Single top measurements and the | V tb | extraction at the LHC
SSingle top measurements and the | V tb | extraction atthe LHC Abideh Jafari for the CMS and ATLAS collaborations
Universit´e catholique de Louvain, Louvain-la-Neuve, BelgiumE-mail: [email protected]
Abstract.
The CMS and ATLAS experiments have performed detailed studies on theelectroweakly produced top quarks at the LHC. These studies range from accurate measurementsof the cross section and | V tb | in different production modes to search for new interactions inthe tWb vertex. Moreover, different properties of the top quark are precisely measured in thiscontext. All measurements are consistent with the standard model and no sign of new physicsis observed.
1. Introduction
At the LHC [1], the top quark is mainly produced in pairs via the strong interaction. To lesserextent, it is produced individually through the electroweak interaction including the tWb. The t -channel process is the dominant single top production mode.The W-associated production (tW),recently observed at the LHC, occurs with a moderate rate whereas the s -channel process is sorare that it has been only possible to set an upper limit for its cross section. In this article wereport part of the latest LHC results on single top quark analyses. The analyses are performedusing the proton-proton collisions recorded by the CMS [2] and ATLAS [3] experiments at 7 and8 TeV center-of-mass energies.
2. The t -channel cross section measurements The event signature of the t -channel single top in its leptonic final state contains a chargedlepton (e or µ ), missing transverse energy ( (cid:54) E T ) and two jets where one is originating from a bquark (b-tagged jet) and the other from the spectator quark. This selection is hereafter referredto as 2J1T. Main backgrounds to this final state are t¯t, W boson produced in association withjets (W+jets) and QCD multijet events.At 7 TeV center-of-mass energy, ATLAS has performed a comprehensive analysis [4] using4 . − of the LHC data. For the signal, the 2J1T and 3J1T event samples are dividedinto two categories based on the lepton charge. An additional selection on the lepton-jettopology is applied to reject the QCD background. The modelings of simulated backgroundsare validated in control regions with less stringent b tagging criteria. The presence of b jets,as defined in the signal region, is vetoed. The b tagging efficiency is controlled in a 3J2Tregion. A neural network output is fitted to data simultaneously in the four signal categories toextract the t -channel cross section. The QCD multijet background is estimated using data-driven techniques while other background expectations are derived using simulation. Thesystematic uncertainties are evaluated with pseudo-experiments. The measurement, with the a r X i v : . [ h e p - e x ] D ec otal uncertainty, yields σ tq = 46 ± σ ¯tq = 23 ± m t = 172 . R t = 2 . ± .
18. The jet energy scale (JES) is thedominant systematic uncertainty for the cross sections while for R t , the parton distributionfunction (PDF) uncertainty is the largest one. Figure 1 shows the R t measurement comparedto predictions using different PDFs. The total cross section, 68 ± | V tb | value, | V tb | = 1 . ± . | V tb | (cid:29) | V td | , | V ts | . A lower limit of | V tb | > .
88 is obtained under the assumptionof | V tb | ≤
1. These results are comparable with the earlier CMS single top measurement at7 TeV [5].A fiducial t -channel cross section measurement is provided by the ATLAS collaboration using20 . − of the 8 TeV data [6]. A binned maximum likelihood fit is performed to the neuralnetwork output in the 2J1T region where backgrounds are treated as nuisance parameters.The t¯t and W+jets modelings are validated in the 2J2T and 2J0T regions, respectively.The fiducial phase space is defined close to that of the reconstructed and selected data set.The particle-level objects are constructed from stable particles in the final state, with a verysimilar definition to the reconstructed objects. The fiducial cross section within the detectoracceptance is measured to be σ fid = 3 . ± .
05 (stat . ) ± .
47 (syst . ) ± .
09 (lumi . ) pb. Systematicuncertainties are obtained from pseudo-experiments with the dominant contributions from JESand signal generator. The fiducial measurement is extrapolated to the full phase space usingdifferent Monte Carlo generators (Fig. 1). The inclusive cross section is determined to be σ t = 82 . ± . . ) ± . . ) ± . ± . . ), using aMCatNLO+ Herwig . Thisleads to | V tb | = 0 . +0 . − . with a lower bound of | V tb | > .
