en
stringclasses
8 values
ru
stringclasses
8 values
## 1 Introduction The experimental evidence for neutrino mass, acquired in recent decades, provides concrete evidence for physics beyond the Standard Model (SM) (see e.g. [1]). Although the requisite new degrees of freedom cannot yet be determined it is clear that additional particles are likely to exist, in order to generate the masses. Similarly, there is by now a large amount of evidence for an additional galactic constituent, an unknown substance referred to as dark matter (see e.g. [2]). This may or may not require new degree’s of freedom, but the hypothesis that the dark matter is comprised of a stable (or long lived) new particle species provides a simple explanation for this observed feature of the Universe. Given that these two indicators for beyond-SM physics can be explained by extending the particle spectrum of the SM, it is natural to ask if the requisite new particles can be related. Could the mechanism of neutrino mass be related to the existence of a stable dark-matter candidate? A particularly simple model realizing this idea was proposed by Ma in 2006 [3]. This model extends the SM to include an additional SM-like scalar doublet and gauge-singlet fermions, all of which are odd under a discrete \(Z_{2}\) symmetry. The extended field content allows for radiative neutrino mass, generated at the one-loop level, while the lightest beyond-SM field is absolutely stable. One thus arrives at a simple synergetic model of radiative neutrino mass and dark matter. In this work we generalize Ma’s approach. We present a class of related models, all of which generate neutrino mass via a loop diagram with the same topography as Ma’s, whilst simultaneously admitting stable dark-matter candidates. The loop-diagram employed by Ma contains a mass insertion on the internal fermion line (see Figure 1), and our generalizations fall naturally into two categories; those which break lepton-number symmetry via a Majorana mass insertion, and those with a Dirac mass insertion, such that lepton-number symmetry is broken at a vertex. Although the basic mechanism is very similar in both cases, this difference modifies one’s expectations for the beyond-SM field content and the associated phenomenology. It turns out that, in both cases, this approach is very general and many realizations are possible. However, restricting attention to models in which the beyond-SM multiplets are no larger than the adjoint representation significantly reduces the possibilities. As we shall see, there are only two such (minimal) models with a Majorana mass-insertion, both of which are known [3, 4]. We find seven additional models with a mass insertion of the Dirac type, all of which achieve radiative neutrino mass and dark-matter candidates. We detail these models, finding a subset that are compatible with direct-detection experiments. There are cases with an inert singlet, an inert doublet, and an inert triplet; the models therefore provide a natural setting for inert \(N\)-tuplet theories of dark matter, such that radiative neutrino mass is also achieved. Interestingly, three of the new models admit a simple extension that can explain the origin of the (formerly imposed) discrete symmetry. By upgrading the discrete symmetry to a gauged \(U(1)\) symmetry, and extending the field content by a single SM-singlet scalar, the discrete symmetry can arise as an accidental symmetry of the low-energy Lagrangian, after \(U(1)\) symmetry breaking takes place. This provides a simple explanation for the discrete symmetry. Though we focus on models with representations no larger than the adjoint, we also briefly discuss cases where the beyond-SM fields can be quadruplet and/or quintuplet representations of \(SU(2)_{L}\). We present the candidate models in these cases, and mention some key issues, based on the lessons learned from our preceding studies. Despite the use of larger multiplets, these models can still be of interest; in addition to allowing for radiative neutrino mass, the exotics with larger electric-charges in these multiplets can enhance the \(2\gamma\) and/or \(\gamma+Z\) signal from dark-matter annihilation when they appear inside loops [5]. This can provide a simple connection between the mechanism of neutrino mass and the astrophysical gamma-ray signal [6]. Before proceeding we note that the connection between radiative neutrino mass and dark matter has been explored in a number of different models, including Ref. [7], which precedes Ma’s work; for other examples see [8]. Previous works on inert-singlet models [9], inert-doublet models [10], and inert-triplet models [11] are also well known. Additional relevant works dealing with inert-multiplet dark matter and/or radiative neutrino mass are cited in the text. Note also that Refs. [12, 13] have detailed the one-loop realizations of the \(d=5\) operator for neutrino mass. Inert scalar dark matter can also help cure the little hierarchy problem found in low-scale seesaws [14]. The present work follows on from the generalized tree-level seesaws presented in Ref. [15]. The plan of this paper is as follows. In Section 2 we discuss Ma’s model and present the generalizations that similarly achieve radiative neutrino mass and dark-matter candidates. Section 3 considers the case of fermionic dark-matter, while Section 4 discusses scalar dark-matter. One of the generalized models is presented in more detail in Section 5. In Section 6 we show that some of the models allow a simple extension, such that the discrete symmetry appears as an accidental symmetry in the low-energy theory. Models with exotics forming larger \(SU(2)\) representations are discussed in Section 7 (and explicitly displayed in the Appendix). We conclude in Section 8. [FIGURE:S1.F1][ENDFIGURE] ## 2 Radiative Neutrino Mass and Dark Matter We are interested in the class of models that generate neutrino mass radiatively by the diagram in Figure 1. Here \(\mathcal{F}\) is a beyond-SM fermion and \(S_{1,2}\) are new scalars (which can be identical in some cases). A basic feature of this diagram is that the three vertices can all involve two beyond-SM fields. Consequently one can always consider a discrete \(Z_{2}\) symmetry whose action on the beyond-SM multiplets is \[\{\mathcal{F},\;S_{1},\;S_{2}\} \rightarrow -\ \{\mathcal{F},\;S_{1},\;S_{2}\}\;,\] (1) while all SM fields transform trivially. The lightest field within the multiplets \(\mathcal{F}\) and \(S_{1,2}\) will thus be stable, and provided this field is electrically neutral and colorless, one arrives at a dark-matter candidate. These comments are generic for all models of this type; the connection between loop masses and dark matter is simple to realize in this class of models. Figure 1 produces Majorana neutrino masses, so the loop diagram must contain a source of lepton number violation. Choosing the convention for lepton-number symmetry such that the new fermion \(\mathcal{F}\) has the same value as the SM leptons, there are two ways to explicitly break lepton-number symmetry; it can be broken at either the mass insertion or at one of the vertices. The simplest models, in terms of the requisite number of beyond-SM multiplets, are those with a lepton number violating (Majorana) mass insertion. In this case one has \(\mathcal{F}_{L}\equiv\mathcal{F}_{R}^{c}\) and minimal cases occur for \(S_{1}=S_{2}\equiv S\). Thus, only two beyond-SM multiplets are required. The general loop-diagram for this subset of models is given in Figure 2. We consider this case first. ### Models with a Majorana Mass Insertion We seek models that achieve neutrino mass via Figure 2 and give rise to dark-matter candidates. Clearly the fermion must form a real representation of the SM gauge symmetry, \(\mathcal{F}_{R}\sim(1,R_{\mathcal{F}},0)\), in order to allow a bare Majorana mass. As we are considering dark-matter candidates we do not consider colored fields. Note also that \(R_{\mathcal{F}}\) must be odd-valued to ensure there is no fractionally charged particles (the lightest of which would be stable and thus cosmologically excluded). Odd-valued \(R_{\mathcal{F}}\) also ensures that \(\mathcal{F}_{R}\) contains an electrically neutral component, so no additional constraint is imposed by this demand. With this information one can obtain the viable combinations of \(\mathcal{F}_{R}\) and \(S\) that generate Figure 2. The basic Lagrangian terms are \[\mathcal{L} \supset i\bar{\mathcal{F}_{R}}\gamma^{\mu}D_{\mu}\mathcal{F}_{R}-\frac{M _{\mathcal{F}}}{2}\;\overline{\mathcal{F}_{R}^{c}}\mathcal{F}_{R}+|D^{\mu}S|^{ 2}-M_{S}^{2}|S|^{2}+\lambda\bar{L}\tilde{S}\mathcal{F}_{R}+\lambda_{\text{ \tiny SH}}(S^{\dagger}H)^{2}+\mathrm{H.c.},\] (2) where \(L\) (\(H\)) is the SM lepton (scalar) doublet and \(\tilde{S}\) denotes the charge-conjugate of \(S\). It turns out that the possible combinations for \(\mathcal{F}\) and \(S\) are not restricted by quantum numbers; one can consider increasingly large multiplets, presumably up to some unitarity limits [16], and realize a model with Figure 2 and a dark-matter candidate. However, if we restrict our attention to models with \(R_{\mathcal{F}},R_{S}\leq 3\), such that no new multiplet is larger than the adjoint representation, there are only two possibilities. The first case is Ma’s original proposal, which employs an additional (inert) scalar doublet \(S\sim(1,2,1)\), and a gauge-singlet fermion \(\mathcal{F}_{R}\sim(1,1,0)\)[3]. This model is the prototype for the class we are considering. The second model also employs the scalar doublet \(S\sim(1,2,1)\), but instead utilizes the triplet fermion \(\mathcal{F}_{R}\sim(1,3,0)\)[4], familiar from the Type-III seesaw [17]. Thus, both of the models are known in the literature, and there are no additional possibilities unless one considers larger multiplets. In each of these models the dark-matter candidate can be a neutral component of \(S\) or a Majorana fermion. There is, however, an important difference between the singlet case, \(\mathcal{F}\sim(1,1,0)\), and the other model; the singlet does not participate in weak interactions and is therefore brought into thermal contact with the SM sector via the Yukawa coupling. For fermionic dark-matter, this can produce conflict between the need to keep the Yukawa coupling large to ensure thermal dark-matter, and the need to suppress the Yukawa coupling to limit the size of flavor changing effects. This issue does not arise in the triplet fermion model, as the fermions can maintain equilibrium with the SM sector via weak interactions in these cases, even if the Yukawa couplings are suppressed. For an analysis of Ma’s model, incorporating recent LHC data on the Higgs, see e.g. Ref. [18]. Also note that loop effects can induce observable interactions between dark matter and experimental detectors in Ma’s model [19]. [FIGURE:S2.F2][ENDFIGURE] ### Models with a Dirac Mass Insertion Having exhausted the minimal models with a lepton-number violating mass insertion, we now consider models with a Dirac mass insertion; i.e. the beyond-SM fermion has nonzero hypercharge. In this case the general mass-diagram has the form shown in Figure 1. The fields \(\mathcal{F}_{R}\) and \(\mathcal{F}_{L}\) are no longer related by charge conjugation, so the mass insertion is of the Dirac type and \(\mathcal{F}\) is a vector-like fermion. In addition one requires \(S_{1}\neq S_{2}\). We again consider a \(Z_{2}\) symmetry under which the SM fields transform trivially but the new fields are odd. The Lagrangian contains the following pertinent terms: \[\mathcal{L} \supset i\bar{\mathcal{F}}\gamma^{\mu}D_{\mu}\mathcal{F}\ -\ M_{\mathcal {F}}\;\overline{\mathcal{F}}\mathcal{F}\ +\ \sum_{i=1,2}\left\{|D^{\mu}S_{i}|^ {2}\ -\ M_{i}^{2}|S_{i}|^{2}\right\}\] (3) \[+\ \lambda_{1}\;\bar{L}\mathcal{F}_{R}S_{1}\ +\ \lambda_{2}\;\bar {L}\mathcal{F}^{c}_{L}\tilde{S}_{2}\ +\ \lambda_{\text{\tiny SH}}\;S_{1}\tilde {S}_{2}H^{2}+\mathrm{H.c.},\] where, in our convention, lepton number symmetry is broken by a Yukawa coupling. With the \(Z_{2}\) symmetry present, there are no terms in the scalar potential that are linear in _just one_ of the new scalars \(S_{1,2}\). Therefore the beyond-SM scalars do not acquire an induced VEV and parameter space exists for which \(\langle S_{1,2}\rangle=0\), so the \(Z_{2}\) symmetry remains exact. In selecting viable multiplets one must ensure that no new multiplet contains a fractionally charged field, to avoid a (cosmologically excluded) stable charged field. To ensure that the lightest \(Z_{2}\)-odd field is a neutral dark-matter candidate one must demand that the new multiplets contain at least one neutral field. Note that the neutral field does not have to appear as an explicit propagating degree of freedom inside the loop diagram; it is sufficient merely that the loop-diagram exists and that the particle content includes a neutral field that can play the role of dark matter. With these conditions in mind we search for viable combinations of the beyond-SM multiplets that realize Figure 1. We find that the size of the beyond-SM multiplets is not restricted by our demands; one can consider increasingly large multiplets and realize the loop diagram. However, if attention is restricted to models in which none of the beyond-SM multiplets are larger than the adjoint representation, only seven distinct models are found. These are listed in Table 1. Of the seven models, one employs the exotic lepton triplet \(\mathcal{F}\sim(1,3,-2)\), studied in Refs. [20, 21], three contain an exotic vector-like (and SM-like) lepton doublet \(\mathcal{F}\sim(1,2,-1)\)[21, 22, 23], and two contain a charged lepton doublet \(\mathcal{F}\sim(1,2,-3)\)[21, 24]. There is also a model with a SM-like charged singlet fermion, \(\mathcal{F}\sim(1,1,-2)\), which already appeared in Ref. [25]. Neutrino masses take a standard calculable form in these models. For example, in models \((A)\) and \((B)\) only singly-charged exotics propagate in the loop, and the SM neutrino mass matrix is given by³ [FOOTNOTE:3][ENDFOOTNOTE] \[(\mathcal{M}_{\nu})_{\alpha\beta} \simeq \frac{\left[(\lambda_{2}^{*})^{a}_{\alpha}(\lambda_{1}^{*})^{a}_{ \beta}+(\lambda_{2}^{*})^{a}_{\beta}(\lambda_{1}^{*})^{a}_{\alpha}\right]}{32 \pi^{2}}\,\frac{\lambda_{\text{\tiny SH}}\langle H\rangle^{2}}{M_{>}^{2}-M_{<} ^{2}}\left[\frac{M^{2}_{>}\,M_{\mathcal{F},a}}{M_{\mathcal{F},a}^{2}-M_{>}^{2} }\,\log\frac{M_{\mathcal{F},a}^{2}}{M_{>}^{2}}\ -\ (M_{>}\to M_{<})\right]\] Here \(M_{\mathcal{F},a}\) is the mass for the charged component of the exotic fermion \(\mathcal{F}_{a}\), and summation is implied for the repeated index \(a\) (which labels the exotic-fermion generations).