Search for TeV-scale gravity signatures in high-mass final states with leptons and jets with the ATLAS detector at $\sqrt{s}=13$ TeV

(1)

Contents lists available atScienceDirect

Physics Letters B

www.elsevier.com/locate/physletb

Search for TeV-scale gravity signatures in high-mass ﬁnal states with leptons and jets with the ATLAS detector at √

s = ^{13 TeV}

.TheATLAS Collaboration

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

Articlehistory:

Received8June2016

Receivedinrevisedform7July2016 Accepted11July2016

Availableonline15July2016 Editor:W.-D.Schlatter

A search for physics beyond the Standard Model, in ﬁnal states with at least one high transverse momentumchargedlepton(electronormuon)andtwoadditionalhightransversemomentumleptonsor jets,isperformedusing3.2 fb⁻¹ofproton–protoncollisiondata recordedbytheATLASdetectoratthe LargeHadronColliderin2015at√

s=^{13 TeV.}^Theûpperêndôf^thedistributionofthescalarsumofthe transversemomentaofleptonsandjetsissensitivetotheproductionofhigh-massobjects.Noexcessof eventsbeyondStandardModelpredictionsisobserved.Exclusionlimitsaresetformodelsofmicroscopic blackholeswithtwotosixextradimensions.

©²⁰¹⁶^TheÂuthor(s).^Published^byÊlsevier^B.V.^Thisîsânôpenâccessârticleûnder^the^CC^BY^license (http://creativecommons.org/licenses/by/4.0/).FundedbySCOAP³.

1. Introduction

Models of TeV-scale gravity postulate that the fundamen- tal scale of gravity, MD, in a higher-dimensional space–time is muchlowerthanismeasuredinourfour-dimensionalspace–time.

In large extra-dimensional models (e.g. the model proposed by Arkani-Hamed,DimopoulosandDvali(ADD)[1,2])thereare n ad- ditional ﬂat extra dimensions, assumed to be compactiﬁed on a toruswithacommonradiusmuchlargerthan1/M_D.Anotherclass ofmodels (e.g.that ofRandall andSundrum(RS)[3,4])uses one extra dimension in a highly warpedanti-de-Sitter space. Both of these types of model can address the large difference between the scale of electroweak interactions, O(⁰.1 TeV), and that of gravity, the Planck scale, MPl=O(10¹⁶ TeV), in a natural way.

Interesting signatures are expected in these models in the form ofnon-perturbativegravitationalstatessuch asmicroscopicblack holes [5,6].Such ﬁnal statescould be produced inproton–proton (pp) interactions at the Large Hadron Collider (LHC) [7]. In the absenceofafulltheoryofquantumgravity,predictionsforproduc- tioncross-sectionsanddecaysofblackholesrelyonsemi-classical approximationswhichare expectedtobevalidifthemassofthe black hole is well above MD andalso higher than the Hawking temperature[8].Astrongriseintheproductionrateofsuchstates isexpectedwhen theenergyscaleoftheinteractions reachesthe order of M_D. Since the gravitational interaction couples to the

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

energy–momentum tensor rather than gauge quantum numbers, ﬁnalstatesareexpectedtobepopulated“democratically”, accord- ingtothenumberofavailableStandardModeldegreesoffreedom.

For this reason, it is expected that a significant fraction of final states would contain leptons. This search exploitsthis feature to enhancethesignalcontributionincomparisonwiththedominant background atthe LHC,which arisesfromquark andgluon scat- teringprocessesforminghadronicfinalstates.Finalstateswithat least three hightransverse momentum (p_T) objects are selected, of which at least one must be an electron or muon (leptons in what follows) and the others can be either leptons or hadronic jets. The discriminatingvariable usedin thissearch,

pT, isthe scalar sumof the transverse momenta of high pT objects in an event. Thesignalisexpectedtoappearathigh

pT.Searchesby ATLAS[9–12]andCMS[13,14]duringRun1oftheLHCdidnotre- veal anysigniﬁcant excessesoverexpectedbackgroundlevels. An ATLAS analysis [15] of Run-2 data at 13 TeV also found no evi- denceofneweffectsinmultijetﬁnalstates.Thisworkextendsthe reach of the analysisin Ref. [12], performedat a centre-of-mass energyof8 TeV,with3.2 fb⁻¹ ofdatarecordedby ATLASin2015 at13 TeV.Thissearchispotentiallysensitivetootherformsofnew physicsathigh-massandinvolvingtheelectroweaksector.

