Triple Photon Production at the Tevatron in Technicolor Models

Physics Letters B 549 (2002) 154–158 www.elsevier.com/locate/npe

Triple photon production at the Tevatron in technicolor models

A. Zerwekh a,C.Diba,R.Rosenfeldb

a Department of Physics, Universidad Técnica Federico Santa María, Valparaíso, Chile b Instituto de Física Teórica, UNESP, Rua Pamplona 145, 01405-900 São Paulo, SP, Brazil Received 9 August 2002; received in revised form 30 September 2002; accepted 19 October 2002 Editor: H. Georgi

Abstract We study the process pp¯ → γγγ as a signal for associated photon–technipion production at the Tevatron. This is a clean signature with relatively low background. Resonant and non-resonant contributions are included and we show that technicolor models can be effectively probed in this mode.  2002 Elsevier Science B.V. All rights reserved.

1. Introduction Another interesting possibility is that the electroweak symmetry breaking is triggered by some new The origin of fermion masses and mixings is strong interaction, generally called technicolor, and in one of most important issues in particle physics. this case the lightest boson could be a pseudoscalar, Unfortunately, these parameters are inputs in the well- like a pion, named technipion. In fact, in these mod- tested Standard Model (SM). Fermion masses are els of dynamical symmetry breaking a whole new set possibly related to the electroweak symmetry breaking of resonances related to the technicolor sector is pre- mechanism, which is not known at the moment and dicted [2]. is the top priority of present and future experiments. It is important to find experimental signatures that In the SM, a scalar electroweak doublet with self- can distinguish these different models of symmetry interactions described by an ad hoc quartic potential breaking. A compilation of experimental signatures is responsible for the symmetry breaking, leaving a for different technicolor models, like multiscale and scalar physical boson, the Higgs boson (J PC = 0++), top-color assisted walking technicolor, can be found as a remnant. Favorite extensions of the SM, like the in [3]. minimal supersymmetric standard model (MSSM) [1], In this Letter we focus on the signature arising from also predict the existence of a heavy pseudoscalar associated photon–technipion production. This is anal- boson (J PC = 0−+), in addition to a light scalar boson. ogous to the associated gauge-higgs boson production. ¯ → () ± () The process pp ΠT (γ,Z,W ),whereΠT is a E-mail addresses: alfonso.zerwekh@fis.utfsm.cl isospin triplet (singlet) technipion, can be enhanced (A. Zerwekh), cdib@fis.utfsm.cl (C. Dib), [email protected] by low-lying technicolor resonances like the techni- (R. Rosenfeld). rho and the techni-omega. These processes have been

