# Full text of "Asymmetries in t tbar production: LHC versus Tevatron"

## See other formats

Asymmetries in tt production: LHC versus Tevatron J. A. Aguilar-Saavedra, M. Perez- Victoria Departamento de Fisica Tedrica y del Cosmos and CAFPE, Universidad de Granada, E-18071 Granada, Spain The measurement of a charge asymmetry in tt production at LHC constitutes more than an independent confirmation of the forward-backward asymmetry found at Tevatron. Indeed, both measurements together can be used to identify the source of the asymmetry. This is demonstrated for the case of new Z' , W' vector bosons and colour-sextet and triplet scalars, exchanged in t, u channels respectively, and a very heavy axigluon in the s channel. In particular, current LHC measurements disfavour Z' , W' models above the 2a level. PACS numbers: 12.60.-i,14.65.Ha Keywords: top quark; hadron colliders o O Q in 43 : Oh- i . Or <D ■ 43 ' ^1- > o in o X The top quark is singled out among the other quarks by its large mass and short lifetime, making the study of its production and decay properties specially clean. Furthermore, thanks to these particular features, it can be a sensitive probe of new physics beyond the Standard Model (SM) . Actually, some observations at the Fermilab Tevatron might already be a hint of new physics. The CDF and DO Collaborations have measured the values A FB = 0.158 ±0.075 [1], A FB = 0.196 ±0.065 0, respec- tively, for the forward-backward (FB) asymmetry in top quark pair production. Both are above the SM predic- tions, e.g. Afg = 0.051 - 0.089 The CDF Col- laboration also reports a clear enhancement of the asym- metry at high tt invariant masses, A-p-Q — 0.475 ± 0.114 for m t i > 450 GeV (more than three standard deviations above the SM prediction A F ^ = 0.088 - 0.12), whereas DO does not find a statistically significant mass depen- dence. On the other hand, the CMS Collaboration has re- cently presented a measurement of the charge asymme- try in tt production at the CERN Large Hadron Collider (LHC), using 1.09 ib -1 of data 0. The reported value, A c = -0.016 ± 0.030 (stat)^;^ (syst), is still domi- nated by the statistical uncertainty, and a much better precision is expected in the near future. Systematic un- certainties are also expected to improve with a better knowledge of the detector. Clearly, the measurement of the charge asymmetry at LHC provides an independent test of the excess observed at Tevatron. We recall here that the Tevatron asymmetry Ap B mentioned above is defined as the relative difference (normalised to the total number) between the number of events with cos 8 > and cos 6* < 0, with 8 the angle between the top quark and initial proton in the centre of mass frame. At LHC, the charge asymmetry Ac measured by the CMS Col- laboration is the relative difference between events with \r) t \ > \rjt \ and \i] t \ < \ru\, with r\ t {r\A the pseudo-rapidity of the top (anti)quark, rj = — logtan6*/2, in the labora- tory frame. This definition takes advantage of the larger average longitudinal boost of t quarks in pp collisions as- sociated to a FB asymmetry at the parton level. Many SM extensions have been proposed to accommo- date the FB asymmetry measured by the CDF Collabora- tion. These models introduce new particles which can be exchanged in s, t or u channels in the processes qq tt, with q = u,d. While the presence of narrow s-channel resonances in tt production could be eventually spotted by an examination of the tt invariant mass distribution with sufficient statistics (perhaps also requiring a centre of mass energy of 14 TeV), this is more difficult for new particles exchanged in t or u channels, as for example new Z' ', W' vector bosons (t channel), or colour-sextet and triplet scalars (u channel). In this paper we show that, combining the measure- ments of the charge asymmetry at LHC and the FB asym- metry at Tevatron, it is possible to discriminate among the different models, already disfavouring some of them. To arrive at this conclusion, it is necessary to go beyond the usual analyses with a few selected benchmark points, and instead scan over all the allowed values of the cou- plings and masses. It is also crucial to impose