Evidence for single top-quark production in the s -channel in proton–proton collisions at $\sqrt{s}=8$ TeV with the ATLAS detector using the Matrix Element Method

(1)

Contents lists available atScienceDirect

Physics Letters B

www.elsevier.com/locate/physletb

Evidence for single top-quark production in the s-channel in proton–proton collisions at √

s = ^{8 TeV with} ^the ^ATLAS ^detector ^using the Matrix Element Method

.ATLASCollaboration

a r t i c l e i n f o a b s t ra c t

Articlehistory:

Received20November2015

Receivedinrevisedform17February2016 Accepted3March2016

Availableonline8March2016 Editor:W.-D.Schlatter

This Letter presents evidence for single top-quark production in the s-channel using proton–proton collisions at a centre-of-mass energy of 8 TeV with the ATLAS detector at the CERN Large Hadron Collider.The analysis isperformedoneventscontainingone isolatedelectronormuon,large missing transversemomentumandexactlytwob-taggedjetsinthefinalstate.Theanalyseddatasetcorresponds to an integrated luminosity of 20.3 fb⁻¹. The signal is extracted using amaximum-likelihood fitof a discriminant which is based on the matrix element method and optimized in order to separate single-top-quark s-channel events fromthe main background contributions, whichare top-quark pair production and W boson production in association with heavy-flavour jets. The measurementleads to an observed signal significance of 3.2 standard deviations and a measured cross-section ofσs= 4.8±⁰.8(stat.)⁺₋¹₁^._.⁶₃(syst.) pb, whichis consistent with the Standard Model expectation. The expected significancefortheanalysisis3.9standarddeviations.

©²⁰¹⁶^CERN^for^the^benefitôf^theÂTLASCollaboration.PublishedbyElsevierB.V.Thisisanopen accessarticleundertheCCBYlicense(http://creativecommons.org/licenses/by/4.0/).FundedbySCOAP³.

1. Introduction

In proton–proton (pp) collisions, top quarks are produced mainlyin pairs viathe stronginteraction, butalsosingly viathe electroweak interaction through a Wtb vertex. Therefore, single top-quark production provides a powerful probe for the elec- troweakcouplings ofthe topquark. In theStandard Model(SM), three different production mechanisms are possible in leading- order(LO) QCD: an exchangeof a virtual W boson eitherinthe t-channel orin the s-channel(see Fig. 1), orthe associated production ofa top quark anda W boson. Among other interesting features,s-channelsingletop-quarkproductionissensitivetonew particles proposed in several models of physics beyond the SM, such aschargedHiggsbosonor W boson production[1].Italso plays an important role in indirectsearches fornew phenomena that could be modelled as anomalous couplings in an effective quantumﬁeld theory [2].Furthermore, s-channelproduction,like theothertwoproductionchannels,providesadirectdetermination of the absolutevalue of the Cabibbo–Kobayashi–Maskawa (CKM) matrixelement V_tb.

Singletop-quarkproductionwasﬁrstseenby theCDFandDØ collaborations in combined measurements of the s-channel and t-channel [3,4]. Recently, the s-channel alone was observed in a

E-mailaddress:atlas.publications@cern.ch.

Fig. 1. Feynmandiagram inleading-order QCDfor thedominanthard scattering processinthes-channelofsingletop-quarkproduction.

combination of the results from both collaborations [5]. At the LargeHadronCollider(LHC)[6]theproductionofsingletopquarks was observed both in the t-channel and in associated W t pro- ductionby theCMS[7,8] andATLAScollaborations[9,10].Forthe s-channel, results ofa search at√

s=8 TeV using an integrated luminosityof20.3 fb⁻¹werepublishedbyATLAS[11].Thatanaly- siswasbasedonaboosteddecisiontree(BDT)eventclassifierand led toan upperlimit of14.6 pb at the95% confidencelevel.The obtainedcross-sectionwas σ_s^BDT=⁵.0±⁴.3 pb withan observed signalsignificanceof1.3σ.

