and it does so by targeting proteins in the intrinsic pathway of coagulation, particularly FXIa. To confirm the involvement of FXIa in streaming the anticoagulant effect of inhibitor 1, we measured the effect of adding variable concentrations of inhibitor 1 on FXIa-induced clotting of FXIdeficient human plasma. Figure 4C shows that FXI-deficient human plasma clotted at 136.9 s, yet adding 2.4 nM or 4.8 nM of human FXIa accelerated its clotting so as to take place at 56.9 s or 42.4 s, respectively. However, the addition of 233-467 μM of inhibitor 1 significantly delayed the FXIa-induced clotting of FXI-deficient human plasma by 1.04-1.94-fold when clotting was induced by 2.4 nM FXIa or by 1.11-2.72-fold when clotting was induced by 4.8 nM FXIa. Overall, these results further establish that the anticoagulant activity of inhibitor 1 in human plasma is attributed to its effect on FXIa of the intrinsic coagulation pathway. ## Allosteric Inhibition of FXIa by Benzyl Tetraphosphonate 1 To understand the mechanistic basis of molecule 1 inhibition of human FXIa, Michaelis-Menten kinetics of S-2366 hydrolysis by the wild type full-length FXIa was performed in the presence of inhibitor 1 at pH 7.4 and 37 °C. Figure 5 shows the initial rate profiles in the presence of inhibitor 1 (0-100 μM). Each profile displays a characteristic rectangular hyperbolic trend, which could be fitted using equation 2 to give the apparent K M and V MAX (Table 4). The K M for S-2366 did not significantly change (0.41 � 0.04 mM-0.28 � 0.09 mM) in the presence or absence of inhibitor 1. Nevertheless, the V MAX decreased steadily from 116 � 4 mAU/min in the absence of inhibitor 1 to 45.7 � 4.2 mAU/min at 100 μM of inhibitor 1. Thus, the inhibitor appears to bring about structural changes in the active site of FXIa which do not affect the formation of Michaelis complex but lead to a significant disruption in FXIa catalytic activity. This indicates that molecule 1 is an allosteric inhibitor of human FXIa. ## Inhibition of FXIa: Activation of the Physiologically Relevant Substrate FIX by Tetra-Phosphonate 1 Although tetraphosphonate 1 inhibited the hydrolysis of chromogenic tripeptide substrate S-2366 by FXIa, yet we have aimed at establishing its physiological relevance by evaluating its effect on the physiological substrate of FXIa i. e. FIX. During coagulation, FXIa binds to and activates FIX (Figure 1) by successively cleaving two peptide bonds of Arg145-Ala146 and Arg180-Val181 so as to eventually generate FIXaβ. Subsequently, FIXaβ along with factor VIIIa forms the intrinsic tenase complex, in the presence of calcium ions and phospholipids, to activate factor X to FXa which eventually amplifies thrombin generation. To establish the physiological relevance of benzyl tetraphosphonate 1 inhibitory activity toward FXIa, we evaluated FXIa activation of FIX in the presence and absence of inhibitor 1 using SDS-PAGE (Figure 6A) and Western blotting (Figure 6B). The figures show that inhibitor 1 dose-dependently (0-1000 μM) inhibited the formation of the intermediate FIX i. e. FXIα as well as the fully activated FIX i. e. FIXaβ (heavy chain (FXIaβ-HC) and light chain (FXIaβ-LC)) as indicated by the absence of the corresponding bands. These results suggest that the inhibitory activity of tetraphosphonate 1 toward FXIa is physiologically relevant and it takes place at a concentration range (� 250 μM) similar to that used in the chromogenic substrate hydrolysis assays. ## Effect of Benzyl Tetraphosphonate 1 on FXI(a) Interactions with Macromolecules To understand the effect of benzyl tetraphosphonate 1 beyond the inhibition of FXIa catalytic activity, we investigated three intermolecular interactions involving FXIa and its zymogen FXI. In this direction, it has been reported that FXIa can be inhibited by antithrombin, an endogenous serpin, in a reaction that is accelerated in the presence of heparin by a template-or bridging-based mechanism. In this interaction, a denaturation-resistant complex between the FXIa active site and the reactive center loop of antithrombin is formed. In this complex, the light chain of FXIa (FXIa-LC) acylates antithrombin by forming a covalent bond. Figure 7A shows SDS-PAGE evidence of the 90 kDa FXIa-LC -antithrombin complex (lane 3) formation in the absence of inhibitor 1. In contrast, the presence of increasing concentrations of inhibitor 1 (50-1000 μM) inhibited the formation of such complex. In fact, the formation of the complex appears to be significantly diminished at the highest concentration tested of 1000 μM. At concentrations of 250 μM and 1000 μM, the band of 90 kDa complex on SDS-PAGE significantly diminished and that of FXIa-LC appeared again at 30 kDa (lanes 5 and 6). These results suggest that inhibitor 1 disrupts the formation of FXIa-antithrombin complex, potentially by competing with heparin for its anionbinding sites on the catalytic domain and/or apple 3 domain. Furthermore, it is well documented that the zymogen FXI is activated by the action of thrombin or FXIIa in processes that are accelerated in the presence of inorganic polyphosphates or dextran sulfate. Specifically, dextran sulfate is a sulfated polysaccharide that appears to accelerate the activation reactions via template-dependent mechanism in the which the polysaccharide engages with the anion binding sites on both the substrate i. e. FXI as well as the activators i. e. thrombin or FXIIa. Figure 7B shows that inhibitor 1 inhibited the activation of FXI by thrombin, as shown by the decreased intensity of FXIa-HC and FXIa-LC bands (lanes 6-10), over the concentration range of 2-1000 μM. Likewise, Figure 7C shows that inhibitor 1 inhibited the activation of FXI by FXIIa, as shown by the decreased intensity of FXIa-HC and FXIa-LC bands (lanes 7-9), over the concentration range of 50-1000 μM. Taken together, benzyl tetraphosphonate 1 has been found to inhibit the dextran sulfate-mediated activation of FXI by thrombin as well as by FXIIa at comparable concentrations to those determined in the above experiments. This suggests that inhibitor 1 potentially competes with dextran sulfate for anionbinding sites on the catalytic domain and/or apple 3 domain. Overall, the above results suggest that benzyl tetraphosphonate 1 is likely to bind to an allosteric site on FXI(a) rather than the active site. Similar to sulfated non-saccharide mimetics of heparins, the binding site is likely to be the anion-binding site in the catalytic domain. ## Molecular Modeling of Potential Binding of Benzyl Tetraphosphonate 1 to the Anion -Binding Site in the Catalytic Domain of FXIa To identify a plausible binding mode for benzyl tetraphosphonate 1, we performed molecular docking studies by considering the anion-binding site on the catalytic domain of FXIa. The rationale for considering this site is that its lysine and arginine residues have been implicated in FXI(a) interactions with the negatively charged functional groups of several macromolecules including the sulfate groups of heparins and the phosphate groups of inorganic polyphosphates. The anion binding site on the catalytic domain of FXIa has also been implicated in the action of SPGG and that of SCI, two small molecule heparin mimetics that act as allosteric inhibitors of human FXIa. The docking studies of benzyl tetraphosphonate 1 onto the anion-binding site were carried out using Glide, as described in the experimental part. Initial coordinates for FXIa catalytic domain were taken from the crystal structure of PDB ID: 2FDA. The studies revealed that one phosphonate group on one end of the inhibitor structure interacts via electrostatic/hydrogen bond with K529 and T523 residues (Figure 8). The studies also revealed that the central urea carbonyl oxygen interacts via hydrogen bond with N524 residue and that another phosphonate group on the other end of inhibitor 1 interacts with K535 residue (Figure 8). In particular, the K529 residue (chymotrypsin number is K170) and the K535 (chymotrypsin number is K175) have been implicated in heparin-mediated inhibition of FXIa by antithrombin as well as in polyphosphate-mediated activation of FXI. Impor- tantly, these results highlight the essential requirement of at least two phosphonate groups for the inhibitor action toward FXIa. Although these results are to be experimentally confirmed via crystallography studies and/or mutagenesis studies, how-ever, the above computational exercise is important as the identified putative binding site can be used to guide subsequent efforts to optimize the potency and selectivity of inhibitor 1. ## Discussion Oral as well as parenteral anticoagulants are widely used to prevent and/or treat thromboembolic diseases. In addition to having potent and selective pharmacodynamic profile, ideal anticoagulants should have predictable pharmacokinetics and do not require continuous monitoring and/or dose adjustment. They should also be rapidly reversed via the use of affordable and effective antidotes. In contrast to the existing ones, ideal anticoagulants should be devoid of hepatotoxicity, osteoporosis, or thrombocytopenia. They should also be safe for use in compromised patients with a high risk of thrombosis as in pregnant patients and cancer patients. Importantly, they should induce no bleeding complications. While a significant progress has been made towards satisfying the above criteria, bleeding risk continues to be a significant side effect of heparins and warfarin. Newer direct anticoagulants inhibiting thrombin or FXa are increasingly replacing heparins and warfarin, yet they appear to be still associated with significant risks of major bleeding. Accordingly, patients who may benefit from anticoagulation therapy do not receive it or receive a lower dose which is often not effective. This is the case of patients of atrial fibrillation and chronic kidney diseases. Thus, the search for safer anticoagulant drugs with little-to-none bleeding risk continues. Towards this goal, several FXI(a)-targeting agents are in development including small molecules (BMS-986177 and EP-7041), monoclonal antibodies (AB023, MAA868, and Osocimab), aptamers (FELIAP), and sulfated glycosaminoglycan mimetics (SPGG and SCI). Osocimab and BMS-986177, in particular, are currently in advanced clinical trials. Among the above molecules, SPGG and SCI were developed as small molecule allosteric inhibitors of FXIa. They have been projected to target the anion-binding site in the catalytic domain. In fact, their structural features were designed in relevance to heparin, a heterogenous mixture of sulfated glycosaminoglycans that is known to allosterically inhibit FXIa by targeting the anion-binding sites in the catalytic domain and the apple 3 domain. Inorganic polyphosphates are also known to engage with FXI(a) anion-binding sites. However, the inorganic polyphosphates facilitate the autoactivation of the zymogen FXI as well as its activation in the presence of thrombin or FXIIa, and thus, leads to a procoagulant outcome. Inspired by the anionic nature of heparin and inorganic polyphosphates, several aptamers have also been designed to potentially disrupt the inorganic polyphosphates' procoagulant role as well as the procoagulant properties of FXI(a). These include aptamers 12.7 and 11.16 (RNA aptamers which bind to FXI(a) anion binding site on the catalytic domain resulting in allosteric inhibition) and FELIAP (DNA aptamer which binds to or near active site of FXIa). In this study, we report on a small molecule (~1180.74 Da), benzyl tetraphosphonate 1, that directly inhibits the catalytic activity of FXIa with an IC 50 value of ~7.4 μM (K i value is estimated to be ~7.4 μM, given the testing conditions), as determined in the corresponding chromogenic substrate hydrolysis assay. Importantly, the in vitro inhibition of FXIamediated hydrolysis of the chromogenic tripeptide substrate i. e. S-2366 has been found to translate into an in vitro inhibition of FXIa-mediated hydrolysis of its physiological substrate i. e. FIX. Thus, the activity of inhibitor 1 appears to be physiologically relevant. Furthermore, the inhibitor has demonstrated significant selectivity over other procoagulant serine proteases including thrombin, FIXa, and FXa as well as over the transglutaminase FXIIIa, as determined by the corresponding in vitro enzyme assays. Inhibitor 1 has also demonstrated a substantial anticoagulant effect by targeting FXIa in human plasma as demonstrated by its relatively selective effect on APTT as well as by its ability to abolish the FXIa-induced clotting of FXIdeficient human plasma. Despite the direct effect of the inhibitor, its biological effects beyond the direct inhibition of FXIa activity are very interesting. In this arena, the molecule has been found to inhibit the dextran sulfate-mediated activation of the zymogen FXI by thrombin and FXIIa at concentrations comparable to those that inhibit FXIa hydrolysis of S-2366 as well as FIX. This has suggested that the benzyl tetra-phosphonate 1 may bind to a cluster of basic amino acids presented by both the enzyme FXIa and its zymogen FXI. This cluster is likely to be the anion binding sites, particularly the one in the catalytic domain. In fact, the ability of the inhibitor to disrupt FXIa-antithrombin complex formation in the presence of heparin lends support to the projection of the inhibitor targeting the anion-binding site on the catalytic domain. In line with these observations is also the fact that the molecule only affected the V MAX parameter of FXIa hydrolysis of the chromogenic substrate but not the K M parameter, as determined by Michaelis-Menten kinetics experiment. As a result, the above studies indicate that benzyl tetraphosphonate 1 is allosteric inhibitor of FXIa and it binds to similar anion-binding sites on both FXIa and FXI. In fact, molecular modeling studies show that the inhibitor can favorably bind to two key basic residues in this binding site. The two residues are K529 and K535 which have previously been implicated in heparin-mediated inhibition of FXIa by antithrombin as well as in polyphosphate-mediated activation of FXI. In contrast to the previous small molecule allosteric inhibitors of FXIa i. e. SPGG and SMI, which were associated with ~100 % inhibition of the enzyme activity, inhibitor 1 induces sub-maximal inhibition of FXIa at saturation, with an efficacy value of ~68 %. Such phenomenon is not possible for orthosteric inhibitors and can only be exhibited by allosteric inhibitors. This further supports the allosteric behavior of inhibitor 1 which, in turn, is important because it permits partial enzyme inhibition (modulation) rather than complete inhibition. Partial allosteric inhibition may translate into a lesser bleeding risk as previously proposed for partial allosteric inhibitors of thrombin. Importantly, targeting allosteric sites on FXIa also better permits the realization of selective enzyme modulation, considering the fact that other coagulation proteases share significant active site similarity. Thus, the allosteric nature of inhibitor 1 will likely result in a better safety profile. Overall, the discovery of this inhibitor is extremely promising for the prospect of discovering more clinically relevant partial modulators of FXIa. ## Conclusions and Outlook In this study, we have identified benzyl tetraphosphonate 1 as the first of its kind inhibitor of human FXIa. The inhibitor demonstrates substantial potency and selectivity toward FXIa and over other coagulation factors of thrombin, FIXa, FXa, and FXIIIa. The molecule not only inhibits the catalytic activity of FXIa, but it also inhibits the activation of its zymogen by thrombin and FXIIa. Importantly, the inhibitor represents an important step forward towards designing allosteric inhibitors of FXIa with submaximal efficacy. This is certainly important to modulate the catalytic activity of an enzyme belonging to a superfamily of largely conserved enzyme members so as to achieve a higher level of enzyme inhibition selectivity, and subsequently, a higher margin of safety. Furthermore, inhibitor 1 also exhibits a significant anticoagulant activity as demonstrated by extending the APTT of human plasma. Accordingly, we put forward this inhibitor as a lead molecule to develop a new generation of effective anticoagulants that are devoid of bleeding complications so as to be safely used in a wide range of patients populations, particularly those who are at a high risk of bleeding. Future studies will focus on establishing the structureactivity relationship of inhibitor 1 to further enhance its potency and selectivity. The structural manipulation will focus on optimizing the number and the position of phosphonate groups as well as on the length and the substituents of the urea-based linker. The druggability of phosphonate derivatives can also be further advanced by preparing the corresponding prodrugs to achieve meaningful oral bioavailability. It is also worth to mention here that the toxicity profile of inhibitor 1 has been evaluated in three cell lines of breast (MCF-7), intestine (CaCo-2), and kidney (HEK-293) (unpublished data). Initial results suggest that 10 μM of the inhibitor does not significantly affect the proliferation of the above cell lines. Testing at higher concentrations will be performed and reported in due time. Lastly, a recent report has suggested that drugs that target the contact activation enzymes including FXIa may serve as potential therapeutics for patients with COVID-19. Earlier, targeting FXI(a)/FXIIa interface by pharmacological means has been shown to prevent coagulopathy, systemic inflammation, and mortality in experimental sepsis. In nonhuman primates, inhibition of contact activation also prevented death from Staphylococcus aureus-induced systemic inflammatory response syndrome. Overall, these studies suggest that the new class of allosteric FXIa inhibitors can be further developed as adjunct therapy for COVID-19 and similar microbial outbreaks that are typically associated with excessive coagulopathy and inflammation. ## Experimental Section Materials Phosphonate derivative (1) and phosphate derivatives (2-4) were purchased from Santa Cruz Biotechnology (Dallas, TX). Reagents for clotting assays including thromboplastin D, APTT reagent, and CaCl 2 solution were all from Fisher Scientific (Pittsburgh, PA). Chemicals used to prepare enzyme assay buffers were from Milipore-Sigma (Burlington, MA), Fisher Scientific, or Bio-Rad laboratories (Hercules, CA). N,N-dimethyl-casein, dansylcadaverine, and dithiothreitol for FXIIIa assay were also from Milipore-Sigma. All types of plasmas were purchased from George King Bio-Medical, Inc. (Overland Park, KS). Antithrombin, coagulation zymogens, and coagulation enzymes including thrombin, FIXa, FXa, FXIa, and FXIIIa were from Haematologic Technologies, Inc. (Essex Junction, VT). Chromogenic substrates: Spectrozyme TH, Spectrozyme FIXa, and Spectrozyme FXa were obtained from Biomedica-Diagnostics (Windsor, NS Canada). Factor XIa chromogenic substrate (S-2366; Lpyroglutamyl-L-prolyl-L-arginine p-nitroaniline hydrochloride) was obtained from Diapharma (West Chester, OH). These substrates have a nitro-anilino chromophore and they were designed based on the physiological substrate of the corresponding clotting factor to ensure its specificity. Heparin was from Milipore-Sigma. F11 (mouse monoclonal antibody from Abnova TM ) for plasma studies, dextran sulfate (MW ca > 500,000), and Coomassie Brilliant Blue for gel electrophoresis were also from Fisher Scientific. Human FXIIa (αform) and antibodies for western blot were from Enzyme Research Laboratories (South Bend, IN). The buffers used for enzyme assays were: a) 50 mM Tris-HCl buffer, pH 7.4, containing 100-150 mM, NaCl, 0.1 % PEG8000, and 0.02 % Tween 80 for human thrombin, FXa, and FXIa; b) 20 mM Tris-HCl buffer, pH 7.4, containing 100 mM NaCl, 2.5 mM CaCl 2 , 0.1 % PEG8000, 0.02 % Tween 80, and 33 % v/v ethylene glycol for human FIXa; and c) 50 mM TrisHCl buffer, pH 8.0, containing 10 mM CaCl 2 and 100 mM NaCl for human FXIIIa. ## Inhibition of FXIa in Chromogenic Substrate Hydrolysis Assay by Phosphonate/Phosphate Derivatives (1-4) Direct inhibition of human FXIa was measured by the corresponding chromogenic substrate hydrolysis assay, as reported earlier , at pH 7.4 and 37 °C. Each well of the 96-well microplate contained 85 μL of the buffer to which 5 μL of molecules 1-4 (or high pure water) and 5 μL of FXIa (0.765 nM) were sequentially added. Following 10-min incubation, 5 μL of FXIa substrate (345 μM) was rapidly added and the residual FXIa activity was measured from the initial rate of increase in absorbance at the wavelength of 405 nm. Stocks of the potential inhibitors were serially diluted. Relative residual FXIa activity at each concentration of the inhibitor was calculated from the ratio of FXIa activity in the presence and absence of the inhibitor. Logistic eq. 1 was used to fit the concentration dependence of residual FXIa activity so as to obtain the potency (IC 50 ) and efficacy (Î"Y%) of inhibition. In this equation, Y is the ratio of residual FXIa activity in the presence of inhibitor to that in its absence, Y M and Y 0 are the maximum and minimum possible values of the fractional residual FXIa activity, IC 50 is the concentration of the inhibitor that leads to 50 % inhibition of enzyme activity, and HS is the Hill slope. Y M , Y 0 , IC 50 , and HS values are determined by nonlinear curve fitting of the data. ## Effect of Benzyl Tetraphosphonate 1 on Other Coagulation Factors The inhibition potential of benzyl tetraphosphonate 1 against thrombin, FIXa, and FXa was also evaluated using the corresponding chromogenic substrate hydrolysis assays reported in our previous studies. Briefly, to each well of a 96-well microplate containing 185 μL of 20-50 mM Tris-HCl buffer, pH 7.4, containing 100-150 mM NaCl, 0.1 % PEG8000, and 0.02 % Tween80 at either 25 °C (thrombin) or 37 °C (FIXa and FXa) was added 5 μL of 0-1 mM benzyl tetraphosphonate 1 (or high pure water) and 5 μL of the enzyme. The final concentrations of the enzymes were 6 nM (thrombin), 89 nM (FIXa), and 1.09 nM (FXa). Following 10-min incubation, 5 μL of Spectrozyme TH (final conc. 50 μM), Spectrozyme FIXa (850 μM), or Spectrozyme FXa (125 μM), was rapidly added and the residual enzyme activity was measured from the initial rate of increase in absorbance at the wavelength of 405 nm. Relative residual enzyme activity as a function of the concentration of the inhibitor was calculated. Likewise, to measure the effect of benzyl tetraphosphonate 1 on human FXIIIa, a bi-substrate, fluorescence-based trans-glutamination assay was performed as we reported previously. Generally, 1 μL of molecule 1 was diluted with 87 μL of pH 7.4 buffer (50 mM Tris-HCl, 1 mM CaCl 2 , 100 mM NaCl, and 2 mg/mL N,N-dimethylcasein) and 5 μL dithiothreitol (20 mM) at 37 °C followed by the addition of 2 μL of human FXIIIa (0.3 μM) and incubation for 10 min. The activity of FXIIIa was monitored following the addition of 5 μL of dansylcadaverine (2 mM) by measuring the initial rate of increase in fluorescence emission (λ Ex. = 360 nm and λ Em. = 490 nm). As with the above enzymes, relative residual FXIIIa activity as a function of the concentration of the inhibitor was calculated. In all of the above enzyme experiments, data were plotted using equation 1 above to obtain the corresponding IC 50 values, only if 50 % or more of enzyme inhibition was obtained. ## Effect of Benzyl Tetraphosphonate 1 on Clotting Times in Human Plasmas Clotting times (APTT and PT) were measured using the BBL Fibrosystem fibrometer (Becton-Dickinson, Sparles, MD), as reported in our previous studies. For the APTT assay, 10 μL of benzyl tetraphosphonate 1 was mixed with 90 μL of citrated human plasma and 100 μL of prewarmed APTT reagent (0.2 % ellagic acid). After incubation for 4 min at 37 °C, clotting was initiated by adding 100 μL of prewarmed 25 mM CaCl 2 , and the time to clotting was recorded. For the PT assay, thromboplastin-D was prepared according to the manufacturer's directions by adding 4 mL of distilled water, and then, the resulting mixture was warmed to 37 °C. A 10 μL of benzyl tetraphosphonate 1 was then mixed with 90 μL of citrated human plasma and was subsequently incubated for 30 sec at 37 °C. Following the addition of 200 μL of prewarmed thromboplastin-D preparation, the time to clotting was recorded. In the two assays, seven concentrations of the inhibitor were used over the concentration range of 0-500 μM to establish a concentration vs effect curve. The concentrations vs clotting times data were fitted to a quadratic trend line, which was eventually used to determine the concentration of the inhibitor necessary to increase the clotting time by 1.5-fold. Clotting times in the absence of an anticoagulant was determined in a similar fashion using 10 μL of deionized water and was found to be 34.4 � 0.1 sec for APTT and 14.5 � 2.1 sec for PT. To establish the FXIa-dependent effect of the inhibitor in human plasma, the APTT assay was repeated using FXIdeficient human plasma to which human FXIa (0, 2.4, and 4.8 nM) was added in the absence and the presence of benzyl tetraphosphonate 1 (0, 233, 327, and 467 μM). ## Michaelis-Menten Kinetics for Chromogenic Substrate (S-2366) Hydrolysis by Human FXIa in the Presence of Benzyl Tetraphosphonate 1 The initial rate of S-2366, a chromogenic tripeptide substrate, hydrolysis by purified human FXIa was obtained from the linear increase in absorbance at the wavelength of 405 nm corresponding to the consumption of < 10 % of the chromogenic substrate, as reported in our previous studies. The initial rate was measured as a function of various concentrations of the substrate (0-2000 μM) in the presence of a fixed concentration of benzyl tetraphosphonate 1 in 20 mM Tris HCl buffer, pH 7.4, containing 100-150 mM NaCl, 0.1 % PEG8000, and 0.02 % Tween80 at 37 °C. The experiment was conducted at five concentrations of the inhibitor: 0, 5, 10, 25, 50, and 100 μM. The data was fitted using the standard Michaelis Menten equation 2 to determine the K M (the affinity of the substrate to the active site of FXIa) and V MAX (the maximum hydrolysis reaction velocity). ## Effect of Benzyl Tetraphosphonate 1 on FXIa-Mediated Activation of FIX The experiments were done as reported in our earlier studies. FIX (6.2 μM) was incubated with FXIa (10 nM) in the presence of inhibitor 1 (0, 50, 250, and 1000 μM) in 50 mM HEPES supplemented with 5 mM CaCl 2 , at room temperature. Samples were incubated for 30 min. Following the 30-min incubation, the reactions were quenched using sodium dodecyl sulfate (SDS) polyacrylamide gel electrophoresis (SDS-PAGE) loading buffer containing dithiothreitol and electrophoresed on a 10 % SDS-polyacrylamide gel. Protein bands were visualized by staining with Coomassie Brilliant Blue. For Western blot experiment, plasma FIX (200 nM) was incubated at room temperature with human FXIa (40 nM) in 50 mM HEPES buffer supplemented 5 mM CaCl 2 . After 30-min incubation, SDS-PAGE loading buffer containing dithiothreitol was added, fractionated on 10 % polyacrylamide-SDS gels, and then transferred to nitrocellulose membrane. The primary antibody was goat anti-human FIX polyclonal IgG, and the secondary antibody was horseradish peroxidase-conjugated anti-goat IgG. Detection was by chemiluminescence. The relative positions of FIX and FIXa bands were confirmed using Western blots of known standards for each protein. ## Effect of Benzyl Tetraphosphonate 1 on FXIa -Antithrombin Complex Formation The effect of inhibitor 1 on the complex formation between FXIa and antithrombin was performed in HEPES buffer supplemented with 5 mM of CaCl 2 , as reported earlier. Briefly, FXIa (300 nM) was pre-incubated with inhibitor 1 (0, 50, 250, and 1000 μM) at 37 °C for 5 minutes, and then combined with 2 μM purified human antithrombin in the presence of 2 μM sodium heparin for a further 30 minutes. At the incubation time, samples were quenched using SDS-PAGE loading gel buffer containing dithiothreitol and subjected to electrophoresis on 10 % SDS-PAGE. Protein bands were visualized by staining with silver stain. ## Effect of Benzyl Tetraphosphonate 1 on FXI Activation by Thrombin Thrombin-mediated FXI activation was analyzed by SDS-PAGE, as reported earlier. [49, FXI (700 nM) was incubated with α-thrombin (70 nM) with or without dextran sulfate (10 μg/ml) in the presence of different concentrations of inhibitor 1 (2, 20, 50, 250, and 1000 μM) in HEPES buffer supplemented with 5 mM of CaCl 2 . Samples were incubated for 60 mins. After the incubation period, reactions were quenched using argatroban (2 μM) and polybrene (6 μg/ml). samples were placed into reducing sample buffer, size fractionated on 10 % polyacrylamide-SDS gel and stained with silver staining. ## Effect of Benzyl Tetraphosphonate 1 on FXI Activation by FXIIa FXIIa-mediated activation of FXI activation was analyzed by SDS-PAGE, as reported earlier. [49, FXI (700 nM) was incubated with α-FXIIa (200 nM) with or without dextran sulfate (2 μg/ml) in the presence of different concentrations of inhibitor 1 (50, 250, and 1000 μM) in HEPES buffer supplemented with 5 mM of CaCl 2 . Samples were incubated for 60 mins. After the incubation period, reactions were quenched using corn trypsin inhibitor (1 μM) and polybrene (6 μg/ml). samples were placed into reducing sample buffer, size fractionated on 10 % polyacrylamide-SDS gel and stained with silver staining. ## Molecular Modeling Studies of FXIa and Benzyl Tetraphosphonate 1 The docking studies were carried out using Glide of Schrodinger Suite 2017-1. Initial coordinates for FXIa were taken from the crystal structure of FXIa in complex with α-ketothiazole argininebased ligand (PDB: 2FDA). The protein structure was prepared by removing the crystallographic water molecules and the crystal ligand, and by adding hydrogen atoms consistent with the physiologic pH of 7.0 using Maestro 11.1 of Schrodinger Suite. The protein molecule was then energy minimized with an RMSD cutoff value of 0.3 Ã for all heavy atoms. Initial coordinates for benzyl tetraphosphonate 1 were built and energy minimized using the Schrodinger Suite. The basic residue of K529, R530, R532, K535, and K539 in the catalytic domain and the surrounding area of these amino acids were specified as the ligand binding site. The grids for the target protein were generated using the OPLS3 forcefield. The grid center was set to be the centroid of the above basic residues, with a cubic grid box of 10 Ã on each side. No constraints were used in the grid generations. The docking calculations were done using the default parameters under the stand precision mode. All the poses were subjected to post-docking minimization. The best-docked structure based on the docking score was selected for subsequent analysis. The chymotrypsin-based numbering for the presented amino acids is as follows: T523 = T164, N524 = N165, K529 = K170, R530 = R171, R532 = R173, K535 = K175, and K539 = K179.

chemsum

{"title": "Discovery of Benzyl Tetraphosphonate Derivative as Inhibitor of Human Factor Xia", "journal": "Chemistry Open"}

emission_and_absorption_tuning_in_mr-tadf_b,n-doped_heptacenes:_towards_ideal-blue_hyperfluorescent_

## Abstract: Developing high-efficiency purely organic blue organic light-emitting diodes (OLEDs) that meet the stringent industry standards is a major current research challenge. Hyperfluorescent device approaches achieve in large measure the desired high performance by combining the advantages of a high-efficiency thermally activated delayed fluorescence (TADF) assistant dopant with a narrowband deep-blue multi-resonant TADF (MR-TADF) terminal emitter. However, this approach requires suitable spectral overlap to support Förster resonance energy transfer (FRET) between the two.Here we demonstrate colour tuning of a recently reported MR-TADF B,N-heptacene core through control of the boron substituents. While there is little impact on the intrinsic TADF properties -as both singlet and triplet energies decrease in tandem -this approach improves the emission colour coordinate as well as the spectral overlap for blue hyperfluorescence OLEDs (HF OLEDs). Crucially, the redshifted and more intense absorption allows us to pair this MR-TADF emitter with a high-performance TADF assistant dopant and achieve maximum external quantum efficiency (EQEmax) of 15% at colour coordinates of (0.15, 0.10). The efficiency values recorded for our device at a practical luminance of 100 cd m -2 are among the highest reported for HF TADF OLEDs with CIEy ≤ 0.1. ## Introduction Organic light-emitting diode (OLED) display technology is evolving at a brisk rate, with high-end ultrahigh definition (UHD) 4K and 8K OLED displays already in the market. These high-resolution displays must meet stringent emission colour standards (BT.2020-2), 1 which are defined according to the Commission International de l'Éclairage (CIE) 1931 as (0.13, 0.05), (0.17, 0.80) and (0.71, 0.29) for blue, green, and red, respectively. 2 At present this colour requirement is met through the use of absorptive filters or microcavities to deliver saturated blue, green, and red emission. 3,4 Unfortunately this approach necessarily results in a reduced efficiency of the devices as unwanted emission wavelengths are rejected. An alternative and attractive solution to the issue of colour purity is to develop materials with intrinsically narrowband emission. Commercial OLED displays currently use organic triplet-triplet annihilation (TTA) emitters for blue pixels. By virtue of the operational exciton harvesting mechanism, the internal quantum efficiency of blue OLEDs is presently capped at 62.5%. Organic thermally activated delayed fluorescence (TADF) materials, by contrast, can harvest 100% of the electrically generated excitons to produce light and achieve higher device efficiencies. 5 Similarly, organometallic phosphorescent OLEDs can generate high-efficiency OLEDs, although with significant intrinsic limitations for blue-emitting device in terms of their stability. These limitations arise as photon energies approach metal-ligand bond dissociation energies, and as thermal population of metal-centred states (particularly for d6 metal complexes) leads to severe non-radiative decay. The TADF mechanism involves the thermal upconversion of non-emissive triplet excitons into singlets via reverse intersystem crossing (RISC). A small energy gap between T1 and S1 (DEST) is necessary to achieve TADF, and this is ensured when there is spatial separation of the frontier orbitals, which is normally achieved in a highly twisted donor-acceptor (D-A) architecture. However, D-A TADF emitters typically emit from charge transfer (CT) states where there is a large reorganization in the excited state that leads to broad (70-100 nm) emission. 6,7 These broad emission bands interact poorly with the aforementioned colour purity filters, while also making it difficult to engineer deep-blue emitters and suitably high-triplet hosts. 11 Multiresonant TADF (MR-TADF) compounds are an alternative class of TADF materials that are typically based on p-and n-doped nanographene (Figure S1). 12,13 Because of their rigid structure, these compounds show narrowband emission, typically with full width of half maxima (FWHM) of around 20-30 nm. Thus, MR-TADF emitters show potential to generate the required deep blue emission demanded by industry, and support the development of stable, efficient, pure blue narrowband emitting organic materials that are expected to revolutionize OLED displays. 14 MR-TADF compounds show TADF due to the alternating pattern of electron density between the ground and excited states that leads to small DEST, combined with upper-triplet crossings from thermally-populated Tn states back to the emissive S1 state. The excited states thus possess a distinct short range charge transfer (SRCT) character, 13,19,20 and MR-TADF compounds are endowed with high singlet radiative decay (kr) rates of around 10 7 s -1 . 5,19 Despite these advantages, the RISC rates reported for MR-TADF materials typically lag ~100 times slower than those of leading D-A or D-A-D TADF materials. 13,21 This limitation frequently leads to inefficient triplet harvesting and low performance of the OLEDs at practical luminance. To circumvent this limitation, MR-TADFs have recently found a parallel application as terminal emitters in hyperfluorescent (HF) OLEDs. Indeed, truly ground-breaking device performances have been achieved when pairing a high-RISC D-A assistant dopant with narrowband MR-TADF terminal emitters. This performance is in some cases supported by spontaneous alignment of the MR-TADF with the substrate, thus improving optical outcoupling of the device, 24 although the specific loss mechanisms that occur under electrical excitation, including charge trapping and Dexter transfer, 25 remain poorly understood. High-performance HF OLEDs rely on efficient energy transfer between the D-A TADF assistant dopant that is responsible for exciton harvesting and the MR-TADF terminal emitter. Förster resonance energy transfer (FRET) is understood to be the main mechanism for this transfer, and FRET efficiency is proportional to the overlap between the emission spectrum of the D-A TADF assistant dopant and the absorption spectrum of the MR-TADF terminal emitter. As a result, the compatibility of many MR-TADF materials is severely limited by the low molar extinction coefficient of their lowest energy absorption bands -those that predominately overlap with the emission spectrum of the D-A TADF assistant dopant. Expanding the FRET compatibility of a specific terminal emitter requires increasing the spectral overlap between the assistant dopant and terminal emitter. This in turn requires development of deeper-blue D-A TADF assistant dopants, but such emitters remain persistently elusive -a circumstance that has stimulated the rapid development of the alternate hyperfluorescence approach. 26 Clearly then, synthetic control over both the emission and absorption spectra of MR-TADF materials is crucial to enable and optimise their use in HF OLEDs with available D-A TADF co-hosts. We previously reported a linear B,N-doped ladder type heptacene that emits at 390 nm (near UV) with a FWHM of 31 nm (240 meV) in THF solution. 30 The material displayed weak TADF due in part to the large ΔEST of 0.31 eV, along with significant TTA contribution reflected in the extended ms-timescale of the emission decay. 30 Due to the near UV emission and subsequent paucity of suitable OLED hosts, the investigated device performance was poor. 31 In the present report we show how replacement of the hydroxyl groups for mesityl substituents in α-3BNMes (Figure 1) leads to a desired red-shifting of the emission towards an ideal-blue emission colour coordinate, as well as a red-shifting of the absorption that supports HF compatibility with available D-A TADF assistant dopants. Together, these changes in optical properties compared to α-3BNOH allow α-3BNMes to be used in high-performance deep-blue HF OLEDs that show 15% maximum external quantum efficiencies (EQEmax) and colour coordinates of (0.15, 0.10). These results represent highly competitive device performance metrics at this colour coordinate that are enabled by the finely-tuned emission and absorption spectra of α-3BNMes (Table S2). ## Synthesis α-3BNMes was synthesized in three steps (Scheme S1), where the key borylation step proceeds in 57% yield. The identity and purity of α-3BNMes were established from a combination of 1 H NMR spectroscopy (S2), high-resolution mass spectrometry (S4), HPLC (S5), GPC trace analyses (S7) and single crystal X-ray diffraction analysis (Figure 2). Like the parent compound (α-3BNOH), α-3BNMes shows high thermal stability, revealed by thermogravimetric analysis (TGA), with a decomposition temperature (Td), defined as the 5% weight loss of the material, at 503 °C (Figure S8). We investigated the structure of α-3BNMes by growing single crystals via slow evaporation of the compound in THF. Full datasets were collected from many different crystals, and we report here our best result. The data obtained are adequate to demonstrate connectivity and gross structure but do not merit discussion of bond lengths (Figure 2). Like the parent α-3BNOH, in α-3BNMes the heptacene core remains planar, and the aryl substituents attached to boron and nitrogen align nearly orthogonal to main acene core. The electron density distribution patterns of these orbitals is typical for MR-TADF compounds (Figure 3). 13 The HOMO density is mainly localized on the nitrogen and carbon atoms positioned ortho to them, while in the LUMO, electron density is mainly localized on the boron atoms and the carbons ortho to them. We applied spin component scaling second order approximate coupled-cluster (SCS-CC2) to accurately predict the nature and energies of the excited states and the ΔEST (Figure 3b and Table S1). 13,19,32 α-3BNMes shows a gratifyingly smaller ΔEST of 0.20 eV compared to that of α-3BNOH (ΔEST of 0.29 eV), but at the expense of a slightly smaller oscillator strength (f) for the transition to S1 of 0.08 compared to α-3BNOH (f of 0.09). 30 The predicted S1 energy is also stabilized to 3.40 eV compared to 3.69 eV in α-3BNOH. The difference density plot of the S1 state shows the characteristic alternating pattern of increasing and decreasing electron density on adjacent atoms that is characteristic of MR-TADF compounds. 32 The patterns revealed in the difference density plots for T1 and T2 differ slightly from that of S1. This difference in the nature of the excited singlet and triplet states will result in enhanced spin-orbit coupling and assist in RISC following El Sayed's rules. 33 The absorption spectrum was simulated by capturing transitions to the first five singlet excited states at the SCS-CC2/cc-pVDZ level of theory using DFT calculated ground state. Good agreement between the measured and simulated spectral shapes was obtained (Figure S9). Similar trends in the difference density pictures of the lowest-lying singlet excitations were obtained for α-3BNMes (Figure S10) compared to those of a-3BNOH , 30 explaining the near identical shapes of their absorption spectra (Figure 4). ## Optical properties Steady-state photophysical properties of both α-3BNMes and the previous α-3BNOH in dilute THF are shown in Figure 4. In the case of α-3BNMes, a bathochromic shift of 330 meV compared to α-3BNOH is observed in both the main π-π* absorption band at 3.50 eV (354 nm) and the quasi degenerate S1, S2 SRCT excited states at 2.94 eV (421 nm). By replacing the strongly mesomerically electron-donating hydroxyl groups with inductively electron-withdrawing mesityl substituents, both the HOMO and LUMO levels are stabilized although the LUMO to a greater extent. This results in a net stabilization of the excited states in α-3BNMes, and red-shifts both its absorption and emission spectra. The molar extinction coefficient for the high-energy band (S4) is also increased by a factor of two, although this still appears at very short wavelengths, and thus is unsuitable for FRET and HF OLED applications using available D-A TADF assistant dopants. Nonetheless, this demonstrates that significant control over the absorption spectrum is indeed possible. Promisingly, the HF device-relevant SRCT absorption bands are increased by a factor of 1.4 along with a useful red-shift to wavelengths beyond 400 nm in α-3BNMes. Although these device-relevant SRCT bands remain smaller and at higher energies than in v-DABNA (Figure S11, limiting HF compatibility with available D-A TADF co-hosts), they still represent a significant improvement compared to α-3BNOH. It remains unclear how the structure of v-DABNA leads to its lowest-energy absorbance band becoming dominant, compared to the S4 band in the heptacene systems. The solution photoluminescence spectrum of α-3BNMes shows a narrow emission centred at lPL = 2.80 eV (442 nm), with FWHM of 190 meV (30 nm) and a small Stokes shift of 140 meV (20 nm). Gratifyingly, the replacement of the hydroxyl groups with the mesityl substituents leads to a yet narrower, deep-blue emission, shifting the CIE coordinates of the PL spectrum from (0.17, 0.01) to (0.15, 0.04). These CIE 1931 values are considered ideal for blue OLEDs as defined by BT.2020-2. 1 Together, these THF solution results establish the absorption-and emission-tuning abilities of the boron substituents towards synthetic control of the singlet states of the B,N-heptacene core. We find that the ΔEST are very similar in both materials, indicating that mesityl substitution of the boron atoms results in similar changes to both the singlet and triplet state energies of α-3BNMes compared to α-3BNOH. Time-resolved photoluminescence decays of α-3BNMes doped at 1 wt% in PMMA matrix are shown in Figure 5b. Similar to previous reports of the photophysics of α-3BNOH 30 and many other reported MR-TADF materials, the delayed fluorescence from α-3BNMes is weak and long-lived compared to D-A TADF materials, 8,9,34,35 indicating only a moderate rate of rISC. When measured at lower temperatures, the delayed emission component (after 100 ns) is suppressed, evidencing the thermally activated mechanism of this emission. At longer delay times (beyond 0.1 ms) and at temperatures below 150 K the emission decay rate changes significantly, which along with a strong spectral red-shift identifies the phosphorescence regime from which the phosphorescence spectrum (Figures 5a, 5c) is extracted. Exponential fitting of the room temperature decay reveals three different components: the prompt fluorescence has a lifetime, tp, of 10 ns, and there are two main components to the delayed emission with lifetimes of 9.08 μs and 7.06 ms (Figure S12a). The shorter of the delayed lifetimes, corresponding to a regime of dual emission, is attributed to combined monomer and a red-shifted aggregate emission (Figures 5c, S12c), most probably associated with excimer emission as described by Stavrou et al., 15 which is common in planar MR-TADF emitters. Notably, the contribution to the emission decay from this species is minimal and energetically is at the same positions as for α-3BNOH. 30 The longest delayed component is attributed to pure monomer emission with mono-excitonic origin, in contrast with α-3BNOH where it was found to have bi-excitonic TTA contribution (Figure S13). As in other recent studies of MR-TADF materials, 15,18 we find that changing the host matrix has only a modest influence on kRISC (Figure S12b). Finally, the photoluminescence quantum yield (PLQY) of α-3BNMes in 1% mCP doped film was determined to be 63%, thus nearly double that of α-3BNOH (35% in 1 wt% mCP), and is sufficiently high to support OLED applications. With increasing doping concentration, no significant differences were observed in the current densityvoltage-luminance (jVL) curvers, the spectra, or EQE as a function of current density in both host environments. Consequently, we do not anticipate qualitatively different behaviour in MR-TADF only devices with lower doping concentration, which were thus not investigated. More interestingly, no broadening of the electroluminescence (EL) spectrum was observed upon increasing concentration, indicating suppression of excimer formation even at high concentrations, a very common phenomenon of other MR-TADF emitters in films. Maximum EQEs of 1.7 % at ~ 100 cd m -2 were achieved in devices with each of the two hosts. We suggest that this poor performance is due to the weak TADF contribution of the material, leading to low efficiency combined with additional efficiency roll-off from the known instability of DPEPO, 36,37 especially at high current densities (Figure S14c). Nonetheless, the CIE coordinates are very attractive at (0.15, 0.08), which are the same as in the photoexcited film and similar to that determined in THF solution. To compensate for the poor intrinsic exciton harvesting performance of α-3BNMes while still taking advantage of its ideal emission spectrum we then applied it as a terminal emitter in HF OLEDs in combination with a D-A(-D) TADF sensitizer. In order to ensure adequate spectral overlap necessary for the energy transfer, we employed 2,8-bis(2,7-di(tert-butyl)-9,9-dimethylacridin-10(9H)yl)dibenzo[b,d]thiophene (DtBuAc-DBT) as a TADF co-host, previously reported to give ~11% EQEmax and suitably high-energy blue emission with CIE coordinates of (0.16, 0.17). 8 This D-A-D TADF assistant dopant was co-evaporated at 25 vol% in the EML, alongside 1 vol% α-3BNMes and 74 vol% DPEPO. The device architecture was chosen to be the same as the one previously reported, 8 with the addition of a-3MesBN as the terminal emitter, using ITO The resulting HF OLEDs possess good efficiency, with an EQEmax of 15% and an EQE100 (at 100 cd/m 2 ) of 10.2%, with a FWHM of 290 meV (49 nm) and CIE of (0.15, 0.10), which is enabled by triplet harvesting by the D-A-D sensitizer, together with narrow deep blue emission from the α-3BNMes (Fig- ure 6). Although the EQEmax is improved at 15% in the HF OLED (compared to 11% for the DtBuAc-DBT device) the larger roll-off at higher current densities results to an EQE300 of 6.2% and 8.8% for the HF and TADF OLEDs, respectively. The increase in EQEmax is a compromise between detrimental factors (e.g., the moderate PLQY of α-3BNMes as the terminal emitter, competing unproductive Dexter energy transfer, in-situ charge trapping, etc.) and beneficial effects (e.g., α-3BNMes alignment/anisotropic emission, and good FRET overlap outcompeting DtBuAc-DBT non-radiative decay), although some of these factors are deceptively difficult to quantify by established experimental methods. 25 The overall HF OLED performance is nonetheless competitive with other recent leading reports at similar colour coordinates (Figure S16 and Table S2). The EL spectrum of the HF OLEDs is nearly identical to the α-3BNMes spectrum, indicating efficient FRET with only a small contribution of the D-A-D TADF sensitizer at ~525 nm. This efficient FRET occurs despite a seemingly small FRET overlap (Figure S15) and highlights the importance and challenge of both emission and absorption spectral tuning for developing high-efficiency blue HF OLED material combinations. While it is incredibly challenging to further deepen the emission colour of available D-A-D TADF assistant dopants, the need to do so can be avoided in HF OLEDs by instead redshifting the absorption spectrum of the MR-TADF terminal emitters, as we demonstrate here. ## Conclusions We have achieved emission and absorption colour tuning in a deep blue non-triangulene type MR-TADF compound by altering the boron substituents in a B, N-doped heptacene. Compared to the parent UV emitter α-3BNOH, these changes imbue α-3BNMes with ideal CIE coordinates for blue OLEDs and suitable absorption spectrum for FRET compatibility with existing D-A TADF co-hosts. The resulting HF OLEDs achieved an EQEmax of 15% and deep-blue colour coordinates of (0.15, 0.10), compared to <1% efficiencies for the unassisted α-3BNOH. In light of slow progress towards the development of truly deep-blue D-A TADF emitters, we advance that controlling the absorption spectrum of terminal MR-TADF emitters to expand their compatibility in HF devices is a more fruitful approach.