existing constraints from experimental data, in order to bound their range of variation. The most obvious and robust constraints on the tt asymmetries result from tt production itself. At Teva- tron, the total cross section has been precisely measured, a = 7.50 ± 0.48 pb {H, which limits the possible size of new physics contributions. Total cross section mea- surements at LHC will not be so restrictive because tt production is dominated by gg fusion and the systematic and theoretical uncertainties leave more room for possi- ble departures in qq — > tt. On the other hand, the tt cross section at high invariant masses is sensitive to new physics and sets constraints on the masses and couplings of any new particles giving rise to the tt asymmetry [9j . Of course, there are additional restrictions on the extra particles, as for example the production of like-sign top pairs and dijets. They are not considered here because they can be evaded in specific models I3~12|- Further- more, we do not attempt to reproduce the tt invariant mass distribution at Tevatron for the models considered, as this distribution is reasonably similar to the measured 2 one [l| for most of the parameter space allowed by other constraints, nor we consider the tij cross section at LHC, which can be restrictive in certain parameter space re- gions of t-channel models (l3| . The simplified analysis presented here suffices for our purpose. Taking into ac- count only the constraints from the Tevatron cross sec- tion and the LHC tail, we find that different SM exten- sions give predictions for the asymmetries corresponding to different, often disjoint regions in the (Afb, Ac) plane, rendering model discrimination feasible. The inclusion of additional constraints will only shrink the allowed regions and strengthen our conclusions. We are also conservative in the interpretation of the ti production limits. There are some discrepancies between different state-of-the-art predictions for the SM ti total cross section at Tevatron, with some results quite close to the measured one. for example a = 7.46±g'jo pb but also significantly smaller ones, a — 6.30 ± 0.191q23 [l5| . While the former value requires small new physics con- tributions or large new amplitudes A ncw ~ —2 Asm, the latter allows for moderate contributions to both the cross section and the asymmetry. Thus, when requir- ing agreement with the Tevatron ti cross section we let the SM contribution be anywhere between these two val- ues, which makes our constraints much looser (and hence the allowed regions larger) than if we stick to one of ei- ther predictions. Taking into account the uncertainties in these theoretical predictions as well as in the experi- mental measurement, we require in our analysis that new physics contributions to ti production lie inside the inter- val [—0.8, 1.7] pb. For the LHC cross section at the high- mass tail, no dedicated analysis is available yet. Still, an examination of the invariant mass distributions that have been released [16] shows that large excesses over the SM prediction are already excluded. Following Refs. 0, [ItJ we take the cross section for m t i > 1 TeV as a constraint, requiring that its value is at most three times the SM pre- diction. All possible vector bosons and scalars contributing to qq — > ti have been classified in Ref. [9j according to their transformation properties under the SM gauge group SU(3) C x SU(2) L x U(l)y. There are ten possible new vector bosons and eight types of scalars, but perhaps the most interesting extensions are new colour-singlet or octet vector bosons, and colour-triplet or sextet scalars. Their transformation properties and general interaction Lagrangians with the quarks are collected in Table [Jj We use standard notation with left-handed doublets qn and right-handed singlets um, dm; t 1 are the Pauli matrices, A a the Gell-Mann matrices normalised to tr(A a A & ) = 26 a b and ip c = C-tp 7 ', with C the charge conjugation matrix. The subindices a, 6, c denote colour, and e a bc is the totally antisymmetric tensor. In our analysis we take five illustrative examples rep- resenting a large fraction of the models proposed in the literature to explain