Search for high-mass dilepton resonances using 139 fb Of Collision Data Collected at TeV with the ATLAS detector

Article

Reference

Search for high-mass dilepton resonances using 139 fb Of Collision Data Collected at TeV with the ATLAS detector

ATLAS Collaboration

ADORNI BRACCESI CHIASSI, Sofia (Collab.), et al.

Abstract

A search for high-mass dielectron and dimuon resonances in the mass range of 250GeV to 6TeV is presented. The data were recorded by the ATLAS experiment in proton–proton collisions at a centre-of-mass energy of √s = 13 TeV during Run 2 of the Large Hadron Collider and correspond to an integrated luminosity of 139 fb$^{−1}$. A functional form is fitted to the dilepton invariant-mass distribution to model the contribution from background processes, and a generic signal shape is used to determine the significance of observed deviations from this background estimate. No significant deviation is observed and upper limits are placed at the 95% confidence level on the fiducial cross-section times branching ratio for various resonance width hypotheses. The derived limits are shown to be applicable to spin-0, spin-1 and spin-2 signal hypotheses. For a set of benchmark models, the limits are converted into lower limits on the resonance mass and reach 4.5TeV for the E$_{6}$-motivated Zψ′ boson. Also presented are limits on Heavy Vector Triplet model couplings.

ATLAS Collaboration, ADORNI BRACCESI CHIASSI, Sofia (Collab.), et al . Search for

high-mass dilepton resonances using 139 fb Of Collision Data Collected at TeV with the ATLAS detector. Physics Letters. B , 2019, vol. 796, p. 68-87

DOI : 10.1016/j.physletb.2019.07.016

Available at:

http://archive-ouverte.unige.ch/unige:128780

Disclaimer: layout of this document may differ from the published version.

1 / 1

Contents lists available atScienceDirect

Physics Letters B

www.elsevier.com/locate/physletb

Search for high-mass dilepton resonances using 139 fb

of pp collision data collected 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:

Received18March2019

Receivedinrevisedform23May2019 Accepted5July2019

Availableonline10July2019 Editor:M.Doser

A search for high-massdielectron and dimuonresonances inthe mass range of250 GeV to 6 TeV is presented.ThedatawererecordedbytheATLASexperimentinproton–protoncollisionsatacentre-of- massenergyof√

s=13 TeV duringRun2oftheLargeHadronColliderandcorrespondtoanintegrated luminosity of139 fb⁻¹.Afunctionalformisfittedtothedileptoninvariant-massdistributiontomodel the contribution from background processes, and a generic signal shape is used to determine the significance ofobserveddeviationsfromthisbackground estimate.Nosignificantdeviationisobserved and upper limits areplacedatthe95% confidencelevel onthe fiducialcross-sectiontimesbranching ratioforvariousresonancewidth hypotheses.Thederivedlimits areshowntobeapplicabletospin-0, spin-1andspin-2signalhypotheses.Forasetofbenchmarkmodels,thelimitsareconvertedintolower limitsontheresonancemassandreach4.5 TeV fortheE6-motivatedZ_ψ boson.Alsopresentedarelimits onHeavyVectorTripletmodelcouplings.

©^{2019TheAuthor.PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense} (http://creativecommons.org/licenses/by/4.0/).FundedbySCOAP³.

1. Introduction

Searches in the dilepton (dielectron and dimuon) final state havealongandillustrioushistorywiththediscovery ofthe J/ψ mesonin 1974 [1,2] and ϒ mesonin 1977 [3] aswell as the Z bosonin1983 [4,5].Asthesewerekeystepswhichledtothees- tablishment of the Standard Model (SM) of particle physics, the studyofthesamefinalstatecouldhelptopavethewaytoabet- terunderstandingofthephysicsprocessesbeyondit.