chemsum

{"title": "Emission and Absorption Tuning in MR-TADF B,N-Doped Heptacenes: Towards Ideal-Blue Hyperfluorescent OLEDs", "journal": "ChemRxiv"}

a_new_four-parameter_cubic_equation_of_state_for_predicting_fluids_phase_behavior

## Abstract: A new four-parameter cubic equation of state (EoS) is generated by incorporating the critical compressibility factor (Zc) apart from the critical pressure (Pc) and temperature (Tc).One free parameter in the denominator of the attractive term and two parameters in the alpha function are adjusted using the experimental data of saturated liquid density, vapor pressure, and isobaric liquid heat capacity of 48 components including hydrocarbons and non-hydrocarbons.Applying this equation of state, saturated liquid density, saturated vapor density, and vapor pressure of pure components are accurately reproduced compared with experimental values. Furthermore, the predicted properties including derivatives of alpha function, such as enthalpy of vaporization, entropy of vaporization and isobaric heat capacity of liquid, also have decent accuracy. The global average absolute relative deviation (AAD) of saturated liquid density, saturated vapor density, saturated vapor pressure, enthalpy of vaporization, entropy of vaporization, and isobaric heat capacity of liquid in a wide reduced temperature (Tr) range of subcritical region reproduced by this work are 4.33%, 4.18%, 3.19%, 2.26%, 2.27%, and 5.82%, 2 respectively. Substantial improvement has been achieved for the isobaric liquid heat capacity calculation. ## INTRODUCTION Equation of state is of fundamental instrument for thermodynamic properties prediction of both pure substance and mixtures as well as simulating fluids phase behavior for practical application. Cubic EoS is an outstanding member of the EoS family equipped with balanced feature of simplicity, accuracy, and computational speed. Therefore, cubic EoSs are widely applied in reservoir modelling, petrochemical industries, chemical design, and separation processes. Since the time van der Waals proposed the first cubic equation of state, numerous equation of state have been deduced by either revising the repulsive term or modifying the attractive term in order to improve its predictive capability. Among them, Soave (SRK EoS) and Peng and Robinson (PR EoS) developed probably the most successful cubic equations of state for vapor liquid equilibrium calculation. However, both of these two generic cubic EoSs assumed a particular critical compressibility factor for all substances, 0.333 and 0.307, respectively, which violates the facts of component dependent critical compressibility. Therefore, the predicted densities differ significantly from their experimental values, so do some related properties. Many attempts have been made to counter this weakness. Equations of state proposed by Schmit and Wenzel , Patel and Teja , and Trebble and Bishnoi introduced a component dependent empirical critical compressibility. While this parameter, in general, does not equal to the experimental value of critical compressibility of fluids, the new free parameter made the deviation of the thermodynamic properties to be adjustable, and subsequently improved the accuracy of the EoSs. On the other hand, the work conducted by Guennec and Privat recently showed the volume translated-consistent Peng-Robinson (tc-PR) and the volume translated-consistent Redlich-Kwong (tc-RK) EoSs could noticeably eliminate the apparent discrepancy of the saturated densities reproduced by the original PR and RK EoSs. Instead of using an empirical critical compressibility factor as those three parameter cubic EoSs, the present work integrate the experimental critical compressibility factor, and introduce a fourth parameter as a free one to be determined by minimizing the deviation of the saturated liquid density. Moreover, the Melhem type of alpha function is found to be suitable for the attractive term of the new EoS, ## THE NEW CUBIC EQUATION OF STATE The equation of state proposed in this work has the following form: Where R is the universal gas constant, a(T) is function of temperature, while b, w, and u are temperature independent constant for particular component. The arrangement of the cubic equation chosen here is not fresh, identical forms having been chosen earlier by Schmit and Wenzel , Adachi et al. , Jan et al. . By imposing the van der Waals constraints at the critical point, the following set of equations are obtained: where In this particular work, the following Melhem type of alpha function is employed for the proposed EoS: Equations ( 1) -( 8) complete the description of the new cubic EoS. 48 pure components including hydrocarbons and non-hydrocarbons are selected to adjust the undetermined parameters in the new EoS. b data source from Perry's Chemical Engineers' Handbook, Eighth Edition. c data source from Guennec and Privat. ## OBJECTIVE FUNCTION FOR IDENTIFYING OPTIMAL u, m AND n A proper objective function is indispensable for obtaining the optimal value of those undetermined parameters. The common form of the objective functions usually took the vapor pressure and saturated liquid density into consideration, and different weighting factor set (w1, w2) were used for authors, such as Dashtizadeh et al. , Nasrifar et al. , Bonyadi et al. set (0.8, 0.2), while Haghtalab et al. sets (0.7, 0.3), for vapor pressure term and saturated liquid density term, respectively. In contrast, in the literature of Guennec et al. , the properties, for instance enthalpy of vaporization and heat capacity, were incorporated into the objective function due to the enthalpy and heat capacity calculation involve the first and second derivatives of the alpha function with respect to temperature, while the vapor pressure and density calculation only involve the alpha function itself. Therefore, if the derivatives of alpha are not involved in the objective function, it may lead to parameters in the alpha function unsuitable for reproducing properties containing alpha derivatives. In this work, objective function is also applied in order to acquire optimal value for those undetermined parameters in the new EoS, but the form of the objective functions used here are slightly different from those aforementioned, and details are described below. An interesting feature of the EoS proposed in this work is that the deviation of the saturated liquid density has weak correlation with the alpha value as long as the value belongs to the neighborhood of optimal alpha value, while strongly depend on the value of parameter u. Therefore, the procedure for identifying optimal u is decoupled from the procedure for identifying parameters in the alpha function. In the present work, the following objective function one (OF1), eq 10, is chosen to optimize the value of u, while the discrete alpha value derived from the phase equilibrium condition when experimental value of pressure is used as input parameter at each data point. The calculation is confined by the principle of iso-fugacities of two phases in equilibrium condition, eq 9. The principle denotes as below: For the present work, the Genetic Algorithm in MATLAB toolbox has been utilized to optimize the regulable parameters. With this technique, the optimal values of u, m and n can be acquired respectively. The generalized form of u denotes as eq 12: For the nonpolar and slightly polar substance pool, where methanol, ethanol, water, acetone, and ammonia were excluded, the following generalized form, eq 13 and 14, could be used for the parameters m and n in the alpha function. And the relationship between m and n was elaborated in literature by Forero G. et al. . 𝑚 = 76.8570(𝜔𝑍 𝑐 2 ) 2 + 10.8880𝜔𝑧 𝑐 2 + 0.1486 In the case of polar substances (methanol, ethanol, water, acetone, and ammonia) included in ## RESULTS AND DISCUSSION In order to assess the performance of the newly proposed cubic EoS, calculation of some thermodynamic properties, including saturated liquid density, saturated vapor density, vapor pressure, enthalpy of vaporization, entropy of vaporization, and isobaric heat capacity of liquid, of pure substances are carried out. The thermodynamic relations for using eq 1 are provided in Appendix. Results are compared with the Jan et al. (JT) EoS, the Adachi et al. (ALS) EoS, the generalized tc-PR EoS with Twu88 alpha function, and the generalized tc-RK EoS with Twu88 alpha function. While addressing those substances listed in Table 2 with tc-PR and tc-RK EoS, the Twu91 alpha function is incorporated instead of Twu88 alpha function. The JT and ALS EoSs are selected as comparison is because they have the same form as the EoS proposed in this work, while developed with different methodology. The generalized tc-PR and tc-RK EoSs are involved due to their accuracy and the significance of original PR and RK EoSs. Results are categorized in Tables 3-5. The calculated average absolute relative deviations (AAD) for each component presented in Table 1 and the global average absolute relative deviation are reported. In Table 3, the predicted saturated liquid density and saturated vapor density of 48 pure substances are compared with experimental data, as well as the predictions given by the JT, ALS, tc-PR and tc-RK EoSs. In regard to

chemsum

{"title": "A New Four-Parameter Cubic Equation of State for Predicting Fluids Phase Behavior", "journal": "ChemRxiv"}

computational_prediction_of_the_supramolecular_self-assembling_properties_of_organic_molecules:_flex

## Abstract: Two families of organic molecules with different backbones have been considered. The first family, composed by a substituted central phenyl is considered as flexible. The second one, based on a macrolactam-like unit, is considered as rigid. They have however a common feature, three amide moieties (as substituents for the phenyl and within the cycle for the macrolactam-like molecule) that allow hydrogen bonding when molecules are stacked. In this study we propose a computational protocol to unravel the ability of the different families to self-assemble into organic nanotubes. Starting from the monomer and going towards larger assemblies like dimers, trimers, and pentamers we applied different theoretical protocols to rationalize the behavior of the different assemblies. Both structures and thermodynamics were investigated to give a complete picture of the process. Thanks to the combination of a quantum mechanics approach and molecular dynamics simulations along with the use of tailored tools (non covalent interaction visualization) and techniques (umbrella sampling), we have been able to differentiate the two families and highlight the best candidate for self-assembling purposes. ## Introduction Supramolecular chemistry is a key concept for many edifices or mechanisms that are essential for life . So it is for some applications going from material sciences to medicine through information storage . It has been defined by Lehn, one of its founders, as the chemistry of intermolecular bonds and interactions . Supramolecular chemistry involves a wide variety of weak interactions of different strength. Metal-ligand interactions are the strongest one with an interaction energy of about 80 kcal.mol −1 . Ionic, ion-dipole and dipole-dipole interactions are slightly less stabilizing with an interaction energy ranging from 15 to 50 kcal.mol −1 . Aromatic interactions, encompassing π-π, π-cation and π-hydrogen bond, lead to a bounding of 3 kcal.mol −1 . Finally, hydrogen bonds and van der Waals forces are the less energetic ones with interactions energies of 1-10 kcal.mol −1 and lower than 1 kcal.mol −1 , respectively. As the previously mentioned interactions are non-covalent, it means that the interactions can be reversible and thus confer a kind of flexibility to the system. Nevertheless, increasing the number or combining weak interactions can lead to really stable assemblies and one can cite for example the structure of ice, DNA or the synthetic supramolecular polymer, namely the nylon . In both cases, hydrogen bonds (H-Bonds) are at the origin of the large stability of the supramolecular assemblies (SMA). The association and dissociation processes within large SMA is ensured by the inter-and intramolecular H-Bonds, the best example being the secondary and tertiary structures of proteins that are dictated by those weak interactions . In particular, H-Bonds based on an amide pattern (N-H ••• O=) are frequently found. The low to medium strength of this type of H-Bond, around 2 kcal.mol−1, along with its flexibility, allow SMA to assemble and disassemble easily. One of the specificity of the amide-based H-Bonds is their directionality. It has been used to build SMA that are characterized by a stacking of the molecules leading to a one-dimensional columnar SMA . Many studies have reported SMA structures where the building block was a substituted benzene with various number of amide moieties for the formation of organogels or liquid crystals . The number and the orientation of the amide moieties have been shown to be crucial for the effectiveness of the formation of the stacking . Based on the idea of an amide-substituted benzene for the formation of columnar stacking, we have designed four benzenetricarboxamide molecules (B4s family, Figure 1). Amide moieties are perpendicular to the plane of the benzene. A difference in the orientation of the three amide moieties was observed for B4s family, while two of them were oriented in the same direction, the last one was found antiparallel (see Figure S3 in ESI †). The remaining positions of the central benzene were substituted by -CH 3 (B4, B4c, B4p) or -I (B4pI) groups to maintain the amide moieties perpendicular to the plane of the molecule. In addition to the substitution of the central benzene, amides were also substituted by -C 6 H 13 alkane chain (B4), -C 6 H 11 cyclic group (B4c), and CH 2 -Ph moiety for both B4p and B4pI. A different backbone, based on a macrolactam-like unit, has been considered (B9s family, Figure 1). It also encompassed three amide moieties that are conjugated to an alcene to avoid their rotation. Two substitution patterns have been added, a -CH 3 for B9m and a -CH 2 OCOCH 3 for B9. The global idea is to compare those two apparently C3-symmetric backbones for an efficient formation of columnar stacking involving amide-based H-Bonds. To address this challenge we propose a computational approach. Recent studies have already shown that a fully theoretical approach, based on molecular dynamics simulations and quantum chemical calculations, was able to unravel hydrogen bond interactions between organic molecules within supramolecular assemblies . Hence, the multiscale approach that will be developed in this study will address for the different small organic molecules (See Figure 1) their self-assembling property and the stability of the possible columnar stackings, considered as organic nanotubes. The multiscale approach will involve calculations based on quantum mechanics but also classical molecular dynamics simulations combined with tailored tools to characterize the H-Bonds. Umbrella sampling is particularly relevant for the analysis of supramolecular interactions as it allows the calculation of a binding energy through the derivation of a potential of mean force . Non-covalent interactions (NCI) visualization is also a valuable tool that can illustrate the underlying weak interactions such as hydrogen bonds or van der Waals interactions within SMA . ## Single point calculations All the calculations have been performed using Gaussian16 package within the density functional theory (DFT) framework. We used the ωB97X-D range-separated hybrid exchange correlation functional (XCF) combined with the 6-311+G(d,p) atomic basis set . This XCF is known as one of the most efficient to consider structures and energies of assemblies involving hydrogen bond interactions . The solvent (water) was modeled using an implicit solvation model, namely the polarizable continuum model (PCM) . In order to describe large assemblies, we relied on a QM/QM' ONIOM approach. Within this model it is necessary to define 2 subsystems. The model system, composed of the central phenyl and the amide groups for B4s and the macrolactam-like unit for B9s, will be described at both the high level of theory (ωB97X-D/6-311+G(d,p)) and the low level (HF/3-21G(d)) of theory. The real system, encompassing the model system and the lateral groups of both families, will only be treated at the low level of theory. A charge embedding framework was also added to the hybrid QM/QM' ONIOM scheme . ## Non-covalent interactions It has been possible to visualize non-covalent interactions (NCI) through the use of NCIPLOT code . NCI analysis gives rise to an index that is based on the calculated electronic density and its reduced gradient, represented as a two-dimensional plot. For a given system, there will be a drastic change in the reduced density gradient (RDG) between the atoms that are interacting, leading to density critical points. The latter can be represented on the molecular structure as an isosurface to indicate the region where a weak interaction is occuring. Nevertheless, both attractive (H-Bonds, van der Waals) and repulsive (steric repulsion) interactions can be spotted thanks to this index. By looking at the second derivative of the density and to the sign of its eigenvalue it is possible to distinguish attractive and repulsive interactions. Hence the density and the sign of the eigenvalue of the density second derivative give information about the strength and the type of interaction respectively and one can visualize them via isosurfaces. The electronic density used to compute the NCI index is the one calculated at the same level of theory as the one presented in Section 2.1.1. ## Classical Molecular Dynamics Molecular dynamics (MD) simulations were run with the generalized AMBER force field (GAFF) within GROMACS 2018.3 package . As GAFF is not directly implemented in GROMACS, we used acpypi script to convert the files from AMBER to GROMACS formalism. For each molecule, the atomic charges were derived following the parametrization procedure in GAFF, that is to say using HF/6-31G(d) RESP charges. The validity of the force field was checked by comparing the structures obtained after an optimization process in vacuum with GAFF and with DFT at the ωB97X-D/6-311+G(d,p) level. Results are provided in the Supporting Information section. During the simulations, the system is composed by n organic molecules (n being equal to 1, 2, 3, 5, and 10 and the organic molecules being either from B4s or B9s families). The general philosophy of our molecular dynamics simulations is represented schematically on Figure 2. The size of the simulation box and the number of water molecules depend on the system under investigation. All those information for each system under investigation are gathered in Supporting Information. To describe the electrostatic interactions, periodic boundary conditions were imposed along with a cut-off of 10 and the use of the Particle Mesh Ewald (PME) method . Following a steepest descents minimization, each system was equilibrated in two steps. For the first step, a simulation in the canonical ensemble (NVT) during 100 ps was carried out. The temperature was set to 310 K using the Berendsen weak coupling method . Organic molecules and solvent were coupled to separate temperature coupling baths. For the following second step, simulation under constant pressure (NPT) was performed. To maintain an isotropic pressure of 1 bar, we relied on the previously mentioned weak-coupling Berendsen method. The production phase was then carried out in the same NPT ensemble. Temperature was controlled thanks to the Nosé-Hoover thermostat while the isotropic character of the pressure was maintained via the Parinello-Rahman barostat. Combining this thermostat and barostat ensures the presence of a true NPT ensemble. Simulation time, if not explicitly precised, was set to 10 ns. ## Umbrella sampling Within the umbrella sampling (US) approach we have considered dimers, trimers and pentamers. The systems were placed in a rectangular box that can allow a pulling simulation (e.g a simulation box that is too small will lead to an interaction with the periodic images). In particular, the z length had to be large enough in order to satisfy the minimum image convention. The solvent molecules (water), were described through the TIP3P model. The first step consisted in an NPT equilibration of 100 ps, as it was described above. For the proper pulling simulation, restraints were applied to one of the monomers for dimers, to a dimer for trimers and to a tetramer in the case of the pentamer. The molecules that were restrained were thus considered as immobile references. The molecule that was not restrained was then pulled away from the immobile one, along the z-axis over 500 ps at a rate of 0.1 nm.ps −1 with a spring constant of 250 kJ.mol −1 nm −2 . The final COM (center of mass) distance between the two considered assemblies that was obtained was 4 nm. Snapshots were extracted from this pulling simulation in order to be as many starting points for the different umbrella sampling windows. For COM distances under 1 nm, a separation of 0.05 nm was considered between each window and then, for COM distances above 1 nm and up to 2.5 nm, the spacing between the windows was 0.1 nm. An example of the corresponding histogram is provided in Supporting Information. It allows a smoother and more accurate description of the interaction at small COM distances. This approach leads to around 25 windows. For each window, a 10 ns simulation was performed, resulting in a total simulation time of 250 ns for the US approach, for each assembly of each molecule. The analysis of the results was done using the Weight Histogram Analyzis Method (WHAM), implemented in the GROMACS 2018.3 package . A schematic representation of the umbrella sampling approach is provided in Figure 3. 3 Results and discussion ## Monomers Key structural parameters (see Figure 1 for their definitions) have been selected to study the different monomers. For both B4s and B9s families, three dihedral angles were defined (α, β , and γ), illustrating the relative position of the amide groups with respect to the plane of the molecule. We followed their evolution along the 10 ns MD simulations. The average values are reported in Table 1. Table 1: Average values of the selected dihedral angles (in degrees) defined on Figure 1 along with the RMSD value (in ) for each of the molecules. Standard deviations are provided in parentheses. -90 ( 13) 91 ( 14) 88 ( 13) 2.33 (0.48) B4c -90 ( 13) 89 ( 14) 89 ( 13) 2.08 (0.48) B4p -89 ( 13) 90 ( 13) 90 ( 13) 3.07 (0.56) B4pI -88 ( 14) 90 ( 12) 89 (13) 3.12 (0.51) B9 95 ( 7) 95 ( 6) 96 (8) 0.44 (0.12) B9m 83 ( 7) 83 ( 7) 84 (8) 0.80 (0.13) For the B4s family, as previously mentioned, all the amide moieties are not oriented in the same direction. As illustrated by α that is negative, there is one amide moiety that is antiparallel to the two other ones for B4s. This feature will thus induce a difference for the stacking behavior of the next steps. with perfectly perpendicular amide groups. One has to notice that for the RMSD calculation, the hydrogen atoms were not considered. The results are gathered in Figure 4 and Table 1. this family (0.44 and 0.80 for B9 and B9m respectively). Fluctuations are also small with deviations equal to 0.12 and 0.13 for B9 and B9m respectively. On the other hand, for B4s molecules, the average RMSD is ranging from 2.08 to 3.12 for B4c and B4pI respectively, indicating quite large structural modifications during the simulation. Standard deviations confirm this trend with values around 0.50 for the four molecules of B4s family. Studying the monomers of the different families has highlighted the fact that one family can be considered as rigid (B9s) while the other one appears to be more flexible (B4s). This feature may have an impact if one wants to build larger supramolecular assemblies as it will ease, or not, the formation of hydrogen bonds. The next sections are dedicated to the comparison of the two families for the formation and the stabilities of different SMA. ## Qualitative approach To study the possible self-assembling behavior of the different molecules, we decided to consider the smallest and thus simplest supramolecular entity, namely a dimer. The interaction of the two monomers is ensured by the amide moieties. Indeed it is possible to form a hydrogen bond between the N-H part of the amide of one molecule and the C=O bond of another amide. As previously observed during the study of the monomers, the amide groups are perpendicular to the plan of the molecule and it is thus possible to stack monomers via a network of three hydrogen bonds. The dimers we built are represented on Figure 5. 2. By looking at the average value of d along the MD simulations it is already possible to have a qualitative idea of the stability of the different assemblies. If all the dimers involving monomers from the B4s family are stable, it is not the case for the B9s one. Indeed, the d values that are reported in Table 2 illustrate the fact that the two monomers involved within a dimer are remaining close to each other. If we go deeper in the analysis, we observe that the average value of d, ranging from 3.77 to 4.23 for B4s, is close to the crystalline characteristic value (4.8 ) for such a stacking involving amide-based H-Bonds . One has to notice that the standard deviations presented in Table 2, around 0.2-0.3 , illustrate the fact that the dimers are not "broken" along the simulation. The same conclusion can be drawn for B9m with an average value of 3.82 . As the dimer involving B9 monomers were not stable, it was not relevant to measure d along the simulation. One has to notice that if h 1 , h 2 , and h 3 are equivalent for B9s family, they are not for B4s dimers. Because of the asymmetry in the orientation of the amide moieties, only h 2 and h 3 are equivalent while h 1 is unique. 3) -a For the calculation of the average value and the associated standard deviation, only the areas highlighted on Figure 6 are considered. The same procedure is used for the other molecules, averages are made when the bonds are effective during the corresponding time. Looking at those different values gives insights into the interactions involved in the stability of the dimers. We have represented on Figure 6 the evolution of h i distances for B4. It is clear from Figure 6(a) that the hydrogen bond h 1 is always effective during the 10 ns simulation. Moreover, with an average value of 1.99±0.24 this hydrogen bond is within the range (1.5-3.5 ) of a hydrogen bond with a medium strength (4 to 15 kcal.mol−1). The same conclusion can be drawn for all the dimers of the B4s family. The analysis is quite trickier for h 2 and h 3 . As represented on Figure 6b and c, those H-Bonds are not always effective. The area that are colored represent the moment where h 2 and h 3 distances are within the 1.5-3.5 range. The time corresponding to the colored area are given in Table 2. If h 2 is effective during almost the 10 ns simulation (8.4 ns), it is not the case for h 3 , that is only effective transiently for 3.1 ns. Nevertheless, the dimer of B4 is stable and the stacking is maintained during all the simulation (d distance gives this information). By looking closer to the structure along the trajectory it has been possible to highlight a particular structure (Figure 5(c)). During the simulation the two monomers do not stay aligned. There is a tilt of one monomer with respect to the other one, inducing an hybrid hydrogen bonding. The keto part of one amide points between two amino moieties of two other amides of the other monomer. One expected H-bond is thus presents (h 2 or h 3 ), even if it is longer than a perfectly aligned H-bond, while the other one is new. We have thus introduced two new H-bonds, namely h' and h" (Figure 5(c)). Their evolutions are represented on Figure 6d and e. One can observe that either h' or h" is effective during the simulation. Their average values along with their effective time are gathered in Table 2. It is possible to say that the H-Bonds complement each other. When h 2 is not effective, one can observe that h' is. When h 3 is not effective, h" is taking over. We decided to combine those four H-Bonds on the same graphics (Figure 6(e)). It appears that there is always at least one but most of the time two hydrogen bonds (h 2 , h 3 , h', or h") that is present within the dimer, ensuring its stability along with h 1 . The Boltzmann population ratio for the normal (with h 1 , h 2 and h 3 ) and the hybrid (with h 1 and h 2 /h" or h 3 /h') dimer (see Figure S4 in ESI †) is always around 40/60 for all the dimers of the B4s family. Indicating that the hybrid dimer is more favorable than the normal one. For the other dimers of the B4s family, the same behavior is observed with a compensation of h 2 and h 3 by h' and h" to ensure the global stability of the dimers. For B9m dimer, h 1 , h 2 , and h 3 are effective during all the trajectory but no hybrid H-Bond can be observed as there is no sufficient flexibility for the amide moieties. To summarize the findings about dimers, we have been able to highlight (i) the (non-)stability of B4s family and B9m (B9), (ii) the effectiveness of the hydrogen bond network to build supramolecular assemblies and (iii) an hybrid H-bond pattern allowed by the flexibility and the possibility for B4s dimers to orient their amide moieties. ## Quantitative approach: Umbrella Sampling To go further in assessing the stability of the different dimers, we have undertaken US simulations. It allows us to retrieve the binding free energy, ∆G, along a reaction coordinate, x, that represents the preferential direction for the stacking pattern. Using around 25 sampling windows along this axis, one can construct a one-dimensional potential of mean force (PMF) profile for each system under study, leading to a binding energy, E bind . US simulations for larger assemblies, namely a trimer and a pentamer, were also performed. For each PMF profile, the minimum energy is associated to a particular distance, d com , that can be roughly compared to the d distance discussed in the previous section as it is the distance between the center of masses of the different assemblies. All those values, E bind and d com , for each system and for each molecule, are gathered in Table 3. For the following discussion, the comparison and evolution of E bind will be made on the absolute value. The question that motivated the new simulations involving trimers and pentamers is the following: How is the binding energy evolving when the supramolecular assembly is getting larger? More specifically, is it getting harder to add a monomer to a dimer, a tetramer? The answer to this question will help us to understand the self-supramolecular assembling behavior of B4s and B9s into larger assemblies. The same umbrella sampling approach presented before was used to study first the interaction within a trimer. Two subsystems were considered, a dimer and a monomer and we were looking for the binding energy of the monomer with the dimer. For the pentamer, the binding energy is computed for the interaction of a tetramer and a monomer. The PMF profile for trimers and pentamers, for each molecule, is provided on three possible different behavior are observed, a slight increase of E bind for B4 (+4) and B4pI (+2), a decrease for B4c (-7) and no evolution for B4. One can nevertheless notice that the formation of trimers still remain favorable in all the cases. For B9 and B9m, E bind is slightly increasing (+2) and decreasing (-6) respectively. In conclusion, adding a monomer to an already formed dimer is more favorable for B4s than for B9s. Trimers of B4s appear even more stable than dimers of B9s. Going further and considering pentamers for both families leads to a unique conclusion. The binding energy is always decreasing when compared to the energies obtained for dimers and trimers. Nevertheless, the values obtained for B4s (15, 12, 8, and 22 kcal.mol −1 for B4, B4c, B4p, and B4pI respectvively) are still higher than the ones obtained for B9s (7 and 4 kcal.mol −1 for B9 and B9m respectively), indicating that adding a monomer to a small oligomer of B9s family will be less favorable than for the B4s family. ## Pentamers Classical molecular dynamics simulations involving pentamers have been performed for each of the molecules. Two behaviors were observed for the two families. For B9s, almost no H-bond interactions were maintained throughout the trajectory leading to a non-stability for both of the assemblies. Though, dimers were observed ponctually (see Figure S8 and S9 in ESI †). The conclusions that were drawn in the previous sections are thus confirmed with (1) a low but still possible stability for dimers and (2) an unfavorable binding energy for systems involving a large number of monomers. For B4s family the conclusions are different with respect to B9s. Indeed, for the entire B4s family, the pentamers appeared as stable along the trajectory (Figure 8 for B4p and Figure S5, S6, and S7 in ESI † for B4, B4c, B4pI respectively). The H-Bond network is at the origin of this stability. As for the dimers, h 1 is always effective and can thus be considered as the backbone of the entire supramolecular assembly. Within the pentamer, amide moieties are no longer perpendicular to the plane of the molecules, leading on average to a tilt of 15 • (8(a)). The fact that the amide moieties are quite flexible also allows the formation of the previously mentioned hybrid bonds. During the simulation, there is an alternation of h 2 /h' and h 3 /h" bonds (Figure 8(c)) which implies that the monomers are stacked in staggered rows. The interactions between each pair of monomers and the stability of this interaction is provided by the formation of one expected H-Bond (h 1 ) and an hybrid scheme composed of two H-Bonds (h 2 /h' or h 3 /h"). This finding confirms the importance of the slight flexibility of the amide moieties for the global stability of SMA. It also highlights the fact that the stability of the SMA is dynamic as there is a constant compensation of the H-bonds. ## Aggregation mode Once the strength of the different supramolecular assemblies (dimers to pentamers) have been considered from both a qualitative and quantitative point of view, we decided to have a look at the formation of those assemblies. In the previous sections, we reported data about already formed supramolecular assemblies, in this section, we will study the aggregation and association process, aiming at answering the question: How do we go from monomers to larger assemblies? Different simulations with similar starting configurations were set up as follows to address different problems: -Non-interacting monomers (no H-Bond between them) placed in the center of the simulation box for a long (100 ns) simulation time. -Non-interacting monomers (no H-Bond between them) placed in the center of the simulation box for 100 short (1 ns) simulation times. The first type of simulation will allow us to know if SMA can self-assemble spontaneously and if this SMA will be stable along a long simulation time. The second type of simulation will provide information on the frequency of formation of such SMA and more precisely the frequency of formation of H-Bonds. For simplicity, results and detailed analysis on B4p are presented here and in Supporting Information for the others. ## One long simulation We used PACKMOL to generate a starting configuration encompassing 10 monomers that are loosely compacted, meaning they are relatively close to each other but with no hydrogen bonds or particular interaction between them. They were then placed in the center of a simulation box for a 100 ns long simulation time. To ensure that the monomers are not interacting at the beginning of the simulation and more particularly that no hydrogen bond is effective for the starting configuration, we represented the radial distribution function (RDF) for the oxygen and hydrogen atoms involved in those interactions (see Figure 9 g(r) On Figure 10 we have provided a representative structure of a SMA of B4p, involving 8 monomers, that has been formed during the simulation. One has to notice that the two remaining monomers are involved in a dimer that is not interacting with the octamer. To illustrate the interactions between the different monomers we also provided interaction surfaces extracted from the NCI analysis for each pair of monomers. Various If we go deeper in the analysis of the different interactions, it is possible to observe that all the interactions are not formed simultaneously but sequentially. Indeed the first interaction that is effective within the SMA is the h 1 bond, with a formation at the nanosecond timescale (see Figure 11). We then have the formation of the previously mentioned interactions (h 2 and h') at a slightly higher timescale (few nanoseconds), followed by N-H ••• π interactions and O 1 -H 2 /H 3 after tens of nanosecond. If h 1 is the first interaction being observed, it is also the most stable one as it is effective during almost the entire trajectory (Figure 11(b)). For other molecules of the B4s family, there is always the formation of a SMA involving at least five monomers (B4 and B4pI). B4p, with 8 monomers involved is the most efficient one while B4c forms a seven-members SMA (see Figure S10, S11, and S12 in ESI †). The same kind of interactions are present within all the SMA with a predominance for the h 1 bond, completed with the previously mentioned interactions. On the opposite, when considering B9 and B9m molecules, no large SMA were detected and only poorly stable dimers were observed. As a conclusion one can say that the self-assembling process is (1) efficient for B4s family but not for the B9s one, (2) quite fast, of the order of the nanosecond, (3) can be a long process as the interactions are added sequentially, and (4) dynamic in the sense that some interactions (N-H ## Many short simulations We observed in the previous section that, when a H-Bond is formed, it is then quite stable along the trajectory. The analysis we propose to perform in this section aims at retrieving the frequency of formation of the H-Bond. To do so, we generated 100 different starting configurations encompassing 10 monomers close enough to each other but with no H-Bond between them and performed a 1 ns simulation. We then extracted the final structure of each simulation and counted the H-Bonds between all the monomers. One has to notice that no distinction was made between all the possible H-Bonds (9 in total). On Figure 12 we have represented, for the 100 simulations and for B4p and B9m only, the total number of H-Bonds that have been observed between each pair of dimers. The results for B4, B4c, B4pI and B9 are also provided in ESI † (Figure S13 and S14). ..,10) for B4p (bottom left, green) and B9m (up right, gray) for a total of 100 simulations. The structures that are considered for the count are the final ones obtained at the end of the 1 ns simulation. The size of the dots is proportionnal to the number of H-Bonds that have been detected. For the largest dots, we provided the corresponding exact number of interactions. One can observe immediately on Figure 12 that the number of H-Bonds present within B4p aggregate is larger than the ones of B9m. Due to the particular arrangement of the molecules, some preferential interactions are observed. For example, for B4p, the formation of H-Bond between monomer 1 and monomer 2 is almost systematic. Indeed, the count reveals 143 H-bonds between those 2 monomers for a maximum of 300. Nevertheless, one can also observe that monomer 1 is also involved in interactions with monomer 5 (52) and in a lesser extent with monomer 6 (31) and 7 (18). So does monomer 2 with monomer 7 (17), 8 (43) and 10 (29). For B4p, there are six main interactions involving M1-M2, M3-M4, M4-M10, M10-M7 M5-M8, and M6-M9. For information, all the monomers are staying close to each other, as a loose aggregate, during all the simulations. This can be explained by the fact that other interactions (C-H ••• π and N-H ••• π) are formed and thus enforce the stability of the entire supramolecular assembly. For B9m, no systematic interaction was observed with a maximum count observed for a M6-M9 interaction. This result may also be due to the fact that B9m monomers are not staying "packed" during the simulation. The loose aggregate is thus not even stable for B9m molecules. ## Conclusion By defining a complete theoretical protocol based mainly on molecular dynamics simulations and aiming at studying the ability of organic molecules to form supramolecular assemblies and their resulting stabilities, we have been able to provide some hints for an effective self-supramolecular assembling process. A total of 6 molecules (B4, B4c, B4p, B4pI, B9, and B9m) divided into 2 families (B4s and B9s), bearing a different backbone but with three amide moieties in common have been considered. The study of the monomers allowed us to validate our molecular dynamics approach and also to understand the properties of the molecules when they are isolated. Considering the dimers has permitted to define the hydrogen bonds network. Expected hydrogen bonds (h 1 , h 2 , and h 3 ) were always observed for both families. Nevertheless, they were not always effective all together and an hybrid scheme (h' and h") has been highlighted. This hybrid H-bond network was observed only for one of the two families, namely B4s. The calculation of the binding energies clearly showed that dimers and even trimers or pentamers of B4s family were more favorable than those of the B9s. Nevertheless, the evolution of the binding energy going from dimers to pentamers indicates that small oligomers (e.g trimers) may be more stabilized than larger SMA (e.g pentamers). Looking at the the formation of the assemblies starting from a loose aggregate allowed us to observe that not only H-bonds can ensure the stability of the aggregate but also N-H • • • π or C-H • • • π in a lesser extent. Finally, the most important point to consider for the self-assembling process is the dynamical behavior of the stability. If SMA of B4s family are more stable than those of B9s it is because they have a relative flexibility. One can also notice that the stability is not something that is frozen but that is also dynamical. We have shown within this study that considering relatively flexible molecules, instead of rigid ones, is a better strategy for the conception of supramolecular assemblies. ## Supporting Information See supporting information for: Validation of GAFF parameters; General parameters for molecular dynamics simulations; Parameters for the umbrella sampling approach; Particular structure of B4s and B9s families; Hybrid hydrogen bond network; Pentamers for B4, B4c, B4pI, B9, and B9m; Interaction maps for B4, B4c, B4pI and B9.

chemsum

{"title": "COMPUTATIONAL PREDICTION OF THE SUPRAMOLECULAR SELF-ASSEMBLING PROPERTIES OF ORGANIC MOLECULES: FLEXIBILITY vs RIGIDITY A PREPRINT", "journal": "ChemRxiv"}

visible-light_induced_singlet_nucleophilic_carbenes:_rapid_and_mild_access_to_fluorinated_tertiary_a

## Abstract: Singlet nucleophilic carbenes (SNCs) that contain only one heteroatom donor remain underexplored and underutilized in chemical synthesis. To discover new synthetic strategies that harness these SNCs as reactive intermediates, aromatic or aliphatic siloxy carbenes represent excellent model substrates as they can be readily generated photochemically from stable acyl silane precursors. We herein report the discovery that photochemically generated siloxy carbenes undergo 1,2-carbonyl addition to trifluoromethyl ketones, followed by a silyl transfer process to afford benzoin-type products. This new transformation is a rare example of the use of ketones as trapping reagents for SNC intermediates and delivers an efficient, user-friendly and scalable process to access fluorinated tertiary alcohol derivatives driven by only light, circumventing the use of catalysts or additives. Carbenes are neutral reactive intermediates that possess six valence electrons, four of which occupy two sigma-bonding orbitals while the remaining two occupy either one or both of the non-bonding orbitals. 1 Carbenes are important reactive intermediates in organic synthesis, where triplet carbenes (unpaired non-bonding electrons) exhibit reactivity akin to radical intermediates and singlet carbenes (paired non-bonding electrons) can display either electrophilic or nucleophilic reactivity. 2 Electrophilic singlet carbenes, typically generated from diazo precursors, have been widely exploited in organic synthesis in processes including C-H insertion and cyclopropanation reactions. 3 Highly stabilized singlet nucleophilic carbenes (exemplified by NHC-based organocatalysts and ligands) have also been extensively utilized in synthesis with chiral NHC derivatives employed to catalyse a variety of chemical reactions with high chemo-and enantioselectivity. 4 Comparatively, the application of singlet nucleophilic carbenes (SNCs) containing only one heteroatom substituent in chemical synthesis remains limited. Examples of such SNCs include cyclic oxacarbenes generated photochemically from cyclobutanone, 5 aryl alkoxy carbenes, 6 aryl amino carbenes 7 and siloxy carbenes. 8 When considering the reaction of singlet carbenes with ketones, electrophilic carbenes undergo an O-insertion process whereby the oxygen atom of the carbonyl group donates electrons into vacant 2p orbital of the carbene to generate a carbonyl ylide (see Figure 1). The ylide is available to react in subsequent processes including dipolar cycloadditions to generate chemotypes including oxiranes and furans. 9 While the reactivity of nucleophilic carbenes with aldehydes, esters and acyl fluorides is well-established to generate species including Breslow, acyl azolium and diazolium enolate intermediates, 4e-h the reaction of nucleophilic carbenes with ketones remains relatively unexplored. To date, few reports have described the reaction of stabilized singlet nucleophilic carbenes (e.g. NHCs) with ketones. 10 For nucleophilic carbenes containing one heteroatom donor, there exists only two single examples from 1971 (Figure 1), where Brook and co-workers reported that a siloxy carbene reacted with acetone to afford an oxirane or with cyclohexanone to afford a silyl enol ether derivative (Figure 1). To gain new insight into this underexplored mode of reactivity for carbene intermediates, we set out to conduct a detailed investigation into the reaction of SNC intermediates with ketones. While oxacarbenes, aryl alkoxy carbenes, and aryl amino carbenes would all be suitable SNCs for such a study, we focused on siloxy carbenes as model substrates for studying partially stabilized SNCs as these intermediates can be readily generated from stable acyl silane precursors via irradiation with visible light, circumventing the inclusion of catalysts or additives in the reaction mixture. 1d,12 We envisaged that if successful, the photochemical addition of singlet nucleophilic carbene intermediates to ketones would deliver a new metal-free approach to the synthesis of tertiary alcohol derivatives, an important chemotype in organic synthesis and pharmaceutical sciences. 13 Our initial studies involved irradiation of benzoyltrimethylsilane (1a) with blue LEDs (427 nm, 40W) in the presence of two equivalents of various ketones including acetone, cyclobutanone, cyclohexanone, acetophenone and 2,2,2-trifluoroacetophenone in dichloromethane. Only the reaction with 2,2,2-trifluoroacetophenone led to quantifiable amounts of product, whereby after only 2 hours, complete consumption of the acyl silane was observed and trifluoromethyl alcohol derivative 3a was isolated in 82% yield. A solvent screen was subsequently performed which revealed n-hexane to be the optimal solvent for this reaction (affording 3a in near quantitative yield), however THF, diethyl ether and chloroform also proved suitable. The inclusion of 4 molecular sieves also improved the reaction by limiting hydrolysis of the acyl silane to the corresponding aldehyde which can occur when the carbene reacts with adventitious water in the reaction mixture. Furthermore, experimental controls revealed that in the absence of light no reaction was observed. We subsequently investigated the UV/Vis absorption profile of both the individual reaction components and the reaction mixture (see supporting information). The trifluoromethyl ketone exhibits absorption at wavelengths below 300 nm whereas acyl silane 1a absorbs light in the visible region (max = 420 nm) facilitating selective excitation of the carbene precursor by irradiation at 427 nm. Due to the absence of spectral shifts in the absorbance profile of the mixture, there was also no evidence for the formation of an electron-donor acceptor (EDA) complex between the two reagents. A key consideration when employing siloxy carbene intermediates is that following nucleophilic carbene addition to a electrophile such as a ketone, an oxonium ion is formed that contains two potential sites of subsequent reaction. Following carbene addition to the trifluoromethyl ketones, the zwitterionic species formed can react in two possible manners, either via oxyanion addition to the base of the oxonium ion to afford an oxirane (for an example see Figure 1), or the oxyanion can abstract the silyl group from the oxonium ion to regenerate the carbonyl system, the latter of which is operating during the reaction described herein. The structure of the ketone product was confirmed by 13 C NMR analysis with key spectral characteristics including a resonance at  = 192 ppm assigned to the new ketone motif, a quartet at  = 121 ppm (J = 286 Hz) assigned to the CF3 group, and a quartet at  = 83 ppm (J = 26 Hz) assigned to the new tetrasubstituted carbon center formed which is bonded to the CF3, silyl ether and ketone functional groups. A shift of the CF3 resonance from  = -69 to -74 ppm from the 2,2,2-trifluoroacetophenone starting material to the product can also be observed following analysis by 19 F NMR. With high-yielding conditions in hand, the photochemical reaction process was explored using different acylsilane and trifluoromethyl ketone substrates (Scheme 1). Initially, variations in the silyl group were explored with the trimethyl, triethyl and tert-butyldimethyl silyl analogues all affording the product in high yield (3b-d), however the tert-butyldimethyl silyl analogue required a longer reaction time (6h) to achieve full conversion and was isolated in a reduced yield (88%). The corresponding triisopropyl silyl analogue failed to react under the standard conditions. Variations in the aryl component of 1 were then explored with 3and 4-monosubstituted aroyl silanes containing methyl, methoxy, chloro or fluoro substituents all performing well (Scheme 1, 3e-h, 3k-m, 78-99%). Di-and tri-substituted aroyl silanes also afforded the corresponding fluorinated tertiary alcohol derivatives 3i and 3j in excellent yield. Trifluoroacetophenone derivatives containing additional substituents in either the 3-or 4-aryl position were then subjected to reaction with acyl silane 1a under the standard photochemical conditions to afford the corresponding benzoin-type adducts in exceptional yield (Scheme 1, 3n-3t, 87-99% yield). To further explore the reactivity of photochemically generated nucleophilic carbene intermediates with ketones, a series of aroyl silanes 1 were reacted with an aliphatic ketone in the form of 1,1,1trifluoroacetone 4 (Scheme 2). This reaction also proceeded with high efficiency, affording to corresponding ketones 5a-f in near quantitative yield. Again, diverse substitution patterns on the aromatic ring of the aroyl silanes were tolerated (Scheme 2). To this point, attempts to react the carbene intermediate generated from aliphatic acyl silanes with 2a or 4 via direct irradiation (using 370, 390 and 427 nm LEDs) or triplet energy transfer (employing 2 mol% of [Ir(dF(CF3)ppy)2(dtbbpy)]PF6 at 440 nm) proved unsuccessful. 8g,h To further probe this photochemical process a series of additional experiments were conducted (Figure 2). Aroyl silanes were reacted with 2-chloro-2,2-difluoroacetophenone to demonstrate that other halogenated ketone derivatives are applicable in this reaction process (Figure 2a). Subsequently, a competition experiment between aroyl silanes containing either electron withdrawing or electron donating substituents was conducted by irradiating a solution containing 0.25 mmol each of aroyl silanes 3g and 3i in the presence of 0.25 mmol of 2,2,2-trifluoroacetophenone for 2 hours at 427 nm. Scheme 2. Photochemical reaction of siloxy carbenes with trifluoroacetone. Reaction conditions: acyl silane (0.25 mmol), 1,1,1-trifluoroacetone (0.50 mmol), n-hexane (1.0 mL), 4 MS (~250 mg) irradiated at 427 nm for 2h. The conversion to the product was analyzed using 1 H NMR which revealed that the carbene generated from the 4-chlorophenyl acyl silane trapped 68% of the available ketone with the remaining 32% trapped by the carbene generated from the electron rich acyl silane. This result infers that either formation of the carbene from acyl silane 3i is more facile or that the carbene generated from 3i is in fact more reactive. According to the Beer-Lambert law, light transmittance decreases exponentially as distance from the photon source increases. It is well established that the use of flow chemical technology (where reactions are conducted in microchannels or microtubing) can significantly increase the efficiency of photochemical processes leading to shorter reaction times and decreased by-product formation. 14 Thus, the opportunity to increase the efficiency and scalability of the process described herein was explored in flow by pumping a 0.25M solution of the reagents in n-hexane through a 4 mL reactor coil irradiated by 427 nm LEDs (40 W). Initially, optimization of the reaction time was conducted with 99% and 100% conversion achieved at 15-and 20-minute reaction (residence) times, respectively. Attempts to reduce the residence time to 10 minutes led to a decreased conversion of 96%, and irradiation of the 4 mL reactor coil with two 427 nm LEDs for 10 mins failed to increase the conversion. Finally, a reaction time of 5 minutes employing two photoreactors resulted in 75% conversion of the acyl silane to the product. With optimized flow conditions in hand, a 0.25M solution of anhydrous n-hexane containing 7 mmol of acyl silane and two equivalents of the 2,2,2-trifluoroacetophenone was pumped in a continuous fashion through the 4 mL reactor coil employing a residence time of 15 minutes. Following the completion of this process, evaporation of the volatile components and recrystallisation from n-hexanes afforded 2.32 g of 3a in 94% isolated yield (Figure 2c). A crystal structure of 1a confirmed the structure of the tertiary alcohol derivative generated from the photochemical bond-forming process (Figure 2f) and the fluorinated silyl ether products could be further manipulated through cleavage of the silyl group to afford the tertiary alcohol, and reduction to produce the diol (Figure 2d). Finally, the difference in reactivity between an electrophilic carbene and nucleophilic carbene intermediate with the same trifluoromethyl ketone is depicted mechanistically in Figure 2e. In 2017, Jiang and co-workers reported that the electrophilic Pd-carbene intermediate generated from an aryl alkyl tosyl hydrazone underwent an O-insertion process with the carbonyl group of 2,2,2-trifluoroacetophenone to afford a carbonyl ylide which further reacted to generate the oxirane product. 15 For the new transformation described herein, reaction of the photochemically generated nucleophilic carbene intermediate occurs at the carbon-atom end of the trifluorocarbonyl group which gives rise to a zwitterionic intermediate affording the benzoin-type product after silyl transfer. In summary, we have discovered a new process for the formation of tertiary alcohol derivatives employing fluorinated ketone derivatives as trapping reagents in the presence of photochemically generated nucleophilic carbene intermediates. To note, the formation of related products has been achieved employing reactions that proceed via acyl anion intermediates generated from the addition of an NHC catalysts to an aldehyde or a cyanide reagent to acyl phosphonates. 16 Advantageously, the protocol described herein proceeds In addition, this new transformation represents a unique example of SNCs reacting with ketones and importantly, providing additional insights into the properties and reactivity of the relatively underexplored class of singlet nucleophilic carbene intermediates that contain one-heteroatom donor.