0370-2693/02/$ – see front matter  2002 Elsevier Science B.V. All rights reserved. PII:S0370-2693(02)02896-4 A. Zerwekh et al. / Physics Letters B 549 (2002) 154–158 155 studied in [4] and the importance of the process in- volving the final state photon has been stressed in [5]. + − → () The process e e γΠT was analyzed for LEP [6,7] and future linear colliders [7]. The limits from rare Z decays at LEP [6] are not applicable in our case since we will consider a technipion heavier than the Z boson. Recently, Lane et al. [8] re-studied this Fig. 1. Triangle anomaly giving the continuum contribution to the process taking into account both continuum and res- ¯ → () pp ΠT γ process. onance contributions, but concentrating on the dominant bb¯ decay mode. In this Letter we study the possibility of using ¯ → () → the process pp γΠT γγγ, which is a clean signature with relatively low background even in a hadronic environment, to put some constraints in some technicolor models. We also include both resonant and non-resonant contributions in our analysis and Fig. 2. Techni-omega and techni-rho contributions to the pp¯ → () perform a simulation of the significance level of this ΠT γ process. signature. SU(NTC), are given by: 4e2 2. The model SΠ γγ = √ NTC, T 6 − 2 The coupling of the technipion to two gauge bosons = 2 1√ 4sin θw SΠTZγ 2e NTC, (3) is mediated by the Adler–Bell–Jackiw anomaly [9] 6sin2θw () arising from a techniquark triangle. The ΠT B1B2 coupling can be parametrized as: 4e2 =− √ SΠ γγ NTC, T 3 6 SΠTB1B2 µ ν α β AΠ B B = √ µναβ ε ε k k , (1) 2 T 1 2 2 1 2 1 2 4e tan θw π F = √ 4 2 ΠT SΠ Zγ NTC. (4) T 3 6 where ε1,2 and k1,2 are the polarization vectors and Consequently, the decay of neutral technipions into momenta of the gauge bosons B1,2, respectively. FΠT is the technipion decay constant, which is related to the two photons is induced entirely by the anomaly. In contrast, the associated production of a photon with technipion coupling to the axial current. The group- a neutral technipion is mediated by both Π( )γγ and theoretical factor SΠ B B is given by [10]: T T 1 2 () ΠT γZ anomalous vertices, as well as by possible {Q ,Q } = 1 2 s-channel vector resonances, the isosinglet techni- SΠTB1B2 g1g2 Tr QΠT , (2) 2 omega (ωT), and the isotriplet techni-rho (ρT). These further contributions are depicted in Fig. 2 and can be where g and g are the corresponding gauge coupling 1 2 treated as a generalization of vector meson dominance. constants and Q , Q and Q are the charges 1 2 ΠT From the viewpoint of perturbation theory, the ano- under the gauge groups and isospin, respectively, of malous couplings appear only at the one-loop level. the technifermions circulating in the loop. For our () The resonances, considered as techniquark bound purposes we will be concerned only with the ΠT γγ states, are a sum to all orders in technicolor interac- () and ΠT γZ couplings, since they provide the only tions and therefore include one-loop effects. However, ¯ → () contributions to the process pp γΠT ,shownin no ambiguity of double-counting arises when we con- Fig. 1, and the corresponding group-theo factors, for sider both the anomaly and resonance contributions, a one-family technicolor model with gauge group as these are due to very different energy scales, the 156 A. Zerwekh et al. / Physics Letters B 549 (2002) 154–158 former being a low-energy effect and therefore is not 3. Simulation of the process important at the resonance mass scale. In the absence of isospin violation, the techni- The inputs to our codes are the relevant masses of omega mixes with the isoscalar part of the electroweak () ΠT , ωT, ρT, the technipion decay constant FΠT and current, the Bµ field, whereas the techni-rho mixes the resonance widths Γρ and Γω . In order to reduce 3 T T with the isotriplet part, the Wµ field. In terms of the the number of parameters, we will use in our calcu- physical fields of the photon and the Z-boson, the lations the reference set of values m () = 110 GeV mixing strengths are given by: ΠT and m = m . We also adopt N = 4andF = ωT ρT TC ΠT 82 GeV, as appropriate in multiscale walking techni- = α + gωT−γ (QU QD), color, but the results are relatively insensitive to this αT choice since the couplings of the vector resonances α =− + ( ) → gωT−Z (QU QD) tanθW , (5) and the branching ratio BR(ΠT γγ)are indepen- αT dent of FΠT . The vector resonance widths were ob- and tained from PYTHIA version 6.125 [11]. We used the parton distribution function CTEQ6 α gρ −γ = , [12] with√ both momentum and factorization scales√ T α T set at sˆ and a total center-of-mass energy of s = = α 2000 GeV. We convoluted the relevant parton distrib- gρT−Z cot2θW , (6) αT ution functions with the amplitudes described above. Total luminosities of 2 fb−1 (run 2a) and 30 fb−1 (op- where α is the fine structure constant and α is related T timistic run 2b) were considered. to the technicolor coupling constant g and can be T The irreducible background was generated using estimated by a naïve scaling from QCD: the program CalcHEP 2.1 [13]. The main irreducible ¯ 2 contribution comes from uu,¯ dd → γγγ. gT 3 αT = = 2.9 . (7) We have also estimated the reducible background 4π N TC pp¯ → jjγ and pp¯ → e+e−γ , with jets and electrons Finally, the relevant amplitudes for the decays being mis-identified as photons. Due to our high → () ρT,ωT γΠT are given by, in the notation of [5]: pT cuts on the final state particles and the mis- identification probabilities (of the order of 10% for M → ( ) −3 0 VT(q) G(p1)ΠT (p2) e → γ and 10 for jet → π , η → γ ), we can safely eV neglect these reducible backgrounds. = VTGΠT µ ∗ν α β µναβ ε (q)ε (p1)q p1 , (8) A Gaussian smearing for the final state photon MV √ energy with σE/E = 0.20/ E [14] was applied to where MV is a mass parameter usually taken to be both signal and background. 200 GeV and = VωTγΠT cosχ, 4. Results V () = (QU + QD) cosχ , ωTγΠT = + In Figs. 3 and 4 we show for illustrative purposes VρTγΠT (QU QD) cosχ, the differential cross sections for signal and back- = Vρ γΠ cosχ . (9) T T ground as a function of the 2-photon and 3-photon in- = = In the equation above χ and χ are mixing angles be- variant mass, respectively, for mωT mρT 350 GeV = tween the isospin eigenstates and the mass eigenstates. and m () 110 GeV. In both figures one can clearly ΠT In our computations we use a value of sin χ = sin χ = see a signal that stands out above the background. 1/3andQU + QD = 5/3 [5]. In order to compute the The 2-photon distribution in Fig. 3 shows a peak cen- fermionic widths of the technipions we use the cou- tered around the technipion mass (which we chose at pling constant g ¯ = m /F . 110 GeV). Since this is a 2-photon invariant mass dis- ΠTf f f ΠT A. Zerwekh et al. / Physics Letters B 549 (2002) 154–158 157