Standard Modelpredictionsareavailable fortheproductionof single top quarksinnext-to-leading-order (NLO)QCD [12–14]in- cludingresummednext-to-next-to-leadinglogarithmic(NNLL)corrections for soft gluon emissions [15–17]. For the s-channel the predictedtotalinclusivecross-sectionforpp collisionsatacentre- http://dx.doi.org/10.1016/j.physletb.2016.03.017

0370-2693/©²⁰¹⁶^CERN^for^the^benefitôf^theÂTLASCollaboration.PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense (http://creativecommons.org/licenses/by/4.0/).FundedbySCOAP³.

(2)

of-massenergy√

s=^{8 TeV is} σ_s^th=⁵.61±⁰.22 pb,whileforthe t-channelitis σ_t^th=⁸⁷.76⁺₋³₁^._.⁴⁴₉₁pb,and σ_{W t}^th =²².37±¹.52 pb for associated W t production. The given uncertainties include varia- tions of the renormalization and factorization scales, as well as anestimate oftheuncertaintyofthepartondistributionfunction (PDF)neededforthecalculation.

In this Letter, a measurement of single top-quark s-channel production in pp collisions with √

s=^{8 TeV at} ^the ^LHC ^is ^pre- sented. Each of the two other single-top-quark production pro- cesses,t-channel and W t production, is treatedasa background processassumingitscross-sectionaspredictedbyNLO+^NNLL^QCD calculations. In the SM the top quark decays almost exclusively into a W boson anda b-quark. This analysisconsiders only the leptonicdecays(e or μ)oftheW boson,sincethefullyhadronic ﬁnalstatesaredominatedbyoverwhelmingmulti-jetbackground.

Someoftheeventscontaining a W bosondecayingintoa τ lepton which subsequently decays leptonically are also selected. At LO the ﬁnal state contains two jets with large transverse mo- menta: one jet originating fromthe decay of the top quark into a b-quark (“b-jet”), and another b-jet from the Wtb vertex pro- ducingthetopquark. Thustheexperimentalsignature consistsof anisolatedelectronormuon,largemissingtransversemomentum, E^miss_T , dueto the undetected neutrino fromthe W boson decay, andtwojetswithlargetransversemomentum, pT,andwhichare bothidentiﬁedascontainingb-hadrons (“b-tagged”).Theelectron andmuon channelsinthisanalysis aremerged regardless ofthe leptonchargeinordertomeasurethecombinedproductioncross- sectionoftopquarksandtopantiquarks.

IncontrasttotheaforementionedBDT-basedanalysis[11],the signal extractionin this analysisis basedon the matrix element (ME) method [18,19]. The same data set is used in both analy- ses.Thisanalysistakesadvantageofenhancedsimulationsamples whichreduce thestatisticaluncertaintyandgiveabetterdescrip- tion of the data. Furthermore,updated calibrations for the 2012 dataareused,resultinginareductionofsystematicuncertainties.

The event selection is improved by adding a veto on dileptonic events,whichleadstoasigniﬁcantsuppressionofthebackground fortop-quarkpair(t¯t)production(seeSection5).Thecombination ofall thesemeasures resultsin asigniﬁcant improvement inthe sensitivitytothes-channelprocess.Approximatelyhalfofthisim- provementcan be attributedto thechange in methodfrom BDT toME.Inparticular,theBDT technique appliedtothisanalysisis limitedbythesamplesizesavailableforthetraining,whiletheME approachisnotsensitivetothislimitation.

2. TheATLASdetector

TheATLASdetector[20]isamulti-purposedetectorconsisting ofa tracking system, calorimeters andan outer muon spectrometer. The inner tracking system contains a silicon pixel detector, a silicon microstrip tracker anda straw-tube transition radiation tracker.Thesystemissurroundedbyathinsolenoidmagnetwhich produces a 2 T axial magnetic ﬁeld, and it provides charged- particletracking as well as particle identiﬁcationin the pseudorapidity¹ region |η|<2.5. The central calorimetersystem covers therange of|η_|<1.7 and isdivided into a liquid-argonelectro- magneticsamplingcalorimeterwithhighgranularityandahadron calorimeterconsistingofsteel/scintillatortiles.Theendcapregions