⁴ The masses \(M_{>,<}\) refer to the charged scalar mass-eigenstates, which are linear combinations of the charged scalars \(S_{1}^{+}\) and \(S_{2}^{+}\). The mixing results from the \(\lambda_{\text{\tiny SH}}\)-term in Eq. (3), which takes the explicit form \(\lambda_{\text{\tiny SH}}\tilde{H}^{\dagger}S_{1}S_{2}^{\dagger}H\subset \mathcal{L}\), for models \((A)\) and \((B)\). If all the exotics are at the TeV scale one requires dimensionless couplings of \(\mathcal{O}(10^{-3})\) to obtain \(m_{\nu}\sim 0.1\) eV. The scenario with all exotics at the TeV scale is most interesting from a phenomenological perspective. However, strictly speaking one only requires the lightest exotic to have a mass of \(\lesssim\mathcal{O}(\mathrm{TeV})\) in order to realize a dark-matter candidate. The other exotics can be much heavier, allowing larger dimensionless couplings. [FOOTNOTE:4][ENDFOOTNOTE] Before moving on to discuss dark matter in detail, we note that, of the models in Table 1, only models \((A)\), \((B)\) and \((D)\) are expected to produce (dominant) radiative neutrino masses in the absence of the \(Z_{2}\) symmetry, as we briefly discuss in Appendix A. [TABLE:S2.T1][ENDTABLE] ## 3 Inert Fermionic Dark Matter We now turn our attention to the dark-matter candidates in these models. It is _a priori_ possible that both fermionic and scalar dark-matter candidates are possible, as in Ma’s original proposal [3]. In this section we consider fermionic dark-matter. Note that not all the models contain neutral beyond-SM fermions; specifically, model \((A)\) has \(\mathcal{F}\sim(1,1,-2)\), and models \((F)\) and \((G)\) use \(\mathcal{F}\sim(1,2,-3)\). In these cases all beyond-SM fermions are charged and only scalar dark-matter is possible. One can already exclude the parameter space with light fermions in these models, namely \(M_{\mathcal{F}}<M_{1,2}\), due to the appearance of a stable charged-fermion. On the other hand, models \((B)\) through \((E)\) all contain neutral fermions and can, in principle, admit fermionic dark-matter. However, all of the fermion multiplets in these models have nonzero hypercharge, which can lead to strong constraints from direct-detection experiments. More precisely, if the dark-matter abundance is generated by a standard thermal WIMP one can exclude Dirac-fermion dark-matter with nonzero hypercharge, due to the strong constraints from e.g. XENON100 [26]. Thus, it is important to determine whether the neutral fermion is Dirac or Majorana. At tree-level the fermion \(\mathcal{F}\) remains a Dirac particle. However, its coupling to the SM neutrinos, which obtain Majorana masses via Figure 1, leads to a small radiative Majorana-mass. For the case of \(\mathcal{F}\sim(1,3,-2)\), the typical diagram is shown in Figure 3. Similar diagrams can occur for models \((C)\) through \((E)\), though the scalar \(S_{1}\) is real in these cases. The loop-induced Majorana mass will, in general, split the Dirac fermion \(\mathcal{F}\) into a pair of Majorana fermions. However, one can already see that the mass-splitting will be very small. The sub-loop in Figure 3 is the same loop-diagram that generates SM neutrino masses in Figure 1. Thus, in the limit that SM neutrino masses vanish, \(m_{\nu}\to 0\), the Majorana mass for \(\mathcal{F}\) will also vanish. We therefore expect \(\Delta M_{\mathcal{F}}\propto m_{\nu}\), where \(\Delta M_{\mathcal{F}}\) is the Majorana mass for \(\mathcal{F}\). This is born out by explicit calculations. For example, with \(M_{\mathcal{F}}\ll M_{S}\), where \(M_{S}\) denotes an approximate common mass for the beyond-SM scalars, one obtains \[\Delta M_{\mathcal{F}} \sim \frac{\lambda_{1}^{2}\lambda_{11\text{\tiny H}}}{16\pi^{2}}\, \frac{\langle H\rangle^{2}}{M_{S}^{2}}\times m_{\nu},\] (5) where the Lagrangian contains the term \(\lambda_{11\text{\tiny H}}^{*}(H^{\dagger}S_{1})^{2}\subset\mathcal{L}\) to generate the uppermost vertex in Figure 3. The beyond-SM neutral fermions therefore form pseudo-Dirac particles with a tiny splitting. [FIGURE:S3.F3][ENDFIGURE] Direct detection experiments give strong constraints on spin-independent elastic-scattering events that can occur when fermionic dark-matter couples to the \(Z\) boson [26]. These constraints can be avoided if the Dirac fermion has a mass-split, as the resulting pair of Majorana fermions has non-diagonal couplings to the \(Z\) boson (to leading order). Provided the mass split exceeds the average kinetic-energy of the local dark-matter particles, \(Z\)-boson exchange with SM detectors is highly suppressed, as the heavier fermion is kinematically inaccessible. However, for the models with neutral fermions in Table 1 the mass split satisfies \[\frac{\Delta M_{\mathcal{F}}}{M_{\mathcal{F}}} < \frac{m_{\nu}}{M_{\mathcal{F}}}\ \ \lesssim\ \ 10^{-12}\quad\quad \mathrm{for}\quad M_{\mathcal{F}}=\mathcal{O}(\mathrm{TeV}),\] (6) which is (much) too small to evade direct-detection bounds, given typical DM speeds of \(v_{\text{\tiny DM}}\sim 10^{-3}\). We conclude that none of the models in Table 1 are viable when the lightest beyond-SM field is a fermion, due to either a cosmologically excluded stable charged-particle or a dark-matter candidate that contradicts direct-detection constraints. The entire region of parameter space in which a fermion is the lightest beyond-SM state is thus excluded for these models. ## 4 Inert Scalar Dark Matter With the above information we can restrict our attention to the limit \(M_{\mathcal{F}}\gg M_{1,2}\) for the models in Table 1, for which the stable particle is a scalar. In this limit the models are effectively inert \(N\)-tuplet models with the additional feature of realizing radiative neutrino mass. In this section, we consider the viability of the scalar dark-matter candidates in the different models. We first consider models \((A)\) and \((B)\), whose common features allow them to be discussed together. Both these models have a single (candidate) dark-matter multiplet, which is an inert SM-like scalar doublet, \(S_{1}\sim(1,2,1)\); i.e. an inert doublet [10]. Also, in both models the second scalar is a charged doublet, \(S_{2}=(S_{2}^{++},S_{2}^{+})^{T}\), whose components must be heavier than the dark matter. Inert-doublet dark matter is well-studied in the literature, and it is known that a viable dark-matter abundance can be realized [10]. The inert-doublet leads to three new scalars, which we denote as \(H^{\prime\pm}\), \(H^{\prime 0}\) and \(A^{0}\), and either of the last two can be the dark matter. As per usual for an inert-doublet model, the neutral components of \(S_{1}\) cannot mix with the SM Higgs in models \((A)\) and \((B)\), due to the discrete symmetry. However, the charged scalar \(S_{1}^{+}\) will mix with \(S_{2}^{+}\), as mentioned already. If this mixing is large, the phenomenology of the lightest charged-scalar will differ from that of \(H^{\prime+}\) in a standard inert-doublet model. For small mixing the lightest charged scalar will correspond mostly to \(H^{\prime+}\) and the phenomenology of \(S_{1}\) will be well approximated by a standard inert-doublet. Note that one cannot take the limit \(\lambda_{\text{\tiny SH}}\to 0\) without turning off the radiative neutrino mass in Figure 1. The demand that radiative neutrino mass is realized therefore requires nonzero mixing between \(S_{1}^{+}\) and \(S_{2}^{+}\). However, given that \(S_{2}\) must be heavier than the dark-matter, one generally expects the mixing to be of order \(\langle H^{0}\rangle^{2}/M_{2}^{2}\) which is \(\lesssim 10^{-1}\) for \(M_{2}\gtrsim\) TeV. Thus, \(S_{1}\) can be well-approximated by a standard inert-doublet. The inert-doublet model contains five main regions of parameter space in which the observed relic-abundance is obtained [27]. Four of these have a light particle-spectrum that can be probed at the LHC. The discovery of an SM-like scalar with mass of roughly \(125\) GeV at the LHC allows one to update the viable parameter space and phenomenology of the inert-doublet models. Recent analysis, incorporating the LHC data, shows that the low-mass regions for the dark-matter candidate can already be in tension with constraints from XENON100 [26] and WMAP [28], while the heavier region with \(M_{\text{\tiny DM}}\gtrsim 500\) GeV is essentially unaffected [29]. Specifically, the surviving region for lighter dark-matter lies close to the Higgs-resonance/\(WW\)-production threshold [29]. The region of parameter space with \(M_{\text{\tiny DM}}=\mathcal{O}(10)\) GeV is particularly interesting for the present models as, in this case, the additional beyond-SM multiplets can be light enough to appear at the LHC. This was discussed already in Ref. [25] for model \((A)\), where it was shown that the charged scalar-doublet \(S_{2}\sim(1,2,3)\) can produce observable signals when the inert-doublet dark matter is light. Although the region of parameter space with heavier dark-matter (\(M_{\text{\tiny DM}}\gtrsim 500\) GeV) will not be accessible at the LHC, it is expected that XENON-1T will probe this parameter space, potentially giving observable direct-detection signals [30]. One deduces that viable dark matter is possible in both models \((A)\) and \((B)\), and that the combined (projected) LHC and XENON-1T data sets are expected to probe the viable parameter space in these models. We note that, in general, models with hypercharge-less dark matter are not as strongly constrained by direct-detection experiments. Such candidates do not couple directly to the \(Z\) boson so interactions with detectors do not arise at tree-level. Provided the mass-splitting between the charged and neutral components of the dark-matter multiplet exceeds the average kinetic-energy of the dark matter in the local halo, interactions with the \(W\) boson are also highly suppressed (or absent). Even if the neutral and charged components of a dark-matter multiplet are degenerate at tree-level, an \(\mathcal{O}(100)\) MeV split is induced radiatively, which is sufficient to ensure scattering via \(W\) boson exchange is suppressed/absent. These comments apply to model \((D)\), in which the sole dark-matter candidate is the neutral component of the inert real-triplet, \(S_{1}\sim(1,3,0)\). The neutral component of this multiplet does not interact with the \(Z\) boson, and the charged component can be sufficiently split (by radiative effects) to ensure the neutral state is the lightest field. This alleviates potential tension with direct-detection experiments. The possibility of inert real-triplet dark matter is well known in the literature [11, 31, 32, 33]. The neutral component of \(S_{1}\) is a viable cold dark-matter candidate that saturates the observed relic abundance of \(\Omega_{\text{\tiny CDM}}\hat{h}^{2}\simeq 0.11\)[28] for \(M_{\text{\tiny DM}}\approx 2.5\) TeV [11]. If the real-triplet is lighter it can only comprise part of the dark-matter abundance and additional candidates are needed. In model \((D)\) the dark-matter abundance must be comprised solely of \(S_{1}\), so that \(M_{\text{\tiny DM}}\approx 2.5\) TeV is a necessary requirement for this model. Unfortunately this makes it difficult to directly produce the exotic states at the LHC; both \(\mathcal{F}\sim(1,2,-1)\) and \(S_{2}\sim(1,1,2)\) must be heavier than \(2.5\) TeV to ensure the dark matter is the lightest exotic, pushing them beyond projected experimental reach. Model \((D)\) is, however, a viable model of dark matter and radiative neutrino mass for \(M_{\text{\tiny DM}}\approx 2.5\) TeV. Next we turn our attention to models \((F)\) and \((G)\), which both employ \(\mathcal{F}\sim(1,2,-3)\) and \(S_{1}\sim(1,3,2)\). In both models \(S_{2}\) is comprised purely of electrically-charged fields, so the neutral component of \(S_{1}\) is the sole dark-matter candidate. This complex neutral-field cannot mix with the SM scalar, due to the \(Z_{2}\) symmetry, so its particle and antiparticle sates remain degenerate. The dark-matter abundance is therefore comprised of both states, posing a serious difficulty for these models. Due to the nonzero hypercharge for \(S_{1}\), the neutral field can scatter off SM-detectors via tree-level \(Z\)-boson exchange. This process is strongly constrained by direct-detection data sets. Previous works show that one requires a mass of \(\sim 2.6\) TeV to obtain the correct abundance, however, the spin-independent cross-section exceeds \(10^{-37}\)cm\({}^{2}\) in the regions of parameter space compatible with the LEP experiments [32, 33]. Such a large cross section is incompatible with the constraints from, e.g., XENON100 [26]. Thus, although one can successfully generate the requisite dark-matter abundance, direct-detection constraints prove fatal for models \((F)\) and \((G)\) and both models can be excluded. It remains for us to consider models \((C)\) and \((E)\). These models admit two distinct scalar dark-matter candidates and thus allow more possibilities, as we shall see in following section. ## 5 Models with a Real Scalar and a Complex Triplet Models \((C)\) and \((E)\) both employ the SM-like fermion \(\mathcal{F}\sim(1,2,-1)\) and the complex scalar triplet \(S_{2}\sim(1,3,2)\). Furthermore, in both cases \(S_{1}\) is a real scalar. These models differ from the other cases as both scalars now possess a neutral component, giving two dark-matter candidates. We saw that models \((F)\) and \((G)\) could be excluded precisely because the dark-matter abundance was comprised of the neutral component of the complex scalar triplet. This difficulty is avoided in models \((C)\) and \((E)\), however, due to allowed mass-mixing between the neutral components of \(S_{1}\) and \(S_{2}\). In this section we discuss model \((C)\) in some detail, to elucidate the possibilities. The analysis of model \((E)\) is rather similar, due to the related field content, and we limit ourselves to some brief comments on this model at the end of the section. Model \((C)\) contains the beyond-SM scalars \(S_{1}\sim(1,1,0)\equiv S\) and \(S_{2}\sim(1,3,2)\equiv\Delta\). The full scalar-potential can be written as \[V(H,S,\Delta) = \frac{-\mu^{2}}{2}|H|^{2}+\frac{\tilde{M}_{S}^{2}}{2}S^{2}+\tilde {M}_{\Delta}^{2}\mathrm{Tr}(\Delta^{\dagger}\Delta)+\lambda_{1}|H|^{4}+\lambda _{2}\,[\mathrm{Tr}(\Delta^{\dagger}\Delta)]^{2}\] (7) \[+\lambda_{2}^{\prime}\,\mathrm{Tr}(\Delta^{\dagger}\Delta\Delta^{ \dagger}\Delta)+\frac{\lambda_{3}}{2}S^{4}+\lambda_{4}\,|H|^{2}\,\mathrm{Tr}( \Delta^{\dagger}\Delta)+\lambda_{4}^{\prime}\,H^{\dagger}\Delta^{\dagger}\Delta H\] \[+\lambda_{5}|H|^{2}S^{2}+\lambda_{6}\,S^{2}\,\mathrm{Tr}(\Delta^{ \dagger}\Delta)+\lambda_{\text{\tiny SH}}S\left\{\tilde{H}^{\dagger}\Delta^{ \dagger}H+H^{\dagger}\Delta\tilde{H}\right\},\] where the overall phase of \(\Delta\) has been used to choose \(\lambda_{\text{\tiny SH}}\) real without loss of generality.⁵ The discrete symmetry \(\{S,\,\Delta\}\rightarrow-\{S,\,\Delta\}\) ensures there is no mass-mixing between the SM scalar and the beyond-SM fields. This symmetry also forbids terms linear in a single beyond-SM scalar, like \(H\Delta^{\dagger}H\) or \(SH^{2}\), which would otherwise induce a non-zero VEV for \(\Delta\) and \(S\) after electroweak symmetry breaking. Consequently, parameter space exists in which neither \(S\) nor \(\Delta\) acquire a VEV. The scalar \(S\) and the neutral components of \(\Delta\) will mass-mix, however, due to the \(\lambda_{\text{\tiny SH}}\) term in the potential. [FOOTNOTE:5][ENDFOOTNOTE] Expanding the neutral SM-scalar around its VEV, and expanding the neutral component of \(\Delta\) as \[H^{0}=\frac{1}{\sqrt{2}}(v+h^{0}+i\chi^{0})\quad\mathrm{and} \quad\Delta^{0}=\frac{1}{\sqrt{2}}(\Delta_{R}+i\Delta_{I}),\] (8) respectively, the mass-mixing Lagrangian for the neutral scalars is \[\mathcal{L} \supset -\frac{1}{2}\mathcal{S}^{T}\mathcal{M}^{2}\mathcal{S}.\] (9) Here the basis vector is \(\mathcal{S}=(S,\Delta_{R},\Delta_{I})^{T}\), and the squared-mass matrix has the form \[\mathcal{M}^{2}=\left(\begin{array}[]{ccc}\hidden@noalign{}\hfil \tilde{M}_{S}^{2}+\lambda_{5}v^{2}&\frac{\lambda_{\text{\tiny SH}}}{2\sqrt{2}} v^{2}&0\\ \hidden@noalign{}\hfil\frac{\lambda_{\text{\tiny SH}}}{2\sqrt{2}}v^{2}&\tilde{ M}_{\Delta}^{2}+\frac{\lambda_{4}}{2}v^{2}&0\\ \hidden@noalign{}\hfil 0&0&\tilde{M}_{\Delta}^{2}+\frac{\lambda_{4}}{2}v^{2} \end{array}\right).\] (13) Thus, the CP-odd scalar \(\Delta_{I}\) is a mass eigenstate with mass \(\tilde{M}_{\Delta}^{2}+\lambda_{4}v^{2}/2\), while the CP-even scalars \(S\) and \(\Delta_{R}\) mass-mix to produce two physical scalars that are linear combinations of these fields. The dark-matter candidate will be one of the neutral-scalar mass eigenstates. To determine which one, we must find the masses for the mixed CP-even states. Let us define \(M_{S}^{2}=\tilde{M}_{S}^{2}+\lambda_{5}v^{2}\) and \(M^{2}_{\Delta}=\tilde{M}_{\Delta}^{2}+\lambda_{4}v^{2}/2\), which are the CP-even mass eigenstates in the limit \(\lambda_{\text{\tiny SH}}\to 0\). In this limit \(\Delta_{R}\) and \(\Delta_{I}\) are degenerate and form a single complex-scalar with mass \(M^{2}_{\Delta}\). For nonzero \(\lambda_{\text{\tiny SH}}\), the CP-even mass eigenvalues can be written as \[M_{\pm}=\frac{1}{2}\left\{M_{S}^{2}+M^{2}_{\Delta}\pm\left[(M_{S }^{2}-M^{2}_{\Delta})^{2}+\frac{\lambda_{\text{\tiny SH}}^{2}}{2}v^{4}\right]^ {1/2}\right\},\] (14) where the eigenstates are related to the original fields as \[\left(\begin{array}[]{c}\hidden@noalign{}\hfil S_{+}\\ \hidden@noalign{}\hfil S_{-}\end{array}\right)=\left(\begin{array}[]{cc} \hidden@noalign{}\hfil\cos\theta&\sin\theta\\ \hidden@noalign{}\hfil-\sin\theta&\cos\theta\end{array}\right)\,\left(\begin{ array}[]{c}\hidden@noalign{}\hfil S\\ \hidden@noalign{}\hfil\Delta_{R}\end{array}\right).