chemsum

{"title": "Visible-Light Induced Singlet Nucleophilic Carbenes: Rapid and Mild Access to Fluorinated Tertiary Alcohol Derivatives", "journal": "ChemRxiv"}

exfoliated_black_phosphorous-mediated_cuaac_chemistry_for_organic_and_macromolecular_synthesis_under

## Abstract: The development of long-wavelength photoinduced copper-catalyzed azide-alkyne click (CuAAC) reaction routes is attractive for organic and polymer chemistry. In this study, we present a novel synthetic methodology for the photoinduced CuAAC reaction utilizing exfoliated two-dimensional (2D) few-layer black phosphorus nanosheets (BPNs) as photocatalysts under white LED and near-IR (NIR) light irradiation. Upon irradiation, BPNs generated excited electrons and holes on its conduction (CB) and valence band (VB), respectively. The excited electrons thus formed were then transferred to the Cu II ions to produce active Cu I catalysts. The ability of BPNs to initiate the CuAAC reaction was investigated by studying the reaction between various low molar mass alkyne and azide derivatives under both white LED and NIR light irradiation. Due to its deeper penetration of NIR light, the possibility of synthesizing different macromolecular structures such as functional polymers, cross-linked networks and block copolymer has also been demonstrated. The structural and molecular properties of the intermediates and final products were evaluated by spectral and chromatographic analyses. ## Introduction For the last decade, click chemistry has been recognized as an indispensable part of synthetic chemistry due to its easiness of application, efficiency to produce the targeted products with very high yields and little or no byproducts under a variety of conditions, and high interconnected group tolerance. Since the introduction of click chemistry by Sharpless and Mendal , many studies have been dedicated to better understanding of the concept and expanding its scope to be applied in various fields of chemistry including bioconjugation , drug discovery , materials science and so on . The development of the use of light in click chemistry has set a milestone as a new and effective method for the synthesis of macromolecules . Initiation of this reaction photocatalytically provides many advantages for the synthetic methodologies including bioconjugation, labeling, surface functionalization, dendrimer synthesis, polymer synthesis, and polymer modification by adding spatial and temporal control . In recent years, heterogeneous photocatalysts have been performed in many photosynthetic reactions since they provide a more reasonable and easy way to synthesize the targeted products compared to the classical homogenous photocatalysts. In this respect, 2D materials offer great potential due to converting the inexhaustible energy of sunlight into chemical and electrical energy along with having a less environmental impact. After the discovery of the photocatalytic effect of 2D materials under UV light the heterogeneous photocatalysts have been successfully applied in both small-and large-scale synthesis such as organic reactions , free radical polymerization (FRP) , controlled radical polymerization (CRP) , CuAAC chemistry , and thiol-ene chemistry . However, most of the conventional 2D materials have a wide bandgap that requires UV light irradiation for their activation. Since 94% of the rays from the sun are not sufficient to activate these conventional semiconductor materials, many strategies have been proposed to design photocatalysts that can harvest in a wide spectrum of sunlight, especially in the NIR region . In particular, the development of new photocatalyst systems that absorb the incident light from the sun at much longer wavelengths have aroused widespread interest . However, the most of the NIR photocatalysts applied exhibit relatively low catalytic efficiency due to their low absorption characteristics and require complicated synthetic procedures. In this respect, it is worth to mention that elemental 2D materials with a proper bandgap and charge mobilities have been shown to act as photocatalysts in several reactions . Exfoliated black phosphorus (BP), the most stable allotrope of phosphorus, has been shown as a highly efficient photocatalyst possessing superior features in many respects . BP, a vital semiconductor 2D material with excellent physicochemical properties such as high carrier mobility, tunable optical absorption, and novel electronic band structure, fills the gap between graphene and wide bandgap semiconductors . Furthermore, BP shows a layer thickness tunable bandgap ranging between 0.3 and 2.1 eV. Therefore, BPNs can efficiently be applied as a photoredox catalyst with broadband solar absorption [34, . The use of 2D materials for the photoinitiated electron transfer reactions with Cu II catalysts for the photoinduced atom transfer radical polymerization (ATRP) and CuAAC reactions prompted us to develop a new photoredox system that works under NIR irradiation for the CuAAC reaction. In this work, we report a new synthetic strategy to the photochemical reduction of Cu II to Cu I for the CuAAC reaction using BPNs as the photo-initiator under NIR light. ## Results and Discussion The detailed preparation and characterization of the initial BP crystals and BPNs were previously reported . BPNs were tested as NIR photoinitiator for the CuAAC reactions of low molar mass compounds and polymers possessing antagonist azide and alkyne functionalities (Figure 1). The optical absorption spectra of BPNs, copper(I) chloride (Cu I Cl, 0.05 mmol) and copper(II) chloride (Cu II Cl 2, Initially, the model reaction between benzyl azide (Az-1) and phenylacetylene (Alk-3) in the presence of copper(II) chloride/ N,N,N',N',N''-pentamethyldiethylenetriamine (Cu II Cl 2 / PMDETA) and exfoliated BPNs under the white LED irradiation was performed (Figure 3). The reaction was followed by 1 H NMR spectroscopy during the click process. The decrease of the acetylene proton at 4.42 ppm and appearance of the new signal at 8.67 ppm corresponding to the triazole moiety con-firmed successful click reaction under white LED exposure conditions after 4 h (Figure 3a). Kinetic studies conducted by 1 H NMR analysis confirmed that the click reaction between benzyl azide and phenylacetylene resulted in almost complete conversion within 4 h white LED irradiation (Figure 3b). In this connection, it should be pointed out that the reaction proceeds also in dark almost at the same rate (Supporting Information File 1, Figure S4). This is an expected observation because there is no back reaction to reform Cu(II). Similar observations were reported by the other photoinduced CuAAC reactions . In order to demonstrate the functional group tolerance, the extent of the reaction was investigated on various alkyne groups using benzyl azide under both white LED and NIR light irradiation. The results presented in Table 1 revealed that NIR-lighttriggered click reactions produced the corresponding products with slightly higher yields favored by the higher penetration of NIR light to the reaction media containing heterogeneously dispersed BPNs. Compared with propargyl alcohol (Alk-1) and propargyl acrylate (Alk-4), the rate of clicking slightly decreased in the case of propargylamine (Alk-2), but still gave high yields. Therefore, it can be concluded that Alk-2 and Alk-1 exhibit relatively lower efficiency probably due to the additional coordination of the Cu I catalyst. Notably, the reaction with Alk-4 gave higher yields with both light sources. In the light of previous studies, a photoinduced electron transfer mechanism presented in Scheme 1 can be proposed. Upon the light irradiation, BPNs absorb the light and generate a single electron which was transferred from the conduction band to the Cu II complex to form Cu I capable of catalyzing the click reaction in a conventional manner. Scheme 1: Proposed mechanism for photoinduced CuAAC reaction using exfoliated BPNs. The applicability of the described click reaction to synthetic polymer chemistry was also demonstrated. For this purpose, polymer functionalization by using alkyne functional poly(εcaprolactone) (PCL-Alk) and 9-(azidomethyl)anthracene (Az-2) as click components was investigated. The detailed 1 H NMR spectrum of the resulting anthracene functional polymer (PCL-Anth) exhibited the characteristic signals of triazole and benzylic protons at 5.5 ppm and 8.70 ppm, respectively (Figure 4a). The obtained polymer has similar absorption characteristic to bare anthracene (Figure 4b). The fluorescence spectrum of diluted solution of PCL-Anth in THF excited at λ exc = 350 nm showed the characteristic emission bands of the excited (singlet) anthracene at 595, 655, and 725 nm (Figure 4c). These observations clearly confirmed the successful chain-end functionalization. In addition, block copolymer formation via NIR activated CuAAC process between the polymers having antagonist click components, namely, polystyrene azide (PS-Az) and PCL-Alk, was investigated. At the end of irradiation in the presence of exfoliated BPNs and Cu II Cl 2 /PMDETA, polystyrene-b-poly(εcaprolactone) (PS-b-PCL) is selectively formed (Scheme 2). Figure 5a displays the GPC traces of precursors PS-Az, PCL-Alk, and the block copolymer PS-b-PCL. As it can be seen, the trace of Ps-b-PCL block copolymer was clearly shifted to higher molecular weight region without contamination of the precursor polymers. The 1 H NMR spectrum of the block copolymer displayed the characteristic peaks of both macromolecular segments. Additionally, the methylene protons adjacent to the triazole ring at 7.48 ppm were noted (Figure 5b). These results indicated that structurally diverse polymers formed by different polymerization mechanisms can readily be linked just by a simple NIR-induced CuAAC reaction. The macromolecular scope was further extended to the preparation of cross-linked materials. Thus, the formulations containing bisphenol A di(3-azido-2-hydroxypropan-1-ol) ether (Az-3), and 1-(prop-2-yn-1-yloxy)-2,2-bis((prop-2-yn-1-yloxy)methyl)butane (Alk-5) as multifunctional click components were irradiated in the presence of BPNs and Cu II ligand under NIR light. The gelation was completed after 24 h (Scheme 3). The photocuring process was also followed by differential scanning calorimetry (DSC). The DSC thermogram shows two exo- thermic peaks at 220.38 and 241.74 °C, corresponding to the photo click cure reaction in two stages (Figure 6a). Since a complete reaction of all the azide groups could not occur during the dynamic ramping of temperature, the residual azide groups decomposed at higher temperature. The IR spectrum of the cross-linked polymer further demonstrates the formation of a triazole ring by the decrease of the azide peak at 2100 cm −1 (Figure 6b). Representative TEM images recorded at different magnifications of the resulting cross-linked polymer are shown in Figure 7. From the TEM images, it can be concluded that the process leads to the formation of BPNs-embedded cross-linked polymers. The darker regions circled with yellow dashed line in Figure 6a were attributed to the BPNs while the other relatively lighter regions were ascribed to the cross-linked polymer. To further prove the existence of BPNs in the cross-linked structure, a high-angle annular dark-field scanning TEM (HAADF-STEM) image and the associated elemental mapping images for C, N, and P were recorded and depicted in Figure 7c and 7d. The elemental mapping images adequately demonstrated the presence and the distribution of P atoms that are attributed to BPNs in the cross-linked polymer in addition to C and N atoms (Figure 7d). In contrast to the cross-linked polymer, the distribution of BPNs in the block copolymer structure could not be visualized by TEM, HAADF-STEM, and elemental mapping images (Supporting Information File 1, Figures S5 and S6). This behavior is expected since BPNs are immobilized between the interconnected chains in the cross-linked structure. ## Conclusion In conclusion, we have demonstrated the use of BPNs as an efficient photoinitiator for the photoinduced CuAAC reactions under white LED and NIR light irradiation. The described method is applicable to organic and macromolecular syntheses. NIR irradiation appeared to be more efficient compared to the while LED due to the higher penetration in the dispersed media. In macromolecular syntheses, polymer chain-end functionalization, block copolymer formation of structurally different polymers and cross-linking polymerization can successfully be achieved by using suitably selected click components. This new method would dramatically extend the applications of photoin-duced CuAAC reactions, particularly when the components are light sensitive at short wavelength region and spatial control is required. Characterizations 1 H NMR spectra were recorded at room temperature at 500 MHz on an Agilent VNMRS 500 spectrometer. Gel permeation chromatography (GPC) measurements were performed on a TOSOH EcoSEC GPC system equipped with an auto sampler system, a temperature-controlled pump, a column oven, a refractive index (RI) detector, a purge and degasser unit and a TSKgel superhZ2000, 4.6 mm ID × 15 cm × 2cm column. ## Experimental Materials Tetrahydrofuran was used as an eluent at a flow rate of 1.0 mL/min at 40 °C. The refractive index detector was calibrated with polystyrene standards having narrow molecular-weight distributions. The data were analyzed using Eco-SEC analysis software. A Hitachi HT7700 (TEM) with EXALENS (120 kV) working at a high-resolution (HR) mode was used to obtain transmission electron microscopy (TEM) images, high-angle annular dark field (HAADF) scanning transmission microscope (STEM) images and the associated EDS elemental mapping images. ## Synthesis of black phosphorus crystals and preparation of its nanosheets Black phosphorus (BP) was prepared using a modified lowpressure chemical vapor transport method . For the synthesis, 500 mg of red phosphorus, 20 mg of Sn and 10 mg of SnI 4 were placed into a quartz ampoule with the dimensions of 20 cm length and 1.5 cm width. The air was evacuated by vacuum, and the ampoule was left to dry at least for 30 min under vacuum. The sealed ampoule was placed horizontally in a muffle furnace. The applied heating program was as follows: firstly, the temperature raised to 893 K in 5 h and kept at this temperature for 5 h. Next, the temperature was lowered to 758 K in the span of 6 h and the temperature was kept at this temperature for 2 h. Finally, the oven was cooled to 393 K in 5 h, and it was left for natural cooling afterwards. After the heating process, the ampoule was cracked in dry toluene and the crystalline BP was separated. In order to remove surface impurities, the BP crystals were transferred into absolute ethanol and sonicated for 30 minutes. The sonicated crystals were carefully transferred to a Schlenk tube and dried under vacuum. The Schlenk tube was filled with argon and crushed under inert atmosphere. The produced BP crystals were stored under vacuum. BP nanosheets were prepared by the liquid phase exfoliation of BP crystals. A specific amount of BP was dispersed thoroughly in DMSO by a sonication bath (200 W) for 10 h at 6 °C. The resulting BP nanosheets dispersion was kept under an inert atmosphere for the further use. ## Preparation of azide and alkyne derivatives Synthesis of benzyl azide (Az-1) A literature procedure was used . Product was obtained pale yellow oil, yield 96%. ## Synthesis of (azidomethyl)anthracene (Az-2) A literature procedure was used . 9-Hydroxymethylanthracene (7.40 mmol, 1 equiv) was added to DCM (50 mL) and cooled to 0 °C. Then, SOCl 2 (1.5 equiv) was slowly introduced to the reaction media and allowed to warm up to room temperature while being stirred for 1 h. The solvent was removed under vacuum and the residue redissolved in DMF (10 mL). Following dissolution of the compound, NaN 3 (1.5 equiv) was added, and the reaction was stirred at 50 °C. After 1 h, the reaction mixture was allowed to cool down, diluted with water and extracted with EtOAc. The combined organic phases were washed with brine, dried with anhydrous MgSO 4 , filtered, and concentrated under vacuum. Brownish yellow crystalline solid, yield = 93%. Synthesis of bisphenol A di(3-azido-2-hydroxypropan-1-ol) ether (Az-3) Diazido monomer, bisphenol A di(3-azido-2-hydroxypropan-1ol) (Az-3) was synthesized according to a described method . Az-3 was obtained as light yellowish viscous oil and was directly used without further purification, yield 98%. 1 Synthesis of 1-(prop-2-yn-1-yloxy)-2,2-bis((prop-2yn-1yloxy)methyl)butane (Alk-5) A literature procedure was followed . The crude obtained product was then purified using column chromatography to give a clear oil, yield 70%. 1 Synthesis of ω-azido terminated polystyrene (PS-Az) ω-Bromo functional polystyrene was synthesized by ATRP according to a reported procedure . In a flask equipped with a magnetic stirrer, PS-Br (1 equiv) and sodium azide (5 equiv) were dissolved in 5 mL DMF. The reaction mixture was stirred at room temperature 24 h, then precipitated in 10-fold excess of methanol, filtered and dried in vacuum to yield PS-N 3 . Yield 95% (M n,GPC : 1589 g•mol −1 , M w /M n : 1.13). FTIR: 2096 cm −1 . ## Synthesis of acetylene-terminated poly(ε-caprolactone) (PCL-Alk) Acetylene-terminated PCL-Alk was synthesized according to a modified procedure . To a Schlenk tube, 3-butyn-1-ol was dissolved in ε-caprolactone and heated to 110 °C under nitrogen. After the reaction mixture warmed up homogeneously, one drop of tin octoate was added to the reaction media and the solution was stirred for 3 hours. The obtained polymer was dissolved in chloroform and precipitated in methanol:water (2:1) to yield poly(ε-caprolactone). White solid, (85%) M n,GPC : 1576 g•mol −1 , M w /M n : 1.2. FTIR: 2102 cm −1 . ## Synthesis of organic molecules For the first step of the reaction an appropriate amount of black phosphorus was exfoliated in DMSO-d 6. In a typical experiment, exfoliated BP in DMSO-d 6 (0.5 mL) and azide compound (1 mmol, 1 equiv) were added to a NMR tube containing Cu (II) Cl 2 (0.05 equiv), PMDETA (0.1 equiv). After 5 min, alkyne derivative (1 mmol, 1 equiv) was added slowly to the NMR tube. The reaction tube was irradiated by using a Philips 150 W PAR38E E27 halogen pressure glass type bulb with strong IR-A (NIR) emission. The light intensity inside the reaction tube was ≈200 mW•cm −2 . The light bulb was attached to the top of a photoreactor setup equipped with a large air cooling fan and the reaction temperature was kept constant at room temperature (24−25 °C). 1 H NMR spectra were recorded 4 h later. Synthesis of anthracene functional poly(ε-caprolactone) (PCL-Anth) The same process as in the block copolymerization was applied. Az-2 (19.27 mg, 1 equiv), PCL-Alk (1 equiv), CuCl 2 (1 equiv) and PMDETA (1 equiv) were placed in a Schlenk tube. The tube was degassed by three freeze pump-thaw cycles. Then the tube was irradiated under NIR light for 48 h. After the given time, the mixture was diluted with THF and the copper complex was removed by passing through a neutral alumina column. Excess amount of THF was evaporated by a rotary evaporator. After precipitation of the mixture to cold methanol, the polymer was collected by filtration and dried under vacuum overnight. 1 H NMR was demonstrated in Figure 4. ## Synthesis of polystyrene-b-poly(ε-caprolactone) (PS-b-PCL) Firstly, under dark conditions BP was exfoliated in dry DMF by a sonic bath for 8 h at 10 °C. Subsequently, the solution was transferred into a centrifuge at 2500 rpm for 15 min. Terminally, this exfoliated BPNs with PS-Az (200 mg, 1 equiv), Cu II Cl 2 (1 equiv), PMDETA (1 equiv) and PCL-Alk (1 equiv) were placed in a Schlenk tube. The tube was degassed by three freez-pump thaw cycles. Then the tube was irradiated with NIR light 48 h. At the end of the reaction, the mixture diluted THF and the copper complex was removed by passing it through a neutral alumina column. Excess amount of THF was evaporated by a rotary evaporator. After precipitation of the mixture to cold methanol, the polymer was collected by filtration and dried under vacuum overnight. M n , GPC : 3510 g•mol −1 , M w /M n : 1.10. ## Synthesis of cross-linked polymer Az-3 and Alk-5 was mixed in equal ratio (1 equiv) with Cu II Cl 2 (0.05 equiv) and PMDETA (0.1 equiv) in a small transparent vial and 300 µL BPNs in DMF was added to the vial, then irradiated 4 h. After the gelation was completed, the gel was placed in DCM for 24 h hours, then filtered and dried 24 h in a vacuum oven. ## Supporting Information Supporting Information File 1

chemsum

{"title": "Exfoliated black phosphorous-mediated CuAAC chemistry for organic and macromolecular synthesis under white LED and near-IR irradiation", "journal": "Beilstein"}

the_crystalline_state_as_a_dynamic_system:_ir_microspectroscopy_under_electrochemical_control_for_a_

## Abstract: Controlled formation of catalytically-relevant states within crystals of complex metalloenzymes represents a significant challenge to structure-function studies. Here we show how electrochemical control over single crystals of [NiFe] hydrogenase 1 (Hyd1) from Escherichia coli makes it possible to navigate through the full array of active site states previously observed in solution. Electrochemical control is combined with synchrotron infrared microspectroscopy, which enables us to measure high signal-to-noise IR spectra in situ from a small area of crystal. The output reports on active site speciation via the vibrational stretching band positions of the endogenous CO and CN À ligands at the hydrogenase active site. Variation of pH further demonstrates how equilibria between catalytically-relevant protonation states can be deliberately perturbed in the crystals, generating a map of electrochemical potential and pH conditions which lead to enrichment of specific states. Comparison of in crystallo redox titrations with measurements in solution or of electrode-immobilised Hyd1 confirms the integrity of the proton transfer and redox environment around the active site of the enzyme in crystals. Slowed proton-transfer equilibria in the hydrogenase in crystallo reveals transitions which are only usually observable by ultrafast methods in solution. This study therefore demonstrates the possibilities of electrochemical control over single metalloenzyme crystals in stabilising specific states for further study, and extends mechanistic understanding of proton transfer during the [NiFe] hydrogenase catalytic cycle. ## Introduction Obtaining crystals of redox enzymes in intermediate states relevant to catalysis is a high-profle, yet challenging target. Methods for controlling the redox state of protein crystals include the titration of crystal medium with reductant or oxidant until a desired solution potential is reached, 1,2 exposure of crystals to substrate/inhibitor, 3,4 or crystallisation of protein which has been pre-equilibrated to a desired redox state. 5,6 However, these methods often lack precision in generating pure enzyme states. There is also growing interest in triggering catalytic steps in enzyme crystals which can be coupled with time-resolved serial synchrotron or XFEL crystallography, and to date, such methods have typically relied on photo-triggers for reactivity. Methods for studying crystalline and lyophilised enzyme using gas exchange have also been reported. 10 Verifcation of protein redox states in crystallo presents a further challenge, and to this end, a number of synchrotron macromolecular crystallography beamlines have introduced microspectroscopic methods for secondary characterisation of protein crystals, including UV-visible and Raman spectroscopy. We have previously demonstrated the possibility of electrochemical control over single crystals of hydrogenase I from Escherichia coli (Hyd1) coupled with synchrotron infrared (IR) microspectroscopy for simultaneous reporting on the active site speciation. 14 Vibrational absorption bands of the integral CO and CN ligands at the active site of hydrogenases make this spectroscopic method ideal for elucidating the redox and coordination state of the active site. By applying steps in electrode potential, we were able to achieve uniform and reversible manipulation of Hyd1 in crystallo from the most oxidised to the most reduced levels. We now show how fne potential control over Hyd1 crystals can be used to generate specifc redox levels, enabling us to control and examine transitions between catalytically-relevant redox and protonation states. Hydrogenases are a broad group of enzymes responsible for bidirectional heterolytic activation of dihydrogen (H 2 # H + + H / 2H + + 2e ) at di-iron or nickel-iron bimetallic active sites. 15,16 They have attracted attention for wide-ranging applications in biotechnology: energy applications in microbial H 2 production and bioanodes for H 2 /O 2 fuel cells, through to H 2driven biocatalytic cascades. The active sites of most hydrogenases are 'wired' to a bacterial membrane or to their natural redox partner via a chain of FeS clusters in the protein. In the [NiFe] enzymes, the active site Ni atom is ligated by two terminal cysteine thiolates, with two additional cysteines bridging to the Fe atom. The Fe is further coordinated by one CO and two CN ligands (Scheme 1). During catalysis, the Ni formally cycles through Ni I/II/III , whereas the Fe remains formally Fe II , presumably stabilised by buffering of electron density from the combination of pi-acceptor and sigma-donor properties of the coordinated CO and CN ligands. 16 The CO and CN stretching bands in the mid-IR (n CO and n CN respectively) respond sensitively in wavenumber position to changes in electron density at the active site, and even to protonation and changes in hydrogen-bonding in the vicinity of the active site. 22 Since protons (H + ) are the product/substrate of hydrogenases, the activity and speciation of these enzymes are greatly pH-dependent. 23,24 It is possible, electrochemically, to step through the range of catalytically-relevant redox levels of hydrogenase by controlling the electron transfer and protoncoupled electron transfer (PCET) steps, as shown in Scheme 1. A catalytic cycle for [NiFe] hydrogenases has been proposed by combining insight from spectroscopic, computational, structural and activity studies. 16 Viewed in the direction of H 2 oxidation (Scheme S1 †), it is generally accepted that H 2 binds at the Ni II redox level, 'Ni a -SI', the most oxidised catalytic state (subscript 'a' denoting an 'active' catalytic species), however, any Michaelis complex with H 2 has evaded detection to date. Heterolytic cleavage of H 2 leaves a hydridic H in a bridging position between the Ni and Fe, 25 and a proton on a nearby base, the identity of which remains hotly debated. The resulting state is generally termed Ni a -R, though several Ni a -R sub-states exist, likely reflecting sequential proton movement away from the active site. 22 An electron must next be transferred from the active site to the FeS cluster relay chain to form the Ni a -C state which still contains a bridging hydride, but is formally Ni III . From the Ni a -C state, the bridging hydride is lost as a proton to a nearby base (not necessarily the same base that accepts the initial proton from H 2 cleavage at the Ni a -R level 32 ) leaving the Ni reduced formally by two electrons to Ni I (generally termed 'Ni a -L', noting that there are also multiple Ni a -L sub-states, again likely reflecting differential proton location in the region of the active site). Since the transition from Ni a -C to Ni a -L simply requires relocation of electron density and a proton, these two states have been described as tautomeric and exist at the same redox level (Scheme 1). Although Ni a -L was frst observed as a low temperature photo-product of Ni a -C, evidence for Ni a -L as a catalytic intermediate has accumulated from a series of steady state and transient spectroscopic studies which support earlier theoretical mechanistic proposals. 22, Finally, the Ni must be oxidised back to Ni II via electron transfer to the FeS cluster chain, to re-generate the Ni a -SI state ready to bind the next molecule of H 2 . Hyd1 belongs to the group of so-called O 2 -tolerant hydrogenases, along with membrane bound hydrogenase (MBH) enzymes from Ralstonia eutropha (Cupriavidus necator), Aquifex aeolicus, and Hydrogenovibrio marinus. 15 Hydrogenases within this group differ principally in the structure and potential of the electron-relay FeS cluster proximal to their active site, and the unusually high potential of this cluster has been linked to their O 2 -tolerance. Like other [NiFe] hydrogenases, Hyd1 is isolated in a mixture of oxidised inactive states that require reductive activation. 43,44 Scheme 1 Skeletal structure of the active site redox states for [NiFe] hydrogenases ordered by redox level. Dashed arrows represent the H 2 binding and activation step during catalytic H 2 oxidation. States are colour-coded to match data throughout this work. Catalytically active states are labelled "Ni a -X", where X ¼ SI, C, L or R. n CO band positions refer to Hyd1, pH 5.9. The predominant state is Ni-B, which has a bridging OH ligand 16,45 (Scheme 1), and is reversibly re-formed following oxidation, particularly at low H 2 . 46 Despite a wealth of spectroscopic, structural, and biophysical studies on hydrogenases from diverse organisms, many details of the mechanism of H 2 activation remain uncertain. Individual proton and electron transfer events, how they are temporally linked to the catalytic mechanism and/or formation and reactivation of inactive states, the identities of proton donors and acceptors, and the identity of key catalytic intermediates are all questions that remain unanswered for both the [NiFe] and [FeFe] hydrogenases. One of the challenges is how to unify the understanding gained from measurements made on different physical sample types. Spectroscopy of hydrogenases is typically performed in solution, and solution IR spectroelectrochemical 'redox titrations' are well established for hydrogenases, with use of smallmolecule redox mediators to facilitate diffusion-controlled electron transfer. 16 Frozen samples are required for nuclear resonance vibrational spectroscopy and most EPR measurements, while crystalline material is required for structure determination. Electrochemistry on flms of electrodeimmobilised protein (Protein Film Electrochemistry, PFE) has been used widely in studying hydrogenases, and we have previously introduced a complementary IR spectroelectrochemical approach termed Protein Film Infrared Electrochemistry (PFIRE) which provides chemical/structural insight to complement information from PFE alone. 32,50,51 Here, we compare the potential-dependence of IR-detected equilibrium active site states observed for Hyd1 under electrochemical control in solution, on an electrode (PFIRE) and in single crystals. Signifcantly, we now show that, within single crystals of Hyd1, it is possible to achieve control over the full manifold of states observed in solution for this enzyme. A related report by Morra et al. demonstrates electrochemical manipulation of [FeFe] hydrogenase I from Clostridium pasteurianum using similar methods, 78 and these studies present the possibility for using electrochemical control over single protein crystals to establish samples in the solid state for further structural study. ## Purication of Hyd1 Hyd1 was prepared aerobically according to a published procedure. 52 For PFIRE and solution-based experiments, no further purifcation was required, however for crystallisation it was essential to remove any aggregated protein and bound cytochrome-b subunit via size exclusion chromatography (SEC) followed by hydroxyapatite chromatography, as described previously. 26 Fractions containing highly pure Hyd1 (HyaAB) were identifed by SDS-PAGE, pooled and buffer exchanged into SEC buffer (20 mM Tris, pH 7.2, 150 mM NaCl, 0.02% (w/v) DDM detergent, 1 mM dithiothreitol) by repeated spin concentration and dilution (using Vivaspin 20 mL, 50 kDa molecular weight cut-off centrifugal concentrators until a 1500-2000 fold dilution of the phosphate buffer used during hydroxyapatite chromatography was achieved). For crystal growth, protein samples were concentrated to 5 mg mL 1 , as judged by Bradford assay. 53 Crystals of Hyd1 were acquired according to previously established protocols, 26 using the sitting drop vapour diffusion technique, where 1.5 mL of protein solution was mixed with an equal amount of crystallisation buffer (either 100 mM Bis-Tris, pH 5.5-5.9, 200 mM Li 2 SO 4 , 150 mM NaCl, PEG 3350 (19-21% w/v) or 100 mM Tris$HCl, pH 8.0, 200 mM Li 2 SO 4 , 150 mM NaCl, 19-21% PEG 3350) followed by streak seeding with old smaller crystals of Hyd1. Incubation under an anaerobic atmosphere (<0.3 ppm O 2 ) at 20 C resulted in crystals appearing within 24 hours. ## Single-crystal IR microspectroscopic-electrochemical experiments An adaptation of our previously-reported cell design 14,54 was used for single-crystal microspectroscopic electrochemistry, and is described in more detail in the ESI (Fig. S1 †). The microspectroscopicelectrochemical cell contained a miniature Ag/AgCl reference electrode (3 M KCl, 2 mm diameter, eDAQ), a graphite ring counter electrode (cut from a graphite tube, Goodfellow), and a glassy carbon working electrode (4 mm diameter, Alfa Aesar). The working electrode was polished to high reflectivity (ca. 10-20% in the mid-IR) using increasingly fne grades of silicon carbide paper (2500 and 4000 grit, Kemet). The polished electrodes were washed by ultrasonication in ultrahigh purity water (MilliQ, 18 MU cm) prior to cell assembly. The reference electrode was removed during the polishing process to avoid damage and contamination. A solution containing the redox mediators 2,6-dichloroindophenol, phenazine methosulfate, indigo carmine, anthroquinone-2-sulfonate, and methyl viologen, each at 1 mM concentration (Table S1 †) was prepared in N 2 -degassed crystal stabilisation buffer (for experiments conducted at pH 5.9 this was 100 mM Bis-Tris, pH 5.9, 200 mM Li 2 SO 4 , 150 mM NaCl, 22% v/v PEG 3350, whereas for experiments conducted at pH 8.0 the buffer used was 100 mM Tris, pH 8.0, 200 mM MgCl 2 , 150 mM NaCl, 22% PEG 3350). A 3 mL aliquot of the mediator solution was added to each well of a crystallisation plate containing Hyd1 crystals (crystals were stored in $3 mL mother liquor, and the size and number of crystals varied between wells). Gentle pipetting suspended the crystals without damaging crystal integrity. The resulting 6 mL mixture containing redox mediators and Hyd1 crystals was then deposited onto the glassy carbon working electrode of the microspectroscopic-electrochemical cell. An additional 12 mL of redox mediator solution in crystal stabilisation buffer was then pipetted onto the Ag/AgCl reference electrode and graphite counter electrode such that the cell was flled with approximately 18 mL of ca. 0.66 mM mediator solution. A CaF 2 window (UV grade, 30 mm diameter, 1 mm thickness, Crystran) was sealed onto the cell surface using a PTFE gasket (Harrick, 25 mm thick) and silicone sealant (Dowsil, SE 9187L Silicone RTV) to maintain an anaerobic environment within the cell. Assembly of the IR microspectroscopic-electrochemical cell, crystal handling, and mediator solution preparation, were carried out in a N 2 -flled glovebox (Plas-Labs Inc., 815 PGB series, <20 ppm O 2 ). The addition of redox mediators facilitates diffusion-controlled transfer of electrons through the electrolyte to enable electron transfer between the working electrode and the crystalline protein (a representative cyclic voltammogram of the mediator solution is shown in Fig. S2 †). Solvent channels within the Hyd1 crystal have radii between 5.4-6.6 (calculated using pdb 6FPO and MAP_CHANNELS 55 ) and are thus large enough to allow diffusion of redox mediators throughout the crystal (Fig. S3 †). IR microspectroscopic-electrochemical experiments were carried out on the MIRIAM beamline B22 at Diamond Light Source, UK, using a Vertex 80V Fourier transform IR spectrometer coupled to a Hyperion 3000 IR microscope (Bruker) with a high-sensitivity photovoltaic mercury cadmium telluride (MCT) 50 mm pitch detector cooled to 77 K using liquid N 2 . A transflection geometry was used to obtain IR spectra (i.e. the microscope is used in reflection mode and detected light that passes through the cell twice), using a 36 objective and 15 15 mm 2 knife-edge aperture in the detection beampath. Each spectrum was recorded as an average of 1024 interferograms working at 80 kHz scanner velocity and at 4 cm 1 resolution (ca. 160 s measurement time). Data acquisition was performed using Bruker OPUS software (version 7.5). Electrochemical measurements were acquired using an AutoLab 128N potentiostat (Metrohm) controlled by Nova software (version 1.10). The miniature Ag/AgCl reference electrode was calibrated against a saturated calomel reference electrode (SCE, BAS), and potentials quoted in the text are adjusted to mV vs. the standard hydrogen electrode (SHE) using the conversion E (mV vs. SHE) ¼ E (mV vs. SCE) + 241 mV at 25 C. 56 Baseline correction, and all subsequent data analysis was carried out using OriginPro software (OriginLab Corp., version 9.1). Baseline correction was applied using an interpolated spline function, and careful comparison with 2 nd derivative and difference spectra was used to avoid distortion of peak shapes. Baseline corrected spectra are presented in the main text, and representative raw spectra are shown in the ESI. † ## Solution infrared spectroscopic-electrochemical measurements Electrochemically-controlled IR redox titrations of solution phase Hyd1 were recorded using our previously-reported methods. 57,58 Briefly, a 3D carbon particle network electrode 59 containing Hyd1 trapped within a mixture of the polymer electrolyte Nafon (Sigma, titrated to pH 6 in phosphate buffer) and carbon black particles (XC72R, DUPONT) was prepared on the surface of an ATR-IR accessory (GladiATR, Pike Technologies) housed in an anaerobic, dry glovebox (Glove Box Technologies). The 3D network electrode was sealed into an electrochemical cell containing a carbon rod working electrode connection, saturated calomel reference electrode, and a Pt wire counter electrode. A closed loop of N 2 -purged electrolyte was pumped through the cell to prevent build-up of any trace H 2 produced by Hyd1. For more details see ESI. † IR spectra were recorded using an Agilent 680-IR spectrometer controlled by ResPro 4 software, as an average of 1024 interferograms (ca. 360 s measurement time). Electrochemical control was provided by an Autolab 128N potentiostat (Metrohm), and potentials (E) are reported relative to SHE using the conversion E (mV vs. SHE) ¼ E (mV vs. SCE) + 241 mV at 25 C. 56 Infrared spectroscopic-electrochemical measurements of electrode-adsorbed Hyd1 (PFIRE) The IR spectroscopic data collected from electrode-adsorbed Hyd1 are reproduced using data from Hidalgo et al. 51 The PFIRE method is briefly described in the ESI. † In order to aid comparison to our single crystal method, we report the absorbance of individual Ni a -R sub-states separately in this manuscript, whereas they were summed to give a 'total' Ni a -R absorbance in Hidalgo et al. We have also reassigned some of the Ni a -L absorbances relative to the original manuscript such that both the Ni a -L I,II,III and Ni a -R I,II,III sub-states are labelled in order of decreasing wavenumber of the active site CO stretch, n CO , in line with other literature. 35,36, ## Initial characterisation and electrochemical reduction of crystalline Hyd1 Fig. 1A shows a visible image, at 36 magnifcation, of a single Hyd1 crystal lying on the working electrode surface (fne scratches are also visible in the glassy carbon surface, and another crystal can be seen to the left of the image, roughly vertically oriented). The 15 15 mm 2 area used to record IR spectra is shown with a black square. Prior to electrochemical manipulation of the crystal, an IR spectrum was recorded at the open circuit potential (OCP) imposed by the oxidised mediator cocktail (typical OCP values were +209 to +274 mV vs. SHE). Fig. 1B shows a representative IR spectrum of crystalline Hyd1 recorded at pH 5.9 and an OCP of +209 mV before any electrochemical manipulation, showing the CN and CO stretching regions, n CN and n CO , respectively (for raw data see Fig. S4 †). Crystals prepared from aerobically purifed 'as-isolated' Hyd1 contain a mixture of oxidised inactive states as is common for [NiFe] hydrogenases, 16 predominantly the Ni-B state (Fig. 1B, n CO 1943 cm 1 ) with minor contributions from another oxidised species with n CO 1937 cm 1 . The identity of this minor component is unknown, but similar species have been observed in other hydrogenases and attributed to readily-activated species at the same redox level as Ni a -SI. 16,22 Very intense absorbances are observed from crystalline Hyd1 due to the high effective protein concentration within the crystals (ca. 8 mM of active site, see ESI †). Furthermore, the transflection geometry of the microspectroscopic-electrochemical cell means that the effective IR pathlength through the sample is of the order of 30-50 mm, approximately double the crystal thickness. In the case of the crystal sample shown in Fig. 1A we can estimate the molar extinction coefficient of the Ni-B n CO band as approximately 4000 M 1 cm 1 , in good agreement with the extinction coefficient reported for this state of the active site in the large subunit of Ralstonia eutropha MBH in solution. 67 Crystalline Hyd1 was subjected to electrochemical reduction by applying a potential of 597 mV for a minimum of 1 hour, until no further changes in n CO and n CN bands were observed over a period of 10 minutes. Electrochemical reduction of crystalline Hyd1 is somewhat analogous to reductive activation used in PFE and PFIRE measurements on [NiFe] hydrogenase, 32,68 and as shown in Fig. 1C 2 shows a series of spectra of a single Hyd1 crystal recorded as a function of applied potential at pH 5.9, as both baselinecorrected spectra (Fig. 2A) and a 2D 'heatmap' plot (Fig. 2B). Fig. 2 focusses on the n CO region only, data including the n CN region are shown in Fig. S6, † and raw data are shown in Fig. S7. † These spectroscopic data were recorded following electrochemical reduction at 597 mV, as a series of small steps (25-100 mV per step) towards more positive potentials were applied between 597 mV and +203 mV. After each step the potential was held until spectroscopic equilibration was achieved, as judged by no further changes to IR spectra (or a minimum of 8.5 minutes; corresponding chronoamperometry data shown in Fig. S8 †). The changes in Hyd1 active site speciation during this in crystallo oxidative redox titration can be seen from the potentialdependent shift in n CO bands in Fig. 2. These correlate well with the ladder of redox states shown in Scheme 1. The Ni a -R species (1922 and 1914 cm 1 ) dominate at the most reducing potentials, converting to the Ni a -C (1951 cm 1 ) and Ni a -L (1877 and 1866 cm 1 ) states at intermediate potentials, and then forming the most oxidised catalytically active state Ni a -SI (1929 cm 1 , maximum intensity at 122 mV). At the most oxidising potentials the oxidised, inactive state Ni-B dominates (1943 cm 1 ). Signifcantly, all previously established active site states for Hyd1 are observed in crystallo, including multiple sub-states of Ni a -R and Ni a -L. Fitting of the n CO bands to Gaussian band profles and extracting the ftted peak absorbances allows plotting of titration curves of each active site state as a function of potential, as shown in Fig. 3A (representative spectral peak ftting shown in Fig. S9 †). In order to correlate states observed in Hyd1 crystals with the potential dependence of states observed in more conventional spectroscopic-electrochemical studies, Fig. 3 compares equilibrium potential-controlled redox titrations of the Hyd1 active site in single crystals (Fig. 3A), of Hyd1 in solution (Fig. 3B) and of electrode-adsorbed Hyd1 (Fig. 3C, recorded under an Ar atmosphere, reproduced using data from Hidalgo et al.). 51 The assignments of the n CO and n CN bands for each active site redox species in the crystalline state are consistent with those observed in both solution and electrode-adsorbed IR spectra of Hyd1 (Table S2 †). At pH 5.9, there is little or no catalytic H + reduction by Hyd1 (ref. 24 and 51) and as such the data in Fig. 3, recorded under an inert atmosphere, reflect essentially non-turnover behaviour of the Hyd1 active site. The titration curves measured from crystalline, solution, and electrode-adsorbed samples are remarkably similar: the potentials at which the maximum intensity for each redox species is observed are consistent throughout. This result confrms the integrity of the dynamic behaviour around the active site of Hyd1 in single crystals, thus showing that observations made in the crystalline state of Hyd1 are mechanistically relevant to the enzyme in solution, and to PFE studies of enzyme activity. ## Potential-controlled single-crystal redox titration of Hyd1, pH 8.0, monitored by IR microspectroscopic electrochemistry Hydrogenases catalyse H 2 activation via a series of exquisitelytimed electron transfer and proton-coupled electron transfer steps (Scheme S1 †). Studies of hydrogenase activity and spectroscopic properties have been reported at a range of pH in order to establish details of the proton inventory during catalysis. 35,36,69 Crystals of Hyd1 are stable over a relatively wide pH range, 32 and by pre-soaking crystals in pH-adjusted crystal additional species with n CO at 1938 cm 1 is evident at potentials above +100 mV, and accounts for the apparent loss of Ni-B (1942 cm 1 ) at these potentials in Fig. 4B. The potential dependence of this 1938 cm 1 species (Fig. S13 †) is similar to the potential of the [Fe 4 S 3 ] 5+/4+ proximal cluster transition 39,70 and could be related to formation of the superoxidised proximal cluster (Table S4 †). Further investigation of this behaviour is a target for future studies. pH-dependent behaviour of the active site and surroundings Fig. 5A compares IR spectra in the n CO region at both pH 5.9 and pH 8.0, extracted from redox titrations at 222 mV (pH 5.9, Fig. 3A) and 299 mV (pH 8.0, Fig. 4), potentials at which the intensities of the Ni a -C and Ni a -L states are maximal. We observe three main differences in both the titration data and spectra at pH 5.9 in comparison to pH 8.0: (1) The relative populations of Ni a -C and total Ni a -L, and of the individual Ni a -L II/III sub-states, are pH dependent. (2) The equilibrium midpoint potential for transitions between each redox level shifts to more negative potentials at pH 8.0. (3) The n CO peak positions for all redox states shift to lower wavenumbers at pH 8.0 relative to pH 5.9 (Fig. 5A and Table S3 †). The redox titration data reported in Fig. 3 (at pH 5.9) and Fig. 4 (at pH 8.0) explicitly show contributions from two Ni a -R sub-states present in Hyd1, Ni a -R II and Ni a -R III . The peak positions at each pH are provided in Table S3. † For simplicity in Fig. 3 (n CO at 1877 cm 1 at pH 5.9) and Ni a -L III (n CO at 1866 cm 1 at pH 5.9). We have previously demonstrated a pH-dependent tautomeric equilibrium between the Ni a -C and Ni a -L species in Hyd1 (Scheme 1). 23 Here we observe the same shift in equilibrium towards Ni a -L at higher pH (Fig. 5B), showing that Ni a -C/Ni a -L tautomerism is maintained in the crystalline state. This observation is critical, as it suggests that crystallisation does not perturb proton transfer equilibria in the vicinity of the [NiFe] active site, in addition to the unperturbed electron transfer redox equilibria demonstrated by the equilibrium redox titrations in Fig. 3 and 4. In addition to providing evidence of Ni a -C/Ni a -L tautomerisation in the crystalline state, it is clear from Fig. 5 that the relative proportions of each Ni a -L sub-state also vary with pH, consistent with the behaviour of electrode-adsorbed Hyd1 (Fig. S14 †) where the population of Ni a -L II remains roughly constant above pH 6. 71,72 The mechanistic role of the Ni a -L substates as sequential intermediates in proton transfer to/from the [NiFe] active site has been demonstrated in phototriggered potential jump measurements on soluble hydrogenase 1 (SH1) from P. furiosus, 36 and cryogenic photolysis of the [NiFe] hydrogenase from D. vulgaris Miyazaki F. 33,34 The most common representation of 'Ni a -L' invokes protonation of a terminal cysteine-S ligand to Ni at the active site. Evidence of cysteine-S protonation in the Ni a -L I sub-state has been reported S3. † For IR spectra of the n CN and n CO regions across the full potential range of 600 to +200 mV see Fig. S10. † (B) The speciation curves illustrate how the absorbance of the n CO peaks of Hyd1 active site species vary with potential at pH 8.0. in the D. vulgaris Miyazaki F [NiFe] hydrogenase, where H/D labelling suggested the presence of an S-H stretching vibration in Ni a -L I . 37 We have previously noted that the Ni a -L I sub-state does not accumulate signifcantly, if at all, in O 2 -tolerant [NiFe] hydrogenases such as Hyd1, 22 and this behaviour is maintained in the crystalline state. Computational modelling studies of the active site suggest that deprotonation of a terminal cysteine thiol ligand to Ni causes n CO to shift to lower energy by ca. 30 cm 1 . 73 This shift upon deprotonation matches the difference in n CO observed between Ni a -L I and Ni a -L II/III for a range of [NiFe] hydrogenases, 22,72 leading us to postulate that a terminal cysteine thiol is not present in either of the Ni a -L II or Ni a -L III sub-states. The high Hyd1 concentration (8 mM) within single crystals allows us to test this hypothesis further through direct observation of the S-H stretching region (ca. 2450-2600 cm 1 ). Difference spectra are particularly sensitive to changes in cysteine-S protonation between individual redox states. 37 Potential-induced single crystal difference spectra (Fig. S15, † calculated from raw in crystallo microspectroscopy data) suggest that there is no change in cysteine-S protonation between the Ni a -R II/III , Ni a -C, Ni a -L II/III , and Ni a -SI redox states in Hyd1. Therefore we fnd no evidence of S-H bond formation in the Ni a -L II or Ni a -L III , and Ni a -R II or Ni a -R III sub-states of Hyd1 (Fig. S15 †). Whilst the apparent lack of an S-H resonance in crystalline Hyd1 does not conclusively rule out cysteine thiol formation in Ni a -L II/III or Ni a -R II/III , the high S/N and intensity of spectra recorded from concentrated, crystalline Hyd1 would provide the ideal scenario for detecting any low-intensity S-H resonances. It is generally accepted that a glutamate residue (E28 in Hyd1 numbering) close to the active site is critical for proton transfer during Ni a -L formation from Ni a -C, 32,36,37,74 and the primary proton acceptor during this transition in P. furiosus SH1 has been shown to have a pK a of approximately 7. 35 It is therefore possible that deprotonation of E28 is required for enrichment of Ni a -L III at pH 8.0. However the Ni a -L II sub-state has a considerably lower apparent pK a $ 5 (Fig. S14 †), implying that deprotonation of E28 is not required for Ni a -L II formation from Ni a -C in E. coli Hyd1. The spectra of Hyd1 in Fig. 5A show an apparent peak shift and broadening of the Ni a -C n CO band upon change of pH. Peak ftting of these data (Table S3 †) suggests that the Ni a -C peak actually contains contributions from two distinct n CO resonances for Ni a -C, at 1951 cm 1 and 1947 cm 1 , with the lower wavenumber species enriched at pH 8.0. This is consistent with the observations of Greene et al., who noted a pH equilibrium between two forms of Ni a -C in P. furiosus SH1 with an apparent pK a of 6.8. 36 The mechanistic relevance of this is not clear, although Greene et al. have postulated that protonation/deprotonation of glutamate E28 could account for the pH-dependent shift in the n CO position of Ni a -C. 36 In Fig. 5A we also observe a pH-dependent shift in the n CO band of Ni a -SI (Table S3 †). ## The transition from Ni a -C to Ni a -SI We have previously shown that redox transitions involving chemical steps such as proton transfer appear to be retarded in the crystalline state. 14 By continuously recording IR spectra during equilibration after each potential step in an electrochemical redox titration, we can monitor these kinetic aspects of equilibration in the crystals. Fig. 6A shows a series of difference spectra of a Hyd1 crystal at pH 5.9, following equilibration after a small oxidative potential step from 197 mV to 172 mV, i.e. a positive potential step from where Ni a -C and the Ni a -L states are maximal (Fig. 3A). The difference spectra are presented as 172 mV minus 197 mV, and the raw experimental data and baseline corrected spectra are shown in Fig. S16 and S17. † The corresponding change in absorbance of the Previous studies have shown the involvement of the Ni a -L substates as an on-pathway intermediate between Ni a -C and Ni a -SI during catalysis. 23,33,36 Fast kinetic methods, capable of probing redox chemistry of the [NiFe] active site with sub-turnover frequency time resolution, were necessary to conclusively confrm the catalytic competence of the Ni a -L states. Here we are able to access similar information without the need for fast time resolution. The electrochemical control afforded by the microspectroscopicelectrochemical cell, in combination with solution redox mediators, provides a source or sink of electrons that are available to the crystalline protein on a timescale that is clearly faster than some chemical steps in crystallo. This is evident due to the relatively fast equilibration of the Ni a -C and Ni a -R states (which differ only in active site redox state rather than protonation in Hyd1, Scheme 1), relative to Ni a -L and Ni a -SI (which additionally require a proton transfer step) in Fig. 6. We have previously noted that O 2 -tolerant [NiFe] hydrogenases do not accumulate either the Ni a -R I or Ni a -L I sub-states, and instead favour the Ni a -R II/III and Ni a -L II/III substates. 22 We fnd no evidence for cysteine-S protonation in the Ni a -R II/III and Ni a -L II/III sub-states in crystallo, and the faster rate of the Ni a -R / Ni a -C transition implied by the data in Fig. 6 and S17 is consistent with this transition involving only electron transfer in the case of Hyd1. The fact that the Ni a -L to Ni a -SI transition is apparently rate limiting during this potential step is consistent with involvement of both proton and electron transfer. We have previously discussed possible mechanistic implications of the unusual high-potential [4Fe3S] 5+/4+/3+ cluster proximal to the active site found in O 2 -tolerant [NiFe] hydrogenases, in particular concerning whether proton-coupled electron transfer between Ni a -L and Ni a -SI occurs via a concerted or stepwise mechanism. 22 From the work of the groups of Hirota and Dyer it is known that onwards formation of Ni a -SI from Ni a -C, via Ni a -L, requires the proximal iron-sulfur cluster to be capable of receiving an electron, i.e. to be in an oxidised state. 33,68 The potential of both the [4Fe3S] 4+/3+ and [4Fe3S] 5+/4+ transitions of the proximal cluster in Hyd1 are relatively high, +3 mV and +230 mV respectively at pH 6 (see Table S4 †), 38,39,70 and so the proximal cluster of Hyd1 will largely be in the [4Fe3S] 3+ state at equilibrium at the potentials applied in Fig. 6. In the O 2 -sensitive hydrogenases the proximal cluster is a standard cubane [4Fe4S] 2+/+ cluster and its potential is closer to the potential of the H + /H 2 couple at neutral pH. 38 The electron transfer necessary for the Ni a -R / Ni a -C and Ni a -C / Ni a -SI transitions is therefore hindered in Hyd1 relative to O 2 -sensitive [NiFe] hydrogenases, and we suggest that this may be responsible for the fact that the Ni a -R I and Ni a -L I sub-states do not accumulate in Hyd1. In combination with our earlier hypothesis that cysteine-S protonation is not present in either of the Ni a -L II and Ni a -L III sub-states of Hyd1, we tentatively suggest two possible proton-coupled electron transfer mechanisms for the conversion between Ni a -L and Ni a -SI. In the frst mechanism, concerted proton and electron transfer occurs during Ni a -SI formation as previously reported by Dyer and co-workers. 35 In the second mechanism, likely prevalent in O 2 -tolerant hydrogenases such as Hyd1, proton and electron transfer is stepwise due to electron transfer between the active site and proximal cluster becoming limiting in the presence of the unusual high potential [4Fe3S] cluster. Proton transfer is relatively unaffected, as the active site structure and surrounding amino acids are highly conserved, and so H + can leave the active site ahead of electron transfer during the Ni a -L / Ni a -SI transition. Scope for crystal structures of well-defned states Prolonged exposure to the mediator cocktail and application of potential have no effect on the ability of Hyd1 crystals to diffract X-rays (Fig. S19 and Table S5 †), suggesting electrochemical manipulation of Hyd1 crystals offers the exciting prospect of producing molecular models for intermediates of catalysis that have so far been inaccessible to structure determination. The exquisite control of electrochemical potential afforded by the electrode allows crystals to be precisely poised under conditions that favour formation of only the intermediate of interest, for example the most reducing potentials applied allowed accumulation of pure Ni a -R. This advantage contrasts strongly with the reduction of [NiFe] hydrogenase crystals by H 2 which generates complex mixtures of states that are less suitable for structure determination. 10,25 Manipulation of pH offers a further dimension to the control over speciation of the crystalline enzyme. Such control offers a more rational approach to obtaining structures for catalytic intermediates than has previously been possible and eliminates the need for low activity variants, 75 inhibitors, 76 or transition-state analogues 77 that have been mainstays of classical (pre-XFEL) time-resolved structure determination from single crystals. Furthermore, our technique offers the possibility of fnally linking the spectral fngerprints of each intermediate to a defned confguration of the active site and spectral changes occurring during turnover with specifc atomic motions. ## Conclusions Working with single crystals of Hyd1, we have described a method that allows complete control over the redox state of crystalline proteins. The measurements reported confrm that protein crystals can be viewed as a dynamic system where all known states and intermediates are reliably accessible. High protein concentrations in the crystalline state allow us to record spectra with high signal/noise ratios, facilitating assignment of the n CO bands of each active site state. The active site n CO and n CN band positions and redox chemistry in single crystals of Hyd1 are consistent with previously reported behaviour, providing compelling evidence that crystallisation does not change the immediate environment or chemical properties of the active site relative to solution-phase protein. All known states of the Hyd1 active site can be generated under fne potential control, and detailed redox titrations recorded from single crystals match those for Hyd1 in solution or adsorbed on an electrode. These single crystal measurements thus bridge the gap between structural, spectroscopic, and activity-based biophysical methods, and provide confrmation that the behaviour of proteins in a range of physical states are comparable. The pH of the crystals is also readily manipulated, and all aspects of proton transfer, including Ni a -C/Ni a -L tautomerism are retained in the crystalline state. Detailed electrochemicallycontrolled redox titrations of the Hyd1 active site demonstrate the importance of single crystal microspectroscopy as a complementary method to protein crystallography, and could be used for spectroscopic characterisation post-X-ray diffraction to provide confrmation of the redox state solved. This aspect is particularly important given the often complex mixtures of redox states that are present over wide potential windows. Likewise, the ability to generate pure redox states across a narrow potential window, as demonstrated here, allows infrared microspectroscopic-electrochemical methods to deliver a roadmap for how to enrich and prepare individual redox states in crystallo for downstream structure determination. Crystal structures can thus be directly relevant to redox states probed during the catalytic cycle. In addition to electrochemical navigation of the redox states, through careful control over pH it is possible to access different sub-states (e.g. of Ni a -R and Ni a -L), providing an extra dimension of control for future crystallographic studies. Of further signifcance, through electrochemical control over single crystals we are able to access retarded reaction steps that are otherwise hidden in steady-state catalytic studies. The use of both positive and negative potential steps reveals details of proton-coupled electron transfer to and from the active site, and has allowed us to hypothesise sequential, rather than concerted, proton and electron transfer during the Ni a -L / Ni a -SI transition in O 2 -tolerant [NiFe] hydrogenases such as Hyd1. This is in contrast to concerted proton and electron transfer observed in O 2 -sensitive [NiFe] hydrogenases. The method reported here, already extended to crystals of C. pasteurianum [FeFe] hydrogenase I, 78 is likely to have general relevance in structure-function studies of complex redox metalloenzymes.