Variousmodels predictresonances whichdecayintodileptons andcanbe categorised accordingtotheir spin. Anewhigh-mass spin-0 resonance, H, introduced as part of an extended scalar sector insome models, such asthe Minimal SupersymmetricSM (MSSM) [6],hashigherdecayrateintoapairofmuonsratherthan electrons.Themajorityofsearchesfornewneutralhigh-massres- onanceshavefocusedonanewspin-1vectorboson,generallyre- ferredtoasZ,thatappearsinmodelswithextendedgaugesym- metries. Typical benchmark models include the Sequential Stan- dardModelZ_SSMboson [7],whichhasthesamefermioncouplings as the SM Z boson, a Zχ and a Z_ψ boson of an E₆-motivated GrandUnification model [8], or a Z_HVT bosonof the Heavy Vec- tor Triplet model [9]. In the first two models, the Z boson is a singlet,associatedwithanewU(1)gaugegroup,andgenerallyits couplingstotheSM W andZ bosonsareassumedtobezero.The

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

Z_HVT bosonisaneutralmemberofanewSU(2)gaugegroup,i.e.

partofatripletandcannotexistwithouttwonewchargedheavy bosons, W_HVT ^± , with which it is nearly degenerate in mass. New spin-2 resonances, excited states of the graviton, are introduced in theRandall–Sundrum model [10] with a warpedextra dimen- sion. Inexperimental terms the described scenarios wouldresult inalocalexcessofsignalcandidatesoverasmoothlyfallingdilep- tonmassspectrum.Thissearchhasacleanexperimentalsignature withafullyreconstructablefinalstateandexcellentdetectioneffi- ciency.

This Letter presents a search for a new resonance decaying into twoelectrons ortwomuonsin 139 fb⁻¹ ofdatacollected in proton–proton (pp) collisions at theLHC at acentre-of-mass en- ergy√

s=^{13 TeV.Previoussearcheswith36.1 fb−1 of ppcollision} data at √

s=13 TeV conducted by the ATLAS and CMS experi- ments [11,12] showednosignificantexcessandledtolowerlimits of upto 3.8 TeV forthe massofthe Z_ψ boson.The analysispre- sented in this Letter, compared with that published in Ref. [11], benefits from: a factor of four increase in integratedluminosity;

severalimprovementsinthereconstructionsoftware,includingthe use of a new dynamical, topological cell-clusteringalgorithm for electron reconstruction [13] and an improved treatment of the relativealignmentoftheinnertrackerandthemuontrackingde- tectorsinthemuonreconstruction;theuseofinvariant-massside- bandsoftheexpectedsignalindatatoconstrainthefitparameters of the background distribution, which is described by a smooth functional form instead of relying on simulation; and a generic https://doi.org/10.1016/j.physletb.2019.07.016

0370-2693/©^{2019TheAuthor.PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense}(http://creativecommons.org/licenses/by/4.0/).Fundedby SCOAP³.

Table 1

Theeventgeneratorsusedforsimulationofthesignalandbackgroundprocesses.TheacronymsMEandPSstandformatrixelement andpartonshower.Thetop-quarkmassissetto172.5 GeV.

ME Generator and ME PDFs PS and non-perturbative effect with PDFs Background process

NLO Drell–Yan Powheg-Box[23,24], CT10 [25],Photos Pythiav8.186 [26], CTEQ6L1 [27,28],EvtGen1.2.0 tt¯ Powheg-Box, NNPDF3.0NLO [29] Pythiav8.230, NNPDF23LO [30],EvtGen1.6.0 Single tops-channel,W t Powheg-Box, NNPDF3.0NLO Pythiav8.230, NNPDF23LO,EvtGen1.6.0 Single topt-channel Powheg-Box, NNPDF3.04fNLO,MadSpin Pythiav8.230, NNPDF23LO,EvtGen1.6.0 Diboson (W W,W ZandZ Z) Sherpa2.1.1 [31], CT10 Sherpa2.1.1, CT10

Signal process

LO Drell–Yan Pythiav8.186, NNPDF23LO Pythiav8.186, NNPDF23LO,EvtGen1.2.0 Randall–SundrumG^∗→ ^Pythiav8.210, NNPDF23LO Pythiav8.210, NNPDF23LO,EvtGen1.2.0 MSSMgg→H→ ^Powheg-Box, CT10 Pythiav8.212, CTEQ6L1,EvtGen1.2.0

signallineshapedescribedbyanon-relativisticBreit–Wignerfunc- tionconvolvedwiththedetectorresolution,whichsimplifiesrein- terpretationsoftheresult.

