Asymmetric Dark Matter and Baryogenesis from SU(2)_{\ell}
Bartosz Fornal, Yuri Shirman, Tim M. P. Tait, Jennifer Rittenhouse West
UUCI-HEP-TR-2016-20
Asymmetric Dark Matter and Baryogenesis from SU (2) (cid:96) Bartosz Fornal,
1, 2
Yuri Shirman, Tim M. P. Tait, and Jennifer Rittenhouse West Department of Physics and Astronomy, University of California, Irvine, CA 92697, USA Department of Physics, University of California, San Diego, 9500 Gilman Drive, La Jolla, CA 92093, USA (Dated: July 11, 2017)We propose a theory in which the Standard Model gauge symmetry is extended by a new SU (2) (cid:96) groupacting nontrivially on the lepton sector which is spontaneously broken at the TeV scale. Under this SU (2) (cid:96) the ordinary leptons form doublets along with new lepton partner fields. This construction naturally contains adark matter candidate, the partner of the right-handed neutrino, stabilized by a residual global U (1) χ symmetry.We show that one can explain baryogenesis through an asymmetric dark matter scenario, in which generationof related asymmetries in the dark matter and baryon sectors is driven by the SU (2) (cid:96) instantons during a firstorder phase transition in the early universe. I. INTRODUCTION
Evidence for the existence of dark matter (DM) composing roughly of the matter in the Universe is extremely compellingand includes measurements of galactic rotation curves, gravitational lensing, cosmic microwave background anisotropy and X -ray emission from elliptical galaxies. The Standard Model (SM) of particle physics in its current form does not account for thepresence of DM. However, if DM couples appreciably to visible matter, it is reasonable to expect that it represents an extensionof the SM built on similar principles, i.e. those of a gauge theory. One such possibility is to postulate a larger gauge symmetrygroup. Ideally, one would like this expanded symmetry to play a role in explaining why DM is stable on the scale of the age ofthe Universe, such that it does not decay into SM states.Another problem associated with the SM is its inability to explain the current matter-antimatter asymmetry of the Universe.Two of the Sakharov conditions [1], the requirement of a first order phase transition and sufficient amount of CP violation, arelikely not fulfilled in the SM alone. There are many ideas for how to fix this by introducing new fields in the theory. Similarlyto the approach used for the DM problem, one class of potential solutions relies on extensions of the Standard Model gaugegroup. For example, it was shown [2] that the breakdown of a new gauge group can overcome the SM difficulties and provide aframework for successful baryogenesis.It is tempting to combine these ideas, positing that the DM interactions with the SM are responsible both for generating itsabundance as well as generating the observed baryon asymmetry. This is the philosophy behind asymmetric dark matter (ADM)models [3–8], in which the DM and baryon asymmetries are intimately related. A typical feature of these models is that thenatural scale for the DM mass is ∼ GeV , which requires light messengers in order to realize a large enough DM annihilationcross section to annihilate away the symmetric component. In this paper we construct an ADM model that naturally containssuch particles.As a starting point it is natural to consider gauged extensions of the SM global symmetries. The possibility of gauging leptonand baryon number was considered in [9–14]. Unfortunately, models based on this approach offer only a limited possibilityof explaining the primordial baryon asymmetry [15–17]. More recently, a baryogenesis mechanism based on a non-Abelianextension of the SM baryon number and color in SU (4) gauge group was proposed in [18, 19]. Such an extension successfullyunifies DM with the SM baryons, but ultimately relies on unspecified UV physics represented by higher-dimensional operatorsto generate the asymmetry.In this paper we follow the general approach of [19] and extend the SM gauge group by an additional SU (2) (cid:96) gauge symmetryunder which the SM leptons transform nontrivially, promoting them to SU (2) (cid:96) doublets along with additional partner fields. Alepton number assignment is extended to the partner fields, thus generalizing the SM lepton number. The SU (2) (cid:96) symmetry isspontaneously broken via a vacuum expectation value (vev) of a new leptonic Higgs field Φ . All new matter fields introduced inthe model obtain vector-like masses after SU (2) (cid:96) breaking. The lightest of these particles will be stable due to residual globalsymmetries and provides a dark matter candidate of ADM framework. During the SU (2) (cid:96) phase transition, SU (2) (cid:96) sphaleronsgenerate both DM and lepton number, with the latter later converted to baryon number by electroweak sphalerons. A similar useof sphalerons is described in the aidnogenesis scenario [20]. II. MODEL