1 ATLASusesaright-handedcoordinatesystemwithitsoriginatthe nominal interactionpoint(IP)inthecentreofthedetectorandthez-axisalongthebeam pipe.Thex-axispointsfromtheIPtothecentreoftheLHCring,andthe y-axis pointsupward.Cylindricalcoordinates(r,φ)areusedinthe transverseplane,φ beingtheazimuthalanglearoundthebeampipe.Thepseudorapidityisdeﬁnedin termsofthepolarangleθasη= −^{ln tan}(θ/2).

are equipped with liquid-argon calorimeters for electromagnetic and hadronic energy measurements up to |η_|₌4.9. The outer muon spectrometerisimmersedina toroidalmagneticﬁeld pro- vided by air-core superconducting magnets and comprises tracking chambers forprecise muon momentum measurements up to

|η_|₌2.7 and trigger chamberscovering the range |η_|<2.4.The combinationofallthesesystemsprovideseﬃcientandprecisere- constructionofleptonsandphotonsintherange|η_|<2.5.Jetsand E^miss_T arereconstructedusingenergydeposits overthefull cover- ageofthecalorimeters,|η|<4.9.Athree-leveltriggersystem[21]

isusedtoreduce therecordedrateofuninterestingeventsandto selecttheeventsinquestion.

3. Dataandsimulationsamples

Thedataforthisanalysiswascollected withthe ATLASdetec- torattheLHCin2012atacentre-of-mass energyof8 TeV using single-electronorsingle-muontriggers.Theappliedtriggerthresh- olds ensure a constant eﬃciency with respect to the energy of the lepton candidates used in thisanalysis. Each triggered event includes on average about 20 additional pp collisions (pile-up) fromthesamebunch-crossing.Onlyeventsrecordedunderstable beamconditionsareselectedandalleventshavetopassstringent dataqualityrequirementsandneedtocontainatleastonerecon- structed primary vertex withat least ﬁve associated tracks. The datausedbythisanalysiscorrespondstoanintegratedluminosity of20.3±⁰.6 fb⁻¹ [22].

The samples used for the simulation of the single-top-quark s-channel signal events, as well as the ones for the tt,¯ single- top-quark t-channel and W t backgrounds, were produced using the NLO generator Powheg-Box (v1_r2129) [23] with the CT10 PDFs[24].Thepartonshower, hadronizationandunderlyingevent were simulatedwith Pythia (v6.42) [25]usingthePerugia 2011C set of tuned parameters [26]. For generator, parton shower and fragmentation modeling studies, alternative simulation samples are employed. In caseof the s-channel signal and the t-channel backgroundthe MadGraph5_aMC@NLO(v2.0)generatorwas used [27], while for the tt and¯ W t backgrounds it was the MC@NLO (v4.03) [28] generator. In both cases the CT10 PDFs were used andthegeneratorswereinterfacedto Herwig (v6.52)[29]forpar- ton showeringand hadronization,and Jimmy (v4.31) [30] for the underlyingevent.Theimpactsofscalevariationsaswellasuncer- tainties onthe initial andﬁnal state radiation (ISR/FSR) insignal events were studied usingsamples generated with the Powheg- Boxgenerator,againconnectedto Pythia,forvariousvaluesofthe factorizationandrenormalizationscales.

TheprocessesforW bosonproductioninassociationwithjets (W+^jets)^were^modelled^by^the^LOmulti-parton Sherpa generator (v1.4.1)[31] togetherwithCT10PDF sets.Thisgeneratormatches the partonshower to themulti-leg LOmatrix elements by using the CKKW method[32]. The Sherpa generator was used for the complete event generation including the underlying event, using thedefaultsetoftunedparameters.Thebackgroundcontributions from Z boson production in association withjets ( Z+^jets) ^were simulatedusingtheLO Alpgen (v2.14)generator[33]coupledwith Pythia(v6.42)andCTEQ6L1PDFsets[34].Thelatterisalsousedto testtheW+jets modeling.Thedibosonprocesses(W W ,W Z ,Z Z ) were simulatedusingthe Herwig (v6.52)and Jimmy (v4.31) gen- eratorswith theAUET2 tune [35] andthe CTEQ6L1 PDF set.The single-boson anddiboson sampleswere normalized to their production cross-sectionscalculated at next-to-next-to-leading order (NNLO)[36] andNLO[37],respectively.Themulti-jet background wasmodelledbyadata-drivenmethodasdescribedinSection4.

