MSUHEP-060915 hep-ph/0609179 Phenomenology of Littlest Higgs Model with T-parity: including effects of T-odd fermions Alexander Belyaev, Chuan-Ren Chen, Kazuhiro Tobe, C.-P. Yuan Department of Physics and Astronomy, Michigan State University, East Lansing, MI 48824, USA (Dated: February 2, 2008) 7 0 We study the collider phenomenology of a Littlest Higgs model with T-parity. 0 2 We first stress the important role of the T-odd SU(2) doublet fermions (introduced n a to make the model T-parity invariant) in high energy scattering processes, such J 7 as qq¯ W+W where W are the T-odd partners of W-bosons. Because the 1 → H H− H± mass of the T-odd SU(2) doublet fermions cannot be too heavy to be consistent 2 v 9 with low energy data, they can be copiously produced at the CERN Large Hadron 7 1 Collider (LHC). Therefore, we study the collider phenomenology of the model with 9 0 emphasis on the contributions of the T-odd fermion to the production of the heavy 6 0 T-parity partners (either bosons or fermions) of the usual particles at the LHC. The / h p production cross sections and the decay branching ratios of the new heavy particles - p are classified and various experimental signatures are discussed. e h : v i X I. INTRODUCTION r a The standard model (SM) is an excellent low energy description of the elementary parti- cles. The absence of any significant deviations from the SM predictions on the electroweak precision measurements suggests that the cutoff scale of the SM, as a low energy effective theory, is as large as, or larger than, 10 TeV [1]. However, having such a relatively high cutoff scale in the SM, the Higgs boson receives a large radiative correction to its mass pa- rameter and therefore, the SM requires unsatisfactory fine-tuning to yield a correct scale of the electroweak symmetry breaking. This fine-tuning problem of the Higgs mass parameter (known as the “Little hierarchy problem”) has been one of the driving forces to consider physics beyond the SM. Moreover, a recent finding of the necessity of dark matter candidate also provides a strong motivation to seek for physics beyond the SM. 2 It has been shown recently that the collective symmetry breaking mechanism in the Little Higgs models [2] can provide an interesting solution to the Little hierarchy problem and the Littlest Higgs (LH) [3] model is the most economical Little Higgs model discussed in the literature. However, the original version of the LH model suffers from precision electroweak constrains [4] and the value of f, which characterizes the mass scale of new particles in the model, is forced to be larger than about 4 TeV. Since the cutoff scale of the model is about 4πf, the fine tuning between the cutoff scale and the weak scale will be needed again for a too large value of f. The Littlest Higgs model with T-parity (LHT) [5, 6, 7, 8] is one of the attractive Little Higgs models. It provides a possible dark matter candidate [9] and furthermore, all dangerous tree-level contributions to low energy electroweak (EW) observables are forbidden by T-parity and hence the corrections to low energy EW observables are loop-suppressed and small. As a result, the relatively low new particle mass scale f is still allowed by data, e.g., f > 500 GeV [7]. TheLHTpredictsheavyT-oddgaugebosonswhichareT-paritypartnersoftheSMgauge bosons. Moreover, in order to implement T-parity in the fermion sector, one introduces the heavy T-odd SU(2)-doublet fermions, which are T-parity partners of the SM SU(2)-doublet fermions and unique to Little Higgs models