\] (21) Here, the mixing angle is \[\tan 2\theta = \frac{\lambda_{\text{\tiny SH}}v^{2}}{\sqrt{2}(M^{2}_{S}-M^{2}_{ \Delta})}.\] (22) In the limit where the singlet-scalar is heaviest, \(M_{S}^{2}\gg M^{2}_{\Delta}\), the mass eigenvalues are approximately \[M_{+}^{2}\simeq M_{S}^{2}+\frac{\lambda_{\text{\tiny SH}}^{2}v^{ 4}}{8M_{S}^{2}}\quad\mathrm{and}\quad M_{-}^{2}\simeq M^{2}_{\Delta}-\frac{ \lambda_{\text{\tiny SH}}^{2}v^{4}}{8M_{S}^{2}}\quad\mathrm{for}\quad M_{S}^{2 }\gg M^{2}_{\Delta}.\] (23) Noting that \(M_{-}^{2}<M^{2}_{\Delta}\), reveals that \(S_{-}\) is the lightest exotic state and is thus the DM candidate. Simple expressions are obtained for the mass eigenvectors in this limit: \[\left.\begin{array}[]{c}\hidden@noalign{}\hfil S_{+}\simeq S+ \frac{\lambda_{\text{\tiny SH}}v^{2}}{2\sqrt{2}M_{S}^{2}}\Delta_{R}\\ \\ \hidden@noalign{}\hfil S_{-}\simeq\Delta_{R}-\frac{\lambda_{\text{\tiny SH}}v^ {2}}{2\sqrt{2}M_{S}^{2}}S\end{array}\right.\quad\quad\mathrm{for}\quad M_{S}^{ 2}\gg M^{2}_{\Delta},\] (27) so that the lightest scalar \(S_{-}\) is comprised mostly of \(\Delta_{R}\). Thus, for \(M_{S}^{2}\gg M^{2}_{\Delta}\) the dark matter is comprised of \(S_{-}\), which mostly consists of the CP-even part (\(\Delta_{R}\)) of the neutral field in the scalar triplet \(S_{2}\). The mass-splitting between \(S_{-}\) and the CP-odd state \(\Delta_{I}\) is \(|\Delta M^{2}|=\lambda_{\text{\tiny SH}}^{2}v^{4}/8M_{S}^{2}\). Provided this splitting exceeds the dark-matter kinetic-energy, \(\sqrt{|\Delta M^{2}|}>\mathrm{KE}_{\text{\tiny DM}}\), the state \(\Delta_{I}\) will not be kinematically accessible via tree-level processes in direct-detection experiments. This significantly weakens the bounds on dark matter arising from a complex triplet. This also gives an upper bound on the mass parameter \(M_{S}^{2}\), beyond which the splitting between the dark matter and \(\Delta_{I}\) is so small that tree-level scattering via \(Z\) exchange is expected in present-day experiments. One finds \[M_{S} < \frac{1}{2\sqrt{2}}\,\frac{|\lambda_{\text{\tiny SH}}|\,v^{2}}{ \mathrm{KE}_{\text{\tiny DM}}}\ \simeq\ \left(\frac{|\lambda_{\text{\tiny SH}} |}{10^{-2}}\right)\,\left(\frac{10^{-3}}{v_{\text{\tiny DM}}}\right)^{2}\left( \frac{2.6~{}\mathrm{TeV}}{M_{\text{\tiny DM}}}\right)\times 10^{2}~{}\mathrm{ TeV}.\] (28) Thus, the heavier state \(S_{+}\) cannot be made arbitrarily heavy if the dark matter is to avoid exclusion via, e.g., the XENON100 data. With \(M_{\text{\tiny DM}}\equiv M_{-}=\mathcal{O}(\mathrm{TeV})\), the mass-split between \(S_{-}\) and \(\Delta_{I}\) is smaller than \(M_{\Delta}\) (the mass of \(\Delta_{I}\)). Once the temperature drops below \(M_{\Delta}\), the heavier state \(\Delta_{I}\) will decay, with the decay products necessarily containing \(S_{-}\), due to the conserved discrete-symmetry. The expression \(|\Delta M^{2}|=\lambda_{\text{\tiny SH}}^{2}v^{4}/8M_{S}^{2}\) shows that the mass-splitting between \(S_{-}\) and \(\Delta_{I}\) is bounded as \(\sqrt{|\Delta M^{2}|}\lesssim|\lambda_{\text{\tiny SH}}|\times 4\) GeV, given that \(M_{\Delta}\gtrsim 3\) TeV is needed to achieve the correct relic abundance and we are working with \(M_{S}>M_{\Delta}\). In this mass range \(\Delta_{I}\) can decay as \(\Delta_{I}\to S_{-}+Z^{*}\to S_{-}+\bar{f}f\), where \(f\) is a SM fermion with mass \(m_{f}<|\lambda_{\text{\tiny SH}}|\times 2\) GeV. Therefore, even if charged SM fermions are not kinematically available, final-states containing neutrinos will be accessible unless \(\lambda_{\text{\tiny SH}}\) is exceptionally small. After \(\Delta_{I}\) has decayed away, the primordial plasma is comprised of \(S_{-}\) and the SM fields. \(S_{-}\) can maintain equilibrium with the SM sector via gauge interactions and via the \(\lambda_{4}\) and \(\lambda_{4}^{\prime}\) (hereafter \(\lambda_{4}\)) quartic terms in Eq. (7).⁶ When the quartic interactions are dominant, the model is similar to an inert real-triplet model; the dark-matter abundance will be obtained for \(M_{\text{\tiny DM}}\simeq 2.5\) TeV, in line with the analysis of Ref. [11]. As one makes \(\lambda_{4}\) smaller, the tree-level mass splitting between the charged and neutral components of \(\Delta\) diminishes and coannihilation channels like \(\Delta^{-}\Delta^{++}\to W^{+}\gamma\) become available. At this point, making \(\lambda_{4}\) smaller does not modify the requisite dark-matter mass as gauge interactions dominate. The analysis of Ref. [33] for an inert complex-triplet finds that \(M_{\text{\tiny DM}}\gtrsim 2.8\) TeV is required for the entire region of parameter space.⁷ We thus expect that \(M_{\text{\tiny DM}}\gtrsim 2.5\) TeV will be required even when the gauge interactions dominate the quartic interactions during freeze-out for the present scenario. With this value fixed, model \((C)\) becomes a viable model of neutrino mass and dark matter. It will be difficult, however, to produce the exotics in this model, given that the lightest exotic mass is \(\gtrsim 2.5\) TeV.⁸ [FOOTNOTE:6][ENDFOOTNOTE] [FOOTNOTE:7][ENDFOOTNOTE] [FOOTNOTE:8][ENDFOOTNOTE] We now briefly discuss the alternative limit with \(M_{S}^{2}\ll M^{2}_{\Delta}\). In this case the CP-even mass eigenvalues are \[M_{+}^{2}\simeq M_{\Delta}^{2}+\frac{\lambda_{\text{\tiny SH}}^{ 2}v^{4}}{8M_{\Delta}^{2}}\quad\mathrm{and}\quad M_{-}^{2}\simeq M^{2}_{S}- \frac{\lambda_{\text{\tiny SH}}^{2}v^{4}}{8M_{\Delta}^{2}}\quad\mathrm{for} \quad M_{\Delta}^{2}\gg M^{2}_{S}.\] (29) We see that the dark-matter candidate remains as the lightest CP-even eigenstate \(S_{-}\), with mass \(M_{-}\). The mass eigenstates are now \[\left.\begin{array}[]{c}\hidden@noalign{}\hfil S_{+}\simeq\Delta_ {R}+\frac{\lambda_{\text{\tiny SH}}v^{2}}{2\sqrt{2}M_{\Delta}^{2}}S\\ \\ \hidden@noalign{}\hfil S_{-}\simeq S-\frac{\lambda_{\text{\tiny SH}}v^{2}}{2 \sqrt{2}M_{\Delta}^{2}}\Delta_{R}\end{array}\right.\quad\quad\mathrm{for}\quad M _{\Delta}^{2}\gg M^{2}_{S},\] (33) so the dark matter is comprised mostly of the singlet-scalar \(S\). Singlet-scalar dark matter is well known [9] and detailed analysis show that \(M_{\text{\tiny DM}}\gtrsim 80\) GeV is compatible with direct-detection constraints and WMAP data for a Higgs mass of \(m_{h}\simeq 125\) GeV [37]. The viable region of parameter space can be probed by XENON1T, excepting a small resonant window with \(M_{\text{\tiny DM}}\simeq 62\) GeV, where the dark-matter-Higgs coupling can be very small. Lighter dark-matter with \(M_{\text{\tiny DM}}\lesssim 60\) GeV is ruled out by LHC bounds on invisible Higgs decays [37]. We see that model \((C)\) has viable parameter space in which it behaves like an inert-triplet model or an inert singlet model. This analysis is sufficient to demonstrate that model \((E)\) also has viable regions of parameter space. In model \((E)\) one has \(S_{1}\sim(1,3,0)\), while the second scalar remains as \(S_{2}\sim(1,3,2)\). The scalar potential for this model contains a term similar to the \(\lambda_{\text{\tiny SH}}\) term in Eq. (7), which mixes the neutral component of \(S_{1}\) with the CP-even neutral component of \(S_{2}\). If \(S_{2}\) is heaviest the model behaves like an inert real-triplet model, while if \(S_{1}\) is heaviest the lightest scalar is mostly comprised of \(\Delta_{R}\) (the neutral CP-even part of \(S_{2}\)). Direct-detection constraints can be evaded due to the mass mixing, and the model is again an effective model of inert-triplet dark matter. In both cases we expect that a viable dark-matter abundance and viable neutrino masses can be obtained, though the dark matter will be heavy, with \(M_{\text{\tiny DM}}\gtrsim 2.5\) TeV (neglecting resonant regions). In terms of the observational prospects for the beyond-SM multiplets at the LHC, the limit \(M_{\Delta}^{2}\gg M^{2}_{S}\) in model \((C)\) appears to be the most optimistic scenario for the models in Table 1 (excepting the resonant regions, which also allow lighter fields). In this limit the dark matter can be relatively light, \(M_{\text{\tiny DM}}\simeq 100\) GeV, and thus the exotic states \(\Delta\) and \(\mathcal{F}\) can both be of order a few hundred GeV. In principle it could be possible to observe all three beyond-SM multiplets in this limit. For the other viable models in Table 1 the dark matter has to be relatively heavy: \(M_{\text{\tiny DM}}\gtrsim 500\) GeV for the inert doublet models, and \(M_{\text{\tiny DM}}\gtrsim 2.5\) TeV for the inert triplet cases. This pushes the additional beyond-SM fields beyond the reach of the LHC. Note that any mass degeneracy between charged and neutral members of an inert multiplet at tree-level is lifted by radiative effects, making the charged components heavier than the neutral components. The heavier members of a given multiplet can decay to lighter members of the same multiplet via the weak interactions, e.g. \(\mathcal{F}^{+}\to W^{+}+\mathcal{F}^{0}\), where the \(W\) can be virtual. A heavier multiplet can also decay to a lighter multiplet via the Yukawa coupling; e.g. \(\mathcal{F}^{-}\to S^{0}+\ell^{-}\) if \(M_{\mathcal{F}}\gg M_{S}\). Because of the discrete symmetry the new fields can only be pair produced in colliders, and conservation of the \(Z_{2}\) charge means final states resulting from exotic decay chains necessarily include stable electrically-neutral fields that will escape the detector. ## 6 On the Origin of the Discrete Symmetry Following Ma’s original proposal, we employed a discrete \(Z_{2}\) symmetry to ensure stability of the lightest beyond-SM field appearing in the neutrino-mass diagram. One can argue that the use of a discrete symmetry is not completely satisfying, either because it seems ad hoc, or because of the view that quantum gravity effects are not expected to conserve global symmetries. This motivates one to consider whether a simple explanation for the discrete symmetry can be found. The simplest possibility is to replace the discrete symmetry with a gauged \(U^{\prime}(1)\) symmetry, which would not be broken by quantum gravity effects. With enough additional ingredients one can presumably achieve this goal for all the models we have discussed. However, we would like to know which models allow for a minimal extension, such that \(Z_{2}\to U^{\prime}(1)\), and a single SM-singlet scalar \(\eta\) is added to the particle spectrum to break the \(U^{\prime}(1)\) symmetry. Writing the full gauge group as \(\mathcal{G}_{\text{\tiny SM}}\times U(1)^{\prime}\), where \(\mathcal{G}_{\text{\tiny SM}}\) is the SM gauge group, we have the following transformation properties for the beyond-SM fields⁹ [FOOTNOTE:9][ENDFOOTNOTE] \[\eta\sim(1_{\text{\tiny SM}},Q_{\eta})\,,\quad S_{1,2}\sim(Q^{ \text{\tiny SM},}_{1,2},Q)\,,\quad\mathcal{F}\sim(Q^{\text{\tiny SM}}_{ \mathcal{F}},-Q),\] (34) where the “SM” superscript denotes the charges under \(\mathcal{G}_{\text{\tiny SM}}\), given in Table 1. Inspection of Eq. (3) shows that all Lagrangian terms needed to generate neutrino mass are allowed by the \(U^{\prime}(1)\) symmetry. However, in the case of models \((A)\) and \((B)\), which are inert-doublet models, the enhanced symmetry prevents the additional term \((S_{1}^{\dagger}H)^{2}\). This term is not needed for neutrino mass but is required to split the neutral components of \(S_{1}\sim(1,2,1)\) in order to avoid direct-detection constraints [10]. Thus, models \((A)\) and \((B)\) are not compatible with this minimal symmetry extension. On the other hand, we find that models \((C)\), \((D)\) and \((E)\), which have one scalar forming a real representation of the SM gauge symmetry, remain as viable models of dark matter provided \(Q_{\eta}=-2Q\). This relationship is needed to lift a mass-degeneracy of neutral beyond-SM fields. For example, consider model \((C)\), which now has the following terms in the scalar potential \[V(H,S,\Delta,\eta) \supset \lambda_{\text{\tiny SH}}\left\{S\tilde{H}^{\dagger}\Delta^{ \dagger}H+S^{*}H^{\dagger}\Delta\tilde{H}\right\}\ +\ \frac{\mu_{\eta}}{2} \left\{S^{2}\eta+S^{*2}\eta^{\dagger}\right\},\] (35) in addition to the terms in Eq. (7). All other terms containing \(\eta\) depend only on the modulus \(|\eta|^{2}\). We have used the relative phase of \(S\) and \(\Delta\) to choose \(\lambda_{\text{\tiny SH}}\) real and the phase of \(S\) to choose \(\mu_{\eta}\) real. Note that the symmetry breaking \(U^{\prime}(1)\to Z_{2}\) is achieved by nonzero \(\langle\eta\rangle\), motivating the discrete symmetry as an accidental subgroup of the gauged \(U^{\prime}(1)\) symmetry. In the basis \(\mathcal{S}=(S_{R},\Delta_{R},S_{I},\Delta_{I})^{T}\), the squared-mass matrix has the form¹⁰ [FOOTNOTE:10][ENDFOOTNOTE] \[\mathcal{M}^{2}=\left(\begin{array}[]{cccc}\hidden@noalign{}\hfil \tilde{M}_{S}^{2}+\lambda_{5}v^{2}+2\mu_{\eta}\langle\eta\rangle&\frac{\lambda _{\text{\tiny SH}}}{4}v^{2}&0&0\\ \hidden@noalign{}\hfil\frac{\lambda_{\text{\tiny SH}}}{4}v^{2}&\tilde{M}_{ \Delta}^{2}+\frac{\lambda_{4}}{2}v^{2}&0&0\\ \hidden@noalign{}\hfil 0&0&\tilde{M}_{S}^{2}+\lambda_{5}v^{2}-2\mu_{\eta} \langle\eta\rangle&\frac{\lambda_{\text{\tiny SH}}}{4}v^{2}\\ \hidden@noalign{}\hfil 0&0&\frac{\lambda_{\text{\tiny SH}}}{4}v^{2}&\tilde{M}_ {\Delta}^{2}+\frac{\lambda_{4}}{2}v^{2}\end{array}\right).\] (40) Observe that the entries for the CP-even and CP-odd states are identical in the limit \(\mu_{\eta}\to 0\). This would produce degenerate states so that, in the case where the dark matter is comprised mostly of \(\Delta\), the dark matter would be ruled out by XENON100 (it would be an inert complex-triplet). For non-zero \(\mu_{\eta}\), however, the CP-even and CP-odd states are non-degenerate and viable dark-matter is achieved. When the dark matter is mostly (or completely, for model \((D)\)) comprised of a real representation of \(\mathcal{G}_{\text{\tiny SM}}\), the splitting achieved by nonzero \(\mu_{\eta}\) also ensures direct detection signals resulting from mixing between \(Z^{\prime}\) and \(Z\) are suppressed.¹¹ [FOOTNOTE:11][ENDFOOTNOTE] There will be an additional scattering process for the dark matter due to the mixing between \(\eta\) and the SM scalar, which gives a standard Higgs portal interaction. Given that the coupling for this interaction is not needed to achieve the observed dark-matter abundance, one can always choose this coupling to be small enough to comply with constraints. Thus, with this simple gauge extension, we can explain the origin of the discrete symmetry for models \((C)\), \((D)\) and \((E)\), while retaining the desirable features of radiative neutrino mass and a viable dark-matter abundance. Note that Ma’s original proposal [3], and the variant using a real triplet fermion [4], are not compatible with this minimal symmetry upgrade; in the case of scalar dark-matter, the \((S^{\dagger}H)^{2}\) term is precluded, meaning direct-detection experiments rule the model out, similar to the minimal gauge extension of models \((A)\) and \((B)\). Furthermore, one encounters gauge-anomalies given that \(\mathcal{F}_{R}\) is a chiral field in Refs. [3, 4] — additional model building is therefore needed to explain the origin of the \(Z_{2}\) symmetry in these cases. We do not pursue this matter here. ## 7 Beyond the Adjoint Representation In the preceding sections we studied generalizations of Ma’s 2006 model with radiative neutrino mass and stable dark-matter candidates. In doing so we restricted our attention to beyond-SM multiplets no larger than the adjoint representation. As mentioned already, one can generate neutrino mass via Figure 1 and obtain dark-matter candidates in models with larger multiplets. We briefly discuss this matter in the present section. First consider the case with a Majorana mass insertion, as in Figure 2. Allowing for \(SU(2)\) representations as large as the quintuplet-rep. we find two additional models. Both of these employ the quadruplet scalar \(S\sim(1,4,1)\), with the real fermion being either a triplet \(\mathcal{F}\sim(1,3,0)\), or a quintuplet \(\mathcal{F}\sim(1,5,0)\). The latter model was detailed in Ref. [38].