B Decay Anomalies and Dark Matter from Vectorlike Confinement Arxiv

B decay anomalies and dark matter from vectorlike conﬁnement

James M. Cline Niels Bohr International Academy & Discovery Center, Niels Bohr Institute, University of Copenhagen, Blegdamsvej 17, DK-2100, Copenhagen, Denmark and McGill University, Department of Physics, 3600 University St., Montr´eal,Qc H3A2T8 Canada

Lepton ﬂavor universality violating B → K`` and K∗`` decays tentatively observed by LHCb can be explained by leptoquark exchange. We explore a simple model for the B anomalies with a

composite leptoquark from new strong dynamics at the TeV scale, a conﬁning SU(NHC ) hypercolor

interaction. The new matter fields, fundamentals under SU(NHC ), are heavy vectorlike fermions Ψ, S, and an inert scalar doublet φ. Ψ is colored under QCD while S is neutral, and the hyperbaryon SN is a dark matter candidate. The model is tightly constrained by meson-antimeson oscillations, lepton flavor violation, and LHC searches for resonant production of the exotic bound states. The dark matter may be detectable through its magnetic dipole moment. If mS is sufficiently small, composite leptoquarks and heavy lepton partners can be pair-produced at an observable level at LHC.

1. INTRODUCTION physics loop effects [14–20]. It could also be induced by exchange of a heavy Z0 vector boson, inspiring the con- While the Standard Model (SM) continues to ac- struction of many models of this kind [21–54]. By Fierz curately describe particle interactions at the shortest rearrangement, (2) takes the form suggestive of vector probed distance scales, there is an encouraging hint of leptoquark exchange. Global analyses identify two pos- new physics from decays of neutral B mesons to K or sible vector leptoquarks U1 and U3 and one scalar S1 as K∗ and charged leptons. The LHCb collaboration mea- viable candidates, in the notation of ref. [55] where the sures the ratios of branching ratios [1, 2] subscript denotes the SU(2)L representation. Models involving leptoquark exchange have been studied in refs. B(B → K(∗)µ+µ−) [15, 56–87]. R(∗) = (1) K B(B → K(∗)e+e−) In the present work we offer a simple, UV complete model in which the vector U1 leptoquark arises from con- to be (20−30)% below the SM prediction for both K and fining dynamics of an SU(NHC ) gauge theory, referred to K∗ final states. Uncertainties from the hadronic matrix as hypercolor. It is an example of vectorlike confine- elements cancel in the ratios, greatly reducing the the- ment [88] in which the constituent fermions have bare oretical errors and making these ratios particularly in- masses that can be freely chosen. Our proposal dif- teresting probes of new physics that could violate lepton fers from previous models for RK(∗) with composite lep- flavor universality. Although the discrepancies between toquarks [57, 71, 78, 81] where the constituent masses theory and experiment in the individual measurements were assumed to be much lighter than the confinement are below 3 σ, combining them increases the significance scale ΛHC . In that case, approximate chiral symmetry to 4 σ [3–13]. This conclusion remains true when other is important, and the lightest bound states are pseudo- observables are included in the fit, like the branching Nambu-Goldstone bosons (pNGBs). ratio for B → φµ+µ−, which also has a deficit, and s Here we instead assume that the constituent masses M arXiv:1710.02140v4 [hep-ph] 19 Dec 2017 angular decay distributions that are subject to hadronic are larger than Λ (but not too much larger) so that chi- uncertainties. HC ral dynamics plays no role, yet the system is not quirky Model-independent fits to the data show that the ad- [89] as occurs if Λ M. One advantage of this choice dition of the single effective operator HC is that the strong dynamics is somewhat more calculable, c via a nonrelativistic potential model that works reason- O = (¯s γ b )(¯µ γαµ ) (2) bLµL Λ2 L α L L L ably well for QCD -onium states like J/ψ and Υ. More- over collider constraints on production of the new bound to the effective Hamiltonian is sufficient to give a good states are under better control in this regime, where res- fit to the observations, with [4] onant production is suppressed by the wave function of the constituents, and pair production is limited by the c 1.0 × 10−3 large bound state masses. We also thereby avoid compli- 2 = 2 (3) Λ TeV cations associated with composite Higgs scenarios. The form (2) arises in the standard model from W ex- Another distinctive feature of our model is that it pro- change at one loop, and can be realized in models of new vides a dark matter candidate, which is intimately linked 2

L L L LQQ SU(3) SU(2) U(1) U(1) SU(N) Z S S ψ L y em HC 2 ψ S ψ Ψ 3 1 2/3 2/3 N −1 Q Q L L QQ (a) (b) (c) S 1 1 0 0 N −1 φ 1 2 −1/2 (0, −1) N¯ −1 Figure 1: Tree-level contributions to ﬂavor-violating processes from exchange of composite vector bosons. Table I: Quantum numbers of new physics particles

and so one can anticipate that the matching of the ob- to the B decay phenomenology. In particular, one of the served b → sµ+µ− FCNC will give rise to other FCNC hyperquarks (denoted by S) is a singlet under the SM observables. The observed B decay anomaly is medi- gauge group. It is a constituent of the composite lep- ated by tree-level exchange of a composite leptoquark toquark, and the baryon-like SNHC bound state is the Φ = ΨS¯, while the other kinds of bound states ΨΨ¯ and dark matter. Exceptionally, S can be lighter than Λ HC SS¯ give analogous contributions to neutral meson mixing without entailing a pNGB, since the approximate U(1) A and flavor-violating decays of charged leptons, depicted flavor symmetry associated with S is anomalous. in fig. 1. A further difference relative to previous models of com- To fully specify the model, we must define the flavor posite leptoquarks is that we invoke a scalar hyperquark basis of the SM fields in (4). For simplicity we assume φ, a doublet under SU(2)L. It is motivated by the need that the mass matrices of the charged leptons and down- to couple the new physics to left-handed quarks and lep- type quarks are already diagonal in this basis, so that tons, eq. (2). Previous models achieved this by taking all the mixing giving rise to the CKM matrix is due to the fermionic hyperquarks to be doublets, but it is more diagonalizing the up-type quark masses. Then going to straightforward to have a dark matter candidate if this the mass eigenbasis of the quarks, is avoided, and the massless limit of S can be safely ˜ ¯ ˜ ¯ taken without introducing new relatively light scalars λiQi → λj u¯L,iVij, dL,j that would be easily produced in colliders. In our model ≡ λ˜0 u¯ , λ˜ d¯ (5) both fermionic hyperquarks S, Ψ, are isosinglets, while Ψ i i i i ¯ carries color. The SΨ bound state is the leptoquark. Thus the couplings to up-type quarks are given by λ˜0 = After defining the model in section 2 and reducing it i V λ˜ . to an effective theory suitable for addressing the B decay ij j It is worth noting that the individual lepton flavor anomalies in section 3, we identify (without fully delin- symmetry for generation i becomes exact in the limit eating) some regions of parameter space that are consis- λ → 0: these couplings are only multiplicatively renor- tent with constraints on flavor changing neutral currents i malized, up to neutrino mass insertions. Therefore if we (FCNCs) in section 4, dark matter direct detection (sec- only wish to explain lepton flavor universality violation tion 5), and collider searches (section 6). The examples in the muon sector, it is technically natural to set the we focus on have potential for associated new signals in couplings λ , λ → 0. all of these categories of observables. A summary and 1 3 If S is lighter than φ and Ψ, the hyperbaryon bound conclusions are given. Appendices provide details of the state Σ = SN is a dark matter candidate. Its stability potential model for estimating bound state properties, is protected by the accidental Z symmetry under which the formalism for computing dipole moments from par- 2 all the new matter fields are odd. It is a consequence tial compositeness of the light fermions, and the magnetic of the SU(N ) gauge symmetry, analogous to baryon dipole moment of the dark matter constituent fermion. HC number conservation in the SM. Moreover the fields Ψ and S can consistently be assigned normal baryon and lepton number, respectively. 2. MODEL

Our model introduces three matter fields that are fun- 3. LOW-ENERGY EFFECTIVE THEORY damentals of hypercolor (HC): a Dirac fermionic DM particle S, a vectorlike fermion Ψ that carries SM color, and The Ψ¯ S bound states of our model have the quan- a scalar φ that is an SU(2)L doublet. The quantum num- tum numbers of leptoquarks and contribute to the de- bers are shown in table I. Gauge symmetry allows these cays B → K(∗)µ+µ−. The leptoquarks can be either fields to couple to left-handed SM quarks and leptons pseudoscalar Π or vector ρµ, with decay constants only through the interactions ¯ µ µ h0|(Sγ γ5Ψ)|Πi = fΠ pΠ ˜ ¯ a A ¯ ∗A a µ L = λi Qi,a φAΨ + λi SAφa Li (4) ¯ µ h0|(Sγ Ψ)|ρλi = fρ mρλ (6) where a(A) is the SU(2)L (SU(N)HC ) index and i is the where λ labels the helicity state of the vector. These cur- flavor index. These interactions explicitly break the ap- rents couple to the SM fields via Qγ¯ µL and so the inter- proximate flavor symmetries of the standard model (SM), actions of the pseudoscalar are suppressed by small quark 3 and lepton masses through the equations of motion, due 4. FLAVOR CONSTRAINTS µ µ µ to the momentum factor pΠ = pq + pl . This can also be understood in terms of the helicity suppression of the The fit of ref. [4] implies that the RK(∗) anomalies can amplitude for pseudoscalar decay to approximately chiral be explained by taking states in analogy to charged pion decay. Hence we are interested in the vector leptoquark for explaining the B g g∗ 1 × 10−3 22 32 = (13) decay anomalies. 2 2 mρ TeV We seek an effective description of the leptoquark interaction with the SM fields, To translate this into a constraint on parameters of the model, we must determine ψ(0) and the bound state ij ¯ µ gρ ρµ (Qiγ Lj) (7) mass. For this purpose we use a potential model that is described in appendix A. To simplify this prelimi- ij ∼ 1 The coupling gρ can be determined by matching the de- nary analysis, we assume that mΨ = mΦ. We con- cay rate for ρµ → QiL¯j in the effective theory and the sider two possibilities for the dark constituent mass, underlying model. In the effective theory, the rate is ΛHC . mS . mφ ≡ M and mS ΛHC . mφ, such that S is respectively nonrelativistic or relativistic within |gij|2 the bound state. Two version of the potential model are Γ(ρ → Q L¯ ) = ρ m (8) µ i j 24π ρ given to treat these cases. We avoid the quirky regime where ΛHC M [89] since collider constraints are ex- neglecting the quark and lepton masses. In the UV the- pected to become more stringent (and difficult to quan- 2 ory, this rate can be computed as Γ = σvrel|ψ(0)| , where tify), as we will discuss in section 6. ψ is the wave function for the ΨS¯ bound state and σ is In either case, the coefficients of the dimension-6 oper- the cross section for ΨS¯ → QiL¯j annihilation, ators (12) are proportional to