with T-parity. Therefore, having the relatively low new particle mass scale f, the CERN Large Hadron Collider (LHC) will have a great potential to directly produce the T-parity partners of the SM particles, and hence it is important to probe the LHT at the LHC. In previous works on studying the phenomenology of the LHT [6], the effects of T-odd SU(2) doublet fermions were not included. A preliminary study on the phenomenology of these T-odd SU(2)-doublet fermions in the LHT was reported in Ref. [10]. Although, moti- vated by the dark matter consideration, Ref. [11] studied some interesting processes which includetheeffectsofT-oddSU(2)doubletfermions, acompletestudyonthephenomenology of these T-odd fermions in the LHT has not yet been presented. In this paper, we first stress the important role of the T-odd fermions in high energy scattering processes relevant to the LHC, such as qq¯ W+W , where W is the T-parity → H H− H± partner of W-boson. We show that it is necessary to include the contribution from the t-channel process, via the exchange of these T-odd heavy fermions, to render its scattering amplitude with a good high energy behavior, so that its partial-wave amplitudes respect the unitaritycondition. Wealsoshowthatitsnumericaleffectcannotbeignoredforstudyingthe 3 collider phenomenologyattheLHC.Furthermore, sincethecurrent experimental constraints of the four-fermion contact interactions place an upper bound on the T-odd SU(2) doublet fermion masses [7], we find not only that the T-odd fermion contribution to qq¯ W+W is → H H− quantitatively important, but also that the direct pair production rateof the T-oddfermions could be significant at the LHC. To illustrate this point, we classify all production processes ofthenewheavy particlespredictedbytheLHT,andcalculatethecorresponding production cross sections and decay branching ratios, including the effects induced by the T-odd SU(2) doublet fermions. The rest of this paper is organized as follows. In Sec. II, we briefly review the model we study here, a Littlest Higgs model with T-parity. In Sec. III, we discuss the high energy behavior of uu¯ W+W process to illustrate the importance of the T-odd → H H− SU(2) doublet fermion contribution to high energy scattering processes, in order to restore the unitarity of partial wave amplitudes. In Sec. IV, we show our numerical results of the phenomenological study on the Littlest Higgs model with T-parity at the LHC energy. Our conclusion is given in Sec. V. II. A LITTLEST HIGGS MODEL WITH T-PARITY Inthissection, webriefly review theLittlest HiggsModelwith T-paritystudied in[5, 6, 7] andpresentournotationofthemodel. TheLittlestHiggsmodelisbasedonanSU(5)/SO(5) non-linear sigma model [3]. A vacuum expectation value (VEV) of an SU(5) symmetric tensor field (Σ ) breaks the SU(5) to SO(5) at the scale f with 0 0 0 0 1 0 0 0 0 0 1  Σ = 0 0 1 0 0 . (1) 0     1 0 0 0 0      0 1 0 0 0      A subgroup [SU(2) U(1) ] [SU(2) U(1) ] of the SU(5) is gauged, and at the scale f 1 1 2 2 × × × it is broken into the SM electroweak symmetry SU(2) U(1) . The 14 Nambu-Goldstone L Y × bosons Πa associated with the global symmetry breaking decompose under SU(2) U(1) L Y × as 10 30 21/2 31 and they are parametrized by the non-linear sigma model field ⊕ ⊕ ⊕ Σ = ξ2Σ as the fluctuations around the VEV in the broken directions, where ξ = eiΠaXa/f 0 and Xa are the generators of the broken symmetry. The components 10 and 30 in the 4 Nambu-Goldstone boson multiplet are eaten by the heavy gauge bosons associated with the gauge symmetry breaking. The SU(2) doublet 2 is considered to be the Higgs doublet. 