the ti asymmetry, which also involve TABLE I. Some vector bosons and scalar representations me- diating qq — i> ti. Label Rep. Interaction Lagrangian +gt j d m ^d R] ) B M B\ (1,1)! -9ijdm^u Rj Bl? +h.c. e M (8, 1) - [g^qu^^-qLj + gijUR^^-URj +gt J d Ra ^d Rj ) Ql ^ 4 (3, 1)_4 -gijEabcURAbURjc + n - c - Q 4 (6, 1)_4 [u R i a U Rjb + U R ibU Rja ] fi 4ai>t + h.c. the three possibilities of new particle exchange in the s, t or u channels. Flavour-violating Z' boson If], 18- 2^1: A neutral vec- tor boson £> M exchanged in the t channel in uu — > ti. We take its Z'tu couplings to be right-handed, g™ 3 ^ 0, as preferred by B physics constraints. Our results are in- dependent of this choice, however. For a real Z' boson the contribution to the FB and charge asymmetries is strongly constrained by the absence of like-sign top pair production [23]. However, the relation between tt and ti production can be evaded by placing the new boson in a complex representation of a flavour group [10|. W boson [H-[26||: A charged boson Bj\ with right- handed couplings g\3 exchanged in the t channel in dd — > ti. Charged bosons with left-handed couplings can also appear in SU(2)^ triplets but this possibility is again disfavoured by B physics constraints. Axigluon 0, [27l - [29| : A colour octet vector with axial couplings g\ = — g% = —gfi, produced in the s channel, qq —> ti. We consider this new particle to be heavy enough not to be produced on shell; otherwise its presence would generally be noticed by a bump in the ti invariant mass distribution [38| and the discrimination from i-, u-channel resonances would be straightforward. The exception to this rule is given by colour octets below the ti production threshold [39j or very broad [39L |4(]1 | but in those cases the predictions are very model-dependent and deserve a separate study j^. We assume the ax- igluon is only produced in uu and dd initial states, which give the largest fraction of the ti cross section at the Teva- tron and the LHC, and neglect additional contributions, for example from ss annihilation. In any case, includ- ing these small contributions hardly affects our results. No assumptions are necessary about the relative size of first- and third-generation couplings, since the axigluon is taken heavy and the cross section is propo rtional to the product of couplings. (Other models [30h33I | in which the couplings of the new colour octet are not purely axial give very similar results for the relation between Aq and ^4fb , but the asymmetries generated are smaller relative to the 3 increase in cross section.) Colour-triplet scalar fill 34-3(|: A colour triplet w 4 with flavour- violating tu couplings 1713 , necessarily right- handed, exchanged in the u channel in uu — > tt. Notice that the antisymmetry in colour indices implies that di- agonal couplings to uu, tt identically vanish. Colour-sextet scalar [TU 12, 34-37 1: A colour sextet O , also with right-handed flavour- violating tu couplings f/13, and exchanged in the u channel. In contrast with w 4 , for the sextet there may be diagonal uu, tt couplings, albeit not related to the flavour-violating ones. They can potentially give rise to large (unobserved) tt signals unless suppressed by some flavour symmetry [llj, [l2 1 . The predictions of these models for the asymmetries Apb and Ac are found by performing a comprehen- sive scan over the allowed parameter space, with par- ticle masses between 100 GeV and 10 TeV, except for the axigluon, which is assumed to be very heavy and its amplitude replaced by a four-fermion interaction (4l| . The interval of the scan is adjusted as necessary to ob- tain a smooth variation of the predictions with the mass. The couplings are scanned uniformly in the range al- lowed by the Tevatron cross section limits, i.e. requir- ing Act = [—0.8, 1.7] pb. This constraint fixes the maxi- mum size of the coupling for each mass considered. (The resulting allowed range for the coupling may be a sin- gle interval or the union of two, due to the competition between the interference and quadratic contributions to the cross section.) The total number of parameter space points sampled ranges between more than 2000 for the Z' boson to