Fig. 3. 2-photon invariant mass distribution for signal (upper his- Fig. 4. 3-photon invariant mass distribution for signal (up- = = = togram) and background (lower histogram) for mωT mρT 350 per histogram) and background (lower histogram) for mωT − −1 GeV and m = 110 GeV for L = 30 fb 1. The bin size used in mρ = 350 GeV for L = 30 fb . The bin size used in these his- Π( ) T T tograms is 1.1GeV. these histograms is 0.43 GeV. tribution in 3-photon events, the width of the peak does Table 1 not correspond to the technipion width, but it contains Cross sections (before cuts), number of events (after cuts), signal/background ratio and significance of the signal for L = − − also the combinatoric error from the selection among 2fb 1 and 30 fb 1 (first and second figures, respectively) for dif- the three photons. Indeed, for a technipion much nar- ferent techni-resonance masses rower than the techni-vector meson, the widths in both mω ,ρ (GeV) σ (fb) Events S/B Significance figures are comparable. The 3-photon distribution in T T 210 18.22 12−175 38.621.2−82.3 Fig. 4 shows a peak centered around the techni-vector 250 9.22 9−135 19.013.1−50.7 meson mass. In this case, the width in the histogram 300 4.83 4.3−64.713.17.5−29.1 reflects the resonance width together with the photon 350 2.70 2.6−38.810.35.2−20.0 energy resolution that we use in the simulation. In ad- 400 1.83 0.92−13.85.02.2−8.4 − − dition, the distribution away from the peak is domi- 450 1.06 0.8 12.27.82.5 9.8 500 1.00 0.2−3.33.50.9−3.4 nated by the anomaly contribution. As it is compara- 550 0.78 0.2−3.43.50.9−3.5 ble to the background, the non-resonant contribution 600 0.52 0.06−0.91.60.3−1.2 cannot be detected. In order to further suppress the background, the following cuts were used: signal as a function of the techni-resonances for the two luminosities. MωT Mγγγ ∈ Mω − ,Mω + 20 GeV , We now comment on the dependence on some of T 10 T the parameters that we held fixed in our analysis. First MΠT of all, except for the first value of the techni-resonance Mγγ ∈ MΠ − ,MΠ + 10 GeV , T 10 T mass considered in Table 1, the dominant process () ¯ → → pTγ 70 GeV. (10) is pp ωT γΠT ,sinceformρT > 2m ( ) the ΠT strong decay mode ρ → Π Π dominates. There- In Table 1 we present our results for the to- T T T fore, for most of the parameter space the techni-omega tal number of 3-photon events for a given techni- 1 contribution is dominant. The dependence on (QU + resonance mass and for 2 different integrated lumi- − − Q ), χ, χ and M is evident from the techni-omega nosities, namely L = 2fb 1 and 30 fb 1. D V coupling to either the isoscalar singlet or the isoscalar We can see that resonances up to 350 GeV can be − triplet pseudoscalar. The techni-omega would totally found at the 5σ level even with L = 2fb 1.Foran accumulated luminosity of L = 30 fb−1, resonances as heavy as 550 GeV can be detected at the 3σ level. 1 We would like to thank the referee for bringing this point to our In Fig. 5 we show the statistical significance of the attention. 158 A. Zerwekh et al. / Physics Letters B 549 (2002) 154–158