2. ATLASdetector

ATLAS [14–16] is a multipurpose detector with a forward–

backwardsymmetriccylindricalgeometrywithrespecttotheLHC beamaxis.¹ The innermostlayers consist oftracking detectorsin thepseudorapidityrange|η_|<2.5.Thisinner detector(ID)issur- roundedby a thinsuperconducting solenoid that provides a 2 T axial magnetic field. It is enclosed by the electromagnetic and hadroniccalorimeters,whichcover|η_|<4.9.Theoutermostlayers ofATLAS consist ofan external muon spectrometer (MS) within

|η_|<2.7, incorporating three large toroidal magnetic assemblies witheightcoils each. Thefield integral of thetoroidsranges be- tween2.0and6.0 Tmformostoftheacceptance.TheMSincludes precision tracking chambers and fast detectors for triggering. A two-level trigger system [17] reduces the recorded event rateto anaverageof1 kHz.

3. Dataandsimulation

The dataset used in this analysis was collected during LHC Run 2instablebeamconditionsandwithalldetectorsystemsop- eratingnormally.Theeventqualitywascheckedtoremoveevents with noise bursts or coherent noise in the calorimeters. Events in the dielectron channel were recorded using a dielectron trig- gerbasedonthe ‘veryloose’or ‘loose’identification criteria [17]

withtransverseenergy(ET)thresholdsbetween12and24 GeV for bothelectrons,dependingonthedata-takingperiod.Eventsinthe dimuonchannel are required to passat least one oftwo single- muontriggers:thefirstrequiresatransversemomentum(pT)ofat least50 GeV,whilethesecondhasathresholdloweredto26 GeV butrequiresthemuoncandidatetobeisolated [17].Theintegrated luminosity of the dataset is determined to be 139.0±².4 fb⁻¹, followingamethodologysimilar tothat detailedinRef. [18], and usingthe LUCID-2 detectorfor the baseline luminosity measure- ments [19], from calibration of the luminosity scale using x-y beam-separationscans.

While the search in this analysis is carried out entirely in a data-drivenway,simulatedeventsamplesforthesignalandback-

1 ATLASusesaright-handedcoordinatesystemwithitsoriginatthe nominal interactionpoint(IP)inthecentreofthedetectorandthez-axisalongthebeam pipe.Thex-axispointsfromtheIPtothecentreoftheLHCring,andthe y-axis pointsupwards.Cylindricalcoordinates(r,φ)areusedinthetransverseplane,φ beingtheazimuthalanglearoundthez-axis.Thepseudorapidityisdefinedinterms ofthepolarangleθasη= −^{ln tan}(θ/2).Angulardistanceismeasuredinunitsof R≡

(η)²+(φ)².

groundprocesses are used todetermine appropriate functions to fit the data, study background compositionsand to evaluate the signalefficiency.Themainbackgroundsindecreasingorderofim- portance areDrell–Yan (DY), top-quarkpair(tt¯), single-top-quark anddiboson production.Multi-jet and W+jets processes in the dielectronchannel areestimatedwithadata-drivenmethod [11].

Multi-jetandW+jets processesinthedimuonchannelaswellas processeswith τ-leptonsinbothchannelshaveanegligibleimpact andarenotconsidered.TheMonteCarlo(MC)eventgeneratorsfor the hard-scatterprocess, showeringandpartondistribution func- tions(PDFs)arelistedinTable1.The‘afterburner’generatorssuch asPhotos[20] forthefinal-statephotonradiation(FSR)modelling, MadSpin [21] to preserve top-quark spin correlations, and Evt- Gen [22], used forthe modellingof c- and b-hadron decays,are alsoreported.

The DY [32] and diboson [33] samplesare generatedin slices ofdileptonmasstoincrease thesamplesizeinthehigh-massre- gion.Next-to-next-to-leading-order(NNLO)correctionsinquantum chromodynamic (QCD) theory and next-to-leading-order (NLO) corrections in electroweak (EW) theory, are calculated and ap- pliedto the DY events.Thecorrections arecomputedwith VRAP v0.9 [34] andtheCT14NNLOPDFset [35] inthecaseofQCDef- fectswhereastheyarecomputedwithMCSANC[36] inthecaseof quantum electrodynamic effects dueto initial state radiation, in- terference between initial andfinal state radiation,and Sudakov logarithm single-loopcorrections. The top-quarksamples [37] are normalisedtothecross-sectionscalculatedatNNLOinQCDinclud- ing resummation of the next-to-next-to-leading logarithmic soft gluontermsasprovidedbyTop++2.0 [38].

