Measurement of the $t\bar{t}$ production cross-section using $e\mu$ events with b-tagged jets in $\mathit{pp}$ collisions at $\sqrt{s}=13$ TeV with the ATLAS detector

Contents lists available atScienceDirect

Physics Letters B

www.elsevier.com/locate/physletb

Measurement of the t ¯ t production cross-section using e μ events with b-tagged jets in pp collisions at √

s = ^{13 TeV with the ATLAS detector}

.TheATLAS Collaboration^

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

Articlehistory:

Received9June2016

Receivedinrevisedform10August2016 Accepted10August2016

Availableonline16August2016 Editor:W.-D.Schlatter

Thispaperdescribesameasurementoftheinclusivetopquarkpairproductioncross-section(σtt¯)witha datasampleof3.2fb⁻¹ofproton–protoncollisionsatacentre-of-massenergyof√_s

=^{13TeV,collected} in 2015 by the ATLAS detector at the LHC. This measurement usesevents with an opposite-charge electron–muonpairinthefinalstate.Jetscontainingb-quarksare taggedusinganalgorithmbasedon trackimpactparametersandreconstructedsecondaryvertices.Thenumbersofeventswithexactlyone andexactlytwob-taggedjetsarecountedandusedtodeterminesimultaneouslyσ_t¯t andtheefficiency to reconstruct and b-tag ajet fromatop quark decay, therebyminimising theassociated systematic uncertainties.Thecross-sectionismeasuredtobe:

σ_t¯t=818±8(stat)±27(syst)±19(lumi)±12(beam)pb,

where thefouruncertainties arisefromdata statistics,experimentalandtheoreticalsystematiceffects, theintegratedluminosityandtheLHCbeamenergy,givingatotalrelativeuncertaintyof4.4%.Theresult isconsistentwiththeoreticalQCDcalculationsatnext-to-next-to-leadingorder.Afiducialmeasurement correspondingtotheexperimentalacceptanceoftheleptonsisalsopresented.

©^{2016TheAuthor(s).PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense} (http://creativecommons.org/licenses/by/4.0/).FundedbySCOAP³.

1. Introduction

Thetopquarkistheheaviestknownfundamentalparticle,with amassm_t whichismuchlargerthananyoftheotherquarks,and closetothescaleofelectroweaksymmetrybreaking.Thestudyof itsproductionanddecaypropertiesforms acorepartoftheLHC physicsprogramme.AttheLHC,topquarksareprimarilyproduced inquark–antiquarkpairs(tt),¯ andtheprecisepredictionofthecor- respondinginclusivecross-sectionissensitivetothegluon parton distributionfunction(PDF)andthetopquarkmass,andpresentsa substantialchallengeforQCDcalculationaltechniques.Physicsbe- yondtheStandardModelmayalsoleadtoanenhancementofthe tt production¯ rate.

Calculations of the t¯t production cross-section at hadron col- liders are available at full next-to-next-to-leading-order (NNLO) accuracyinthestrongcouplingconstant αS,includingtheresum- mation of next-to-next-to-leading logarithmic (NNLL) soft gluon terms [1–5]. In this paper a reference value of 832⁺₋⁴⁰₄₆ pb at a centre-of-mass energyof√

s=¹³TeV assuming m_t=¹⁷².5 GeV isused, correspondingto a relativeprecision of ⁺₋⁴₅^._.⁸₅%. Thisvalue wascalculatedusingthetop++ 2.0program[6].Thecombined

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

PDF and αS uncertainties of ±^{35 pb were calculated using the} PDF4LHCprescription[7]withtheMSTW200868%CLNNLO[8,9], CT10NNLO[10,11]andNNPDF2.35fFFN[12]PDFsets,andadded inquadraturetothefactorisationandrenormalisationscaleuncer- tainty of⁺₋²⁰₂₉ pb.Thecross-sectionat√

s=^{13TeV ispredictedto} be3.3timeslargerthanthecross-sectionat√

s=^8TeV.