Almost all the generated eventsamples were passed through the full ATLAS detector simulation [38] based on Geant4 [39]

(3)

and then processed with the same reconstruction chain as the data.Theremainingsamples,whichconsistofsingle-top-quarks- andt-channel samples for scale variation studies, aswell as for t-channel and t¯t modeling studies, are passed through the ATL- FAST2 simulationof the ATLAS detector,which uses a fastsimu- lationforthe calorimeters [40].The simulated eventswere over- laid withadditional minimum-biaseventsgenerated with Pythia to simulatetheeffect ofadditional pp interactions. Allprocesses involving top quarks were generated using a top-quark mass of 172.5 GeV.

4. Backgroundestimation

Thetwo mostimportantbackgroundsare tt and¯ W+^{jets pro-} duction.Theformerisdifficulttodistinguishfromthesignalsince tt events¯ contain real top-quark decays. In its dileptonic decay mode, tt events¯ can mimic the final-state signature of the signal if one of the two leptons escapes unidentified, whereas the semileptonic decay mode contributes to the selected samples if onlytwoofitsfourjetsareidentifiedorifsomejetsare merged.

The W+jets eventscancontributetothebackgroundifthey con- tainb-jetsintheﬁnalstateorduetomis-taggingofjetscontaining other quark ﬂavours. Single top-quarkt-channel production also leads toa sizeablebackgroundcontribution,while associated W t productionhasonlyasmalleffect.

Alesssignificant backgroundcontributionismulti-jet produc- tionwherejets,non-promptleptonsfromheavy-flavourdecays,or electrons from photon conversions are mis-identified as prompt isolated leptons. This background is estimated by using a data- driven matrixmethod[41],where theprobability to mis-identify an isolated electron ormuon inan eventis obtainedby exploit- ingsumrulesbasedondisjointcontrolsamples,one almostpure electronormuonsampleandanothercontainingahighfractionof mis-identified leptons due to a relaxed lepton-isolationcriterion.

For both decay channels the amount of multi-jet background is below2%intheﬁnalselection.Otherminorbackgroundsarefrom Z+^{jets and}^dibosonproduction.

Apart from the data-driven multi-jet background, all samples are normalizedto their predictedcross-sections. The samples for single top-quark production are normalized to their NLO+^NNLL predictions(seeSection1),whileforalltt samples¯ arecentcalcu- lationwithTop++(v2.0)atNNLOinQCDincludingresummations ofNNLLsoftgluontermsof σ_t^th_¯_t =²⁵³⁺₋¹³₁₅ ^{pb is}^used^for^the^nor- malization [42–47].

5. Eventreconstructionandselection

Fortheselectionofs-channelﬁnalstates,asinglehigh-pT lepton,eitherelectronormuon,exactlytwob-taggedjetsandalarge amountofE^miss_T arerequired.

Electrons are reconstructed asenergy deposits in the electromagnetic calorimeter matched to charged-particle tracks in the inner detector and must pass tight identiﬁcation requirements [48,49]. The transverse momentum of the electrons must satisfy p_T>30 GeV and be in the central region with pseudorapidity

|η|<2.47,excludingtheregion 1.37<|η|<1.52,whichcontains alargeamountofinactivematerial.Muoncandidatesareidentiﬁed usingcombinedinformationfromtheinnerdetectorandthemuon spectrometer[50,51].Theyarerequiredto have p_T>30 GeV and

|η_|<2.5.Boththeelectronsandmuonsmustfulﬁll additionaliso- lationrequirements, asdescribed in Ref. [41], in orderto reduce contributions from non-prompt leptons originating from hadron decays,andfakeleptons.

