Contents lists available atScienceDirect
Physics Letters B
www.elsevier.com/locate/physletb
Search for pair-produced resonances decaying to jet pairs in proton–proton collisions at √
s = 8 TeV
.CMSCollaboration
CERN,Switzerland
a r t i c l e i n f o a b s t ra c t
Articlehistory:
Received24December2014 Receivedinrevisedform16April2015 Accepted21April2015
Availableonline24April2015 Editor: M.Doser
Keywords:
CMS Physics Dijets
Results are reportedofageneralsearchfor pairproductionofheavyresonances decayingtopairs of hadronic jets inevents with at least fourjets. The study is basedon upto 19.4 fb−1 of integrated luminosity fromproton–proton collisionsatacenter-of-massenergyof8 TeV,recordedwiththe CMS detector atthe LHC. Limits are determined on the production ofscalar top quarks (top squarks) in theframeworkofR-parityviolatingsupersymmetryandontheproductionofcolor-octetvectorbosons (colorons).FirstlimitsattheLHCareplacedontopsquarkproductionfortwoscenarios.Thefirstassumes decaytoabottomquarkandalight-flavorquarkandisexcludedformassesbetween200and385 GeV, and thesecondassumesdecaytoapairoflight-flavorquarksandisexcludedformassesbetween200 and 350 GeVat95% confidencelevel. Previous limits oncolorons decayingto light-flavor quarks are extendedtoexcludemassesfrom200to835 GeV.
©2015CERNforthebenefitoftheCMSCollaboration.PublishedbyElsevierB.V.Thisisanopenaccess articleundertheCCBYlicense(http://creativecommons.org/licenses/by/4.0/).FundedbySCOAP3.
1. Introduction
Wepresenttheresultsofasearchforpairproductionofheavy resonances decayingto pairs oflight- and heavy-flavor quarks in multijetevents.Theanalysisisbasedondatasamplescorrespond- ingtoasmuchas19.4±0.5 fb−1[1]ofintegratedluminosityfrom proton–proton collisions at √
s=8 TeV, collected with the CMS detector[2] at the CERN LHC in 2012. Events that have at least fourjetswithhightransversemomentum(pT)withrespecttothe beamdirectionareselectedandinvestigatedforevidenceofpair- produceddijetresonances.
Many models of particle physics beyond the standard model (SM) incorporate particles that decay into fully hadronic final states. Supersymmetric (SUSY) models are SM extensions,which simultaneouslysolvethehierarchyproblemandunifyparticle in- teractions[3,4].In natural SUSY models, wherethere is minimal fine-tuning, thetop quark superpartner(top squark) andthe su- perpartners of the Higgs boson (higgsinos) are required to be light[5–9].NaturalSUSYisunderconstrainedincertainR-parityvi- olating(RPV)scenarios[10].R-parityisaquantumnumberdefined as R= (−1)3B+L+2S, where B and L are the baryon and lepton numbers,respectively, and S is thespin. The RPVsuperpotential, W ,isdefinedas
E-mailaddress:[email protected].
W =1
2λi jkLiLjEkc+ λi jkLiQjDck+1
2λi jkUciDcjDck, (1) where λare thecouplings, i, j,k are thegeneration indices,c is the charge conjugation, L and Q are the doublet superfields of the lepton andquark, respectively, and E, D, and U arethe sin- gletsuperfieldsofthe lepton,down-typeandup-typequarks,re- spectively.ModelsthatincorporateRPVmayallowbaryonnumber violation through a non-zero λUDD coupling, and one such un- constrainedscenario [11] isthatofthe hadronicallydecaying top squark,t→qq.If thetop squarks arepair-produced inhadronic collisions andthendecayvia suchan RPVprocess,the finalstate would consist of four jetswith no momentum imbalance in the transverseplane.