We extend the gauge symmetry of the Standard Model to: SU (3) c × SU (2) W × U (1) Y × SU (2) (cid:96) . (1) a r X i v : . [ h e p - ph ] J u l Field SU (2) (cid:96) SU (2) W U (1) Y ˆ l L = (cid:32) l L ˜ l L (cid:33) − / e R = (cid:32) e R ˜ e R (cid:33) − ν R = (cid:32) ν R ˜ ν R (cid:33) l (cid:48) R − / e (cid:48) L − ν (cid:48) L , Φ TABLE I: Fields and their representations under the gauge symmetries SU (2) (cid:96) × SU (2) W × U (1) Y . The SM leptons reside in upper components of SU (2) (cid:96) doublets, while the new fermions (denoted by a tilde) reside in lowercomponents of SU (2) (cid:96) doublets: ˆ l L ≡ (cid:18) l L ˜ l L (cid:19) , ˆ e R ≡ (cid:18) e R ˜ e R (cid:19) , ˆ ν R ≡ (cid:18) ν R ˜ ν R (cid:19) . (2)To maintain cancellation of SU (2) W × U (1) Y anomalies and allow new fermions to acquire Dirac masses after SU (2) (cid:96) sym-metry breaking we also introduce a set of leptons that are neutral under SU (2) (cid:96) , l (cid:48) R , e (cid:48) L , ν (cid:48) L . (3)Finally, we introduce two Higgs doublets charged under SU (2) (cid:96) . The second doublet allows us to introduce two ingredientsnecessary for a realistic ADM model: sufficient CP violation to catalyze the production of an adequate baryon asymmetry tomatch observation, and an annihilation channel allowing for a sufficient annihilation of the symmetric component of the DM.The quantum numbers for all the relevant particles, including two SU (2) (cid:96) Higgs doublets (discussed below) are summarized inTable I.To allow for a spontaneous breaking of SU (2) (cid:96) , we introduce the following scalar potential for SU (2) (cid:96) doublets Φ i : V (Φ , Φ ) = m | Φ | + m | Φ | + ( m Φ † Φ + h . c . ) + λ | Φ | + λ | Φ | + λ | Φ | | Φ | + λ | Φ † Φ | + (cid:104) ˜ λ Φ † Φ | Φ | + ˜ λ Φ † Φ | Φ | + ˜ λ (Φ † Φ ) + h . c . (cid:105) (4)where we have imposed a U (1) symmetry (discussed below) to ensure that the dark matter is ultimately stable. In addition,while we do not explicitly show Φ i interactions with the SM Higgs, such interactions are generically present at tree level andare induced radiatively even if not present. This interaction plays an important role in allowing the lightest component of Φ todecay through induced mixing with the SM Higgs, with constraints discussed in Sec. IV.The potential contains four complex parameters: m , ˜ λ , ˜ λ , and ˜ λ . For generic parameters, one phase can be rotated awayby redefining the phase of the combination Φ † Φ (the only combination appearing in the potential), leaving three physical phasecombinations [21].It is easy to choose parameters so that SU (2) (cid:96) is completely broken by Φ , vevs. The potential, Eq. (4), is structurallyidentical to a two Higgs doublet model (with the global U (1) playing the role of the SM’s gauged U (1) Y hypercharge) andadmits a similarly rich array of mass eigenstates for the physical bosons. The vacuum can be parameterized by v (cid:96) = (cid:112) v + v and tan β = v /v , where v and v are the vevs of the two doublets, respectively, (cid:104) Φ (cid:105) = √ (0 , v ) T and (cid:104) Φ (cid:105) = √ (0 , v ) T .There is a spectrum of five physical scalar Higgs bosons which are mixtures of the original CP-even and CP-odd components ofthe doublets Φ and Φ .The Yukawa interactions consistent with the gauge symmetries are given by L Y = (cid:88) i (cid:16) Y abl ¯ˆ l aL Φ i l (cid:48) bR + Y abe ¯ˆ e aR Φ i e (cid:48) bL + Y abν ¯ˆ ν aR Φ i ν (cid:48) bL (cid:17) + y abe ¯ˆ l aL H ˆ e bR + y abν ¯ˆ l aL ˜ H ˆ ν bR + y (cid:48) abe ¯ l (cid:48) aR H e (cid:48) bL + y (cid:48) abν ¯ l (cid:48) aR ˜ H ν (cid:48) bL + h . c . , (5) SymmetriesExact & Approximate Lepton Basis Low EnergyField U (1) U (1) U (1) (cid:48) U (1) U (1) L U (1) χ U (1) D U (1) L U (1) χ ˆ l L = (cid:32) l L ˜ l L (cid:33) −
11 1 0ˆ e R = (cid:32) e R ˜ e R (cid:33) −
11 1 0ˆ ν R = (cid:32) ν R ˜ ν R (cid:33) − −
11 0 1 l (cid:48) R e (cid:48) L ν (cid:48) L − i , i = 1 , − − −
20 0 0
TABLE II: Charges under global U (1) symmetries. The first two columns represent charges under exact global symmetries of the Lagrangian.The next four columns represent charges under exact and approximate global symmetries when y ν and y (cid:48) ν can be neglected. The last threecolumns represent charges under exact and approximate global symmetries in low energy physics. where a and b are flavor indices. After electroweak symmetry breaking the Yukawa matrices y abe and y abν lead to the usual leptonmass matrices (with Dirac neutrino masses). The newly introduced Yukawa matrices Y l , Y e , Y ν , y (cid:48) e are responsible for generationof Dirac mass terms between the lepton partners and spectators after SU (2) (cid:96) breaking. In addition, the full mass matrix for newfields includes mixing between electroweak singlets and doublets induced by the SM Higgs vev: √ (cid:0) ˜ ν L ν (cid:48) L (cid:1) (cid:32) Y l v (cid:96) y ν vy (cid:48)† ν v Y † ν v (cid:96) (cid:33) (cid:32) ν (cid:48) R ˜ ν R (cid:33) + 1 √ (cid:0) ˜ e L e (cid:48) L (cid:1) (cid:32) Y l v (cid:96) y e vy (cid:48)† e v Y † e v (cid:96) (cid:33) (cid:32) e (cid:48) R ˜ e R (cid:33) + h . c . , (6)where each field should be understood as a three-component vector in flavor space, with the Yukawa couplings × complexmatrices, and the vevs v and v (cid:96) belong to the SM Higgs and the new Higgses, respectively.The couplings y e and y ν are required to reproduce the SM flavor structure, and are thus tiny. If in addition y (cid:48) ν,e v (cid:28) Y (cid:96),ν,e v (cid:96) ,there is little mixing between doublets and singlets, and negligible contributions to precision electroweak observables. We findit convenient to choose parameters in this technically natural regime. III. BARYOGENESIS