78 at 95% confidence level (CL) when | V tb | ≤ t R NNPDF 2.3MSTW2008 (68% CL)HERAPDF 1.5GJR08 (VF)CT10 (+ D0 W asym.)CT10ABM11 (5 flav.)Measurement resultNNPDF 2.3MSTW2008 (68% CL)HERAPDF 1.5GJR08 (VF)CT10 (+ D0 W asym.)CT10ABM11 (5 flav.) sys. ⊕ stat. stat. =7 TeVs dt = 4.59 fb L ∫ ATLAS )t( -ch. t s (t)/ -ch. t s = -ch. t R1 1.2 1.4 1.6 1.8 2 2.2 -1 = 8 TeV, L = 19.7 fbsCMS, CT10wCT10ABM11MSTW2008NNPDF 2.3HERAPDFCMS 0.19 (syst.) – – [pb] t σ
40 50 60 70 80 90 100
NLO+NNLL (MSTW2008)=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 acceptance correction from:Data corrected with Preliminary
ATLAS =8 TeVs dt = 20.3 fb L ∫ Figure 1.
The top-anti-top cross section ratio measurement by ATLAS (top left) at 7 TeV [4]and by CMS (top right) at 8 TeV [7] together with the extrapolated ATLAS fiducial cross sectionwith different generators at 8 TeV (bottom) [6].The CMS experiment has exploited the discriminating feature of the pseudorapidity of thenon-b-tagged jet ( η j (cid:48) ) to measure the t -channel cross section at 8 TeV [7]. Events with onecharged lepton (e or µ ) in the 2J1T sample are split into two categories based on the leptonharge. Each event category is further divided into a signal region (SR) with a top quarkmass of 130 < m (cid:96) b ν <
220 GeV, and two side bands (SB) with m (cid:96) b ν out of this range.The QCD multijet events are rejected with a selection on (cid:54) E T and the transverse mass ofthe W boson ( m WT ) in the electron and muon channels, respectively. Using a data sampleequivalent to 19 . − , a template fit is performed on the η j (cid:48) distribution in SR with backgroundstreated as constrained nuisance parameters. Data events in SB are used to extract the W+jetstemplate. The shape for t¯t is also corrected, using data-driven correction factors from a 3J2Tcontrol region. The shape and normalization of QCD multijets are determined with other datadriven techniques. Those systematic uncertainties that are not marginalized in the fit, areevaluated with pseudo-experiments. The cross sections for the top quark and top antiquarkproduction are σ t = 53 . ± . ± . σ ¯t = 27 . ± . ± . σ tot = 83 . ± . ± . R t = 1 . ± .
10 (stat . ) ± .
19 (syst . ), where PDF as the largestuncertainty. Figure 1 shows the CMS R t measurement compared with different PDF predictions.The inclusive cross section at 8 TeV is found to be larger than the one at 7 TeV by R / =1 . ± .
08 (stat) ± .
12 (syst . ). The | V tb | measurements at 7 TeV and 8 TeV are combined andresult in | f LV V tb | = 0 . ± .
038 (exp . ) ± . . ) where f LV accounts for deviations from theSM in the tWb coupling (see Sec. 5).
3. The cross section measurements of W -associated production The tW production has been observed in CMS in the dilepton final state, using 12 . − ofthe 8 TeV data [10]. A maximum likelihood fit is performed on a BDT output over all leptonflavor combinations (ee, e µ and µµ ) and in three signal regions (1J1T, 2J1T and 2J2T). Theshapes for signal and backgrounds are taken from simulation. Systematic uncertainties aretreated as constrained nuisance parameters in the fit except for the luminosity and theoryuncertainties which are unconstrained. The scale uncertainty and the t¯t modeling are thedominant systematic uncertainties. A significance of 6 . σ is observed for the signal, to becompared with the expected values of 5 . ± . σ . This corresponds to σ tW = 23 . ± . | V tb | = 1 . ± .
12 (exp . ) ± .
04 (theo . ). Constraining | V tb | ≤ | V tb | > .
78 at 95% CL.A similar measurement is carried out in ATLAS, in the e µ final state, using 20 . − ofthe 8 TeV data [11]. A fit to the BDT output is simultaneously performed in the 1J1T and2J ≥
1T regions. Templates for signal and backgrounds are taken from simulation and systematicuncertainties evaluated by means of pseudo-experiments with the main contribution comingfrom the tW and t¯t modeling. The measurement has reached to an observed significance of4 . σ where 4 . σ was expected. This corresponds to σ tW = 27 . ± . . ) ± . . ) pb and | f LV V tb | = 1 . ± .
12 (exp . ) ± .
03 (theo . ). Assuming | f LV | = 1 and | V tb | ≤ | V tb | > . σ tW = 25 . ± . . ) ± . . ) ± . . ) pb for the tW cross section. The combinedmeasurement for the V tb CKM matrix element is found to be | f LV V tb | = 1 . ± .