2. ATLASdetector

ATLAS[16]isamultipurposedetectorwithaforward–backward symmetric cylindrical geometry and nearly 4π coverage in solid http://dx.doi.org/10.1016/j.physletb.2016.07.030

0370-2693/©²⁰¹⁶^TheÂuthor(s).^Published^byÊlsevier^B.V.^Thisîsânôpenâccessârticleûnder^the^CC^BY^license(http://creativecommons.org/licenses/by/4.0/).Fundedby SCOAP³.

(2)

angle.¹ The inner detector(ID) utilises fine-granularity pixel and microstrip detectors over the pseudorapidity range |η_|<2.5 to provide precise track parameter and secondary vertex measure- ments.ForRun2oftheLHC,anewpixellayerhasbeenaddedata radiusof3.3 cm[17].Agas-filledstraw-tubetrackercomplements the silicontracker at larger radii. The tracking detectors are im- mersedina2 Tmagneticfieldproducedbyathinsuperconducting solenoid. The electromagnetic(EM) calorimetersemploy lead absorbers and use liquid argon as the active medium. The barrel EMcalorimetercovers|η|<1.5 andtheend-capEMcalorimeters cover1.4<|η_|<3.2.Hadroniccalorimetryintheregion|η_|<1.7 isperformedusingsteelabsorberswithscintillatortilesastheac- tive medium. Liquid-argon calorimetry with copper absorbers is used in the hadronic end-cap calorimeters, which cover the region 1.5<|η_|<3.2. The forward calorimeters (3.1<|η_|<4.9) usecopper andtungsten asabsorber with liquidargon as active material.The muonspectrometer(MS)measures thedeflectionof muon trajectoriesin the region |η_|<2.7, using threestations of precision drift tubes (with cathode strip chambers in the inner- moststationfor|η_|>2.0) located inatoroidal magneticfield of approximately0.5 T and1 T in the centraland end-capregions, respectively. The muon spectrometer is also instrumented with separate trigger chamberscovering |η|<2.4. Events are selected usinga first-leveltrigger implementedin custom electronics,de- signedtoreducetheeventratedownto100 kHzusingasubsetof detectorinformation[18].Softwarealgorithms withaccesstothe fulldetectorinformation arethen usedto yielda recordedevent rateofabout1 kHz.

3. Analysis 3.1.Signalsimulation

Signal samples are generated by using the Charybdis2 1.0.4 generator [19] to simulate the production and decay of rotating blackholes inmodels withn=^2, ⁴ând⁶ êxtra^dimensionsând valuesofMDrangingfrom2 TeVto5 TeV.Blackholesareassumed to be produced over a continuous rangeof mass values above a thresholdMth, setsoastoavoidthe theoreticaluncertainties as- sociated with the region close to M_D. The analysis is guided by twobenchmarksignal models,thefirstofwhichhas MD=^{2 TeV} andM_th=^{7 TeV,}^resultingⁱⁿâcross-sectionof0.72 pb.Thesec- ondhas MD=^{4 TeV,} ^Mth=^{6 TeV,}ândâcross-sectionof0.93 pb.

Inthesesimulations,noinitial-stategravitationalradiationisper- mitted,whiletheﬁnal decayofthe black-holeremnant produces avariable numberof particles,whose multiplicity is drawn from aPoissondistributioninaccordancewiththe Charybdis2 default.

TheCTEQ6L1 partondistribution functions(PDFs)usedare taken from Ref. [20], while the ﬁnal-state fragmentation and parton showeringismodelledusing Pythia8[21].Thedetectorresponseis modelledusinga fastsimulationoftheresponseofthecalorime- ters[22] and Geant4 [23] forother parts of thedetector. Events fromminimum-biasinteractions arealsosimulatedwith Pythia8.

Theyareoverlaid onthe simulatedsignalandbackgroundevents accordingto theluminosity proﬁle ofthe recorded data.Interac- tionswithin the samebunchcrossing asthehard-scatteringpro- cessandinneighbouringbunchcrossingsareboth simulatedand arereferredtoaspile-up.

1 ATLASusesaright-handedcoordinatesystemwithitsoriginatthe nominal interactionpoint(IP)inthecentreofthedetectorandthez-axisalongthebeam direction.Thex-axispointsfromtheIPtothecentreoftheLHCring,andthey-axis pointsupward.Cylindricalcoordinates(r,φ)areusedinthetransverseplane,φbe- ingtheazimuthalanglearoundthebeamdirection.Thepseudorapidityisdeﬁned intermsofthepolarangleθasη= −^{ln tan}(θ/2).Objectseparationsaremeasured usingR=

(φ)²+ (η)².