In this section, we discuss the SU (2) (cid:96) phase transition, which provides the out of equilibrium condition necessary to evolve anonzero baryon number. A. Non-perturbative dynamics
Non-perturbative dynamics lead to simultaneous lepto- and DM-genesis through sphaleron processes during the SU (2) (cid:96) phasetransition. To understand the generation of lepton and DM numbers during the SU (2) (cid:96) transition we first need to analyze theglobal symmetries of the model. The theory posesses two anomaly-free global U (1) symmetries consistent with the gaugestructure and Yukawa interactions in Eq. (5) (see the first panel of Table II). In a realistic model, neutrino Yukawa couplings y ν and y (cid:48) ν must be small, therefore an additional approximate global symmetry U (1) (cid:48) exists. The U (1) (cid:48) global symmetry isanomalous under SU (2) (cid:96) and thus will be broken by instanton-generated interactions. For our purposes it is convenient toconstruct linear combinations of the U (1) and U (1) (cid:48) symmetries that will correspond to generalized lepton and DM numbers.Charge assignments under these symmetries, U (1) L and U (1) χ respectively, are shown in the middle panel of Table II.The U (1) symmetry is spontaneously broken by the Φ and Φ vevs. However, a global U (1) D subgroup of SU (2) (cid:96) × U (1) survives. This unbroken U (1) D is a diagonal combination of the U (1) and the U (1) group generated by the τ generator of SU (2) (cid:96) . The charges of the fields in the low energy theory are shown in the last panel of Table II. Note that the charges oflight fields under U (1) D are given by the sum of lepton and DM charges and thus U (1) D is not visible in low energy physics.However, U (1) D distinguishes between the SM leptons and new particles and thus will be responsible for the stability of theDM.Both U (1) L and U (1) χ are individually anomalous under SU (2) (cid:96) interactions. On the other hand, the sum of lepton andDM numbers will be conserved since it corresponds to an anomaly free U (1) symmetry. This means that SU (2) (cid:96) instantonsgenerate effective interactions in low energy theories that break U (1) L and U (1) χ individually while conserving the sum ofthe two charges, L + χ . For illustrative purposes, it is convenient to consider a one flavor toy model. The SU (2) (cid:96) instantonsgenerate an effective 4-fermion interaction that involves all of the doublets of SU (2) (cid:96) . Applying results from [22], the one-familysphalerons can be represented as a dimension six operator, O eff ∼ (cid:15) ij (cid:104) ( l iL · ¯ ν R )( l jL · ¯ e R ) − ( l iL · ¯ ν R )(˜ l jL · ¯˜ e R ) + ( l iL · ˜ l jL )(¯ ν R · ¯˜ e R ) − ( l iL · ˜ l jL )(¯˜ ν R · ¯ e R ) + (˜ l iL · ¯˜ ν R )(˜ l jL · ¯˜ e R ) − (˜ l iL · ¯˜ ν R )( l jL · ¯ e R ) (cid:105) , (7)where the dots denote Lorentz contractions, i, j are SU (2) W indices and the SU (2) (cid:96) indices have been expanded out.It is easy to see that instanton-induced interactions in Eq. (7) violate lepton and DM numbers by ∆ L = − and ∆ χ = 1 respectively. Consider, for example, the last term in Eq. (7). It is responsible for processes ν L ˜ e L → ˜ ν R e R and ˜ ν L e L → ˜ ν R e R .Since l L , ˜ l L , e R , ˜ e R have L = 1 while ν R , ˜ ν R have χ = 1 , the instanton-induced interactions violate lepton and DM numbers by ∆ L = − and ∆ χ = 1 .The generalization to the three generation model is straightforward, leading to a 12 fermion operator. At zero temperaturethe instanton operator is exponentially suppressed, but at high temperatures the SU (2) (cid:96) symmetry is restored, and lepton- andDM-number violating interactions are unsuppressed. The combined effect of all instanton-induced interactions (7) is calculatedby numerically solving the diffusion equations (see Sec. III B 2).One might be worried that the lepton and DM numbers will be immediately washed out since both U (1) L and U (1) χ areexplicitly broken by Yukawa interactions. However, the right-handed neutrinos and their partners reach chemical equilibriumlong after the SU (2) (cid:96) phase transition because of the smallness of their Yukawa couplings. In the case of the SM neutrinos, thisis implied by the small observed neutrino masses. In the case of the neutrino partners, this requires Γ( H ↔ (cid:96) (cid:48) ν (cid:48) ) be much lessthan the Hubble scale at the SU (2) (cid:96) phase transition, which will be satisfied provided y (cid:48) ν (cid:46) − for u ∼ TeV. As a result boththe lepton and DM number asymmetries survive until the electroweak transition. At that point the electroweak sphalerons seethe effective lepton number deficit and transfer it into baryons through the Dirac leptogenesis mechanism [23, 24].