techni-vectors as well as the technipion masses from the 3- and 2-photon invariant mass distributions, respectively.

Acknowledgements

A.Z. and C.D. received partial support from Fonde- cyt (Chile) grants No. 3020002 and 8000017, respectively. R.R. would like to thank CNPq and PRONEX for partial ﬁnancial support. The authors also acknowl- − Fig. 5. Statistical signiﬁcance of signal for L = 30 fb 1 (dots) edge support from Fundacion Andes (Chile) and Fun- −1 and L = 2fb (solid line) as a function of the masses of the daço Vita (Brazil) grant C-13680/4. techni-resonances. decouple for Q + Q = 0. For general values the U D References signal cross section scales roughly as:

2 [1] See, e.g., S.P. Martin, in: G.L. Kane (Ed.), Perspectives in 8e 2 + 2 2 2 NTC(QU QD) cos χ Supersymmetry, World Scientific, Singapore, hep-ph/9709356. 3MV [2] For a recent review, see C.T. Hill, E.H. Simmons, hep- 2 ph/0203079, and references therein, Phys. Rep., submitted for 8e + 2 + 4 2 publication. 2 NTC(QU QD) cos χ . (11) 27MV [3] For a recent compilation of technicolor signatures, see http://d0server1.fnal.gov/users/gll/technicolor.html. [4] E. Eichten, K. Lane, J. Womersley, Phys. Lett. B 405 (1997) 5. Conclusions 305. [5] K. Lane, Phys. Rev. D 60 (1999) 075007. [6] G. Rupak, E.H. Simmons, Phys. Lett. B 362 (1995) 155. In this Letter we have examined the triple photon [7] V. Lubicz, P. Santorelli, Nucl. Phys. B 460 (1996) 3. production at the Tevatron as a signature for techni- [8] K. Lane, K.R. Lynch, S. Mrenna, E.H. Simmons, hep- color models. We have included both resonant and ph/0203065. non-resonant contributions, but the former are domi- [9] S.L. Adler, Phys. Rev. 177 (1969) 2426; J.S. Bell, R. Jackiw, Nuovo Cimento A 60 (1969) 47. nant in a hadron machine, where the center-of-mass [10] J. Ellis, M.K. Gaillard, D.V. Nanopoulos, P. Sikivie, Nucl. energy of the process is not fixed. The relatively low Phys. B 182 (1981) 529. background enables one to obtain large significance [11] T. Sjöstrand, et al., Comput. Phys. Commun. 135 (2001) 238. levels. We found that technicolor models can be effec- [12] J. Pumplin, et al., JHEP 0207 (2002) 12. tively probed in this mode and, with an accumulated [13] A. Pukhov, et al., hep-ph/9908288, Preprint INP MSU 98- L = −1 41/542. luminosity of 30 fb , resonances as heavy as [14] See, e.g., Report of the Tevatron Higgs Working Group, 550 GeV can be detected or excluded at the 3σ level. M. Albrow, et al., hep-ph/0010338. Using this mode we can have information on both the