1/2 The doublet 21/2 and the triplet 31 Higgs bosons remain in the low energy effective theory, which are introduced through: 0 H Φ 2×2 √2 iπ+ iφ++ iφ+ ΠaXa =  H† 0 HT  with H = − and Φ = − − √2 ,(2) √2 √2  h+iπ0   iφ+ iφ0+iφP   Φ H∗ 0  √2 − √2 − √2  † √2 2×2        where 0 is a two by two matrix with zero components and the superscript T denotes 2 2 × taking transpose. Here we only show the doublet Higgs H(21/2), where π+ and π0 are eaten by the SM W- and Z-bosons, respectively, and the triplet Higgs Φ(31), which forms a symmetric tensor with components φ , φ , φ0 and φP [12]. Since the non-linear sigma ±± ± model field Σ transforms as Σ VΣVT under the SU(5) rotation V, its gauge-invariant → kinetic term is given by f2 = Tr(D Σ) (DµΣ), (3) µ † L 8 wherethecovariantderivativeD forthe[SU(2) U(1) ] [SU(2) U(1) ]gaugesymmetry µ 1 1 2 2 × × × is defined as D Σ = ∂ Σ i g¯ (Wa QaΣ+ΣQaWa )+g¯ (B Y Σ+ΣY B ) . (4) µ µ − A Aµ A A Aµ A′ Aµ A A Aµ A=1,2 X (cid:2) (cid:3) Here Wa and B (A = 1,2) are gauge bosons, and g¯ and g¯ are gauge couplings for Aµ Aµ A A′ SU(2) and U(1) gauge symmetries, respectively. They are related to the SM gauge A A couplings g for SU(2) and g for U(1) as 1/g2 = 1/g¯2 + 1/g¯2 and 1/g2 = 1/g¯2 + 1/g¯2. L ′ Y 1 2 ′ 1′ 2′ The generators for SU(2) (denoted as Qa) and for U(1) (denoted as Y ) are explicitly A A A A expressed as σa/2 0 0 0 0 0 2 2 2 2 2 2 2 2 × × × Qa =  0T 0 0T , Qa =  0T 0 0T , (5) 1 2 2 2 2 2  0 0 0  0 0 σa /2   2 2 2 2 2   2 2 2 ∗   × ×   × −  Y = diag.(3,3, 2, 2, 2)/10, Y= diag.(2,2,2, 3, 3)/10, (6) 1 2 − − − − − where σa is the Pauli matrix, 0 = (0,0)T and “diag.” denotes a diagonal matrix. 2 5 A. Gauge boson sector T-parity [5, 8] is naturally introduced in this framework. It exchanges [SU(2) U(1) ] 1 1 × and [SU(2) U(1) ] symmetries. For example, Wa Wa and B B under T-parity. 2× 2 1µ ↔ 2µ 1µ ↔ 2µ The Lagrangian Eq. (3) is invariant under T-parity if g¯ = g¯ and g¯ = g¯ and Σ transforms 1 2 1′ 2′ ˜ as Σ Σ = Σ ΩΣ ΩΣ with Ω = diag.(1,1, 1,1,1). Note that the doublet Higgs H 0 † 0 → − (triplet Higgs Φ) is even (odd) under T-parity. The T-even combinations of the gauge fields are SM SU(2) gauge bosons (Wa) and U(1) hypercharge gauge boson (B ), defined as L µ Y µ Wa = W1aµ+W2aµ and B = B1µ+B2µ. The T-odd combinations are T-parity partners of the µ √2 µ √2 SM gauge bosons. After taking into account electroweak symmetry breaking, the masses of the T-parity partners of the photon (A ), Z-boson (Z ) and W-boson (W ) are given by H H H g f 5v2 v2 M = ′ 1 SM + , M M = gf 1 SM + . (7) AH √5 − 8f2 ··· ZH ≃ WH − 8f2 ··· (cid:20) (cid:21) (cid:20) (cid:21) Here v is the electroweak breaking scale, v 246 GeV, so that at tree level the SM SM ≃ SM gauge boson masses can be expressed as M = gv and M = √g2+g′2v for W- W 2 SM Z 2 SM boson and Z-boson, respectively. Because of the smallness of g , the T-parity partner of ′ the photon A tends to be the lightest T-odd particle in this framework. Since the lightest H T-odd particle is stable, it can be an interesting dark matter candidate [9]. Because of the T-parity, SM gauge bosons do not mix with the T-odd heavy gauge bosons even after the electroweak symmetry breaking. Consequently, the low energy EW observables are not modified at tree level. Since the new heavy T-odd particles always contribute to loops in pairs, the loop corrections to the EW observables are typically small. As a result, the new particle mass scale f can be as low as 500 GeV [7], and hence, T-odd