TTop quark production at ATLAS and CMS
Luca Lista, on behalf of the ATLAS and CMS collaborations
INFN Sezione di Napoli, Comp. Univ. M. S. Angelo, via Cintia, 80126, Naples, Italy
A review of the main recent results on top quark production from the ATLAS and CMSexperiments is presented. Results on both electroweak single top quark production and strongtop pair production are presented.
Top quark production proceeds at hadron colliders via strong or electroweak processes. Theformer results in the production of a top-antitop pair, while the latter results in the productionof a single top quark or antiquark, and may proceed via a W exchange in the t or s channel, orin associated production with a W boson. Top-pair production at LHC is enhanced comparedto Tevatron thanks to the larger gluon-fusion contribution. While at Tevatron the channel withthe lowest cross section was the associated tW production, which has been only observed atLHC, at LHC the single-top s channel has the smallest cross section, and is the hardest channelto probe. Single top quark production in the t channel has been measured by ATLAS and CMS both at7 and 8 TeV. The most recent ATLAS analysis based on an integrated luminosity of 20 fb − collected at 8 TeV adopted a neural network discriminant to separate the t -channel signal fromthe backgrounds, using 14 discriminating variables. One electron or muon is required togetherwith two hardonic jets, one of which has been identified as stemming from a b quark. Themultijet background is the hardest to model in simulation, and has been determined from datausing a fit to the distribution of the transverse missing energy. ATLAS measures a fiducial crosssection within the acceptance corresponding to a kinematic selection specified in Ref. to be: σ fid. t -ch. = 3 . ± . ± . ± . , (1)in agreement with theory predictions from different Monte Carlo (MC) generators, as displayedin Fig. 1. The fiducial cross-section measurement can be extrapolated to the entire phase a r X i v : . [ h e p - e x ] M a y [pb] fidt σ =60 GeV µ AcerMC+Pythia6 =172.5 GeV µ AcerMC+Pythia6 2)+Pythia6 → Powheg(2 3)+Pythia8 → Powheg(2 3)+Pythia6 → Powheg(2 3)+Herwig → aMC@NLO(2 result ATLAS
Predicted fiducial cross-section: =60 GeV µ AcerMC+Pythia6 =172.5 GeV µ AcerMC+Pythia6 2)+Pythia6 → Powheg(2 3)+Pythia8 → Powheg(2 3)+Pythia6 → Powheg(2 3)+Herwig → aMC@NLO(2 sys. ⊕ stat. stat. =8 TeVs -1 dt = 20.3 fb L ∫ Preliminary
ATLAS
Figure 1 – Comparison of ATLAS measurement of fiducial single-top production cross section with different MonteCarlo generators. space using acceptance estimates from generators. Assuming the prediction of the aMC@NLO generator plus Herwig parton shower , the inclusive single-top production cross section is: σ t -ch. = 82 . ± . ± . ± . ± . , (2)in agreement with next-to-leading order (NLO) theory predictions. CMS measured the single-top production cross section in the t channel from a fit to the distribution of the pseudorapidityof the light jet accompanying the top quark . The shapes of the distributions for the W+jets andthe t¯t backgrounds are determined from control samples in data. Events are selected requiringthe presence of one electron or muon together with tho jets, one of which is compatible with ab jet, and the reconstructed top-quark mass must be in the range 130 −