Spin-1 signal templates are obtained by a matrix-element reweighting [11] of theleading-order (LO) DY samples generated in slices of dilepton mass. These signal templates are used only for cross-section and efficiency calculations. The relative natural width(Z/m_Z) forthebenchmark modelsconsideredvariesbe- tween 0.5%for Z_ψ and3% for Z_SSM .Interference effectsbetween the resonant signal andthe background processes are neglected.

Higher-order QCD correctionsfor all the spin-1 signals are com- puted withthe samemethodologyasfortheDY background.For the HVTmodel,thesecorrectionsare not applied, whichensures consistenttreatmentwiththeothersignalchannelsinaneventual combination, similar to that described in Ref. [39]. Electroweak correctionsarenotappliedtothesignalsamplesduetotheirlarge model dependence. Spin-0 signal efficiencies are obtained from samples of the MSSM gluon–gluon fusion production of a heavy Higgsbosondecayingintodileptonpairs,gg→^H→,produced in themass rangemH =400–1000 GeV and withrelative natural width(H/mH) varying betweenzeroand 20%.Spin-2 signal ef- ficiencies are obtained from Randall–Sundrum graviton G^∗→ samples produced in the mass range mG^∗ =750–5000 GeV and with coupling strengths, k/m_Pl, of 0.1, 0.2 and0.3, where k is a

scalethatdefinesthewarpfactoroftheextradimensionandm_Pl isthereducedPlanckmass.

Simulatedeventsamplesinclude theeffectof multiple pp in- teractions in the same or neighbouring bunch crossings. These effects are collectively referred to as pile-up. The simulation of pile-up collisions was performed with Pythia v8.186 using the ATLAS A3 set of tuned parameters [40] and the NNPDF23LO PDF set,and weighted to reproducethe average numberof pile- up interactions per bunch crossing observed in data. The gener- ated events were passed through a full detector simulation [41]

based on Geant 4 [42]. Spin-0 and spin-2 MC signal samples wereproducedwithafastparameterisationofthecalorimeterre- sponse [43].

Verylargegenerator-level-onlyMCsamples(withmorethan55 timesthe dataevents) forNLO DY events areused forthe back- groundstudiesdescribedinSection6.Thesesamplescouldnotbe producedwiththefulldetectorsimulationduetothelargenum- berofeventsrequired.

4. Eventselection

Theselection ofdilepton eventscloselyfollowsthat described inRef. [11].Aneventisselectedifatleastone ppinteractionver- texisreconstructed.Theprimaryvertexischosentobethevertex withthehighestsummedp²_T oftrackswithtransversemomentum p_T>0.5 GeV whichareassociatedwiththevertex.

Electron candidates are reconstructed from ID tracks that are matched to clusters of energy deposited in the electromagnetic calorimeter with energy deposition consistent with that of an electromagnetic shower [44]. Reconstructed electrons must have ET>30GeV,satisfy |η_|<2.47 in ordertopass throughthe fine- granularityregionoftheEMcalorimeter,andbeoutsidetherange 1.37<|η|<1.52 corresponding tothe transitionregion between thebarrelandendcapEMcalorimeters.Thecalorimetergranularity inthe excluded transitionregionis reduced,andthe presenceof significantadditionalinactivematerialdegradestheelectroniden- tificationcapabilitiesandenergyresolution.The‘medium’electron working point used for the final selection has an identification andreconstruction efficiency forprompt electrons above 92% for ET>80GeV.

MuoncandidatesareidentifiedbymatchingIDtrackstotracks reconstructed in the MS [45]. Muon candidates must have p_T>

30GeV and |η|<2.5. To ensure optimal muon momentum res- olution at high pT, the ‘high pT’ identification working point is used.Itrequiresatleastthreehitsineach ofthreelayersofpre- cision tracking chambers in the MS, and specific regions of the MSwherethealignmentissuboptimalarevetoedasaprecaution.