chemsum

{"title": "The crystalline state as a dynamic system: IR microspectroscopy under electrochemical control for a [NiFe] hydrogenase", "journal": "Royal Society of Chemistry (RSC)"}

direct_in_situ_observation_of_the_electron-driven_synthesis_of_ag_filaments_on_α-ag2wo4_crystals

## Abstract: In this letter, we report, for the first time, the real-time in situ nucleation and growth of Ag filaments on a-Ag 2 WO 4 crystals driven by an accelerated electron beam from an electronic microscope under high vacuum. We employed several techniques to characterise the material in depth. By using these techniques combined with first-principles modelling based on density functional theory, a mechanism for the Ag filament formation followed by a subsequent growth process from the nano-to micro-scale was proposed. In general, we have shown that an accelerated electron beam from an electronic microscope under high vacuum enables in situ visualisation of Ag filaments with subnanometer resolution and offers great potential for addressing many fundamental issues in materials science, chemistry, physics and other fields of science. ## I n science, discovering new ways of thinking about facts is more important than obtaining these facts 1 . The investigation of nanocrystal growth is a rich research field that impacts both fundamental and applied science because controlling nanoscale sizes and morphologies can directly affect functional applications 2 . By observing nanocrystal structures microscopically, insight into the growth mechanism can be used to design a method to control nucleation, which is one of the most challenging processes in nanoscience and nanotechnology 3 . With ever-increasing temporal and spatial resolution allowing for atomic scale nanomaterial characterisation, the advent of highly sophisticated electron-and photon-based spectroscopies and scanning probe microscopies is primarily responsible for the developments in this field. In particular, measurements using a transmission electron microscopy (TEM) heating holder for in situ analysis provided an in-depth understanding of the crystal growth process, and further application of this method has attracted considerable interest . Recently, de Jonge and Ross 7 reviewed in situ liquid TEM characterisation because it facilitates the study of step-by-step nanoscale evolution 8,9 . In this manuscript, for the first time, we report the real-time in situ formation and growth of Ag filaments on a-Ag 2 WO 4 crystals using an accelerated electron beam under high vacuum. Noble metal nanoparticle preparation is an interdisciplinary subject that is attracting intense research and development due to both the fundamental and applied scientific value of nanometer scale metals . In particular, the synthesis and surface chemistry of Ag nanoparticles have been extensively reported 13 . Very recently, Xia et al. 14 outlined the current developments in the shape-controlled synthesis of Ag nanocrystals, and Li et al. 15 reviewed research focused on Ag nanowire preparation using a soft solution method as well as applications using the Ag nanowires. Recently, we have obtained silver tungstate (a-Ag 2 WO 4 ) crystals using various methods (coprecipitation, sonochemistry and hydrothermal treatment), and their corresponding photoluminescence properties have been studied 16 . An overview of results reported from the various growth experiments reveals the absence of research on nanocrystal growth by electron irradiation. In this research, we used an electron beam to grow Ag nanofilaments from a-Ag 2 WO 4 crystals. X-ray diffraction (XRD), field emission scanning electron microscopy (FE-SEM), selected-area electron diffraction (SAD), TEM and high-resolution microscopy (HRTEM) have been employed to study these materials. Through the use of these complementary techniques as well as first-principles calculations based on density functional theory (DFT), we determined the electronic structure of the a-Ag 2 WO 4 bulk. We investigated the formation of Ag nanofilament, which was followed by a subsequent growth process from the nano-to micro-scale. This observation facilitated an in-depth investigation of the physicochemical property behaviour and possible applications in various fields (i.e., sensors, catalysis, optical devices and bio systems). ## Results TEM images obtained at 5 s intervals, which show the interesting growth of the Ag filaments from the a-Ag 2 WO 4 matrix, are displayed in Figures 1(a-h) (indicated by blue arrows). The red arrows indicate another region that illustrates particle absorption by the matrix. To the best of our knowledge, this phenomenon (growth and re-absorption of Ag) is new for this material and occurs in all of the regions that the electron beam irradiated. Importantly, when the TEM microscope is used, the growth process is faster and more intense than when the SEM microscope is used due to the higher electron energy and dosage associated with TEM. As shown in the earlier stage of growth for another series of images (Supplementary Movie S1), an initial superficial Ag particle is formed from the a-Ag 2 WO 4 matrix which follows the growth of the filaments over time. In addition, to thoroughly investigate these nucleation and growth processes, the different regions of the sample where the Ag filaments are present were analysed using TEM and HRTEM. Figures 2a-c show a time-resolved series of FE-SEM images obtained under high vacuum (1 3 10 25 Pa) and the crystal morphology simulation during the growth of Ag filaments stimulated by the electron beam on the a-Ag 2 WO 4 surface. The a-Ag 2 WO 4 crystal has a hexagonal rod-like elongated shape (see Fig. 2a), and the corresponding morphology was modelled using the crystallographic data listed in the Supplementary Information (see Table SI1). Figure 2a shows a FE-SEM image of the a-Ag 2 WO 4 crystals that were acquired after a rapid approach and focus adjustment (time zero). This image reveals that after receiving a small electron dose, the a-Ag 2 WO 4 crystal surface contains a small amount of Ag nanoparticles. Figure 2b confirms that after 6 min of exposure to a 30 kV electron beam, the metallic Ag nanoparticles on the a-Ag 2 WO 4 crystal surface begin to grow. A reasonable amount of electrons induces the appearance of several defects in the surface which produces a continuous axial flow of metallic Ag particles. FE-SEM analyses revealed that this axial Ag growth process is highly reproducible. Figure 2c shows that increasing the exposure time to 10 min produces two effects: 1) Ag nanoparticles change to filamentary Ag; and 2) the nucleation of new Ag nanoparticles on the crystal surface. Figures 2a-c indicate that the Ag filament formation process from unstable a-Ag 2 WO 4 crystals is a reduction process converting the [AgO 2 ], [AgO 4 ], [AgO 6 ] or [AgO 7 ] clusters into Ag 0 . Therefore, this redox process promotes the transformation of a-Ag 2 WO 4 crystals with an ordered structure to a disordered structure. Recently, using the chemical reduction method, Tsuji et al. 17 monitored the rapid transformation of various Ag nanostructures from Ag 1 ions to Ag 0 in solution by time-dependent surface plasmon resonance. In another example, using in situ TEM images, Yasuda et al. 18 observed indium oxide reduction at 820uC for metallic indium and intermetallic species (PdIn 3 ). In the present work, using energetic electrons and in absence of an external heat treatment, we observed the formation of Ag nanorods in the solid state. This process does not follow an epitaxial growth mechanism. Figures 3a-d show low-magnification and HRTEM images of the Ag-Ag 2 WO 4 interface. The beam effect on the original a-Ag 2 WO 4 crystal allows the Ag to form particles and/or filaments and induce amorphisation of the a-Ag 2 WO 4 crystals. To fully understand these structural and chemical changes in the a-Ag 2 WO 4 crystals, X-ray energy dispersive spectroscopy (EDS) analysis was performed at several points along the filamentary grown region (crosses inside the yellow circle in Fig 3c ), and these results are presented in Figures 3e-h. The region adjacent to the Ag filament (Region 1) is composed of tungsten oxide with a small amount of Ag. The small amount of Ag in this region is expected once the Ag atoms migrate to form the filaments. The C and Cu contributions observed in the EDS analysis are from the Lacey Cu grid. At the interface (Region 2) where the matrix and filament are superposed, the primary elements are W, Ag and O. In this region, we observe contributions from the metallic Ag particle/filament and amorphous matrix. Therefore, the relative intensity of Ag increased compared to region 1. In the filament (Regions 3 and 4), the EDS analysis indicates that the filament is primarily composed of Ag atoms with a small amount of W and O. No significant change in the chemical composition was observed along the filament. The EDS and HRTEM results confirm that the crystalline Ag with a minor amount of W and O atoms are segregated indicating that a small portion of the W and O atoms from the matrix was pulled into the filament during the fast Ag filament growth. Figure 4 displays the structure modification effect in the a-Ag 2 WO 4 crystal induced by the electron beam. The early characterisation step indicated (see Fig. 4a) that the Ag filaments have not begun to grow from the a-Ag 2 WO 4 matrix. The corresponding electron diffraction pattern (SAD; see Fig. 4b) shows a set of rings indicating the polycrystalline nature of the a-Ag 2 WO 4 crystals (indexed as the orthorhombic structure). These results are in agreement with the XRD results (see Fig. S1. in Support Information). However, after focusing the electron beam on the sample for a few seconds, the Ag nanofilaments begin to grow in several regions of the a-Ag 2 WO 4 crystal (see Fig. 4c). In addition, the regions adjacent to the filamentary growth, denoted by the red circle, tend to disrupt, which was confirmed by the amorphous diffraction pattern (see Fig. 4d) and is in agreement with the high-resolution images of the matrix after filamentary growth (Figs. 3b,d). This result indicates that the Ag mass transport modifies the original structure of the compound, which no longer present long-range order. ## Discussion This behaviour can be explained by the structural and electronic information recently published by our group 16 structure). The corresponding electronic structure dictates its stability and activity 16 . In addition, this core-shell arrangement has a core that is composed of internal [WO 6 ], [AgO 6 ] and [AgO 7 ] clusters, whereas the external portion is formed by [AgO 4 ] and [AgO 2 ] clusters. When an electron beam irradiates this material, these distorted clusters produce a lattice distortion that is propagated along the material, altering the electronic distribution along these polar cluster networks. Because the absorption of one electron is typically a quantum phenomenon, first-principles calculations were performed to verify the polar nanodomains in a-Ag 2 WO 4 at the atomic level. The details of our calculations are reported in the Supplementary Information (see Figs. S2-S5). The theoretical results indicate that the [AgO 4 ] clusters are the most positively charged, whereas the [AgO 2 ] clusters are the most negatively charged. Therefore, it is feasible that electron absorption occurs at the [AgO 4 ] cluster. Our method utilises irradiation with electrons, which generates a redox environment. We propose the following mechanism: upon electron irradiation, the external and positively charged [AgO 4 ] clusters become polarised to form highly reactive moieties that can be rapidly used to reduce the adjacent angular [AgO 2 ] clusters undergoing a disproportionation rearrangement. This procedure produces [AgO 6 ] clusters and metallic Ag, which flows to the surface and results in local amorphisation of the a-Ag 2 WO 4 crystal. Based on these theoretical results, a reasonable mechanism has been proposed to explain the experimental results (supporting information) obtained in this work. For the first time, we have demonstrated that metallic Ag nanoparticles/nanofilaments grow in situ from a-Ag 2 WO 4 crystals. In this experiment, an external stimulus, such as an accelerated electron beam from FE-SEM/TEM measurements, is capable of initiating the nucleation and growth of Ag filaments. XRD, FE-SEM, SAD, TEM and HRTEM techniques have been used to extensively characterise the material. To complement these experimental measurements, structural and electronic properties have been estimated using first-principles calculations. Based on these results, we have identified that the driving force for this redox process can be attributed to order-disorder effects of the constituent clusters of a-Ag 2 WO 4 in the short-, intermediate-and long-range structures. These interlinking patterns are responsible for the physical/chemical properties of the material. In addition, we have proposed a possible formation mechanism where: irradiated electrons are absorbed by the external and higher charged tetrahedral [AgO 4 ] clusters, followed by a subsequent disproportionation reaction with the angular [AgO 2 ] clusters. This procedure produces [AgO 6 ] clusters and metallic Ag that migrate to the surface, resulting in the local amorphisation of a-Ag 2 WO 4 . We have confirmed that this material is an ideal platform with outstanding potential for use in biological, plasmonics and catalytic applications, and these studies are currently in progress. ## Methods Synthesis. The a-Ag 2 WO 4 crystals were prepared at 90uC in 1 min by the injection of precursors ions into hot aqueous solutions. A typical a-Ag 2 WO 4 crystal synthesis procedure is described below: First, 1 3 10 23 mols of tungstate sodium dihydrate (Na 2 WO 4 .2H 2 O) (99.995% purity, Sigma-Aldrich) and 2 3 10 23 mols of silver nitrate (AgNO 3 ) (99.8% purity, Sigma-Aldrich) were dissolved separately in 50 mL of deionised water. The first solution was transferred to a 250 mL glass flask and heated to 90uC under constant stirring for 10 min. Then, the second solution that contained 50 mL of AgNO 3 at room temperature was pumped by a syringe and injected into the hot aqueous solutions (90uC), and a suspension was rapidly formed with a temperature reduction to 70uC. The following suspension was immersed in a beaker with 50 mL of deionised water at 5uC. These a-Ag 2 WO 4 crystals were obtained as a fine white powder precipitated at the bottom of the glass flask. The resulting suspensions were washed several times with deionised water to remove any remaining sodium ions. Finally, these white powder precipitates were collected and dried with acetone at room temperature for 4 h. Characterisations. The a-Ag 2 WO 4 crystals were characterised by their XRD patterns using a D/Max-2500PC diffractometer (Rigaku, Japan) with Cu-Ka radiation (l 5 1.5406 A ˚) in the 2h range from 10u to 110u with a scanning velocity of 1u/min and a step of 0.02u. The shape and size of the a-Ag 2 WO 4 crystals were observed by FE-SEM through a Carl Zeiss microscope (Model Supra 35) operated at 30 kV and by TEM with a CM200 model microscope (Philips) operated at 200 kV. The a-Ag 2 WO 4 microcrystals were characterised using SAD and HRTEM. The samples used to obtain the TEM images were prepared by drying droplets of the as-prepared samples from an acetone dispersion sonicated for 10 min and deposited on the Cu grids.

chemsum

{"title": "Direct in situ observation of the electron-driven synthesis of Ag filaments on \u03b1-Ag2WO4 crystals", "journal": "Scientific Reports - Nature"}

photoredox-mediated_hydroalkylation_and_hydroarylation_of_functionalized_olefins_for_dna-encoded_lib

## Abstract: DNA-encoded library (DEL) technology features a time-and cost-effective interrogation format for the discovery of therapeutic candidates in the pharmaceutical industry. To develop DEL platforms, the implementation of water-compatible transformations that facilitate the incorporation of multifunctional building blocks (BBs) with high C(sp 3 ) carbon counts is integral for success. In this report, a decarboxylative-based hydroalkylation of DNA-conjugated trifluoromethyl-substituted alkenes enabled by single-electron transfer (SET) and subsequent hydrogen atom termination through electron donoracceptor (EDA) complex activation is detailed. In a further photoredox-catalyzed hydroarylation protocol, the coupling of functionalized, electronically unbiased olefins is achieved under air and within minutes of blue light irradiation through the intermediacy of reactive (hetero)aryl radical species with full retention of the DNA tag integrity. Notably, these processes operate under mild reaction conditions, furnishing complex structural scaffolds with a high density of pendant functional groups. ## Introduction The identifcation of specifc binding molecules remains a central theme in drug discovery efforts in academic and industrial laboratories. 1 The global pharmaceutical industry invests over $186 billion annually on research and development to meet the ever-increasing demands for safe and improved therapeutics. 2 In recent years, DNA-encoded library (DEL) technology (Fig. 1) has emerged as a novel interrogation platform to accelerate the advancement of small-molecule modulators of biomolecular targets. 3 Conceptualized by Brenner and Lerner in 1992, 4 DEL platforms confer unprecedented capabilities to overlap the versatility of chemical synthesis with the powerful features of genetic coding, allowing simultaneous testing of combinatorial libraries of exceptional magnitude (>10 6 to 10 12 discrete members). 3 Through "split and pool" synthetic cycles, diverse BBs are encoded by a unique DNA tag functioning as a molecular identifer. 3 Following synthesis, DEL libraries are incubated with immobilized target proteins, after which non-binders are washed away. The chemical structures of the potent ligands are decoded using polymerase chain reaction (PCR) amplifcation and posterior next-generation DNA sequencing. 3 In this vein, DELs feature a more effective and inexpensive discovery format ($0.0002 per library member) compared with conventional high-throughput screening (HTS) methods ($1000 per library member). 5 To be successful, on-DNA chemistries are required to incorporate BBs from readily available chemicals bearing multifunctional handles for further diversifcation under mild, dilute, and aqueous conditions. 6 In light of these considerations, the development of reliable transformations that operate through novel reactivity modes and employ an abundant, diverse set of starting materials would expedite progress in this feld. As part of a program centered on the development of catalytic tools to yield novel structural scaffolds, we recently reported the synthesis of gem-difluoroalkenes, 7,8 carbonyl mimics that display in vivo resistance toward metabolic processes, 9 through photoinduced radical/polar crossover defluorinative alkylation. 7,8,10 As a complementary approach to build chemical diversity, we became interested in pushing the limits of photochemical paradigms to access benzylically tri-fluoromethylated compounds, bioactive structural motifs in medicinal settings (Scheme 1A). 11 Specifcally, fluorine incorporation is a powerful strategy invoked by the pharmaceutical and agrochemical industries to alter a molecule's chemical, physical, and biological properties, such as its pK a , dipole moment, and molecular conformation. 12,13 As a consequence of these factors, fluorinated scaffolds are prevalent in more than 25% of marketed drugs. 11c As an important representative, the trifluoromethyl (-CF 3 ) group renders increased metabolic stability, lipophilicity, and binding selectivity when embedded in therapeutic candidates. 11, 14 Typically, the trifluoromethyl group can be installed through nucleophilic, electrophilic, or radical routes. 15 Although these strategies undoubtedly expand chemical space, these protocols remain elusive in the context of late-stage functionalization and the incorporation of sensitive functional groups in complex environments under DEL-like conditions. An underexplored opportunity to achieve C(sp) 3 trifluoromethylation is the direct hydroalkylation of trifluoromethyl-substituted olefns. Specifcally, the carbofunctionalization of these electrophilic unsaturated systems occurs readily at room temperature with exquisite functional group compatibility, 16 thus rendering the incorporation of pharmaceutically relevant cores and complex alkyl fragments feasible in a library setting. However, given the established propensity of trifluoromethyl-substituted alkenes to undergo intramolecular E1cB-type fluoride elimination in metalcatalyzed cross-couplings that proceed through the intermediacy of a-CF 3 -metal species 17 via the nucleophilic addition of organometallic reagents 18,19 or in the presence of traditional photoredox catalysts irrespective of the nature of the radical precursor (Scheme 1B), 16,19a,20 hydrofunctionalization 21 efforts remain challenging. In particular, the hydroalkylation of trifluoromethyl-substituted alkenes using unactivated alkyl counterparts presents a formidable, yet potentially powerful scenario to rapidly access unprecedented benzylically tri-fluoromethylated building blocks from commodity radical progenitors with a high content of C(sp 3 ) carbons. To address this challenge and unlock access to benzylically trifluoromethylated motifs from commodity chemicals in DEL synthesis, we report a decarboxylative-based, radical-mediated hydroalkylation of DNA-tagged trifluoromethyl-substituted alkenes enabled by the merger of electron donor-acceptor (EDA) complex photoactivation 22,22j,22k and hydrogen atom transfer (HAT) chemistry (Scheme 2). 23 Under blue light irradiation, a commercially available electron donor, Hantzsch ester (HE, diethyl 1,4-dihydro-2,6-dimethyl-3,5pyridinedicarboxylate), functions as a strong photoreductant to induce C(sp 3 ) radical generation from commercially available carboxylic acid derivatives. 24 As part of its dual role, HE subsequently serves as a suitable hydrogen atom donor, impeding the formation of anionic intermediates upon radical addition as well as circumventing the necessity for alkylmetal complexes, species intrinsically primed to undergo b-F elimination in reactions with trifluoromethyl-substituted alkenes. 17e,20, 25 In this vein, the utility of this EDA paradigm is partially driven by its ability to deliver complex trifluoromethyl-substituted hydrofunctionalized products with high C(sp 3 ) carbon counts selectively under mild and open-air conditions. As a complement to the hydroalkylation protocol, a radicalmediated intermolecular hydroarylation of electronically unbiased olefns was developed (Scheme 2). 26 Because alkenes are plentiful, structurally diverse, and versatile commodity feedstocks, readily available from petrochemical and renewable resources, they are ideal precursors for C-C bond formation in DELs, and the strategy developed is based on photoinduced reductive activation of DNA-conjugated (het)aryl halides to deliver reactive (het)aryl radical species that can be harnessed in useful synthetic operations followed by hydrogen atom termination. 23 ## Discussion Recently, synthetic processes driven by EDA complex photochemistry have gained considerable traction, including protocols resulting in borylation, sulfonylation, and thioetherifcation. 22,27 To harness the synthetic potential of EDA complex photoactivation toward DEL platforms, we examined the feasibility of the proposed decarboxylative hydroalkylation using on-DNA trifluoromethyl-substituted alkene 1A and unactivated primary redox-active ester (RAE) 2a as model substrates (Table 1). Under blue Kessil irradiation (l max ¼ 456 nm), efficient conversion to the desired benzylic trifluoromethylsubstituted product 3a was observed using 50 equivalents of the radical precursor under ambient reaction conditions within minutes of illumination. In contrast to radical-mediated alkylation promoted by metal reductants 17e,28 or external photoredox catalysts, 29 this open-to-air EDA paradigm provides an exceedingly low barrier to practical implementation in highthroughput settings and circumvents side reactivity stemming from singlet oxygen generation through triplet-energy transfer. 29 To examine the influence of the dihydropyridine (DHP) backbone on the efficacy of this photochemical manifold, the reaction was conducted with four different DHP derivatives to gain a deeper understanding of their dual reactivity profle in EDA complex photoactivation and HAT catalysis. The C4substituted DHP (HE A, entry 7) thus exhibited no reactivity under the reaction conditions, whereas cyano substitution at C3 and C5 of the DHP (HE B, entry 8) led to diminished product formation. As expected, 4,4 0 -dimethyl HE C (entry 9) failed to promote the reaction, presumably because of competitive back electron transfer (BET) 22a that restores the ground-state EDA complex from its radical ion pair in the absence of a probable photooxidative aromatization event. Notably, commercially available and bench-stable HE 30 displayed optimal performance (84% yield, entry 1), accommodating water as a co-solvent and high dilution conditions (0.3-0.6 mM), with only trace amounts of the corresponding gem-difluoroalkene detected. Using UV/vis absorption studies, a bathochromic shift of the reaction mixture in 8 : 1 DMSO/H 2 O (0.6 mM) was observed, with a wavelength band tailing to 500 nm (see ESI †). This is indicative of the formation of a new molecular aggregate between the electron-defcient aliphatic RAEs and the electron-rich HE. Using Job's method 31 of continuous variation, we determined a molar donor-acceptor ratio of 1 : 1 for the colored EDA complex (see ESI †). Further spectrophotometric analysis at 450 nm revealed an association constant (K EDA ) of 1.2 M 1 of HE with 1-methylcyclohexane-N-hydroxyphthalimide ester using the Benesi-Hildebrand method, 32 highlighting a plausible association event of charge-transfer complexes prior to SET events under the conditions of the reaction. Finally, control experiments validated the necessity of all reaction parameters for effective C(sp 3 )-C(sp 3 ) bond formation. Next, we examined the scope of redox-active carboxylate derivatives using on-DNA trifluoromethylated alkene 1A (Scheme 3). In general, a broad palette of primary aliphatic systems that lack any radical stabilizing factors exhibited excellent reactivity. The method further benefts from broad functional group tolerance, facilitating the introduction of bifunctional handles including ketones (3a, 3f, 3q), aryl halides (3c, 3e, 3n), a terminal alkyne (3d), esters (3j, 3o), substituted alkenes (3k, 3l, 3m, 3o), free alcohols (3o, 3p), as well as medicinally-relevant heteroaromatic scaffolds (3a, 3n). In addition, Boc-and Fmoc-protected amines served as competent substrates. This is crucial in DEL settings, where library members should ideally bear multifunctional BBs that allow subsequent derivatization. The scope was further extended to the modifcation of biologically active molecules displaying a high density of pendant functional groups, including the herbicide 2,4-dichlorophenylacetic acid (3e), long-chain fatty acids (3k-3m), the anti-inflammatory agent indomethacin (3n), mycophenolic acid (3o), as well as various steroids (3p, 3q). In particular, this method provides a clear advantage in terms of scope over previously reported on-DNA photoinduced decarboxylative alkylation protocols, which are largely limited to aheteroatom-stabilized radicals, 7,33 or exclusively restricted to secondary and tertiary radicals. 22k,34 More specifcally, complementary decarboxylative methods employing zinc nanopowder as a reductant under strictly deoxygenated conditions fail to incorporate primary systems on DNA, 28 presumably because of the higher reduction potentials associated with the radical precursor. Most importantly, these methods largely proceed through anionic intermediates, where in the case of the trifluoromethyl-substituted olefns, there is a predominant propensity for intramolecular E1cB-type fluoride elimination 16a,19a,20 to afford the corresponding gem-difluoroalkenes (rather than trifluoromethyl-substituted alkanes) via radical/ polar crossover pathways. 10 In a similar manner, secondary and tertiary radical architectures are harnessed effectively to afford functionalized synthetic frameworks, including scaffolds derived from proteinogenic and non-proteinogenic amino acids (3u-3y), a glycoside (3z), and lipid-lowering agent gemfbrozil (3za) (Scheme 3). The reaction conditions proved general for both acyclic and cyclic carboxylate derivatives, including bridged bicyclics (3zf-3zh), as well as strained ring systems, such as cyclobutanes (3s, 3t) and a cyclopropane (3zc). Notably, trifluoromethylsubstituted bicyclo[1.1.1]pentane (BCP) product 3zg was obtained in good yield. These BCP derivatives serve as bioisosteres for arenes, internal alkynes, and tert-butyl groups in medicinal chemistry settings. 35 With respect to the scope of trifluoromethyl-substituted alkenes, a diverse array of DNA headpieces (DNA-HPs) led to the desired benzylic trifluoromethyl-substituted products without compromising yields (Scheme 4). In general, both electron-withdrawing and electron-donating groups are well tolerated under the developed conditions. Substitution at the para-, meta-, and ortho-positions of the HPs' aryl moieties was explored, whereby efficient decarboxylative photocoupling took place. Furthermore, comparable reactivity was observed for unactivated primary, secondary, tertiary, as well as stabilized benzylic-, a-oxy-, and a-amino radical species, further underscoring the versatility of this photochemical EDA paradigm. The commercial availability and structural diversity of carboxylic acids render them particularly attractive for use as multifunctional BBs in DEL libraries. To validate the modularity of this approach further, we developed a telescoped, one-pot photoinduced decarboxylative alkylation protocol through in situ formation of aliphatic RAEs with N-hydroxyphthalimide tetramethyluronium hexafluorophosphate (HITU), 28 a benchstable solid that can be readily prepared on kilogram scale (Scheme 5). This reagent features great versatility in reaction scope, accommodating a wide array of functional groups including ketone (3f), carbamate (3zi), aryl chloride (3zl), and Boc-protected amines (3zn, 3zo). Using 24-well plates, microdosing of the carboxylic acid, DIPEA, and HITU in DMSO is accomplished under air, followed by 3 h of activation time. The in situ formed RAEs can then be treated directly with a solution of HE and the corresponding DNA headpiece, reaching synthetically useful yields after 10 min of illumination (Scheme 5, workflow). Notably, this HITU-mediated alkylation performs equally well using unactivated-and a-heteroatom-stabilized radical progenitors, presenting a direct route toward C-C bond formation through oxidative quenching modes, an underexplored challenge in DEL-based environments. 28 As an extension of the hydroalkylation chemistry, an on-DNA multicomponent reaction (MCR) was In recent years, MCRs 36 have emerged as a powerful tool to furnish novel scaffolds with inherent molecular complexity from abundant feedstocks. Through sequential bond formation, MCRs enable the sampling of uncharted chemical space to accelerate drug discovery efforts. 36 An underexplored realm in DEL synthesis is the development of olefn dicarbofunctionalization reactions. 3a Specifcally, alkenes serve as versatile BBs that possess functional group-rich handles for derivatization. However, in addition to chemo-and regio-selectivity concerns associated with DEL reactions that rely on high loadings of reagents, these processes are further complicated by the generation of undesired two-component coupling products. Keeping these considerations in mind and taking advantage of the electronically distinct nature of trifluoromethyl-substituted alkenes, a polarity-reversing radical cascade/ trifluoromethylation of olefns has been developed through EDA complex photoactivation between HE and Umemoto's reagent, 22a a commercially available trifluoromethylating agent (Scheme 6). In particular, this open-to-air charge-transfer manifold harnesses electrophilic trifluoromethyl radicals for subsequent addition to electron-neutral or electron-rich alkenes, abundant yet currently underexplored partners in photoinduced DEL synthesis. 3a, 37 The resulting nucleophilic, open-shell radical intermediates may then engage in chemoselective coupling with on-DNA trifluoromethyl-substituted alkenes. As part of its dual role, the HE also functions as a hydrogen atom donor to furnish bis-trifluoromethylated products of signifcance in medicinal settings. 11 Remarkably, the scope of the olefn partner proved general, tolerating diverse functional groups including a free alcohol (11a) and unprotected glycoside 11d. In this vein, we anticipate this mode of catalysis will help inform the design and implementation of unique synthetic disconnections toward complex, bioactive targets in DELs. Having developed suitable conditions for the hydroalkylation of unsaturated DEL platforms, attention was turned toward hydroarylation transformations. Recently, research efforts have validated Ni/photoredox dual manifolds in DEL platforms using carboxylic acids, 7,33b,33e 1,4-dihydropyridines (DHPs), 7 a-silylamines, 8 and alkyl bromides 8,38 as radical precursors. Given our long-standing interest in the design of complex (hetero)aryl scaffolds with high C(sp 3 ) carbon counts, 39 we sought to expand reactivity in DEL synthesis through intermolecular radicalmediated hydroarylation of functionalized olefns to generate alkylarenes (Scheme 7). To develop a complementary approach toward C(sp 3 )-C(sp 2 ) bond formation, we reasoned that singleelectron reduction of DNA-bound, halogenated aryl subunits 26 would grant access to reactive (het)aryl radical species in a regioselective fashion. Subsequent addition to alkenes followed by hydrogen atom termination would deliver unprecedented structures from readily available substrates. Inspired by pioneering work by Beckwith 40 and related, precedented milestones, 26 we hypothesized that photoinduced electron transfer from highly reducing transition-metal-based complexes would enable this strategy under mild reaction conditions. However, because of the high redox potentials associated with organohalides and the propensity of aryl radicals to undergo reduction through rapid HAT, 26a the adaptation of this mechanistic proposal in DEL environments posed challenges. Importantly, aryl radicals have been shown to induce DNA strand damage, 41 underscoring the requirement for a regulated generation of these high-energy intermediates and the necessity for wellorchestrated addition reactions. To achieve chemo-and regioselectivity, the following criteria was considered: (i) the rate of (het)aryl radical addition to unsaturated systems must be competitive with C-X bond reduction stemming from undesired HAT pathways. (ii) The rate of hydrogen atom abstraction by the resulting alkyl radical must be competitive with its addition to another equivalent of the alkene. (iii) The rate of single-electron reduction of the aryl halide should take place preferentially over that of the alkyl radical intermediate. Specifcally, the choice of both photocatalyst and hydrogen atom donor influences product distributions. We determined that 300 equiv. of the olefnic substrate and a 1 : 200 photocatalyst-to-HAT reagent ratio was optimal for reactivity. Toward this end, the combination of fac-Ir(ppy) 3 and HE enabled the construction of alkylated arenes under air within minutes of blue light irradiation. Control experiments demonstrated that all reaction components are necessary for aryl radical generation. With optimized conditions established, we surveyed DNAtagged (het)aryl halides with norbornene as the alkyl source (Scheme 7). Aryl iodide 13A bearing a chloride substituent afforded the desired product with the electrophilic crosscoupling handle intact, delivering linchpins for further functionalization. Electron-neutral iodobenzene 13B as well as derivatives bearing electron-donating groups (13C) or electronwithdrawing groups (13G) served as excellent substrates. Further extension to less activated aryl bromides was also Scheme 6 On-DNA multicomponent trifluoromethylation promoted by photoactive EDA complex activation. Note: several diastereomers are expected to form under these conditions. possible (13D). Notably, electrophilic pyridyl radicals were employed as coupling partners, giving rise to functionalized heteroaromatics (19a, 19j-19n). Finally, in addition to the strained bicyclic norbornene, a broad array of functionalized alkenes was examined. In general, unactivated alkenes bearing unprotected alcohols (15b, 15c, 19j), an ester (19k), ketones (15i, 19l), and an epoxide were all accommodated. In addition, this photochemical paradigm was extended to the modifcation of activated styrene derivatives (15e, 15g) in synthetically useful yields. Even benzylically trifluoromethylated product 15g could be used as a substrate to afford product with complete retention of the bromide handle, presumably because of the high loading of alkene precursor compared to the photoredox catalyst, precluding an overreduction event of the halide. From the standpoint of DEL synthesis, which benefts from minimal reagent input (e.g., 10-25 nmol of HP per transformation), such equivalencies can be leveraged to achieve selectivity and unique reactivity trends that are otherwise untenable in traditional small molecule synthesis. In particular, these halogenated alkenes can further grow DEL libraries through transition metal-catalyzed cross-coupling efforts. ## DNA compatibility with EDA complex photoactivation Because the integrity of the DNA barcode is essential to a successful protein target selection, mock ligations and qPCR amplifcations were performed to evaluate the ability of the RAE hydroalkylation conditions to be used in an actual library production. A model DNA conjugate composed of a representative headpiece ligated to a 4-cycle tag and equipped with a 2base 5 0 overhang was subjected to the standard hydroalkylation conditions. This model DNA conjugate was also subjected to control reactions where either Hantzsch ester or light was omitted. Post reactions, these DNA conjugates were further elongated by ligation to install the necessary PCR primers and quantifed by qPCR. There was no signifcant difference in qPCR amplifcation across the various experiments, suggesting full DNA integrity (see ESI †). These fndings further underscore the utility of EDA paradigms as a general blueprint toward selective on-DNA alkylation under open-to-air conditions. ## DNA compatibility with aryl radical intermediates Mindful of well-established precedent of DNA strand damage in the presence of reactive aryl radical species, 41 the hydroarylation conditions were studied to evaluate the resulting DNA integrity. Again, a model DNA conjugate composed of a representative headpiece ligated to a 4-cycle tag and equipped with a 2-base 5 0 overhang was reacted using the standard conditions. This model DNA conjugate was also subjected to control reactions where either Hantzsch ester, photocatalyst, or light was omitted. Post reactions, these DNA conjugates were further elongated by ligation to install the necessary PCR primers and quantifed by qPCR. There was no signifcant difference in qPCR amplifcation across the various experiments, suggesting full DNA integrity (see ESI †). These results emphasize the mild nature of the developed photoredox paradigm, whereby the formation of reactive aryl radical intermediates in a regulated fashion facilitates productive on-DNA alkylation. ## Conclusions In summary, this report demonstrates a mechanistically-driven proof-of-concept for the implementation of charge-transfer complex activation as an enabling technology to introduce diverse C(sp 3 )-hybridized architectures from commodity chemicals in DEL platforms (including unactivated primary, secondary, tertiary, as well as stabilized benzylic, a-alkoxy, and a-amino systems). Specifcally, this EDA paradigm was utilized to achieve the selective decarboxylative-based hydroalkylation of trifluoromethyl-substituted alkenes through radical/HAT crossover to access complex benzylic trifluoromethylated scaffolds, unlocking a complementary reactivity outcome to established carbodefluorinative protocols mediated by an external photoredox catalyst. Furthermore, a general intermolecular hydroarylation protocol of electronically unbiased olefns through selective C-X bond activation in DNA-tagged organohalides is reported. Remarkably, this photochemical paradigm delivers reactive (hetero)aryl radical species in a regulated fashion without compromising the DNA integrity. Notably, these open-to-air processes are chemoselective, operate under mild and dilute reaction conditions, and are completed within minutes, rendering them suitable for late-stage functionalization and high-throughput settings in the pharmaceutical industry. We anticipate these fndings will expedite drug discovery research and provoke further development in radicalmediated DEL synthesis.