¹² In both models we expect either scalar or fermionic dark-matter is possible, as in Ma’s original proposal; the neutral fermion does not couple to the \(Z\) boson and can therefore remain consistent with direct-detection constraints. [FOOTNOTE:12][ENDFOOTNOTE] Generalizing the models with a fermion mass insertion of the Dirac type (i.e. generalizing the models in Table 1), we find more variants are possible. For completeness we list these in the Appendix, but here offer the following comments. As with the models in Table 1, we find that fermionic dark-matter can be ruled out for all models with larger gauge representations. The fermions remain as pseudo-Dirac particles with tiny splittings, set by the SM neutrino masses. Such small splittings permit unsuppressed tree-level scattering with SM detectors via \(Z\)-boson exchange, which is ruled out by XENON100. We thus rule out the parameter space in which the fermion is the lightest beyond-SM state, for the same reasons as discussed in Section 3. For the case of scalar dark-matter, one has to consider the individual models, as was needed for the models in Table 1. Some models can be immediately ruled out for the same reasons that models \((F)\) and \((G)\) could be excluded; for example, model \((L)\) in Table 2 can be excluded as it gives an inert complex-triplet model. Similarly model \((R)\) in Table 3 is ruled out, as \(S_{2}\) contains only charged components and \(S_{1}\) has nonzero hypercharge. The other models appear to be compatible with direct-detection constraints, provided the neutral components of the scalars mix when the lightest scalar has nonzero hypercharge, much as models \((C)\) and \((E)\) were viable. For example, model \((M)\) contains \(S_{1}\sim(1,4,1)\) as the only beyond-SM scalar with a neutral component. However, the Lagrangian allows a term \(\lambda(S_{1}^{\dagger}H)^{2}\subset\mathcal{L}\) that can split the components of the neutral scalar, allowing one to avoid direct-detection constraints (this is analogous to the splitting obtained in an inert-doublet model). Finally, we note that the use of larger multiplets may have an additional phenomenological benefit. Ref. [5] shows that large multiplets that mediate interactions between dark matter and the SM can enhance loop-induced annihilation of dark matter into \(2\gamma\) and \(\gamma+Z\) final states, without requiring non-perturbatively large couplings. This occurs because the larger multiplets admit fields with larger electric-charges, naturally enhancing loop-processes with final-state photons. It does not appear to be possible to realize the astrophysical gamma-ray signal [6] in the models presented in the Appendix, but simple extensions do seem compatible with this idea. For example, model \((N)\) in Table 3 employs \(\mathcal{F}\sim(1,4,-1)\), \(S_{1}\sim(1,5,0)\) and \(S_{2}\sim(1,5,2)\). When \(S_{1}\) is the lightest beyond-SM state the dark matter is comprised (mostly) of the neutral component of \(S_{1}\). There is a one-loop contribution to processes like \(DM+DM\to 2\gamma,\gamma+Z\), containing virtual \(S_{2}\) states in the loop, that is enhanced by the presence of the multiply charged component in \(S_{2}\). Note that the dark-matter mass is required to be either \(M_{\text{\tiny DM}}\sim 130\) GeV or \(M_{\text{\tiny DM}}\sim 144\) GeV, in order to generate the gamma-ray excess via dark matter annihilations into either \(2\gamma\) or \(\gamma+Z\) final states, respectively. However, dark-matter comprised of \(S_{1}\sim(1,5,0)\) is expected to have a mass \(M_{\text{\tiny DM}}\gtrsim 5\) TeV in order to achieve the correct relic abundance [11], which is too large to explain the astrophysical signal. If one adds a singlet scalar \(S\), that is also odd under the \(Z_{2}\) symmetry, to the model, then the region of parameter space where \(S\) is the dark matter is compatible with \(M_{\text{\tiny DM}}\sim 130\) GeV or \(M_{\text{\tiny DM}}\sim 144\) GeV. The components of \(S_{1}\) and/or \(S_{2}\) can then be \(\mathcal{O}(100)\) GeV and the loop-processes advocated in Ref. [5] are present in the model, thereby enhancing the astrophysical gamma-ray signal.¹³ In this example there is a simple connection between the astrophysical signal and the mechanism of neutrino mass, with the large multiplets that enable the latter also enhancing the former. It could be interesting to take these ideas further to see if the dark matter can be realized as one of the fields in the neutrino loop-diagram, rather than an extra degree of freedom, or to study the phenomenology of the model just described. [FOOTNOTE:13][ENDFOOTNOTE] ## 8 Conclusion We studied a class of models with radiative neutrino mass and stable dark-matter candidates. Neutrino mass was generated by a one-loop diagram with the same topography as that proposed by Ma [3]. We generalized Ma’s approach, detailing all variants with beyond-SM fields no larger than the adjoint representation. In the case where the neutrino mass diagram contained a Majorana mass insertion, only two models were found, both of which were known. When the mass-insertion was of the Dirac type, such that lepton-number symmetry was broken by a vertex, we found a number of additional models. Fermionic dark-matter was excluded in all of these models, while two of the models were completely excluded due to direct-detection constraints. The remaining models allowed radiative neutrino mass and achieved a viable (scalar) dark-matter abundance. There were cases with an inert singlet, an inert doublet, and an inert triplet, providing a natural setting for inert \(N\)-tuplet models of dark-matter, with the additional feature of achieving radiative neutrino mass. Interestingly, some of the models allowed a simple extension, such that the (formerly imposed) discrete symmetry emerged as an accidental low-energy symmetry. We briefly discussed models with larger beyond-SM multiplets, showing that viable scenarios exist. With simple extensions, the large multiplets enabling neutrino mass can also enhance present-day astrophysical gamma-ray signals, allowing a simple connection between the mechanism of neutrino mass and the astrophysical gamma-ray signal. ## Acknowledgments The authors thank Y. Kajiyama, K. Nagao, H. Okada, T. Schwetz, A. Strumia, and K. Yagyu. SSCL is supported in part by the NSC under Grant No. NSC-101-2811-M-006-015 and in part by the NCTS of Taiwan. KM is supported by the Australian Research Council. ## Appendix A Mass Without the \(\mathbf{Z_{2}}\) Symmetry Of the models presented in Table 1, only models \((A)\), \((B)\) and \((D)\) are expected to produce (dominant) radiative neutrino masses in the absence of the \(Z_{2}\) symmetry. The other models contain the triplet scalar \(S_{1,2}\sim(1,3,2)\), which Yukawa-couples to the SM leptons, and acquires a VEV due to the term \(\mu HS_{1,2}H\subset V(H,S_{1},S_{2})\), in the absence of the discrete symmetry. Thus, tree-level neutrino masses of the standard Type-II seesaw [40] form are expected to dominate the loop effect when the \(Z_{2}\) symmetry is discarded.¹⁴ For models \((A)\), \((B)\) and \((D)\), on the other hand, tree-level neutrino masses do not arise if the \(Z_{2}\) symmetry is removed, while the loop-diagram in Figure 1 persists. Note that, if the \(Z_{2}\) symmetry is turned off, the fermion \(\mathcal{F}\) is not needed in order to generate nonzero radiative neutrino masses in model \((D)\)[41]. However, the spectrum obtained without \(\mathcal{F}\) is of the simplified-Zee form [41], which is incompatible with the observed mixing pattern [42]. Thus, the fermion \(\mathcal{F}\sim(1,2,-1)\) is required to obtain a _viable_ mixing pattern in the absence of the \(Z_{2}\) symmetry. [FOOTNOTE:14][ENDFOOTNOTE] Model \((B)\) has a similar particle content to the model presented in Ref. [43], modulo the replacement \(S_{2}\sim(1,2,3)\to S_{2}\sim(1,4,3)\). This difference precludes the tree-level mass found in Ref. [43] so model \((B)\) is purely a model of radiative masses, which could be studied without the discrete symmetry and dark matter.¹⁵ [FOOTNOTE:15][ENDFOOTNOTE] ## Appendix B Models with Larger Multiplets In the text we found seven models that employ beyond-SM multiplets in either the fundamental or adjoint representation of \(SU(2)_{L}\), and had an internal Dirac mass-insertion. In addition to these minimal models, one can realize radiative neutrino mass and dark-matter candidates with larger multiplets. We present the additional minimal models that arise if one permits multiplets forming the quadruplet (isospin-\(3/2\)) representation of \(SU(2)_{L}\) in Table 2. The labeling scheme follows on from Table 1 in the text. If one allows for quintuplet multiplets there are additional models, shown in Table 3 (also see Ref. [44] for a detailed example). The first case listed as model \((M)\) was presented in Ref. [15]. [TABLE:A2.T2][ENDTABLE] [TABLE:A2.T3][ENDTABLE] ## References * [1] M. C. Gonzalez-Garcia, M. Maltoni, J. Salvado and T. Schwetz, JHEP **1212**, 123 (2012) [arXiv:1209.3023 [hep-ph]]. * [2] A. H. G. Peter, arXiv:1201.3942 [astro-ph.CO]. * [3] E. Ma, Phys. Rev. D **73**, 077301 (2006) [hep-ph/0601225]. * [4] E. Ma and D. Suematsu, Mod. Phys. Lett. A **24**, 583 (2009) [arXiv:0809.0942 [hep-ph]]. * [5] J. Kopp, E. T. Neil, R. Primulando and J. Zupan, Phys. Dark. Univ. **2** (2013) 22 [arXiv:1301.1683 [hep-ph]]. * [6] C. Weniger, JCAP **1208**, 007 (2012) [arXiv:1204.2797 [hep-ph]]. * [7] L. M. Krauss, S. Nasri and M. Trodden, Phys. Rev. D **67**, 085002 (2003) [hep-ph/0210389]. * [8] C. Boehm, Y. Farzan, T. Hambye, S. Palomares-Ruiz and S. Pascoli, Phys. Rev. D **77**, 043516 (2008) [hep-ph/0612228]; M. Lindner, D. Schmidt and T. Schwetz, Phys. Lett. B **705**, 324 (2011) [arXiv:1105.4626 [hep-ph]]; F. -X. Josse-Michaux and E. Molinaro, Phys. Rev. D **84**, 125021 (2011) [arXiv:1108.0482 [hep-ph]]; S. Kanemura, T. Nabeshima and H. Sugiyama, Phys. Rev. D **85**, 033004 (2012) [arXiv:1111.0599 [hep-ph]]; M. Gustafsson, J. M. No and M. A. Rivera, arXiv:1212.4806 [hep-ph]; M. Aoki, J. Kubo and H. Takano, arXiv:1302.3936 [hep-ph]; Y. Kajiyama, H. Okada and T. Toma, arXiv:1303.7356 [hep-ph]. * [9] V. Silveira and A. Zee, Phys. Lett. B **161**, 136 (1985); J. McDonald, Phys. Rev. D **50** (1994) 3637 [hep-ph/0702143 [HEP-PH]]; C. P. Burgess, M. Pospelov and T. ter Veldhuis, Nucl. Phys. B **619** (2001) 709 [hep-ph/0011335]. * [10] R. Barbieri, L. J. Hall and V. S. Rychkov, Phys. Rev. D **74** (2006) 015007 [hep-ph/0603188]; L. Lopez Honorez, E. Nezri, J. F. Oliver and M. H. G. Tytgat, JCAP **0702** (2007) 028 [hep-ph/0612275]. Q. -H. Cao, E. Ma and G. Rajasekaran, Phys. Rev. D **76** (2007) 095011 [arXiv:0708.2939 [hep-ph]]; S. Andreas, M. H. G. Tytgat and Q. Swillens, JCAP **0904** (2009) 004 [arXiv:0901.1750 [hep-ph]]; L. Lopez Honorez and C. E. Yaguna, JCAP **1101** (2011) 002 [arXiv:1011.1411 [hep-ph]]; M. Gustafsson, S. Rydbeck, L. Lopez-Honorez and E. Lundstrom, Phys. Rev. D **86** (2012) 075019 [arXiv:1206.6316 [hep-ph]]; S. Kashiwase and D. Suematsu, Phys. Rev. D **86** (2012) 053001 [arXiv:1207.2594 [hep-ph]]. * [11] M. Cirelli, N. Fornengo and A. Strumia, Nucl. Phys. B **753**, 178 (2006) [hep-ph/0512090]; M. Cirelli, A. Strumia and M. Tamburini, Nucl. Phys. B **787**, 152 (2007) [arXiv:0706.4071 [hep-ph]]. * [12] E. Ma, Phys. Rev. Lett. **81** (1998) 1171 [hep-ph/9805219]. * [13] F. Bonnet, M. Hirsch, T. Ota and W. Winter, JHEP **1207**, 153 (2012) [arXiv:1204.5862 [hep-ph]]. * [14] M. Fabbrichesi and S. Petcov, arXiv:1304.4001 [hep-ph]. * [15] K. L. McDonald, JHEP **1307**, 020 (2013) [arXiv:1303.4573 [hep-ph]]. * [16] K. Hally, H. E. Logan and T. Pilkington, Phys. Rev. D **85**, 095017 (2012) [arXiv:1202.5073 [hep-ph]]. * [17] R. Foot, H. Lew, X. G. He and G. C. Joshi, Z. Phys. C **44**, 441 (1989). * [18] S. -Y. Ho and J. Tandean, arXiv:1303.5700 [hep-ph]. * [19] D. Schmidt, T. Schwetz and T. Toma, Phys. Rev. D **85**, 073009 (2012) [arXiv:1201.0906 [hep-ph]]. * [20] C. -K. Chua and S. S. C. Law, Phys. Rev. D **83**, 055010 (2011) [arXiv:1011.4730 [hep-ph]]; A. Delgado, C. Garcia Cely, T. Han and Z. Wang, Phys. Rev. D **84**, 073007 (2011) [arXiv:1105.5417 [hep-ph]]; S. S. C. Law and K. L. McDonald, Phys. Lett. B **713**, 490 (2012) [arXiv:1204.2529 [hep-ph]]; I. Baldes, N. F. Bell, K. Petraki and R. R. Volkas, arXiv:1304.6162 [hep-ph]; G. Bambhaniya, J. Chakrabortty, S. Goswami and P. Konar, arXiv:1305.2795 [hep-ph]. * [21] E. Del Nobile, R. Franceschini, D. Pappadopulo and A. Strumia, Nucl. Phys. B **826**, 217 (2010) [arXiv:0908.1567 [hep-ph]]. * [22] A. Joglekar, P. Schwaller and C. E. M. Wagner, JHEP **1212** (2012) 064 [arXiv:1207.4235 [hep-ph]]. * [23] C. Arina, R. N. Mohapatra and N. Sahu, Phys. Lett. B **720** (2013) 130 [arXiv:1211.0435 [hep-ph]]. * [24] S. S. C. Law, JHEP **1202** (2012) 127 [arXiv:1106.0375 [hep-ph]]. * [25] M. Aoki, S. Kanemura and K. Yagyu, Phys. Lett. B **702**, 355 (2011) [Erratum-ibid. B **706**, 495 (2012)] [arXiv:1105.2075 [hep-ph]]. * [26] E. Aprile _et al._ [XENON100 Collaboration], Phys. Rev. Lett. **107** (2011) 131302 [arXiv:1104.2549 [astro-ph.CO]]; Phys. Rev. Lett. **109**, 181301 (2012) [arXiv:1207.5988 [astro-ph.CO]]. * [27] E. M. Dolle and S. Su, Phys. Rev. D **80** (2009) 055012 [arXiv:0906.1609 [hep-ph]]. * [28] E. Komatsu _et al._ [WMAP Collaboration], Astrophys. J. Suppl. **192** (2011) 18 [arXiv:1001.4538 [astro-ph.CO]]. * [29] A. Goudelis, B. Herrmann and O. Stål, arXiv:1303.3010 [hep-ph]. * [30] M. Klasen, C. E. Yaguna and J. D. Ruiz-Alvarez, arXiv:1302.1657 [hep-ph]; M. Klasen, C. E. Yaguna, J. D. Ruiz-Alvarez, D. Restrepo and O. Zapata, arXiv:1302.5298 [hep-ph]. * [31] P. Fileviez Perez, H. H. Patel, M. J. Ramsey-Musolf and K. Wang, Phys. Rev. D **79**, 055024 (2009) [arXiv:0811.3957 [hep-ph]]; * [32] T. Hambye, F. -S. Ling, L. Lopez Honorez and J. Rocher, JHEP **0907** (2009) 090 [Erratum-ibid. **1005** (2010) 066] [arXiv:0903.4010 [hep-ph]]. * [33] T. Araki, C. Q. Geng and K. I. Nagao, Phys. Rev. D **83** (2011) 075014 [arXiv:1102.4906 [hep-ph]]; T. Araki, C. Q. Geng and K. I. Nagao, Int. J. Mod. Phys. D **20** (2011) 1433 [arXiv:1108.2753 [hep-ph]]. * [34] R. Foot, A. Kobakhidze, K. L. McDonald and R. R. Volkas, Phys. Rev. D **76** (2007) 075014 [arXiv:0706.1829 [hep-ph]]; R. Foot, A. Kobakhidze, K. L. McDonald and R. R. Volkas, Phys. Rev. D **77**, 035006 (2008) [arXiv:0709.2750 [hep-ph]]. * [35] Y. Kajiyama, H. Okada and K. Yagyu, arXiv:1303.3463 [hep-ph]. * [36] S. Kanemura and H. Sugiyama, Phys. Rev. D **86** (2012) 073006 [arXiv:1202.5231 [hep-ph]]. * [37] A. Djouadi, O. Lebedev, Y. Mambrini and J. Quevillon, Phys. Lett. B **709**, 65 (2012) [arXiv:1112.3299 [hep-ph]]. * [38] K. Kumericki, I. Picek and B. Radovcic, Phys. Rev. D **86**, 013006 (2012) [arXiv:1204.6599 [hep-ph]]; I. Picek and B. Radovcic, Phys. Lett. B **719**, 404 (2013) [arXiv:1210.6449 [hep-ph]]. The same model is mentioned in Y. Liao, JHEP **1106**, 098 (2011) [arXiv:1011.3633 [hep-ph]]. * [39] M. Cirelli and A. Strumia, New J. Phys. **11** (2009) 105005 [arXiv:0903.3381 [hep-ph]]. * [40] W. Konetschny and W. Kummer, Phys. Lett. B **70**, 433 (1977); T. P. Cheng and L. F. Li, Phys. Rev. D **22**, 2860 (1980); M. Magg and C. Wetterich, Phys. Lett. B **94**, 61 (1980); J. Schechter and J. W. F. Valle, Phys. Rev. D **22**, 2227 (1980); G. Lazarides, Q. Shafi and C. Wetterich, Nucl. Phys. B **181**, 287 (1981); C. Wetterich, Nucl. Phys. B **187**, 343 (1981); R. N. Mohapatra and G. Senjanovic, Phys. Rev. D **23**, 165 (1981). * [41] S. S. C. Law and K. L. McDonald, arXiv:1303.6384 [hep-ph]. * [42] X. -G. He, Eur. Phys. J. C **34**, 371 (2004) [hep-ph/0307172]. * [43] K. S. Babu, S. Nandi and Z. Tavartkiladze, Phys. Rev. D **80**, 071702 (2009) [arXiv:0905.2710 [hep-ph]]. * [44] I. Picek and B. Radovcic, Phys. Lett. B **687**, 338 (2010) [arXiv:0911.1374 [hep-ph]]; K. Kumericki, I. Picek and B. Radovcic, Phys. Rev. D **84**, 093002 (2011) [arXiv:1106.1069 [hep-ph]].