˜2 2 2 |ψ(0)|2 |λi λj |(mS + mΨ) ζ ≡ (14) σvrel = NHC (9) 3 2 2 mR 96π(mS mΨ + mφ) that encodes the effects of the confining dynamics, where ignoring the initial velocities of the bound particles, and mR is the mass of the exchanged resonance. ζ is dimen- assuming only the spin-1 configuration of S and Ψ con- sionless and thus only depends upon ratios of mass or tributes in the sum over spins. This gives energy scales (and NHC ). In the case where all masses are approximately equal to scale M, there is just one 1/2 ˜ ∗ NHC λiλj (mS + mΨ) ij such ratio, r = M/ΛHC . Using the nonrelativisitic po- gρ = 2 ψ(0) (10) 4 mρ (mS mΨ + mφ) tential model in the region M/ΛHC > 1, we numerically

determine ζ(r) for NHC = 2, 3, 4, shown in fig. 2 (solid Once the effective interaction (7) is specified, we can curves). It reaches a maximum value of ∼ 0.0037 for ∼ integrate out the leptoquark to generate the dimension-6 r = 3 if NHC = 3, 4, and is smaller if NHC = 2. For operator shown in fig. 1(a), the case mS ΛHC , the low scale mS becomes irrele- vant and we can can still express ζ as a function of the ∗ g g 0 0 ij i j a µ b µ same r = M/ΛHC , using the relativistic version of the ¯ 0 ¯ δL = − 2 (Qi γ Qi ,b)(Lj0 γ Lj,a) (11) mρ potential model. ζ is smaller for these models, ∼ 0.002 for r ∼ (2 − 3). after Fierz transforming [90], where a, b are SU(2) in- Since ζ is small, the best case for avoiding large cou- L ˜ dices. This contribution must interfere destructively with plings λ2, λ2,3 in the underlying model is to have mφ ∼ + − ∗ m ∼ M and Λ ∼ (0.3 − 0.4)M for N = 3 or 4. the SM contribution to b → sµ µ , requiring g32g22 to Ψ HC HC be approximately real and positive. Then with ζ = 0.0037, eq. (13) implies In addition to the composite leptoquarks, there are ¯ ¯ M 2 3 composite ΨΨ ≡ ρΨ and SS ≡ ρS vector bosons, whose 2 ˜ ˜ |λ2 λ2 λ3| = 0.3 (15) exchanges are shown in fig. 1(b,c). The ensuing operators TeV NHC are From the potential model, the bound state mass mρ at ∗ hijhi0j0 a µ b µ Λ/M = 0.4 and NHC = 3 is predicted to be 1.75 times the δL = − (L¯ γ L 0 )(L¯ 0 γ L ) 2 m2 i i ,b j j,a naive value of 2M that ignores the contribution from the ρS ∗ hypercolor flux tube. For M = 1 TeV therefore, mρ = kijki0j0 a µ b µ − (Q¯ γ Q 0 )(Q¯ 0 γ Q ) (12) 3.5 TeV. 2 m2 i i ,b j j,a ρΨ where hij and kij are determined analogously to (10), ˜ ∗ ∗ ˜ ˜∗ 1 with λiλj → λiλj and λiλj , respectively. Also this choice maximizes the wave function at the origin 4

0.004 In the case where mS ΛHC , the couplings must be increased by ∼15%, to compensate for the smaller ζ factor. From the partially relativistic version of the po- 0.003 tential model, the mass of the resonance is found to be N=4 3 R N=2 N=3

approximately m /

2 m ≈ m ∼ 0.002 S Ψ mρ = (0.56 + 1.8 NHC )ΛHC + mΨ (16)

ψ(0)| m | ≈ 0 valid for m /Λ 3; for smaller values m tends to- S Ψ HC & ρ 0.001 ward a constant independent of mΨ.

0 4.1. Meson-antimeson mixing 2 4 6 8 10 r = M / ΛHC

Diagrams of the type ﬁg. 1(c) contribute to the oscil- 2 3 lations of neutral mesons: K0-K¯ 0, D0-D¯ 0, B0-B¯0 and Figure 2: The function ζ = |ψ(0)| /mR (eq. (14)) that en- 0 ¯0 codes the strong dynamics dependence of the exchange of Bs -Bs , placing constraints on diﬀerent combinations of bound state resonances between SM fermion bilinears. Solid ˜ the λi couplings by similar reasoning as led to eq. (15). curves are for all constituents having same mass M, while

For the ﬁducial model parameters with ζ = 0.0037 and dashed are for the case of mS ΛHC .

NHC = 3, we find the upper limits given in table II. upper limit These are inferred by comparing the operator coefficients meson quantity fiducial value 2 (units M/TeV) (units M/TeV) kij/2mρ2 with the upper limits from ref. [16] for Bs mix- Ψ K0 |λ˜ λ˜ | 1.3 × 10−3 1 × 10−3 ing and refs. [91? ] for K, D and Bd mixing (which gives 1 2 0 ˜0 ˜0 −3 −4 the most stringent bounds in each case). The bound on K D |λ1λ2| 2 × 10 7 × 10 0 and D mixing are conservative in the sense that we have B |λ˜1λ˜3| 0.026 0.0066 0 ˜ ˜ assumed that the contribution to the imaginary part of Bs |λ2λ3| 0.066 0.066 the amplitude, which is much more strongly constrained, is as large as that of the real part. Table II: 3rd column: bounds from meson-antimeson mixing As an example of parameters that can satisfy these on quark couplings. M is the (assumed) universal mass of constraints, we take M = 1 TeV and the constituents. Last column: values following from fiducial parameter choices, eq. (17). ˜ ˜ ˜ λ1 = −0.01, λ2 = 0.1, λ3 = 0.66, λ2 = 2.1 ˜0 ˜0 ˜0 (λ1 = 0.014, λ2 = 0.13, λ3 = 0.66) (17) be made lower than 800 GeV because of LHC constraints, as we will see in section 6. Our main goal in this paper The primed couplings (pertaining to the up-type quarks) is to establish some region of viability for the model. A are determined by the unprimed ones through eq. 5. The thorough exploration of the parameter space will become resulting contributions to the products of couplings rele- more strongly motivated if the B decay anomalies are vant to meson mixing are shown in table II (last column). confirmed with greater significance. ˜ ˜ The relative minus sign between λ1 and λ2 leads to an ˜ 0 accidental cancellation in Vijλj such that the D mixing contribution is just below the limit; without this sign, we 4.2. Lepton flavor violating decays ˜0 ˜0 would get λ1λ2 = 0.004, saturating the limit. The fact that all the mixing constraints are close to As mentioned above, nothing obliges us to turn on the being saturated, with the exception of Bs, can be under- couplings λ1, λ3 to e and τ leptons in this model, apart stood from the lack of any approximate flavor symmetry from aesthetic considerations. If these couplings are non- such as minimal flavor violation [94] in this model. By + − vanishing, then the diagram of fig. 1(b) contributes to requiring a large enough b → sµ µ transition and try- lepton-flavor violating (LFV) decays τ → 3` (where ` is ing to avoid couplings much larger than 1, we are pushed µ or e) and µ → 3e. We can derive upper bounds on into this restricted region of parameter space. |λ |, |λ | by comparing the amplitude from exchange of One way to relieve the tension is to take smaller values 1 3 the ρS = SS¯ vector meson, of M, say 800 GeV. Then√ the couplings could on average ∗ 2 be smaller by a factor of 0.8 than in eq. (17). If we keep λiλ |λj| 3 ζ j (`¯ γµ` )(`¯ γµ` ) (18) the leptonic coupling large, λ2 = 1.1, the quark couplings 4 2M 2 j i j j can all be reduced by 0.8, and the last column in table √ µ ¯ µ II is reduced by 0.64, leading to a 60% reduction in the to the SM amplitude 2 2 GF (¯νiγ ì)(`jγ νj) for the cor- actual amplitude. Allowing for larger leptonic coupling responding allowed decay, giving a ratio of branching ra- −9 2 6 can further ameliorate the situation, which is thus not as tios of Rij = 1.8 × 10 |λi| |λj| . From τ → 3µ we re- −7 −12 marginal as the limiting case makes it appear. M cannot quire Rτµ < 1.2 × 10 and from µ → 3e, Rµe < 10 , 5

k, µ −Q/2 + q In addition, as shown in appendix B, the diagrams of −Q/2+q+k ﬁg. 3 give rise to dimension-6 transition moment opera- Q/2+q−k tors of the form −Q/2 + q F fi Q/2 + q λ i Q−k Q Q/2+q µν f −eF f F i f ¯ ¯ Q λ i Q−k L = qq µ Qf σµν Fq + q` µ F`σµν Lf (21) µ 2 q ` (a) (b) k, 2MF −Q/2 + q −Q/2 + q where qf is the electric charge of particle f, a quark or lepton. Ref. [97] noted that these lead to transition mag- Q/2 + q Q/2+q−k netic moments among the light fermion states upon di- f F fi F i λ Q λ i Q−k Q i Q−k agonalization of the mass matrices, which take the form ! ! (c) (d) m 0 f f¯ F¯ f L (22) R R µ M F Figure 3: Diagrams for transition magnetic moment from f F L heavy composite to light fundamental fermion. Equal constituent masses are assumed for routing of momenta. Cross where f, F denote the SM and heavy fermions, respec- represents the bound-state wave function. tively, with mf being the SM mass matrix and MF the heavy composite mass. To leading order for small mixing, (22) is diagonalized by separate left- and right-handed 1 θL,R giving transformations OL,R = T with mixing angles −θL,R 1 that are vectors in ﬂavor space, |λ1| < 0.23, |λ3| < 0.9 (19) mf µf µf This implies that the anomalous contribution to the b → θR = 2 , θL = (23) + − 2 MF MF se e amplitude (which goes as |λ1| ) is less than 0.01 + − times that of b → sµ µ . After transforming to the mass eigenbasis, a transition moment for the light fermions is generated,

q µi µj mj f f f f ¯ µν 4.3. Radiative transitions L = e 4 (fL,i σµν fR,j)F (24) 2MF Another potentially important class of FCNCs is the 2 4 It is proportional to |ψ(0)| /MF = ζ/MF rather than just radiative decays such as b → sγ, τ → µγ. They arise ζ (recall eq. (14)). The same potential model can be used through the mass mixing of SM fermions with heavy to compute MF as we used for the bosonic bound states, composite fermions from the confining sector. These since quantum mechanics is insensitive to the spins of are vectorlike doublet leptons and quarks, φS ≡ F` and the constituent particles, and we are ignoring the spin- φΨ ≡ Fq, that naturally have magnetic moments com- spin interactions in our simple treatment. Then for the ing from their charged constituents. They contribute to 2 4 fiducial parameters with ΛHC = 0.4 M, |ψ(0)| /MF = dipole operators of the light fermions in the mass eigen- ζ/(3.5 M). basis, including transition magnetic moments. The mass mixing in this kind of model was previously considered in ref. [95]. Here we make the connection 4.3.1. FCNC quark transitions to transition moments more precise, using methods that were previously applied to the rare decays of Z → Υγ From the general formula (24), the operator contribut- in the SM in ref. [96]. The contributions look like loop ing to b → sγ is diagrams (fig. 3) but the integral is over the relative momentum of the bound constituents and is weighted by ˜ ˜ 2 eλ2λ3|ψ(0)| mb µ s¯L/qγ bR (25) the wave function in momentum space. Details are given 6M 4 M in appendix B. There we show that the mass-mixing be- Fd tween the SM and heavy fermions takes the form This implies that the conventionally normalized Wilson coefficient of the dipole operator O7 is ˜ ¯ a ¯ ∗ a ψ(0) ˜ ¯ ¯ λf Qf,aφ Ψ + λiSφaLi → √ λiQiFq + λiF`Li √ M 2 ˜ ˜ 2π λ2λ3 ζ ∼ −4 i ¯ i ¯ C7 = 2 = 7 × 10 (26) ≡ µqQiFq + µ`F`Li (20) 10.5 GF VtsVtbM where ψ is the wave function of the bound state and M using the parameter values (17) with ΛHC = 400 GeV, is the mass of the constituents, that we continue to take well below the experimental limit of ∼ 0.02 [98]. Under to be equal for simplicity. For unequal masses, M is the scaling of parameters described at the end of section replaced by twice the reduced mass. 4.1, this prediction remains unchanged. 6

6 Λ 0 HC = 400 GeV 5 N = 4 -1 4

Σ Σ g XENON100 limit 3 N = 3 composite

10 -2

(TeV) (Banks et al.) dark matter Σ log m 2 -3 1 N = 2 PandaX-II 0 -4 0.4 0.5 0.6 0.7 0.8 0.9 1 1 1.5 2 2.5 3 3.5 4 m (TeV) S log10(mΣ / GeV)

Figure 4: Left: hyperbaryonic dark matter mass mΣ versus constituent mass mS for NHC = 2, 3, 4 and ΛHC = 400 GeV. Heavy dots show the value of mΣ in the limit mS → 0. Right: predicted gyromagnetic moment for the NHC = 3 model (orange dashed) and current constraint from direct detection [105] (solid blue), based on rescaling of earlier limit from ref. [103] (solid red). Dashed red is an estimate of the improved relative sensitivity at low mass in recent experiments, compared to the XENON 100 [104] limit.