B. Phase transition
1. Finite temperature effective potential
The one-loop effective scalar potential at nonzero temperature can be written schematically in terms of the background field u as V ( u, T ) = V tree ( u ) + V ( u,
0) + V temp ( u, T ) , (8)where the first of the individual contributions is the tree-level part, V tree ( u ) = − m u + 14 λ u , (9)where the mass parameter m and quartic λ schematically indicate combinations of those parameters from the scalar potential,Eq. (4). The remaining terms on the right-hand side correspond to the zero temperature Coleman-Weinberg correction and theone-loop finite temperature contribution.To calculate the Coleman-Weinberg term, we implement the cut-off regularization scheme and assume that the minimum ofthe one-loop potential and the SU (2) (cid:96) Higgs mass are the same as their tree-level values (see, e.g. [25]). The zero-temperatureone-loop correction takes the form, V ( u ) = 164 π (cid:88) i n i (cid:26) m i ( u ) (cid:20) log (cid:18) m i ( u ) m i ( v (cid:96) ) (cid:19) − (cid:21) + 2 m i ( u ) m i ( v (cid:96) ) (cid:27) , (10)where the sum is over all particles charged under SU (2) (cid:96) and n i denoted the number of degrees of freedom, with an extra minussign for the fermions. FIG. 1: Plot of the finite temperature effective potential V ( u, T c ) for v (cid:96) = 2 TeV , λ = 2 × − , g (cid:96) = 1 and T c = 200 GeV . Using the well-known formula for the one-loop finite temperature correction [25], the temperature-dependent piece is V temp ( u, T ) = T π (cid:88) i n i (3 ∓ (cid:90) ∞ dx x (cid:104) log (cid:16) ∓ e − √ x + m i ( u ) /T (cid:17) − log (cid:0) ∓ e − x (cid:1)(cid:105) . (11)In the expression above the sum is again over all fields, n i is the number of degrees of freedom including a factor of − forfermions; the minus signs are for bosons whereas the plus signs are for fermions.At this point, one can analyze the shape of the full effective potential for various temperatures. For successful baryogenesis thephase transition has to be strongly first order, v (cid:96) ( T c ) /T c (cid:38) , which favors small values of the effective quartic λ . The constrainton v (cid:96) coming from the LEP-II experiment is v (cid:96) (cid:38) . [17]. Finally, the critical temperature of the phase transition shouldnot be lower than ∼
175 GeV , in order to occur before the electroweak phase transition. Figure 1 shows the plot of V ( u, T c ) forsample parameter values fulfilling those constraints: g (cid:96) = 1 , v (cid:96) = 2 TeV and λ = 2 × − , leading to T c ∼
200 GeV .
2. Bubble nucleation and diffusion equations
A first order phase transition takes place at the critical temperature T c . Bubbles of true vacuum are nucleated and then expand,eventually filling the entire universe. Following [2] and denoting the bubble radius by R and the width by L w , we examine anansatz for the bubble profile, u ( r ) = 12 u c (cid:20) − tanh (cid:18) r − RL w (cid:19)(cid:21) . (12)In this case the width of the bubble scales as L w ∼ /T . The expanding bubble is assumed to be large, so that to a goodapproximation we can analyze its evolution in one dimension, along the z axis, taken to be perpendicular to the bubble wall. Weplace the bubble wall at z = 0 with the broken phase on the z > side. We adopt a bubble wall velocity of v w ≈ . c .In the presence of CP violation, the SU (2) (cid:96) instantons produce lepton and DM number asymmetries. In order to estimate theirmagnitude, a set of coupled diffusion equations is solved for particle number densities [2, 26]. Since only leptons are affectedby the presence of the new gauge group, there are 12 relevant equations in our case (see Appendix A) involving the followingparticle number densities, n ( l ) = n ( e L ) + n ( ν L ) , n ( e ) = n ( e R ) , n ( ν ) = n ( ν R ) , n (˜ l ) = n (˜ e L ) + n (˜ ν L ) , n (˜ e ) = n (˜ e R ) , n (˜ ν ) = n (˜ ν R ) ,n ( l (cid:48) ) = n ( e (cid:48) R ) + n ( ν (cid:48) R ) , n ( e (cid:48) ) = n ( e (cid:48) L ) , n ( ν (cid:48) ) = n ( ν (cid:48) L ) , n ( h ) = n ( h + ) + n ( h ) , n (Φ u ) = n (Φ u ) , n (Φ d ) = n (Φ d ) . (13)There are nine constraints on the particle number densities corresponding to the Yukawa equilibrium conditions, as well asfour constraints coming from the instanton equilibrium requirement (see Appendix B). However, not all of those constraints FIG. 2: SM lepton and DM particle number densities as a function of the spatial position z assuming the bubble wall located at z = 0 . are independent. Only seven of the Yukawa equilibrium conditions and one of the instanton equilibrium conditions are linearlyindependent.The diffusion equations contain diffusion constants for each particle species. For particles charged under the SM, theirmagnitude has been estimated in Ref. [27]. Carrying out a similar calculation and taking the SU (2) (cid:96) gauge coupling to be g (cid:96) ≈ we obtain the following estimates, D l = D ˜ l ∼ D e = D ˜ e ∼ D ν = D ˜ ν ∼ D Φ u = D Φ d ∼ /T , D h = D l (cid:48) ∼ /T , D e (cid:48) ∼ D ν (cid:48) ∼ /T . (14)Finally, we arrive at four coupled equations for the l, ˜ ν, φ u and h particle number densities, v w [4 n (cid:48) ( l ) + 4 n (cid:48) (˜ ν ) − n (cid:48) (Φ u ) − n (cid:48) ( h )] − T [22 n (cid:48)(cid:48) ( l ) + 4 n (cid:48)(cid:48) (˜ ν ) − n (cid:48)(cid:48) (Φ u ) − n (cid:48)(cid:48) ( h )] = 0 , (15) v w (cid:2) n (cid:48) ( l ) + n (cid:48) (Φ u ) (cid:3) − T [2 n (cid:48)(cid:48) ( l ) + n (cid:48)(cid:48) (Φ u )] = γ θ ( L w − | z | ) , (16) v w (cid:2) − n (cid:48) ( l ) + 6 n (cid:48) (˜ ν ) + n (cid:48) (Φ u ) − n (cid:48) ( h ) (cid:3) − T (cid:2) − n (cid:48)(cid:48) ( l ) + 6 n (cid:48)(cid:48) (˜ ν ) + n (cid:48)(cid:48) (Φ u ) − n (cid:48)(cid:48) ( h ) (cid:3) = γ θ ( L w − | z | ) , (17) v w (cid:2) n (cid:48) ( h ) (cid:3) − T [13 n (cid:48)(cid:48) ( h )] = 0 , (18)where the primes denote derivatives with respect to z , and γ , γ are the CP-violating sources for the SU (2) (cid:96) Higgs induced bythe bubble wall. The values for γ and γ in the two Higgs doublet model have been derived in [28] and are given by γ i ( z ) ≈ ˜ λ π Γ φ i T m m φ i ( T ) ∂ t z φ i , (19)with ˜ λ and m free parameters from the scalar potential. One can easily choose parameter values such that γ ≈ γ ≈ × − GeV , which yields the observed ratio of the baryon to entropy ratio of ∼ − , as discussed in the subsequentsection.
3. Lepton and dark matter asymmetries
We are interested in the particle number densities corresponding to the conserved global U (1) charges given in Table II, i.e., ∆ L ( z ) = n ( l ) + n (˜ l ) + n ( l (cid:48) ) + n ( e ) + n (˜ e ) + n ( e (cid:48) ) , ∆ χ ( z ) = n ( ν ) + n (˜ ν ) + n ( ν (cid:48) ) . (20)Using the equilibrium conditions we arrive at, ∆ L ( z ) = 3 (cid:104) n ( l ) + n (˜ ν ) − n (Φ u ) − n ( h ) (cid:105) , (21)and ∆ χ ( z ) = (cid:104) n ( l ) + n (˜ ν ) − n (Φ u ) + 12 n ( h ) (cid:105) . (22)Figure 2 shows the solution of the diffusion equations assuming T c = 200 GeV and γ = γ = 5 × − GeV . The ratio ofthe produced lepton and DM asymmetries is, (cid:12)(cid:12)(cid:12)(cid:12) ∆ L ∆ χ (cid:12)(cid:12)(cid:12)(cid:12) = 3 , (23)roughly independent of the numerical values of v w , T c , γ , and γ .
4. Baryon asymmetry
The particle number densities in Fig. 2 are normalized to the entropy s ≈ (2 π / g ∗ T , with g ∗ ∼ . For the above setof parameters, the ratio of the lepton number density and the entropy is roughly ∆ L/s ∼ × − . The lepton asymmetry is n L /s ≈ × − .The SU (2) (cid:96) instantons shut off after the SU (2) (cid:96) breaking concludes and the DM asymmetry freezes in. However, abovethe electroweak phase transition the SM sphalerons are active and they convert part of the SM lepton asymmetry to a baryonasymmetry. The baryon asymmetry generated by the sphalerons is [29], ∆ B = 2879 ∆ L . (24)This result only slightly depends on the lepton partner masses, with the effect minimized if those masses are below the elec-troweak scale. The final generated baryon asymmetry to entropy ratio is therefore, n B s ≈ − , (25)in the correct ballpark to match observations. IV. DARK MATTER
The DM is a mixed state, largely composed of the lightest ˜ ν R . Through interactions with the SM Higgs, it picks up a smallcomponent of electroweak doublet, χ L = ν (cid:48) L + (cid:15) ˜ ν L ,χ R = ˜ ν R + (cid:15) ν (cid:48) R , (26)with (cid:15) ∼ y ν v/ ( Y ν v (cid:96) ) (cid:28) .In standard ADM models, the baryon and DM asymmetries are of similar size, depending on the exact form of the operatorsmediating them. Assuming the DM is relativistic at the decoupling temperature, this implies a DM candidate with a mass at ∼ GeV scale . In particular, the relation between the DM mass and the relic abundances is given by, m χ = m p Ω DM Ω B (cid:12)(cid:12)(cid:12)(cid:12) ∆ B ∆ χ (cid:12)(cid:12)(cid:12)(cid:12) . (27) We note that it is also possible to realize an ADM scenario with a DM mass of several TeV via Xogenesis [30]. In that case the DM mass is approximatelyten times the decoupling temperature and must be roughly m Z (cid:48) / to have a sufficiently large annihilation cross section. FIG. 3: Dark matter annihilation channels to the pseudoscalar component of Φ . From Eqs. (23) and (24) we obtain | ∆ B/ ∆ χ | ≈ , which gives, m χ (cid:39) . (28)A mass of few GeV makes it difficult for the symmetric DM component to efficiently annihilate away, a generic challengein ADM scenarios. We circumvent this issue by arranging for a light Higgs boson with ∼ GeV mass into which the DM canannihilate efficiently. Provided the light scalar has a significant CP-odd component, the Yukawa interactions in Eq. (5) and theDM content (26) imply that, to leading order in (cid:15) , the coupling takes the form, L DM ≈ Y χ ¯ χ γ φ χ . (29)This provides a natural DM annihilation channel as shown in Fig. 3. Using this interaction we obtain, ( σv ) NR = Y χ m χ v π (2 m χ − m φ ) (cid:32) − m φ m χ (cid:33) / . (30)Writing the thermally averaged annihilation cross section as (cid:104) σ A v (cid:105) = σ ( T /m χ ) where σ = Y χ m χ π (2 m χ − m φ ) (cid:32) − m φ m χ (cid:33) / , (31)the present energy density of the symmetric component of the DM particles is [31, 32], Ω χ h (cid:39) (cid:18) . × − GeV (cid:19) σ √ g ∗ (cid:18) m χ T f (cid:19) , (32)where T f is the freeze-out temperature and g ∗ is the number of relativistic degrees of freedom. For m φ ≈ , the remnantsymmetric component will be subdominant to the asymmetric component produced by the SU (2) (cid:96) phase transition provided theYukawa coupling satisfies, Y ν (cid:38) . . (33)Light scalar bosons of mass ∼ GeV are somewhat unexpected from a potential whose overall energy scale is characterized by v (cid:96) ∼ TeV. However, they can be realized provided the quartics are all small (which is also favored by the need for a strongly firstorder phase transition) and/or tan β is large, indicating that v (cid:29) v . While a detailed analysis of the scalar sector is beyondthe scope of this work, it is a generic prediction that there will be light ( ∼ GeV) scalar particles with weak ( ∼ − ) couplingsto leptons. Such particles are typically not currently constrained by low energy experiments, but may be accessible in the future[33].The coupling of φ to the SM Higgs via λ | Φ | | H | terms in the scalar potential allows for thermal equilibrium between φ andthe SM provided Γ( φ → µ ¯ µ ) is greater than the expansion rate at temperatures T ∼ . This holds true for λ (cid:38) − .While the level of scalar mixing corresponding to the lower limit is too small to be observable at the LHC, larger values couldbe detectable.The primary mediator for dark matter scattering with heavy nuclei is the Z (cid:48) gauge boson, which does not interact with quarksand therefore does not appear in tree-level DM direct detection diagrams (though contributions do appear at one loop). As aresult, the corresponding bound on v (cid:96) set by the null search results from the CDMSlite experiment [34] is much less stringentthan the collider constraint from LEP-II of v (cid:96) (cid:38) . . The DM direct detection diagrams involving the SM gauge bosonsare shown in Fig. 4. The calculation of the spin-independent direct detection cross section closely follows the results of [19]for the electroweak diagrams, which, combined with the CDMSlite bounds, provides an upper limit on the doublet admixtureparameter (cid:15) (cid:46) . , consistent with the assumption that y (cid:48) ν,e v (cid:28) Y (cid:96),ν,e v (cid:96) . FIG. 4: Diagrams contributing to DM interacting with quarks.
V. CONCLUSIONS
We have investigated a novel extension of the Standard Model, in which SU (3) × SU (2) × U (1) is supplemented by anon-Abelian gauge group SU (2) (cid:96) , in which the leptons are promoted to doublets by introducing lepton partner fermions. Thedoublet partners pair up with extra SU (2) (cid:96) singlet fields needed to cancel the anomalies and develop vector-like masses after SU (2) (cid:96) breaking. We find that the global symmetries produce a viable dark matter candidate, which is the lightest of the leptonpartners and stable due to a residual U (1) χ global symmetry. In addition, it can explain baryogenesis through the breakdown ofthe new gauge group and naturally fits into an asymmetric dark matter framework.We explicitly analyze the details of the dynamics of SU (2) (cid:96) phase transition, during which nonperturbative interactionsmediated by SU (2) (cid:96) instantons violate both lepton and dark matter numbers, leading to a correlated asymmetry in both sectors.We find that the wide (and relatively unconnected to electroweak observables) parameter space easily allows for a first orderphase transition with sufficient asymmetry generated in the leptons to lead (through the electroweak sphalerons) to a baryonasymmetry in line with observation. Provided the mass of the dark matter is a few GeV, its correlated abundance will also matchcosmological observations.Both the lepton partners and the additional neutral gauge bosons are colorless, and thus are not efficiently produced at theLHC. The most stringent constraints are bounds on the Z (cid:48) , which contributes to e + e − at LEP-II. It would be interesting to seehow a future high energy e + e − collider could shed light on such a scenario, and whether it could preclude enough of parameterspace to say something definitive about its potential realization as the mechanism for baryogenesis.It is also of interest to explore similar extensions of the Standard Model based on generalizations of baryon number. This iscurrently work in progress [35]. Acknowledgments
This research was supported in part by the NSF grant PHY-1316792. BF also acknowledges partial support from the DOEgrant de-sc0009919 and is grateful to the Institute of Advanced Studies at the Nanyang Technological University in Singaporefor hospitality during the completion of this work. TMPT is grateful for helpful conversations with M. Buckley, Y. Kahn, G.Krnjaic, M. Perelstein, M. Peskin, and C.E.M. Wagner.