3.2. Eventselection

Events are selected from a sample with an integrated luminosity of 3.2±⁰.2 fb⁻¹. The luminosity estimate is derived following thesame methodology asthat detailedin Ref. [24], from a calibration of the luminosity scale using a pair of x– y beam- separation scans performed in August 2015. The event selection uses the lowest-threshold single-leptontriggers available in each data-taking period withgood operational conditions. The single- electron trigger uses a minimum thresholdof E_T=^{60 GeV.} ^The minimum threshold used for the single-muon trigger is pT= 50 GeV. All the ﬁnal-state objects are required to satisfy ba- sic criteriato ensure that they are well reconstructed andorigi- natefromtheprimary interaction.Candidateelectronsandmuons are required to have pT>10 GeV andpseudorapidity |η_|<2.47 (electrons) or |η|<2.5 (muons). They are also required to satisfy baselineidentiﬁcationcriteria(the “Loose”operatingpoint of Ref. [25] for electrons andthe “Medium” criteriaof Ref. [26] for muons). Jetsofhadronsare reconstructed usingthe anti-kt algorithm with a radius parameter of 0.4 [27] and are required to be of at least “loose” quality [28] and to have a calibrated [29]

pT>20 GeV and |η_|<2.8. Jets containing b-hadrons are iden- tiﬁed usingthe “b-tagging”techniques described in Refs. [30,31].

Toavoiddouble-countingofreconstructedobjects,electrons shar- ing an inner detectortrack witha muon are removed. Following this, jetcandidates thatare notb-taggedare removed iftheyare within R<0.2 of an electron candidate. Finally, lepton candi- dates are removed if they lie within R<0.4 of a survivingjet candidatethat isnottaggedasoriginatingfrompile-up[32].The remaining electrons are required to satisfy the “Tight” operating point ofRef. [25]. Leptonsare requiredto beisolated fromother activityusing a relativelyloose criterion designedtopass 99% of leptonsfrom Z decays[26,33].Eventsaresortedintoelectronand muon channelsaccordingto theﬂavour ofthehighest pT lepton.

Two signal regions (SRs) are deﬁned, requiring a leading lepton withp_T>100 GeV andatleasttwootherobjects(leptonsorjets) with pT>100 GeV,with

pT>2 TeV or3 TeV,where pT in- cludes allobjects intheeventwith pT>60 GeV.The ﬁrstsignal region(namedSR-2TeV)allowsthesearchtocovertheparameter spaceneartheexistinglimits,whilethesecond(namedSR-3TeV) providessensitivityatthehighest

pT accessible.TheSR-3TeV selection gives eﬃciency×acceptance values for the benchmark signalmodelsof19%(forthemodelat MD=^{2 TeV,}^Mth=^{7 TeV)} and8%(forthemodelatMD=^{4 TeV,}^Mth=^{6 TeV).}

3.3. Backgrounds

ThedominantbackgroundsoriginatefromW and Z bosonpro- ductionassociatedwithhadronicjets(W+^{jets and} ^Z+^jets)^and from tt production.¯ For these backgrounds, the distributions in kinematic quantities are predicted by Monte Carlo(MC) simulations, which are normalised to data in dedicated control regions (CRs). Each CR uses selections which enhance the contribution of the relevant background while maintaining a negligible expected signal contribution.Single-top-quark anddiboson produc- tionprocessesgivesmallcontributionsthatareestimateddirectly fromsimulations,withnormalisationstakenfromRefs.[34,35]and fromthegenerator,respectively.Thebosonicbackgroundprocesses are simulated using Sherpa 2.1 [36], while POWHEG [37–39] in conjunction with Pythia6 [40] is used for top quark production processes. All these background simulations use the CT10 PDF set[41].Thedetectorresponseismodelledusing Geant4.Theelec- tron channel also contains background eventsfrom hadronicjets whichareincorrectly reconstructedaselectrons.Thisbackground, called “multijet”, is estimated from the data using a sample of

(3)

Table 1

DeﬁnitionsofthesignalregionsandofthecontrolregionsusedintheestimateoftheW+^jets, Z+jets andt¯t backgrounds.