almost 10000 for oj 4 . Our computations are per- formed by including the new particles and four-fermion interactions in the leading-order generator Protos [42j |. The new physics contributions to A-p-Q (for m t j > 450 GeV) and Ac (inclusive) are presented in Fig. [T] (left), for the five models studied, taking into account the con- straints on the tt cross section and tail mentioned above. We only show the regions where is positive, as is the excess found by the CDF and DO Collaborations. To a good approximation, the total asymmetries Afb, Ac are obtained by summing the SM contributions, A|g = 0.088±0.013 (43, A C M = 0.0130±0.0011 Q, to A^ and A c ew , respectively. This amounts to considering the dom- inant contributions to the total asymmetry, which are: (i) interference between new physics and tree-level SM con- tributions, as well as purely new physics ones; (ii) the interference between the next-to-leading order and tree- level SM. As one can observe, current LHC data already bring interesting implications for the models discussed. A salient feature of our analysis is that for the Z' boson the positive asymmetries (which require a large coupling) have minimal values A£° B W > 0.32, A c cw > 0.04 allowed by tt cross section constraints. Hence, the present LHC measurement of Ac disfavours this model at 2.2tr (97% confidence level) . The same measurement also disfavours the W' at 2a (95% confidence level) if the new physics contribution to the Tevatron asymmetry is moderate, A^™ > 0.12, as it is preferred by the CDF measure- ment, Af^ = 0.387 ±0.115, and also hinted by the more recent one by the DO Collaboration. The rest of models predict smaller asymmetries at LHC, and are less con- strained by the present measurement of Ac . Notice that the difference between a W boson and the scalar and ax- igluon models stems from the different uu and dd parton densities. At Tevatron (pp collider) both u, d from the proton and u, d from the antiproton are valence quarks, so that dd is roughly 1/4 smaller than uu. At LHC (pp) both u, d are sea quarks and dd is only 1/2 smaller than uu, resulting in a slope twice larger for the W allowed region. Further discrimination can be achieved by the mea- surement of Ac at high invariant masses, for example rriu > 600 GeV for which the SM cross section is only six times smaller than the total rate and statistics will be good. The result is shown in Fig. [T] (right). For a Z' boson exchanged in the t channel the asymmetry en- hancement is much more pronounced than for the rest of models, and an unfolded measurement at high mass can definitely probe this model. (The same comment ap- plies to W bosons.) Moreover, although apparently the scalars and the axigluon have similar predictions also for high m t t, the fact is that model parameters giving close (A^, A c cw ) points in the left-hand plot correspond to different A c ew in the right-hand one. This can be under- stood by recalling that for light oj 4 /£1 4 scalars exchanged in it-channel the top quarks are preferrably produced in the direction of the initial antiquark. A positive Afb at Tevatron can only be generated for scalar masses above few hundreds of GeV, so that the enhancement of the it- channel propagator in the backward direction is less pro- nounced. For this reason, Ac at LHC is small for large m t i except when u 4 /tt 4 are heavy and it even decreases with mtt f° r light scalars and/or high m t i- These arguments provide a strong motivation for the analysis of the m t j dependence of the charge asymmetry at LHC. To demonstrate its relevance we select one point from Fig. [T] (left), A™™ 0.13, A c cw ~ 0.016 and three models yielding these values: (i) a heavy axigluon |29| (see Ref. [J for details on the effective operators); (ii) a colour sextet [II HS-^; (iii) a colour triplet \x&\M- [3^|. We plot in Fig.