Similarly for c → uγ, we predict a transition magnetic LHC constraints prevent taking MF` to be smaller than ˜ ˜0 moment like (25) with λi → λi, eq. (5). Using the values the TeV scale, leaving little room for increasing aµ. (17), the contribution to the c → uγ amplitude is ∼ 10−2 smaller than the SM contribution. Likewise, the contribution to d → sγ is orders of magnitude below the 5. COMPOSITE DARK MATTER 0 + − limit from KS → π ` ` decays (see section 7.7 of [99]). The η0-like SS¯ bound state is unstable to self- annihilation into µ+µ− through φ exchange, but the hy- 4.3.2. FCNC leptonic decays perbaryonic bound state Σ ≡ SN is a stable dark matter candidate. Determination of its thermal relic density can ¯ The radiative decays ì → `jγ proceed analogously to depend upon the annihilation of the constituent SS par- those of the quarks, due to mixing with the heavy com- ticles above the SU(NHC ) confinement transition, as well ∼ ¯ posite lepton whose mass is MF` = 3.5 M for our fiducial as ΣΣ annihilations following confinement. For the para- parameter choice. The induced transition magnetic mo- meters of interest, with ΛHC ∼ 400 GeV and mS . 1 TeV, ment is it has been shown [95, 102] that the resulting relic density is too small (by a factor of ∼ 1000). However since eλ λ ζ m µ = i j i (27) hyperbaryon number is conserved,2 it is possible that an ij M M F` asymmetry between S and S¯ was produced in the early 2 3 universe, so that Σ can be asymmetric dark matter. We and the partial decay width is δΓ = µijmi /8π [100]. This gives limits make this assumption here, without attempting to account for the origin of the asymmetry. −4 |λ1| < 7.5 × 10 , |λ3| < 0.56 (28) From the nonrelativistic potential model, the DM mass can be numerically determined as a function of mS. In respectively, from µ → eγ and τ → µγ that are more the range 1 < mS/ΛHC < 10, we find the approximate stringent than (19). fit ∼ 2 mΣ = 3.8(1 − NHC ) + NHC ΛHC + (0.3 + 0.8 NHC ) mS 4.3.3. Muon anomalous magnetic moment (30) The dependence on mS is shown in fig. 4(left), including The contribution to the muon anomalous magnetic mo- the limit mS → 0 (heavy dots on y-axis), obtained from ment is the relativistic version of the potential model. Although S has no direct interactions with nuclei, it 2 ± ± (g − 2) 2 mµ ζ gets a magnetic moment at one loop (with φ or µ in aµ = = |λ2| (29) 2 MF` M

−11 For our benchmark values, we get aµ ∼ 10 , which is 300 times too small [101]. It will be seen below that 2 note that S, Ψ and φ∗ can be assigned hyperbaryon number +1 7

_ _ g g q φ γ, l * scalar HC mesons, through the diagrams in fig. 5. We γ, Z,W Z,W q φ q q φ* l ψ work in a region of parameter space where ΛHC is not q (b) q (a) too much lower than the constituent masses M, which _ _ g g (γ) q ψ q ψ avoids the complications of quirks [89]. In the quirky _ _ q gψ g q gψ g (γ) regime, there are many excited quantum states to scatter (c) (d) into, that can efficiently radiate hypergluons and settle to the ground state before self-annihilating. Then partons Figure 5: Resonant production of HC bound states leading to with energy above that of the resonance could still be dileptons (a), dijets (b-d) or diphotons (d). effective for production. On the other hand for ΛHC . M, the resonances are well-separated and can only radiate QCD gluons or elec- 2 2 troweak gauge bosons since the HC glueball mass is too the loop), depending upon R ≡ mS/mφ heavy, ∼ 7ΛHC [106] for NHC = 3. In this regime, the 2 parton center of mass energy must match the mass of the e|λ2| mS µS = f(R) (31) resonance, and we can use similar techniques for comput- 32π2 m2 φ ing resonant production of bound states as for J/ψ and + − with Υ in e e collisions. This approximation misses events in which the bound state is produced by partons above Z 1 u2 u(1 − u) the threshold for the resonance, plus QCD radiation to f(R) = du + (32) 1 − Ru 1 − R + Ru take away the excess energy. We defer study of such 0 higher-order corrections. in the approximation of neglecting mµ (see appendix C for details). The loop function diverges logarithmically 6.1. Production cross section as R → 1 in this approximation; for example if mS = 0.9 mφ, f(0.81) = 1.3, while for mS = 0.8 mφ, f(0.64) = 0.93. In general, the production cross section from colliding If Σ has spin, it inherits a magnetic moment from its protons to form a narrow resonance R from partons pi, p can be expressed as S constituents. This will be the case if NHC is odd. If j N = 3, the wave function of SaSbSc is in hyper- HC abc 16π2 (2J + 1)N Γ(R → p p ) color space, and if the spatial wave function is s-wave, σ(pp → R) = c i j (33) total antisymmetry demands that the spins be aligned, mRs (2s1 + 1)(2s2 + 1)Nc1Nc2 like the ∆ baryon in QCD. The quark model then pre- Z 1 dx · [f (x)f (m /sx) + {i ↔ j}] dicts that Σ has a magnetic moment that is three times x i j R that of S. mR/s

The gyromagnetic ratio of Σ is given (for odd NHC ) by where fi is the parton distribution (PDF) for pi, 2si + 1 gΣ = 4NHC µsMΣ/e. Ref. [103] derived an upper limit and Nci are the number of spin states and colors respec- on g versus DM mass based upon early XENON100 data tively of the incoming particles, 2J + 1 and Nc likewise [104]. We update their limit by comparing the relative for the resonance, and Γ is the partial width for the de- sensitivities of that search to the recent PandaX-II result cay that is inverse to the production process. (In the case [105] in the high mass regime, to constrain our model. of incoming gluons, the doubling of PDFs in the second ∼ Allowing mS to vary between ΛHC = 400 GeV (the value line is correct, compensating for the factor of 1/2 in the that maximizes ζ if M = 1 TeV) and 1 TeV, we rescale phase space for decay of R into identical particles.) The 2 1/2 λ2 by (ζmax/ζ) where ζmax = 0.0037 for the fiducial generalization to resonances carrying SU(2)L rather than model parameters, and ζ is smaller for mS < M. This QCD quantum numbers is obvious. keeps the predicted B decay signals constant while vary- ing mS, if all the couplings scale in the same way. Using (30), gΣ and mΣ are determined as a function of mS. The 6.2. Predicted widths result for NHC = 3 is shown in fig. 30(right). It translates to an upper limit mS . 800 GeV. (ζ is not appreciably The decay widths of bound states have been quantita- decreased at this value of mS.) tively addressed in regimes where the constituent masses are either much heavier than the confinement scale [89] or much lighter [88, 107]. We are interested in the case 6. COLLIDER CONSTRAINTS where ΛHC ∼ 0.4 M (with the possible exception of ¯ mS ΛHC , but this is not relevant here because the SS

Because of hypercolor conﬁnement, only HC-neutral bound states cannot be produced at LHC). For NHC = 3, bound states of the new particles φ, Ψ, S can be produced the potential model predicts that the kinetic energy of the in high-energy collisions. We expect the dominant mode constituent is 0.3 of its mass energy, so the nonrelativis- for discovery to be resonant production of vector and tic approximation is not very bad. Ultimately, lattice 8 calculations should be done to make more quantitative 6.2.2. φφ¯ resonance predictions. ∗ We ignore decays of bound states into hypergluons, There are four kinds of ρφ = φφ vector bound states, since we work in a regime where any hyperhadrons are three components in an SU(2)L triplet and one singlet. too heavy to be produced. Even if mS < ΛHC , the Since the SU(2)L dynamics are not important at the TeV ΠS = SS¯ pseudoscalar is not a light pseudo-Nambu scale, we average over the isospins of the bound state Goldstone boson, because the approximate chiral U(1) constituents. The angular momentum of ρφ is purely ﬂavor symmetry is anomalous, like for the η0 of QCD. orbital, so the constituents are in a relative p-wave for

Likewise the glueballs of the SU(NHC ) are also too heavy. which ψ(0) = 0. Therefore the width depends upon the Ref. [89] calculates (in terms of |ψ(0)|2) the decay derivative of ψ at the origin. Following [109], we find widths assuming that the bound state constituents do not that the SU(2)L contribution to the width is form a resonance of definite spin, but these are straight- 2 3 3π |∇~ ψ(0)| forward to rescale for the physical eigenstates of spin. Γ(ρ → W ∗ → q q¯ ) = N α2 (37) φ i j 2 HC 2 m4 Doing this gives results in agreement with the treatment ρφ of bound state decays in ref. [108]. In our model, the ΨΨ¯ states can be in the QCD color singlet or octet rep- for decay into a single generation of approximately massless quarks. For decays into leptons, the result is 1/3 resentations. Given that ΛHC ΛQCD, this distinction is unimportant for the dynamics since the lifetime of the times smaller (from lack of QCD color), giving a branch- bound state is much shorter than the QCD hadronization ing ratio of 3/4 to quarks and 1/4 to leptons. Hence this time scale. process leads to both dijet and dilepton signals. For decays into muons, there is an additional, potentially larger contribution from S exchange (that interferes with the W exchange contribution, but we ignore this interference for 6.2.1. ΨΨ¯ resonances purposes of estimation), given by 2 2 ~ 2 + − 2π |λ2| |∇ψ(0)| ¯ Γ(ρ → µ µ ) = NHC Starting with the ΨΨ bound states, there is the pseu- φ 3 4π (m2 + 1 m2 )2 doscalar Π that can be produced by gluon fusion (di- ρφ 4 S Ψ (38) agram (d) of fig. 5) or the vector ρΨ coming from qq¯ annihilation, diagram (c). The respective decay widths are [89] 6.2.3. Composite fermion