chemsum

{"title": "Photoredox-mediated hydroalkylation and hydroarylation of functionalized olefins for DNA-encoded library synthesis", "journal": "Royal Society of Chemistry (RSC)"}

success_of_montreal_protocol_demonstrated_by_comparing_high-quality_uv_measurements_with_“world_avoi

## Abstract: the Montreal protocol on Substances that Deplete the ozone Layer has been hailed as the most successful environmental treaty ever (https://www.unenvironment.org/news-and-stories/story/ montreal-protocol-triumph-treaty). Yet, although our main concern about ozone depletion is the subsequent increase in harmful solar UV radiation at the earth's surface, no studies to date have demonstrated its effectiveness in that regard. Here we use long-term UV Index (UVI) data derived from high-quality UV spectroradiometer measurements to demonstrate its success in curbing increases in UV radiation. Without this landmark agreement, UVi values would have increased at mid-latitude locations by approximately 20% between the early 1990s and today and would approximately quadruple at midlatitudes by 2100. In contrast, an analysis of UVI data from multiple clean-air sites shows that maximum daily UVI values have remained essentially constant over the last ~20 years in all seasons, and may even have decreased slightly in the southern hemisphere, especially in Antarctica, where effects of ozone depletion were larger. Reconstructions of the UVi from total ozone data show evidence of increasing UVI levels in the 1980s, but unfortunately, there are no high-quality UV measurements available prior to the early 1990s to confirm these increases with direct observations.Concern about ozone depletion arose primarily because of its potential to increase UV-B radiation, and the consequent effects on health and the environment 1 . Observations of unexpected springtime decrease in stratospheric ozone over Antarctica 2 -commonly referred to as the "ozone hole" -led to the rapid adoption of the Montreal Protocol on Substances that Deplete the Ozone Layer in 1987 3 . It has been shown that without this treaty and its subsequent Amendments and Adjustments, ozone holes would by now also be occurring over the North Pole, resulting in highly elevated UVI in the Arctic 4 . Even at mid-latitudes, the UVI would by now have increased markedly, and would more than double at some latitudes by the middle of the 21 st century [5][6][7] . However, due to the success of the Montreal Protocol, ozone decreases appear to have been brought under control 8 , though any effects on surface UV irradiances have not previously been confirmed by measurements.We attempt to redress this issue by using an analysis of long-term UV measurements from instruments that meet the stringent requirements of the Network for the Detection of Atmospheric Composition Change (NDACC, www.ndacc.org). Observatory global irradiance measurements were available every 30 min prior to September 2010, and every 15 min after that date. At Villeneuve d' Ascq, scans are performed every 30 min. We considered whether to use peak daily values, peak values during a specified time window (e.g., ±1 hour of local noon), or mean values over the noon period (e.g., ±0.5 hour of local noon). Variations in seasonal mean UVIs calculated from the three quantities agreed to within ±3%, which is smaller than the measurement uncertainty of ±5% 10 . By using peak values instead of mean values, cloud effects on seasonal averages are reduced and time-series derived from measurements therefore agree slightly better with our modelling results, which do not take attenuation from clouds into account. The following analysis is therefore based on seasonal means of the daily maximum UVI observed within ±1 hour of local noon (12:00 UTC at Greenwich; 00:45 UTC at Lauder, etc). Data gaps were treated as described by Bernhard 18 . In brief, single missing days were "filled in" by calculating the average of the UVI of adjacent days. Data gaps lasting for more than one day were corrected by taking climatological variations in SZA and ozone into account. If more than 30 days were missing within a 90-day period, seasonal means were excluded from further analysis. Seasonal changes in UVI are large at mid-to high-latitudes. For latitudes poleward of 45°, UVI values in winter are less than 10% of the summer peaks. And in polar regions, the UVI reduces to zero during the polar night. Additionally, there are large day-to-day differences in UVI due to changing cloud conditions and types. To minimize these effects, we consider seasonal means of approximately 90 days (December-February, March-May, June-August, and September-November), which are also more relevant from an environmental perspective. Measurements of UVI from the sites in Table 1 are compared with values calculated for clear skies for four different ozone scenarios: (1) ozone values from the NIWA/BS global total-column ozone assimilation, based on measurements between 1978 and today 19 , as described at www.bodekerscientific.com. (2) ozone values projected in the World Avoided scenario from 1974 to 2065 6 , which was calculated using the GEOS-CCM (Goddard Earth Observing System -Chemistry Climate model (NASA, USA)). ( 3) ozone values projected in the World Avoided scenario from 1974 to 2100 7 which was calculated with the NIWA-UKCA (NIWA-UK Meteorological Office Climate Assessment) model. (4) ozone values projected for the World Expected scenario, as calculated with the NIWA-UKCA model. This scenario is the REF-C2 experiment described by Morgenstern et al. 20 . In this simulation, ozone-depleting substances follow the "WMO (2010) A1" scenario, which assumes compliance with the Montreal Protocol. Differences between the two CCM models are summarized in Table 2. Details of both models are given elsewhere 16, . We note that future changes are dependent on the RCP scenario, and in particular depend on future changes in CO 2 , CH 4 , and N 2 O 25 . For example, a strong overshoot in ozone would be expected for RCP 8.5, and by the end of the 21 st century, differences between N 2 O and CH 4 scenarios may account for differences in ozone larger than 5%. For all scenarios, the clear-sky UVI at noon was interpolated from a 5D table of UVI as a function of ozone amount, solar zenith angle (SZA), altitude (pressure), aerosol optical depth, and surface albedo that had been pre-computed with the discrete-ordinate 26 implementation in the TUV radiative transfer model 27 . Subsequent corrections were made to account for seasonal variations in Earth-Sun separation. In the World Avoided scenarios (Scenarios (2) and ( 3)) and World Expected scenario (Scenario (4)), the UVI was calculated for sea level with no aerosols and low surface albedo. Although both SO 2 and NO 2 can affect UVI, they have not been included in the clear-sky calculation. With the exception of San Diego, Thessaloniki, and Melbourne, sites considered in this paper are "clean-air" sites where the effect of absorption by these trace gases can be considered negligible. To better approximate the measurements, corresponding UVI calculations using the ozone data at each site (Scenario (1)) were computed with TUV for the altitude at each site, assuming an aerosol optical depth of 0.05 (at wavelength 0.5 µm) with a single scattering albedo of 0.90 for all sites, and a representative (annually invariant) surface albedo, ranging from 0.05 at snow-free sites, to 0.98 at the South Pole. As shown below, the comparison between the measured UVI data and the UVI data calculated from the assimilated ozone dataset (Scenario (1)) demonstrates that year-to-year variability in the UVI can be estimated with sufficient accuracy from ozone changes. The observed good agreement gives us confidence that UVI calculated from ozone for the period prior to the 1990s, when no direct UVI measurements are available, can be used to infer changes in UVI over this period. The comparisons with the two World Avoided scenarios were included to demonstrate the divergence between what has actually happened compared with what would have happened without the Montreal Protocol, using two independent model calculations. ## Results from Model Simulations We begin by comparing projected ozone fields from the two World Avoided models (Scenarios (2) and ( 3)). Predicted ozone values from these two models are compared for selected latitudes in Fig. 1. The two models are in general agreement in long-term behaviour, but also show significant differences over shorter time scales, especially at low latitudes. Close agreement is not expected at time scales of less than 5 years because there is a random component to the way modes of variability such as the quasi biennial oscillation (QBO) manifest themselves in the model runs. The systematic differences between the two models are within the usual range for chemistry-climate models 28 . Modelled ozone data for Scenarios (2), ( 3) and (4) were compared with measured ozone data (Scenario (1)) and results are provided as Supplementary Data. In general, modelled and measured total ozone column amounts (TOCs) agree reasonably well. For the reference period 1978-1987 (i.e., the period between the year when ozone measurements from space became available globally and the year when the Montreal Protocol came into effect), TOCs calculated with the NIWA-UKCA model (both World Expected and World Avoided) for sites between 45°S and 45°N -which represents 71% of the globe -are on average 8 ± 4% (±1σ) higher than the measured TOCs. The maximum bias for this latitude range is 14%. TOCs calculated with the GEOS World Avoided model are slightly smaller. It overestimates the measured TOC by 2 ± 5% on average, with a maximum bias of 12%. Deviations between the modeled and measured TOCs become larger at higher latitudes, especially in the southern hemisphere. In this latitude range, differences can exceed 20% with biases from GEOS model exceeding those from the NIWA-UKCA model. These larger biases are partly a consequence of the low ozone amounts there and partly due to the difficulty to correctly model the destruction of ozone by heterogeneous chemical processes, which strongly depend on the temperature of the lower stratosphere. Furthermore, specific sites are not necessarily well represented by the zonal means used in the model. This is especially true for high-latitude southern hemisphere sites (i.e., Ushuaia and Palmer Station) in spring when the position of the ozone hole is frequently displaced toward the South American quadrant. Some of the discrepancy may also be due to interpolation errors in the ozone assimilation during the polar night, but this is irrelevant for changes in the UVI. Even though some uncertainties remain in the models, the agreement in geographic and seasonal patterns gives confidence that they can be used to project future changes. Finally, we also note that there may be errors in the model projections if factors such as changes in the Brewer-Dobson circulation are not properly parameterised. Previous studies have also shown that the CCM models tend to overestimate ozone amounts. For example, the GEOS-CCM used previously 6 has a high-ozone bias at mid to high latitudes 29 , which would lead to UVI estimates that are too low by approximately 10% at latitude 45° S. The green lines in Fig. 1 are the mean differences between the two models (smoothed), which shows that they depend on both latitude and time. In the longer-term, projections by the two models are in good agreement. The reasonable level of consistency between these models adds further confidence to those predictions but also illustrates some significant differences. It is interesting to note that by the year 2100, projected ozone declines are largest at high latitudes, and smallest in the tropics. By that time, the lowest mean ozone amounts would be in polar regions, and the highest in the tropics. time series for lauder. We now compare these UVI projections from the CCM models with observed changes in UVI and with those expected from clear-sky calculations using the assimilated ozone values. Results for Lauder are shown in Fig. 3. Figure 3a displays the time series of daily ozone derived from the NIWA/BS ozone assimilation for the period since satellite measurements became available. In Fig. 3b, the peak UVI measured within one hour of local noon is compared with the UVI calculated for clear-sky conditions using the NIWA/BS ozone data. There is good agreement in the seasonal range between measured and modelled UVI. Because of the large seasonal and day-to-day variability, it is difficult to see any long-term effects (other than in the upper envelope of data) in Fig. 3b. The lower four panels of Fig. 3 redress this by considering seasonal means, with each data point representing the mean over different 3-month periods. Because of the previously discussed biases in calculated and modelled ozone, all data shown in Fig. 3c-f were normalized. The measured UVI values were normalized to the average of all values, while the calculated UVIs were normalized to the average of only those years with measurements. The normalized peak UVI changes derived from measurements (blue lines) are compared with those calculated from the assimilated ozone dataset (black lines). Also shown in these panels are the corresponding UVI changes that would have occurred according to the World Avoided calculations reported by the GEOS-CCM model (broken red lines) and the NIWA-UKCA model (broken magenta lines). UVI values in the World Expected are also shown (broken green lines). The heavier solid lines for each model run include smoothing with an approximating spline. Both World Avoided datasets and the World Expected dataset were normalized to the average of the calculated UVI values of the years 1978 to 1987; the decade that immediately preceded the signing of the Montreal Protocol. Corresponding plots for all other sites are in Supplementary Data. ## Model In the period prior to 2020 shown in Fig. 3, projected changes are generally smaller with the NIWA-UKCA model than with the GEOS-CCM model. However, this is not everywhere the case. At northern mid-latitudes, the two models are in close agreement (see Supplementary Data). They are also similar at high southern latitudes during summer, while the NIWA-UKCA values are larger at high northern latitudes. Reasons for these model differences are outside the scope of the present study. Green lines indicate the smoothed difference between the two models (i.e., NIWA-UKCA minus GEOS-CCM). Plots for latitude 15° are representative of all equatorial latitudes, while those for 75° are representative for higher polar latitudes. ## UVi in summer. Possible changes in UV radiation during summer are most relevant from an environmental and health perspective. In Fig. 4, we compare measured, calculated, and projected UVIs for summer at all 17 sites in Table 1. The figure shows that the World Avoided trends depend greatly on latitude and that differences between the two model runs are also latitude-dependent. For example, at latitude 40°N, there is close agreement between the two models (for other seasons, see Supplementary Data). With the exception of Hoher Sonnblick, the measured data from all sites closely match the calculated values and are much closer to the World Expected curves than the World Avoided curves. However, there is a significant divergence between the measured values and World Expected values at higher latitudes: measured and calculated UVI ratios are more or less constant over time while the World Expected curves have a broad maximum during the first decade of the 21 st century (last two panels of Fig. 4). This suggests that the impact of the large spring-time ozone depletions persists for too long into the following summer in the model. This is a well-known limitation of CCMs 28 for which the causes are not fully understood, and is an area of current research. The measurements at Hoher Sonnblick (Panel f) show a continuation of the previously reported upward trend in summertime UVI 30 , which was attributed to changes in aerosols or cloud cover. This increase is not present in all seasons, so it is not a calibration artefact. However, there are several seasonal gaps in the data, and the year-to-year variability is also much greater at this mountainous site (see Supplementary Data), possibly due to variable snow cover. At some other sites, the length of the time series is too short for reliable trend estimates. Only nine of the 17 sites have more than 20 years of data coverage. Figure 4 also indicates that UVI ratios calculated from the assimilated ozone data are generally less than unity prior to the 1990s and before the start of UV measurements. The slope in the calculated data is generally consistent with the slope in the datasets of the three CCM models. These model calculations imply that considerable changes in summer UVI occurred between 1978 and 1990 (about 5% at northern mid-latitude sites, up to 10% at southern mid-latitude sites and up to 20% at the three Antarctic sites). However, direct UVI observations needed to confirm these changes do not exist. After accounting for differences in elevation between the sites, we note that latitude-for-latitude, the mean summer UVI values tend to be larger at the southern hemisphere (SH) sites compared with the northern www.nature.com/scientificreports www.nature.com/scientificreports/ hemisphere (NH) sites, as has been predicted and observed previously . This NH/SH difference is as expected. For example, Fig. 2 shows that at 45°S, the peak UVI value is larger than at 45°N by approximately 13%. Approximately half of this difference is due to ozone differences and half is due to Sun-Earth separation differences. The much larger NH/SH differences reported in the past also include effects of clouds and aerosols, which were not considered in the CCM calculations. The NH/SH asymmetry is projected to be only slightly smaller by year 2100. UVi in spring at high latitudes. In Fig. 5, we compare normalized measured, calculated, World Avoided, and World Expected UVIs in spring for the Arctic and Antarctic sites included in this study, where the largest long-term changes in UVI have been observed. Note the expanded vertical axis scale compared with that in Fig. 4. For these high-latitude sites, the "calculated" and "measured" data sets are in almost perfect agreement, particularly in Antarctica, indicating that ozone variation is the largest contributor to UVI changes. Variability of clouds and albedo is a minor contributor in comparison. The "hump" in the World Expected data noted previously for the summer data (Fig. 4), which peaks in the early part of the 21 st century, is also present in the spring data (Fig. 5). However, in the spring period, there is also a hint of this in the measured and calculated data for the two Antarctic sites and perhaps also at Summit. Data from Barrow show more variability, as one would expect from an Arctic coastal site that is affected by changes in albedo from snow and sea ice. summer (e) and autumn (f) determined from measurements of the NDACC instruments (blue), calculated from total ozone column (black), and projected by the two World Avoided (red and magenta) and World Expected (green) CCM model runs. UVI changes were normalized as described in the text and the "UVI ratios" shown in Panels (c-f) are ratios relative to these normalizations. Note that for the season that spans two years (summer at this southern hemisphere site), the year label refers to the year at the start of the season. For example, the summer of Dec 2017 to Feb 2018 is plotted at 2017. ## Decadal trends in UVi. Due to the less complete data coverage early in the period, our statistical analysis is restricted to the period since 1996. This period also excludes effects from the eruption of Mt Pinatubo, which led to significant effects on ozone and UV at some locations . The period since 1996 also includes approximately two complete 11-year solar cycles, which cause variations in ozone of ±2% 39,40 that would lead in turn to a modulation of similar relative magnitude, but of opposite sign, in the UVI. Even for the period since 1996, measured data are unfortunately incomplete at some sites, and only a small number of sites have full data coverage through to the end of 2018 (see Table 1 www.nature.com/scientificreports www.nature.com/scientificreports/ (BSI). Their time series from Ushuaia and San Diego are also restricted to the period before 2010 and 2008 respectively. There was also a gap from 2003 to 2008 in Melbourne data. Of the 17 sites, only three (Lauder, Palmer, and Thessaloniki) have near-complete data coverage for the full period from 1996 to 2018. However, six others (Barrow, Hoher Sonnblick, Boulder, Mauna Loa, Arrival Heights and South Pole) have only a few years of missing data over that period. Greatest confidence should be placed on data from these nine stations. There are other possible forcing mechanisms that drive interannual variability in UVI. These include variability due to the QBO or the El Nino Southern Oscillation (ENSO). However, over time scales as long as 22 years, any such effects are small, and have not been considered in the statistical analysis. In these seasonal averages, where each data point is separated from the previous value by a period of 9 months, any statistical auto-correlation effects should also be small. In the trend analysis, we have assumed that any long-term changes over the period 1996 to 2018 could be represented by a linear change, which is a reasonable assumption (e.g., see Fig. 4). Results of the statistical analysis are shown in Fig. 6. The results are not particularly sensitive to the start year, resulting in a similar picture (not shown) if the start-date for the analysis is changed by 1 or 2 years. Note that the error bars shown for the trend estimates in Fig. 6 are 2-σ uncertainties of the regression model, which also include small random uncertainties in the measurements that propagate to the 90-day averages analysed here. However, systematic errors due to potential long-term drifts in the measurements (e.g., due to the transfer of irradiance scales between calibration standards) are not included. Figure 6 shows the following: 1. The two World Avoided Simulations give similar trends: With the exception of Palmer Station in winter, there is good agreement between the trends derived from the two World Avoided simulations. (At Palmer Station, the sun is only 2° above the horizon (SZA min = 88°) at the winter solstice, so peak UVI values during this period are small.) The calculated UVI is also sensitive to small modelling and sampling differences (see Supplementary Data). Also, as shown in Fig. 1 (lower left panel), there are significant differences in the seasonal variability between the two model projections in this latitude region and period. 2. UV Trends are large in the World Avoided Simulations: In the World Avoided (without the Montreal Protocol), UVI levels would have increased over the last 22 years by approximately 50% per decade at high southern latitudes in spring, and by 30% per decade in summer. Increases of up to 20% per decade would have been seen at northern high latitudes. At midlatitudes, the increases would have been approximately 5-10% per decade. Such changes would have been clearly detectable in the measurement data. 3. Measurements Differ from the World Avoided Simulations: In the spring and summer, the observed trends for all nine sites with good data coverage (solid symbols in Fig. 6) are significantly different from the World Avoided trends at all mid-and high-latitudes in the southern hemisphere, and at high latitudes in the northern hemisphere. At northern mid-latitudes, differences between measured trends and trends from the World Avoided simulations are significant in summer at all sites except Hoher Sonnblick, but in other seasons the results are more mixed. 4. Measurements follow the World Expected Simulations: For all sites, trends calculated from the measurements are consistent with the World Expected scenario, though uncertainties are large at Hoher Sonnblick, and other sites (e.g., San Diego, Saint Denis and Obs. Haute-Provence), where there are only a few years of observation. While variability in UVI is close to that predicted by changes in ozone at the southern hemisphere sites, the situation is more complex at northern mid-latitude sites, such as Hoher Sonnblick, where the effects of changes in aerosols and clouds are more important in some seasons. www.nature.com/scientificreports www.nature.com/scientificreports/ confirmed, the downward trend at Arrival Heights would remain statistically significant. At the South Pole and Arrival Heights, both measured and calculated UVIs appear to have decreased during spring (the period most affected by the ozone hole), however, trends calculated from these changes are not yet statistically significant. At Lauder, small downward trends in measured UVI are observed for summer and autumn, which are on the verge of being statistically significant. These trends are generally consistent with trends derived from the NIWA/BS ozone dataset, although throughout the southern hemisphere, trends are systematically more negative for measurements than for calculations for all seasons. Where data are available in the early 1990s, most sites show evidence of increasing UVI in the earliest part of the record when ozone was declining. However, the short length of the data record prior to 2000, and possible interference from the eruption of Mt Pinatubo, hinder our ability to ascribe changes in UVI to changes in ozone. In most cases, we find that calculated and measured trends have been small and are in most cases not significantly different from zero. With the exception of one site (Hoher Sonnblick), they also agree with each other to within their error bars, and with trends calculated with the World Expected model. At clean-air sites in the southern hemisphere, the UVI has followed the World Expected scenario within the limits of the measurement uncertainty. Differences between measurements and the two World Avoided models are already highly significant in the Arctic, and at southern hemisphere sites, especially in Antarctica. The situation is more complex at mid-latitudes in the northern hemisphere, where the effects of changes in aerosol and clouds mask effects of ozone. For winter and autumn, changes observed at mid-latitude sites are mixed. Measurements show a tendency towards decreasing UVI in the southern hemisphere and increasing UVI in the northern hemisphere. However, few of the observed long-term changes since 1996 are statistically significant. A statistically significant decrease in measured summertime UVI has been observed at the Antarctic site Arrival Heights, and the reduction in UVI at Lauder is close to being statistically significant at the 95% level. However, at both sites, trends calculated from ozone remain close to zero, showing that the reductions in UVI must be due to other factors (e.g., changes in cloud, aerosol, or surface albedo). Measurements during the last few years at Lauder also show signs of decreases in UVI, which also cannot be attributed to ozone changes. The Montreal Protocol has been effective in curbing increases in harmful UVI. Without the Montreal Protocol, UVI values at northern and southern latitudes <50° would by now be 10 to 20% larger in all seasons compared to UVIs observed during the early 1990s. These changes would have had grave consequences for public health and would have led to increases in skin cancer occurrences 41 . For latitudes >50°S, UVI values would have increased over the same period between 25% (Ushuaia in summer and autumn) to more than 100% (South Pole in spring and summer). UVI values in the future remain uncertain because of: a. possible volcanic eruptions, which could temporarily exacerbate ozone depletion as long as chlorine levels remain elevated; b. interactions with other aspects of climate change, such as changes in clouds and aerosols; c. slowly varying natural modes of variability 42 , such as the Interdecadal Pacific Oscillation 43 , which may affect cloud cover but are not included in the study; d. effects of increasing greenhouse gases, which will lead to stratospheric cooling and changes in dynamics that may subsequently cause ozone to increase above levels observed in 1980 outside polar regions; e. non-compliance to the Montreal Protocol, such as found in a recent study 44 . The reduced number of operational sites measuring UV irradiance is concerning. The value of time series data increases with the length of the record, as shown by the smaller error bars for the longer-term measurement sites in Fig. 6. Unfortunately, there are only a few high-quality NDACC sites with data from the early 1990s that are still operational. It is important that they continue to be supported in case of unexpected future changes.

chemsum

{"title": "Success of Montreal Protocol Demonstrated by Comparing High-Quality UV Measurements with \u201cWorld Avoided\u201d Calculations from Two Chemistry-Climate Models", "journal": "Scientific Reports - Nature"}

a_three-dimensional_tetraphenylethylene-based_fluorescence_covalent_organic_framework_for_molecular_

## Abstract: The development of highly-sensitive recognition of hazardous chemicals, such as volatile organic compounds (VOCs) and polycyclic aromatic hydrocarbons (PAHs), is of significant importance because of their widespread social concerns related to environment and human health. Here, we report a three-dimensional (3D) covalent organic framework (COF, termed JUC-555) bearing tetraphenylethylene (TPE) side chains as an aggregation-induced emission (AIE) fluorescence probe for sensitive molecular recognition. Due to the rotational restriction of TPE rotors in highly interpenetrated framework after inclusion of dimethylformamide (DMF), JUC-555 shows impressive AIE-based strong fluorescence. Meanwhile, owing to the large pore size (11.4 Å) and suitable intermolecular distance of aligned TPE (7.2 Å) in JUC-555, the obtained material demonstrates an excellent performance in the molecular recognition of hazardous chemicals, e.g., nitroaromatic explosives, PAHs, and even thiophene compounds, via a fluorescent quenching mechanism. The quenching constant (K SV ) is two orders of magnitude better than those of other fluorescence-based porous materials reported to date. This research thus opens 3D functionalized COFs as a promising identification tool for environmentally hazardous substances. ## Introduction The demand for sensing environmentally and biologically important molecules has attracted wide attention in exploiting fluorescent probes. 1,2 During the detecting process, molecular motions of a movable fluorescent probe can be transformed by the fixing analyte or micro-environment, thereby leading to significant alterations in visual signals. Since the pioneer work of Aggregation-Induced Emission (AIE) phenomenon in 2001 by Tang group, 6 AIE molecular rotors play an essential role in the behavior of fluorescence emission. 7 AIE-based fluorescent molecules are non-emissive in their monomers, but turn into highly emissive in molecular aggregates due to the restriction of intramolecular motions. 2,8 Recently, tetraphenylethylene (TPE) and its derivatives, as one of the most important AIE luminogens, have been widely reported, which can act as AIE-active fluorescent probes for chemical sensors. In particular, TPE-based molecules have been demonstrated to promote exciton migration and enhance luminescence activity in porous materials, e.g., supramolecular coordination complexes, 12-14 porous polymers, and metal−organic frameworks (MOFs). Covalent organic frameworks (COFs) are a class of charming crystalline porous materials, which are constructed from organic building blocks linked by covalent bonds. Over the past decade, COFs has driven considerable research efforts in various application fields including gas adsorption and separation, heterogeneous catalysis, organic electronics, and many others. Interestingly, some studies have also proved the combination of COFs and TPE-based AIE molecular rotors to enhance the fluorescence intensity. 6,7, 43 For example, Jiang group has reported highly emissive boronate-linked COFs with TPE monomers, and showed the sensitive fluorescence in the presence of ammonia vapor. 44 Wang group has obtained an AIEgen-based COF by introducing TPE rotors, which emits yellow fluorescence upon excitation with a photoluminescence quantum yield of 20%. 45 Recently, Zhao and co-workers have realized that the COF nanosheets could display signal amplification effect in biomolecular recognition of amino acids and small pharmaceutical molecules (Ldopa). 46 Despite the above processes, TPE-based COFs for molecular recognition is limited; in particular, highly-sensitive recognition of hazardous chemicals by TPE-based COFs has yet to be reported. Here, we report a 3D TPE-based COF (termed JUC-555) as an AIE fluorescence probe for highly-sensitive molecular recognition. Different from all previously reported structures, where TPE function as the linkers in the skeletons, JUC-555 features dangling TPE in the pores. When the pores were filled with a suitable solvent such as DMF, the rotational restriction of TPE rotors in 10-fold interpenetrated diamondoid (dia) framework induced by confinement endows JUC-555 with exceptional AIE-based fluorescent (the on-state). In addition, owing to the large pore size (11.4 ) and suitable intermolecular distance of TPE (7.2 ) in JUC-555, the obtained material indicates an impressive performance in the special recognition of hazardous chemicals, e.g., nitroaromatic explosives, PAHs, and even thiophene compounds, via a fluorescent quenching mechanism (the off-state). The difference between the on-state and the off-state is exceptional with quenching constant (K SV ) approaching close to 10 9 . To the best of our knowledge, these results showed two orders of magnitude higher sensitivity to those of fluorescence-based porous materials reported to date, such as Eu-MOF, 47 NUS-25 nanosheet et al., 48 demonstrating 3D functionalized COFs as a promising identification tool for environmentally and biologically important analytes. 1a). To construct a 3D architecture, we further choose a typical tetrahedral organic linker, tetra(4-aminophenyl)methane (TAPM, Fig. 1a) as a 4-connected building unit. Based on the Schiff-base chemistry, the condensation of BFTP and TAPM results in a novel 3D COF (JUC-555, Fig. 1a). Since the BFTP as a linear building unit shares the same length (~16.9 ) with that of 4,7-bis(4-formylbenzyl)-1H-benzimidazole (BFBZ) from LZU-79, the newly synthesized COF is expected to have 10-fold interpenetrated dia network (Fig. 1b-d). 49 The structure of JUC-555 (Fig. 1c) was determined by powder X-ray diffraction (PXRD, Supplementary Fig. S2) combined with structural simulations (Supplementary Fig. S3). After a geometrical energy minimization by using the Materials Studio 7.0 software package 30 based on the 10-fold interpenetrated dia net (Fig. 1d) with disordered TPE side chains, the unit cell parameters of JUC-555 were obtained (Fig. 1b, a S1) nearly equivalent to the predictions with good agreement factors (Rp = 3.39% and ωRp = 4.30%). Some peaks after 2 ˃ 15° were strong due to disordered TPE molecules in the channel or dynamic effect of 3D framework. 50,51 It should be noted that a similar structure with 10-fold interpenetrated dia net (LZU-79) has been proved by its single crystal, and the PXRD pattern of JUC-555 was well consistent with that from LZU-79 (Supplementary Fig. S4). 49 According to these results, JUC-555 showed a microporous framework with a diameter of about 11.4 (Fig. 1c), which has been demonstrated by the nitrogen (N 2 ) adsorption-desorption isotherm at 77 K (Supplementary Figs. S5-7). Aligned TPE side chains in the pores of JUC-555 after optimization displayed a maximum layer to layer distance of 7.2 and a minimum layer to layer distance of 3.7 (Fig. 1e). Furthermore, JUC-555 was stable in various organic solvents and water (Supplementary Fig. S8), and was thermally stable up to 400 °C under nitrogen according to the thermogravimetric analysis (Supplementary Fig. S9). FT-IR and solid state 13 C NMR shown in Supplementary Figs. S10 and S11, respectively, also confirmed the successful transformation of aldehyde and amine groups to the C=N bonds in JUC-555. ## AIE characteristics of JUC-555 Different from previous reports where TPEs were integrated as the linkers of the crosslinked skeletons, our TPE units in BFTP are designed as the side chains, which endows TPE rotors with more flexibility. In addition, upon inclusion of molecules of different sizes, the degree of molecular congestion in the pores can be tuned, which offers an ability to tune the photoluminescence of TPEs due to the AIE effect. As shown in Supplementary Fig. S12, the ultraviolet−visible (UV−vis) spectrum of the BFTP monomer exhibited an absorption band at 352 nm, whereas that of JUC-555 exhibited a peak centered at 339 nm. Upon excitation, the solid samples of the BFTP compound emitted blue luminescence with peak maxima at 466 nm and JUC-555 emitted brilliant blue luminescence with peak maxima at 482 nm. JUC-555 is yellow as a solid, and shows a greenish yellow color under UV irradiation of 365 nm. As shown in Supplementary Fig. S13, the emission color of BFTP monomer and JUC-555 show CIE coordinates of (0.16, 0.20) in the blue region and (0.21, 0.35) in cyan color, respectively. To have a better understanding of the dynamic behavior of BFTP rotor in JUC-555, we explored the AIE fluorescent properties of BFTP monomer and JUC-555 by using the mixed solvent of THF and water with different water fractions (f w ). As shown in Supplementary Fig. S14, the BFTP monomer exhibited a typical AIE characteristic. When the water component of the mixed solvent was increased to 60%, the fluorescence emission showed a sudden enhancement, and the highest fluorescence intensity was obtained at f w = 90%, which is 14.5-fold higher than that in pure THF solution (f w = 0%). Such AIE characteristic can be attributed to the rotational restriction of the phenyl rings (AIE molecular rotor) in the aggregated state in poor solvents. However, this fluorescence enhancement is much weaker for JUC-555 (a merely 1.35fold increase at f w = 90% as shown in Supplementary Fig. S15) because of the lack of rotational restriction of AIE molecular rotors in the channel. We also measured the fluorescence of JUC-555 in different organic solvents (Supplementary Fig. S16). It's quite amazing that JUC-555 showed exceptionally stronger luminescence in DMF than in other organic solvents. Photoluminescence quantum yield (PLQY) measurements also confirmed that JUC-555 in DMF showed much higher QY than those in other solvents (Supplementary Fig. S17). A quantitative comparison of the PLQY between the monomer and JUC-555 under different conditions was also investigated (Fig. 2a). BFTP monomer showed a moderate quantum yield of 8% as a solid and a very low quantum yield of 0.2% in DMF solution. When water content in DMF was raised to 90%, a quantum yield of 6% was achieved due to AIE effects of the monomer. As for JUC-555, the solid showed only a low QY of 0.6% with a lifetime of 2.187 ns (Supplementary Fig. S18), but a high QY of 13.2% was achieved in DMF solution. Dynamic Light Scattering (DLS) results show that the mean particle size of JUC-555 in DMF is around 1 μm (Supplementary Fig. S19). Such JUC-555 in DMF solution was very stable under ambient conditions (Supplementary Fig. S20). Comparing to the emission spectra of JUC-555 in other solvents, a strong red-shift to 550 nm in DMF along with an over 20-fold luminescence were observed (Supplementary Fig. S16). Such enhancement could also be explained by AIE effects due to the inclusion and molecular recognition of DMF in the pores. The conformation of DMF molecules in COF channels was simulated by Materials Studio 7.0. As shown in the optimized geometry in Fig. 2b, DMF molecules were perfectly aligned in the channels with a comfortable packing coefficient of 27.96% and the intermolecular π-π distance is 3.4 . Compared with that of 3.7 in empty JUC-555, a much tighter packing of TPE phenyl is expected. The O•••H distance of 2.2 and O•••H-N angle of 154.9 o shown in Fig. 2C indicate intermolecular H-bonding between the imidazole moiety of JUC-555 and DMF solvent molecules, which is consistant with the observation in the IR spectra (Supplementary Fig. S21) where the N-H vibration in imidazole moiety of JUC-555 and the C=O stretching in included DMF shifted to lower frequency and the peak being broadened when compared that that of the monomer and DMF, respectively. Such intermolecular H-bonding and the increased molecular crowding induced by inclusion of DMF in the pores rigidify the molecular conformation and impede the intramolecular motions, hence endowing JUC-555 in DMF with an over 20-fold luminescence increase. 52 Based on the strong luminescence feature of JUC-555 in DMF, we then tested it as fluorescent sensors. Firstly, we focused on the detection of explosives, nitroaromatic compounds, which are recognized as one of the major classes of dangerous as well as their highly explosive nature. For efficient nitro explosive detection, it requires high sensitivity and unveils quick response even at low concentrations. When JUC-555 was treated with nitrobenzene (5.0 × 10 -8 M), the fluorescence quenching processes were obvious (Fig. 3a). The quenching constant (K SV ) was calculated to be 7.57 × 10 8 M −1 (Fig. 3b), which shows two orders of magnitude higher quenching constant value for nitrobenzene than all previously reported porous material to our best knowledge (Fig. 3c, Table S8). Similar quenching by other nitro aromatic compounds (Supplementary Fig. S22-26) were also experiential by JUC-555 with similar K SV constants as summarized in Table S2. The K SV of 1,2-dinitrobenzene, 1,3-dinitrobenzene and 1,4-dinitrobenzene are 7.59 × 10 8 M -1 , 5.81 × 10 8 M -1 and 5.14 × 10 8 M -1 respectively (Fig. 3b). Large Ksv value means a higher sensitivity and stronger interactions between nitro explosives and JUC-555. A blue-shift of 10~15 nm was observed in the quenching process by nitroexplosive. In addition, the detection limit of JUC-555 for nitrobenzene, 1,2-dinitrobenzene, 1,3dinitrobenzene and 1,4-dinitrobenzene is able to reach 0.1673 nM (20.60 ppt), 0.1685 nM (28.32 ppt), 0.1773 nM (29.81 ppt) and 0.1786 nM (30.03 ppt) respectively (Table S3). S4) are higher than the those of previously reported porous material sensors. For example, when JUC-555 was treated with acenaphthylene (5.0 × 10 -5 M), the fluorescence quenching processes were shown in Supplementary Fig. S37. The quenching constant (K SV ) representing the binding affinity was calculated to be 5.35 × 10 5 M −1 (Supplementary Fig. S27, Table S5) and it is more than two orders of magnitude higher than that of NUS-25, which was reported as a chemical sensor for the specific detection of acenaphthylene. Similar quenching phenomenon of other PAHs was also experiential by JUC-555 with similar K SV constants. The K SV for benzene, naphthalene, p-xylene, toluene, o-xylene, m-xylene, fluorene, anthracene, phenanthrene, acenaphthylene, pyrene and triphenylene are 2.56 x 10 6 M -1 , 2.15 x 10 6 M -1 , 1.62 x 10 6 M -1 , 1.52x 10 6 M -1 , 1.41 x 10 6 M -1 , 1.41 x 10 6 M -1 , 8.28 x 10 5 M -1 , 8.10 x 10 5 M -1 , 7.39 x 10 5 M -1 , 7.00 x 10 5 M -1 , 5.04 x 10 5 M -1 , and 3.62 x 10 5 M -1 , respectively (Supplementary Figs. S28-40). A blue-shift of 15 nm was observed in the quenching process by benzene, and blue-shifts of 5 nm were observed in the quenching processes by naphthalene, fluorene, anthracene, acenaphthylene, and pyrene. In addition, the detection limit of JUC-555 for benzene, naphthalene, fluorene, anthracene, phenanthrene, acenaphthylene, pyrene and triphenylene are able to reach 200. 42 S6. The percentages of fluorescence quenching caused by PAHs range from 78.7% for benzene to 39.2% for triphenylene as shown in Supplementary Fig. S41. Sulfur in gasoline is a considerable source of sulfur oxide emissions, which have been one of the major causes of environmental pollution formed during the combustion of sulfur-containing fuels. Since sulfur is present in gasoline in forms of different thiophene derivatives, it is very important to develop a sensitive way to detect such sulfur containing compounds in gasoline. We find JUC-555 is extraordinary sensitive for thiophene derivative recognition. Fluorescence quenching was observed when JUC-555 was exposed to various benzothiophene and dibenzothiophenes (Supplementary Figs. S42-45). The quenching constant (K SV ) for thianaphthene, dibenzothiophene, and 4,6-dimethyldibenzothiophene was calculated to be 8.06 × 10 5 M −1 , 5.74 × 10 5 M -1 , and 5.92 × 10 5 M -1 (Fig. 4). A blue-shift of 5~10 nm was observed in the quenching process. In addition, the detection limit of JUC-555 for thianaphthene, dibenzothiophene, and 4,6dimethyldibenzothiophene are able to reach 94.488 nM (12.680 ppb), 133.499 nM (24.599 ppb), and 146.823 nM (31.171 ppb), respectively, as summarized in Table S7. It should be noted that the correlation between the the Ksv and packing coefficient follows a normal Guassian distribution (Fig. 4b). µ = 0.3253, sigma = 0.0653 for PAHs, and µ = 0.3253, sigma = 0.0653 for S-PAHs were fitted. Such a nice fitting indicates the inclusion is mainly determined by supramolecular recognition mechanism with the exception of benzne, anthracene and triphenylene where the first two were too slim, while the last was two bulky in size. ## Fluorescence quenching mechanism The high sensitivity of JUC-555 toward hazardous molecules prompted us to conduct a detailed study on the quenching mechanism. We selected three examples of the harmful substances, nitrobenzene, fluorene and thianaphthene, to explain the mechanism. First of all, no overlap between the absorption of the analyte and emission of JUC-555 (Supplementary Fig. S46) indicates that the quenching is not a resonance energy transfer (RET) mechanism. The blue shifts of 0~15 nm in fluorescence spectra (Tables S4-7) suggest a mechanism of photoinduced electron transfer (PET) during the fluorescence quenching process. We also conducted timeresolved photoluminescence (TRPL) measurements to provide more evidences regarding the quenching processes caused by hazardous molecules. The TRPL decay curves in Supplementary Figs. S47 and S48 show that the lifetimes (τ 0 ) of JUC-555 in DMF decrease gradually from 3.733 ns to 3.477 ns upon 0 to 50 μL of nitrobenzene titration (5 × 10 -8 M in DMF), indicating the electron transfer from JUC-555 to nitrobenzene. We also performed structural optimization of JUC-555 upon inclusion of nitrobenzene using Materials Studio 7.0 to elucidate the reason for the extreme high Ksv for nitrobenzene. The optimized conformation of nitrobenzene in the channels of JUC-555 was simulated. As shown in Fig. 5, It is clear that nitrobenzene molecules (4.29 × 6.10 ) fit well into the pores of JUC-555 (5.7 × 5.7 ) with a comfortable packing coefficient of 26.6% and perfectly aligned insertions of nitrobenzenes between the layers of JUC-555 (layer to layer distance of 7.2 ) can be observed. There are strong C-H-π interactions between the TPE units and nitrobenzene molecules. We also studied the quenching dynamic of JUC-55 by PAHs with different sizes. As shown in Supplementary Figs. S53 and S54, it's evident that the fluorescence quenching process is size dependent with quenching constants of 0.02239, 0.00418 and 0.00241 for benzene, naphthalene and triphenylene, respectively. As shown in Supplementary Fig. S55, the measurement Ksv for nitrobenzene is still in the range of experimental errors after 5 cycles, indicating the fluorescence sensing of nitrobenzene is also very robust. DFT calculations of the energy levels of JUC-555 and nitrobenzene derivatives (Supplementary Fig. S56) show that the lowest unoccupied molecular orbital (LUMO) of JUC-555 fragment (-2.47 eV, Supplementary Fig. S57) is higher than those of the nitroaromatic compounds tested (-2.50 eV for nitrobenzene, -3.11 eV for o-dinitrobenzene, -3.19 eV for m-dinitrobenzene and -3.55 eV for pdinitrobenzene, respectively), which support a photoinduced electron transfer mechanism of the fluorescence quenching of JUC-555. ## Outlook In conclusion, we have designed and synthesized a novel 3D COF (JUC-555) where TPE-based AIEgens were integrated as side chains in the pores. The synergy between the confinement effect of ordered channels and the AIE effect of dangling TPEs render JUC-555 with strong molecular recognition capability and fluorescence sensing ability. Depending on what types of molecules to be included into the pores, the fluorescence of JUC-555 can be switched between the on-state when rotation restriction enhances AIE effects and the off-state when photoinduced electron transfer induces fluorescence quenching. These phenomena were exploited to identify hazardous chemical molecules such as nitroaromatic explosives, PAHs, and thiophene-based PAHs with much higher sensitivity than those of most previously reported porous materials. This research thus starts a new path to identify dangerous molecules in environment through the combined effects of molecular recognition and fluorescence quenching. Continuing work on chemical sensing of metal ions using JUC-555 is in progress. 2D AIE-based COFs with similar design principle will be pursued in due course. ## Online content Any methods, additional references, Nature Research reporting summaries, source data, extended data, supplementary information, acknowledgements, peer review information; details of author contributions and competing interests; and statements of data and code availability are available at https://doi.org/ (60 µg mL -1 ) at intervals of 10 min. Fluorescence spectra were recorded after the addition of aromatic nitro compounds, PAHs, and S-PAHs solutions. Excitation wavelength of 339 nm was used. Fluorescence quenching was analyzed using the Stern−Völmer equations derived for 1:1 complex to determine the binding mode: The quenching percentage was estimated using the formula (I 0 -I)/I 0 × 100%, where I 0 is the original maximum peak intensity and I is the maximum peak intensity after exposure to aromatic nitro compounds, PAHs, and S-PAHs solutions. Computational methods for HOMO-LUMO energy calculations. To quantitatively evaluate the interactions between JUC-555 and nitro explosives molecules, the electronic properties of a JUC-555 fragment were calculated using DFT. Initially, the structure of JUC-555 was optimized by Forcite using Materials Studio 7.0 to remove geometric distortions. Then, a JUC-555 fragment was used in DFT calculations. The nitro explosives molecules and the JUC-555 fragment were optimized using the M06 functional with 6-31G(d) basis set. The HOMO and LUMO energy levels of the JUC-555 fragment and nitro explosives molecules were calculated using the M06 functional with 6-311G** basis set. All the DFT calculated were perfromed using Gaussian 16. GCMC simulation. These simulations were performed with the Sorption module of Materials Studio 7.0. All GCMC simulations included a 4,000,000-cycle equilibration period followed by a 4,000,000-cycle production