[2]the charge asymmetry as a function of the cut to™ on the tt invariant mass. The differences in the behaviour are striking and illustrate the general trend. In order to reproduce the same values for Apg^, A™™, the colour sextet and triplet must have different mass and coupling, because the interference with the SM has opposite sign [9j, [35 1 . But it is the mass that mainly determines the variation of the asymmetry with to™ 11 . As we have mentioned aboce, for a relatively light u- channel particle (for instance the w 4 benchmark in Fig. [5]) the asymmetry does not grow with m t i due to the effect of the it-channel propagator which prefers backward top 4 0.08 0.06 0.25 0.04 0.02 M B (m ;; >450GeV) A pB (m rr >450GeV) FIG. 1. Left: allowed regions for the new physics contributions to the FB asymmetry at Tevatron and the inclusive charge asymmetry at LHC. Right: the same, with the charge asymmetry for m t i > 600 GeV. 0.15 1 i«0.10 §„ : C/A~ = 0.33TeV"~ £2 4 : g 13 = 1.17 M= 950GeV CO 4 : g n = 0.936 M= 340 GeV 400 500 600 700 800 900 1000 FIG. 2. Dependence of the charge asymmetry on the m t i cut, for a point with A 1 ^ ~ 0.13, y4g cw ~ 0.016 (inclusive). quarks, while the numerator prefers forward top quarks. For a heavier it-channel particle (as for example the Q benchmark in the same figure) the it-channel propagator effect is attenuated and the asymmetry reaches higher values. Thus, a more precise measurement of the inclu- sive charge asymmetry and an unfolded measurement at high mass will be of great help discriminating these mod- els. Additional information can eventually be obtained from more subtle observables, such as the polarisation of the ti pair [44J. Besides, we have also checked that for a central charge asymmetry with \r)tt\ < 1 S EH the results are quite similar, while for a forward one some discrimination power is lost. The allowed regions in Fig.[T]have been obtained, as ex- plained above, by imposing a "minimal" set of constraints: the ti cross section at the Tevatron and the high invari- ant mass cross section at the LHC. Hence, these allowed regions contain all the possible predictions for the asym- metries in viable models^ Additional constraints could be imposed, for example the tij cross section at the LHC, which is important for a certain range of masses in Z' , W' models [j^ j. or the ti tail at the Tevatron. Doing this is not necessary in our analysis, since the regions we obtain with our minimal constraints are already disjoint, as we have shown. At any rate, we expect that the range of predictions for viable models will not be much smaller than the allowed regions shown in Fig.Q] For example the tij cross section constraint is important only for a nar- row Z\ W mass range above the top quark mass, where on-shell associated production, e.g. gu tZ 1 — > tiu, is large. Also, the constraints on new physics from the Tevatron tail are loosened by the smaller detection effi- ciency for the new contributions (4(| . Though systematic scans of the parameter space for viable models have not been performed elsewhere, some sample points studied in detail [ll|, [H, EH 0, Efi| suggest that most of the parameter space allowed by our constraints gives viable models. In summary, in this paper we have investigated the re- lation between the ti asymmetries at Tevatron and LHC. If the excess found by the CDF and DO Collaborations 1 Notice that the bulk contribution to the total cross section at Tevatron comes from the region with m t j ~ 400 — 500 GeV, where detection efficiency of the new physics is not very different from that of SM tt production. At LHC the efficiency lose at the tail for light t-channel particles is not very pronounced Q], and in any case the agreement between the SM prediction and the experimental measurement suggests much more stringent limits than the ones considered here. 5 corresponds to new physics, the most robust probe to in- vestigate its origin is the study of tt production at LHC, searching for a charge asymmetry and an enhanced tt tail. We have shown how the measurements of the Tevatron and LHC asymmetries can be used to identify the source of these excesses. In particular, with present data the models with Z' and W bosons are already disfavoured at the 95% confidence level. The results presented here also provide a strong motivation for the detailed study of the mtt dependence of the charge asymmetry at LHC, which will be possible thanks to the good statistics ex- pected at this top quark factory. This work has been partially supported by projects