Jetsare reconstructedbyusingtheanti-kt algorithm[52] with aradiusparameterof0.4 forcalorimeterenergyclusterscalibrated

withthe localcluster weighting method[53]. Forthejet calibra- tion an energy- and η-dependent simulation-based scheme with in-situ corrections based on data [54] is employed. Only events containing exactly two jetswith p_T>40 GeV forthe leading jet and p_T>30 GeV for thesecond leading jet,aswell as|η|<2.5 for both jets are selected. Events involving additional jets with pT>25 GeV and |η_|<4.5 are rejected. Both jetsmust be iden- tifiedasb-jets.Theidentificationisperformedusinganeuralnet- work which combines spatial and lifetimeinformation fromsec- ondary vertices of tracks associated with the jets. The operating point of the tagging algorithm used in this analysiscorresponds to a b-tagging efficiency of 70% and a rejection factor for light- flavour jetsofabout140,whiletherejectionfactorforcharmjets isaround5[55,56].

The missingtransverse momentumiscomputedfromthevec- tor sumofall clustersof energydepositsin thecalorimeterthat are associatedwithreconstructedobjects,andthetransversemo- mentaofthereconstructedmuons[57,58].Theenergydepositsare calibratedatthe correspondingenergyscale oftheparentobject.

Since E^miss_T is a measure for the undetectable neutrino originating from the top-quark decay, in this analysis only events with E^miss_T >35 GeV areaccepted.Furthermore,thetransversemass²of the W boson, m_T^W, needs to be larger than 30 GeV tosuppress multi-jetbackground.

The main background at thisstage of the selection originates from top-quark pair production, which is in turn dominated by dileptonict¯t events.Toreducethisbackground,avetoisappliedto alleventscontaininganadditionalreconstructedelectronormuon identiﬁed withloose criteria. Theminimum required p_T ofthese leptons is 5 GeV. By this measure the tt background¯ is dimin- ished by 30% while reducing the signal by less than half a per cent.Aftertheapplicationofall eventselectioncriteria, asignal- to-background ratio of 4.6% is reached. The event yields for all samplesinthesignalregionarecollectedinTable 1.

Apartfromthesignal region,two moreregionsare definedto validate themodeling,onevalidationregionfort¯t productionand acontrol regionforthe W+jets background.Thelatterisusedto constrainthenormalizationoftheW+jets backgroundinthefinal signalextraction,asexplainedinmoredetailinSection8.Thetwo regions are definedin thesame wayasthe signal region,except thatneitherthevetooneventswithadditionaljetsnortheoneon dileptoneventsareapplied.Top-quarkpairproductionisenriched byselectingeventscontainingexactlyfourjetswithp_T>25 GeV, twoofthemb-taggedatthe70%workingpoint.The W+^{jets con-} trolregionisdefinedusingalessstringentb-tagrequirement(80%

workingpoint);inordertoensurethatthisregionisdisjointfrom the signal region, it is requiredthat atleast one ofthe two jets failstomeetthesignalregionb-taggingcriteriaatthe70%workingpoint.

6. Matrixelementmethod

The ME method directly uses theoretical calculationsto com- pute a per-event signal probability. This technique was used for theobservationofsingletop-quarkproductionattheTevatron[59, 3,4]. The discrimination betweensignal and backgroundis based on the computation of likelihood values P(^X|^Hproc) for the hy- pothesis that a measured event with ﬁnal state X is of a cer- tain process type Hproc. Those likelihoods can be computed by

2 Thetransversemass,m^W_T ,iscomputedfromtheleptontransversemomentum, p_T,andthedifferenceinazimuthalangle,Δφ,betweentheleptonandthemissing transversemomentumasm^W_T =

2E^miss_T p_T

1−cos(Δφ( ˆE_T^miss,pˆ_T)) .

Evidence for single top-quark production in the s -channel in proton–proton collisions at $\sqrt{s}=8$ TeV with the ATLAS detector using the Matrix Element Method

Physics Letters B

Evidence for single top-quark production in the s-channel in proton–proton collisions at √

s = 8 TeV with the ATLAS detector using the Matrix Element Method

s = ^{8 TeV with} ^the ^ATLAS ^detector ^using the Matrix Element Method