Selection Control regions

Signal regions Z+^jets ^W+^jets ^t¯t

pT 750–1500 GeV >2000(3000)GeV

Number of objects ≥^{3 objects} ≥^{3 objects}

(leptons or jets) with pT>60 GeV with pT>100 GeV

Leading lepton Isolated Isolated

(electron or muon) with pT>60 GeV with pT>100 GeV

m 80–100 GeV n/a

E^miss_T n/a >60 GeV n/a n/a

Number of leptons =2, opposite sign

=¹ ≥¹

same ﬂavour

Number of b-tagged jets n/a =⁰ ≥²

Number of jets n/a ≥⁴ n/a

eventsselectedwithloosenedidentiﬁcationcriteriausingtheMa- trixMethod[42].Therateofbackgroundmuonsfromhadronicjets isnegligible.

ThebackgroundCRselectioncriteriaaresummarisedinTable 1.

All of the CRs select events with 750<

pT<1500 GeV, including atleastthree objects with pT>60 GeV of whichone is requiredtobe alepton. The Z+^{jets CR}additionallyrequiresex- actlytwo leptons withthesameﬂavour andopposite charge and an invariant mass m in the range 80–100 GeV. The W +^jets CRrequires events withexactly one lepton and a missing trans- versemomentum E^miss_T [43]exceeding60 GeV.InthisCR,inorder to suppress background from top quark production, noneof the jets may be b-tagged. The tt CR¯ also requires exactly one lepton,buttheremustbeatleastfourjetsofwhichatleasttwoare b-tagged.Inordertouseinformationabouttheshapeofthe

pT distributiontomoreaccuratelyconstrainthe normalisationofthe W+^jets, ^Z+^{jets and}^t¯t backgroundsintheSRs,eachcontrolre- gionisdividedintothree250-GeV-widebins.

3.4. Systematicuncertainties

Thesystematicuncertainties inthesignalandbackgrounds in- cludethose dueto the limitednumbers ofsimulatedevents and tothemeasurementofintegratedluminosity.Experimentaluncer- tainties arising from the trigger efficiencies, lepton identification and reconstruction procedures, the b-tagging algorithm and the energycalibrationofleptons andjets,aswell aseffectsfromthe jet energy resolution, are also takeninto account. Potential mis- modelling by the MC simulations of the W +^jets, ^Z+^{jets and} tt backgrounds¯ isquantifiedbycomparingthenominalagainstal- ternative simulated samples andPDF sets. Forthe W+^{jets and} Z +jets backgrounds, simulated by Sherpa, the default renormalisation, factorisation and resummation scales are doubled or halved. The matrix element andparton shower are matched us- ingtheCKKW[44]scheme,forwhichthedefaultscaleof20 GeV is changed to 15 GeV and to 30 GeV. For tt,¯ uncertainties in the hard scatter andfragmentation are estimatedby comparison withalternativegenerators andpartonshower models.Variations of the renormalisation scale and of the amount of initial- and final-stateradiation are performedwithin the nominalgenerator.

Sincetheoverallnormalisations ofthebackgroundsarewell con- strainedbytheﬁtstothedatadescribedbelow,onlyvariationsin shape as a function of

pT are relevant. The systematic uncer- tainty in the predictedyields in both channels ofSR-2TeV and SR-3TeV is dominated by the limited sizes of the Monte Carlo samples.ThetotaluncertaintiesintheSRsaremainlyofstatistical origin.

4. Results

Results are extracted from profile likelihood fits using three backgroundnormalisationparametersfortheW+^jets,^Z+^{jets and} t¯t backgrounds.Thesenormalisationparametersarefreelyfloating inthefits.Nuisanceparametersareincludedinthefitstodescribe the systematic uncertainties, takinginto account the correlations acrosstheprocessesandregionsinvolvedineachfit.Abackground likelihoodfittoallcontrolregionsofbothleptonchannels,assum- ing no signal contribution,is usedto predict theexpectedyields invalidationregions(VRs)andtotestthehypothesisthatthedata iswelldescribed withnosignalintheseregions.TheVRsarede- finedusingthesameeventselectionsasthesignalregions,butin therange1500<

pT<2000 GeV.AsintheCRs,anysignalcon- taminationintheVRsisexpectedto besmall, basedonprevious analyses[12]andonsignalsimulations.Comparisonsbetweenthe data andthe predictions in the control regions, wherethe back- groundpredictions are adjusted by thebackground likelihoodﬁt, maybeseeninFig. 1.TheMCsimulationprovidesagooddescrip- tionoftheCRdata,withscalefactorsof0.81±⁰.07,1.01±⁰.08 and0.95±⁰.08 for W+^jets, ^Z+^{jets and}^tt respectively.¯ No sig- niﬁcant deviation fromthe background predictionis observed in the VRs.