FPA2010-17915 (MICINN), FQM 101 and FQM 437 (Junta de Andalucia) and CERN/FP/116397/2010 (FCT). [1] T. Aaltonen et al. [CDF Collaboration], Phys. Rev. D 83, 112003 (2011). [2] V. Abazov et al. [DO Collaboration], larXiv: 1107.49951 [hep-ex] . [3] P. Ferrario and G. Rodrigo, Phys. Rev. D 78, 094018 (2008). [4] W. Bernreuther and Z. G. Si, Nucl. Phys. B 837, 90 (2010). [5] W. Hollik and D. Pagani, arXiv:1107.2606 [hep-ph]. [6] J. H. Kuhn and G. Rodrigo. larXiv :1109.6830 [hep-ph]. [7] CMS Collaboration, note CMS PAS TOP-11-014. [8] T. Aaltonen et al. [CDF Collaboration], CDF note 9913. [9] J. A. Aguilar-Saavedra and M. Perez- Victoria, JHEP 1105, 034 (2011). [10] S. Jung, A. Pierce and J. D. Wells, Phys. Rev. D 83, 114039 (2011). [11] Z. Ligeti, M. Schmaltz and G. M. Tavares, JHEP 1106, 109 (2011). [12] B. Grinstein , A. L. Kagan, M. Trott and J. Zupan, Phys. Rev. Lett. 107, 012002 (2011). [13] M. I. Gresham , I. -W. Kim and K. M. Zurek, larXiv: 1107.43641 [hep-ph]. [14] U. Langenfeld, S. Moch and P. Uwer, Phys. Rev. D 80, 054009 (2009). [15] V. Ahrens, A. Ferroglia, M. Neubert, B. D. Pecjak and L. L. Yang, JHEP 1009, 097 (2010). [16] CMS Collaboration, note CMS PAS TOP-10-007. [17] C. Delaunay, O. Gedalia, Y. Hochberg, G. Perez and Y. Soreq, JHEP 1108, 031 (2011). [18] S. Jung, H. Murayama, A. Pierce and J. D. Wells, Phys. Rev. D 81, 015004 (2010). [19] V. Barger, W. Y. Keung and C. T. Yu, Phys. Rev. D 81, 113009 (2010); V. Barger, W. Y. Keung and C. T. Yu, Phys. Lett. B 698, 243 (2011). [20] J. Cao, L. Wang, L. Wu and J. M. Yang, Phys. Rev. D84, 074001 (2011). [21] E. L. Berger, Q. H. Cao, C. R. Chen, C. S. Li and H. Zhang, Phys. Rev. Lett. 106, 201801 (2011). [22] B. Bhattacherjee, S. S. Biswal and D. Ghosh, Phys. Rev. D 83, 091501 (2011). [23] J. A. Aguilar-Saavedra and M. Perez- Victoria, Phys. Lett. B 701, 93 (2011) [24] K. Cheung, W. Y. Keung and T. C. Yuan, Phys. Lett. B 682, 287 (2009); K. Cheung and T. C. Yuan, Phys. Rev. D 83, 074006 (2011). [25] Q. H. Cao, D. McKeen, J. L. Rosner, G. Shaughnessy and C. E. M. Wagner, Phys. Rev. D 81, 114004 (2010). [26] J. Shelton and K. M. Zurek, Phys. Rev. D 83, 091701 (2011). [27] P. Ferrario and G. Rodrigo, Phys. Rev. D 80, 051701 (2009) . [28] P. H. Frampton, J. Shu and K. Wang, Phys. Lett. B 683, 294 (2010). [29] Y. Bai, J. L. Hewett, J. Kaplan and T. G. Rizzo, JHEP 1103, 003 (2011). [30] A. Djouadi, G. Moreau, F. Richard and R. K. Singh, Phys. Rev. D 82, 071702 (2010); A. Djouadi, G. Moreau and F. Richard, Phys. Lett. B701, 458-464 (2011). [31] C. Delaunay, O. Gedalia, S. J. Lee, G. Perez and E. Pon- ton, Phys. Rev. D 83, 115003 (2011); Phys. Lett. B703, 486-490 (2011). [32] G. Burdman, L. de Lima and R. D. Matheus, Phys. Rev. D 83, 035012 (2011). [33] E. Alvarez, L. Da Rold and A. Szynkman, JHEP 1105, 070 (2011), [34] J. Shu, T. M. P. Tait and K. Wang, Phys. Rev. D 81, 034012 (2010). [35] A. Arhrib, R. Benbrik and C. H. Chen, Phys. Rev. D 82, 034034 (2010). [36] I. Dorsner, S. Fajfer, J. F. Kamenik and N. Kosnik, Phys. Rev. D 81, 055009 (2010); Phys. Rev. D 82, 094015 (2010) . [37] K. M. Patel and P. Sharma, JHEP 1104, 085 (2011). [38] R. Barcelo, A. Carmona, M. Masip and J. Santiago, Phys. Rev. D 84, 014024 (2011). [39] J. A. Aguilar-Saavedra and M. Perez- Victoria, Phys. Lett. B 705, 228 (2011). [40] R. Barcelo, A. Carmona, M. Masip and J. Santiago, larXiv:1106.4054l [hep-ph]. [41] F. del Aguila, J. de Bias and M. Perez- Victoria, JHEP 1009, 033 (2010); J. A. Aguilar-Saavedra, Nucl. Phys. B 843, 638 (2011); C. Degrande, J. M. Gerard, C. Grojean, F. Maltoni and G. Servant, JHEP 1103, 125 (2011). [42] J. A. Aguilar-Saavedra, Nucl. Phys. B 804, 160 (2008). [43] J. M. Campbell and R. K. Ellis, Phys. Rev. D 60 (1999) 113006. [44] J. Cao, L. Wu and J. M. Yang, Phys. Rev. D 83, 034024 (2011) ; D. Choudhury, R. M. Godbole, S. D. Rindani and P. Saha, Phys. Rev. D 84, 014 023 (2011); D. Krohn, T. Liu, J. Shelton and L. T. Wang. larXiv: 1 105.37431 [hep- ph[; D. W. Jung, P. Ko and J. S. Lee, Phys. Lett. B 701, 248 (2011). [45] J. L. Hewett, J. Shelton, M. Spannowsky, T. M. P. Tait and M. Takeuchi, Phys. Rev. D84, 054005 (2011). [46] M. I. Gresham, I. W. Kim and K. M. Zurek, Phys. Rev. D 83, 114027 (2011).