Measurementsof σ_t¯t havebeenmadeat√

s=^{7 and8 TeV by} both ATLAS [13–15] and CMS [16–18]. The most precise ATLAS measurements of σtt¯ at these collision energies were made us- ing events with an opposite-charge isolated electron and muon pair and additional b-tagged jets [13]. This paper documents a measurement of σ_t¯t at √

s=^{13 TeV using the same final state} andanalysistechnique. Wherever possible,the analysisbuildson the studies and procedures used in the earlier publication [13].

A fiducial measurement determining the cross-section in the re- gion corresponding tothe experimental leptonacceptance isalso presented.

ThedataandMonteCarlosimulationsamplesare describedin Section 2,followedby theobjectandeventselection inSection3 andthemethodfordeterminingthett cross-section¯ inSection4.

The evaluation of backgrounds isdiscussed in Section 5and the systematicuncertaintiesinSection 6.Finally,theresultsandcon- clusionsaregiveninSection7.

http://dx.doi.org/10.1016/j.physletb.2016.08.019

0370-2693/©^{2016TheAuthor(s).PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense}(http://creativecommons.org/licenses/by/4.0/).Fundedby SCOAP³.

Table 1

SummaryofMonteCarlosamplesusedtomodelthesignalandbackgroundprocesses.The‘Calculation’

columncorrespondstotheorderofthematrixelementcalculationintheMonteCarlogenerator.

Process Generator+parton shower Calculation

tt¯ Powheg-Box v2+^Pythia6

NLO Powheg-Box v2+^Herwig++

Madgraph5_aMC@NLO+^Herwig++

W t single top Powheg-Box v1+^{Pythia6 NLO}

Powheg-Box v1+^Herwig++

Z+^{jets Sherpa2.1.1} NLO (up to two partons)

Diboson Sherpa2.1.1 NLO (up to two partons)

Powheg+^Pythia8 NLO

t-channel single top Powheg-Box v1+^Pythia6 NLO

W+^jets Powheg-Box v2+^{Pythia8 NLO}

tt¯+^W/Z MadGraph+^{Pythia8 LO}

2. Dataandsimulationsamples

The analysis is performed using the full 2015 proton–proton (pp)collisiondatasampleat√

s=^{13TeV with25 nsbunchspac-} ing recordedby the ATLAS detector[19,20]. Thedata correspond toanintegratedluminosity of3.2fb⁻¹ afterrequiringstableLHC beamsandthat all detectorsubsystems were operational. Events arerequired topass eithera single-electronor single-muontrig- ger,withthresholdssettobealmostfullyefficientforleptonswith transversemomentump_T>25GeV passingofflineselections.Each event includes the signals from on average about 14 additional inelasticpp collisionsinthesamebunchcrossing(knownaspile- up).

Monte Carlo simulated event samples are used to optimise the analysis, to compare to the data, and to evaluate signal and background efficiencies and uncertainties. The samples used in the analysis are summarised in Table 1. The main t¯t signal and backgroundsamples were processed through the ATLAS detector simulation [21] based on GEANT4 [22]. Some of the systematic uncertaintieswere studiedusing alternativet¯t samplesprocessed throughafastersimulationmakinguseofparameterisedshowers in the calorimeters [23]. Additional simulated pp collisions gen- erated with Pythia8.186 [24] were overlaid to model the effects fromadditionalcollisionsinthesameandnearbybunchcrossings.

Allsimulatedeventswereprocessedusingthesamereconstruction algorithms and analysischain as the data, andsmall corrections wereapplied toleptontrigger andreconstruction efficiencies and resolutionstoimprovetheagreementwiththeresponseobserved indata.

The baseline tt simulation¯ sample was produced at next-to- leading order (NLO) in QCD using the matrix-element genera- tor Powheg-Box v2 [25–27] with CT10 PDFs [10], interfaced to Pythia6[28]withthePerugia2012setoftunedparameters(tune) [29]forpartonshower, fragmentationandunderlyingeventmod- elling. The h_damp parameter, which gives a cutoff scale for the first gluon emission, was set to mt, a value which was chosen to give good modelling ofthe t¯t system p_T at √

s=^{7 TeV [30].}

The EvtGen[31]packagewasusedtobettersimulatethedecayof heavy-flavourhadrons.