chemsum

{"title": "A three-dimensional tetraphenylethylene-based fluorescence covalent organic framework for molecular recognition", "journal": "ChemRxiv"}

towards_density_functional_approximations_from_coupled_cluster_correlation_energy_densities

## Abstract: Semi-)local density functional approximations (DFAs) are the workhorse electronic structure methods in condensed matter theory and surface science. The correlation energy density c (r) (a spatial function that yields the correlation energy E c upon integration) is central to defining such DFAs. Unlike E c , c (r) is not uniquely defined, however. Indeed, there are infinitely many functions that integrate to the correct E c for a given electron density ρ. The challenge for constructing useful DFAs is thus to find a suitable connection between c (r) and ρ. Herein, we present a new such approach by deriving c (r) directly from the coupledcluster (CC) energy expression. The corresponding energy densities are analyzed for prototypical two-electron systems. To explore their usefulness for designing DFAs, we construct a semilocal functional to approximate the numerical CC correlation energy densities. Importantly, the energy densities are not simply used as reference data, but guide the choice of the functional form, leading to a remarkably simple and accurate correlation functional for the Helium isoelectronic series. ## I. INTRODUCTION There is no doubt that density functional theory (DFT) has had an unrivalled impact on computational chemistry and physics This is because modern realizations of DFT (density functional approximations, DFAs) tend to offer the best compromise between accuracy and computational cost for most applications This is especially true for semilocal DFAs, where E xc only depends on properties of the electron density, such as the local density and its gradient. Such methods are sometimes referred to as "pure" density functionals, as opposed to, e.g., hybrid fuctionals which are based on a generalized Kohn-Sham scheme. 9 Indeed, the early adoption of semilocal DFAs in the quantum chemistry community can be largely attributed to the remarkable accuracy with which, e.g., the semilocal BLYP 10,11 functional describes energy differences in molecules at a much lower cost than post-Hartree Fock methods such as second-order Møller-Plesset perturbation theory (MP2). 12 Even though BLYP and other popular semilocal functionals based on the generalized gradient approximation (GGA) were developed in the 1980-90s, they are still widely used. More recent functionals like those of the ωB97 and Minnesota families (both based on Becke's 1997 power-series approximation) are also commonly applied in chemistry, although mostly in their hybrid variants Similarly, in the solid-state community, the ubiquotous semilocal PBE 16 functional is still the most frequent choice. Here, more recent alternatives, like the constraint-based SCAN 17 functional of Perdew and co-workers and the Bayesian (m)BEEF 18,19 methods are a) Electronic mail: johannes.margraf@ch.tum.de also gaining traction. Of course, there have been highly significant developments beyond semilocal methods. Most prominently, the already mentioned hybrid functionals (e.g. B3LYP or PBE0) complement semilocal DFA exchange with 'exact' Hartree-Fock exchange. 20,21 This makes the functional depend on the occupied Kohn-Sham (KS) orbitals, and not just on the electron density. Particularly in their more recent range-separated variant, these methods are able to extend the applicability of DFT into areas where "pure" DFAs have difficulties, e.g. charge-transfer states or reaction barrier heights. In (gas-phase) molecular chemistry, these methods have become the de facto standard, whereas they are still too computationally demanding for routine application to condensed matter or nanosized systems. The higher computational demand of hybrids is a direct consequence of the fact that the exchange energy now depends on the occupied KS orbitals, and not just on the total electron density. This is even more critical for correlation functionals beyond the semilocal approximation, which depend on the unoccupied (virtual) KS orbitals as well. Such 'higher-rung' functionals are typically based on the random-phase approximation (RPA) or second-order perturbation theory (double-hybrid functionals). This strongly improves their thermochemical accuracy, and allows for the description of van-der-Waals interactions. The virtual orbital dependence of these methods translates to a quite unfavourable formal scaling with the basis-set size (typically O(N 5 ) or worse, compared to O(N 3 ) for GGAs). This is further aggravated by the fact that they additionally require larger (correlation consistent) basis sets, though this deficiency is less critical for the more recent range-separated correlation approaches. 31,32 Such DFAs are consequently not really comparable with 'lower-rung' GGAs, in terms of applicability. Instead, they compete with wavefunction methods such as MP2 or CC. Improving correlation functionals without resorting to virtual orbitals is therefore an exciting prospect and the focus of this work. To this end, we adhere to a purist approach to DFT. In general, the exchange-correlation energy is only dependent on the electron density ρ, and can be determined via numerical integration of a spatial function: Here, xc [ρ](r) is the exchange-correlation energy density. The notation xc [ρ](r) implies that the energy density is both a spatial function (i.e. it has a single scalar value at a given point in space) and a functional of the electron density. In the most general case, the exchangecorrelation energy density on a given point r depends on the electron density at all other points. Semilocal approximations like the GGA use a more convenient formulation, where xc (r) only depends on local quantities like the local electron density ρ(r) or its gradient ∇ρ(r). Furthermore, the exchange and correlation components are usually treated separately, leading to expressions for x [ρ](r) and c [ρ](r). We will focus on the latter. Within this paradigm, there are two classic approaches to designing DFAs. On one hand, there is the constraintbased philosophy championed by Perdew, Burke, Levy and others. Here, exact conditions for the DFA are derived from theoretical considerations of model densities such as the homogeneous electron gas or spherical two-electron densities. 36,37 On the other hand, the property-based approach postulates a parametric form for the exchange-correlation energy density, which is then fitted to accurate reference properties of real molecular or condensed phase systems (often based on higher level calculations). In this contribution, we follow a new route to constructing "pure" DFAs, namely by deriving a correlation energy density from ab initio coupled cluster (CC) wavefunctions. This can be thought of as an intermediate strategy between the constraint and property-based philosophies. On one hand, the DFA is constructed to reproduce high quality benchmark calculations, as in the property-based approach. On the other hand, it is not based on a predefined fit function. Instead, the functional form emerges naturally from the shape of the correlation energy densities of meaningful model systems, as in the constraint-based approach. This paper is organized as follows: In the theory section, we discuss the meaning of the exchange and correlation energies in DFT and WFT and motivate why we expect the CC correlation energy density ( CC c ) to be a useful model for a correlation functional. Then the formalism for computing CC c is presented. In the results section, we analyze the properties of CC c for prototypi-cal two-electron systems. The usefulness of these energy densities is then illustrated by constructing an accurate DFA to the CC correlation energy of the He isoelectronic series. ## II. THEORY We denote occupied molecular orbitals (MOs, φ(r)) with the indices i, j, k . . ., virtual MOs by a, b, c . . . and general MOs by p, q, r . . .. All calculations are performed in a one-electron basis of atom-centered, normalized basis-functions χ µ (r), with indices µ, ν, σ . . .. Following common practice in the CC community, the basisfunctions are referred to as atomic orbitals (AOs). For clarity, it should be noted that the term "exchangecorrelation energy density" is often used in the literature for the correlation energy per particle. The exchangecorrelation energy per volume (as used in this paper) is in that case often referred to as the exchange-correlation kernel. The latter can be converted into the former by dividing through the electron density. ## A. Exchange and Correlation in WFT and DFT The concepts of exchange and correlation are fundamental to both WFT and DFT. In WFT methods, the correlation energy E c is defined with respect to the Hartree-Fock (HF) energy, and simply describes the difference between HF and the exact non-relativistic energy (i.e. the full configuration interaction limit) in a given basis. 42 Meanwhile, the exchange energy E x emerges naturally from the HF formalism, due to the antisymmetry of the wavefunction. 43 In DFT, exchange and correlation in principle describe the same physical phenomena, but the energies are not referenced to HF. Instead, the KS equations use the variational principle to obtain (given the exact functional) the exact density. 2 Accordingly, the exact exchange and correlation energies are referenced to that density, and not to the HF one. One would thus not expect the WFT and DFT E xc to be numerically identical unless the HF density is exact, which is only true in some special cases like the homogeneous electron gas and for one-electron systems like the hydrogen atom. From a DFT perspective, the WFT correlation energy thus contains implicit corrections to the classical and exchange energies, which otherwise carry some error due to the approximate HF density. To understand these differences in detail, it is helpful to consider the individual components to the DFT and CC total energies. In DFT, all energy contributions are written as functionals of the exact ground state density ρ 0 : with the non-interacting kinetic energy functional T s and the contributions of the external (U ) and the Hartree (J) potentials. Equivalently, these terms can be expressed as functionals of the occupied KS orbitals {φ KS i }, which is particularly useful for the kinetic energy. In CC, similar components are computed in terms of the HF orbitals {φ HF p }: Here, K[{φ HF i }] is the HF exchange energy. Given the exact exchange-correlation functional and full CC expansion, both expressions lead to the same energy (E DFT tot = E CC tot ). It is therefore tempting to equate the last term in the DFT expression with the last two terms of the CC formula leading to: However, this is an approximation because i |φ HF i | 2 does not yield the exact ground-state density. Accordingly, for E CC tot to be exact, E CC c must also contain corrections to all other terms in the energy expression: where , and so on. When constructing a correlation functional based on CC reference data, we are essentially hoping for a high accuracy of eq. ( 4). In particular that, and Indeed, these conditions are related, since the exact DFT exchange can be computed analogously to the HF case, but using {φ KS i } instead of {φ HF i }, leading to The difference between the WFT and DFT correlation energies thus boils down to the difference between {φ KS i } and {φ HF i }. While the exact KS orbitals are generally not available (because a general expression for the exact E xc [ρ 0 ] is unknown), it has been observed that Brueckner theory offers an excellent approximation to {φ KS i }. 44,45 Very briefly, the idea behind the Brueckner CC approach is to rotate the HF orbitals in such a way that the T 1 contribution to the correlation energy vanishes. This is equivalent to introducing a (non-local) correlation potential into the HF equations. 46 If the chosen CC expansion is exact (see below), the total energies of the canonical and Brueckner CC methods are identical. However, the individual components on the r.h.s. of eq. ( 3) change. Specifically, the sum of the first four terms (the reference energy) becomes less negative, while the last term (the correlation energy) becomes more negative by the same amount. In the following we apply the CC singles and doubles expansion (CCSD) to two electron systems. We can use canonical and Brueckner CC calculations to numerically estimate the accuracy of eq. ( 4) for this case. For the He atom, the correlation energy difference between a canonical CCSD and Brueckner CCD calculation is 2.6 × 10 −5 E h (see the Supporting Information, table S1). Consequently, the approximation made in eq. ( 4) is very good in this particular case. In a more general vein, it can be noted that HF electron densities are often surprisingly good. Indeed they are often better than self-consistent GGA densities as observed by Bartlett, Burke and others. Note, however, that the above discussion is no longer valid if semilocal exchange functionals are used (in particular for molecular systems). Semilocal correlation functionals cannot describe the type of static (left-right) correlation that is evident, e.g. when dissociating the hydrogen molecule in a spin-restricted calculation. As was observed by Handy and others, this contribution is instead emulated by GGA exchange functionals. 50 The case is different for atomic systems, however. Many classic GGA functionals are based on the approximate equivalence of exchange and correlation in DFT and WFT for atoms. For example, Becke's 1988 exchange functional was fitted to HF exchange energies of atoms, and the Lee-Yang-Parr (LYP) correlation functional is derived from the Colle-Salvetti formula, which expresses the WFT correlation energy of the Helium atom in terms of the corresponding HF density matrix. 10,11 Even functionals which are not based on WFT at all (such as the already mentioned SCAN functional and the "nearly correct asymptotic property" NCAP functional) show reasonably good numerical agreement with the WFT based exchange and correlation energies of noble gas atoms. 17,51 It has also been found empirically that WFT and DFT correlation energies are compatible, as reflected in the success of double hybrid functionals, which describe E c as a linear combination of GGA and MP2 correlation. 25 ## B. Correlation Energy Densities from WFT The connection between WFT and DFT has long been the subject of intensive research. Most prominently, such efforts have been directed at the exchange-correlation potential, V xc . These studies have underscored the limitations of most semi-local approximations to V xc , particu- larly those that are the functional derivatives of common DFAs 59,60 . Such ab initio potentials are also essential components of some of the higher-rung DFAs methods mentioned above. Knowledge of V xc does not provide a route to the corresponding functional E xc , however. The latter requires an expression for the exchange correlation energy density xc (r), as given in eq. ( 1). Unfortunately, an inherent difficulty with defining xc (r) is that it is not unique. In principle, the only condition is that integrating this function over all space yields the exchange-correlation energy. Adding any function that integrates to zero to an ansatz for xc (r) therefore yields equally valid energy densities that may look completely different (see Fig. 1). 64 In this sense, xc (r) is arbitrary. However, not all possible energy densities are mappable to the electron density in an efficient way. A systematic way for defining xc (r) for different systems from ab initio calculation allows exploring this mapping, and therefore represents a promising starting point for designing new DFAs. One strategy to this end is relating xc (r) to the exchange-correlation hole potential. 56,65 This offers a systematic route to calculating xc (r), given that the one-and two-particle density matrices are known. This has, e.g., been done for configuration interaction wavefunctions with singe and double excitations (CISD). 56 More recently, Vyboishchikov used modified "local" twoelectron integrals to calculate the correlation energy density c (r) at the MP2 and CISD level. 66 These functions were used to construct a simple local correlation functional for spherically confined atoms. ## C. CC Correlation Energy Densities In the following we introduce a new method to calculate an c (r) from first principles, namely one that integrates to the CC correlation energy. The approach has several advantages: (1) By virtue of being CC-based, it is automatically size-extensive (unlike truncated CI). ( 2) Only integrals and amplitudes that are available in any standard CC code are required. (3) The c (r) obtained in this manner is by construction topologically similar to the electron density, making it amenable to semilocal approximations (see below). In CC, the ground-state wavefunction Ψ CC is defined with respect to a reference determinant ψ 0 as: 67 By truncating T at double (N=2), triple (N=3), or quadruple (N=4) excitations one obtains specific CC methods, abbreviated as CCSD, CCSDT, and CCSDTQ respectively. An important feature of these methods is that they are exact for systems with a number of electrons smaller or equal to the highest excitation level (i.e. CCSD is exact for two-electron systems). Irrespective of the truncation, the CC correlation energy only depends on the single and double amplitudes (t a i and t ab ij ), while higher than double excitations contribute to the energy indirectly, by coupling with T 1 and T 2 . The correlation energy is calculated as: ) with τ ab ij = t ab ij + 1 2 t a i t b j , and the antisymmetrized twoelectron integrals in MO basis defined as These integrals are obtained from the corresponding AO integrals and the MO coefficients which define ψ 0 , formally via: We are now looking to transform the coupled cluster correlation energy into a form resembling the DFT expression: We start from the AO-CC approach of Ayala and Scuseria, which is based on an MO to AO transformation of the T-amplitudes: 70 Given these AO amplitudes, the correlation energy can be calculated as: We now partition the energy into atomic or AO contributions, using: Because the AO basis-functions are normalized, the CC correlation energy can now be written as an integral over space: This defines the CC correlation energy density as: As noted above, CC c (r) is topologically similar to the electron density, in the sense that it is a linear combination of atomic densities. As shown in Fig. 1, the shape of the correlation energy density is in principle arbitrary. However, an energy density that is similar to the electron density can be much more easily approximated by a (semi-)local approach. Using eqs. ( 15), ( 16), ( 18) and ( 20), CC c (r) can be calculated for any system, as long as a standard calculation is possible. In the following some exemplary calculations for atomic two-electron systems are performed at the CCSD level, using a custom Python program interfaced with the Psi4 program package. 71,72 Calculations for two-electron ions were performed with a modified uncontracted cc-pV5Z basis set for Helium, where the scaling factor of the orbital exponents was optimized individually for each ion (abbreviated u-5Z). 73 In all other DFT calculations, the pcseg-3 basis set of Jensen is used. 74 Additional CCSD calculations on He-Zn were performed with the core-polarized cc-pwCV(5+d)Z basis. 75 DFT correlation energies are calculated by numerical quadrature on Lebedev-Treutler (75,302) grids. 76 All DFT calculations (also for PBE) are performed nonself-consistently using HF densities with the same code. ## III. RESULTS As model systems, we calculate CC c (r) for the twoelectron ions from H − to Ne 8+ (see Fig. 2). In all cases, the correlation energy density decays in an approximately exponential fashion as a function of the distance from the nucleus, with the individual curves being highly system dependent. Specifically, CC c (r) decays slowly for the very diffuse H − ion and quickly for Ne 8+ . It is furthermore notable that the correlation energy density for He is quite similar to the one obtained by Vyboishchikov's 'local 2e-integral' approach, despite the different mathematical ansatz. 66 In the supporting information we also include the respective plots for the PBE and LDA correlation functionals (Figs. S1 and S2). While both energy densities are qualitatively similar to Fig. 2, there are important differences. In the LDA case, the functions decay at approximately the same rate as CC c (r), but they are less curved and display larger values at the nuclear cusp. In contrast, the PBE curves overall decay more quickly and display a more complex shape, with a fast initial decay close to the nucleus followed by a slower asymptotic decay. From a DFT perspective, the more interesting dependence is between CC c (r) and ρ (Fig. 3). As the atomic electron densities are monotonically decaying, there is a unique mapping between the two for each ion. Specifically, | CC c (r)| increases approximately parabolically with ρ. Unsurprisingly, the curves are again somewhat system dependent, however. This simply means that a LDA-like correlation functional cannot represent CC c (r) exactly for all systems. If it is to be useful for defining DFAs, it should at least be approximately possible to effectively map CC c (r) to ρ, however. Furthermore, this mapping should ideally only use readily available local features of the electron density, such as ρ(r) or the reduced density gradient s = |∇ρ(r)| 2(3π 2 ) 1/3 ρ(r) 4/3 . To explore whether this is possible in the presented formalism, we construct a simple GGA functional to approximate CC c (r). To this end, only datapoints with s < 5 were taken into account, following the observation of Burke, Perdew and coworkers that the energetically relevant range is 0 < s < 3. 77 As can be seen in Fig. 4, a simple linear fit allows an accurate description of all datapoints with s < 0.1 (i.e. those with approximately "homogeneous electron gas"-like conditions). This is reminiscent of the Wigner functional, 78,79 which is linear in ρ to leading order, but allows some more flexibility in the low density regime: where c 1 and c 2 are coefficients to be defined. Eq. ( 21) forms the local baseline functional for our GGA (with c 1 = −0.0468 and c 2 = 0.023). As shown in Fig. 5, the residual error of W c [ρ(r)] is strongly dependent on the reduced gradient s. The largest errors are found in the regime between 0 < s < 2. For the full GGA functional, we now choose the enhancement-factor ansatz: Plotting CC c / W c vs. s, gives insight into the numerical distribution of an ideal enhancement factor (Fig. 6). Interestingly, all ions from He to Ne 8+ approximately fall on a curve, whereas the H − datapoints deviate significantly. This reflects the well-known inability of GGAs to adequately describe atomic anions. 80 Specifically, semilocal DFAs only attach a fractional electron to an atom in a complete basis-set due to the self-interaction error. 81,82 This is an inherent limitation of the GGA functional form, not of the CC reference calculations. 83 We therefore exclude H − when fitting parameters, though it is retained in the analysis, for comparison. The distribution of the numerical enhancement factor in Fig. 6 suggests that F (s) should have a sigmoidal form with the asymptotic behaviour: We therefore base F (s) on the "complementary" logistic function: with coefficients c 3−5 . Combining equations ( 21), ( 22) and ( 25), the final functional, which we call ccDF, thus has the simple 5parameter form: One could optimize these parameters to directly reproduce the numerical F (s) as closely as possible. However, this strategy is not optimal, as F (s) only enters the energy expression as a scaling factor for W c [ρ(r)]. Consequently, it has little effect on the total energy, whenever W c [ρ(r)] is small. A more promising approach is therefore to use total correlation energies (E c ) as reference data. A least-squares fit of the GGA parameters to the correlation energies of He to Ne 8+ yields: The resulting enhancement factor is a good fit to the numerical F (s) (solid line in Fig. 6), and the ccDF functional accurately reproduces the CCSD correlation energies of He to Ne 8+ (Fig. 7). This figure also includes the PBE correlation energies. Unsurprisingly, ccDF more closely reproduces the CCSD correlation energies than PBE, given that it was fitted to this data. It is, however, notable that this functional achieves very high total accuracies of 10 −3 E h or better (except for H − , see above), given its simple functional form. More importantly, both functionals display the correct qualitative behaviour: As Z increases, the correlation energy converges to a constant value. As discussed in the Theory section, exact numerical agreement between DFT and WFT correlation energies should generally not be expected. Neither is it necessary for chemical applications. For example, both MP2 and PBE correlation energies will often deviate from more accurate CC values by 10% or more, yet both methods are still quite accurate in terms of energy differences. In fact, even the CCSD/u-5Z values we used for fitting ccDF are only converged to within several milli-Hartree, since the complete basis-set limit for absolute correlation energies of isolated atoms is notoriously difficult to reach. 84 Still, a useful DFA should reproduce the qualitative behaviour of accurate WFT reference values. Having established the accuracy of ccDF for twoelectron systems, the question arises whether this functional form can also be applied in the many-electron case. To this end, we computed the correlation energies for the closed-shell neutral atoms from He to Kr (table 1), for which highly accurate reference energies are available. 84,85 Here, ccDF and PBE show qualitatively different behaviour. For He and Be, both functionals recover >90% of the correlation energy. For all other systems, PBE continues to recover 85-100% of the correlation energy while the ccDF values range from 60-70%. This behaviour can readily be explained by considering the spin-polarized form of the Wigner functional, upon which ccDF is based: Here, ρ α and ρ β are the up and down-spin densities, respectively. By construction, this functional only describes correlation between electrons of opposite spin (i.e., the correlation energy for fully spin-polarized systems is zero). Though this is not widely appreciated, the LYP functional actually suffers from the same problem, since the first term in its expansion is exactly eq. ( 27). Obviously, closed shell two-electron systems like He only display opposite spin correlation. Similarly, Be possesses filled 1s and 2s orbitals, so that there is only weak core-valence correlation between same-spin electrons, and the bulk of the correlation energy is of opposite-spin nature. ccDF describes these systems quite accurately. For all other systems, ccDF underestimates the total correlation energy by about one third, presumably due to the missing same-spin contribution. Importantly, this is in good agreement with the relative contribution of same-spin correlation for general many-electron systems, as estimated by Grimme and Head-Gordon in the construction of the spin-component-scaled (SCS) and scaled-opposite-spin (SOS) MP2 methods. 91,92 For instance, SOS-MP2 simply scales the opposite-spin correlation energy by 1.3 to approximate the full correlation energy. To further corroborate this interpretation, we turn to the spin component decomposition of the CC energy, which allows computing the opposite-spin contribution to the CC correlation energy as: 93 As shown in Fig. 8, E ccDF c indeed correlates with E CCSD c,OS quite well. This indicates that the physics of opposite-spin correlation are essentially transferable between two-and many-electron system. However, neither this transferability nor the GGA approximation should be expected to be perfect. Future work will therefore focus on developing general correlation functionals based on CC. We consider the presented results to be very encouraging for this endeavour. ## IV. CONCLUSIONS In this paper, we have introduced a new approach to calculating CC correlation energy densities CC c (r) for atomic systems. These densities are derived from an AObased formulation of CC and exactly integrate to the respective CC correlation energy. The properties of CC c (r) were discussed for a range of atomic two-electron systems. As these energy densities are by construction topologically similar to the electron density, they are well suited to be approximated by DFAs. As a proof-of-principle, a CCSD based GGA functional was constructed for the He isoelectronic series. By analysis of the numerical CCSD functional, we find that a remarkably simple enhancement-factor formula can be fitted to yield highly accurate correlation energies for these systems. Despite only being fitted to two-electron systems, we find that the ccDF functional also provides reasonable estimates of the opposite-spin correlation energies of many-electron atoms. This indicates that CC c (r) provides a robust physical basis for the design of DFAs, and that the He isoelectronics form an interesting set of model densities. However, it should be emphasised that the proposed GGA functional is mainly intended as a proof-ofprinciple, and should not be applied to general systems as is. Most importantly, it should at least be augmented with a corresponding same-spin functional. 88 Furthermore, the proposed form of CC c (r) is only one possible choice. An expression based on the one-and two-particle density matrices may in fact be preferable, as it would allow using the "gold-standard" CCSD(T) method as reference, which includes perturbative triple contributions. In contrast, our current approach can only be used with full coupled cluster methods (CCSD,CCSDT,CCSDTQ, etc.), of which all but CCSD display prohibitive computational scaling for all but the simplest systems. Moving beyond CCSD is a prerequisite to obtain a good description of electron correlation from systems with more than two electrons. Importantly, the present framework is general enough to be applied to more complex functional forms (e.g. truly non-local functionals), and this will be the subject of future work. An especially promising route lies in the use of CC energy densities to train "machine-learned" functionals. 94 The fact that CC c (r) can guide the design of a simple and accurate functional form like the GGA indicates that it contains the necessary information to this end.

chemsum

{"title": "Towards Density Functional Approximations from Coupled Cluster Correlation Energy Densities", "journal": "ChemRxiv"}

hydrogen_bond_networks_near_supported_lipid_bilayers_from_vibrational_sum_frequency_generation_exper

## Abstract: We report vibrational sum frequency generation (SFG) spectra in which the C-H stretches of lipid alkyl tails in fully hydrogenated single-and dual-component supported lipid bilayers are detected along with the O-H stretching continuum above the bilayer. As the salt concentration is increased from ~10 µM to 0.1 M, the SFG intensities in the O-H stretching region decrease by a factor of 2, consistent with significant absorptive-dispersive mixing between χ (2) and χ (3) contributions to the SFG signal generation process from charged interfaces.A method for estimating the surface potential from the second-order spectral lineshapes (in the OH stretching region) is presented and discussed in the context of choosing truly zero-potential reference states. Aided by atomistic simulations, we find that the strength and orientation distribution of the hydrogen bonds over the purely zwitterionic bilayers are largely invariant between sub-micromolar and hundreds of millimolar concentrations. However, specific interactions between water molecules and lipid headgroups are observed upon replacing phosphocholine (PC) lipids with negatively charged phosphoglycerol (PG) lipids, which coincides with SFG signal intensity reductions in the 3100 cm -1 to 3200 cm -1 frequency region. The atomistic simulations show that this outcome is consistent with a small, albeit statistically significant, decrease in the number of water molecules adjacent to both the lipid phosphate and choline moieties per unit area, supporting the SFG observations. Ultimately, the ability to probe hydrogen-bond networks over lipid bilayers holds the promise of opening paths for understanding, controlling, and predicting specific and non-specific interactions between membranes and ions, small molecules, peptides, polycations, proteins, and coated and uncoated nanomaterials. ## I. Introduction. The structure of water over lipid membranes is of interest for a variety of reasons that are rooted in fundamental scientific interest and connect all the way to biological function and technological applications. Specific questions pertain to whether there exist populations of interfacial water molecules that can undergo hydrogen-bond (H-bond) interactions with certain membrane constituents that can be strengthened or weakened with variations in ionic strength, or, as indicated by molecular dynamics simulations, 2 whether some population of water molecules exists that may interact specifically with certain lipid headgroups over others. While interface-specific vibrational spectroscopic approaches, particularly those that are based on sum frequency generation (SFG), are in principle well suited for probing water near membranes, this method has been largely limited to probing lipid monolayers 1, chemically asymmetric bilayers, or the use of D 2 O as opposed to H 2 O. Indeed, the use of SFG spectroscopy for probing fully hydrogenated lipid bilayers is now just emerging. Part of the reason for this relatively new application of vibrational SFG spectroscopy to probe chemically unmodified lipid bilayers is rooted in the symmetry-breaking requirement of the method, 26 which has limited its use largely to asymmetric bilayers consisting of a deuterated and a hydrogenated leaflet, or lipid monolayers, as stated above. SFG signals generated by asymmetric membranes (deuterated leaflet on one side and hydrogenated leaflet on the other side, or aliphatic lipid tail on one side and polar headgroup on the other) are strong enough to be detectable using lowrepetition rate, low peak power laser systems most commonly used in the field. Two studies known to us also report SFG spectra of unlabeled symmetric lipid bilayers, demonstrating their low signal yields when compared to labeled bilayers. Our recent work has shown, in the C-H stretching region, that commercially available broadband optical parametric amplifier laser systems running at modest (kHz) repetition rates can overcome these limitations, with reasonably high signal-to-noise ratios obtained in just a few minutes of spectral acquisition time. Here, we report how to apply this approach to probe the C-H stretches of the alkyl tails in fully hydrogenated single-and dual-component supported lipid bilayers (SLBs) along with the O-H stretching continuum of the H-bond network system in the electrical double layer above them. The approach probes lipid tail order and disorder while also informing on changes in the H-bond network strength that result from changes in the bulk ionic strength up 100 mM NaCl. Moreover, by varying the lipid bilayer composition from 100% zwitterionic lipid to an 8:2 mixture of zwitterionic and negatively charged lipids, we identify specific H-bond interactions between water molecules and the lipid headgroup choline moieties that manifest themselves in spectral intensity changes in the 3100 cm -1 to 3200 cm -1 range. II. Methods. A. Bilayer Preparation. 1,2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC) and 1,2dimyristoyl-sn-glycero-3-phospho-(1-rac-glycerol) (DMPG) were purchased from Avanti Polar Lipids and used without further purification. Lipid bilayers from small unilamellar vesicles of pure DMPC, lipid mixtures containing 90 mol% DMPC and 10 mol% DMPG, and 80 mol% DMPC and 20 mol% DMPG were prepared by the vesicle fusion method, as described earlier, 29, 31-34 on 3 mm thick calcium fluoride windows (ISP Optics, CF-W-25-3). Prior to use, the calcium fluoride window was sonicated in HPLC-grade methanol (Fisher Scientific) for 30 min, rinsed with ultrapure water (18.2 Ω•cm resistivity; Millipore), and dried with N 2. The window was then plasma cleaned (Harrick Plasma Cleaner, 18W) for 10 min. Experiments were carried out at room temperature (21 ± 2 o C). All SLBs were formed at 0.01 M Tris buffer and 0.1 M NaCl in the presence of 0.005 M CaCl 2 •2H 2 O at pH 7.40 ± 0.03. 33 Following bilayer formation, SLBs were rinsed with Ca-free buffer to remove excess vesicles. The spectra were recorded at two different ionic strengths. Before the preparation of aqueous solutions, Millipore water was left overnight to equilibrate with atmospheric CO 2 . The solution pH was measured for each salt concentration and the pH was adjusted to 7.4 with minimal NaOH and HCl before the solutions were flowed across the interface resulting in ionic strengths of ~10 µM and 0.1 M for the Millipore solution and NaCl solution, respectively. B. Vibrational Sum Frequency Generation Spectroscopy. Details of our SFG approach and experimental setup for probing condensed matter interfaces in the C-H stretching region have been reported previously. Here, we adapt this approach to extend our spectral range into the O-H stretching region, as described in detail in the Supporting Information (see part numbers of the optical elements in Supplementary Figure S1). Briefly, 90% of the output from a Ti:Sapphire amplifier laser system (Spectra Physics Solstice, 3 mJ/pulse, 795 nm pulses, 1 kHz repetition rate, 120 femtosecond pulse duration) pumps a travelling-wave optical parametric amplifier to generate a broadband tunable IR beam tuned to the C-H and O-H regions (2800-3600 cm -1 ), while the remaining portion is sent down the visible upconverter beam line, where it is attenuated using a variable density filter and spectrally narrowed using an etalon. The IR and visible beams are focused to a ~30 µm beam waist at the interface, where they overlap at the CaF 2 /water interface at 38° and 30° from the surface normal, respectively. The beams approach the interface from the CaF 2 side and the SFG signal is detected in reflection. The resultant SFG signal is dispersed on to a spectrograph (Acton SP-2558) and liquid nitrogen-cooled CCD camera. During the SFG experiments, the IR line was purged with dry house N 2 to avoid water absorption bands that appear in this stretching region. All SFG spectra were collected using the near total internal reflection geometry and the ssp polarization combination (s-polarized SFG, s-polarized 800 nm light, p-polarized IR light). All SFG spectra were recorded in triplicates and normalized to the ppp-polarized SFG response obtained from a gold window. To cover the full spectral range of interest, multiple spectra are collected at different IR center wavelengths before being combined into a single spectrum. Further details regarding spectral acquisition and analysis procedures are provided in the Supporting Information (See Figure S2). C. FRAP Measurements. Two-dimensional diffusion coefficients, which can serve as a metric for bilayer quality, were estimated using fluorescence recovery after photobleaching (FRAP). FRAP measurements and sample preparation were carried out in a manner consistent with our previous approach. 29 For these experiments, vesicles composed of DMPC or a 9:1 mixture of DMPC/DMPG lipids were doped with 0.1 mol% TopFluor PC (Avanti Polar Lipids, 810281). After forming the SLB as described in Section IIA, the cell was flushed with 20 mL of 0.1 M NaCl, 0.01 M Tris buffer (pH 7.4). In a second set of experiments, the flow cell was flushed with 20 mL of pH-adjusted Millipore water with no added salt. For SLBs formed from a 9:1 mixture of DMPC/DMPG lipids, we find diffusion coefficients on the order of 0.5 ± 0.2 µm 2 /s (13 replicates over two samples) after rinsing with 0.1 M NaCl, 0.01 M Tris buffer, which is consistent with our previously reported two-dimensional diffusion coefficients 29 and indicates that a well-formed bilayer is produced from the abovementioned method. Upon rinsing with pH-adjusted Millipore water with no added salt, we find that the diffusion coefficient for SLBs formed from 9:1 mixtures of DMPC/DMPG lipids are on the order of 0.03 ± 0.01 µm 2 /s ( 6replicates over 1 sample). The diffusion coefficient for SLBs formed from pure DMPC lipids on calcium fluoride is 0.4 ± 0.2 µm 2 /s (6 replicates over two samples) after rinsing with 0.1 M NaCl, 0.01 M Tris buffer. After rinsing with pH adjusted Millipore water, we find a diffusion coefficient of 0.07 ± 0.02 µm 2 /s (4 replicates over two samples). For SLBs formed from 8:2 mixtures of DMPC/DMPG lipids, the diffusion coefficient on calcium fluoride is 0.07 ± 0.04 µm 2 /s (6 replicates over one sample) after rinsing with 0.1 M NaCl, 0.01 M Tris buffer. Representative traces, along with a detailed procedure used in these experiments, are provided in the Supporting Information (see Figure S3). These results indicate the bilayers transition between the gel and fluid phases, irrespective of the nature of the underlying substrates (CaF 2 vs fused silica). D. Computational Methods. Molecular dynamics (MD) simulations for investigating the structure of the H-bond network near the lipid-water interface were performed using the CHARMM-GUI 42 input generator to set up the DMPC and 9:1 DMPC/DMPG systems. Each system contains a 10 x 10 nm 2 lipid bilayer. For pure DMPC, systems were set up with 0.15 M NaCl or no salt added, both with a hydration level (i.e., water:lipid ratio) of 53. The 9:1 DMPC/DMPG system was set up with 0.15 M NaCl and a hydration level of 65. We performed equilibration and production runs with CHARMM-GUI generated input files using the NAMD 43 package. The CHARMM36 force field was applied for the lipid, water, and ions. The Particle-Mesh-Ewald 46 (PME) method was used for the electrostatic interactions with a realspace cutoff of 1.2 nm. Force switching with a cutoff of 1.2 nm was applied to the van der Waals interactions. The PME grid size was set to 108, 108, and 100 for the X, Y, and Z dimensions in the DMPC simulations, and to 108, 108, and 120 for the 9:1 DMPC/DMPG simulations. RATTLE 47 was applied to constrain all bonds involving hydrogen atoms in length. Langevin dynamics were applied for constant pressure and temperature control. A Nose-Hoover Langevin piston was applied with constant ratio on the X-Y plane and a target pressure of 1 atm. The target temperature was set to be 303.15 K with a damping coefficient of 1.0 ps -1 . For the DMPC systems, the production run lasted for 30 ns with a 2 fs time step; for the 9:1 DMPC/DMPG systems, the production was run for 70 ns. Any unspecified details, including the equilibration process before production runs (see Table S2 for details), are consistent with the standard CHARMM-GUI protocol which have been shown to provide area per lipid and other essential properties of lipid bilayers in good agreement with experiments. 42, ## III. Results and Discussion. A. Single-Component Zwitterionic Supported Lipid Bilayers. Figure 1A shows the ssppolarized SFG response from the pure DMPC bilayer without added salt. At this low ionic strength (~10 µM), we find clear spectral signatures from the C-H oscillators of the alkyl tails, 29, 31, 34 as well as broad contributions from the O-H stretches of the water molecules. The non-zero signals are due to the fact that the molecular environment above and below the bilayer is not fully symmetric, as would be expected for a suspended bilayer. Instead, symmetry breaking occurs due to the presence of the aqueous phase on one side and the solid support on the other. The frequencies corresponding to the signal peaks in the C-H stretching region shown in Figure 1A are comparable to the ones we observe for supported lipid bilayers formed on fused silica substrates (see Supporting Information Figure S4) 29,31,34 The two broad features in the O-H stretching continuum located at ~3200 cm -1 and ~3400 cm -1 are associated with bandwidths (full width at half maximum) of about 200 cm -1 . The peak positions are within 50 cm -1 of what has been reported for water spectra obtained from symmetric bilayers prepared from negatively charged lipids on CaF 2 . 28 The difference is attributed to the fact that our current experiments use bilayers formed from purely zwitterionic lipids. Replacing the H 2 O phase with D 2 O while maintaining low ionic strength, shown in Supporting Information Figure S5, leads to the C-H oscillators retaining their frequencies while the O-H stretching continuum is entirely absent. This experiment indicates that 1) there are no exogenous photon sources contributing to the SFG response from the bilayer under water (H 2 O), and 2) that H 2 O that may be possibly trapped between the bilayer and the substrate is readily exchanged or associated with too little SFG intensity to be detectable by our method. Control experiments assessing the possible role that CaF 2 dissolution could have on the spectra (see Supporting Information Figure S6) show that the presence of the bilayer eliminates any flowdependent changes in the SFG signal intensity produced by the interfacial water molecules. The O-H stretching continuum can be viewed as a display of the various O-H … O distances sampled in the water network probed by the SFG spectrometer. As shown, for instance, by Lawrence and Skinner, 54 frequencies around 3200 cm -1 correspond to O-H stretches associated with water molecules in tighter H-bond networks, where distances between the donor hydrogen and acceptor oxygen atoms (H … O) are as short as 1.6 or less. Towards 3400 cm -1 , the spectrum samples water molecules in a considerably looser H-bond network, having H … Odistances as long as 2.1 or so. Towards 3550 cm -1 , H … O distances can be as long as 2.4 or more. At the very end of the spectrum, near 3700 cm -1 , would be the O-H stretch of non-Hbonded water molecules, those that "straddle the interface". 55 Such signals are not identified within our signal-to-noise ratio, even though they have been reported to be present in Langmuir monolayers prepared from DPPC lipids. 56 Figure 1B shows the SFG spectrum from the supported lipid bilayer in comparison with that of two other aqueous CaF 2 interfaces, namely that of bare CaF 2 in contact with ~10 µM ionic strength water adjusted to pH 7.4, as well as bare CaF 2 in contact with water vapor in He flow adjusted to 80% relative humidity (see Supporting Information Figure S7). The SFG response from the bare CaF 2 /water interface is in reasonable agreement with published data. 53, We find that the peak positions from the bilayer/water interface is blue-shifted by around 25 cm -1 when compared to those obtained from the bare CaF 2 /water interface. Additionally, the SFG spectrum from the CaF 2 /water vapor interface exhibits a blue-shifted SFG spectrum when compared to the bilayer/water or CaF 2 /water interfaces, consistent with the expectation that its hydrogen-bonding environment is looser than in the case of bulk water in contact with the solids. Upon increasing the ionic strength in the bulk aqueous phase, the sodium and chloride ions can modify the H-bond network of water molecules in the bulk in ways that are the subject of much past and ongoing scientific attention and discussion. NaCl, whose anion and cation fall right in the middle of the familiar Hofmeister series, are not necessarily expected to modify the H-bond network over lipid bilayers at the relatively modest concentrations (0.1 M) employed here. Moreover, ion-specific interactions with the lipids used in our work are unlikely to be strong under the conditions of our experiments. Indeed, Figure 2 shows that the spectral changes we observe in response to changes in the ionic strength are largely uniform over the entire frequency region probed in our experiment (1000 cm -1 ). Between 3000 cm -1 and 3600 cm -1 , the ratio of the SFG spectral intensities at low (~10 µM) and high (0.1 M) ionic strength is computed to vary only slightly, from 1.7 at 3000 cm -1 to 2.3 at 3600 cm -1 and back to 2.0 at 3700 cm -1 (average of 2.1 ± 0.2 over all frequencies). We find this slight frequency dependence of the SFG intensity ratio to be indicative of a minor influence that the relatively modest salt concentrations used here even under what we term "high salt" have on the various contributors to the H-bond network. This interpretation is borne out in molecular dynamics simulations as well, which are described next. To further explore the molecular details near the bilayer/water interface, we performed MD simulations for a DMPC lipid bilayer with and without 0.15 M NaCl salt. We focus here on the analysis of the interfacial water structure, specifically the orientation of interfacial water molecules and the O•••O distance of neighboring water molecules. The water orientation is characterized with the dipole angle, θ, which is defined as the angle between the dipole vector of water and the membrane normal pointing towards the bulk. The distribution of water orientations is analyzed as a function of distance from the membrane-water interface, i.e., we plot the two-dimensional distribution 63 (Figure 3): in which 𝜃 is the dipole angle defined above, z is the normal distance of the water oxygen from the bilayer center, 𝜌(𝑧) is the number density of water, and 𝑠𝑖𝑛𝜃 is the angular Jacobian factor. The distribution shown in Figure 3 is normalized to that of the bulk value. According to the mass density distribution (see Figure S8), the lipid-water interface is identified at z ~ 20 . As shown in Figure 3 (left column), in all cases studied, the water orientation distribution shifts towards smaller dipole angles near the interface, while the opposite shift is observed for the small amount of water molecules that penetrate below the lipid/water interface to interact with the lipid glycerol groups (for a snapshot, see Figure S9). The distribution approaches the bulk value at ~8-10 away from the lipid-water interface. Nevertheless, the distribution of water orientation angles remains broad even at the interface, which is likely due to the dynamic nature of the lipid headgroup (see Figure S10). As a result, no statistically significant difference is observed between the two DMPC cases studied, suggesting that the impact of salt on the water orientation at the interface is subtle compared to the effect of thermal fluctuations. Regarding the distributions of the nearest O•••O distances among water molecules, which reports on the hydrogen bonding strength, our results in Figure 3 (right column) suggest again that the impact of salt on the distance dependent orientation distributions of the water molecules is small for the salt concentrations investigated. Rather than being due to changes in the H-bonding network, we find that the SFG signal intensity reductions that coincide with raising the salt concentration from 10 µM to 0.1 M are consistent with absorptive-dispersive mixing between χ (2) and χ (3) contributions to the SFG signal generation process from charged interfaces, according to Here, the first two terms are the non-resonant and resonant 2 nd -order susceptibility and the 3 rd term is given by the inverse Debye screening length, 𝜅, the inverse of the coherence length of the SFG process, ∆𝑘 ! , and the interfacial potential, Φ(0), multiplied by the 3 nd -order susceptibility. 71 We recently showed 66-68 that for an exponential distance dependence of 𝛷(z), the χ (3) phase angle, ϕ, equals 𝑎𝑟𝑐𝑡𝑎𝑛 ∆𝑘 ! 𝜅 . Using "primitive ion" models, 72 such as Gouy-Chapman theory, we estimate at the low (resp. high) salt concentration investigated here that 𝜅 is 1 x 10 7 (resp. 1 x 10 9 ) m -1 . For our experimental geometry, ∆𝑘 ! of 2.4 x 10 7 m -1 and invariant with salt concentration. The resulting phase angle is shown in Figure 4A. At high salt concentration, eqn. 2 becomes simply additive, 73-74 75 whereas constructive and destructive interference occurs when the phase angle deviates from zero. In the absence of phase-resolved measurements, which are proving to be considerably challenging at buried liquid-solid interfaces such as the ones studied here, it is difficult to quantitatively examine the interfacial potential, even if one uses the 3 rd order (𝜒 !"#$ ! ) term recently reported by Wen et al. 64 that should be quite universally applicable for aqueous interfaces. Moreover, it is perhaps not possible to prepare, in an experiment, a truly "zero potential" reference state: even the fully protonated reference state of a carboxylic acid monolayer, commonly used as a reference state in surface potential measurements, 64, is subject to dipolar potentials. In the absence of 1) phase resolved data and 2) a true zero potential -and/or zero charge density -reference state, quantitative knowledge of the interfacial potential at two different solution or bilayer conditions from which a difference in surface potential, i.e. ΔΦ, can be calculated is difficult to obtain, though methods to acquire this knowledge remain a topic of keen interest to us that we will discuss in forthcoming work. For now, we offer the following method for estimating surface potential changes from the second-order spectral lineshapes (in the OH stretching region): an examination of Equation 2reveals that even if, as suggested by the MD simulations discussed above, the H-bond network close to the interface remains invariant or nearly invariant (implying a constant 𝜒 !"#$ ! ) upon changes in ionic strength, changes in the SFG signal intensity can still arise from the potential-dependent χ (3) term. These changes take the form of a complex multiple of the 𝜒 !"#$ ! term, which is given mainly by the 3 rd order optical properties of bulk water. Unfortunately, given the difficulties discussed above, our lack of phase-resolved measurements and our lack of access to a reference state of true Φ(0)=0, precludes us from comprehensively accounting for the phaseangle dependent χ (2) /χ (3) mixing, and thus quantitatively determining the interfacial potential from the SFG spectra reported here. Yet, surprisingly good qualitative agreement is obtained between the difference of the measured intensity spectra for the low and high salt conditions from Figure 2 and the calculated 𝜒 !"#$ ! intensity spectrum derived from the real and imaginary data reported by Wen et al. (see Figure 4). This agreement supports our conclusion that the spectral changes are not indicative of large changes in the H-bonded network of water molecules but rather result from the χ (3) -potential dependent term. Moreover, as shown Supplementary Information equations S1-S5, under conditions where the SFG responses are dominated by the χ (3) term, i.e. χ (3) Φ >> χ (2) , an estimate of the difference in surface potential, ΔΦ, can be readily provided if the magnitude of the SFG intensity difference, ΔI SFG , observed for conditions of varying ionic strength, bulk solution pH, analyte concentration, or surface composition, is known (see Supporting Information Eqn S5). ## B. Dual-Component Supported Lipid Bilayers Formed from Zwitterionic and Negatively Charged Lipids. Motivated by recent reports that the major contribution in the 3000 cm -1 to 3200 cm -1 frequency region originates from polarized water molecules that bridge phosphate and choline in the zwitterionic lipid headgroup (n. b.: that work focused on lipid monolayer/water interfaces as opposed to lipid bilayer/water interfaces, which are probed in the present study), 2 we proceeded to add negatively charged lipids to the zwitterionic system studied. Mixing in negatively charged lipids, such as DMPG, is then expected to reduce the population of polarized water molecules that interact specifically with the zwitterionic PC headgroup. Figure 5 shows that this response is indeed observed. At 0.1 M NaCl, the three systems we surveyed (100% zwitterionic DMPC, 9:1 DMPC/DMPG, and 8:2 DMPC/DMPG) showed no significant changes in the 3400 cm -1 frequency region. Yet, as the percentage of negatively charged lipids increases, the SFG spectral intensity in the 3200 cm -1 region decreases, indicating the theoretical result obtained for lipid monolayer/water interfaces may also hold for lipid bilayer/water interfaces. Triplicate measurements are shown in the Supporting Information (see Figure S11). Results from our MD simulation for 9:1 DMPC/DMPG with 0.15 M NaCl (Figure 3, bottom row) reveal similar trends when compared to the pure DMPC case, suggesting that the impact of a small amount (10%) of anionic lipids on the structure and orientation of water at the interface is minor, in the background of thermal fluctuations. Yet, computing the number of water molecules adjacent to lipid phosphate, choline, and those close to both phosphate and choline (see Figure S12 for the relevant radial distribution functions), in a manner consistent with the analysis by Morita and coworkers, 2 we find that mixing in DMPG lipids leads to a small, albeit statistically significant, decrease in the number of water molecule adjacent to both the lipid phosphate and choline moieties per area, as shown in Table 1. These computational results support the observations that the SFG signal intensities seen in the experimental spectra between 3100 cm -1 to 3200 cm -1 are due to local water molecules that specifically interact with the phosphate and choline moieties of the DMPC lipids. 2 As shown in Figure S13, these water molecules are also subject to a fairly broad molecular orientation distributions (with the second moment of the dipole angle θ in the range of 34-38°) due to thermal fluctuations at the lipid/water interface. As the salt concentration is increased from ~10 µM to 0.1 M, the SFG intensities in the O-H stretching region decrease by a factor of 2. This observed salt concentration-dependent change in the SFG signal intensity is consistent with significant absorptive-dispersive mixing between χ (2) and χ (3) contributions to the SFG signal generation process from charged interfaces. ## IV. Conclusion. In conclusion Surprisingly good qualitative agreement is obtained between the difference of the measured intensity spectra for the low and high salt conditions from Figure 2 and the calculated 𝜒 !"#$ ! intensity spectrum derived from the real and imaginary data reported by Wen et al. (Figure 4). This agreement supports our conclusion that the spectral changes are not indicative of large changes in the H-bonded network of water molecules but rather result from the χ (3)potential dependent term. As shown in the TOC graphic, at low (resp. high) salt concentration, the surface potential is high (resp. low), thus modulating the SFG response according to the functional form that gives rise to the χ (3) phase angle, 𝜑. Moreover, our analysis provides a method for estimating the difference in surface potential, ΔΦ, from the magnitude of the SFG intensity difference, ΔI SFG , observed for conditions of varying ionic strength, bulk solution pH, analyte concentration, or surface composition, is known (see Supporting Information Eqn S5). The TOC graphic indicates that specific interactions between water molecules and lipid headgroups are observed as well: Replacement of PC lipids with negatively charged PG lipids coincides with SFG signal intensity reductions in the 3100 cm -1 to 3200 cm -1 frequency region. Our atomistic simulations show that this outcome is consistent with a small, albeit statistically significant, decrease in the number of water molecules adjacent to both the lipid phosphate and choline moieties per unit area, supporting the SFG observations. This result further supports recent molecular dynamics simulations indicating that the major contribution in the 3000 cm -1 to 3200 cm -1 frequency region originates from polarized water molecules that bridge phosphate and choline in the zwitterionic lipid headgroup. 2 Ultimately, the ability to probe H-bond networks over lipid bilayers holds the promise of opening paths for understanding, controlling, and predicting specific and non-specific interactions membranes with solutes such as ions 34 and small molecules such as peptides, 70 or larger species such as polycations, 30,38 and coated and uncoated nanomaterials. 31,33, The lines represent the data that have been binned by over nine points in x and y between 3000 cm -1 and 3600 cm -1 .