heavy gauge bosons can be copiously produced at the LHC. B. T-odd SU(2) doublet fermion sector To implement T-parity in the fermion sector, one introduces two SU(2) fermion doublets q (i = 1,2) for each SM fermion doublet [5, 6, 7]. Here q are the doublet under SU(2) i i i (i = 1,2), and T-parity exchanges q and q . The T-even combination of q is the SM 1 2 i fermion doublet and the other T-odd combination is its T-parity partner. To generate a heavy mass for the T-odd fermion doublet, we introduce the following interaction, as 6 suggested in Ref. [5, 6, 7]: ¯ ¯ = κf(Ψ ξΨ +Ψ Σ Ωξ ΩΨ )+hermitian conjugate (h.c.). (8) κ 2 c 1 0 † c L − Here the fermion SU(2) doublets q and q are embedded into incomplete SU(5) multiplets 1 2 Ψ andΨ asΨ = (q ,0,0 )T andΨ = (0 ,0,q )T, andthedoublets q andq areexplicitly 1 2 1 1 2 2 2 2 1 2 written as q = σ (u ,d )T = (id , iu )T with A = 1,2. Under the global SU(5), A − 2 LA LA LA − LA the multiplets Ψ and Ψ transform as Ψ V Ψ and Ψ VΨ , where V is an SU(5) 1 2 1 ∗ 1 2 2 → → rotation matrix. A multiplet Ψ is also introduced as Ψ = (q ,χ ,q˜)T, which transforms c c c c c non-linearly under SU(5): Ψ UΨ where U is an unbroken SO(5) rotation matrix in c c → non-linear representation of SU(5). The object ξ and the non-linear sigma model field Σ ( ξ2Σ ) transform like ξ VξU = UξΣ VTΣ and Σ VΣVT, respectively, under 0 † 0 0 ≡ → → SU(5). T-parity transformation laws are defined as follows: Ψ Σ Ψ , Ψ Ψ , and 1 0 2 c c ↔ − → − ξ Ωξ Ω. Thus, q q and Σ Σ˜ Σ ΩΣ ΩΣ under T-parity. One can verify that † 1 2 0 † 0 → ↔ − → ≡ the interaction in Eq. (8) is invariant under T-parity. From the interaction in Eq. (8), one can see that the T-odd fermion doublet q (q + 1 − ≡ q )/√2 = (id , iu )T gets a Dirac mass, with q˜ (id , iu )T, as 2 L− − L− c ≡ R− − R− v2 M √2κf, M √2κf 1 SM + . (9) d− ≃ u− ≃ − 8f2 ··· (cid:18) (cid:19) One may think that assuming a large κ value, these T-odd fermions will decouple and hence we may ignore any effects induced by the T-odd SU(2) doublet fermions. However, as pointed out in Ref. [7], there is non-decoupling effect in some four-fermion operators whose coefficients become larger as the magnitude of κ increases. The constraint on the four-fermion contact interaction contributing to the e+e qq¯scattering sets an important − → upper bound on the T-odd fermion masses M as [7] q− 2 f M < 4.8 TeV. (10) q− 1 TeV (cid:18) (cid:19) HerewehaveassumedauniversalκvaluetoallT-oddfermioncouplingsgeneratedbyEq.(8) Therefore, the effect of T-odd fermions to high energy collider phenomenology may not be negligible, and actually it is quantitatively important as we will discuss in later sections. The interaction terms in Eq. (8) in general contain flavor indices, and large flavor mixings can cause flavor-changing-neutral-current (FCNC) problem [13]. For simplicity, we assume the flavor diagonal and universal κ in this study. 7 q q U U U U u d 1 2 L1 L2 R1 R2 R+ R+ Y 1/30 2/15 8/15 2/15 8/15 2/15 1/3 1/6 1 − Y 2/15 1/30 2/15 8/15 2/15 8/15 1/3 1/6 2 − TABLE I: U(1) charges Y for fermions. The SM hypercharge is given by Y = Y +Y . A A 1 2 In the multiplet Ψ , there are other extra T-odd fermions. For those fermions, we simply c assume Dirac masses, as suggested in Ref. [5, 8]. Furthermore, we assume that their Dirac masses are so large (as large as about 3 TeV) that these extra T-odd fermions are decoupled, but remains to be small enough not to generate the naturalness problem in the Higgs mass parameter. Thus, in our following analysis, we will not consider any effects induced