## 1 Введение Экспериментальные данные о массе нейтрино, полученные за последние десятилетия, являются конкретным доказательством физики, выходящей за пределы Стандартной модели (СМ) (см. например [1]). Хотя не удалось определить необходимые новые степени свободы, ясно, что скорее всего существуют дополнительные частицы, чтобы генерировать массы. Аналогично, существует обширное количество доказательств наличия дополнительного галактического компонента, неизвестного вещества, называемого тёмной материей (см. например [2]). Это может или не может требовать новых степеней свободы, но гипотеза о том, что тёмная материя состоит из стабильного (или долгоживущего) нового вида частиц, предоставляет простое объяснение этой наблюдаемой особенности Вселенной. Учитывая, что эти два показателя физики за пределами СМ могут быть объяснены расширением спектра частиц СМ, естественно задать вопрос, можно ли связать необходимые новые частицы. Может ли механизм образования массы нейтрино быть связанным с существованием стабильного кандидата на тёмную материю? Особенно простая модель, реализующая эту идею, была предложена Ма в 2006 году [3]. В этой модели основному набору Стандартной Модели добавляются дополнительный Стандартно-Модельный скалярный двойт и синглетные фермионы, которые все являются нечетными относительно дискретной симметрии \(Z_{2}\). Расширенный набор полей позволяет образованние радиационной массы нейтрино на однопетлевом уровне, в то время как самое легкое поле за пределами Стандартной Модели является абсолютно стабильным. Таким образом, получается простая синергетическая модель радиационной массы нейтрино и темной материи. В данной работе мы обобщаем подход Ма. Мы представляем класс связанных моделей, все они порождают массу нейтрино с помощью петляной диаграммы с той же топологией, что и у Ма, и при этом допускают стабильных кандидатов на тёмную материю. Петлевая диаграмма, используемая Ма, содержит вставку массы на внутренней фермионной линии (см. рисунок 1), и наши обобщения естественно подразделяются на две категории: те, которые нарушают симметрию лептонного числа путем вставки массы Майораны, и те с вставкой массы Дирака, при которой симметрия лептонного числа нарушается в вершине. Хотя базовый механизм очень похож в обоих случаях, эта разница модифицирует ожидания относительно содержимого полей за пределами Стандартной Модели и связанной феноменологии. Оказывается, что в обоих случаях этот подход очень общий и существует множество реализаций. Однако ограничение внимания на моделях, в которых мультиплеты внесмешенного типа не превышают представления адъюнктов, значительно сокращает возможности. Как мы увидим, существует только две такие (минимальные) модели с вставкой массы Майораны, которые уже известны [3, 4]. Мы находим семь дополнительных моделей с вставкой массы типа Дирака, все из которых обеспечивают радиационную массу нейтрино и кандидатов на темную материю. Мы подробно описываем эти модели, находим их подмножество, соответствующее прямому обнаружению экспериментов. Есть случаи с инертным синглетом, инертным дублетом и инертным триплетом; модели представляют естественную среду для инертных теорий \(N\)-tuplet темной материи, в которых также достигается радиационная масса нейтрино. Интересно, три из новых моделей предлагают простое расширение, которое может объяснить происхождение (ранее наложенной) дискретной симметрии. Путем повышения дискретной симметрии до калиброванной симметрии \(U(1)\) и расширения поля на одну синглетную скалярную частицу стандартной модели, дискретная симметрия может возникать как случайная симметрия лагранжиана низкоэнергетической области после нарушения симметрии \(U(1)\). Это предоставляет простое объяснение для дискретной симметрии. Хотя мы сосредотачиваемся на моделях с представлениями не большими, чем адъюнкт, мы также кратко рассматриваем случаи, когда дополнительные поля могут быть квадруплетной и/или квинтуплетной представлениями \(SU(2)_{L}\). Мы представляем кандидаты моделей в этих случаях и упоминаем некоторые ключевые вопросы, основываясь на уроках, извлеченных из наших предыдущих исследований. Несмотря на использование более крупных мультиплетов, эти модели всё равно могут быть интересными; помимо возможности возникновения радиационной массы нейтрино, экзотические частицы с более большими электрическими зарядами в этих мультиплетах могут повысить сигнал \(2\gamma\) и/или \(\gamma+Z\) от аннигиляции темной материи, когда они появляются в петлях [5]. Это может обеспечить простую связь между механизмом массы нейтрино и астрофизическим гамма-лучевым сигналом [6]. Перед тем, как продолжить, мы замечаем, что связь между радиационной массой нейтрино и темной материей была исследована в нескольких различных моделях, включая Ref. [7], предшествующую работу Ma; для других примеров см. [8]. Предыдущие работы по моделям с инертным синглетом [9], инертным двойником [10] и инертным тройником [11] также хорошо известны. Дополнительные соответствующие работы, занимающиеся инертной мультиплетной темной материей и/или радиационной массой нейтрино, приводятся в тексте. Также обратите внимание, что Refs. [12, 13] подробно описывают однопетлевые реализации оператора \(d=5\) для массы нейтрино. Инертная скалярная темная материя также может помочь устранить проблему малой иерархии, обнаруженную в низкомасштабных сисасах [14]. Настоящая работа основана на обобщенных сисасах на дереве уровня, представленных в Ref. [15]. План этой статьи следующий. В разделе 2 мы обсуждаем модель Ма и представляем обобщения, которые также обеспечивают радиационную массу нейтрино и кандидатов на тёмную материю. Раздел 3 рассматривает случай фермионной тёмной материи, в то время как раздел 4 обсуждает скалярную тёмную материю. Один из обобщённых моделей подробно представлен в разделе 5. В разделе 6 мы показываем, что некоторые модели позволяют простое расширение таким образом, что дискретная симметрия появляется как случайная симметрия в теории низких энергий. Модели с экзотиками, образующими более крупные представления \(SU(2)\), обсуждаются в разделе 7 (и явно приводятся в приложении). Мы заключаем в разделе 8. [FIGURE:S1.F1][ENDFIGURE] ## 2 Радиационная масса нейтрино и темная материя Мы интересуемся классом моделей, которые генерируют массу нейтрино радиативно по диаграмме на рисунке 1. Здесь \(\mathcal{F}\) является фермионом за пределами стандартной модели, а \(S_{1,2}\) являются новыми скалярами (которые могут быть идентичными в некоторых случаях). Особенностью этой диаграммы является то, что все три вершины могут включать два поля за пределами стандартной модели. В результате всегда можно рассмотреть дискретную симметрию \(Z_{2}\), действующую на полеты за пределами стандартной модели. \[\{\mathcal{F},\;S_{1},\;S_{2}\} \rightarrow -\ \{\mathcal{F},\;S_{1},\;S_{2}\}\;,\] (1) \[\{\mathcal{Ф},\;S_{1},\;S_{2}\} \rightarrow -\ \{\mathcal{Ф},\;S_{1},\;S_{2}\}\;,\] (1) Все поля SM преобразуются тривиально. Самое легкое поле внутри мультиплетов \(\mathcal{F}\) и \(S_{1,2}\) будет, следовательно, стабильным, и если это поле электрически нейтральное и безцветное, мы получаем кандидата на тёмную материю. Эти замечания являются общими для всех моделей данного типа; связь между массами петель и тёмной материей легко осознаваема в этом классе моделей. Фигура 1 представляет массы майоранских нейтрино, поэтому диаграмма петли должна содержать источник нарушения числа лептонов. Выбрав соглашение о симметрии числа лептонов таким образом, что новый фермион \(\mathcal{F}\) имеет то же значение, что и SM лептоны, существуют два способа явного нарушения симметрии числа лептонов; это может быть нарушено либо при массовой вставке, либо при одном из вершин. Самые простые модели в терминах требуемого количества мультиплетов за пределами SM - это модели с массовой вставкой, нарушающей численность лептонов (Майорановской). В этом случае \(\mathcal{F}_{L}\equiv\mathcal{F}_{R}^{c}\) и минимальные случаи возникают для \(S_{1}=S_{2}\equiv S\). Таким образом, требуется всего два мультиплета за пределами SM. Общая диаграмма петли для этого подмножества моделей приведена на рисунке 2. Сначала рассмотрим этот случай. ### Модели со вставкой массы Майораны Мы ищем модели, которые достигают массы нейтрино через Рисунок 2 и порождают кандидатов в тёмную материю. Чтобы позволить голую мажорановскую массу, фермион должен образовывать реальное представление калибровочной симметрии СМ, \(\mathcal{F}_{R}\sim(1,R_{\mathcal{F}},0)\). Поскольку мы рассматриваем кандидатов в тёмную материю, мы не рассматриваем цветные поля. Также обратите внимание, что \(R_{\mathcal{F}}\) должно быть нечётным, чтобы гарантировать отсутствие частиц с дробным зарядом (самая лёгкая из которых была бы стабильной и, следовательно, исключена космологически). Нечётное значение \(R_{\mathcal{F}}\) также обеспечивает наличие электрически нейтральной компоненты в \(\mathcal{F}_{R}\), поэтому эта требование не налагает дополнительных ограничений. С данной информацией можно получить пригодные комбинации \(\mathcal{F}_{R}\) и \(S\), которые генерируют Рисунок 2. Основными лагранжевыми терминами являются \[\mathcal{L} \supset i\bar{\mathcal{F}_{R}}\gamma^{\mu}D_{\mu}\mathcal{F}_{R}-\frac{M _{\mathcal{F}}}{2}\;\overline{\mathcal{F}_{R}^{c}}\mathcal{F}_{R}+|D^{\mu}S|^{2}-M_{S}^{2}|S|^{2}+\lambda\bar{L}\tilde{S}\mathcal{F}_{R}+\lambda_{\text{ \tiny SH}}(S^{\dagger}H)^{2}+\mathrm{H.c.},\] (2) где \(L\) (\(H\)) - это SM лептон (скалярный) двойник, а \(\tilde{S}\) обозначает сопряженно-заряженный \(S\). Оказывается, что возможные комбинации для \(\mathcal{F}\) и \(S\) не ограничены квантовыми числами; можно рассматривать все большие мультиплеты, предположительно, до некоторых единичных пределов [16], и реализовать модель с Рисунком 2 и кандидатом на темную материю. Однако, если мы ограничимся моделями с \(R_{\mathcal{F}},R_{S}\leq 3\), так что ни один новый мультиплет не будет больше сопряженного представления, то существует только два возможных варианта. Первый случай - это оригинальное предложение Ма, которое использует дополнительный (инертный) скалярный двойник \(S\sim(1,2,1)\) и фермион сингл Ма \(\mathcal{F}_{R}\sim(1,1,0)\)[3]. Эта модель является прототипом для класса, которым мы занимаемся. Вторая модель также использует скалярный двойник \(S\sim(1,2,1)\), но вместо этого использует тройной фермион \(\mathcal{F}_{R}\sim(1,3,0)\)[4], знакомый из Type-III seesaw [17]. Таким образом, обе модели известны в литературе, и нет дополнительных возможностей, если только не рассматриваются более крупные мультиплеты. В каждой из этих моделей кандидатом на роль тёмной материи может выступать нейтральная компонента S либо мажорановский фермион. Однако существует важное отличие между случаем синглета \(\mathcal{F}\sim(1,1,0)\) и другой моделью; синглет не участвует в слабых взаимодействиях и поэтому находится в термическом контакте с Стандартной моделью через юкавское связывание. В случае фермионной тёмной материи это может привести к конфликту между необходимостью сохранения большого юкавского связывания для обеспечения термической тёмной материи и необходимостью подавления юкавского связывания для ограничения размера эффектов, связанных с изменением вкуса. Такая проблема не возникает в модели фермионного триплета, так как фермионы могут поддерживать равновесие с Стандартной моделью через слабые взаимодействия в этих случаях, даже если юкавские связывания подавлены. Для анализа модели Ма с использованием последних данных LHC о бозоне Хиггса см., например, ссылку [18]. Также следует отметить, что петлевые эффекты могут вызывать наблюдаемые взаимодействия между тёмной материей и экспериментальными детекторами в модели Ма [19]. [FIGURE:S2.F2][ENDFIGURE] ### Модели с вставкой массы Дирака. После исчерпания минимальных моделей с нарушением числа лептонов через вставку массы, мы теперь рассматриваем модели с вставкой массы Дирака; то есть фермион за пределами СМ имеет ненулевую гиперзарядность. В этом случае общая массовая диаграмма имеет форму, показанную на рисунке 1. Поля \(\mathcal{F}_{R}\) и \(\mathcal{F}_{L}\) больше не связаны сопряжением зарядов, поэтому вставка массы является типом Дирака и \(\mathcal{F}\) является векторным фермионом. Кроме того, требуется \(S_{1}\neq S_{2}\). Мы снова рассматриваем симметрию \(Z_{2}\), при которой поля СМ преобразуются тривиально, но новые поля являются нечетными. Лагранжиан содержит следующие соответствующие члены: \[\mathcal{L} \supset i\bar{\mathcal{F}}\gamma^{\mu}D_{\mu}\mathcal{F}\ -\ M_{\mathcal {F}}\;\overline{\mathcal{F}}\mathcal{F}\ +\ \sum_{i=1,2}\left\{|D^{\mu}S_{i}|^ {2}\ -\ M_{i}^{2}|S_{i}|^{2}\right\}\] (3) \[+\ \lambda_{1}\;\bar{L}\mathcal{F}_{R}S_{1}\ +\ \lambda_{2}\;\bar {L}\mathcal{F}^{c}_{L}\tilde{S}_{2}\ +\ \lambda_{\text{\tiny SH}}\;S_{1}\tilde {S}_{2}H^{2}+\mathrm{H.c.},\] где, в нашей конвенции, симметрия числа лептонов нарушается юкавским связыванием. При присутствии симметрии \(Z_{2}\) нет членов в скалярном потенциале, линейных только в одном из новых скаляров \(S_{1,2}\). Таким образом, скаляры за пределами СМ не приобретают индуцированного ВЭБ и существует область параметров, для которой \(\langle S_{1,2}\rangle=0\), поэтому симметрия \(Z_{2}\) остается точной. При выборе пригодных множеств необходимо убедиться, что ни в одном из них нет поля с дробным зарядом, чтобы избежать (космологически исключенного) устойчивого заряженного поля. Чтобы гарантировать, что самое лёгкое Z2-нечётное поле является нейтральным кандидатом на роль тёмной материи, необходимо требовать, чтобы новые множества содержали хотя бы одно нейтральное поле. Обратите внимание, что нейтральное поле не обязательно должно появляться в виде явной пропагирующей степени свободы внутри диаграммы петли; достаточно лишь наличия диаграммы петли и содержания частиц, включающего нейтральное поле, которое может играть роль тёмной материи. С учетом этих условий мы ищем жизнеспособные комбинации мультиплетов за пределами Стандартной модели, которые реализуют Рисунок 1. Мы обнаруживаем, что размеры мультиплетов за пределами Стандартной модели не ограничены нашими требованиями; можно рассматривать все более крупные мультиплеты и реализовывать петлевую диаграмму. Однако, если обратить внимание только на модели, в которых ни один из мультиплетов за пределами Стандартной модели не больше адъюнктного представления, найдено только семь различных моделей. Они перечислены в Таблице 1. Из семи моделей, одна использует экзотический трехкратный лептон \(\mathcal{F}\sim(1,3,-2)\), изученный в работах [20, 21], три содержат экзотический векторный (и похожий на Стандартную модель) двойной лептон \(\mathcal{F}\sim(1,2,-1)\) [21, 22, 23], и две содержат заряженный двойной лептон \(\mathcal{F}\sim(1,2,-3)\) [21, 24]. Также есть модель с похожим на Стандартную модель заряженным синглетным фермионом, \(\mathcal{F}\sim(1,1,-2)\), который уже появился в работе [25]. Массы нейтрино принимают стандартную расчетную форму в этих моделях. Например, в моделях \((A)\) и \((B)\) только одиноко заряженные экзотические частицы распространяются в петле, а матрица масс нейтрино СМ задается³ [FOOTNOTE:3][ENDFOOTNOTE] перевести маркдаун с английского на русский. Сохранить формулы: \[(\mathcal{M}_{\nu})_{\alpha\beta} \simeq \frac{\left[(\lambda_{2}^{*})^{a}_{\alpha}(\lambda_{1}^{*})^{a}_{ \beta}+(\lambda_{2}^{*})^{a}_{\beta}(\lambda_{1}^{*})^{a}_{\alpha}\right]}{32 \pi^{2}}\,\frac{\lambda_{\text{\tiny SH}}\langle H\rangle^{2}}{M_{>}^{2}-M_{<} ^{2}}\left[\frac{M^{2}_{>}\,M_{\mathcal{F},a}}{M_{\mathcal{F},a}^{2}-M_{>}^{2} }\,\log\frac{M_{\mathcal{F},a}^{2}}{M_{>}^{2}}\ -\ (M_{>}\to M_{<})\right]\] Здесь \(M_{\mathcal{F},a}\) - это масса заряженной компоненты экзотического фермиона \(\mathcal{F}_{a}\), и подразумевается суммирование по повторяющемуся индексу \(a\) (который обозначает поколения экзотических фермионов).