Alternativet¯t simulationsamplesweregeneratedusing Powheg interfacedto Herwig++[32],and Madgraph5_aMC@NLO[33] in- terfaced to Herwig++. The effects of initial- and final-state ra- diation were explored using two alternative Powheg+^Pythia6 samples:one withhdamp set to 2mt, thefactorisation andrenor- malisation scale varied by a factor of 0.5 and using the Perugia 2012 radHitune,givingmorepartonshower radiation;andasec- ond one with the Perugia 2012radLo tune, hdamp=^mt and the factorisation and renormalisation scale varied by a factor of 2, givinglesspartonshower radiation. The sampleswere simulated

followingthe recommendationsdocumentedin Ref.[34].Thetop quark masswas setto172.5 GeV inall thesesimulationsamples andthet→W b branchingfractionto100%.

Backgroundsinthismeasurementareclassifiedintotwotypes:

thosewithtworealpromptleptonsfromW or Z decays(includ- ing those produced via leptonic decays of τ-leptons), and those whereatleastoneofthereconstructedleptoncandidatesis‘fake’, i.e. anon-promptlepton producedfromthedecayofa bottomor charmhadron,anelectronarisingfromaphotonconversion,ajet misidentifiedasanelectron,oramuonproducedfromanin-flight decayofa pionorkaon.Backgroundscontainingtwo realprompt leptonsincludesingle-topproductioninassociationwithaW bo- son (W t), Z+jets production with Z→τ τ _→eμ,and diboson production(W W ,W Z and Z Z )wherebothbosonsdecaylepton- ically.

The dominant W t single-top backgroundwas modelled using Powheg-Boxv1+^{Pythia6 with the CT10 PDFs and the Perugia} 2012 tune, using the ‘diagram removal’ generation scheme [35].

The Z+jets background was modelled using Sherpa 2.1.1 [36]:

matrix elements (ME) were calculated for up to two partons at NLO andfour partonsatleading order usingthe Comix[37] and OpenLoops [38] matrix-element generators andmerged with the Sherpapartonshower(PS)usingthe ME+^{PS@NLO[39]prescrip-} tion; the CT10 PDF set was used in conjunction withdedicated parton shower tuning in Sherpa. Diboson production with addi- tional jetswas alsosimulated using Sherpa 2.1.1 andCT10 PDFs asdescribedabove;thefour-leptonfinalstate,thethree-leptonfi- nal state with two different-flavour leptons, and the two-lepton finalstateweresimulatedtocoverZ Z , Z W andW W production, andincludeoff-shell Z/γ^∗ contributions.Same-charge W W pro- ductionfromQCDandelectroweakprocesseswas included.Alter- native W t anddibosonsimulationsamples weregenerated using Powheg+^{Herwig++and Powheg}+^Pythia8,respectively,toesti- matethebackgroundmodellinguncertainties.

Themajorityofthebackgroundwithatleastonefakeleptonin the selectedsample arisesfromtt production¯ where onlyone of theW bosonsfromthetopquarksdecaysleptonically,whichwas simulatedasdiscussedearlier.Otherprocesseswithone reallep- tonwhichcancontributetothisbackgroundincludethet-channel single-topproduction, modelledusing Powheg-Boxv1+^Pythia6, and W+^{jets with the W decayingto e}ν, μν or τ ν where the

τ-lepton subsequently decays leptonically. This background was modelled using Powheg-Boxv2+^{Pythia8 with the CT10 PDFs.}

The smallexpectedcontributionfromtt in¯ association witha W or Z boson to the same-charge eμ sample used for background estimation was modelled using MadGraph+^{Pythia8 [40].Other} backgrounds, including processeswith two misidentified leptons, arenegligible.