by these extra T-odd fermions. The U(1) charges Y for fermions are listed in Table I.1 Those charges are determined A A by the gauge invariance of the Yukawa couplings which we will discuss later. In addition to the normal SM gauge interactions, the T-odd fermions interact with their SM partner fermions and the heavy gauge boson as follows: g = W+ (u¯ γ d +u¯ γ d )+h.c. L √2 Hµ L µ L− L− µ L ¯ + [(gc T +g s Y )Z +( gs T +g c Y )A ]f γ f +h.c., (11) H 3f ′ H ′ Hµ − H 3f ′ H ′ Hµ L µ L− f=u,d X where Y = 1/10, and s ( sinθ ) describes the degree of mixing between heavy neutral ′ H H − ≡ gauge bosons with s gg′ vS2M and c cosθ . For clarity, the corresponding Feyn- H ≃ g2 g′2/5 4f2 H ≡ H − man rules are presented in Appendix A. Through these interactions, the T-odd fermion can contribute to heavy gauge boson productions. Also, it can be directly produced via exchang- ing light gauge bosons, heavy gaugebosons, and gluons at high energy hadroncolliders, such as the LHC, as we will discuss in the following sections. C. Yukawa couplings for Top and other fermions In order to cancel the large radiative correction to Higgs mass parameter induced by top- quark, we introduce in the top sector the singlet fields U and U , which are embedded, L1 L2 1 Strictly speaking, these U(1)A charges YA (A = 1,2) for fermions are defined by a sum of the U(1)A charges from the original SU(5) and extra fermion U(1) charges. 8 together with the q and q doublets, into the following multiplets: Q = (q ,U ,0 )T 1 2 1 1 L1 2 and Q = (0 ,U ,q )T. For the top-Yukawa interaction, one can write down the following 2 2 L2 2 T-parity invariant Lagrangian: [5, 6, 7]: λ f = 1 ǫ ǫ (Q¯ ) Σ Σ (Q¯ Σ ) Σ˜ Σ˜ u Lt −2√2 ijk xy 1 i jx ky − 2 0 i jx ky R+ λ f(U¯ U h+U¯ U )+h.c., i (12) − 2 L1 R1 L2 R2 where ǫ and ǫ are antisymmetric tensors, and i, j and k run over 1 3 and x and y over ijk xy − 4 5. u and U (i = 1,2) are SU(2) singlets. Under T-parity, these fields transform as − R+ Ri Q Σ Q , U U and u u . The above Lagrangian contains the following 1 ↔ − 0 2 R1 ↔ − R2 R+ → R+ mass terms: v v2 v2 λ f SM 1 SM + u¯ u + 1 SM U¯ u Lt ≃ − 1 f − 4f2 ··· L+ R+ − 2f2 L+ R+ (cid:20) (cid:18) (cid:19) (cid:18) (cid:19) (cid:21) ¯ ¯ λ f U U +U U +h.c. (13) − 2 L+ R+ L− R− (cid:0) (cid:1) Here we have defined the T-parity eigenstates as q (q q )/√2 = (id , iu ), U = + ≡ 1− 2 L+ − L+ L± U1√∓2U2 and UR± = UR1√∓2UR2. One T-odd Dirac fermion T− (T−L ≡ UL−, T−R ≡ UR−) gets a mass M = λ f (cf. Eq. (13)), and a T-odd combination of the doublets q and q obtains T− 2 1 2 a mass from (cf. Eq. (8)). The left-handed (or right-handed) top quark (t) is a linear κ L combination of u and U (or u and U ), and another independent linear combination L+ L+ R+ R+ is a heavy T-even partner of the top quark (T ): + u c s t X+ = X X X , (X = L,R), (14)      U s c T X+ − X X +X      where the mixings are approximately expressed by v c2(c2 s2)v2 s = s2 SM + , s = s 1 α α − α SM + ) , (15) L α f ··· R α − 2 f2 ··· (cid:20) (cid:21) with s = λ / λ2 +λ2 and c = λ / λ2 +λ2. The masses of the top quark (t) and T-even α 1 1 2 α 2 1 2 heavy top quaprk (T ) are given by p + c4 +s4 v2 λ c2s2 v2 M = λ c v 1 α α SM + , M = 1f 1 α α SM + . (16) t 1 α SM − 4 f2 ··· T+ s − 2 f2 ··· (cid:20) (cid:21) α (cid:20) (cid:21) Note that the T-even heavy top (T ) is always heavier than the T-odd heavy top (T ) + − in the effective theory considered here. The Feynman rules of SM and heavy gauge boson interactions in the top sector are also summarized in Appendix A. We note that the coupling 9 strength of