11 where a lowerbound of | V tb | > .
79 is obtained at 95% CL, under the assumption of | f LV | = 1 and | V tb | ≤
4. The s -channel cross section measurements The s -channel signal is characterized by one charged lepton, (cid:54) E T and two b tagged jets. Using19 . − of the 8 TeV data, CMS has set an observed upper limit of σ s − ch < . s -hannel cross section [13]. A BDT discriminant is constructed in the 2J2T (signal) and 3J2T (t¯tbackground) regions. The signal strength is extracted using a likelihood fit to the BDT outputswhere the t¯t and W+jets backgrounds are constrained in the fit. The QCD multijet backgroundis determined with a similar method to the t -channel analysis. Systematic uncertainties areevaluated using pseudo-experiments. The scale uncertainty is the dominant contribution to thetotal uncertainty. Figure 2 summarizes the LHC single top cross section measurements with7 TeV and 8 TeV data samples. ]Ve [Ts7 7.5 8 8.5 9 9.5 [ pb ] s
10 CMS Preliminary TOPLHCWG + ATLASSingle top-quark production
Oct 2014 t - c hanne l W t p r o d . s - c hanne l arXiv:1406.7844 VeT ATLAS t-channel 7
035 (2012) JHEP12
VeT CMS t-channel 7
142 (2012) PLB716
VeT ATLAS Wt prod. 7
VeT CMS Wt prod. 7 C.L. , 95%VeT ATLAS s-channel 7
ATLAS-CONF-2011-118 ATLAS-CONF-2014-007
VeT ATLAS t-channel 8
090 (2014) JHEP06
VeT CMS t-channel 8
ATLAS-CONF-2013-100
VeT ATLAS Wt prod. 8
VeT CMS Wt prod. 8 C.L. , 95%VeT ATLAS s-channel 8 arXiv:1410.0647
C.L. , 95%VeT CMS s-channel 8
CMS-PAS-TOP-13-009
VeT LHC Wt combination, 8
ATLAS-CONF-2014-052, CMS-PAS-TOP-14-009091503, (2011) 83 PRD
NNLL + NLO s a ¯ PDF ¯ scale , MSTW2008nnloVeG = 172.5 top m Figure 2.
The LHC summary plotfor single-top cross section. The s -channel ATLAS result at 8 TeV ap-peared after this presentation, hencenot included in the report.
5. Search for anomalous tWb couplings
Deviations from SM in the tWb vertex can be expressed in terms of the anomalous couplings, f LV , f RV , f LT and f RT , presented in this Lagrangian: L anom . tWb = − g √ γ µ ( f LV P L + f RV P R )tW − µ − g √ iσ µν q ν m W ( f LT P L + f RT P R )tW − µ + H . C ., (1)in which q is the difference of the top and bottom quark 4-momenta. A direct search for theanomalous couplings is carried out in CMS using 5 . − of 7 TeV data. Signal events withone charged lepton (e or µ ) are categorized into sub-samples, namely: 2J1T, 3J1T and 3J2T.QCD multijet events are rejected by imposing a lower bound on the output of a dedicatedBayesian neural network (BNN). The t¯t modeling is validated in a 4J2T sample where 2J0Tand 3J0T samples are used to verify the W+jets modeling. A ”SM BNN” is constructed todiscriminate between the SM t -channel and SM backgrounds whereas an ”atWb BNN” is trainedfor the anomalous hypothetic scenarios against all SM processes. The anomalous scenarios aresimulated for ( f LV , f LT ) and ( f LV , f RV ) combinations where those couplings that are not present inthe combination are set to zero. For each combination the two BNN discriminants are used asinputs in the statistical analysis. The observed (expected) limits at 95% CL are | f LV | > . . | f LT | < . .
06) for ( f LV , f LT ), and | f LV | > . .
88) and | f RV | < . .
39) for ( f LV , f RV ).
6. Top quark properties in production and decay
In the SM, the top quark is highly polarized in the direction of the spectator quark in the t -channel process. Its spin is also correlated with the angular properties of the decay products. Theangle between the charged lepton and non-b-tagged jet 3-momenta in the top quark rest frame( θ ∗ ) is used in CMS to measure the top quark polarization [8]. The data sample correspondsto 20 fb − and the event selection is similar to the CMS t -channel cross section measurement at TeV with an additional threshold on a boosted decision tree (BDT) output to purify the signalsample. The background-subtracted distribution of cos θ ∗ is then unfolded to particle level. Afit to the unfolded distribution leads to a spin asymmetry of A l = 0 . ± .