Fig. 2 showsthe dataandbackgroundpredictionsfor pT in the electron andmuon channelsfollowing thebackground likeli- hoodfit,withtwosignal modelsoverlaid.Thisfigure usestheSR selectionexceptforthefinalrequirementon

p_T.Thedataarein good agreementwiththebackgroundpredictionacross therange of

pT whichcanbetestedwiththepresentdata,withthesize and pattern of deviations between data and background predic- tion being consistent withstatistical fluctuationsand the size of the systematicuncertainties. Table 2presentsthe dataandback- ground predictions in the signal regions. The number of events observed inSR-3TeVis higherthanthe backgroundestimate in the electron channel with a p-value of 1% when tested against thebackground-onlyhypothesis.Theexcessisnot sufficientlysig- nificant to be considered asevidence of any new physics effect.

The ﬁnal results are therefore derived from the combination of the two channels. The observed numbers of eventsin SR-2TeV andSR-3TeVare192and13respectivelyforthecombinationof theelectronandmuonchannels,tobecomparedwithﬁttedback- groundpredictionsof181±^{11 and}⁹.9±¹.4.

Model-independent cross-section upper limits on any poten- tialnewphysics contributionareobtainedfromfits toall control regions and to signal regions combining the electron andmuon channels,withpotentialsignalcontributionsincludedviaafreely- floatingparameterinthosesignalregions.Model-independentup- per limitsof12.1 fb (3.4 fb)atthe 95%confidence level(CL) are set on the maximum observable cross-section (defined as cross-

(4)

Fig. 1. The

pTdistributionineachofthecontrolregions.The W+^{jets CR}îs^shownⁱⁿ^(a)ând^(b),^the ^Z+^{jets CR}ⁱⁿ^(c)ând^(d),ând^the^t¯t CRin(e)and(f). The electronchannelisshownin(a),(c)and(e),andthemuonchannelin(b),(d)and(f).Thedataareshownaspointswitherrorbars;allexpectedbackgroundsareshownas stackedcolouredhistograms,withthetotalbackgrounduncertaintyshownasashadedband.Thelowerpanelsshowtheratioofthedatatotheexpectedbackground.The t¯t,W+jets andZ+jets backgroundsarenormalisedbythefactors0.95,0.81and1.01asobtainedfromthebackgroundlikelihoodfit.Thesingle-top-quarkanddiboson backgroundnormalisationsaretakenfromthesimulation.Themultijetbackgroundisobtainedusingadata-drivenmethod.Additionally,thelikelihoodfitmayconstrain nuisanceparametersforcertainsystematicuncertainties,alteringthenormalisationandshapeofsomeofthedistributions.

section × âcceptance × êfficiency) âllowed ^forâny^form ôf^new physics in the SR-2TeV (SR-3TeV) region which produces a lepton inconjunction withat least two other objects, each with pT>100 GeV.

Fits including predicted signal yields in all control and sig- nalregions simultaneouslyareusedtoextractexclusionlimitsfor speciﬁcblack-holesignal models.Since thesignalregions overlap in

p_T,theseexclusionﬁtsareperformedfor

p_T>3 TeV,com-

biningtheelectronandmuondata.Conﬁdencelevelsareevaluated usingtheCLs procedure[45].Theresultsare showninFig. 3,to- gether with the corresponding limit from the Run 1 analysis at

√s=^{8 TeV}^[12].^The^impact^on^the^Mthlimitforn=^{6 due}^to^the PDF-induced uncertainties inthe signal cross-section variesfrom

±^{200 GeV to} ±^{100 GeV as} ^MD variesfrom2 TeVto 5 TeV.The limit on Mth is more stringentthan that from the Run 1search by almost 3 TeV at M_D=^{2 TeV and} ^by ^more ^than ^{2 TeV} ^at

Search for TeV-scale gravity signatures in high-mass final states with leptons and jets with the ATLAS detector at $\sqrt{s}=13$ TeV

Physics Letters B

Search for TeV-scale gravity signatures in high-mass ﬁnal states with leptons and jets with the ATLAS detector at √

s = 13 TeV

s = ^{13 TeV}