128 π |ψ(0)|2 ∼ Γ(Π → gg) = N α2 (34) For decay of the composite fermion to qg, taking mφ = Ψ 27 HC s m2 ΠΨ mΨ as before, we find 8 π |ψ(0)|2 Γ(ρ → uu¯) = N α2 (35) α |λ˜(0)|2 Ψ 27 HC s m2 Γ(F → gq ) ∼ N s i |ψ(0)|2 (39) ρΨ q i = HC 2 2mΨ 4π |ψ(0)|2 Γ(ρ → e+e−) = N α2 (36) Ψ 9 HC m2 where i is the generation index of the quark and as usual ρΨ the prime is for couplings to up-type quarks. The squared coupling must be absorbed into sum over quark flavors in The last one, unlike the previous two, involves only the the parton luminosity factor for production via qg fusion. color singlet state so it does not entail an average over QCD colors. We include it because it is potentially relevant for the dilepton final state. However count- 6.2.4. Vertex correction ing channels, it predicts a branching ratio of Bl+l− = (α/α )2/2 ∼ 2.5 × 10−7 into electrons or muons, imply- s = For all of the annihilation decays, a more quantitative ing that the dilepton channel is actually unimportant for estimate can be made by taking account of perturbative this resonance. corrections that dress the annihilation vertex. For charmonium and upsilon, the correction from a gluon loop is known to be important, and we adopt the analogous correction for the HC gluon exchange, with α evaluated at 3 For example, decay via a single virtual gauge boson into fermions HC ff¯ gets a factor of 4/3 to correct for the fact that only the spin-1 the scale of the constituent mass. The correction factors state contributes in the average over spins of the initial con- depend upon the annihilation process, and in analogy stituents. Similarly decays to two gauge bosons occur only for with QCD we take [110] spin-0 resonances, requiring a factor of 4 correction to the spin- averaged rate. Likewise the results of [89] ignore QCD color 2 ( 2 ) (N − 1)αHC (20 − π )/3, Π → gg correlations of the constituents and must be rescaled for states C = 1 − ψ (40) ¯ of definite color. 8π 16/3, ρψ → ff 9

_ gg → Π → jj, qq → ρ → jj 100 0 Ψ Ψ gg → Π → γγ 10 Ψ

-1 10 10

-2 CMS gg

10 (fb) 1 γγ B A (pb) -3 ATLAS 10 ATLAS 0.1 σ × σ × -4 A = 50 % vector 10 pseudoscalarρ Λ = 0.42 m Ψ HC Ψ heavy quark F √s = 13 TeV, N = 3 -5 0.01 HC 10 Π Ψ Λ = 0.4 m √s = 13 TeV, NHC = 3 HC Ψ -6 q 10 500 1000 1500 2000 2500 1 2 3 4 5 6 m (GeV) m (TeV) Π R Ψ

Figure 6: Dashed curves: limits on resonant dijet produc- Figure 7: Production cross section for pseudoscalar bound tion cross section times acceptance versus resonance mass state ΠΨ times branching ratio for decay to two photons. from ATLAS [111] and CMS [112], along with predictions for Dashed curve is ATLAS limit [114], solid curve is predicted ΠΨ pseudoscalar (solid curve), ρψ vector (dotted curve) and value for NHC = 3 and ΛHC /mΨ = 0.4. heavy quark Fq (dot-dashed curve) bound√ states, in the case of NHC = 3, acceptance A = 100% and s = 13 TeV center -1 of mass energy. Predictions for the optimal value (in terms of 10

LHC sensitivity) ΛHC /mΨ = 0.4 are shown. -2 10

ρφ → µµ with αHC evaluated at the scale µ∗, twice the inverse -3 Bohr radius. We have assumed that the known NHC = 10 B (pb) ATLAS dilepton

3 results generalize to other values by rescaling by the σ number of hypergluons. -4 10

-5 6.3. Resonant search constraints 10 500 1000 1500 2000 2500 3000 mρ (GeV) Combining the decay widths with the general formula φ for the production cross section (33), we ﬁnd that it de- Figure 8: Production cross section for pseudoscalar bound pends upon the same combination ζ, eq. (14) that ap- state ρφ times branching ratio for decay to two leptons. peared in the dimension-6 FCNC operators. One ﬁnds a Dashed curve is ATLAS limit [115], solid curve is predicted production cross section for the color octet plus singlet value for NHC = 3 and ΛHC /mΨ = 0.4. states ΠΨ given by

3 2 8 π αsNHC σ(pp → ΠΨ) = ζCLgg (41) The loop correction factor (40) is 0.7 for this process. 3s The corresponding prediction is also plotted in fig. 6.4 where Lgg is the parton luminosity factor for gluon fu- We infer a limit of mΨ > 820 GeV for the fiducial model. sion. The perturbative correction (40) is C = 0.5 at At larger values of mΨ/ΛHC = 5, 10, this limit becomes more stringent, m 1 TeV, approximately independent mΨ = 2.5ΛHC , similar to the value that occurs in QCD Ψ & for J/ψ, and ζ = 0.0037 as in the previous sections. of the value of mΨ/ΛHC . In fig. 6 we plot this prediction along with current con- In addition to the ΠΨ → gg channel, the color singlet straints from ATLAS [111] and CMS [112] resonant dijet pseudoscalar can decay to two photons. Its decay width searches, assuming an acceptance of events passing exper- into gluons is smaller than the color average (34) by a imental cuts of A = 50%, comparable to that of various models tested in the searches. The resonance mass must exceed 2.3 TeV, which is satisfied for our fiducial model 4 with mΠ = 3.6 TeV. Keeping mΨ/ΛHC fixed, this limit It is possible that the true limit from the vector resonance is weaker since we have assumed 100% branching ratio into two would allow mΨ to be no lower than 650 GeV. Similarly for production of the color octet vector ρ jets. In the QCD charmonium and upsilon systems, decay to Ψ three gluons is much more likely, resulting in three or more jets. resonance for NHC = 3 we find In this case the reconstruction of the resonance is challenging, and has not been carried out in experimental searches, to our 64 π3α2N knowledge. However the three-jet process is higher order in α σ(pp → ρ ) = s HC ζCL (42) s Ψ 9 s qq¯ and should therefore be subdominant at the TeV scale. 10

_ _ _ _ q _ + N = 4 L q W N = 3 q γ, Z φ _ q γ, Z φ _ γ, Z φ N = 2 g N = 2 S γ ψ φ* γ S φ γ Λ = 100 GeV γ ψ g q N = 3 HC q φ* q φ* φ* L q W− 0.1 (a) (b) (c) β = 1/4 q CMS _ q _ g _ q ψ _ _ q ψ _ g q ψ γ S L ψ φ* ττ _ _ ψ _ φ S L g γ q g ψ _ q g ψ q g ψ _ N = 4 g q q (pb) (f) CMS

(d) (e) 2

g L g q g L 0.01 µµ * φ* φ* g φ q S γ q ψ q φ γ

_ _ σ × β S _ ψ φ* _ q ψ L q ψ g q ψ L (g) q (h) g (i) q

0.001 Figure 9: Pair production of composite states 200 400 600 800 1000 1200 1400 1600 mρ (GeV) factor of 16/9, and its branching ratio into photons is 2 4 2 Figure 10: Cross section for pair production of composite Bγγ = 9α qψ/2αs [113]. The cross section for pp → leptoquarks times branching ratio squared for decays into µ or ΠΨ → γγ is therefore (9/128)Bγγ times eq. (41). The predicted cross section is shown along with the ATLAS τ plus jet (j). Solid is CMS upper limit for jjµµ [117] or jjττ [118] ﬁnal states, dotted is prediction from qq¯ → ρρ∗, dashed diphoton limit [114] in ﬁg. 7; it constrains m > 1.7 TeV. Π is prediction from gg → ρρ∗, assuming low compositeness The ρ resonance has electroweak production cross sec- φ scale ΛHC = 100 GeV and massless S constituents. tion

2 π3α2N σ(pp → ρ ) = 2 HC ζ L (43) aged over PDFs takes the form φ s p qq¯ Z 1 Z 1 where ζ = |∇ψ(0)|2/m5 = (µ /m )5/96π ∼ 5 × 10−5 σ = dx dx f (x )f (x )Θ(x x s − 4m2)σ ˆ (45) p ρφ ∗ ρφ = 1 2 1 1 2 2 1 2 c from the potential model for p-wave states, evaluated 0 0 at ΛHC = 0.4 mφ. The ATLAS dilepton limit [115], where mc is the mass of the composite state. The largest slightly rescaled to account for dominant decay into of cross sections are for qq¯ → ΨΨ¯ and gg → ΨΨ,¯ ρφ to muons (see ref. [49] for details) is plotted in fig. 8 along with the prediction for branching ratio of 100% 8πα2 N + − σˆ = s HC 2m2 +s ˆ χ (46) into µ µ (this neglects decays into electroweak gauge qq¯ 27s ˆ2 Ψ bosons). The resulting limit m > 2 TeV is stronger ρφ 2 than that from diphotons but weaker than that from di- παs NHC χ 2 σˆgg = − 7 + 31mΨ/s jets for the ΠΨ and ρΨ states respectively. 3s ˆ 4 The cross section for heavy quark F production is ! q 1 + χ + 1 + 4m2 /sˆ + m4 /sˆ2 ln (47) Ψ Ψ 1 − χ NHC αs X (0) σ(pp → F ) = ζ |λ˜ |2L (44) q 2s i gqi i p 2 with χ = 1 − 4mΨ/sˆ ands ˆ = x1x2s. The result assuming couplings (17) and 100% branching We find that the composite mass mc must be as small into dijets is plotted in fig. 6, giving a weaker constraint as possible to get interesting constraints from this process. This occurs for m Λ in the ΨS¯ bound states, than those from ρΨ and ΠΨ resonant production. S HC for which the potential model gives ∼ mc = (0.75 + 1.8NHC )ΛHC + mΨ (48) 6.4. Pair production of bound states These are leptoquark states that decay into q`, and have The wave function at the origin (the ζ factor) sup- been searched for by ATLAS [116] and CMS [117, 118]. presses the predicted resonant production signficantly be- The searches look for two jets and two leptons, either low the cross section that would govern the creation of µ+µ− or τ +τ − respectively, for second or third gener- free pairs of hypercolored constituents. This penalty is ation leptoquarks. The limits are on the production avoided if bound states are pair-produced by hadroniza- cross section times β2 where β is the branching ratio tion in the HC sector, as depicted in fig. 9. Then each into the final state searched for. In our model, β ≤ 1/2 possible kind of final state will occur with a relative prob- since ρ can decay into either a charged lepton or neu- ability not much smaller than 1 since for every pair there trino with equal probability. Moreover the branching ra- 2 2 2 are only three kinds of particles (Ψ, φ, S) that can ap- tio into muons is |λ2| /(|λ2| + |λe| , and that into tau 2 2 2 pear from the vacuum. The only other bottleneck is that is |λ3| /(|λ2| + |λe| , typically giving a further reduc- the incoming partons must have at least enough energy tion in β. (Recall that LFV constraints on λ1 make the to make both bound states. Then the cross section aver- branching into e+e− negligible in our model.) 11