⁴ Массы \(M_{>,<}\) относятся к массовым собственным состояниям заряженных скалярных частиц, которые являются линейными комбинациями заряженных скаляров \(S_{1}^{+}\) и \(S_{2}^{+}\). Смешивание происходит из-за члена \(\lambda_{\text{\tiny SH}}\)-терма в уравнении (3), который принимает явный вид \(\lambda_{\text{\tiny SH}}\tilde{H}^{\dagger}S_{1}S_{2}^{\dagger}H\subset \mathcal{L}\) для моделей \((A)\) и \((B)\). Если все экзотики находятся на тевтонском масштабе, то для получения \(m_{\nu}\sim 0.1\) эВ требуются безразмерные связи порядка \(\mathcal{O}(10^{-3})\). Сценарий, в котором все экзотики находятся на тевтонском масштабе, наиболее интересен с феноменологической точки зрения. Однако строго говоря, чтобы получить кандидата на тёмную материю, достаточно, чтобы у самой лёгкой экзотики была масса порядка \(\lesssim\mathcal{O}(\mathrm{TeV})\). Остальные экзотики могут быть значительно тяжелее, что позволяет большим безразмерным связям. [FOOTNOTE:4][ENDFOOTNOTE] - [Сноска:4][КОНЕЦ СНОСКА] Перед тем, как перейти к более подробному обсуждению темной материи, отметим, что модели (A), (B) и (D) из таблицы 1, ожидается, что они порождают (доминирующие) радиационные массы нейтрино в отсутствие симметрии \(Z_{2}\), о чем мы кратко обсудим в Приложении A. [TABLE:S2.T1][ENDTABLE] ## 3 Инертная фермионная тёмная материя Теперь обратим внимание на кандидатов на темную материю в этих моделях. _А преори_ возможно, что и фермионные, и скалярные кандидаты на темную материю возможны, как в исходном предложении Ма [3]. В этом разделе мы рассмотрим фермионную темную материю. Обратите внимание, что не все модели содержат нейтральные фермионы за пределами Стандартной Модели; конкретно, модель \((A)\) имеет \(\mathcal{F}\sim(1,1,-2)\), а модели \((F)\) и \((G)\) используют \(\mathcal{F}\sim(1,2,-3)\). В этих случаях все фермионы за пределами Стандартной Модели заряжены, и возможна только скалярная темная материя. В этих моделях уже можно исключить пространство параметров с легкими фермионами, а именно \(M_{\mathcal{F}}<M_{1,2}\), из-за появления стабильного заряженного фермиона. С другой стороны, модели (B) через (E) все содержат нейтральные фермионы и, в принципе, могут принимать фермионную темную материю. Однако у всех фермионных мультиплетов в этих моделях ненулевая гиперзарядность, что может привести к сильным ограничениям от экспериментов прямого детектирования. Более точно, если изобилие темной материи порождается стандартной термической WIMP-частицей, то можно исключить дираковскую фермионную темную материю с ненулевой гиперзарядностью из-за сильных ограничений, например, от XENON100 [26]. Таким образом, важно определить, является ли нейтральный фермион дираковым или майорановским. На дереве уровня фермион \(\mathcal{F}\) остается Дираковской частицей. Однако его связь с нейтрино Стандартной Модели, которые получают Майорановкие массы через фигурку 1, приводит к небольшой радиационной Майорановской массе. В случае \(\mathcal{F}\sim(1,3,-2)\), типичная диаграмма показана на рисунке 3. Аналогичные диаграммы могут возникать для моделей (C) до (E), хотя в этих случаях скаляр \(S_{1}\) является вещественным. Луп-индуцированная Майорановская масса, в общем случае, расщепляет дираковский фермион \(\mathcal{F}\) на пару Майорановских фермионов. Однако можно уже видеть, что расщепление массы будет очень малым. Подлупа, показанная на рисунке 3, является той же самой луп-диаграммой, которая порождает массы нейтрино Стандартной Модели на рисунке 1. Таким образом, в пределе, когда массы нейтрино Стандартной Модели исчезают, \(m_{\nu}\to 0\), Майорановская масса для \(\mathcal{F}\) также исчезает. Поэтому мы ожидаем \(\Delta M_{\mathcal{F}}\propto m_{\nu}\), где \(\Delta M_{\mathcal{F}}\) - это Майорановская масса для \(\mathcal{F}\). Это подтверждается явными вычислениями. Например, при \(M_{\mathcal{F}}\ll M_{S}\), где \(M_{S}\) обозначает приблизительно одинаковую массу за пределами-Стандартной Модели скаляров, получается \[\Delta M_{\mathcal{F}} \sim \frac{\lambda_{1}^{2}\lambda_{11\text{\tiny H}}}{16\pi^{2}}\, \frac{\langle H\rangle^{2}}{M_{S}^{2}}\times m_{\nu},\] (5) где Лагранжиан содержит термин \(\lambda_{11\text{\tiny H}}^{*}(H^{\dagger}S_{1})^{2}\subset\mathcal{L}\) для генерации верхнего вершины на Рис. 3. Вне-стандартные нейтральные фермионы, следовательно, образуют псевдо-Дираковы частицы с маленьким расщеплением. [ФИГУРА:S3.F3][КОНЕЦФИГУРЫ] Эксперименты прямого обнаружения серьезно ограничивают упругое рассеяние безучастных осей, которое может происходить, когда фермионная темная материя связывается с бозоном \(Z\) [26]. Эти ограничения можно избежать, если фермион Дирака имеет разделение масс, поскольку полученная пара мажорановых фермионов имеет недиагональные связи с бозоном \(Z\) (ведущего порядка). При условии, что разделение масс превышает среднюю кинетическую энергию местных частиц темной материи, обменом бозоном \(Z\) с детекторами стандартной модели сильно подавляется, поскольку тяжелый фермион кинематически недоступен. Однако для моделей с нейтральными фермионами из таблицы 1 разделение масс удовлетворяет уравнению \[\frac{\Delta M_{\mathcal{F}}}{M_{\mathcal{F}}} < \frac{m_{\nu}}{M_{\mathcal{F}}}\ \ \lesssim\ \ 10^{-12}\quad\quad \mathrm{для}\quad M_{\mathcal{F}}=\mathcal{O}(\mathrm{TeV}),\] (6) который слишком мал, чтобы избежать прямых границ обнаружения, с учетом типичных скоростей темной материи \(v_{\text{\tiny DM}}\sim 10^{-3}\). Мы приходим к выводу, что ни одна из моделей в Таблице 1 не является приемлемой, когда наименьшее поле за пределами СМ является фермионом, из-за либо космологически исключенной стабильной заряженной частицы, либо кандидата на темную материю, противоречащего прямым ограничениям обнаружения. Весь регион параметров, в котором фермион является наименьшим состоянием за пределами СМ, тем самым исключен для этих моделей. ## 4 Инертная скалярная тёмная материя С учетом вышеуказанной информации мы можем ограничить наше внимание пределом \(M_{\mathcal{F}}\gg M_{1,2}\) для моделей, представленных в таблице 1, в которых стабильная частица является скалярной. В этом пределе модели эффективно являются моделями инертных \(N\)-метюль, с дополнительной особенностью реализации радиоактивной массы нейтрино. В этом разделе мы рассмотрим жизнеспособность кандидатов в скалярную темную материю в разных моделях. Сначала мы рассмотрим модели \((А)\) и \((В)\), у которых общие особенности позволяют их обсуждать вместе. Обе эти модели имеют единственный (кандидатный) множественный темный материал, который является инертным скалярным двойником, \(S_{1}\sim(1,2,1)\); то есть инертный двойник [10]. Также в обеих моделях второй скаляр является заряженным двойником, \(S_{2}=(S_{2}^{++},S_{2}^{+})^{T}\), компоненты которого должны быть тяжелее темного материала. Инертный двойник темного материала хорошо изучен в литературе, и известно, что может быть достигнуто приемлемое количество темного материала [10]. Инертный двойник приводит к трем новым скалярам, которые мы обозначаем как \(H^{\prime\pm}\), \(H^{\prime 0}\) и \(A^{0}\), и любые из последних двух могут быть темным материалом. Как и обычно для модели инертного двойника, нейтральные компоненты \(S_{1}\) не могут смешиваться с СМ-гиггсом в моделях \((А)\) и \((В)\) из-за дискретной симметрии. Однако заряженный скаляр \(S_{1}^{+}\) будет смешиваться с \(S_{2}^{+}\), как уже упоминалось. Если это смешивание большое, то феноменология самого лёгкого заряженного скаляра будет отличаться от \(H^{\prime+}\) в стандартной модели инертного двойника. При малом смешивании самый лёгкий заряженный скаляр в основном будет соответствовать \(H^{\prime+}\), и феноменология \(S_{1}\) будет хорошо аппроксимироваться стандартной моделью инертного двойника. Обратите внимание, что нельзя взять предел \(\lambda_{\text{\tiny SH}}\to 0\) без выключения радиационной массы нейтрино на Рис. 1. Требование, чтобы радиационная масса нейтрино была реализована, поэтому требует ненулевого смешивания между \(S_{1}^{+}\) и \(S_{2}^{+}\). Однако, учитывая, что \(S_{2}\) должен быть тяжелее тёмного материала, можно ожидать, что смешивание будет порядка \(\langle H^{0}\rangle^{2}/M_{2}^{2}\), которое для \(M_{2}\gtrsim\) ТэВ равно или меньше \(10^{-1}\). Таким образом, \(S_{1}\) может быть хорошо аппроксимирован стандартной моделью инертного двойника. Модель инертного двойника содержит пять основных областей параметров, в которых наблюдается реликтовая плотность [27]. В четырех из них спектр легких частиц может быть исследован на Большом адронном коллайдере. Обнаружение скалярного частицы, похожей на стандартную модель, с массой около \(125\) ГэВ на Большом адронном коллайдере позволяет обновить пригодное пространство параметров и феноменологию моделей инертного двойника. Недавний анализ, основанный на данных Большого адронного коллайдера, показывает, что области низкой массы для кандидатов в тёмную материю уже могут испытывать напряжение из-за ограничений от XENON100 [26] и WMAP [28], тогда как более тяжелая область с \(M_{\text{\tiny DM}}\gtrsim 500\) ГэВ по сути не затронута [29]. Конкретно, выживающая область для более легкой тёмной материи находится близко к резонансу Хиггса/порогу производства \(WW\) [29]. Область пространства параметров с \(M_{\text{\tiny DM}}=\mathcal{O}(10)\) ГэВ является особенно интересной для представленных моделей, поскольку, в этом случае, дополнительные множества за пределами СМ могут быть достаточно легкими для появления на Большом адронном коллайдере (LHC). Это уже обсуждалось в работе [25] для модели \((A)\), где показано, что заряженное скалярное двойноеt \(S_{2}\sim(1,2,3)\) может порождать наблюдаемые сигналы, когда холодное темное вещество инертного двойника является легким. Хотя область пространства параметров с более тяжелым темным веществом (\(M_{\text{\tiny DM}}\gtrsim 500\) ГэВ) не будет доступна на Большом адронном коллайдере (LHC), ожидается, что XENON-1T будет исследовать эту область параметров, потенциально предоставляя наблюдаемые сигналы прямого обнаружения [30]. Отсюда следует, что жизнеспособное темное вещество возможно как в моделях \((A)\), так и в \((B)\), и ожидается, что комбинированные (проекционные) данные LHC и XENON-1T будут исследовать жизнеспособное пространство параметров в этих моделях. Мы отмечаем, что в общем случае модели с темной материей без гиперзаряда не так сильно ограничены экспериментами по прямому обнаружению. Такие кандидаты не связаны напрямую с бозоном Z, поэтому взаимодействия с детекторами не возникают на уровне дерева. При условии, что разделение массы между заряженными и нейтральными компонентами мультиплета темной материи превышает среднюю кинетическую энергию темной материи в локальном гало, взаимодействия с бозоном W также сильно подавлены (или отсутствуют). Даже если нейтральные и заряженные компоненты мультиплета темной материи в силу дерева являются вырожденными, через обмен бозоном W индуцируется радиативное разделение порядка 100 МэВ, что достаточно для подавления / отсутствия рассеяния через обмен бозоном W. Эти комментарии относятся к модели \((D)\), в которой единственным кандидатом на роль темной материи является нейтральный компонент инертного реального трехкратного скалярного поля \(S_{1}\sim(1,3,0)\). Нейтральный компонент этой кратности не взаимодействует с бозоном \(Z\), а заряженный компонент может быть достаточно рассщеплен (радиационными эффектами), чтобы гарантировать, что нейтральное состояние является самым легким полем. Это уменьшает потенциальное противоречие с экспериментами по прямому обнаружению. Возможность инертного трехкратного скрытого скалярного поля в качестве темной материи хорошо известна в литературе [11, 31, 32, 33]. Нейтральный компонент \(S_{1}\) является жизнеспособным кандидатом на роль холодной темной материи, насыщающим наблюдаемую остаточную плотность \(\Omega_{\text{\tiny CDM}}\hat{h}^{2}\simeq 0.11\)[28] для \(M_{\text{\tiny DM}}\approx 2.5\) ТэВ [11]. Если трехкратное скрытое скалярное поле будет легче, то оно может составлять только часть общего объема темной материи, и потребуются дополнительные кандидаты. В модели \((D)\) объем темной материи должен состоять исключительно из \(S_{1}\), поэтому \(M_{\text{\tiny DM}}\approx 2.5\) ТэВ является необходимым условием для этой модели. К сожалению, это затрудняет прямое производство экзотических состояний на Большом адронном коллайдере (ЛАА); как \(\mathcal{F}\sim(1,2,-1)\), так и \(S_{2}\sim(1,1,2)\) должны быть тяжелее \(2.5\) ТэВ, чтобы гарантировать, что темная материя является самым легким экзотическим состоянием, выходящим за пределы предполагаемого экспериментального достижения. Однако модель \((D)\) является жизнеспособной моделью темной материи и радиационного массы нейтрино при \(M_{\text{\tiny DM}}\approx 2.5\) ТэВ. Затем мы обращаем свое внимание на модели \((F)\) и \((G)\), которые оба используют \(\mathcal{F}\sim(1,2,-3)\) и \(S_{1}\sim(1,3,2)\). В обоих моделях \(S_{2}\) состоит исключительно из электрически заряженных полей, поэтому нейтральная компонента \(S_{1}\) является единственным кандидатом в темную материю. Это комплексное нейтральное поле не может смешиваться с скалярным полем Стандартной Модели из-за симметрии \(Z_{2}\), поэтому его частицы и античастицы остаются вырожденными. Поэтому количество темной материи включает оба состояния, что представляет серьезную сложность для этих моделей. Из-за ненулевого гиперзаряда для \(S_{1}\) нейтральное поле может рассеиваться на детекторах Стандартной Модели через обмен \(\Z\)-бозоном на дереве. Этот процесс сильно ограничен данными непосредственного обнаружения. Предыдущие работы показывают, что для получения правильного количества темной материи требуется масса примерно \(2.6\) ТэВ, однако, спин-независимое сечение превышает \(10^{-37}\) см\({}^{2}\) в областях параметров, совместимых с экспериментами на LEP [32, 33]. Такое большое сечение несовместимо с ограничениями, например, от XENON100 [26]. Таким образом, хотя можно успешно получить необходимое количество темной материи, ограничения на непосредственное обнаружение оказываются фатальными для моделей \((F)\) и \((G)\), и обе модели могут быть исключены. Остается рассмотреть модели \((C)\) и \((E)\). Эти модели допускают два различных скалярных кандидата на темную материю и, следовательно, позволяют больше возможностей, как мы увидим в следующем разделе. ## 5 Моделей с вещественным скаляром и комплексным триплетом Модели \((C)\) и \((E)\) оба используют SM-подобные фермионы \(\mathcal{F}\sim(1,2,-1)\) и комплексный скалярный трехплет \(S_{2}\sim(1,3,2)\). Кроме того, в обоих случаях \(S_{1}\) является вещественным скаляром. Эти модели отличаются от других случаев тем, что оба скаляра теперь имеют нейтральную компоненту, что дает двух кандидатов на темную материю. Мы видели, что модели \((F)\) и \((G)\) могут быть исключены именно потому, что обилие темной материи состояло из нейтральной компоненты комплексного скалярного трехплета. Однако, эта сложность устраняется в моделях \((C)\) и \((E)\) благодаря разрешенному массовому смешиванию между нейтральными компонентами \(S_{1}\) и \(S_{2}\). В этом разделе мы более подробно рассмотрим модель \((C)\), чтобы прояснить возможности. Анализ модели \((E)\) достаточно похож на эту, из-за связанных фермионных компонентов, и мы ограничимся краткими комментариями по этой модели в конце раздела. Модель (C) содержит скаляры за пределами Стандартной Модели \(S_{1}\sim(1,1,0)\equiv S\) и \(S_{2}\sim(1,3,2)\equiv\Delta\). Полный потенциал скаляров может быть записан как \[V(H,S,\Delta) = \frac{-\mu^{2}}{2}|H|^{2}+\frac{\tilde{M}_{S}^{2}}{2}S^{2}+\tilde {M}_{\Delta}^{2}\mathrm{Tr}(\Delta^{\dagger}\Delta)+\lambda_{1}|H|^{4}+\lambda _{2}\,[\mathrm{Tr}(\Delta^{\dagger}\Delta)]^{2}\] (7) \[+\lambda_{2}^{\prime}\,\mathrm{Tr}(\Delta^{\dagger}\Delta\Delta^{ \dagger}\Delta)+\frac{\lambda_{3}}{2}S^{4}+\lambda_{4}\,|H|^{2}\,\mathrm{Tr}( \Delta^{\dagger}\Delta)+\lambda_{4}^{\prime}\,H^{\dagger}\Delta^{\dagger}\Delta H\] \[+\lambda_{5}|H|^{2}S^{2}+\lambda_{6}\,S^{2}\,\mathrm{Tr}(\Delta^{ \dagger}\Delta)+\lambda_{\text{\tiny SH}}S\left\{\tilde{H}^{\dagger}\Delta^{ \dagger}H+H^{\dagger}\Delta\tilde{H}\right\},\] где общая фаза \(\Delta\) использовалась для выбора \(\lambda_{\text{\tiny SH}}\) вещественным без потери общности.⁵ Дискретная симметрия \(\{S,\,\Delta\}\rightarrow-\{S,\,\Delta\}\) гарантирует отсутствие смешивания масс между скалярными полями СМ и полями за пределами СМ. Эта симметрия также запрещает линейные термы с одним полем за пределами СМ, такие как \(H\Delta^{\dagger}H\) или \(SH^{2}\), которые в противном случае вызывали бы ненулевую ВЭД для \(\Delta\) и \(S\) после нарушения электрослабой симметрии. Следовательно, существует параметрическое пространство, в котором ни \(S\), ни \(\Delta\) не приобретают ВЭД. Однако скаляр \(S\) и нейтральные компоненты \(\Delta\) будут смешиваться в результате члена \(\lambda_{\text{\tiny SH}}\) в потенциале. [СИМВОЛ3:5][КОНЕЦСИМВОЛА] Раскрывая нейтральный SM-скаляр вокруг его VEV и раскрывая нейтральную компоненту \(\Delta\) как \[H^{0}=\frac{1}{\sqrt{2}}(v+h^{0}+i\chi^{0})\quad\mathrm{и} \quad\Delta^{0}=\frac{1}{\sqrt{2}}(\Delta_{R}+i\Delta_{I}),\] (8) соответственно, массово-смешивающий лагранжиан для нейтральных скаляров записывается в виде \[ \mathcal{L} \supset -\frac{1}{2}\mathcal{S}^{T}\mathcal{M}^{2}\mathcal{S}. \] (9) \[\mathcal{L} \supset -\frac{1}{2}\mathcal{S}^{T}\mathcal{M}^{2}\mathcal{S}.