Appendix A: Diffusion equations
The rates at which the particle densities change are described by 12 diffusion equations. We denote the common rates forthe SU (2) (cid:96) instanton induced interactions by Γ . Neglecting the neutrino and neutrino partner Yukawas, the equations for theparticle number densities n ( l ) , n (˜ l ) , n ( l (cid:48) ) are given in simplified notation (all particle number densities labeled by their species,e.g. ˜ (cid:96) ≡ n (˜ (cid:96) ) ) by, ˙ l − D l ∇ l = − Γ (cid:104) l + 2˜ l − ν + e ) − ν + ˜ e ) (cid:105) − Γ Y l (cid:18) l − Φ u − l (cid:48) (cid:19) − Γ y e (cid:18) l − h − e (cid:19) , (A1) ˙˜ l − D ˜ l ∇ ˜ l = − Γ (cid:104) l + 2 l − ν + ˜ e ) − ν + e ) (cid:105) − Γ Y l (cid:32) ˜ l − Φ d − l (cid:48) (cid:33) − Γ y e (cid:32) ˜ l − h − ˜ e (cid:33) , (A2) ˙ l (cid:48) − D l (cid:48) ∇ l (cid:48) = − Γ y (cid:48) e (cid:18) l (cid:48) − h − e (cid:48) (cid:19) + Γ Y l (cid:18) l − Φ u − l (cid:48) (cid:19) + Γ Y l (cid:32) ˜ l − Φ d − l (cid:48) (cid:33) . (A3)0For ν, ˜ ν, ν (cid:48) , ˙ ν − D ν ∇ ν = − Γ Y ν (cid:18) ν − Φ u − ν (cid:48) (cid:19) − Γ (cid:16) ν + ˜ e + 2 e − l − l (cid:17) , (A4) ˙˜ ν − D ˜ ν ∇ ˜ ν = − Γ Y ν (cid:18) ˜ ν − Φ d − ν (cid:48) (cid:19) − Γ (cid:16) ν + e + 2˜ e − ˜ l − l (cid:17) , (A5) ˙ ν (cid:48) − D ν (cid:48) ∇ ν (cid:48) = Γ Y ν (cid:18) ν − Φ u − ν (cid:48) (cid:19) + Γ Y ν (cid:18) ˜ ν − Φ d − ν (cid:48) (cid:19) . (A6)For e, ˜ e, e (cid:48) , ˙ e − D e ∇ e = − Γ Y e (cid:18) e − Φ u − e (cid:48) (cid:19) + Γ y e (cid:18) l − h − e (cid:19) − Γ (cid:16) e + ˜ ν + 2 ν − l − l (cid:17) , (A7) ˙˜ e − D ˜ e ∇ ˜ e = − Γ Y e (cid:18) ˜ e − Φ d − e (cid:48) (cid:19) + Γ y e (cid:32) ˜ l − h − ˜ e (cid:33) − Γ (cid:16) e + ν + 2˜ ν − ˜ l − l (cid:17) , (A8) ˙ e (cid:48) − D e (cid:48) ∇ e (cid:48) = Γ y (cid:48) e (cid:18) l (cid:48) − h − e (cid:48) (cid:19) + Γ Y e (cid:18) e − Φ u − e (cid:48) (cid:19) + Γ Y e (cid:18) ˜ e − Φ d − e (cid:48) (cid:19) . (A9)Finally, for the particle number densities Φ u , Φ d and h , ˙Φ u − D Φ u ∇ Φ u = γ + Γ Y l (cid:18) l − Φ u − l (cid:48) (cid:19) + Γ Y ν (cid:18) ν − Φ u − ν (cid:48) (cid:19) + Γ Y e (cid:18) e − Φ u − e (cid:48) (cid:19) , (A10) ˙Φ d − D Φ d ∇ Φ d = γ + Γ Y l (cid:32) ˜ l − Φ d − l (cid:48) (cid:33) + Γ Y ν (cid:18) ˜ ν − Φ d − ν (cid:48) (cid:19) + Γ Y e (cid:18) ˜ e − Φ d − e (cid:48) (cid:19) , (A11) ˙ h − D h ∇ h = Γ y e (cid:18) l − h − e (cid:19) + Γ y e (cid:32) ˜ l − h − ˜ e (cid:33) − Γ y (cid:48) e (cid:18) l (cid:48) − h − e (cid:48) (cid:19) . (A12)where γ and γ are the CP-violating sources. Appendix B: Constraints on particle number densities
The equilibrium conditions emerging from the Yukawa terms are, l − Φ u − l (cid:48) , ˜ l − Φ d − l (cid:48) , ν − Φ u − ν (cid:48) = 0 , ˜ ν − Φ d − ν (cid:48) = 0 ,e − Φ u − e (cid:48) = 0 , ˜ e − Φ d − e (cid:48) = 0 , l − h − e = 0 , ˜ l − h − ˜ e = 0 , l (cid:48) − h − e (cid:48) = 0 . (B1)And those from the instanton-induced interactions can be written as, l − ν − e = 0 , l l − ν − ˜ e = 0 , ˜ l − ˜ ν − ˜ e = 0 , l l − ˜ ν − e = 0 . (B2)As discussed in the main text, two of the Yukawa equilibrium conditions and three of the instanton conditions are linearlydependent on the others. Therefore, there are only four independent particle number densities. For example one can choosethem to be l , ˜ ν , h and Φ u , in which case the other particle densities are given by, ˜ l = 2˜ ν − h , l (cid:48) = l − Φ u , ν = l h , ν (cid:48) = l − Φ u h ,e = l − h , ˜ e = ˜ ν − h , e (cid:48) = l − Φ u − h , Φ d = 2˜ ν − l + Φ u − h . (B3) [1] A. D. Sakharov, Pisma Zh. Eksp. Teor. Fiz. , 32 (1967), [Usp. Fiz. Nauk161,61(1991)]. [2] J. Shu, T. M. P. Tait, and C. E. M. Wagner, Phys. Rev. D75 , 063510 (2007), hep-ph/0610375.[3] S. Nussinov, Phys. Lett.