Theserequirementsrejectabout80%(13%)ofthemuoncandidates in(outside)thebarrel–endcapoverlapregion,1.01<|η|<1.1.The muon‘high pT’workingpointhasan η-averagedefficiencyof69%

at1 TeV which decreases to 64% at 2.5 TeV dueto increased oc- casionalcatastrophic energylossathigh p_T.Additionally, a ‘good muon’ selection requires that the uncertainty in the charge-to- momentumratioofmuoncandidatesislessthana pT-dependent value. Thisselection is fullyefficientbelow 1 TeV, butintroduces anadditionalinefficiencyof7%at2.5 TeV.

Electron(muon)candidate tracksmust beconsistent withthe primary vertexboth along the beamline, where the longitudinal impactparameterz₀isrequiredtosatisfy|^z0sinθ|<0.5 mm,and in the transverse plane, where the transverse impact parameter significance |^d0/σ(d0)| ^{is required to be less than 5 (3). To re-} ducebackgroundfrommisidentifiedjetsaswellasfromlight- and heavy-flavourhadrondecaysinside jets,lepton candidatesarere- quiredto be isolated.Electrons mustpass the‘gradient’ isolation working point which targets an E_T-dependent value of the iso-

lation efficiency, uniformin η, using a combinationof trackand calorimeter isolation requirements [44]. Formuons, the summed scalar pT ofgood-qualitytrackswithpT >1 GeV originatingfrom the primary vertex within a cone of variable size² R around themuon,butexcludingthemuon-candidatetrackitself,mustbe less than 6% of the pT of the muon candidate.The efficiency of this selection is above 99% for both electrons and muons with p_T>60GeV. Corrections are applied to electron (muon) candi- dates to matchtheenergy (momentum)scale andresolutionbe- tween simulation and data. These corrections are derived in an energy independent way for electrons [46]. For muons, the cor- rectionisdeterminedasafunctionofpTupto300 GeV,fromafit to Z→μμdatawithtemplatesderived fromsimulation [45]. At hightransversemomentum,thecalibrationsaredominatedbycor- rectionsextractedfromalignmentstudies,usingspecialrunswith the toroidalmagnetic field off.Corrections tothe leptonefficien- cies in the simulation are derived fromthe data for electron ET (muon pT)up to150(200) GeV [44,45].Thesimulationisusedto extrapolatetohigherelectronE_T(muon p_T)andtostudysystem- aticeffects.

The events are required to contain at least two same-flavour leptons. If additional leptons are present in the event, the two same-flavour leptons with the largest E_T (p_T) in the electron (muon) channel are selected to form the dilepton pair. If two different-flavour pairs are found, the dielectron pair is kept, be- cause ofthebetter resolution andhigherefficiencyfor electrons.

A selected muon pair is required to be oppositely charged. For an electron pair, the opposite-charge requirement is not applied because of the higher probability of charge misidentification for high-E_T electrons. The reconstructed mass of the dilepton sys- tem afterthefull analysis selection,m,is requiredto be above 225 GeV to avoidthe Z boson peak region, whichcannot be de- scribedbythesameparameterisationasthehigh-masspartofthe dileptondistributions.

5. Reconstructeddileptonmassmodelling

The relative dilepton mass resolution is defined as (m − m^true )/m^true , wherem^true is thegenerated dileptonmassat Born level before FSR. The mass resolutionis parameterised asa sum ofa Gaussiandistribution,whichdescribesthedetectorresponse, and a Crystal Ball function composed of a secondary Gaussian distribution witha power-lawlow-mass tail,which accounts for bremsstrahlungeffects inthe dielectronchannel orforthe effect of poorly reconstructedmuons. The parameterisation ofthe rela- tive dileptonmassresolutionasafunctionofm^true isdetermined by a simultaneousfit ofthefunction described above toNLO DY MCevents.TheMCsampleisseparatedin200m^true binsofequal size onalogarithmic scaleinthe rangeof130 GeV to 6 TeV.This procedureisrepeatedtoevaluatetheuncertaintyonthefitparam- eters by shifting individually the lepton energy and momentum scaleandresolutionsbytheiruncertainties.

6. Signalandbackgroundmodelling

A resonant signal is searched forby fitting the data dilepton massdistribution.Thefitfunctionconsistsofasmooth functional formforthe background,andageneric signal shape.The generic signal shapes are constructed from non-relativistic Breit–Wigner functions of various widths convolved with the detector resolu- tion, obtainedasdescribed in theprevious section.The shape of

2 R has a maximum value of 0.3 and decreases as a function of pT as 10 GeV/pT[GeV].