W+t¯b is Veff = V c (see Appendix A) where V is the (t,b) element of the tb tb L tb Cabibbo-Kobayashi-Maskawa (CKM) matrix. For our numerical results shown below, we have assumed V = 1, so that Veff = c = 1 s2, where s is given in Eq. (15). Once tb tb L − L L the Veff is measured experimentally, then thpe parameter space of the model can be further tb constrained. In other word, when the parameter s varies, the effective coupling strength α of W+t¯b also varies under our assumption V = 1, so that the single-top production rate tb at the Tevatron and the LHC also varies. As s 0, it is approaching to the SM W+t¯b α → coupling strength. In the top sector, there are two free parameters λ and λ , which can be replaced by λ 1 2 1 and s as two independent parameters. The experimental value of the top quark mass (M ) α t gives the relation between λ and s as 1 α M 1 t λ = 0.71 (17) 1 v 1 s2 ≥ SM − α for s 0 andM = 175GeV. Moreover, fpollowing the methodpresented inRef.[14], we cal- α t ≥ culated theJ = 1 partialwave amplitudes inthecoupled system of(tt¯, T T¯ , b¯b, WW, Zh) + + states, which are relevant to the top Yukawa coupling, to estimate the unitarity limit of the corresponding scattering amplitudes. From the unitarity limit, we can get a mild constraint on the parameters: s /c 3.3, which corresponds to α α ≤ s 0.96 and λ 2.5, (18) α 1 ≤ ≤ cf. Eq. (17). Its detailed discussion is presented in Appendix B for completeness. We could also discuss the “naturalness” constraint on these parameters. If we calculate the one-loop contribution to the Higgs mass parameter (m ) induced by the top sector, the correction h is described by ∆m2 = c yt2 M2 a M2, where y = √2M /v and c is a constant h 16π2 T+ ≡ H H t t SM of O(1). This correction should not be much larger than the Higgs boson (on-shell) mass squared M2, otherwise fine-tuning is needed. Thus the coefficient a is a measure of the H H “naturalness” of the Higgs mass correction. If we take a¯ (= a /2c) to be smaller than 10, H H we get the upper limit on M as T+ a¯ M H H M 6.7 TeV . (19) T+ ≤ 10 120 GeV r (cid:18) (cid:19) In other word, 10 120 GeV f s 0.11 . (20) α ≥ a¯ M 1 TeV r H (cid:18) H (cid:19)(cid:18) (cid:19) 10 a = 10 M = 120 GeV H H 5 λ naturalness limit 1 f = 1 TeV f = 2 TeV 2 1 unitarity limit 0.7 top quark mass constraint 0.5 0 0.2 0.4 0.6 0.8 1 s α FIG. 1: Allowed region of parameters λ and s . Solid line (red) represents a relation between λ 1 α 1 and s required by top quark mass (M = 175 GeV), cf. Eq. (17). Dashed line (green) shows an α t upper limit on s from the unitarity bound on the J = 1 partial wave amplitude in the coupled α system of (tt¯, T T¯ , b¯b, WW, Zh) states, as expressed in Eq. (18). Dash-dotted lines (blue) show + + that naturalness consideration puts lower limit on s (or equivalently lower limit on λ ), as shown α 1 in Eq. (20), and the shaded region in upper-left area of the figure is excluded for f = 1 TeV. For f = 2 TeV, the excluded region is extended to the dash-dotted line with f = 2 TeV. Here we have assumed a¯ = 10 and M = 120 GeV. H H We summarize these constraints on the parameters of the top sector in Fig. 1. For the first and second generation up-type quark Yukawa couplings, we assume the same forms of Yukawa couplings as those for the top quark, cf. Eq. (13), except that we do not introduce SU(2)-singlet fields U and U for the first and second generations because we LA RA do not require the cancellation of the quadratic divergences induced from the light quark sectors, for their Yukawa couplings are tiny.