07 (stat . ) ± .
15 (syst . )for the muon channel and A l = 0 . ± .
11 (stat . ) ± .
23 (syst . ) for the electron channel. Thetop quark polarization, A l = 0 . ± .
12 (stat . ) ± .
32 (syst . ), is extracted from the combinationof the two measurements using the BLUE method.The W helicity fractions are sensitive to the anomalous couplings in Eq. 1. The helicityangle θ ∗ (cid:96) is defined as the angle between the W boson momentum in the top quark restframe and the momentum of the down-type decay fermion in the rest frame of the W boson.The functional form of the top quark partial decay width can be written as ρ (cos θ ∗ (cid:96) | (cid:126)F ),representing the contributions from the right-handed ( F R ), left-handed ( F L ) and longitudinal( F ) helicity fractions of the W boson. A reweighting method, built on ρ (cos θ ∗ (cid:96) | (cid:126)F ), is employedto measure the W helicity fractions [9], with the same data sample and event selection asin Ref. [7]. All events containing a t → (cid:96) b ν interaction, including t¯t and other single topprocesses, are considered as signal, reweighted and added to backgrounds. The resultingcos θ ∗ (cid:96) distribution is fitted to the data to extract, simultaneously, the W helicity fractions andthe W+jets background contamination. The combination of the muon and electron channelsyields F L = 0 . ± .
028 (stat . ) ± .
032 (syst . ), F = 0 . ± .
039 (stat . ) ± .
037 (syst . ) and F R = − . ± .
019 (stat . ) ± .
011 (syst . ), all consistent with the SM predictions. The measuredfractions are used to set limits on the f LT and f RT anomalous couplings. Acknowledgement
The author would like to thank the ATLAS and CMS collaboration for their incredible workin single top quark studies. Also thanks to FNRS (Fond National de la Recherche Scientifique)from Belgium who financially supported the author for this conference.
References [1] L. Evans and P. Bryant, LHC Machine,
JINST (2008) S08001, doi:10.1088/1748-0221/3/08/S08001.[2] CMS Collaboration, The CMS experiment at the CERN LHC, JINST (2008) S08004, doi:10.1088/1748-0221/3/08/S08004.[3] ATLAS Collaboration, The ATLAS Experiment at the CERN Large Hadron Collider, JINST (2008) S08003,doi: 10.1088/1748-0221/3/08/S08003.[4] ATLAS Collaboration, Comprehensive measurements of t -channel single top-quark production cross sectionsat √ s =7 TeV with the ATLAS detector, (2014), arXiv:1406.7844.[5] CMS Collaboration, Measurement of the t -channel single top quark production cross section in pp collisionsat √ s =7 TeV, JHEP (2012) 035, doi:10.1007/JHEP12(2012)035, arXiv:1209.4533.[6] ATLAS Collaboration, Measurement of the Inclusive and Fiducial Cross-Section of Single Top-Quark t -Channel Events in pp Collisions at √ s =8 TeV, (2014), ATLAS-CONF-2014-007.[7] CMS Collaboration, Measurement of the t -channel single-top-quark production cross section and of the V tb CKM matrix element in pp collisions at √ s =8 TeV, JHEP (2014) 090, doi:10.1007/JHEP06(2014)090.[8] CMS Collaboration, Measurement of top-quark polarization in t -channel single-top production, (2013), CMS-PAS-TOP-13-001.[9] CMS Collaboration, Measurement of the W boson helicity in events with a single reconstructed top quark inpp collisions at √ s =8 TeV, (2014), arXiv:1410.1154.[10] CMS Collaboration, Observation of the Associated Production of a Single Top Quark and a W Boson inpp Collisions at √ s =8 TeV, Phys. Rev. Lett. (2014) 231802, doi:10.1103/PhysRevLett.112.231802,arXiv:1401.2942.[11] ATLAS Collaborations, Measurement of the cross-section for associated production of a top quark and a Wboson at √ s =8 TeV with the ATLAS detector, (2013), ATLAS-CONF-2013-100.[12] ATLAS and CMS Collaborations, Combination of cross-section measurements of associated production of asingle top-quark and a W boson at √ s =8 TeV with the ATLAS and CMS experiments, (2014), ATLAS-CONF-2014-052, CMS-PAS-TOP-14-009.[13] CMS Collaboration, Search for s -channel single top-quark production in pp collisions at √ ss