220 GeV. The inclusiveproduction cross section determined by CMS is: σ t -ch. = 83 . ± . ± . . (3)The result is in agreement with NLO theory predictions, as shown in Fig. 2, together with aprevious CMS measurement at 7 TeV. CMS also determined the ratio of cross sections at 8 and7 TeV to be: R / = 1 . ± . ± . , (4)and the ratio of top and antitop production cross section to be: σ t -ch. , t /σ t -ch. , ¯t = 1 . ± . ± . . (5)A measurement of the same ratio at 7 TeV by ATLAS gave : σ t -ch. (t) /σ t -ch. (¯t) = 1 . +0 . − . . (6)The measurement of associated tW production is limited by the large t¯t background. BothATLAS and CMS use multivariate techniques based on boosted decision trees (BDT) to extractthe tW signal. CMS analysis based on 12.2 fb − collected at 8 TeV delivered the first observationof tW production with a cross section of: σ tW = 24 . +5 . − . pb (7) [TeV]s1 2 3 4 5 6 7 8 9 10 - c hanne l t o t a l c r o ss s e c t i on [ pb ] t -1 CMS, L = 19.7 fb -1 CMS, L = 1.17/1.56 fb -1 D0, L = 9.7 fb -1 CDF, L = 3.2 fb
PDF) ¯ (scale – NLO QCD (5 flavour scheme) Campbell et al., JHEP 10 (2009) 042 PDF) ¯ (scale – NLO+NNLL QCD Kidonakis, Phys. Rev. D 83 (2011) 091503 -channel single-top-quark production t N L O + NN LL Q CD s / s [TeV]s Figure 2 – Comparison of CMS measurements of t -channel single-top production cross sections at 7 and 8 TeVwith next-to-leading order theory predictions. with a significance of 6.1 standard deviations (5.4 expected). ATLAS analysis based on 20.3 fb − has a precision similar to CMS : σ tW = 27 . ± . ± . , (8)and the quoted significance is 4.2 σ (4.0 expected).At LHC the single-top channel with the lowest production cross section is the s channel.CMS searched for the s -channel production in the entire sample at 8 TeV. No significant excessover the expected background was observed, and an upper limit on the production cross sectoinin the s channel was set at the 95% confidence level (CL): σ s -ch. < . . (9)This limit corresponds to 2.1 times the standard model (SM) cross section. The largest un-certainty on the above limit is due to the theory modelling of the large t¯t background. Theuse of more recent NLO MC generators to simulate this background is expected to improvethe precision of future measurements of this very rare channel. ATLAS published the limit on s -channel production at 7 TeV based on the first 0.7 fb − : σ s -ch. < . , (10)corresponding to 5.8 times the SM cross section value. | V tb | determination Single-top production allows a direct probe of the tWb copling at the production vertex. Inparticular, assuming that the branching fraction of the top quark to Wb is equal to one, thesquare root of the measured single-top production cross section, divided by its theory prediction,computed assuming | V tb | = 1, must be equal to | V tb | in the SM. Deviations from such value maybe indications of new physics that modifies the tWb vertex coupling. The available measurementsof single-top production in the t and tW channels allow determinations of | V tb | with differentlevels of precision. CMS also quoted a combination of the | V tb | measurements using the two t -channel cross section measurements at 7 and 8 TeV, which achieves the best present precision(4.1%). Table 1 reports the | V tb | determinations from the available measurements. able 1: | V tb | measurements from single-top production cross section. Experiment √ s Process | V tb | PrecisionATLAS 7 TeV t -ch. 1 . +0 . − . . +0 . − . t -ch. 0 . +0 . − . . ± . ± . t -ch. 1 . ± . ± . . +0 . − . (exp.) +0 . − . (th.) 14.8%8 TeV t -ch. 0 . ± . ± . . ± . ± . t -ch. 0 . ± . ± . Single-top production at LHC offers the opportunity to look for new physics that could affect thetgu, tgc, tZu or tZc vertices, exhibiting anomalous couplings, compared to the SM prediction.Such anomalous couplings would result in flavour-changing neutral currents (FCNC). ATLASstudied FCNC at the top production vertex , which would result in a single top quark producedwithout any associated particles. The study based on the 8 TeV data sample set limits toanomalous tgu and tgc coupling that can be translated into limits to the FCNC branchingfractions of the top quark: B (t → gu) < . × − , (11) B (t → gc) < . × − , (12)at the 95% CL. CMS looked for