\] (9) Здесь базисный вектор представлен в виде \(\mathcal{S}=(S,\Delta_{R},\Delta_{I})^{T}\), и матрица квадрат-масс имеет следующую форму \[\mathcal{M}^{2}=\left(\begin{array}[]{ccc}\hidden@noalign{}\hfil \tilde{M}_{S}^{2}+\lambda_{5}v^{2}&\frac{\lambda_{\text{\tiny SH}}}{2\sqrt{2}} v^{2}&0\\ \hidden@noalign{}\hfil\frac{\lambda_{\text{\tiny SH}}}{2\sqrt{2}}v^{2}&\tilde{ M}_{\Delta}^{2}+\frac{\lambda_{4}}{2}v^{2}&0\\ \hidden@noalign{}\hfil 0&0&\tilde{M}_{\Delta}^{2}+\frac{\lambda_{4}}{2}v^{2} \end{array}\right).\] (13) Таким образом, CP-нечетный скаляр \(\Delta_{I}\) является собственным состоянием с массой \(\tilde{M}_{\Delta}^{2}+\lambda_{4}v^{2}/2\), в то время как CP-четные скаляры \(S\) и \(\Delta_{R}\) масс-смешиваются, образуя два физических скаляра, которые являются линейными комбинациями этих полей. Предполагаемым кандидатом на темную материю будет одно из собственных состояний масс нейтрального скаляра. Чтобы определить, какое именно состояние, нам необходимо найти массы для смешанных CP-четных состояний. Давайте определим \(M_{S}^{2}=\tilde{M}_{S}^{2}+\lambda_{5}v^{2}\) и \(M^{2}_{\Delta}=\tilde{M}_{\Delta}^{2}+\lambda_{4}v^{2}/2\), которые являются массовыми состояниями CP-четных частиц в пределе \(\lambda_{\text{\tiny SH}}\to 0\). В этом пределе \(\Delta_{R}\) и \(\Delta_{I}\) являются вырожденными и образуют одну комплексную скалярную частицу с массой \(M^{2}_{\Delta}\). При ненулевом значении \(\lambda_{\text{\tiny SH}}\) значения массовых состояний CP-четных частиц могут быть записаны как \[M_{\pm}=\frac{1}{2}\left\{M_{S}^{2}+M^{2}_{\Delta}\pm\left[(M_{S }^{2}-M^{2}_{\Delta})^{2}+\frac{\lambda_{\text{\tiny SH}}^{2}}{2}v^{4}\right]^ {1/2}\right\},\] (14) где собственные состояния связаны с исходными полями как \[\left(\begin{array}[]{c}\hidden@noalign{}\hfil S_{+}\\ \hidden@noalign{}\hfil S_{-}\end{array}\right)=\left(\begin{array}[]{cc} \hidden@noalign{}\hfil\cos\theta&\sin\theta\\ \hidden@noalign{}\hfil-\sin\theta&\cos\theta\end{array}\right)\,\left(\begin{ array}[]{c}\hidden@noalign{}\hfil S\\ \hidden@noalign{}\hfil\Delta_{R}\end{array}\right).\] (21) translated into Russian. Equations are preserved in their original form. Здесь угол смешивания сохраняется в оригинальной форме уравнений. \[\tan 2\theta = \frac{\lambda_{\text{\tiny SH}}v^{2}}{\sqrt{2}(M^{2}_{S}-M^{2}_{ \Delta})}.\] (22) translates to: \[\tan 2\theta = \frac{\lambda_{\text{\tiny SH}}v^{2}}{\sqrt{2}(M^{2}_{S}-M^{2}_{ \Delta})}.\] (22) В пределе, когда скалярный синглет является самым тяжелым, \(M_{S}^{2}\gg M^{2}_{\Delta}\), собственные значения масс примерно равны. \[M_{+}^{2}\simeq M_{S}^{2}+\frac{\lambda_{\text{\tiny SH}}^{2}v^{ 4}}{8M_{S}^{2}}\quad\mathrm{и}\quad M_{-}^{2}\simeq M^{2}_{\Delta}-\frac{ \lambda_{\text{\tiny SH}}^{2}v^{4}}{8M_{S}^{2}}\quad\mathrm{для}\quad M_{S}^{2 }\gg M^{2}_{\Delta}.\] (23) Обратим внимание, что \(M_{-}^{2}<M^{2}_{\Delta}\), что позволяет заключить, что \(S_{-}\) является самым лёгким экзотическим состоянием и, следовательно, кандидатом на тёмную материю. Простые выражения могут быть получены для собственных векторов масс в этом пределе: \[\left.\begin{array}[]{c}\hidden@noalign{}\hfil S_{+}\simeq S+ \frac{\lambda_{\text{\tiny SH}}v^{2}}{2\sqrt{2}M_{S}^{2}}\Delta_{R}\\ \\ \hidden@noalign{}\hfil S_{-}\simeq\Delta_{R}-\frac{\lambda_{\text{\tiny SH}}v^ {2}}{2\sqrt{2}M_{S}^{2}}S\end{array}\right.\quad\quad\mathrm{для}\quad M_{S}^{ 2}\gg M^{2}_{\Delta},\] (27) таким образом, самая легкая скалярная частица \(S_{-}\) в основном состоит из \(\Delta_{R}\). Таким образом, для \(M_{S}^{2}\gg M^{2}_{\Delta}\) темная материя состоит из \(S_{-}\), которая в основном состоит из четной части (\(\Delta_{R}\)) нейтрального поля скалярного трехпроекционного \(S_{2}\). Массовое разделение между \(S_{-}\) и состоянием с нечетной частью \(\Delta_{I}\) составляет \(|\Delta M^{2}|=\lambda_{\text{\tiny SH}}^{2}v^{4}/8M_{S}^{2}\). При условии, что это разделение превышает кинетическую энергию темной материи, \(\sqrt{|\Delta M^{2}|}>\mathrm{KE}_{\text{\tiny DM}}\), состояние \(\Delta_{I}\) не будет доступно кинематически через процессы на деревянном уровне в экспериментах по прямому обнаружению. Это значительно ослабляет ограничения на темную материю, возникающие из сложного трехпроекционного скаляра. Это также дает верхнюю границу на параметр массы \(M_{S}^{2}\), после которой разделение между темной материей и \(\Delta_{I}\) настолько мало, что ожидается рассеяние на деревьях через обмен \(Z\) в экспериментах на современном уровне. Однако обнаруживается \[M_{S} < \frac{1}{2\sqrt{2}}\,\frac{|\lambda_{\text{\tiny SH}}|\,v^{2}}{ \mathrm{KE}_{\text{\tiny DM}}}\ \simeq\ \left(\frac{|\lambda_{\text{\tiny SH}} |}{10^{-2}}\right)\,\left(\frac{10^{-3}}{v_{\text{\tiny DM}}}\right)^{2}\left( \frac{2.6~{}\mathrm{TeV}}{M_{\text{\tiny DM}}}\right)\times 10^{2}~{}\mathrm{ TeV}.\] (28) \[ M_{S} < \frac{1}{2\sqrt{2}}\,\frac{|\lambda_{\text{\tiny SH}}|\,v^{2}}{ \mathrm{KE}_{\text{\tiny DM}}}\ \simeq\ \left(\frac{|\lambda_{\text{\tiny SH}} |}{10^{-2}}\right)\,\left(\frac{10^{-3}}{v_{\text{\tiny DM}}}\right)^{2}\left( \frac{2.6~{}\mathrm{TeV}}{M_{\text{\tiny DM}}}\right)\times 10^{2}~{}\mathrm{ TeV}.\] (28) Таким образом, более тяжелое состояние \(S_{+}\) не может быть произвольно сделано тяжелым, если темная материя должна избежать исключения через, например, данные XENON100. С \(\ M_{\text{\tiny DM}}\equiv M_{-}=\mathcal{O}(\mathrm{TeV})\), разность масс между \(S_{-}\) и \(\Delta_{I}\) меньше, чем \(M_{\Delta}\) (масса \(\Delta_{I}\)). Когда температура падает ниже \(M_{\Delta}\), более тяжелое состояние \(\Delta_{I}\) распадается, и распадные продукты обязательно содержат \(S_{-}\) из-за сохраняющейся дискретной симметрии. Выражение \(|\Delta M^{2}|=\lambda_{\text{\tiny SH}}^{2}v^{4}/8M_{S}^{2}\) показывает, что разность масс между \(S_{-}\) и \(\Delta_{I}\) ограничена как \(\sqrt{|\Delta M^{2}|}\lesssim|\lambda_{\text{\tiny SH}}|\times 4\) ГэВ, учитывая, что для достижения правильного остаточного изобилия требуется \(M_{\Delta}\gtrsim 3\) ТэВ, и мы работаем с \(M_{S}>M_{\Delta}\). В этом диапазоне масс \(\Delta_{I}\) может распадаться как \(\Delta_{I}\to S_{-}+Z^{*}\to S_{-}+\bar{f}f\), где \(f\) является фермионом СМ с массой \(m_{f}<|\lambda_{\text{\tiny SH}}|\times 2\) ГэВ. Поэтому, даже если заряженные фермионы СМ кинематически недоступны, конечное состояние, содержащее нейтрино, будет доступно, если только \(\lambda_{\text{\tiny SH}}\) является исключительно малым. После того, как \(\Delta_{I}\) распадется, адронный плазма состоит из \(S_{-}\) и полей СМ. \(S_{-}\) может поддерживать равновесие с сектором СМ через взаимодействия калибровочных полей и через квартетные члены \(\lambda_{4}\) и \(\lambda_{4}^{\prime}\) (в дальнейшем \(\lambda_{4}\)) в уравнении (7).⁶ Когда квартовые взаимодействия преобладают, модель похожа на модель инертного реального трехлинейного поля; изобилие темной материи будет получено при \(M_{\text{\tiny DM}}\simeq 2.5\) ТэВ, в соответствии с анализом Ref. [11]. При уменьшении \(\lambda_{4}\), разность масс на дереве между заряженными и нейтральными компонентами \(\Delta\) уменьшается, и становятся доступными каналы коаннигиляции, такие как \(\Delta^{-}\Delta^{++}\to W^{+}\gamma\). В этот момент, уменьшение \(\lambda_{4}\) не изменяет необходимую массу темной материи, так как калибровочные взаимодействия преобладают. Анализ Ref. [33] модели инертного комплексного трехлинейного поля показывает, что требуется \(M_{\text{\tiny DM}}\gtrsim 2.8\) ТэВ для всего диапазона параметров.⁷ Мы предполагаем, что \(M_{\text{\tiny DM}}\gtrsim 2.5\) ТэВ будет необходимо даже при преобладании калибровочных взаимодействий над квартовыми взаимодействиями во время фризаута для данной ситуации. При этом значении модель \((C)\) становится жизнеспособной моделью массы нейтрино и темной материи. Однако будет сложно создать экзотики в этой модели, учитывая, что самая легкая масса экзотики составляет \(\gtrsim 2.5\) ТэВ.⁸ [СНОСКА:6][КОНЕЦ СНОСКИ] [ССЫЛКА:7][КОНЕЦ_ССЫЛКИ] [FOOTNOTE:8][ENDFOOTNOTE] - Требуется сохранить формулы. Мы теперь кратко обсудим альтернативный предел с \(M_{S}^{2}\ll M^{2}_{\Delta}\). В этом случае собственные массы для четных ЦП-состояний равны \[M_{+}^{2}\simeq M_{\Delta}^{2}+\frac{\lambda_{\text{\tiny SH}}^{ 2}v^{4}}{8M_{\Delta}^{2}}\quad\mathrm{и}\quad M_{-}^{2}\simeq M^{2}_{S}- \frac{\lambda_{\text{\tiny SH}}^{2}v^{4}}{8M_{\Delta}^{2}}\quad\mathrm{для} \quad M_{\Delta}^{2}\gg M^{2}_{S}.\] (29) Мы видим, что кандидат на темную материю остается самым легким собственным состоянием \(S_{-}\) с четной зарядовостью \(CP\), массой \(M_{-}\). Теперь массовые собственные состояния представляют собой \[\left.\begin{array}[]{c}\hidden@noalign{}\hfil S_{+}\simeq\Delta_ {R}+\frac{\lambda_{\text{\tiny SH}}v^{2}}{2\sqrt{2}M_{\Delta}^{2}}S\\ \\ \hidden@noalign{}\hfil S_{-}\simeq S-\frac{\lambda_{\text{\tiny SH}}v^{2}}{2 \sqrt{2}M_{\Delta}^{2}}\Delta_{R}\end{array}\right.\quad\quad\mathrm{для}\quad M _{\Delta}^{2}\gg M^{2}_{S},\] (33) таким образом, большая часть тёмной материи состоит в основном из синглет-скаляра \(S\). Синглет-скалярная тёмная материя хорошо известна [9], и подробный анализ показывает, что \(M_{\text{\tiny DM}}\gtrsim 80\) ГэВ совместима с ограничениями на непосредственное обнаружение и данными WMAP для массы Хиггса \(m_{h}\simeq 125\) ГэВ [37]. Пригодная область параметров может быть исследована с помощью XENON1T, за исключением небольшого резонансного окна с \(M_{\text{\tiny DM}}\simeq 62\) ГэВ, где связь между тёмной материей и Хиггсом может быть очень слабой. Более лёгкая тёмная материя с \(M_{\text{\tiny DM}}\lesssim 60\) ГэВ отвергается ограничениями LHC на наблюдаемый распад Хиггса [37]. Мы видим, что модель (С) имеет жизнеспособное пространство параметров, в котором она ведет себя как модель инертного трехкомпонентного состояния или модель инертного синглета. Этот анализ достаточен, чтобы показать, что у модели (Е) также есть жизнеспособные области пространства параметров. В модели (Е) первый скаляр имеет вид \(S_1 \sim (1,3,0)\), в то время как второй скаляр остается \(S_2 \sim (1,3,2)\). Скалярный потенциал для этой модели содержит член, аналогичный члену \(\lambda_{\text{\tiny SH}}\) в уравнении (7), который смешивает нейтральную компоненту \(S_1\) с CP-четной нейтральной компонентой \(S_2\). Если \(S_2\) является самым тяжелым, модель ведет себя как модель инертного вещественного трехкомпонентного состояния, в то время как если \(S_1\) является самым тяжелым, самый легкий скаляр в основном состоит из \(\Delta_R\) (нейтральная CP-четная часть \(S_2\)). Ограничения на прямое обнаружение могут быть обойдены из-за массового смешивания, и модель снова является эффективной моделью инертного трехкомпонентного темного вещества. В обоих случаях мы ожидаем, что достижимая плотность темного вещества и достижимые массы нейтрино могут быть получены, хотя темное вещество будет тяжелым, с \(M_{\text{\tiny DM}}\gtrsim 2.5\) ТэВ (не учитывая резонансные области). Определение перспектив наблюдений для мультиплетов за пределами Стандартной Модели на Большом Адронном Коллайдере (ЛАК) с использованием ограничения \(M_{\Delta}^{2}\gg M^{2}_{S}\) в модели \((C)\), кажется наиболее оптимистичным сценарием для моделей, перечисленных в Таблице 1 (за исключением областей резонанса, которые также позволяют более лёгкие поля). В этом пределе тёмная материя может быть относительно лёгкой, \(M_{\text{\tiny DM}}\simeq 100\) ГэВ, и, следовательно, экзотические состояния \(\Delta\) и \(\mathcal{F}\) могут быть одновременно порядка нескольких сотен ГэВ. В принципе, возможно наблюдение всех трёх мультиплетов за пределами Стандартной Модели в этом пределе. Для других подходящих моделей в Таблице 1 тёмная материя должна быть относительно тяжелой: \(M_{\text{\tiny DM}}\gtrsim 500\) ГэВ для моделей инертного двойноготета и \(M_{\text{\tiny DM}}\gtrsim 2.5\) ТэВ для моделей инертного тройноготета. Это выходит за пределы области доступности дополнительных полей за пределами Стандартной Модели на ЛАК. Обратите внимание, что любая массовая вырожденность между заряженными и нейтральными членами инертного мультиплета на дереве устраняется радиационными эффектами, делая заряженные компоненты тяжелее нейтральных компонентов. Более тяжелые члены заданного мультиплета могут распадаться на более легкие члены того же мультиплета через слабые взаимодействия, например \(\mathcal{F}^{+}\to W^{+}+\mathcal{F}^{0}\), где \(W\) может быть виртуальным. Более тяжелый мультиплет также может распадаться на более легкий мультиплет через юкавское спаривание; например \(\mathcal{F}^{-}\to S^{0}+\ell^{-}\), если \(M_{\mathcal{F}}\gg M_{S}\). Из-за дискретной симметрии новые поля могут быть производимыми парами на коллайдерах, и сохранение заряда \(Z_{2}\) означает, что конечные состояния, возникающие из экзотических цепочек распада, обязательно включают стабильные электрически нейтральные поля, которые будут уклоняться от детектора. ## 6 Об истоке дискретной симметрии Возможно сделать следующее предположение, аналогичное предложению Ма: мы используем дискретную симметрию \(Z_{2}\), чтобы обеспечить стабильность наиболее легкого поля в диаграмме массы нейтрино вне Стандартной Модели. Можно возразить, что использование дискретной симметрии не полностью удовлетворяет, поскольку оно кажется добавочным, или из-за того, что расчеты квантовой гравитации не должны сохранять глобальные симметрии. Это мотивирует нас рассмотреть возможность найти простое объяснение для дискретной симметрии. Самая простая возможность - заменить дискретную симметрию на калиброванную симметрию \(U^{\prime}(1)\), которая не будет нарушаться квантовыми эффектами гравитации. С достаточным количеством дополнительных компонентов, вероятно, можно достичь этой цели для всех моделей, о которых мы говорили. Однако мы хотели бы знать, какие модели допускают минимальное расширение, такое что \(Z_{2}\) превращается в \(U^{\prime}(1)\), и в частицовом спектре добавляется только один скалярный синглет \(\eta\), чтобы нарушить симметрию \(U^{\prime}(1)\). Представляя полную калибровочную группу как \(\mathcal{G}_{\text{\tiny SM}}\times U(1)^{\prime}\), где \(\mathcal{G}_{\text{\tiny SM}}\) - это калибровочная группа стандартной модели, у нас имеются следующие свойства преобразования для полей за пределами Стандартной модели⁹ [FOOTNOTE:9][ENDFOOTNOTE] \[\eta\sim(1_{\text{\tiny SM}},Q_{\eta})\,,\quad S_{1,2}\sim(Q^{ \text{\tiny SM},}_{1,2},Q)\,,\quad\mathcal{F}\sim(Q^{\text{\tiny SM}}_{ \mathcal{F}},-Q),\] (34) \[\eta \sim (1_{\text{\tiny SM}}, Q_{\eta})\,,\quad S_{1,2} \sim (Q^{\text{\tiny SM}}_{1,2},Q)\,,\quad \mathcal{F} \sim (Q^{\text{\tiny SM}}_{\mathcal{F}}, -Q),\] (34) где верхний индекс "SM" обозначает заряды в рамках \(\mathcal{G}_{\text{\tiny SM}}\), приведенные в Таблице 1. Из анализа уравнения (3) видно, что все слагаемые Лагранжиана, необходимые для генерации массы нейтрино, допустимы симметрией \(U^{\prime}(1)\). Однако в случае моделей \((A)\) и \((B)\), являющихся моделями с инертным дуплетом, усиленная симметрия препятствует дополнительному слагаемому \((S_{1}^{\dagger}H)^{2}\). Это слагаемое необходимо для разделения нейтральных компонентов \(S_{1}\sim(1,2,1)\) и для избегания ограничений прямого обнаружения [10]. Таким образом, модели \((A)\) и \((B)\) несовместимы с этим минимальным расширением симметрии. С другой стороны, мы обнаружили, что модели \((C)\), \((D)\) и \((E)\), в которых один скаляр образует действительное представление симметрии калибровки СМ, остаются жизнеспособными моделями тёмной материи при условии \(Q_{\eta}=-2Q\). Это соотношение необходимо для снятия массовой вырожденности нейтральных полей за пределами СМ. Например, рассмотрим модель \((C)\), которая теперь имеет следующие члены в потенциале скаляра. \[V(H,S,\Delta,\eta) \supset \lambda_{\text{\tiny SH}}\left\{S\tilde{H}^{\dagger}\Delta^{ \dagger}H+S^{*}H^{\dagger}\Delta\tilde{H}\right\}\ +\ \frac{\mu_{\eta}}{2} \left\{S^{2}\eta+S^{*2}\eta^{\dagger}\right\},\] (35) в дополнение к терминам в уравнении (7). Все остальные термины, содержащие \(\eta\), зависят только от модуля \(|\eta|^{2}\). Мы использовали относительную фазу \(S\) и \(\Delta\) для выбора \(\lambda_{\text{\tiny SH}}\) вещественным и фазы \(S\) для выбора \(\mu_{\eta}\) вещественным. Обратите внимание, что нарушение симметрии \(U^{\prime}(1)\to Z_{2}\) достигается ненулевым значением \(\langle\eta\rangle\), что побуждает использовать дискретную симметрию как случайную подгруппу калибровочной симметрии \(U^{\prime}(1)\). В базисе \(\mathcal{S}=(S_{R},\Delta_{R},S_{I},\Delta_{I})^{T}\) матрица квадрата массы имеет следующий вид¹⁰ [FOOTNOTE:10][ENDFOOTNOTE] (SEE LAST LINE FOR TRANSLATION) \[\mathcal{M}^{2}=\left(\begin{array}[]{cccc}\hidden@noalign{}\hfil \tilde{M}_{S}^{2}+\lambda_{5}v^{2}+2\mu_{\eta}\langle\eta\rangle&\frac{\lambda _{\text{\tiny SH}}}{4}v^{2}&0&0\\ \hidden@noalign{}\hfil\frac{\lambda_{\text{\tiny SH}}}{4}v^{2}&\tilde{M}_{ \Delta}^{2}+\frac{\lambda_{4}}{2}v^{2}&0&0\\ \hidden@noalign{}\hfil 0&0&\tilde{M}_{S}^{2}+\lambda_{5}v^{2}-2\mu_{\eta} \langle\eta\rangle&\frac{\lambda_{\text{\tiny SH}}}{4}v^{2}\\ \hidden@noalign{}\hfil 0&0&\frac{\lambda_{\text{\tiny SH}}}{4}v^{2}&\tilde{M}_ {\Delta}^{2}+\frac{\lambda_{4}}{2}v^{2}\end{array}\right).