chemsum

{"title": "Hydrogen Bond Networks Near Supported Lipid Bilayers from Vibrational Sum Frequency Generation Experiments and Atomistic Simulations", "journal": "ChemRxiv"}

correlation_between_optical_activity_and_the_helical_molecular_orbitals_of_allene_and_cumulenes

## Abstract: Helical frontier molecular orbitals (MOs) appear in disubstituted allenes and even-n cumulenes. Chiral molecules are optically active, but while these molecules are single-handed chiral, π-orbitals of both helicities are present. Here we computationally examine whether the optical activity of chiral cumulenes is controlled by the axial chirality or the helicity of the electronic structure. We exploit hyperconjugation with alkyl, silaalkyl, and germaalkyl substituents to adjust the MO helicity without altering the axial chirality. For the same axial chirality, we observe an inversion of the helical MOs contribution to the electronic transitions and a change of sign in the electronic circular dichroism and optical rotation dispersion spectra. While the magnitude of the chiroptical response also increases, it is similar to that of chiral cumulenes without helical π-orbitals. Overall, Helical π-orbitals correlate with the big chiroptical response in cumulenes, but are not a prerequisite for it. Molecules that exhibit structural chirality are optically active. Recently, Hendon et al. 1 found that linear p-conjugated molecules with perpendicular end-groups such as allene and certain functionalised polyynes can have helical molecular orbitals (MOs) when their symmetry is reduced and the molecules become chiral. For instance, in cumulene, 4 being the number of double bonds, the frontier p-MOs come in degenerate pairs and look like those of butadiene as shown in Figure 1. A 1,5-disubstitution of cumulene reduces the symmetry from D2d to C2. The C2 symmetry causes the rectilinear p-MOs (px and py) to mix, thus breaking the degeneracy and forming pairs of helical p-MOs. 6 The frontier MOs are near-degenerate pairs of helical MOs; one of each helicity (pP and pM of P-and M-helicity). Consequently, while the molecule is one of two enantiomers, here the R enantiomer, the electronic structure contains both helicities simultaneously. Given 1,5-disubstituted cumulenes are chiral, they will be optically active, and the R-S enantiomers will have opposite chiroptical response. As can be inferred by parity, the helicities of the MOs will also be opposite between the S and R enantiomers. The relation between helical electronic structure and chiroptical response has been contemplated. Caricato analyzed MO contributions to the rotary strength of S-1,3-dimethylallene (2,3-pentadiene, structurally analogous to dimethyl cumulene in Figure 1), and demonstrated that the major contributions to the specific rotation originate from electronic transitions between a few helical p-orbitals. 8 Substituted allenic systems have been found to have big chiroptical responses. More recently, Ozcelik et al. synthesized and measured the chiroptical response of conformationally locked cyclic butadiyne systems with distinct helical electronic structure. 17 They found the chiroptical response to be among the largest measured, but could not attribute this specifically to the appearance of helical MOs in their molecules. 17 Considering the previous theoretical and experimental results that have been presented, 8,17 we find it imperative to examine the relation between the chiroptical response and helical MOs. In this letter, we exploit computationally (see Supporting Information part A for details) the hyperconjugation of substituents into the helical p-system, as a strategy to provide systematic control of helicity and energetic splitting of the otherwise near-degenerate MOs of 1,5-disubstituted cumulenes. These molecules enable the separation of orbital helicity from the axial chirality of the molecule (S or R), and we probe how changes of helicity in the electronic structure affects the chiroptical properties with time-dependent DFT (see Supporting Information part A for details). We examine the optical transitions, and compute conformation-resolved electronic circular dichroism (ECD) and optical rotation dispersion (ORD) spectra of disubstituted cumulenes. We focus on cumulenes, but the findings apply to allenes as even-n cumulenes belong to the allene family. 1,24 The test systems are selected with the aim of conceptual understanding, although we note that syntheses of allenes with permethylated silicon substituents and cumulenes have been reported. Hyperconjugation and Helical p-systems. Methyl substituents break the symmetry and the frontier MOs are explicitly non-degenerate, however, the HOMO and HOMO-1 are split by less than 1 meV. 6 This near-degeneracy of each MO pair (HOMO-1 and HOMO, LUMO and LUMO+1, etc.) seem to be retained with larger methyl-like substituents such as trimethylsilyl (Figure S1). To break the near-degeneracy, strategies utilizing donation from lone-pairs into the helical psystem have been explored. 33,34 Inspired by these approaches, we explore a strategy that let us control the splitting of the helical MO pairs and control the MO helicity. In an ethyl-substituted cumulene (3,4-heptadiene) there is conformational freedom around the C-C-C-C dihedral angle, as illustrated in Figure 2a. We examine the conformations of Rdiethyl cumulene and find local minima to have C-C-C-C dihedral angles at approximately 0° and ±120°. At ±120° the ethyl groups form separate points of chirality, as mirroring a positive dihedral angle gives the negative. We focus on the diastereomers with both dihedral angles in +120° or -120° configuration, and label them (-)R(-) and (+)R(+), indicating these have R axial chirality but opposite ethyl configuration. Shown in Figure 2b, the helicity of the frontier MOs of diethyl cumulene changes with the ethyl configuration. The order of the helical MO changes, i.e., the helicities of the MOs of (-)R(-) is opposite to those of (+)R(+). While the HOMO of (-)R(-) is a P-helix (pP), that of (+)R(+) is an M-helix (pM). As if the electronic structure is being mirrored; however, (-)R(-) and (+)R(+) are not enantiomers but diastereomers because, for example, mirroring (-)R(-) gives (+)S(+). With ethyl substituents the HOMO and HOMO-1 are split by a slightly bigger 0.012 eV, while the splitting in the (+)R(-) conformations is similar to that of the methyl-substituted cumulene (see Supporting Information part B). The saturated ethyl substituent interacts with the helical p-system through hyperconjugation. We estimate the magnitude of this electronic interaction (hyperconjugation energy) with a bond separation equation as described in Supporting Information part C. 35,36 The hyperconjugation of silicon and germanium into a carbon psystem is stronger because their s-system energetically match better with the p-system. This is evident by changing ethyl to silaethyl, providing more than twice the hyperconjugation energy at 0.40 eV (9.3 kcal/mol) compared to ethyl at 0.19 eV (4.4 kcal/mol). As illustrated in Figure 3a, The conformations of the bis-silaethyl system are the same as with ethyl substituents and provide the same control of the MO helicity. However, with silaethyl substituents the splitting between the HOMO-1 and HOMO is more substantial at 0.11 eV. We assess a range of such silicon and germanium-based ethyl-like substituents in Figure 3b, as well methyl-like substituents for control, and plot their MO splitting as a function of the hyperconjugation energy. There is a near-linear correlation for ethyl-like substituents, and permethylated disilyl provides an MO splitting of just over 0.3 eV (7 kcal/mol). The MO splitting is systematically changed with ethyl-like substituents, and the helicities are controlled by their conformation. Optical Activity. The electronic transitions of cumulene are analogous to those of allene; for a thorough discussion of the spectral assignment of substituted allenes, we refer to the extensive experimental and theoretical work by Rauk et al., 40 Runge and co-workers, 41, 42 and earlier work. 43,44 In cumulene, only the p®p* transition of B2 symmetry is electric dipole allowed, corresponding to a transition dipole along the cumulenic axis. In agreement with prior work, 41 all four p®p* transitions are superpositions (configuration interaction) of px®px* and py®py* and only S0®S4 is electric dipole allowed based on symmetry considerations (see Supporting Information part D). When the symmetry of the molecule is reduced to C2 by substituents, the first three transitions are not symmetry forbidden, however, their oscillator strengths remain negligible in all cases studied here. The S0®S4 transition of cumulene and its dimethyl, diethyl, and bis-silaethyl substituted counterparts are listed in Table 1. The transitions are almost unchanged by the substituents, having similar oscillator strengths and excitation energies. However, we note one important difference between the (-)R(-) and (+)R(+) diastereomers of bissilaethyl cumulene (bottom of Table 1). As the helical MO pairs (pM and pP, pM* and pP*) split energetically, the weighting in the configuration interaction changes. In (-)R(-) the pM®pM* excitation has the bigger coefficient. This is reversed in (+)R(+) where pP®pP* has the bigger coefficient. The helicities that contribute to S0®S4 transition are thus systematically controlled by the change of conformation. We simulate the conformation-resolved UV-Vis and ECD spectra for cumulene, and its dimethyl, diethyl, and bissilaethyl counterparts. The UV-Vis spectra of these these molecules are shown in Figure 4a and 4b for reference, and are similar with a dominant p®p* transition between 200 and 250 nm (cf. Table 1). In Figure 4c, the ECD spectra of R-1,5dimethyl cumulene and the (+)R(-) conformation of 1,5diethyl cumulene are shown. The dominant p®p* transition has negative De. The ECD spectra of the (+)R(-) conformation of the diethyl-substituted system is very similar to that of the dimethyl-substituted system. This indicates that the effect of the ethyl substituents cancels out when they are in the mixed (+)R(-) configuration. 5a, along with that of (1S,4S)-norbonenone for reference as a small molecule with high specific rotation. 45 In agreement with work on 1,3-dimethylallene by Crawford and co-workers, and others, 46-49 these R-enantiomers of cumulene have relatively small negative specific rotation. For the ORD spectra of the (-)R(-) and (+)R(+) conformations of the diethyl and bis-silaethyl substituted cumulenes shown in Figure 5b, two things are notably different. The magnitude of the ORD response is stronger compared to the dimethylsubstituted case, and the sign of the specific rotation is opposite for the (-)R(-) and (+)R(+) conformations. The same effect is present in 1,3-diethyl-substituted allene (Figure S10). As in the ECD spectra, the sign of the chiroptical response correlates with the substituent configuration and the associated change of MO helicity. The magnitudes of the specific rotation in (-)R(-) and (+)R(+) conformations of diethyl cumulene reach a similar magnitude to that of norbornenone, and for the bis-silaethyl substituted cases the magnitude is almost twice that of norbonenone. This suggests that with increasing splitting of the helical frontier MOs (HOMO-1 and HOMO), the specific rotation also increases. In Figure 5c and 5d, we plot the specific rotation at 436 nm and 589 nm against the splitting of the HOMO and HOMO-1 for the substituents listed in Figure 3b. While there is some correlation, for systems with increasing MO splitting and big specific rotation the data is more spread. The big specific rotation of norbonenone has been attributed to the electronic coupling between its two separate pchromophores. 8,23,47, The effect of splitting the degenerate px and py systems into helical MOs is possibly somewhat analogous. While the data suggests a connection between the optical rotation and the helical frontier MOs (and their splitting), it also indicates that it is not the only important factor. This is especially the case for the larger substituents that provide some of the biggest MO splittings in Figure 5c and 5d. Similarly, the correlation with the magnitude and sign of the ECD response also becomes less clear (see Supporting Information part E). With increasingly large substituents, the MOs, and thus the electronic transitions, will have less pcharacter as the conjugated part of the molecule is relatively small. Furthermore, strong hyperconjugation and dispersion interaction between the bulky saturated substituents distort the structures from the idealized ones illustrated in Figure 2a. We compute the vertical change of electron density, Dr, for the studied cumulenes in Supporting Information part D. Chiral features appear in Dr of the S0®S4 transition of (-)R(-) and (+)R(+). However, helicity is not apparent in Dr, possibly because the transitions are superpositions of excitations occurring from one helical MO to another, cf. Table 1. Although helicity does not manifest itself in the density, the chiral features of Dr are opposite for of (-)R(-) and (+)R(+). The data presented here, and in other recent work, 8,17 suggest that a large chiroptical response may be intrinsically connected to helical p-orbitals. cumulenes are structurally similar to cumulenes, but have co-planar end-groups and therefore cannot achieve helical MOs in the ground-state. 6,32 In Figure 6, we examine the ORD spectra of disubstituted cumulenes and compare them with those of the equivalently disubstituted cumulenes. Instead of axial chirality there is E-Z isomerism in the cumulenes, and consequently only the substituents provide chirality. An overview of the structure, frontier MOs, and electronic transitions are included in Supporting Information part F. Shown in Figure 6, the specific rotation of the (-)Z(-) and (-)E(-) conformations of the diethyl and bissilaethyl substituted cumulenes has the same sign as (-)R(-) of the cumulenes. With ethyl substituents, (-)Z(-) matches the magnitude of (-)R(-), and with silaethyl substituents the cumulene-response is slighter stronger than that of the cumulene. Clearly, both chiral cumulenes with and without helical MOs can achieve high specific rotation. In summary, we have developed a hyperconjugation-based strategy that enables systematic increase of the splitting of the helical HOMO and HOMO-1 in allene and even-n cumulenes. The frontier p-MOs change helicity depending on the substituent conformation, and the helicity is thus chemically controlled without changing the axial chirality of the molecule. The sign and magnitude of De in conformation-resolved ECD spectra, and the specific rotation in ORD spectra, change with the helicity of the frontier MOs of cumulenes. There is an apparent correlation between p-orbital helicity with sign and magnitude of the chiroptical response. However, the response is of similar magnitude in chiral cumulenes without helical MOs, and we conclude that helical MOs are not a prerequisite for big chiroptical response in substituted allenes and cumulenes. ## Supporting Information Computational Details. Frontier MOs of cumulenes and Allene. Hyperconjugation Energy. Electronic Transitions. Spectra. Electronic structure and transitions of cumulene. The Supporting Information is available free of charge on the ACS Publications website.

chemsum

{"title": "Correlation between optical activity and the helical molecular orbitals of allene and cumulenes", "journal": "ChemRxiv"}

limits_of_the_quantum_cognition_hypothesis:_31_p_singlet_order_lifetimes_of_pyrophosphate_from_exper

## Abstract: A proposal of quantum cognition advances the hypothesis that quantum entanglement between 31 P nuclei could serve as a means of information storage in the brain. Testing this hypothesis requires an understanding of how long-lived these quantum effects may be. We used NMR spectroscopy and molecular dynamics simulations to study the mechanisms that limit these quantum processes in 18 O-enriched molecules of pyrophosphate, the simplest biomolecule that can sustain quantum-entangled 31 P nuclear spin singlet states. We confirmed that chemical shift anisotropy limits the singlet magnetization order lifetimes in high magnetic fields, and we discovered that rapid rotation of the phosphate groups limits the lifetime in low magnetic fields. These findings represent an important starting point in studying whether quantum cognition can be a true biological phenomenon. ## Introduction It has been proposed that 31 P nuclear spin entanglement may play a role in physiology and biological information storage and transmission (1, 2). Although the notion of quantum processes involving entangled nuclear spin states may appear far-fetched, the hypothesis has not been easy to directly prove or refute. The quantum cognition proposal involves the existence of so-called Posner clusters with the stoichiometric formula Ca9(PO4)6, which through an interplay between rotational and nuclear spin states may exhibit symmetryconstrained quantized states labeled by what was called 'pseudospin'. Such clusters have yet to be detected experimentally. Smaller fragments, such as pyrophosphate, have also been considered as potential actors for 31 P nuclear singlet order (SO) modulation of reaction rates, in particular enzymatic ones (1). In this work, we therefore set out to determine the underlying limitations that would have to be considered in support of such claims, in particular regarding the lifetime of 31 P SO, represented by an exponential decay time constant TS, describing the quantum entanglement memory loss. Nuclear SO in 1 H and 13 C spin pairs has been observed to have very long TS values compared to the spin-lattice relaxation time constant T1 in a variety of compounds, in some cases one to two orders of magnitude higher (3)(4)(5)(6)(7).Very recently, SO between 31 P nuclei has been observed and characterized in large diphosphate compounds (8,9). For the 31 P spins in the compounds studied, however, singlet relaxation has been found to be much more rapid than spin-lattice relaxation, with a major reason being the anticorrelation between the chemical shift anisotropy (CSA) tensors of the two spins (8). Two aspects of this prior work motivated us to examine 31 P-spin SO further. The compounds used previously were particularly bulky and contained large asymmetries between the two spins (either transient or constant). We therefore sought to study the small, highly symmetric molecule pyrophosphate, modified to have slight asymmetry, thereby enabling access to SO. Since the main mechanism in prior work on substituted phosphates appeared to be due to CSA, we wished to perform field-dependent studies. We present here Zeeman and SO relaxation studies over a large field range (2 𝜇T to 9.4 T) to investigate the major relaxation mechanisms as a function of magnetic field, and to determine, in particular, the underlying low-field limit to the SO relaxation time. We further identify the mechanistic contributions to SO relaxation by molecular dynamics (MD) and ab initio computation. ## Synthesis and NMR characterization of unsymmetrically 18 O-labeled pyrophosphate One challenge in the study of SO in the pyrophosphate (PPi) molecule is the lack of inequivalence (either chemical or magnetic), which is needed for creating and reading out SO of the 31 P spins. To overcome this challenge, we unsymmetrically labeled PPi with the 18 O isotope. The increased mass of the 18 O nuclei relative to the abundant 16 O isotope was expected to induce a small chemical shift difference between the neighboring 31 P nuclei, sufficiently large to allow creation and read-out of SO. This strategy was used previously for pairs of 13 C nuclei (10). The tetrasodium salt of the unsymmetrically labelled 18 O-PPi (uPPi) was synthesized as described in the Materials and Methods section below and prepared in D2O under highly alkaline conditions to avoid potentially interfering effects due to proton exchange, which can accelerate SO relaxation (11). An excess of potassium cations, relative to the sodium cations from the tetrasodium salt, was found to promote longer SO lifetimes than if sodium ions alone were present. Similar results were obtained by adding ethylenediaminetetraacetate (EDTA) instead (Fig. S1, Supplementary Materials). The NMR properties of the synthesized uPPi 31 P spin system were extracted from a 31 P pulse-acquire spectrum acquired at 9.4 T by multiplet simulation and fitting using the Spinach MATLAB package (http://spindynamics.org/group/) (12). Fig. 1 displays the fitting results. The unsymmetrical isotopic labelling of the uPPi induces a slight chemical shift difference ∆𝛿PP between the two 31 P nuclei of 0.0663 ppm, or 10.7 Hz at 9.4 T. The 31 P nuclei share a homonuclear J-coupling of magnitude 2 JPP = 21.5 Hz. Thus, the uPPi 31 P spin system is in a strongly coupled regime at 9.4 T. Singlet-triplet mixing can occur at high fields, but this mechanism of SO decay is eliminated when the sample is moved to lower fields. Additional peaks are observed which likely stem from partial labelling of the molecule. We could not fully identify these, but products with partial labelling should not affect the results, since the triplet-singlet transfer is tailored to a particular chemical shift / coupling combination. The isotope composition should not affect relaxation rates due to the small differences in mass. The 31 P R1 values of the unlabeled PPi and the 18 O-labeled uPPi were measured at 9.4 T as 0.107 s -1 and 0.102 s -1 , respectively, with identical solution conditions (pD 14.4, 25 °C). ## NMR field-cycling relaxation measurements of uPPi We then performed field-dependent measurements of both SO relaxation and spin-lattice relaxation, in order to compare and contrast known relaxation mechanisms. We chose to utilize the spin-lock induced crossing (SLIC) pulse sequence (13) for preparing and reading out SO for NMR spectroscopic relaxation measurements. The SLIC pulse sequence used for field-dependent measurements of RS (= 1/TS) is displayed in Fig. 2. Optimization of the power and duration of the SLIC spin-lock pulse confirmed the spin system parameters determined via spectral fitting: the optimal pulse amplitude and duration corresponded with 2 JPP of 20.3 Hz and a ∆𝛿PP of 12.3 Hz (Fig. S2, Supplementary Materials). We performed 31 P relaxation measurements on ~300 mL of the uPPi solution in a 5 mm NMR tube using a 9.4 T Bruker NMR spectrometer equipped with a home-built field shuttling system, to transport the sample rapidly between regions of different magnetic field. The shuttling system included a shielded region above the magnet and therefore enabled access to magnetic field strengths as low as 2 µT. An inversion-recovery sequence was used with the same sample shuttling setup in order to measure R1 (= 1/T1) Conversions between detectable magnetization and singlet order were performed at 9.4 T, and the sample was shuttled to low field for the incremented relaxation delay, then shuttled back for detection. A T00 filter prior to singlet readout was used with two-step phase cycling on the first 90° pulse and receiver to remove undesired coherence pathways. The relaxation measurements are shown in Fig. 3. Generally, the R1 and RS values tracked each other, with R1 experiencing a slight increase in the 2 µT to 200 mT range. RS also tended to be smaller than R1 in the high-field regime, above 4.5 T. Both R1 and RS approached a constant relaxation rate offset of approximately 0.018 s -1 at the lowest field values measured. The measured relaxation trends with magnetic field were well approximated using MD simulations and ab initio calculation (Fig. 3, dashed lines), as described below. ## Molecular dynamics simulation and ab initio calculation of relaxation rate curves In order to study the CSA tensors in uPPi and their contributions to longitudinal and SO relaxation, MD simulations were performed using Gaussian 16 and Amber20 (14) software, as described in the Materials and Methods section. Briefly, the uPPi electronic structure was modeled in Gaussian, then energy minimization was performed over 5000 steps, followed by 20,000 steps of 1 fs to reach the desired temperature and pressure (300 K, 1 bar), and the final production run was performed with an isothermal/isobaric ensemble (NPT, 300 K, 1 bar) with 10 7 steps. The CSA tensors at each 31 P nucleus were then calculated ab initio in Gaussian from 100 randomly selected conformations. Fig. 4 shows average and multiplesnapshot representations of the symmetric portion of the CSA tensors experienced by the 31 P nuclei. As is seen here, the principal component appears almost completely aligned with the bond between phosphorus and the bridging oxygen. Because the -PO3 groups experience fast intramolecular rotation about the bridging P-O bond (see Fig. 4B), the CSA tensors were averaged across the 100 conformations, following molecular alignment along the P-P vector. A more detailed justification for this averaging procedure can be found in the Materials and Methods section. The difference between the average tensors at each 31 P nucleus was computed, and the average and difference tensors were separated into their symmetric and antisymmetric components. The (Frobenius) norms of the tensor components are summarized in Table 1 and were used to calculate the CSA contributions to R1 and RS using the expressions In the equations above, 𝜔 % is the Larmor frequency, ‖𝜎‖ & and ‖∆𝜎‖ & indicate the Frobenius norms of the average and difference tensors, respectively, and 𝜏 ! and 𝜏 ' are the first-and second-rank correlation times, respectively, where 𝜏 ! = 3𝜏 ' assuming isotropic motion. The second-rank correlation time was determined to be 48.6 ps, based upon MD simulation following adjustment using the NMR-measured PPi diffusion coefficient, as described in the Materials and Methods section. It is seen that CSA accounts for the major relaxation effect at high magnetic fields. The symmetric CSA component (Fig. 5, solid lines) contributes the most to R1 and RS at high field strengths, whereas the antisymmetric contribution (Fig. 5, dotted lines) is relatively small for both but much larger for RS than it is for R1. Other smaller yet significant relaxation contributions, largely field-independent, are discussed further below. The spin-rotation contribution to R1 was calculated as follows: From MD simulations, the correlation function ω(0)ω(𝑡) for the angular rotation frequency of the -PO3 entity about the bridging P-O bond of PPi was calculated. An exponential fit was performed to this function, which yielded 𝜔(0) ' and the correlation time 𝜏 -. These values were determined as 3.1 rad 2 ps -2 and 0.0255 ps, respectively. Gaussian 16 was used to compute the spinrotation tensor for 31 P in PPi at the B3LYP/aug-cc-pVTZ level, which produced the value for 𝐶 ∥ /2π = 4.424 kHz, for rotation around the bridging P-O vector, and roughly two equivalent values for the perpendicular rotation 𝐶 / /2π = 1.095 kHz. The spin-rotation tensors are visualized in Fig. S3 in the Supplementary Materials, which indicates that the major component of this tensor also points along the bridging P-O bond similar to the CSA tensor. Given that the motion perpendicular to the P-O bond can be assumed to be very small by comparison (see Fig. 4B, showing the superposition of conformers obtained from MD trajectories), we neglect this portion and calculate the spin-rotation relaxation rate constant by the expression where 𝐼 || = 1.758×10 -45 kg m 2 is the moment of inertia for the -PO3 entity with respect to the bridging P-O axis. This expression can be derived by combining Eq. ( 22) from McClung (15) with Eq. (4.83) from Spiess (16). The spin-rotation relaxation rate constant then becomes R1,SR = 0.0113 s -1 . The rate is essentially independent of the magnetic field due to the extremely short correlation time for the angular frequency correlation function. Spin-rotation is also expected to affect the relaxation of SO in uPPi. We made the following considerations: were the spin-rotation field fluctuations produced by each rotating -PO3 group fully uncorrelated, we would predict RS,SR to be twice as large as R1,SR. However, in this case RS would be larger than R1 at low field strengths, whereas experimentally we observed similar low-field values of R1 and RS. We therefore determined the correlation coefficient 𝛼 for the spin-rotation interaction at each 31 P spin following the discussion about correlated mechanisms of Tayler et al (17), in particular Eq.s (1) and (2). From these considerations, once can obtain 𝑅 ,,,1 /𝑅 !,,1 = 2(1 − 𝛼), and when using the experimental values for 𝑅 ,,,1 and 𝑅 !,,1 we obtain the correlation coefficient 𝛼 = 0.5. Modeling the spinrotation contribution to RS in this manner produced an excellent fit to the experimental data (Fig. 3, dashed line). Other known relaxation contributions to R1 and RS are described below. MD simulations following the procedure of Kharkov et al (18) gave the contribution of intermolecular dipolar relaxation between 31 P and 2 D solvent spins as 5.14×10 -3 s -1 . The 31 P-31 P dipolar relaxation contribution, relevant only for R1, was determined to be 1.60×10 -3 s -1 . The correlation times for these processes range from 20-40 ps, and therefore their contributions are likewise almost completely independent of the magnetic field. The singlettriplet leakage (STL) contribution to SO relaxation cannot easily be determined in closed form, since it depends on the specifics of the relaxation mechanism. This effect was therefore estimated using the Spinach NMR simulation package in MATLAB (12), by simulating SO relaxation with and without the chemical shift difference included and calculating the difference. The contribution is field-dependent but relatively minor, as seen in Fig. 5. Finally, the 1 H-31 P dipolar relaxation contribution arising from the added KOH was estimated from the 2 D-31 P contribution as 0.00025 s -1 , which is negligible compared to other relaxation contributions. ## Discussion The quantum cognition proposal advanced by Fisher (1) involves multiple components requiring independent validation. One of the most prominent challenges of the proposal in its current form involves assessing whether quantum entanglement can survive for an appreciable duration in a "wet" biological environment, even though some level of protection from the environment may be provided by symmetry in a Posner cluster (1). Our approach to studying entangled spin order in unsymmetrically labeled pyrophosphate represents an important initial step in studying what phenomena most strongly limit the lifetime of quantum entanglement between 31 P nuclei. Our R1 and RS measurements show that uPPi high-field relaxation is dominated by the CSA mechanism, similar to the case in other reported diphosphates (8,9). In contrast to previous studies, the RS values observed in the high field regime are slightly lower compared with R1, which correspond well with the symmetric CSA tensor norm being somewhat lower for the difference tensor (Table 1). The norm of the antisymmetric component, however, is significantly larger for the difference tensors than for the individual tensors, with the result being a larger antisymmetric CSA contribution to RS. Still, the antisymmetric contribution to RS is smaller than one fifth of the symmetric contribution. Importantly, we observed that towards low fields, a constant offset in relaxation rate constants is approached for the experimentally measured values of both R1 and RS. The offset at the lowest field, 2 𝜇T, was found to be approximately 0.018 s -1 for both. The same trend and a similar, albeit slightly higher R1 and RS offset were observed from measurements on a 30 mM uPPi sample with 10 mM EDTA added (Fig. S4, Supplementary Materials). We believe this constant contribution at the lowest field to be primarily comprised of spinrotation relaxation, as shown in Fig. 5. Furthermore, at very low field strengths (2 µT to 100 mT), R1 showed a peculiar increase in the rate that was consistently observed across different sample formulations (Fig. S4A). This effect is not understood at this time. The largest values of the T1 and TS time constants appear to be approximately 65 s for uPPi under our experimental setup (in the low field range). We note, however, that many of the experimental conditions used for our relaxation measurements are different from those that would be encountered in a biological system. First, our experiments were performed at a relatively high pD, to limit deuteron exchange, whereas faster exchange at physiological pD values would be expected to reduce the TS and possibly T1 relaxation times (11). In addition, the nature of the counterion played a role in the relaxation measurements, and the longest T1 times were observed with an excess of K + ions relative to the Na + ions from the synthesized uPPi tetrasodium salt. It is worth noting that the intracellular K + concentration tends to be approximately fourfold higher than Na + , whereas Na + is much more abundant in the extracellular space (19). This finding may suggest that the intracellular environment is more conducive to long-lived quantum entanglement, at least for free pyrophosphate. Furthermore, D2O was used as a solvent rather than H2O. We note that if H2O were used as a solvent, this limit would be significantly smaller. We measured an increase in R1 of 0.028 s -1 at 9.4 T when we replaced D2O with 90% H2O plus 10% D2O. Assuming this increase to be field-independent, we would therefore expect a T1 and TS maximum of approximately 26 s if we were to repeat the field-dependent measurements with this solvent. Finally, certain paramagnetic species are abundant within cells and tissues and can contribute to relaxation. Comparison of rates observed in degassed and non-degassed samples, however, showed approximately the same rate constants in the low field region, suggesting that the effect of paramagnetic relaxation due to oxygen is low (Fig. S4, Supplementary Materials). Other paramagnetic impurities were considered, but careful and extensive cleaning of glassware with KOH/iPrOH and HCl did not produce significant changes. Examination of relaxation in the presence of EDTA (to potentially capture paramagnetic impurities) likewise did not show significant changes in the observed rate constants (Fig. S4, Supplementary Materials). In summary, we report measurements of 31 P SO decay in isotope labeling-induced unsymmetric PPi over a wide range of field strengths. We demonstrate that CSA dominates both R1 and RS relaxation at high fields but diminishes at low fields, and that the two rates have similar values from 2 µT to 9.4 T. We observe that both R1 and RS approach a constant value at low field strengths, and that this relaxation appears to be primarily explained by spin-rotation relaxation, with minor (but non-negligible) contributions from intermolecular 31 P-2 D dipolar coupling and intramolecular 31 P-31 P dipolar coupling. The magnitude of the spin-rotation relaxation contribution in this molecular system was an unexpected discovery. Our main experimental finding is that the relaxation rates for 31 P longitudinal magnetization and for 31 P nuclear SO are similar for pyrophosphate in solution, with multiple mechanisms contributing to both relaxation processes. Both relaxation times are of the order of 1 minute in low magnetic field under our experimental conditions, and they decrease rapidly as the magnetic field is increased. In low magnetic fields the 31 P singlet lifetime of pyrophosphate is possibly long enough to sustain the hypothesis that such entangled spin pairs might play a role in quantum cognition (1, 2). As far as the authors know, there is no evidence that cognition is significantly disturbed by high magnetic fields, as would be anticipated from the experimental results described here. ## Materials and Methods Unsymmetrically 18 O-labeled pyrophosphate synthesis and formulation The synthesis of 18 O/ 16 O unsymmetrical pyrophosphate tetrasodium salt 6, henceforth referred to as uPPi, is shown in Fig. 6. Light sensitive silver phosphate salt 1 was prepared from 18 O phosphoric acid by a simple precipitation method (20). Subsequent benzylation in the presence of excess benzyl chloride provided the triester 2 in 75% yield (21). Heating triester 2 in the presence of one equivalent of sodium iodide in acetone accomplished selective mono-deprotection (21), and the resulting dibenzyl phosphate sodium salt 3 was converted to the tetrabenzyl 18 O/ 16 O pyrophosphate 4 by reaction with dibenzyl phosphoryl chloride ( 16 O, obtained by the chlorination of dibenzyl phosphite with NCS in benzene and used directly (22)) in the presence of triethylamine (23). Global debenzylation of the tetrabenzyl pyrophosphate using hydrogen over Pd required prolonged reaction times and was inefficient due to accompanying partial hydrolysis to the orthophosphate. Ultimately, a two-step procedure via the dibenzyl pyrophosphate disodium salt 5 was optimised, with the remaining two benzyl groups removed by hydrogenolysis over Pd in the presence of sodium bicarbonate in 5 hours. This six-step sequence afforded the regioselectively O 18 /O 16 labelled pyrophosphate tetrasodium salt 6 as a white crystalline solid. Isotopic incorporation was confirmed by mass spectrometry to be 96% 18 O4, 96% 18 O3. For NMR experiments, the tetrabasic sodium uPPi was formulated as a 30 mM solution in deuterium oxide plus 10 equivalents of potassium hydroxide, in order to minimize proton exchange, which can accelerate singlet relaxation (11), and minimize interactions with sodium ions in solution, which our results seem to indicate also accelerates relaxation of both longitudinal and spin order (Fig. S1, Supplementary Materials). The final concentrations of Na + and K + counterions were 120 mM and 300 mM, respectively. The pD of the solution was expected to be about 14.4, based upon room-temperature pH electrode measurements of a sample prepared identically but with unlabelled tetrabasic sodium pyrophosphate. The NMR tubes used with the samples were carefully cleaned to avoid any paramagnetic impurities by immersing in a KOH/iPrOH bath overnight followed by HCl immersion overnight, rinsing several times with acetone, and drying with argon gas. More details on sample preparation can be found in the Supplementary Materials. ## Field-dependent NMR spectroscopy All field-dependent NMR measurements were performed at the University of Southampton. For measurements of R1 via inversion-recovery, the uPPi 31 P populations were inverted with a 180° pulse, the sample was shuttled to a region with the desired magnetic field strength, and then the sample was returned to the bore for excitation with a 90° pulse followed by acquisition. SO was prepared with a SLIC spin-lock pulse at 9.4 T within the bore, the sample was shuttled to a region above the magnet for singlet relaxation at the desired field strength, and then returned to the magnet bore for singlet order readout via SLIC. The sample shuttling speed to and from the low field for all measurements was about 1 m/s, and the shuttling time (one-way) was no greater than 1 second. The sensitivity of singlet-triplet conversion due to transmitter offset during SLIC was mitigated by turning off the temperature regulation within the NMR scanner, in order to minimize the change in temperature between the bore and the shuttling region above the magnet. The probe temperature within the bore was measured to be about 22 °C with the temperature regulation off, and the temperature during sample shuttling was not expected to vary more than ±5 °C from the probe temperature. ## Simulation methods MD simulations in Amber20 were performed with the following modifications: PPi was parametrized using ESP charges obtained from Gaussian 16 with B3LYP/6-31G(d), the polyphosphate parameters described by Meagher et al (24), with the missing parameters provided by the GAFF2 force field. Minimization was performed in 5000 steps, Timesteps were 1 fs throughout, and the final isothermal/isobaric ensemble (NPT, 300 K, 1 bar) production run contained 10 7 steps. The simulation was performed at 300 K. 100 snapshots were selected randomly to perform ab initio calculations of CSA tensors with the B3LYP/aug-cc-pVTZ combination and the GIAO method. Fig. S5 in the Supplementary Materials shows the individual tensor norms and eigenvalues of the tensor components for all conformers. To calculate the average CSA tensors across all selected conformations, the molecules were aligned along the P-P vector (i.e. along the x coordinate) with the bridging P-O vector pointing upwards in the x-z plane, as shown in Fig. 4B. The CSA tensors were rotated into this frame and averaged. For the R1 calculation, the Frobenius norms were taken of the symmetric and antisymmetric components of the average tensors. For the RS calculation, the Frobenius norm was calculated for the difference between the average tensors of each 31 P. Tensor visualizations were generated using the Ovaloid function from SpinDynamica v3.6 (25) in Mathematica, as described previously (26,27), and displaying with the MoleculePlot3D function. The CSA tensor averaging procedure described above is strictly valid only in the limit where the internal motion is much faster than the overall tumbling rate. We justify its use as follows: from the MD trajectories the root mean square (rms) angular frequency of the -PO3 rotation around the bridging P-O bond is determined as 1.76 rad/ps. From this value, we can calculate the root-mean square rotation of -PO3 within the reorientation correlation time period determined above (48.6 ps) as 13.6π. We therefore can assume that the -PO3 rotation is much faster than the molecular reorientation, so that averaging the tensors for the two 31 P spins prior to taking the differences between them is the correct approach.