3. Objectandeventselection

Thismeasurementmakesuseofreconstructedelectrons,muons andb-tagged jets. The object and eventselections largely follow thoseusedintheearlierpublication;inparticularthesamekine- matic cutsare usedfor electrons andjets, andvery similar ones areusedformuons.

Electroncandidates are reconstructed from an isolated elec- tromagnetic calorimeter energy deposit matched to a track in the inner detector and passing a medium likelihood-based re- quirement[41,42],withinthefiducialregion oftransverse energy ET>25 GeV and pseudorapidity¹ |η|<2.47. Candidates within thetransition regionbetween thebarreland endcapelectromag- netic calorimeters, 1.37<|η_|<1.52, are removed. The electron candidatesmustsatisfyrequirementsonthetransverseimpactpa- rameter significance calculated with respect to the beamline of

|^d0|/σd₀<5 and onthelongitudinal impactparametercalculated withrespectto theprimaryvertexof| ^z0 sin θ|<0.5 mm.The primaryvertexisdefinedastheonewiththehighestsumofp²_T of tracksassociatedtoit. Electronsarerequiredtobe isolatedusing requirementsonthecalorimeterenergyinaconeofsize R<0.2 around the electron (excludingthe depositfrom the electron it- self)divided by the electron p_T,and on the sum of track p_T in a variable-size cone around the electron direction (again exclud- ingtheelectrontrackitself).Thetrackisolationconesizeisgiven by the smaller of R=^{10 GeV}/pT(e) and R=⁰.2, i.e. acone which increases insize atlow p_T up to a maximum of0.2. Se- lectioncriteria,dependent on pT and η,areapplied toproducea nominalefficiencyof95% forelectrons from Z→^{ee decayswith} pT of 25 GeV which rises to 99% at 60 GeV. The efficiencies in tt events¯ are smaller,duetotheincreasedjetactivity.Toprevent double-countingofelectronenergydepositsasjets,theclosestjet with R<0.2 of a reconstructed electron isremoved. Finally, if thenearestjetsurvivingtheaboveselectioniswithin R=⁰.4 of theelectron, the electronis discarded, toensure it issufficiently separatedfromnearbyjetactivity.

Muon candidates are reconstructed by combining matching tracks reconstructed in both the inner detector andmuon spec- trometer, and are required to satisfy pT>25 GeV and |η|<2.4 [43]. Muons are also requiredto be isolated, usingrequirements similar to those for electrons, with the selection criteria tuned to give similar efficiencies for Z→μμ events. The muon can- didates must satisfy the requirements on the transverse impact parameter significanceandon the longitudinal impactparameter of|^d0|/σd0<3 and| ^z0sinθ|<0.5 mm,respectively. Toreduce thebackgroundfrommuonsfromheavy-flavourdecaysinsidejets, muonsareremoved iftheyare separatedfromthe nearestjet by R<0.4. However, if this jet has fewer than three associated tracks,themuoniskeptandthejetisremovedinstead;thisavoids an inefficiency for high-energy muonsundergoing significant en- ergylossinthecalorimeter.

Jets arereconstructedusingtheanti-kt algorithm [44,45]with radius parameter R =⁰.4, starting from topological clusters of deposited energy in the calorimeters. Jets are calibrated using anenergy- and η-dependentsimulation-basedcalibrationscheme withcorrectionsderivedfromdata.Nocorrectionsforsemileptonic b-hadrondecaysareapplied. Jetsare acceptedwithinthefiducial

1 ATLASusesaright-handedcoordinatesystemwithitsoriginatthenominalin- teractionpoint(IP)inthecentreofthedetectorandthez-axisalongthebeampipe.