\] (40) Обратите внимание, что записи для состояний с чётным и нечётным зарядами CP идентичны в пределе \(\mu_{\eta}\to 0\). Это приведёт к вырожденным состояниям, так что в случае, когда тёмная материя состоит в основном из \(\Delta\), тёмная материя будет исключена по данным эксперимента XENON100 (она будет инертной комплексной тройкой). Однако, при ненулевом \(\mu_{\eta}\) состояния с чётным и нечётным зарядами CP становятся не вырожденными, и достигается возможность существования тёмной материи. Когда тёмная материя состоит в основном (или полностью, для модели \((D)\)) из вещественного представления \(\mathcal{G}_{\text{\tiny SM}}\), разделение, достигаемое ненулевым \(\mu_{\eta}\), также обеспечивает подавление сигналов прямого обнаружения, возникающих из смешивания между \(Z^{\prime}\) и \(Z\).¹¹ [СНОСКА:11][КОНЕЦ СНОСКИ] Будет дополнительный рассеивающий процесс для темной материи из-за смешивания между \(\eta\) и скалярным полем СМ, что приводит к типичному взаимодействию через портал Хиггса. Учитывая, что связь для этого взаимодействия необходима для достижения наблюдаемой обили темной материи, всегда можно выбрать такое значение этой связи, чтобы оно было достаточно малым для соблюдения ограничений. Таким образом, с помощью этого простого расширения калибровки мы можем объяснить происхождение дискретной симметрии для моделей \((C)\), \((D)\) и \((E)\), сохраняя желательные свойства радиационной массы нейтрино и жизнеспособной обили темной материи. Обратите внимание, что исходное предложение Ма [3] и вариант, использующий реальный тройной фермион [4], не совместимы с этим минимальным улучшением симметрии; в случае скалярной темной материи термин \((S^{\dagger}H)^{2}\) исключен, что означает, что прямые эксперименты по обнаружению исключают модель, аналогично минимальному расширению моделей с калибровочным полем \((A)\) и \((B)\). Кроме того, возникают гауссовы аномалии, так как \(\mathcal{F}_{R}\) является хиральным полем в работах [3, 4] - для объяснения происхождения симметрии \(Z_{2}\) в этих случаях требуется дополнительное моделирование. Мы не затрагиваем этот вопрос здесь. ## 7 За пределами присоединенного представления В предыдущих разделах мы изучали обобщения модели Ма 2006 года с радиационной массой нейтрино и стабильными кандидатами на тёмную материю. При этом мы ограничились вниманием на мультиплеты размером не больше адъюнктного представления. Как уже упоминалось, можно получить массу нейтрино с помощью Рисунка 1 и получить кандидатов на тёмную материю в моделях с большими мультиплетами. Мы кратко обсудим этот вопрос в настоящем разделе. Сначала рассмотрим случай с включением массы Майораны, как показано на Рисунке 2. Разрешая представления \(SU(2)\) такие, как пятикратное представление, мы находим еще две модели. Оба из них используют четверкое скалярное поле \(S\sim(1,4,1)\), при этом реальное фермионное поле может быть либо трехкратным \(\mathcal{F}\sim(1,3,0)\), либо пятикратным \(\mathcal{F}\sim(1,5,0)\). Последняя модель была подробно рассмотрена в работе [38]. В обоих моделях возможно наличие как скалярной, так и фермионной темной материи, как и в первоначальном предложении Ма; нейтральное фермионное поле не связано с бозоном \(Z\) и, следовательно, может оставаться согласованным с ограничениями прямого обнаружения. [Сноска:12][КОНЕЦСНОСКИ] Обобщая модели с вставкой массы фермионов типа Дирака (т.е. обобщая модели в таблице 1), мы обнаруживаем, что возможны и другие варианты. Для полноты мы перечисляем их в приложении, но здесь мы предлагаем следующие комментарии. Как и в моделях в таблице 1, мы обнаруживаем, что для всех моделей с более крупными калибровочными представлениями фермионная темная материя может быть исключена. Фермионы остаются псевдо-дираковскими частицами с малыми разделениями, определенными массами нейтрино СМ. Такие малые разделения позволяют неограниченное рассеяние на уровне дерева с детекторами СМ через обмен бозоном \(Z\), что исключается экспериментом XENON100. Мы таким образом исключаем область параметров, в которой фермион является самым легким состоянием вне СМ, по тем же причинам, что и обсуждалось в разделе 3. В случае скалярной темной материи необходимо рассмотреть отдельные модели, как это требовалось для моделей в таблице 1. Некоторые модели могут быть сразу исключены по тем же причинам, по которым можно исключить модели \((F)\) и \((G)\); например, модель \((L)\) в таблице 2 может быть исключена, так как она дает невзаимодействующую комплексную модель тройтета. Аналогично, модель \((R)\) в таблице 3 исключается, так как \(S_{2}\) содержит только заряженные компоненты, а \(S_{1}\) имеет ненулевой гиперзаряд. Другие модели, кажется, совместимы с ограничениями прямого обнаружения, при условии смешивания нейтральных компонент скаляров, когда самый легкий скаляр имеет ненулевой гиперзаряд, подобно тому, как модели \((C)\) и \((E)\) являются жизнеспособными. Например, модель \((M)\) содержит \(S_{1}\sim(1,4,1)\) как единственный скаляр вне СМ с нейтральной компонентой. Однако в Лагранжиане существует член \(\lambda(S_{1}^{\dagger}H)^{2}\subset\mathcal{L}\), который может разделить компоненты нейтрального скаляра, позволяя избежать ограничений прямого обнаружения (это аналогично разделению, полученному в модели инертного двойноготта). Наконец, мы отмечаем, что использование более крупных мультиплетов может иметь дополнительное феноменологическое преимущество. Ссылаясь на работу [5], показывается, что большие мультиплеты, которые опосредуют взаимодействия между темной материей и Стандартной Моделью, могут усиливать процессы аннигиляции темной материи с помощью петель в \(2\gamma\) и \(\gamma+Z\) конечные состояния, без необходимости в непертурбативно больших связях. Это происходит потому, что большие мультиплеты позволяют полям иметь большие электрические заряды, естественно усиливая процессы с петлями и фотонами в конечном состоянии. Кажется, что в моделях, представленных в Приложении, невозможно реализовать астрофизический сигнал гамма-лучей [6], но простые расширения кажутся совместимыми с этой идеей. Например, модель \((N)\) в Таблице 3 использует \(\mathcal{F}\sim(1,4,-1)\), \(S_{1}\sim(1,5,0)\) и \(S_{2}\sim(1,5,2)\). Когда \(S_{1}\) является самым легким за пределами Стандартной Модели состоянием, темная материя состоит (в основном) из нейтральной компоненты \(S_{1}\). В однопетлевом вкладе в процессы типа \(DM+DM\to 2\gamma,\gamma+Z\) имеются виртуальные состояния \(S_{2}\) в петле, которые усиливаются за счет наличия многократно заряженной компоненты в \(S_{2}\). Заметим, что масса темной материи должна быть либо \(M_{\text{\tiny DM}}\sim 130\) ГэВ, либо \(M_{\text{\tiny DM}}\sim 144\) ГэВ, чтобы генерировать избыток гамма-лучей через аннигиляции темной материи в \(2\gamma\) или \(\gamma+Z\) конечные состояния соответственно. Однако, ожидается, что темная материя, состоящая из \(S_{1}\sim(1,5,0)\), должна иметь массу \(M_{\text{\tiny DM}}\gtrsim 5\) ТэВ, чтобы достичь правильной избыточности [11], что слишком велико для объяснения астрофизического сигнала. Если добавить синглетный скаляр \(S\), который также является нечетным относительно симметрии \(Z_{2}\), в модель, то область параметров, где \(S\) является темной материей, совместима с \(M_{\text{\tiny DM}}\sim 130\) ГэВ или \(M_{\text{\tiny DM}}\sim 144\) ГэВ. Компоненты \(S_{1}\) и/или \(S_{2}\) могут быть \(\mathcal{O}(100)\) ГэВ, и петлевые процессы, поддерживаемые в работе [5], присутствуют в модели, таким образом усиливая астрофизический сигнал гамма-лучей. В этом примере имеется простая связь между астрофизическим сигналом и механизмом массы нейтрино, крупные мультиплеты, которые обеспечивают последнюю, также усиливают первую. Было бы интересно продолжить развивать эти идеи, чтобы увидеть, может ли темная материя быть реализована одним из полей в диаграмме петель нейтрино, а не дополнительной степенью свободы, или исследовать феноменологию только что описанной модели. [FOOTNOTE:13][ENDFOOTNOTE] Отобразить примечание и конец примечания в исходном виде. ## 8 Заключение Мы изучали класс моделей с радиационной массой нейтрино и стабильными кандидатами на темную материю. Масса нейтрино была получена с помощью диаграммы однократного рассеяния с той же топологией, что и предложил Ма [3]. Мы обобщили подход Ма, детализируя все варианты с полем выше Стандартной Модели, не превышающим адъюнктное представление. В случае, когда в диаграмме массы нейтрино присутствовало массовое включение Майораны, было найдено только две модели, обе из которых были известны. Когда включение массы было типа Дирака, так что симметрия числа лептонов была нарушена вершиной, было найдено несколько дополнительных моделей. Фермионная темная материя была исключена во всех этих моделях, в то время как две из них были полностью исключены из-за ограничений прямого обнаружения. Оставшиеся модели позволяли радиационную массу нейтрино и достигали жизнеспособного избытка (скалярной) темной материи. Были случаи с инертным синглетом, инертным двойником и инертным тройником, обеспечивающие естественную среду для моделей инертных темной материи N-кортежей с дополнительной возможностью достижения радиационной массы нейтрино. Интересно, что некоторые из моделей позволяли простое расширение, при котором (ранее наложенная) дискретная симметрия возникала как случайная симметрия на низких энергиях. Мы кратко обсудили модели с более крупными полевыми кратностями выше Стандартной Модели, показывая, что существуют жизнеспособные сценарии. С помощью простых расширений большие кратности, позволяющие массу нейтрино, также могут усиливать сигналы гамма-излучения астрофизических источников в настоящее время, позволяя простую связь между механизмом массы нейтрино и сигналом астрофизического гамма-излучения. ## Подтверждения Авторы благодарят Y. Kajiyama, K. Nagao, H. Okada, T. Schwetz, A. Strumia и K. Yagyu. SSCL частично поддерживается NSC в рамках гранта № NSC-101-2811-M-006-015 и частично - NCTS Тайваня. KM поддерживается Австралийским советом по научным исследованиям. ## Приложение А Масса без симметрии \(\mathbf{Z_{2}}\) Из моделей, представленных в Таблице 1, только модели \((A)\), \((B)\) и \((D)\) ожидаются создать (доминирующие) радиационные массы нейтрино в отсутствие симметрии \(Z_{2}\). Другие модели содержат тройной скаляр \(S_{1,2}\sim(1,3,2)\), который акоплирует по Юкаве с лептонами СМ и приобретает ВЭО за счет термина \(\mu HS_{1,2}H\subset V(H,S_{1},S_{2})\) в отсутствие дискретной симметрии. Таким образом, ожидается, что массы нейтрино на уровне дерева стандартной формы смешивания типа II [40] будут превалировать над эффектом петли, когда симметрия \(Z_{2}\) отклоняется. В случае моделей \((A)\), \((B)\) и \((D)\), с другой стороны, массы нейтрино на уровне дерева не возникают, если симметрия \(Z_{2}\) устранена, тогда как диаграмма петли на Рисунке 1 сохраняется. Обратите внимание, что если симметрия \(Z_{2}\) отключена, фермион \(\mathcal{F}\) не требуется для создания ненулевых радиационных масс нейтрино в модели \((D)\)[41]. Однако спектр, полученный без \(\mathcal{F}\), имеет простую форму Zee [41], которая несовместима с наблюдаемой смешивающей структурой [42]. Таким образом, фермион \(\mathcal{F}\sim(1,2,-1)\) необходим для получения "жизнеспособной" смешивающей структуры в отсутствие симметрии \(Z_{2}\). [ССЫЛКА В СНОСКУ:14][КОНЕЦ ССЫЛКИ] Модель \((B)\) имеет схожий состав частиц с моделью, представленной в ссылке [43], за исключением замены \(S_{2}\sim(1,2,3)\to S_{2}\sim(1,4,3)\). Это различие исключает древесную массу, найденную в ссылке [43], поэтому модель \((B)\) является чисто моделью радиационных масс, которую можно изучать без дискретной симметрии и тёмной материи.¹⁵ [FOOTNOTE:15][ENDFOOTNOTE] (ru) ## Приложение Б Модели с более крупными мультиплетами В тексте мы нашли семь моделей, которые используют мультиплеты за пределами СМ либо в фундаментальном, либо в присоединенном представлении \(SU(2)_{L}\) и имеют внутреннюю вставку массы Дирака. Помимо этих минимальных моделей, можно реализовать радиационную массу нейтрино и кандидатов в темную материю с использованием больших мультиплетов. Мы представляем дополнительные минимальные модели, возникающие, если разрешить использование мультиплетов, формирующих квадруплетное ( изоспин \(3/2\)) представление \(SU(2)_{L}\) в Таблице 2. Схема маркировки следует из Таблицы 1 в тексте. Если разрешить использование мультиплетов, состоящих из пяти частиц, возникают дополнительные модели, показанные в Таблице 3 (см. также Ссылку [44] для детального примера). Первый случай, указанный как модель \((M)\), был представлен в Ссылке [15]. [TABLE:A2.T2][ENDTABLE] [TABLE:A2.T3][ENDTABLE] ## Ссылки * [1] M. C. Гонсалес-Гарсия, M. Малтони, J. Сальвадо и T. Швец, JHEP **1212**, 123 (2012) [arXiv:1209.3023 [hep-ph]]. * [2] A. H. G. Peter, arXiv:1201.3942 [astro-ph.CO]. * [3] E. Ma, Phys. Rev. D **73**, 077301 (2006) [hep-ph/0601225]. * [4] E. Ma и D. Suematsu, Mod. Phys. Lett. A **24**, 583 (2009) [arXiv:0809.0942 [hep-ph]]. * [5] J. Kopp, E. T. Neil, R. Primulando и J. Zupan, Phys. Dark. Univ. **2** (2013) 22 [arXiv:1301.1683 [hep-ph]]. * [6] C. Weniger, JCAP **1208**, 007 (2012) [arXiv:1204.2797 [hep-ph]]. * [7] L. M. Краусс, S. Nasri и M. Тродден, Phys. Rev. D **67**, 085002 (2003) [hep-ph/0210389]. * [8] C. Boehm, Y. Фарзан, T. Hambye, S. Palomares-Ruiz и S. Pascoli, Phys. Rev. D **77**, 043516 (2008) [hep-ph/0612228]; M. Lindner, D. Schmidt и T. Швец, Phys. Lett. B **705**, 324 (2011) [arXiv:1105.4626 [hep-ph]]; F. -X. Josse-Michaux и E. Molinaro, Phys. Rev. D **84**, 125021 (2011) [arXiv:1108.0482 [hep-ph]]; S. Канемура, Т. Набэсима и Х. Сугияма, Phys. Rev. D **85**, 033004 (2012) [arXiv:1111.0599 [hep-ph]]; M. Gustafsson, J. M. Но и M. A. Rivera, arXiv:1212.4806 [hep-ph]; M. Аоки, J. Кубо и H. Такано, arXiv:1302.3936 [hep-ph]; Y. Каджияма, H. Окада и Т. Тома, arXiv:1303.7356 [hep-ph]. * [9] V. Сильвейра и A. Зи, Phys. Lett. B **161**, Переведите маркдаун с английского на русский. Сохраните формулы:
"# On the nature of Seyfert galaxies with high [OIII]5007 blueshifts\n\nS. Komossa\n\nMax-Planck-Ins(...TRUNCATED)
"# О природе галактик Сейферта с высокими смещениями в (...TRUNCATED)
"# New analytic models of ”traversable”  wormholes\n\n###### Abstract\n\nThe analytic solution(...TRUNCATED)
"# Новые аналитические модели \"проходимых\" червей\n###### (...TRUNCATED)
"# Properties of Black Hole Radiation from Tunnelling\n\nTimothy Clifton\n\n_Department of Astrophys(...TRUNCATED)
"# Свойства излучения черной дыры от туннелирования\n\nЧ(...TRUNCATED)
"# The role of the \\(\\Delta(1232)\\)-resonance in covariant baryon chiral perturbation theory\n\nJ(...TRUNCATED)
"# Роль резонанса \\(\\Delta(1232)\\) в ковариантной пертурбаци(...TRUNCATED)
"# Evolutionary Stochastic Search\n\nfor Bayesian model exploration\n\nLeonardo Bottolo\n\nInstitute(...TRUNCATED)
"# Эволюционный стохастический поиск\nдля исследования(...TRUNCATED)
"# Blueshift of the surface plasmon resonance in silver nanoparticles: substrate effects\n\nSøren R(...TRUNCATED)
"Пересчет смещения длин волн поверхностного плазмонно(...TRUNCATED)
"# On the arithmetic of the BC-system\n\nAlain Connes\n\nCaterina Consani\n\nA. Connes: Collège de (...TRUNCATED)
"# Об арифметике BC-системы\n\nBC-система является одной из(...TRUNCATED)
README.md exists but content is empty. Use the Edit dataset card button to edit it.
Downloads last month
2
Edit dataset card