chemsum

{"title": "Limits of the quantum cognition hypothesis: 31 P singlet order lifetimes of pyrophosphate from experiment and simulation Authors", "journal": "ChemRxiv"}

impacts_of_simulated_erosion_and_soil_amendments_on_greenhouse_gas_fluxes_and_maize_yield_in_miamian

## Abstract: Erosion-induced topsoil loss is a threat to sustainable productivity. Topsoil removal from, or added to, the existing surface is an efficient technique to simulate on-site soil erosion and deposition. A 15-year simulated erosion was conducted at Waterman Farm of Ohio State University to assess impacts of topsoil depth on greenhouse gas (GHG) emissions and maize yield. Three topsoil treatments were investigated: 20 cm topsoil removal, 20 cm topsoil addition, and undisturbed control. Results show that the average global warming potential (GWP) (Mg CO 2 Eq ha −1 growing season −1 ) from the topsoil removal plot (18.07) exhibited roughly the same value as that from the undisturbed control plot (18.11), but declined evidently from the topsoil addition plot (10.58). Maize yield decreased by 51% at the topsoil removal plot, while increased by 47% at the topsoil addition plot, when compared with the undisturbed control (7.45 Mg ha −1 ). The average GWP of erosion-deposition process was 21% lower than that of the undisturbed control, but that greenhouse gas intensity (GHGI) was 22% higher due to lower yields from the topsoil removal plot. Organic manure application enhanced GWP by 15%, and promoted maize yield by 18%, but brought a small reduction GHGI (3%) against the N-fertilizer application. In agricultural ecosystem, sustainable food production and mitigation of greenhouse gas (GHG) emissions have been concerned by agricultural or environmental scientists, especially under future climate conditions. Accelerated erosion is one of the most prevalent forms of soil degradation in the world 1,2 , which poses major threat to food security 3 and a significant impact on GHG emissions 4 . Erosion translocate sediment and soil organic C laterally across landscapes (0.5-0.6 Pg C year −1 ) 5 , potentially causing approximately 0.8-1.2 Gt C year −1 emissions into the atmosphere, while burying 0.4-0.6 Gt C year −1 by deposition processes 2 . Although erosion-induced GHG emissions are dominated by CO 2 , the fluxes of CH 4 and N 2 O are also considerable 6 . IPCC (Intergovernmental Panel on Climate Change) (2007) stated 7 that the emissions of N 2 O and CH 4 have global warming potentials (GWPs) of 310 and 21 times that of CO 2 , which have not yet been adequately investigated. Soil erosion comprises three stages: detachment, transport/redistribution, and deposition 8 . The first two stages, detachment and transportation, lead to increased mineralization and emission of CO 2 . However, the prevalence of anaerobic conditions at the depositional stage reduces the emission of CO 2 and leads to flux of CH 4 and N 2 O 9 . Up to now, there is no systematically assessment on grain yield and GHG emissions under soil erosion-deposition events. In fact, it is difficult to detect the decline of productivity that results from erosion directly, because the productivity reduction caused by erosion often occurs so slowly that it may not be recognized until crop production is no longer economically viable 10 . Moreover, improved technology often masks productivity decline caused by erosion, leading to increased rather than decreased yields 4,10 . Various indirect methods (e.g., the comparative-plot method, transect method, and desurfacing experiments) have been carried out extensively in the study of erosion-productivity relationships 10 . The simplest method is to artificially remove topsoil (which is also referred to as desurfacing experiments 10,11 ). Previous studies have reported that the yield reduction rate was faster after the top 40 cm soil was eroded, but became slower if the deeper soil was lost 12 . Moreover, desurfacing approach can also help eliminate the inherent variability of topsoil depth and landscape position 13 . Restoration of degraded soils is a high priority in global scale 4 . When soil carbon pool in degraded cropland increased by one ton, crop yield would increase by 20-40 kg ha −1 for wheat, 10-20 kg ha −1 for maize, and 0.5-1 kg ha −1 for cowpeas 2 . Substantial studies have reported the restored productivity of de-surfaced soils by amending with fertilizer or manure . In this study, we investigated the impacts of simulated soil erosion-deposition (after 15 years of establishment) on greenhouse gas (GHG; CO 2 , N 2 O, and CH 4 ) emissions and maize yield, via topsoil depth (TSD) removal and addition 8,17 treatments during the growing season, under N-fertilizer and organic manure amendments with no-till management. ## Results Soil temperature, moisture and GHG emissions. Figure 1 displays the diurnal air temperature and precipitation distribution in 2012; the inset shows the cumulative monthly precipitation and mean monthly temperature in 2012 (OARDC). Mean air temperature showed an increase from April (11.55 °C) to July (26.53 °C), followed by a decrease thereafter. Monthly cumulative precipitation varied between 3.69 and 7.96 cm from April to September, with the highest precipitation in September and lowest precipitation in August. Figure 2 shows the variation of soil temperature and soil moisture content at 0-10 cm impacted by simulated erosion under N-fertilizer and organic manure application during the growing season in 2012. Like air temperature, soil temperature showed an increase from April to July, followed by a decrease thereafter. During the growing season, mean soil temperatures for the three TSD treatments were 25.65, 24.75, and 24.92 °C respectively for top soil removal, undisturbed control, and topsoil addition. Obviously higher soil temperature trend was observed for topsoil removal under N-fertilizer application. While, mean soil moisture content for topsoil removal was significantly higher than other two TSD treatments (P = 0.0372) at 10 cm depth. Figure 3 shows the effects of simulated erosion on CO 2 fluxes under N-fertilizer and organic manure application during the growing season in 2012. All three TSD treatments were persistent CO 2 sources both for N-fertilizer and organic manure application during the study. Soil CO 2 fluxes differed among seasons, which stirred by fertilizer application and reached the maxima at the peak of air temperature and corn growth 18 . Under N-fertilizer application, CO 2 fluxes in all three TSD treatments were below 4 g C m −2 d −1 from April 6 th to June 8 th , 2012. CO 2 fluxes increased sharply from July to August, and observed the maximum fluxes (g C m −2 d −1 ) on August 7 th (9.22), July 3 rd (8.98) and July 20 th (4.55) respectively for topsoil removal, undisturbed control and topsoil addition. For organic manure application, peak CO 2 fluxes (g C m −2 d −1 ) appeared on July 3 rd , which were 10.4, 15.54 and 8.94 respectively for topsoil removal, undisturbed control and topsoil addition. For N-fertilizer application, there was a 19% increase in the topsoil removal treatment and a 67% decrease in the topsoil addition treatment compared with undisturbed control (Table 1). For soil receiving organic manure, no significant difference observed among the three TSD treatments (Table 1). Average cumulative CO 2 emissions for the entire growing season significantly differed among the three TSD treatments (P = 0.0003) (Table 2). Figure 3 shows the effects of simulated erosion on N 2 O fluxes under N-fertilizer and organic manure application during the growing season in 2012. Three TSD treatments were also persistent sources of N 2 O fluxes both for receiving N-fertilizer and organic manure during the study. Soil N 2 O fluxes also displayed a small increase on May 10 th after the fertilizer application, and a peak flux coincided with the peak air and soil temperatures in July. The results indicate that the CO 2 and N 2 O fluxes were stimulated by fertilizer application and soil temperature rise. N 2 O fluxes fluctuated below 3 mg N m −2 d −1 from April to June, and rose rapidly from June 8 th , got the maxima fluxes on July 3 rd both under N-fertilizer application (7.35, 6.59 and 3.90 mg N m −2 d −1 respectively for topsoil removal, undisturbed control and topsoil addition) and organic manure application (8.02, 6.70 and 4.71 mg N m −2 d −1 respectively for topsoil removal, undisturbed control and topsoil addition), then declined to the original level (<3 mg N m −2 d −1 ) thereafter until September (Fig. 3). Our result agree with the statement that the N 2 O emission often characterized by a short time of very high flux rates that make up a substantial part of the total annual loss 19 . Under N-fertilizer application, cumulative N 2 O emission decreased by 19% and 42% respectively for topsoil removal and topsoil addition treatments, compared with the undisturbed control (Table 1). For organic manure application, cumulative N 2 O emission increased by 16% for topsoil removal compared with the undisturbed control, and no significant difference was observed between the undisturbed control and topsoil addition (Table 1). Average cumulative N 2 O emissions for the entire growing season significantly differed among the three TSD treatments (P < 0.0001) (Table 2). Figure 3 shows the effects of simulated erosion on CH 4 fluxes under N-fertilizer and organic manure application during the growing season in 2012. Under N-fertilizer application, CH 4 fluxes in the three TSD treatments were generally low (i.e., averaged <2 mg C m −2 d −1 ) and broadly taken up during the growing season. While, under organic manure application, CH 4 fluxes fluctuated widely, and a peak positive CH 4 flux (5.68 mg C m −2 d −1 ) observed on July 3 rd for topsoil removal. For N-fertilizer application, all three TSD treatments were net CH 4 sinks from the atmosphere during the growing season (Table 1). Cumulative CH 4 uptake decreased for both topsoil removal (by 49%) and topsoil addition (by 67%) compared with the undisturbed control (−1.78 kg C ha −1 growing season −1 ) (Table 1). Under organic manure application, topsoil removal was net CH 4 source, and two other TSD treatments were net CH 4 sinks during the growing season, cumulative CH 4 emission increased by 174% for topsoil removal compared with the undisturbed control (−0.77 kg C ha −1 growing season −1 ) (Table 1). Average cumulative CH 4 emissions significantly differed among the three TSD treatments (p < 0.0001) for the entire growing season (Table 2). Soil temperatures were negatively correlated with soil moisture contents both under N-fertilizer application (P = 0.0232, 0.0397 and 0.0046 respectively for topsoil removal, undisturbed control and topsoil addition), and organic manure application (P = 0.0350 for topsoil addition). Soil temperature was positively correlated with CO 2 fluxes (P = 0.0787) and N 2 O fluxes (P = 0.0918) from the topsoil removed plot under N-fertilizer application. Similar and much stronger correlations were also found between soil temperature and CO 2 fluxes (P = 0.0138 and 0.0453 respectively for topsoil removal and undisturbed control) as well with N 2 O fluxes (P = 0.0194 and 0.0615 respectively for topsoil removal and undisturbed control) under organic manure application. CO 2 fluxes were positively correlated with N 2 O flux both under N-fertilizer (P = 0.0073, 0.0002 and 0.0024 respectively for topsoil removal, topsoil addition and undisturbed control) and organic manure application P = 0.0025 and 0.0023 respectively for topsoil removal and topsoil addition; P < 0.0001 for undisturbed control). The CH 4 fluxes were negatively correlated with soil temperature both under N-fertilizer application (P = 0.0466 for undisturbed, P = 0.0255 for topsoil addition) and organic manure application (P = 0.0734 for topsoil addition). Soil bulk density and SOC, total N content. Figure 4 shows the soil bulk density affected by simulated soil erosion under N-fertilizer and organic manure application. Soil bulk density was higher for topsoil removal and lower for topsoil addition at every soil layer depth from 0-40 cm, significant difference were observed among the three TSD treatments both under N-fertilizer (at 20 and 40 cm soil layer depth), and organic manure (20 cm soil layer depth) application. Soil bulk density for soil receiving organic manure was lower than soil with N-fertilizer application at every soil layer depth from 0-40 cm, with significant difference observed at 0-10 cm soil layer depth. Eq Mg grain yield growing season −1 ) = GWP/grain yield Figure 5 shows the SOC and total N content affected by simulated erosion under N-fertilizer and organic manure application. Significant differences of SOC and total N at the top soil layers displayed inverse patterns from that at the lower layer. At the soil layer of 0-10 and 10-20 cm, the SOC and total N were in the order of undisturbed control > topsoil addition > topsoil removal, while in reverse order of topsoil addition > undisturbed control > topsoil removal at 30-40 cm soil layer depth. For topsoil addition, the migrated topsoil might lose part of C from decomposition due to soil disturbance 20,21 ; meanwhile, the former topsoil buried under the plough depth was preserved from decomposition and mineralization . Soil receiving organic manure had higher SOC and total N content than soil with N-fertilizer application at every soil layer depth from 0-40 cm. Soil GWP, GHGI and maize yield. The GWP (Table 1) increased by 16% for topsoil removal and decreased by 64% for topsoil addition compared with the undisturbed control (17.36 Mg CO 2 Eq ha −1 growing season −1 ) under N-fertilizer application; and GWP decreased by 14% and 21% respectively for topsoil removal and topsoil addition compared with the undisturbed control (19.01 Mg CO 2 Eq ha −1 growing season −1 ) under organic manure application. The mean GWP for topsoil removal and topsoil addition decreased by 24% under N-fertilizer application, and decreased by 18% under organic manure application, compared with the undisturbed control. For N-fertilizer application, the GHGI (Table 1) increased by 113% for topsoil removal and decreased by 71% for topsoil addition, compared with the undisturbed control (2.40 Mg CO 2 Eq Mg −1 grain yield growing season −1 ); the average GHGI of the topsoil removal and topsoil addition was 21% higher than that of the undisturbed control. Under organic manure application, the GHGI increased by 101% for topsoil removal and decreased by 53% for topsoil addition, compared with the undisturbed control (2.48 Mg CO 2 Eq Mg −1 grain yield growing season −1 ); the average GHGI of the topsoil removal and topsoil addition was 24% higher than that of the undisturbed control (Table 1). Figure 6 shows the grain yield and above ground residue affected by simulated erosion under N-fertilizer and organic manure application in 2012. The grain yield significantly decreased (by 45% under N-fertilizer, by 57% under organic manure) for topsoil removal, while increased (by 24% under N-fertilizer, by 69% under organic manure) for topsoil addition compared with the undisturbed control (7.23 Mg ha −1 under N-fertilizer, 7.67 Mg ha −1 under organic manure). The average grain yield of topsoil removal and topsoil addition decreased by 10% under N-fertilizer application, and increased by 6% under organic manure application, compared with the undisturbed control (7.23 Mg ha −1 under N-fertilizer, 7.67 Mg ha −1 under organic manure). The above ground residue decreased (by 28% under N-fertilizer, by 54% under organic manure) for topsoil removal, and increased (by 23% under N-fertilizer, by 41% under organic manure) for topsoil addition, compared with the undisturbed control (6.31 Mg ha −1 under N-fertilizer, 9.46 Mg ha −1 under organic manure. The average above ground residue of topsoil removal and topsoil addition decreased by 2% under N-fertilizer application, and decreased by 7% under organic manure application compared with the undisturbed control (6.31 Mg ha −1 under N-fertilizer, 9.46 Mg ha −1 under organic manure). Significant difference were observed among three TSD treatments both for grain yield (P < 0.001) and above ground residue (P = 0.0256) (Table 2). ## Discussion Topsoil remove and addition can not only significantly affect the greenhouse gases emissions, but also changes the crop yield (Table 1 and Fig. 6), which indicated that the erosion-deposition process can significantly alter the global warming effects, soil nutrient status and soil productivity during the erosion events. The average cumulative CO 2 emission in our study was 3.89 Mg C ha −1 growing season −1 ; the value fell into the range of seasonal CO 2 emission reported by several global studies 20, . In the present study, cumulative CO 2 emission significantly increased following 20 cm topsoil removal and decreased following 20 cm topsoil addition throughout the maize growing season compared with the undisturbed soil under N-fertilizer application (Table 1). The difference of cumulative CO 2 emissions among three TSD treatments could be explained by the soil temperature variation that caused by aboveground coverage shade 28 . Soil temperature was the primary drive to CO 2 flux 21, . The enhanced cumulative CO 2 emission at the eroded site may primarily result from its higher soil temperature of 25.9 °C (from 17.5 to 33.9 °C) with less aboveground coverage shade. The reduced cumulative CO 2 emission at the depositional site probably due to its lower soil temperature of 24.7 °C (from 15.5 to 32.2 °C) owing to its dense above ground coverage shade (Figs 2 and 6). In addition, the reduced cumulative CO 2 emissions at depositional site probably also caused by their lower substrate availability (e.g., SOC) 28 in surface soil (Table 1 and Fig. 5). While, under organic manure application, no significant difference observed among the three TSD treatments for cumulative CO 2 emissions (Table 1), which probably due to their similar average soil temperatures (respectively 22.6, 22.5 and 22.5 °C) (Fig. 2). It was worthwhile to note that the cumulative CO 2 emission for the organic manure applied plot (4.16 Mg C ha −1 growing season −1 ) was much greater than that from the plot receiving N-fertilizer (3.62 Mg C ha −1 growing season −1 ) (P = 0.0345). This was probably due to the greater carbon substrate (SOC) under organic manure application 20 (Fig. 5). Soil moisture content did not respond to the cumulative CO 2 emission in our study, probably because the soil seldom underwent prolonged drought 28 . Similar results also reported by Sheng et al. (2010). Smith et al. reported that the release of CO 2 by aerobic respiration is primarily driven by soil temperature, but becomes moisture-dependent as soil dries out 32 . Among the three TSD treatments, undisturbed plot exhibited the largest cumulative N 2 O emission under N-fertilizer application. This can possibly attribute to the greater SOC and total N content in 0-20 cm soil layer depth (Table 2 and Fig. 5), as N 2 O emission depended on SOC and total nitrogen contents, bulk density, clay fraction and soil moisture content that regulates N 2 O production via microbial de-nitrification as well as nitrification 33 . The second largest cumulative N 2 O emission was at the eroded site under N-fertilizer, which might due to its higher soil temperature 32 and greater soil bulk density (Table 2, Figs 2 and 4). The results are in good line with previous studies conducted with an intensively farmed organic soil in North Central Ohio, which reported that N 2 O flux was positively related to soil temperature and CO 2 flux 34 . The least cumulative N 2 O emission at the depositional site may be determined by its lower soil temperature and lower SOC, total N content (Table 2, Figs 2 and 5). In addition, the eroded site displayed the highest cumulative N 2 O emission under organic manure application, probably ascribed to the anaerobic conditions 35 in soil pore space caused by higher soil moisture content and greater soil bulk density (Figs 2 and 4). The result is in agreement with a previous report that N 2 O emissions as a result of denitrification occurred in compaction treatment 36 . Flessa et al., reported that total N 2 O emission from the potato field during the growing season was 2.0 (kg N ha −1 growing season −1 ) in 1998 19 , and Kumar et al., resulted 2.89 (kg N ha −1 year −1 ) of N 2 O emission from corn-corn cropping rotation in an Alfisol of Ohio in 2012 26 , which were similar to our results of average cumulative N 2 O emission (2.9 kg N ha −1 growing season −1 ) in corn field in 2012. In this study, cumulative CH 4 emission was positive at the eroded site under organic manure application, which was probably due to the hypoxic conditions 35 that may have been caused by the relatively greater soil moisture (Fig. 2). Cumulative CH 4 emissions from other TSD treatments were negative, which indicated the net CH 4 sink during the growing season (Table 1). The low values and net uptake of CH 4 reported in our study were consistent with those reported for cultivated soils 25,37 In our study, the average GWP at the eroded site (18.07 Mg CO 2 Eq ha −1 growing season −1 ) generated roughly the same value with the undisturbed control (18.11 Mg CO 2 Eq ha −1 growing season −1 ), but declined dramatically at the depositional site (10.58 Mg CO 2 Eq ha −1 growing season −1 ). The cumulative GHG emission at the eroded site might be stimulated via the higher soil temperature caused by its less aboveground coverage shade, but limited by its lower available C substrate. The cumulative GHG emission at the depositional site declined because of its lower soil temperature and lower available C substrate in surface. In addition, the cumulative GHG emission may also be influenced by soil bulk density and soil moisture content. The results displayed that eroded site can neither play a net sink nor net source of greenhouse gases emission, while depositional site can be a net sink of GHG emission. It is critical to consider the overall (net) effect of the erosion and deposition processes in comparison with undisturbed soil (the control). Therefore, in this paper, we compared important parameters for undisturbed control with the average values observed for the combination of the topsoil removal and topsoil addition conditions. The average value of GWP of eroded and depositional site was 21% lower than that of the undisturbed control, indicated the erosion-deposition process could be net sink of GHG emission. Average GWP for soil receiving organic manure was increased by 15%, compared with that with N-fertilizer application. This might be due to the greater SOC and total N content and the higher soil moisture content after organic manure application (Figs 2 and 5). Given the limited accessibility to the experimental field, gas fluxes were only collected once every 2 weeks. There might be other peaks in GHG fluxes that were not captured during our sampling regime. More frequent sampling intervals are highly recommended in the further study. In our study, the average maize yield significantly decreased by 51% at eroded site, and significantly increased by 47% at depositional site compared with the undisturbed control (7.45 Mg ha −1 ) (P < 0.0001), which was coincide with maize yields on two Alfisols in central Ohio 38 , and consistent with the finding that the corn yield declined by nearly half (46%) on removal of 20 cm of the topsoil in an eroded farmland of Chinese Mollisols 16 . Crop yield usually adversely affected by the impedance of root growth, water and nutrient deficits, high bulk density, penetrometer resistance, and low field moisture capacity under erosion 10,11,16 . On the one hand, topsoil removal significantly declined the SOC and total N content, increased the soil bulk density, and further reduced the maize yield at the eroded site. On the other hand, the topsoil addition buried the former topsoil under plough depth, preserved SOC from decomposition and mineralization 39 , promoted deep root growth, and consequently resulted to greater crop productivity (Figs 4, 5 and 6). This result agree with the findings that the impacts of erosion on agricultural land are usually negative for eroded sites and may be positive for depositional sites 40,41 . The average value of maize yield of eroded and depositional site declined by 2% compared with the undisturbed control (7.45 Mg ha −1 ), which indicate the erosion-deposition process resulted the equivalent production with the uneroded site. Maize yield at the organic manure applied plot (7.97 Mg ha −1 ) increased by 18% compared with that from the plot receiving N-fertilizer (6.74 Mg ha −1 ). This was probably caused by the improved SOC, total N content and soil moisture content with organic manure application (Figs 2 and 5). The average value of GHGI was 4.96, 2.43 and 0.96 Mg CO 2 Eq Mg −1 grain yield growing season −1 , respectively for eroded site, undisturbed control and depositional site. The eroded site enhanced the GHGI because of its lower grain yield, while the depositional site declined the GHGI due to its lower GWP and higher grain yield. The average GHGI of erosion-depositional site increased by 22% compared to the undisturbed control. The average GHGI for soil receiving organic manure exhibited a small reduction of 3% compared with that of N-fertilizer application. In summary, our results displayed that eroded site can neither play a net sink nor net source of greenhouse gases emission, while depositional site can be a net sink of GHG emission. Eroded site significantly reduced maize yield, while depositional site significantly enhanced the maize yield. The erosion-deposition process declined the GWP, not changed maize yield, and increased the GHGI. Soil with organic manure application enhanced GWP, improved maize yield and slightly reduced GHGI compared with soil receiving N-fertilizer. It was worthwhile to note that our results were merely concluded based on the assumption that the area of erosion equals the area of deposition. However, in natural field, eros ion tends to be dissipated over much wider area, while eroded materials often end in areas not suitable for crop growth (river beds, estuaries, etc.) or water bodies (wetlands, reservoirs, ocean). The weight of GHG emissions from eroded area could have been much larger, and depositional zone may have led to additional CH 4 emissions. This calls for systematic investigation in the future study. ## Methods Study area. The study was conducted in an on-going long-term experiment at Waterman Farm of the Ohio State University, Columbus, OH, USA (N + 40° 1′ 5.52″ E −83° 2′ 29.72″). The experiment was initiated in 1997 on the Crosby soil series (deep, fine, mixed, active, mesic, Aeric Epiaqualf). The deep soil developed on nearly level topography (0% to 2% slopes) is of silt loam texture, poorly drained and derived from glacial till. The mean annual rainfall is 1016 mm and the mean annual air temperature is 11 °C 8 . The experiments were designed in a split-plot arrangement with completely randomized blocks. Three TSD levels were carried out as main plots and two amendment types as subplots. The 18 × 9 m main plots were subdivided into 6 × 4.5 m subplots, with three replications for each treatment combination. The main plots were separated by 2.7 m long border strips 8,38 . Three TSD levels created once at the beginning of the experiment to simulated soil erosion-deposition process, which were: (1) topsoil removal (eroded site) created by physically remove of 20 cm topsoil with a landscape loader; (2) undisturbed control (uneroded site); and (3) topsoil addition (depositional site) achieved by deposit of 20 cm topsoil on soil surface. Two amendments were applied in this study: N-fertilizer and organic manure. For the plots receiving N-fertilizer, 150 kg N ha −1 urea-ammonium nitrate (28% N) was side-banded on the soil surface at the 3 rd to 4 th leaf stage of corn growth. For the plots receiving organic manure, dry matter compost (20 Mg ha −1 ) was uniformly top dressed during April each year. CO 2 , N 2 O and CH 4 flux measurements. Soil-air samples for the assessment of CO 2 , N 2 O and CH 4 fluxes were collected using the static chamber method 42 . Gas chambers were made of polyvinyl chloride (PVC) pipes of 15 cm diameter and 30 cm length. The top lid was made of a PVC cap, and the lower end was trimmed to be inserted into the soil. A machine-trimmed PVC trough was coupled around the outer ring of the pipe, approximately 5 cm from the top. The PVC cap was equipped with a sampling port and a rubber septum on the top, and the cap bottom could be fitted into the trough when the cap was in place 42,43 . The chambers were inserted 10 cm deep into the ground at each sampling point, with three replications for each treatment. Chambers were installed 1 month before gas sampling, and the chambers remained in place with the cap opened during the entire growing season, except for temporary removal during seeding or fertilizer application. Chambers were reinstalled in the same place immediately after the fertilizer and seeding operations were completed 34 . When sampling, closed the chamber lid, taken approximately 10 cm 3 soil-air samples from each chamber headspace at 0 and 30 minutes, and transferred it to crimp sealed pre-evacuated 10 ml vials fitted with butyl rubber septa. The vials were evacuated to a pressure of −172 kPa and prepared 1 day before sampling. Soil-air samples were obtained between 11 AM and 2 PM when fluxes were expected to be maximal 43 biweekly during the entire growing season. Three replications were taken for each treatment. The CO 2 and CH 4 in soil-air samples were analyzed using a GC-2014 gas chromatograph (GC; Shimadzu, Kyoto, Japan) equipped with a thermal conductivity detector for CO 2 , and a flame ionization detector for CH 4 . N 2 O was analyzed on a GC fitted with a 63 Ni electron capture detector 34 . Soil temperature and moisture measurements. Soil samples for measurements of soil temperature and soil moisture content were collected biweekly in conjunction with soil-air sampling 18 . Soil temperatures at 10 cm soil depth were monitored by using a digital thermometer near each chamber simultaneously with gas sampling. Gravimetric soil moisture content was also determined by collecting soil samples close to the chambers at 0-10 cm depth. Analysis of soil properties. Bulk and intact core samples were obtained separately in June 2012 to measure soil properties. Intact core soil samples for bulk density analysis were collected at 0-40 cm depth (10 cm intervals) using a manually-driven core sampler with diameter and height both 5 cm. Gravimetric soil moisture content (SMC) was measured by drying a portion of trimmed core samples at 105 °C for 24 h 19 . Wet bulk density was computed as the ratio of soil wet weight to core volume, and soil bulk density was calculated from the wet bulk density and soil moisture content, ρ b = ρ b ′/(1 + w), where ρ b is soil bulk density, ρ b ′ is soil wet bulk density, and w is the gravimetric moisture content. Total porosity was calculated from the equation 44 : f = 1 − (ρ b ′/ρ s ), where ρ s is the soil particle density and is estimated at 2.65 g cm −3 . Bulk soil samples were air-dried at room temperature, ground with a wooden hammer, and sieved through a 2 mm sieve before physical and chemical analysis 38 . Soil total C and N contents were analyzed by the dry combustion method using a vario Max CN analyzer (Elementar, Hanau, Germany) 18 . The SOC was assumed to be equal to the total C as inorganic C concentration was negligible with the soil pH was below 7 38 . Crop yield. Corn (Zea mays L.) was grown from about mid-May to October in 2012 without any major disturbances, and no tillage operation was performed. Corn plants from the center two rows of each plot were hand harvested. Crop residue after the harvest was left on the soil surface. Corn ears were separated from the stover and weighed. Corn ears were shelled, and grains were weighed separately from other parts of the ear after air drying. Subsamples of grain were weighed and then oven-dried at 60 °C for 48 h to determine the water content 38 . Grain yields are reported in Mg ha -1 at 12% moisture content. Data calculations and statistical analysis. Daily gas fluxes (q) (in units of g CO 2 -C m −2 d −1 or mg N 2 O − N m −2 d −1 or mg CH 4 − C m −2 d −1 ) were computed using Eq. (1) 34,42 : The greenhouse gas intensity (GHGI) was calculated by dividing GWP by crop yield using Eq. ( 4) 46 : = − GHGI GWP/grain yield(Mg CO Eq Mg grain yield growing season ) (4) Statistical analysis was performed using the analysis of variance (ANOVA) procedure available in SAS 8.01 for Windows (1999-2000, SAS Institute Inc., Cary, NC, USA). Mean and interactive effects of treatments were separated using the F-protected least significant difference test. The probability level (P) chosen to designate significance was ≤0.05. Correlation and regression analyses were performed on selected variables at P ≤ 0.1 using the same package.

chemsum

{"title": "Impacts of simulated erosion and soil amendments on greenhouse gas fluxes and maize yield in Miamian soil of central Ohio", "journal": "Scientific Reports - Nature"}

isomeric_anthracene_diimide_polymers

## Abstract: N-type semiconducting polymers are attractive for organic electronics, but desirable electron-deficient units for synthesizing such polymers are still lacking. As a cousin of rylene diimides such as naphthalene diimide (NDI) and perylene diimide (PDI), anthracene diimide (ADI) is a promising candidate; its polymers, however, have not been achieved yet because of synthetic challenges for its polymerizable monomers.Herein, we present ingenious synthesis of two dibromide ADI monomers with dibromination at differently symmetrical positions of the ADI core, which are further employed to construct ADI polymers.More interestingly, the two obtained ADI polymers possess the same main-chain and alkyl-chain structures but different backbone conformations owing to varied linking positions between repeating units. This feature enables their different optoelectronic properties and film-state packing behavior. The ADI polymers offer first examples of conjugated polymer conformational isomers and are highly promising as a new class of n-type semiconductors for various organic electronics applications. ## Introduction N-type organic semiconductors, in particular conjugated polymers, are very crucial for optoelectronic devices, but their development lags far behind that of p-type counterparts due to the lack of electron-defcient building blocks. Among a handful of electron-defcient units, six-membered tetracarboxylic aromatic diimides, typically naphthalene diimide (NDI) and perylene diimide (PDI), attract enormous attention because of their high electron affinity and mobility. A plenty of NDI-and PDI-based polymers have been developed as n-type semiconductors for various organic electronics applications. 3, The aromatic diimides can be categorized into rylene diimides and acene diimides according to the p-conjugation core. 17 The NDI and PDI are just representatives of rylene diimides having extended conjugations along the normal axis of the NDI scaffold (Fig. 1). More derivatives e.g. terrylene and quaterrylene diimides were investigated by Müllen, Langhals 21 and Adachi 22 et al. However, they are not suitable for constructing polymers because of high steric hindrance at the longitudinal direction which is adverse to the charge transport and molecular crystallinity, as claimed for PDI polymers. 23 This issue is absent in another class of aromatic diimides, namely acene diimides with conjugation expansion along the equatorial axis of the NDI, yet their polymers cannot be achieved so far owing to synthetic difficulties. Wang et al. 24 and Yamada et al. 25 have performed pioneering studies on the synthesis of acene diimides such as anthracene-, tetracene-, pentacene-and hexacene-diimides. Among them, the anthracene diimide (ADI) with one ring extension relative to the NDI should possess similar electron affinity and potentially provide interesting optoelectronic properties. 26 Moreover, its polymers have been theoretically predicted to present excellent n-type characteristics for organic electronics. 27 However, access to the ADI unit and especially to its polymers is rather challenging. Although Wudl et al. 28 and Yamada et al. 25 separately reported short alkyl chain-substituted ADIs synthesized from different routes, ADI-based polymers were never materialized since synthetic chemists were plagued by challenges in synthesizing polymerizable monomers of ADI. Long alkyl chains attached to amide N atoms are, on the one hand, indispensable for ensuring the solubility of resultant polymers; on the other hand, functionalized groups on the anthracene such as halogens, boronic esters, or tin salts (bromines usually preferred) are needed for the cross-coupling polymerization. Unfortunately, precursor compounds with long alkyl chains are not easily accessible, and symmetrical dibromination of ADI obviously suffers from great complexity owing to the presence of several pairs of substitutable sites (2,6-, 3,7-, and 4,8-positions, Fig. 1). Establishing new synthetic protocols to tackle these issues is thus of critical signifcance. Herein, we report the synthesis and properties of two 2octyldodecyl-substituted dibromide ADI monomers and their derived polymers. The two monomers are dibrominated at 2,6and 3,7-positions (2,6-2Br-ADI and 3,7-2Br-ADI, Fig. 1), respectively, by different strategies, making the two obtained polymers (PADI-2,6-2T and PADI-3,7-2T) possess isomeric backbone conformations. The two polymer conformers, thereby, present noticeably different molecular confgurations, optoelectronic properties, as well as packing and oriented behavior in the flm state. To the best of our knowledge, this is the frst report on ADI-based polymers with backbone isomerism showing potential as a new class of promising n-type semiconducting materials. ## Results and discussion The two monomers are synthesized with pre-and postbromination approaches, respectively. The 3,7-2Br-ADI is synthesized by a post-bromination route (Scheme 1a), where the 3,7-dibromination is conducted after obtaining the 2octyldodecyl-substituted ADI (6a). The ADI 6 is acquired via a Lewis acid-mediated reaction between acyl chloride 3 and alkyl isocyanate 5 (for their synthesis see the ESI †). This approach was initially reported by Yamada et al. using bismuth triflate as a Lewis acid; 25 however, the procedure in the literature did not work for our reaction. By replacing the bismuth triflate with a classic Lewis acid of AlCl 3 and swapping the addition sequence of Lewis acid and isocyanate, the reaction is successfully realized to give the ADI 6. When the substituent is 2-octyldodecyl, 6a is produced with a low yield (<10%, two steps from 2 to 6a). Instead of branched alkyl chains, linear n-hexyl and n-dodecyl are employed to synthesize ADI 6b and 6c with improved yields. It is worthy of note that the yield of 6c with longer alkyl chains exceeds 50%, higher than that ($20%) of 6b. Based on these observations and the fact of adding AlCl 3 prior to isocyanate, we propose a mechanism of Lewis acid-mediated successive two-step Friedel-Crafts-type reaction (Scheme 2). First, the AlCl 3 coordinates with the acyl chloride 3 to generate an electrophile carbocation 12. Then, with the addition of isocyanate, nucleophilic attack to 12, namely the intermolecular Friedel-Crafts-type amidation, occurs to form intermediate 13, which undergoes the intramolecular Friedel-Crafts acylation to offer the double-cyclized ADI. For the frst-step Friedel-Crafts reaction, the nucleophilic attack can be considerably impacted by the alkyl chain of isocyanate. Therefore, the steric hindrance of branched 2-octyldodecyl results in a low yield of 6a, and the yields of linear alkyl chain-substituted 6b and 6c are much higher. Meanwhile, long linear alkyl chains can provide better solubility of intermediates 13 and 14, allowing for efficient second-step intramolecular Friedel-Crafts acylation, by which the 6c containing the n-dodecyl is synthesized in a higher yield than the n-hexyl-substituted 6b. The proposed mechanism suggests that the ADI and its various derivatives could be obtained in a desirable yield via tuning side chains. In the present work, in order to ensure the solubility of the resulting polymers and to compare with classic NDI and PDI analogs, the 2octyldodecyl-substituted 6a is used to prepare its dibromide monomer. After careful optimization (Table S1 †), it is selectively dibrominated at 3,7-positions with a brominating agent, i.e. 5,5dimethyl-1,3-dibromohydantoin (DBH), in conc. H 2 SO 4 at a mild temperature of 65 C to produce the 3,7-2Br-ADI. On the other hand, the 2,6-2Br-ADI is synthesized by a pre-bromination route (Scheme 1b), where the 2,6-dibromination prior to annulation of diimides is performed for the starting material (1) to get tetrabromide anthracene 7 that follows similar acidation, acylation, and Friedel-Crafts-type reactions to yield the target monomer. The molecular structures of two isomeric monomers are unambiguously confrmed by 1 H NMR spectroscopy (Fig. 2) and high-resolution mass spectrometry (see the ESI †). Indicated by density functional theory (DFT) calculations (Fig. 3a-c), the ADI skeleton is highly coplanar and the 3,7dibromination does not vary such a molecular confguration. Interestingly, the 2,6-dibromination causes a small distortion of the ADI plane, most likely owing to the steric hindrance between the adjacent bromine and the carbonyl group. UV-Vis absorption spectra (Fig. 4a) display that the absorption bands of the two monomers locate in between those of dibromide NDI and PDI. Notably, relative to the absorption of the ADI 6a (Fig. S1a †), the 3,7-2Br-ADI exhibits an evident red shift, while the 2,6-2Br-ADI shows a blue shift. Such observations can be ascribed to the effect of different extensions of frontier molecular orbitals (Fig. S2 †), induced by bromine substitutions at different positions. These results suggest that dibromination positions impact on the molecular geometry, the frontier Scheme 2 Proposed mechanism of the Lewis acid-mediated twostep Friedel-Crafts-type reaction. Fig. 2 1 H NMR spectra of ADI 6a and two dibrominated isomers. The two monomers were then copolymerized with distannyl bithiophene to synthesize two polymers, PADI-3,7-2T and PADI-2,6-2T (Scheme 1c). Both polymers exhibit good solubility in chloroform, toluene, chlorobenzene, etc. at room temperature. Their molecular weights/polydispersity index determined by GPC are 17.8 kDa/2.8 and 14.2 kDa/2.3 (Table 1). Thermogravimetric analysis reveals their excellent thermal stability and the differential scanning calorimetry plot shows no detectable thermal transitions (Fig. S3 †). The DFT simulations demonstrate that the backbone conformations of the two polymers vary profoundly. The PADI-3,7-2T shows a rigid and planar polymer backbone, with small dihedral angles between ADI and bithiophene units (Fig. 3d), whereas the PADI-2,6-2T main chains are more distorted (Fig. 3e). In view of their identical polymer structures, the isomeric backbone conformations are bound to stem from the different linking positions between repeating units and can induce diversity in various properties of the two polymers. In toluene solution, the absorption maximum (l max , 740 nm) of PADI-3,7-2T manifests a bathochromic shift of 40 nm compared with that (700 nm) of PADI-2,6-2T (Fig. 4b, c). In the flm state, the PADI-3,7-2T presents a similar absorption spectrum to that in solution (Fig. 4b), while the PADI-2,6-2T exhibits a red-shifted one compared to its solution absorption (Fig. 4c). These results reflect that both polymers with the same repeating units but different linkage modes afford varied optical properties and aggregation behavior. The LUMO/HOMO levels (Fig. S5 †) measured from CV are 4.02/5.43 and 4.06/ 5.44 eV for PADI-3,7-2T and PADI-2,6-2T, respectively. The deep LUMO levels similar to those of NDI and PDI polymers suggest strong electron affinity of ADI polymers with potential as n-type semiconductors. 23,29 The effect of the different backbone linkage modes on molecular packing and orientation in flms was investigated by grazing incidence X-ray diffraction (GIXD). Both pristine flms show a (010) diffraction ring (Fig. 5), indicating their mixed face-on and edge-on orientation; however, the PADI-3,7-2T flm prefers the face-on one. This packing structure indicates the coexistence of parallel and vertical charge transportation channels in its flms. 30 After thermal annealing (TA), the peak intensities become signifcantly stronger. It is clear that the PADI-2,6-2T flm manifests distinct (h00) lamellar diffraction peaks (q z ¼ h 0.27 A 1 , h ¼ 1-4) with large crystal coherence lengths (CCLs) (Table S2 †) in the out-of-plane direction, signifying a high degree of molecular ordering. In contrast, no signals implying ordered lamellar stacking are found for the PADI-3,7-2T flm. Since the two polymers have the same backbone structures and alkyl chains, their difference in lamellar stacking can be attributed to isomeric backbone conformationinduced disparity of side-chain ordering. ## Conclusions In conclusion, we have presented the synthesis of two dibromide 2-octyldodecyl-substituted ADI with elaborate pre-and post-bromination methods, allowing for the exploration of ADI polymers. The dibromination at different positions of the ADI unit endows two monomers with variable coplanarity which further leads to isomeric backbone conformations of the resultant polymers even though they have the same molecular structures. With the unique conformational isomerism, the PADI-2,6-2T and PADI-3,7-2T exhibit distinct optical properties and packing behavior. The advent of ADI polymers with deep LUMO levels enriches the family of aromatic diimide polymers, which are promising for electron-carrying optoelectronic devices. ## Conflicts of interest There are no conflicts to declare.

chemsum

{"title": "Isomeric anthracene diimide polymers", "journal": "Royal Society of Chemistry (RSC)"}