FCNC associated production of a top quark and a Z bo-son , which resulted in limits to anomalous tZu and tZc compling, translated into the followingbranching ratio limits at the 95% CL: B (t → Zu) < . × − , (13) B (t → Zc) < . . (14) The most precise determinations of t¯t production at LHC is performed using dileptonic events.CMS updated the analysis at 8 TeV using a data sample of 5.3 fb − , obtaining the followingcross section measurement : σ t¯t = 239 ± ± ± : σ t¯t = 237 . ± . ± . ± . . ± . p T values, as reported by ATLAS at 7 TeV. CMS reports at 8 TeV a - G e V t T dp σ d σ -4 -3 DataALPGEN+HERWIGMC@NLO+HERWIGPOWHEG+HERWIG
ATLAS Preliminary -1 L dt = 4.6 fb ∫ = 7 TeVs [GeV] Tt p0 100 200 300 400 500 600 700 800 D a t a M C Figure 3 – Differential t¯t cross section measurement by ATLAS as a function of the top-quark transverse momen-tum, compared with prediction of different Monte Carlo generators. deficit in the leading-order and next-to-leading-order MC predictions at low top-quark p T , whilethe approximate next-to-NLO predictions correctly describe the data. This is shown in Fig. 4. t¯t associated production The production of t¯t pairs associated with more particles, either jets or vector bosons, can bestudied at LHC exploiting the large integrated luminosity accumulated during the run at 7 and8 TeV. CMS measured the production of t¯t plus a b¯b pair , which is an important backgroundto ttH and new physics searches. Applying two different thresholds to the transverse momentaof the reconstructed jets, CMS measured the ratio of t¯tb¯b to t¯t plus two jets at 8 TeV: σ t¯tb¯b /σ t¯t jj = 0 . ± . ± . , p T >
20 GeV , (17) σ t¯tb¯b /σ t¯t jj = 0 . ± . ± . , p T >
40 GeV . (18)Both measurements are compatible with the prediction of MadGraph and Powheg gener-ators, though both have central values slightly larger than the SM predictions. ATLAS measuredat 7 TeV the incusive production cross section of t¯t plus a b or c quark : σ t¯t+b / c+ X /σ t¯t+ ≥ jet = 6 . ± . ± . , (19)assuming the jet has p T >
25 GeV and | η | < .
5. This result is also in agreement with the SMpredictions, using
AlpGen , though, as for the previous CMS measurement, the central valueis somewhat larger than the theory prediction.Measurements of t¯t associated with W or Z were performed by CMS at 7 TeV using twoanalyses: an inclusive search of t¯t plus either W or Z in events with same-sign lepton pairs(Fig. 5), and an exclusive analysis of events with three leptons, looking for t¯tZ. The results havea limited precision, due to the limited number of selected events: σ t¯t V = 0 . +0 . − . (stat.) +0 . − . (syst.) pb , (20) σ t¯tZ = 0 . +0 . − . (stat.) +0 . − . (syst.) pb . (21) eV tT p G e V t T dp σ d σ × DataMadGraphMC@NLO POWHEG Approx. NNLO = 8 TeVs at CMS Preliminary, 12.1 fb + Jets Combined µ e/ (arXiv:1205.3453) Figure 4 – Differential t¯t cross section measurement by CMS as a function of the top-quark transverse momentum,compared with prediction of different Monte Carlo generators and approximate next-to-next-to-leading-orderprediction.
The two measurements have statistical significance of 3.0 and 3.3 standard deviations respec-tively, and are in agreement with the SM predictions.ATLAS measured t¯t γ production at 7 TeV in events with a lepton and a jet . Photon fakerate is difficult to model in simulation, and has been determined from data using a fit to thedistribution of a variable describing the photon’s isolation (Fig. 6). The result is: σ t¯t γ = 2 . ± . ± . ± . , (22)assuming a photon transverse energy greater than 8 GeV. CMS updated the t¯t γ measurementat 8 TeV, and the result is: σ t¯t γ /σ t¯t = (1 . ± . ± . × − , (23)assuming a photon transverse energy greater than 20 GeV, and a separation between the photonand the b jet from the top-quark decay ∆ R > .