The predicted cross sections and observed limits are shown in ﬁg. 10 for a model with ΛHC = 100 GeV and 10 N = 2 N = 4 mS ΛHC , which is a limiting case for lowering the masses of bound states containing S. We have assumed ∼ λ3 ∼ λ2 so that β = 1/4. The most stringent limits come

from the search for final state muons, with production via B (fb) the gg → ΨΨ¯ cross section (47). It requires the lepto- CMS excited muon σ × quark mass to be & 1.1, 1.2, 1.26 TeV for NHC = 2, 3, 4, hence mΨ & 670, 590, 470 GeV respectively. Other bound states not involving S as a constituent, 1 although they have potentially interesting signals, are too heavy to be produced at a significant level relative 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 m (TeV) to current constraints from LHC. The composite lepton Fl ∗ ∗ Fl = Sφ arises from electroweak qq¯ → φφ pair production, with parton-level cross section Figure 11: Cross section for production of composite heavy leptons decaying to µ plus photon. Solid (orange) curve 3/2 2 2 is CMS upper limit, dashed are predicted values from Fl πα2NHC sˆ 1 − 4M /sˆ σˆ = 2 2 2 2 (49) pair production, assuming low compositeness scale ΛHC = 4 (ˆs − mW ) + mW ΓW 100 GeV and massless S constituents. Neighboring solid 0 curves are from production of Fl-leptoquark pairs, eq. (50). in the simplifying approximation g g (sin θW → 0). CMS has searched for excited muons and taus decaying to the normal state plus photon [119]. Figure 11 shows kinds of interactions with SM fermions are allowed, and the limit on excited µ production, which is more strin- these break the approximate flavor symmetries of the SM gent than that for excited τ, versus the predicted cross through the six couplings λ to leptons and λ˜ to quarks. sections (100% branching of F → µγ is assumed). Again i i l The leptoquark is the ΨS¯ vector bound state. we take the extreme choice with Λ = 100 MeV and HC There is a limited region of parameter space allowed by m = 0 to obtain light enough m to fall within the cur- S Fl FCNC constraints on meson-antimeson oscillations (me- rently probed mass range; larger values of Λ and m HC S diated by the ΨΨ¯ vector bound states), direct dark mat- are not excluded. Comparison with fig. 10 shows that ter searches, and LHC searches for resonant dijets, which F is excluded up to similar masses as the leptoquarks, l require the Ψ and φ constituents to be near the TeV scale, m 1.3 − 1.5 TeV, depending upon N . Fl . HC and the new confinement scale Λ must be 100 GeV. The remaining mechanism to produce the lightest HC & (Our benchmark model takes Λ = 400 GeV.) The dark states is diagram (g) of fig. 9, which has partonic cross HC matter is a baryon-like bound state Σ = SNHC with mass section ∼ (1 − 5) TeV. If NHC is odd, Σ can have a magnetic (0) α |λ˜ |2 dipole moment that requires mΣ < 2.5 TeV to avoid di- σˆ = N s i p1 − 4M 2/sˆ (50) HC 24s ˆ rect detection. For very light dark matter constituent and low con- neglecting the quark mass (a good approximation since finement scale with mS ΛHC ∼ 100, the new bound the top quark PDF is very small). This produces the states (leptoquarks and heavy lepton partners) are light heavy lepton Fl in conjunction with a leptoquark, and enough (∼ 1 TeV) to be pair-produced and significantly would be probed by the CMS search [119]. The predicted constrained by ATLAS or CMS searches for leptoquarks cross section (assuming 100% Fl → µγ) is also shown in and excited leptons. The model also predicts produc- fig. 11 and happens to be close to that coming from Fl tion of the leptoquark and heavy lepton in association, pair production. Since the leptoquark is produced singly, leading to the unusual signal of two muons, a jet and a it does not produce the same signal as for the existing photon. The dark matter dipole moment is below direct searches that assume pair production and would require detection constraints for small mS. a new analysis, searching for the unusual final state of The predictions are sensitive to nonperturbative as- two leptons, a jet and a photon. pects of confinement, especially bound state masses and wave functions. We have used a simple potential model that works reasonably well for known examples from 7. SUMMARY AND CONCLUSIONS QCD, but which is not necessarily reliable for the parameters of interest here, and neglects spin interactions We have presented a particularly simple realization of between the fermionic consituents. Lattice studies of the composite leptoquarks to explain the anomalous B de- masses and decay constants would allow for a useful im- cay ratios RK(∗) , as well as providing a composite dark provement. Nevertheless it seems likely that the kind of matter candidate. The new ingredients are an SU(NHC ) model presented here should give rise to other anomalous confining gauge sector under which three types of mat- signals in FCNC, dark matter or collider searches. It is ter fields Ψ, S, φ transform as fundamentals. Only two encouraging that the experimental status of the B decay 12 anomalies should become more clear from future data at e+e−. For J/ψ we find ψ(0)2 = 1/(0.5 fm)3, close to the LHCb and Belle II during the next few years [120]. measured value 1/(0.53 fm)3, while the mass is predicted to be 3.0 GeV, compared to the measured 3.1 GeV. The 2 2 Acknowledgment. I thank I. Brivio, J. Evans, B. model predicts ψ(0)Υ/ψ(0)J/ψ = 4.6, while experimen- Gavela, J. Jiang, D. London, J. Martin Camalich, Y. tally one finds 4.7, by comparing the decay rate formula Kats, M. Redi, M. Trott and for helpful discussions, and (including the correction (40)) to the observed partial A. Urbano for pointing out ref. [96]. This work was sup- widths. The worst agreement is for mΥ, predicted to be ported by NSERC (Natural Sciences and Engineering Re- 8.6 GeV versus the measured value 9.5 GeV. search Council of Canada). For p-wave states with µ5/2 ∗ −µ∗r/2 ψp(r) = √ re (A5) Appendix A: Potential model for bound states 96π we find the energy 1. Nonrelativistic constituents µ2 c σ E = m + m + ∗ − α α µ + 5 (A6) p 1 2 8m 4 HC ∗ µ We first consider bound states where all the con- r ∗ stituents are heavier than ΛHC and hence nonrelativis- For baryons in an s-wave state, we take the ansatz tic. Following [109, 121], we estimate the size of hyper- ψ ∼ e−µ∗(r1+r2+··· ) and the energy is color mesonic bound states using a variational method, N XHC µ2 with hydrogen-like ansätzefor the wave functions. The E = m + ∗ (A7) b i 8 m Coulomb-like contribution to the potential from hyper- i=1 i gluon exchange is [122] 1 5cα αHC µ∗ 35σ + 2 NHC (NHC − 1) − + f 16(NHC − 1) 8µ∗ αHC 1 αHC Vc = − NHC − ≡ −cα (A1) 2r NHC r (note that the Coulomb attraction between qq in an anti-

symmetric hypercolor state is 1/(NHC − 1) times weaker for a hyperquark-antiquark pair in the hypercolor singlet than that between qq¯). Comparison to the nucleon of state. There is a linear confining potential QCD motivates the correction factor f = 0.065, indicat- ing that the string tension is much smaller between qq Vl = σr (A2) than qq¯ states. ∼ 2 where σ = 2(NHC − 1)ΛHC is the string tension and ΛHC is identified with the Λ parameter of the 0-flavor running 2. Heavy-light or relativistic systems coupling in the MS scheme. The coefficient is deduced for the case of N = 3 from lattice studies [123] and HC In the case where m < Λ , one or more constituents its generalization to other values of N is inferred from S HC HC is relativistic. Then in the center of mass system of large-N scaling. The energy of an s-wave state with wave a heavy-light meson we have mass plus kinetic energy function 2 p 2 m1 + p /(2m1) + p + m2 in the Hamiltonian [126]. −µ r/2 µ3/2 Still using the wave function ansatz ψ ∼ e ∗ for s- ∗ −µ∗r/2 ψs(r) = √ e (A3) p 2 2 8π wave states, the expectation value of p + m can be found by Fourier transforming ψ and evaluting the in- is tegral in the momentum eigenstate basis. The result is a complicated analytic function that can be fit to the 2 µ∗ cα σ simpler form Es = m1 + m2 + − αHC µ∗ + 3 (A4) 8mr 2 µ∗ ! D E m 1.6 p 2 2 ∼ p + m = µ∗ 0.84 + 0.52 (A8) where mr is the reduced mass of the constituent parti- µ∗ cles, and αHC implicitly depends on µ∗ since it is the running coupling evaluated at that scale. We thus mini- (At m = 0 the exact result is 8µ∗/(3π).) We find that mize the energy numerically. We use the 4-loop, 0-flavor this approximation is good to better than 1% for m < µ∗. expression for the running coupling given in [124, 125]. The kinetic energy in (A4,A6) is then replaced by For µ∗ < ΛHC the perturbative coupling diverges and we 2 q µ∗ 2 2 cut it off at αHC = 2. (Larger cutoffs lead to numerical m1 + + p + m2 (A9) artifacts in the minimization.) 8m1 This model gives reasonable results for J/ψ and Υ us- For baryons with some constituents light, the appropri- ing mc = 1.3 GeV, mb = 4.2 GeV for the quark masses, ate replacement is obvious. The coupling αHC should be and ΛQCD = 200 MeV. For these systems ψ(0) is mea- evaluated taking into account running with the appropri- surable through the electromagnetic decays via γ∗ to ate number of light constituents. 13

Appendix B: Transition moments from It is pertinent to compare this to the corresponding am- compositeness plitude with no photon, which is just the mass mixing amplitude, Here we semiquantitatively estimate the transition magnetic moment interactions between heavy composite (∼) λi ψ(0) i fermions F and their elementary SM counterparts, f , A(F → fi) = −i √ u¯fi PR uψ(S) = −iµf u¯fi PR uψ(S) R R 2µ following the same formalism used by ref. [96] for radia- (B5) tive decays of the Z boson to quarkonium bound states. This allows us to infer that the transition moment is re- The formalism is appropriate for nonrelativistic systems, lated to the mass mixing in a definite way. which is not a very good approximation in our case since For leptons, the decay amplitude from diagram (a) of we prefer the constituent masses and confinement scale fig. 3 is to be of the same order. Hopefully it gives a reasonable estimate, which would require a dedicated lattice study µ i (Q − k) to improve upon. A(F` → ìγ) = −ieµ` u¯ìL 1 2 2 uSR 4 mF + mφ The amplitude for F → fiγ can be written as ` i µ` µν (∼) Z 4 → −ie u¯ iσ k u (B6) d q m2 ìL ν SR A(F → f γ) = −iλ u¯ O (q, k, Q) χ(Q, k) F` i i fi (2π)4 γ (B1) where in the second line we used the Gordon identity where the momenta are indicated in fig. 3. Q is the 0 µ 0 0 0 µ 0 µ 0 µν momentum of they decaying heavy F , k is that of the (p + p ) u¯LuR = m u¯Rγ uR + m u¯Lγ uL +u ¯Liσ kν uR photon, q is the relative momentum between the heavy (B7) µ fermion constituents and χ(Q, q) is the Bethe-Salpeter and took mφ ' mF` /2. The extra terms going as γ in wave function for the bound state. The operator Oγ the Gordon identity are canceled by diagrams (c) and (d) comes from the part of the diagram that is introduced of fig. 3, which serves as a check on the sign. We ignored by insertion of the photon vertex, which depends upon terms going as kµ that do not contribute to the ampli- what kind of particle F (f) is. For down-like quarks, the tude because of transversality of the photon polarization photon attaches to both Ψ and φ±; for up-like quarks vector. it attaches only to Ψ (since φ0 is the other constituent), Similarly for up- and down-type quarks respectively we while for leptons it attaches only to φ± since S is neutral. obtain For simplicity we take the constituent masses to be i equal so that in the absence of exchanged momenta, each 2e µ0 A(F → u γ) = i u u¯ iσµν k u carries half of the momentum Qµ of the bound state. The u i 3 m2 uiL ν ΨR Fu general form of O can then be written as e µi A(F → d γ) = −i d u¯ iσµν k u (B8) µ d i 3 m2 diL ν ΨR i(−Q + 2q + k) Fd Oγ = −ieqφ 2 2 PR (−Q/2 + q + k) − mφ 0i ˜0 ˜0 i i where µ u = λiΛHC with λi given in eq. (5), and µd = − ieq γµP (B2) ˜ ψ(S) R λiΛHC . After rotating the heavy and light fields to elim- Q// 2 + /q − k/ − mψ(S) inate mass mixing, these lead to magnetic moments involving just the light states. where PR projects onto right-handed chirality. The spinor wave function takes the form