B165 , 55 (1985).[4] D. B. Kaplan, Phys. Rev. Lett. , 741 (1992).[5] D. Hooper, J. March-Russell, and S. M. West, Phys. Lett. B605 , 228 (2005), hep-ph/0410114.[6] D. E. Kaplan, M. A. Luty, and K. M. Zurek, Phys. Rev.
D79 , 115016 (2009), 0901.4117.[7] K. Petraki and R. R. Volkas, Int. J. Mod. Phys.
A28 , 1330028 (2013), 1305.4939.[8] K. M. Zurek, Phys. Rept. , 91 (2014), 1308.0338.[9] A. Pais, Phys. Rev. D8 , 1844 (1973).[10] Y. Tosa, R. E. Marshak, and S. Okubo, Phys. Rev. D27 , 444 (1983).[11] S. Rajpoot, Int. J. Theor. Phys. , 689 (1988).[12] R. Foot, G. C. Joshi, and H. Lew, Phys. Rev. D40 , 2487 (1989).[13] C. D. Carone and H. Murayama, Phys. Rev.
D52 , 484 (1995), hep-ph/9501220.[14] H. Georgi and S. L. Glashow, Phys. Lett.
B387 , 341 (1996), hep-ph/9607202.[15] P. Fileviez Perez and M. B. Wise, Phys. Rev.
D82 , 011901 (2010), [Erratum: Phys. Rev.D82,079901(2010)], 1002.1754.[16] M. Duerr, P. Fileviez Perez, and M. B. Wise, Phys. Rev. Lett. , 231801 (2013), 1304.0576.[17] P. Schwaller, T. M. P. Tait, and R. Vega-Morales, Phys. Rev.
D88 , 035001 (2013), 1305.1108.[18] B. Fornal, A. Rajaraman, and T. M. P. Tait, Phys. Rev.
D92 , 055022 (2015), 1506.06131.[19] B. Fornal and T. M. P. Tait, Phys. Rev.
D93 , 075010 (2016), 1511.07380.[20] M. Blennow, B. Dasgupta, E. Fernandez-Martinez, and N. Rius, JHEP , 014 (2011), 1009.3159.[21] J. F. Gunion and H. E. Haber, Phys. Rev. D72 , 095002 (2005), hep-ph/0506227.[22] D. E. Morrissey, T. M. P. Tait, and C. E. M. Wagner, Phys. Rev.
D72 , 095003 (2005), hep-ph/0508123.[23] K. Dick, M. Lindner, M. Ratz, and D. Wright, Phys. Rev. Lett. , 4039 (2000), hep-ph/9907562.[24] H. Murayama and A. Pierce, Phys. Rev. Lett. , 271601 (2002), hep-ph/0206177.[25] M. Quiros, in High energy physics and cosmology. Proceedings, Summer School, Trieste, Italy, June 29-July 17, 1998 (1999), pp. 187–259, hep-ph/9901312, URL http://alice.cern.ch/format/showfull?sysnb=0302087 .[26] A. G. Cohen, D. B. Kaplan, and A. E. Nelson, Phys. Lett.
B336 , 41 (1994), hep-ph/9406345.[27] M. Joyce, T. Prokopec, and N. Turok, Phys. Rev.
D53 , 2930 (1996), hep-ph/9410281.[28] A. Riotto, Phys. Rev.
D53 , 5834 (1996), hep-ph/9510271.[29] J. A. Harvey and M. S. Turner, Phys. Rev.
D42 , 3344 (1990).[30] M. R. Buckley and L. Randall, JHEP , 009 (2011), 1009.0270.[31] E. W. Kolb and M. S. Turner, Front. Phys. , 1 (1990).[32] K. Griest and D. Seckel, Phys. Rev. D43 , 3191 (1991).[33] G. Krnjaic, Phys. Rev.
D94 , 073009 (2016), 1512.04119.[34] R. Agnese et al. (SuperCDMS), Phys. Rev. Lett.116