Thex-axispointsfromtheIPtothecentreoftheLHCring,andthey-axispoints upwards.Cylindricalcoordinates(r,φ)areusedinthe transverseplane,φ being theazimuthalanglearoundthez-axis.Thepseudorapidityisdefinedintermsof thepolarangleθas η= −^{ln tan}(θ/2).Angulardistanceismeasuredinunitsof R≡

( η)²+ ( φ)^2.

region pT>25GeV and|η|<2.5.Toreducethecontributionfrom jetsassociated withpile-up,jetswith pT<50GeV and|η_|<2.4 arerequiredtopassapile-uprejectionveto[46].

Jets are b-tagged as likely to contain b-hadrons using the MV2c20 algorithm [47], a multivariate discriminant making use of track impact parameters and reconstructed secondary vertices andtuned withthe newdetectorconfiguration,i.e. including the Insertable B-Layer detector (IBL) [20]. Jets are defined as being b-tagged if the MV2c20 weight is larger than a threshold value correspondingtoapproximately70%b-taggingefficiencyforb-jets intt events,¯ althoughtheexactefficiencyvarieswith pT.Insimu- lation,thetaggingalgorithmgivesa rejectionfactorofabout440 against light-quarkandgluonjets, andabout8againstjetsorigi- nating fromcharmquarks.The improvementsofafactorofthree inthe light-quarkrejectionandof60%inthe charm-quarkrejec- tion comparedto theb-taggingalgorithm usedinRef. [13] origi- natefromthegainintrackimpactparameterresolutionfromthe IBL, andimprovements in thetrack reconstruction andb-tagging algorithms[47].

Events arerejectedifthe selectedelectronandmuonaresep- arated by φ <0.15rad and θ <0.15 rad,where φ and θ are the differences in polar and azimuthal angles between the two leptons. This requirement rejects events where a muon un- dergoessignificantenergylossintheelectromagneticcalorimeter, thusleading toareconstructedelectroncandidate.Events passing the aboverequirements,andhavingexactly oneselected electron andone selected muon ofopposite electriccharge sign (OS), de- finetheeμpreselectedsample.Thecorrespondingsame-sign(SS) sample isusedintheestimationofbackgroundfromeventswith misidentifiedleptons. Eventsarethenfurtherclassifiedintothose withexactlyoneorexactlytwob-taggedjets.

4. Extractionofthett cross-section¯

The tt cross-section¯ ismeasuredin thedileptoniceμ channel, where onetop quark decays ast→^{W b}→^eνb and theother as t→^{W b}→μνb.² Thefinalstatesfromleptonic τ decaysarealso included.As inRef.[13], σ_t¯t isdeterminedbycountingthenum- bersofopposite-signeμeventswithexactlyone(N1)andexactly two (N₂) b-tagged jets, ignoring any jets that are not b-tagged which maybe present, dueto e.g. light-quarkorgluon jetsfrom QCD radiation or b-jets from top quark decays which are not b-tagged.Thetwoeventcountscanbeexpressedas:

N₁=^Lσ_t¯teμ2b(1−^Cbb)+^N₁^bkg

N2=^Lσ_t¯teμC_bb²₊N^bkg₂ (1) where L is the integrated luminosity ofthe sample and eμ the efficiency for a t¯t event to pass the opposite-sign eμ preselec- tion. The combinedprobability for a jet fromthe quark q inthe t→^{W q decay to fall within the acceptance of the detector, be} reconstructed as a jet with transverse momentum above the se- lection threshold, and be tagged as a b-jet, is denoted by b. If the decays of the two top quarks and the subsequent recon- struction of the two b-tagged jets are completely independent, the probability to tag both b-jets bb is given by bb=_b². In practice, small correlationsare present forkinematic andinstru- mental reasons,andtheseare takenintoaccount viathe tagging correlation coefficient C_b, definedas C_b=bb/_b² orequivalently Cb=^4Nte^t¯μ N^t2^¯t/(N^t₁^¯t+^2Nt2^t¯)², where N^t_e^¯tμ isthe number ofprese- lected eμ t¯t events and N₁^t¯t and N^t₂^¯t are the numbers of events

2 This notationindicatesthe leptonicdecay ofbotht andt.¯ Charge-conjugate modesareimpliedunlessotherwisestated.