2π ~q 2 Appendix C: DM dipole moment χ(Q, q) = √ δ q − u (Q/2 + q) ψ˜(q) (B3) 2µ 0 2µ ψ(S) We give details for the one-loop contribution to the where µ is the reduced mass of the bound state con- magnetic dipole moment of the elementary dark matter stituents, uψ(S) is the Dirac spinor for the fermionic mem- particle S, with diagrams shown in ﬁg. 12. They turn ˜ ber, and ψ is the Fourier transform of the spatial wave out to be infrared convergent in the limit mµ → 0 for the function. internal muon, hence we make this simpliﬁcation. The For nonrelativistic systems, the wave function is 1PI diagrams (a,b) respectively give strongly peaked at small ~q and it is a good approximation to set q = 0 in O . Then the factor R d3q/(2π)3ψ˜(q) = Z 1 Z 1−xZ 4 / 0 µ / γ 2 d l (l + p/ )γ (l + p/)PR e|λ | 2 dx dy u¯ 0 u ψ(0), the spatial wave function evaluated at the origin, 2 (2π)4 p [(l + xp0 + yp)2 − M 2]3 p and the amplitude becomes 0 0 a

(∼) Z 1 Z 1−xZ 4 0 µ λi ψ(0) 2 d l (2l − p − p ) /lPR e|λ | 2 dx dy u¯ 0 u A(F → fiγ) = −i √ u¯fi Oγ (0, k, Q) PR uψ(S) (B4) 2 4 p 0 2 2 3 p 2µ 0 0 (2π) [(l − xp − yp ) − Mb ] 14

q terms linear in /q. Both of these can be put into the form µ µ l γ /q up to terms going as q , that give a vanishing contri- l+p l+p’ bution when contracted with the external photon polar- p p’ p p’ ization vector. These are directly related to the magnetic ν µ l+p l+p’ l moment since −iσµν q → γ /q in this way. q In diagram (b), we again use the Dirac equation to (a) (b) 0 rewrite pP/ R → mSPL, PLp/ → mSPR, and reexpress pµ, p0µ as linear combinations of (p+p0)µ and qµ, of which only the former contribute to the amplitude. Using the p p’ 0 µ ν µ p p’ Gordon identity, (p + p ) → iσµν q → −γ /q. We find that both diagrams (a,b) contribute with the same sign. q q A check on this relative sign is provided by gauge invari- ¯ µ (c) (d) ance: the coupling Sγ SAµ is forbidden, and must not be generated by loops. The divergent parts of diagrams Figure 12: Contributions to elementary dark matter magnetic (a,b) contribute with the same magnitude and sign to the dipole moment. vector current, and these cancel the contributions from (c,d). Such a cancellation would not occur if (a) and (b) came with opposite signs. 2 2 where, defining u = x + y, Defining R = mS/mφ, the final expression for the mag- µ netic moment (which is the coefficient ofu ¯ 0 γ qu ) is 2 2 2 p / p Ma = (1 − u)[mφ − u mS] 2 2 2 e|λ |2 Mb = u [mφ − (1 − u) mS] (C1) 2 µS = 2 2 (fa(R) + fb(R)) (C3) 32π mφ As usual, one shifts the loop momentum variable to sim- with plify the denominators. Only the finite parts contribute to the dipole operator. To extract the dipole moment, Z 1 u2 fa(R) = du one makes the replacements 0 1 − Ru Z 1 µ µ µ u(1 − u) PLpγ/ → PL(p/ + /q)γ → (PRmS + PL/q)γ fb(R) = du (C4) µ 0 µ µ 0 1 − R + Ru γ p/ PR → γ (p/ − /q)PR → γ (mSPL − /qPR) (C2) In the limit of light dark matter, fa(0) + fb(0) = 1/2. in diagram (a), using the Dirac equation (and similarly 0 µ µ µ µ letting p/ γ → mSγ , γ p/ → mSγ ), and keeping the

[1] R. Aaij et al. [LHCb Collaboration], “Test of b → s`+`− transitions in the light of recent data,” lepton universality using B+ → K+`+`− de- arXiv:1704.05340 [hep-ph]. cays,” Phys. Rev. Lett. 113, 151601 (2014) [7] A. Celis, J. Fuentes-Martin, A. Vicente and doi:10.1103/PhysRevLett.113.151601 [arXiv:1406.6482 J. Virto, “Gauge-invariant implications of the [hep-ex]]. LHCb measurements on lepton-flavor nonuniver- [2] R. Aaij et al. [LHCb Collaboration], “Test of lep- sality,” Phys. Rev. D 96, no. 3, 035026 (2017) ton universality with B0 → K∗0`+`− decays,” doi:10.1103/PhysRevD.96.035026 [arXiv:1704.05672 JHEP 1708, 055 (2017) doi:10.1007/JHEP08(2017)055 [hep-ph]]. [arXiv:1705.05802 [hep-ex]]. [8] M. Ciuchini, A. M. Coutinho, M. Fedele, E. Franco, [3] L. S. Geng, B. Grinstein, S. Jäger,J. Martin Camalich, A. Paul, L. Silvestrini and M. Valli, “On Flavourful X. L. Ren and R. X. Shi, “Towards the discovery of Easter eggs for New Physics hunger and Lepton Flavour new physics with lepton-universality ratios of b → s`` Universality violation,” arXiv:1704.05447 [hep-ph]. decays,” arXiv:1704.05446 [hep-ph]. [9] D. Bardhan, P. Byakti and D. Ghosh, “Role of Tensor [4] G. D’Amico, M. Nardecchia, P. Panci, F. San- operators in RK and RK∗ ,” arXiv:1705.09305 [hep-ph]. nino, A. Strumia, R. Torre and A. Urbano, [10] S. Neshatpour, V. G. Chobanova, T. Hurth, F. Mah- “Flavour anomalies after the RK∗ measurement,” moudi and D. Martinez Santos, “Direct comparison of JHEP 1709, 010 (2017) doi:10.1007/JHEP09(2017)010 global fits to the B → K∗µ+µ− data assuming hadronic [arXiv:1704.05438 [hep-ph]]. corrections or new physics,” arXiv:1705.10730 [hep-ph]. [5] W. Altmannshofer, P. Stangl and D. M. Straub, “Inter- [11] A. K. Alok, D. Kumar, J. Kumar and R. Sharma, preting Hints for Lepton Flavor Universality Violation,” “Lepton flavor non-universality in the B-sector: a arXiv:1704.05435 [hep-ph]. global analyses of various new physics models,” [6] B. Capdevila, A. Crivellin, S. Descotes-Genon, J. Ma- arXiv:1704.07347 [hep-ph]. tias and J. Virto, “Patterns of New Physics in [12] A. K. Alok, B. Bhattacharya, A. Datta, D. Kumar, 15

J. Kumar and D. London, “New Physics in b → sµ+µ− [arXiv:1504.07928 [hep-ph]]. after the Measurement of RK∗ ,” arXiv:1704.07397 [hep- [28] A. Celis, J. Fuentes-Martin, M. Jung and H. Sero- ph]. dio, “Family nonuniversal Z0 models with pro- [13] D. Choudhury, A. Kundu, R. Mandal and tected flavor-changing interactions,” Phys. Rev. D 92, R. Sinha, “Minimal unified resolution to no. 1, 015007 (2015) doi:10.1103/PhysRevD.92.015007 (∗) RK(∗) and R(D ) anomalies with lepton mix- [arXiv:1505.03079 [hep-ph]]. ing,” Phys. Rev. Lett. 119, no. 15, 151801 [29] A. Carmona and F. Goertz, “Lepton Flavor (2017) doi:10.1103/PhysRevLett.119.151801 and Nonuniversality from Minimal Composite [arXiv:1706.08437 [hep-ph]]. Higgs Setups,” Phys. Rev. Lett. 116, no. 25, [14] B. Gripaios, M. Nardecchia and S. A. Renner, “Lin- 251801 (2016) doi:10.1103/PhysRevLett.116.251801 ear flavour violation and anomalies in B physics,” [arXiv:1510.07658 [hep-ph]]. JHEP 1606, 083 (2016) doi:10.1007/JHEP06(2016)083 [30] K. Fuyuto, W. S. Hou and M. Kohda, “Z0- [arXiv:1509.05020 [hep-ph]]. induced FCNC decays of top, beauty, and strange [15] M. Bauer and M. Neubert, “Minimal Lepto- quarks,” Phys. Rev. D 93, no. 5, 054021 (2016)

quark Explanation for the RD(∗) ,RK , and doi:10.1103/PhysRevD.93.054021 [arXiv:1512.09026 (g − 2)g Anomalies,” Phys. Rev. Lett. 116, no. [hep-ph]]. 14, 141802 (2016) doi:10.1103/PhysRevLett.116.141802 [31] C. W. Chiang, X. G. He and G. Valencia, “Z0 model [arXiv:1511.01900 [hep-ph]]. for b → s`` flavor anomalies,” Phys. Rev. D 93, [16] P. Arnan, L. Hofer, F. Mescia and A. Criv- no. 7, 074003 (2016) doi:10.1103/PhysRevD.93.074003 ellin, “Loop effects of heavy new scalars and [arXiv:1601.07328 [hep-ph]]. fermions in b → sµ+µ−,” JHEP 1704, 043 (2017) [32] C. S. Kim, X. B. Yuan and Y. J. Zheng, “Con- doi:10.1007/JHEP04(2017)043 [arXiv:1608.07832 [hep- straints on a Z0 boson within minimal flavor vi- ph]]. olation,” Phys. Rev. D 93, no. 9, 095009 (2016) [17] J. F. Kamenik, Y. Soreq and J. Zupan, “Lepton fla- doi:10.1103/PhysRevD.93.095009 [arXiv:1602.08107 vor universality violation without new sources of quark [hep-ph]]. flavor violation,” arXiv:1704.06005 [hep-ph]. [33] S. M. Boucenna, A. Celis, J. Fuentes-Martin, [18] D. Das, C. Hati, G. Kumar and N. Mahajan, “Scruti- A. Vicente and J. Virto, “Non-abelian gauge ex-