1, which can be translated into: σ t¯t γ = 2 . ± . ± . . (24)CMS searched for the production of two top-antitop quark pairs at 8 TeV. This process hasa cross section of the order of one fb in the SM , , but could be enhanced in several modelsbeyond the SM. The CMS analysis looks for the decay of one top quark with an electron ormuon, and the hadronic decay of the other three top-quark. In order to determine the three-jetcombinations that have large probability to come form a top quark decay, dedicated BDTs aretrained. A second BDT is used to select signal events adding event variables. The analysisshowed no excess over the expected background, and the following upper limit at 95% CL hasbeen set : σ t¯tt¯t <
63 fb . (25) e mm m e E v en t s Data + Ztt + WttNon-prompt / MisIDCharge MisIDWZRare SM
CMS = 7 TeVs at -1 L = 5.0 fb
Figure 5 – Expected and measured event yields in the CMS same-sign dilepton analysis for t¯t W , t¯t Z and thevarious backgrounds, superimposed. [GeV] cone20T p0 2 4 6 8 10 12 14 16 18 20 / b i n γ +jets µ data γ tt bkgtnon-t γ tbkg telectron fakeshadron fakes +jets µ data γ tt bkgtnon-t γ tbkg telectron fakeshadron fakes Preliminary
ATLAS -1 L dt = 1.04 fb ∫ Figure 6 – Result of the isolation fit for the single muon channel. The predicted ttγ signal is shown on top of thedifferent background contributions.
Conclusions
The precision of top production measurements at LHC is steadily improving, and the focus ofATLAS and CMS is shifting towards the precise understanding of top production mechanism inorder to perform detailed comparisons with state-of-the-art QCD predictions, and possibly finddeviations from the standard model. The determination of cross sections in fiducial regions isbeing adopted in order to avoid model-dependent extrapolations that would introduce theoryuncertainty. Next updates of LHC activities in top physics will involve both targeting theultimate precision for upcoming 7 and 8 TeV run-I legacy measurements, and the preparationfor run-II data at higher LHC energy.
References
1. ATLAS Collaboration, ATLAS-CONF-2014-007.2. J. Alwall et al. , JHEP , 1106 (2011).3. G. Corcella et al. , JHEP , 0101 (2001).4. CMS Collaboration, arXiv:1403.7366, sub. to JHEP.5. ATLAS Collaboration, ATLAS-CONF-2012-056.6. CMS Collaboration, arxiv:1401.2942 , sub. to Phys. Rev. Lett.7. ATLAS Collaboration, ATLAS-CONF-2013-100.8. ATLAS Collaboration, ATLAS-CONF-2013-063.9. CMS Collaboration, CMS PAS TOP-12-021.10. CMS Collaboration, JHEP , 024 (2014).11. ATLAS Collaboration, ATLAS-CONF-2013-097.12. ATLAS Collaboration, ATLAS-CONF-2013-099.13. CMS Collaboration, CMS-PAS TOP-12-027.14. CMS Collaboration, CMS-PAS-TOP-13-010.15. F. Maltoni and T. Stelzer, JHEP , 027 (2003).16. S. Alioli et al. , JHEP , 043 (2010).17. ATLAS Collaboration, Phys. Rev. D89 , 072012 (2014).18. M.L. Mangano et al.,
JHEP , 001 (2003).19. CMS Collaboration, Phys. Rev. Lett. , 1720 (2013).20. ATLAS Collaboration, ATLAS-CONF-2011-153.21. V. Barger et al. , Phys. Lett. B687 , 70 (2010).22. M.W.G. Bevilacqua, JHEP1207