nizing R-parity violating interactions in light of RK(∗) tensions for B-decay anomalies,” Phys. Lett. B data,” arXiv:1705.09188 [hep-ph]. 760, 214 (2016) doi:10.1016/j.physletb.2016.06.067 [19] J. Kawamura, S. Okawa and Y. Omura, “Impact [arXiv:1604.03088 [hep-ph]]. of the b → sll anomalies on dark matter physics,” [34] K. Cheung, W. Y. Keung and P. Y. Tseng, arXiv:1706.04344 [hep-ph]. “Lepton-Flavor-violating Z0 using the electron-muon [20] X. G. He and G. Valencia, “Are the B-anomalies evi- channel at the LHC,” Phys. Rev. D 94, no. dence for heavy neutrinos?,” arXiv:1706.07570 [hep-ph]. 7, 075006 (2016) doi:10.1103/PhysRevD.94.075006 [21] R. Gauld, F. Goertz and U. Haisch, “On minimal Z0 [arXiv:1606.06696 [hep-ph]]. explanations of the B → K∗µ+µ− anomaly,” Phys. Rev. [35] S. M. Boucenna, A. Celis, J. Fuentes-Martin, A. Vicente D 89, 015005 (2014) doi:10.1103/PhysRevD.89.015005 and J. Virto, “Phenomenology of an SU(2) × SU(2) × [arXiv:1308.1959 [hep-ph]]. U(1) model with lepton-flavour non-universality,” [22] R. Gauld, F. Goertz and U. Haisch, “An explicit Z’- JHEP 1612, 059 (2016) doi:10.1007/JHEP12(2016)059 boson explanation of the B → K∗µ+µ− anomaly,” [arXiv:1608.01349 [hep-ph]]. JHEP 1401, 069 (2014) doi:10.1007/JHEP01(2014)069 [36] A. Crivellin, J. Fuentes-Martin, A. Greljo and [arXiv:1310.1082 [hep-ph]]. G. Isidori, “Lepton Flavor Non-Universality in [23] A. J. Buras, F. De Fazio and J. Girrbach, “331 mod- B decays from Dynamical Yukawas,” Phys. Lett. els facing new b → sµ+µ− data,” JHEP 1402, 112 B 766, 77 (2017) doi:10.1016/j.physletb.2016.12.057 (2014) doi:10.1007/JHEP02(2014)112 [arXiv:1311.6729 [arXiv:1611.02703 [hep-ph]]. [hep-ph]]. [37] I. Garcia Garcia, “LHCb anomalies from a [24] W. Altmannshofer, S. Gori, M. Pospelov and I. Yavin, natural perspective,” JHEP 1703, 040 (2017) “Quark flavor transitions in Lµ−Lτ models,” Phys. Rev. doi:10.1007/JHEP03(2017)040 [arXiv:1611.03507 D 89, 095033 (2014) doi:10.1103/PhysRevD.89.095033 [hep-ph]]. [arXiv:1403.1269 [hep-ph]]. [38] A. Datta, J. Liao and D. Marfatia, “A light Z0 [25] A. Crivellin, G. D’Ambrosio and J. Heeck, “Ad- for the RK puzzle and nonstandard neutrino dressing the LHC flavor anomalies with horizon- interactions,” Phys. Lett. B 768, 265 (2017) tal gauge symmetries,” Phys. Rev. D 91, no. doi:10.1016/j.physletb.2017.02.058 [arXiv:1702.01099 7, 075006 (2015) doi:10.1103/PhysRevD.91.075006 [hep-ph]]. [arXiv:1503.03477 [hep-ph]]. [39] P. Ko, Y. Omura, Y. Shigekami and C. Yu, “The LHCb [26] D. Aristizabal Sierra, F. Staub and A. Vicente, anomaly and B physics in flavored Z0 models with fla- “Shedding light on the b → s anomalies with a vored Higgs doublets,” arXiv:1702.08666 [hep-ph]. dark sector,” Phys. Rev. D 92, no. 1, 015001 (2015) [40] R. Alonso, P. Cox, C. Han and T. T. Yanagida, doi:10.1103/PhysRevD.92.015001 [arXiv:1503.06077 “Anomaly-free local horizontal symmetry and anomaly- [hep-ph]]. full rare B-decays,” arXiv:1704.08158 [hep-ph]. [27] A. Crivellin, L. Hofer, J. Matias, U. Nierste, S. Poko- [41] C. Bonilla, T. Modak, R. Srivastava and J. W. F. Valle,

rski and J. Rosiek, “Lepton-ﬂavour violating B de- “U(1)B3−3Lµ gauge symmetry as the simplest descrip- cays in generic Z0 models,” Phys. Rev. D 92, tion of b → s anomalies,” arXiv:1705.00915 [hep-ph]. no. 5, 054013 (2015) doi:10.1103/PhysRevD.92.054013 [42] R. Alonso, P. Cox, C. Han and T. T. Yanagida, 16

“Flavoured B−L Local Symmetry and Anomalous Rare [60] S. Sahoo and R. Mohanta, “Lepton ﬂavor violating B B Decays,” arXiv:1705.03858 [hep-ph]. meson decays via a scalar leptoquark,” Phys. Rev. D 93, [43] J. Ellis, M. Fairbairn and P. Tunney, “Anomaly-Free no. 11, 114001 (2016) doi:10.1103/PhysRevD.93.114001 Models for Flavour Anomalies,” arXiv:1705.03447 [hep- [arXiv:1512.04657 [hep-ph]]. ph]. [61] G. Kumar, “Constraints on a scalar leptoquark [44] D. Ghosh, “Explaining the RK and RK∗ anomalies,” from the kaon sector,” Phys. Rev. D 94, no. arXiv:1704.06240 [hep-ph]. 1, 014022 (2016) doi:10.1103/PhysRevD.94.014022 [45] Y. Tang and Y. L. Wu, “Flavor Non-universality [arXiv:1603.00346 [hep-ph]]. Gauge Interactions and Anomalies in B-Meson Decays,” [62] D. Das, C. Hati, G. Kumar and N. Mahajan, “Towards

arXiv:1705.05643 [hep-ph]. a unified explanation of RD(∗) , RK and (g −2)µ anoma- [46] C. W. Chiang, X. G. He, J. Tandean and X. B. Yuan, lies in a left-right model with leptoquarks,” Phys. Rev. ¯ “RK(∗) and related b → s`` anomalies in minimal flavor D 94, 055034 (2016) doi:10.1103/PhysRevD.94.055034 violation framework with Z0 boson,” arXiv:1706.02696 [arXiv:1605.06313 [hep-ph]]. [hep-ph]. [63] X. Q. Li, Y. D. Yang and X. Zhang, “Revis- [47] R. S. Chivukula, J. Isaacson, K. A. Mohan, D. Sen- iting the one leptoquark solution to the R(D(∗)) gupta and E. H. Simmons, “RK anomalies and simpli- anomalies and its phenomenological implications,” fied limits on Z0 models at the LHC,” arXiv:1706.06575 JHEP 1608, 054 (2016) doi:10.1007/JHEP08(2016)054 [hep-ph]. [arXiv:1605.09308 [hep-ph]]. 0 [48] S. F. King, “Flavourful Z models for RK(∗) ,” [64] C. H. Chen, T. Nomura and H. Okada, “Expla- arXiv:1706.06100 [hep-ph]. nation of B → K(∗)`+`− and muon g − 2, and [49] J. M. Cline and J. Martin Camalich, “B de- implications at the LHC,” Phys. Rev. D 94, no. cay anomalies from nonabelian local horizon- 11, 115005 (2016) doi:10.1103/PhysRevD.94.115005 tal symmetry,” Phys. Rev. D 96, 055036 (2017) [arXiv:1607.04857 [hep-ph]]. doi:10.1103/PhysRevD.96.055036 [arXiv:1706.08510 [65] D. Beˇcirević, S. Fajfer, N. Koˇsnik and O. Sumen- [hep-ph]]. sari, “Leptoquark model to explain the B-physics 0+ 0− [50] C. H. Chen and T. Nomura, “Penguin b → s` ` anomalies, RK and RD,” Phys. Rev. D 94, no. and B-meson anomalies in a gauged Lµ − Lτ ,” 11, 115021 (2016) doi:10.1103/PhysRevD.94.115021 arXiv:1707.03249 [hep-ph]. [arXiv:1608.08501 [hep-ph]]. + − [51] S. Baek, “Dark matter contribution to b → sµ µ [66] D. Beˇcirević, N. Koˇsnik, O. Sumensari and anomaly in local U(1)Lµ−Lτ model,” arXiv:1707.04573 R. Zukanovich Funchal, “Palatable Leptoquark [hep-ph]. Scenarios for Lepton Flavor Violation in Exclu- [52] L. Bian, S. M. Choi, Y. J. Kang and H. M. Lee, sive b → s`1`2 modes,” JHEP 1611, 035 (2016) 0 “A minimal flavored U(1) for B-meson anomalies,” doi:10.1007/JHEP11(2016)035 [arXiv:1608.07583 arXiv:1707.04811 [hep-ph]. [hep-ph]]. [53] M. Dalchenko, B. Dutta, R. Eusebi, P. Huang, T. Ka- [67] N. Mileo, A. de la Puente and A. Szynkman, “Impli- mon and D. Rathjens, “Bottom-quark Fusion Processes cations of a Electroweak Triplet Scalar Leptoquark on 0 at the LHC for Probing Z Models and B-meson Decay the Ultra-High Energy Neutrino Events at IceCube,” Anomalies,” arXiv:1707.07016 [hep-ph]. JHEP 1611, 124 (2016) doi:10.1007/JHEP11(2016)124 [54] E. Megias, M. Quiros and L. Salas, “Flavor anomalies [arXiv:1608.02529 [hep-ph]]. from warped space,” arXiv:1709.05100 [hep-ph]. [68] G. Hiller, D. Loose and K. Schönwald, “Lep- [55] I. Dorˇsner, S. Fajfer, A. Greljo, J. F. Kamenik toquark Flavor Patterns & B Decay Anomalies,” and N. Koˇsnik, “Physics of leptoquarks in preci- JHEP 1612, 027 (2016) doi:10.1007/JHEP12(2016)027 sion experiments and at particle colliders,” Phys. [arXiv:1609.08895 [hep-ph]]. Rept. 641, 1 (2016) doi:10.1016/j.physrep.2016.06.001 [69] S. Sahoo, R. Mohanta and A. K. Giri, “Explaining [arXiv:1603.04993 [hep-ph]]. the RK and RD(∗) anomalies with vector lepto- [56] G. Hiller and M. Schmaltz, “RK and future quarks,” Phys. Rev. D 95, no. 3, 035027 (2017) b → s`` physics beyond the standard model op- doi:10.1103/PhysRevD.95.035027 [arXiv:1609.04367 portunities,” Phys. Rev. D 90, 054014 (2014) [hep-ph]]. doi:10.1103/PhysRevD.90.054014 [arXiv:1408.1627 [70] O. Popov and G. A. White, “One Leptoquark to unify [hep-ph]]. them? Neutrino masses and unification in the light of [57] B. Gripaios, M. Nardecchia and S. A. Renner, “Com- (g − 2)µ, RD(?) and RK anomalies,” arXiv:1611.04566 posite leptoquarks and anomalies in B-meson decays,” [hep-ph]. JHEP 1505, 006 (2015) doi:10.1007/JHEP05(2015)006 [71] R. Barbieri, C. W. Murphy and F. Senia, “B-decay [arXiv:1412.1791 [hep-ph]]. Anomalies in a Composite Leptoquark Model,” Eur. [58] B. Allanach, A. Alves, F. S. Queiroz, K. Sinha and Phys. J. C 77, no. 1, 8 (2017) doi:10.1140/epjc/s10052- + − A. Strumia, “Interpreting the CMS ` ` jjE/T Ex- 016-4578-7 [arXiv:1611.04930 [hep-ph]]. cess with a Leptoquark Model,” Phys. Rev. D 92, [72] J. M. Cline, J. M. Cornell, D. London and no. 5, 055023 (2015) doi:10.1103/PhysRevD.92.055023 R. Watanabe, “Hidden sector explanation of B- [arXiv:1501.03494 [hep-ph]]. decay and cosmic ray anomalies,” Phys. Rev. D 95, [59] R. Alonso, B. Grinstein and J. Martin Camalich, “Lep- no. 9, 095015 (2017) doi:10.1103/PhysRevD.95.095015 ton universality violation and lepton flavor conserva- [arXiv:1702.00395 [hep-ph]]. tion in B-meson decays,” JHEP 1510, 184 (2015) [73] C. H. Chen, T. Nomura and H. Okada, “Excesses of doi:10.1007/JHEP10(2015)184 [arXiv:1505.05164 [hep- muon g − 2, RD(∗) , and RK in a leptoquark model,” ph]]. 17

arXiv:1703.03251 [hep-ph]. D0 Mixing,” Phys. Rev. D 80, 055024 (2009) [74] A. Crivellin, D. Müllerand T. Ota, “Simultaneous Ex- doi:10.1103/PhysRevD.80.055024 [arXiv:0906.1879 planation of R(D(∗)) and b → sµ+µ−: The Last Scalar [hep-ph]]. Leptoquarks Standing,” arXiv:1703.09226 [hep-ph]. [93] UTfit collaboration, http://www.utfit.org/UTfit/ [75] G. Hiller and I. Nisandzic, “RK and RK∗ beyond the ResultsSummer2016NP Standard Model,” arXiv:1704.05444 [hep-ph]. [94] G. D’Ambrosio, G. F. Giudice, G. Isidori and A. Stru- [76] Y. Cai, J. Gargalionis, M. A. Schmidt and R. R. Volkas, mia, “Minimal flavor violation: An Effective field “Reconsidering the One Leptoquark solution: flavor theory approach,” Nucl. Phys. B 645, 155 (2002) anomalies and neutrino mass,” arXiv:1704.05849 [hep- doi:10.1016/S0550-3213(02)00836-2 [hep-ph/0207036]. ph]. [95] J. M. Cline, W. Huang and G. D. Moore, “Challenges [77] D. Beˇcirevićand O. Sumensari, “A leptoquark model to for models with composite states,” Phys. Rev. D 94, exp no. 5, 055029 (2016) doi:10.1103/PhysRevD.94.055029 accommodate RK ,” arXiv:1704.05835 [hep-ph]. [78] S. Matsuzaki, K. Nishiwaki and R. Watanabe, “Phe- [arXiv:1607.07865 [hep-ph]]. nomenology of flavorful composite vector bosons in [96] B. Guberina, J. H. Kuhn, R. D. Peccei and R. Ruckl, light of B anomalies,” JHEP 1708, 145 (2017) “Rare Decays of the Z0,” Nucl. Phys. B 174, 317 (1980). doi:10.1007/JHEP08(2017)145 [arXiv:1706.01463 [hep- doi:10.1016/0550-3213(80)90287-4 ph]]. [97] P. G. Esposito, “Constraints on Composite Models from [79] B. Chauhan, B. Kindra and A. Narang, “A Leptoquark Leptonic Flavor Changing Decays,” Nuovo Cim. A 92, explanation for (g − 2)µ, RK , RK? and, IceCube PeV 328 (1986). doi:10.1007/BF02724248 events,” arXiv:1706.04598 [hep-ph]. [98] S. Descotes-Genon, L. Hofer, J. Matias and J. Virto, “Global analysis of b → s`` anomalies,” JHEP [80] W. Altmannshofer, P. S. B. Dev and A. Soni, “RD(∗) anomaly: A possible hint for natural supersymmetry 1606, 092 (2016) doi:10.1007/JHEP06(2016)092 with R-parity violation,” arXiv:1704.06659 [hep-ph]. [arXiv:1510.04239 [hep-ph]]. [81] D. Buttazzo, A. Greljo, G. Isidori and D. Marzocca, “B- [99] J. M. Cline, “Scalar doublet models confront physics anomalies: a guide to combined explanations,” τ and b anomalies,” Phys. Rev. D 93, no. arXiv:1706.07808 [hep-ph]. 7, 075017 (2016) doi:10.1103/PhysRevD.93.075017 [82] A. Crivellin, D. Mueller, A. Signer and Y. Ulrich, [arXiv:1512.02210 [hep-ph]]. “Correlating Lepton Flavour (Universality) Violation [100] C. Giunti and A. Studenikin, “Neutrino electromagnetic in B Decays with µ → eγ using Leptoquarks,” interactions: a window to new physics,” Rev. Mod. arXiv:1706.08511 [hep-ph]. Phys. 87, 531 (2015) doi:10.1103/RevModPhys.87.531 [83] S. Y. Guo, Z. L. Han, B. Li, Y. Liao and X. D. Ma, [arXiv:1403.6344 [hep-ph]]. [101] See the review of the muon anomalous magnetic mo- “Interpreting the RK(∗) Anomaly in the Colored Zee- Babu Model,” arXiv:1707.00522 [hep-ph]. ment in C. Patrignani et al. (Particle Data Group), [84] D. Aloni, A. Dery, C. Frugiuele and Y. Nir, “Testing Chin. Phys. C, 40, 100001 (2016). Minimal Flavor Violation in Leptoquark Models of the [102] A. Mitridate, M. Redi, J. Smirnov and A. Strumia, “Dark Matter as a weakly coupled Dark Baryon,” RK(∗) Anomaly,” arXiv:1708.06161 [hep-ph]. [85] N. Assad, B. Fornal and B. Grinstein, “Baryon Number arXiv:1707.05380 [hep-ph]. and Lepton Universality Violation in Leptoquark and [103] T. Banks, J. F. Fortin and S. Thomas, “Direct Detec- Diquark Models,” arXiv:1708.06350 [hep-ph]. tion of Dark Matter Electromagnetic Dipole Moments,” [86] L. Di Luzio, A. Greljo and M. Nardecchia, “Gauge arXiv:1007.5515 [hep-ph]. leptoquark as the origin of B-physics anomalies,” [104] E. Aprile et al. [XENON100 Collaboration], “First arXiv:1708.08450 [hep-ph]. Dark Matter Results from the XENON100 Ex- [87] L. Calibbi, A. Crivellin and T. Li, “A model of vec- periment,” Phys. Rev. Lett. 105, 131302 (2010) tor leptoquarks in view of the B-physics anomalies,” doi:10.1103/PhysRevLett.105.131302 [arXiv:1005.0380 arXiv:1709.00692 [hep-ph]. [astro-ph.CO]]. [88] C. Kilic, T. Okui and R. Sundrum, “Vectorlike Con- [105] Cui, X., et al. [PandaX-II Collaboration], “Dark Mat- finement at the LHC,” JHEP 1002, 018 (2010) ter Results From 54-Ton-Day Exposure of PandaX-II doi:10.1007/JHEP02(2010)018 [arXiv:0906.0577 [hep- Experiment,” arXiv:1708.06917 [astro-ph.CO]. ph]]. [106] Y. Chen et al., “Glueball spectrum and matrix elements [89] J. Kang and M. A. Luty, “Macroscopic Strings on anisotropic lattices,” Phys. Rev. D 73, 014516 (2006) and ’Quirks’ at Colliders,” JHEP 0911, 065 (2009) doi:10.1103/PhysRevD.73.014516 [hep-lat/0510074]. doi:10.1088/1126-6708/2009/11/065 [arXiv:0805.4642 [107] C. Kilic, T. Okui and R. Sundrum, “Colored Reso- [hep-ph]]. nances at the Tevatron: Phenomenology and Discovery [90] C. C. Nishi, “Simple derivation of general Fierz- Potential in Multijets,” JHEP 0807, 038 (2008) like identities,” Am. J. Phys. 73, 1160 (2005) doi:10.1088/1126-6708/2008/07/038 [arXiv:0802.2568 doi:10.1119/1.2074087 [hep-ph/0412245]. [hep-ph]]. [91] M. Bona et al. [UTfit Collaboration], “Model- [108] M. E. Peskin and D. V. Schroeder, “An Introduction to independent constraints on ∆F = 2 operators and quantum field theory,” the scale of new physics,” JHEP 0803, 049 (2008) [109] J. M. Cline, G. Dupuis and Z. Liu, “LHC constraints doi:10.1088/1126-6708/2008/03/049 [arXiv:0707.0636 on dark matter with (130 GeV) gamma ray lines,” [hep-ph]]. JHEP 1309, 065 (2013) doi:10.1007/JHEP09(2013)065 [92] O. Gedalia, Y. Grossman, Y. Nir and G. Perez, [arXiv:1306.3217 [hep-ph]]. “Lessons from Recent Measurements of D0 - anti- [110] W. Kwong, P. B. Mackenzie, R. Rosenfeld and J. L. Ros- ner, “Quarkonium Annihilation Rates,” Phys. Rev. D 18

37, 3210 (1988). doi:10.1103/PhysRevD.37.3210 right-handed neutrinos in final states with two tau [111] M. Aaboud et al. [ATLAS Collaboration], “Search for leptons and two jets in proton-proton collisions −1 √ new phenomena in dijet√ events using 37 fb of pp col- at s = 13 TeV,” JHEP 1707, 121 (2017) lision data collected at s =13 TeV with the ATLAS doi:10.1007/JHEP07(2017)121 [arXiv:1703.03995 [hep- detector,” arXiv:1703.09127 [hep-ex]. ex]]. [112] CMS Collaboration [CMS Collaboration],√ “Searches for [119] CMS Collaboration [CMS Collaboration],√ “Search for dijet resonances in pp collisions at s = 13 TeV using excited leptons in the ``γ final state at s = 13 TeV,” data collected in 2016.,” CMS-PAS-EXO-16-056. CMS-PAS-EXO-16-009. [113] N. Craig, P. Draper, C. Kilic and S. Thomas, “Shed- [120] J. Albrecht, F. Bernlochner, M. Kenzie, S. Reichert, ding Light on Diphoton Resonances,” Phys. Rev. D 93, D. Staub and A. Tully, “Future prospects for exploring no. 11, 115023 (2016) doi:10.1103/PhysRevD.93.115023 present day anomalies in flavour physics measurements [arXiv:1512.07733 [hep-ph]]. with Belle II and LHCb,” arXiv:1709.10308 [hep-ph]. [114] M. Aaboud et al. [ATLAS Collaboration], “Search for [121] D. Spier Moreira Alves, S. R. Behbahani, P. Schus- new phenomena in high-mass diphoton final states√ using ter and J. G. Wacker, “The Cosmology of Compos- 37 fb−1 of proton–proton collisions collected at s = ite Inelastic Dark Matter,” JHEP 1006, 113 (2010) 13 TeV with the ATLAS detector,” [arXiv:1707.04147 doi:10.1007/JHEP06(2010)113 [arXiv:1003.4729 [hep- [hep-ex]]. ph]]. [115] M. Aaboud et al. [ATLAS Collaboration], “Search for [122] S. Raby, S. Dimopoulos and L. Susskind, “Tumbling new high-mass phenomena in the dilepton final state√ Gauge Theories,” Nucl. Phys. B 169, 373 (1980). using 36.1 fb−1 of proton-proton collision data at s doi:10.1016/0550-3213(80)90093-0 = 13 TeV with the ATLAS detector,” arXiv:1707.02424 [123] M. J. Teper, “Glueball masses and other physical prop- [hep-ex]. erties of SU(N) gauge theories in D = (3+1): A Review [116] M. Aaboud et al. [ATLAS Collaboration],√ “Search for of lattice results for theorists,” hep-th/9812187. scalar leptoquarks in pp collisions at s = 13 TeV [124] See the review of QCD in C. Patrignani et al. (Particle with the ATLAS experiment,” New J. Phys. 18, no. Data Group), Chin. Phys. C, 40, 100001 (2016). 9, 093016 (2016) doi:10.1088/1367-2630/18/9/093016 [125] M. Czakon, “The Four-loop QCD beta-function and [arXiv:1605.06035 [hep-ex]]. anomalous dimensions,” Nucl. Phys. B 710, 485 (2005) [117] CMS Collaboration [CMS Collaboration], “Search for doi:10.1016/j.nuclphysb.2005.01.012 [hep-ph/0411261]. pair-production of√ second-generation scalar leptoquarks [126] A. K. Rai, R. H. Parmar and P. C. Vinodkumar, in pp collisions at s = 13 TeV with the CMS detector,” “Masses and decay constants of heavy-light flavor CMS-PAS-EXO-16-007. mesons in a variational scheme,” J. Phys. G 28, 2275 [118] A. M. Sirunyan et al. [CMS Collaboration], “Search (2002). doi:10.1088/0954-3899/28/8/313 for third-generation scalar leptoquarks and heavy