In additionto top squarkproduction,hadron collidersearches forpairproductionof resonancesdecayinginto jetpairs aresen- sitive to a number of models that predict new particles carry- ing color quantum numbers. Some models predict pair produc- tionthroughgginteractionsofcolor-octetvectors, alsocalledcol- orons (C) [12], which then decay to quark pairs. The associated final state of the signal is characterized by the presence of four high-pTjets.
CDF Collaboration has placed 95%confidence level (CL)exclu- sionlimits[13]ontopsquarkproductionfollowedbyRPVdecays in the mass range 50–90 GeV and on coloronproduction in the mass range50–125 GeV.At theLHC, ATLAS hasplaced limitson scalargluonmassesbetween100and185 GeV[14],andseparately http://dx.doi.org/10.1016/j.physletb.2015.04.045
0370-2693/©2015CERNforthebenefitoftheCMSCollaboration.PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense (http://creativecommons.org/licenses/by/4.0/).FundedbySCOAP3.
formasses between 150 and287 GeV [15]. The CMS search for paired dijet resonances resulted in limitson coloron masses be- tween250and740 GeV[16].However,noneofthesesearcheshas beensensitiveenoughtosetlimitsonhadronicRPVdecaysofdi- rectlyproducedtopsquarks.
Inthispaper,we concentrateon searchesfortop squarks and colorons.The benchmark signalsare those wherethe top squark isthelightestsupersymmetricparticle,andinonescenariodecays into two light quarks, and in the second scenario it decays into ab quark anda light quark [17–22].We separately consider the possibilityofdecayswithinthecoloronmodel(gg→CC→qqqq).
The analysis employs a well-established search strategy with optimized event selections. The distribution of a variable repre- sentative of the top squark mass is investigated for evidence of a signal consistent with localized deviations from the estimated large,steeply falling SM background todata. The estimate ofthe backgroundisperformedwithafittothefallingpartofthemass spectrumin data,anda SM MC analysis isused to optimizethe signalselectionandtoderivesystematicuncertainties.
2. CMSexperiment
ThecentralfeatureoftheCMSapparatus[2]isasuperconduct- ingsolenoidof 6 minternal diameter,providing amagnetic field of3.8 T.Withinthesuperconductingsolenoidvolumeareasilicon pixelandstriptracker,aleadtungstateelectromagneticcalorime- ter (ECAL), and a hadron calorimeter (HCAL), which is made of interleaved layers of scintillator and brass absorber. Muons are measuredingasionizationdetectorsembeddedinthesteelreturn yokeoutsidethesolenoid. Extendedforwardcalorimetry comple- mentsthe coverage providedby thebarrelandendcapdetectors.
Energydeposits fromhadronicjetsare measured using theECAL andHCAL. A more detailed description of the CMS detector, to- gether with a definition of the coordinate system used and the relevantkinematicvariables,canbefoundinRef.[2].
3. Triggeringandobjectreconstruction
One data set, representing 19.4 fb−1, was recorded over the entire2012 data taking period with a multilevel trigger system, whichselectedeventswithatleastfourjetswith pT>80 GeV to be reconstructed from only calorimeter information. In addition, asecond data setwas recordedusing thesame triggerlogic, but witha lower jet pT threshold.Thisthresholdwas decreasedpro- gressivelyfrom50to45 GeVduringthe2012datatakingperiod.
Thelatterdatarepresentonlyasubsetoftheentire2012dataset, correspondingtoanintegratedluminosityof12.4 fb−1.Theanaly- sisisseparatedintotwoparts:adedicated“low-mass”searchwith afocusonthemassregionfrom200to300 GeV,whichtakesad- vantageof thislower jet pT threshold,anda “high-mass” search focusingontopsquarkmassesabove300 GeV,whichusestheen- tire19.4 fb−1 datasetandextendstheexpectedtopsquarkmass searchsensitivityby40 GeV.
TheanalysisisbaseduponobjectsreconstructedusingtheCMS Particle Flow algorithm [23]. This method combines calorimeter information with reconstructed charged particle tracks to iden- tifyindividual particlessuch asphotons,leptons, andneutraland chargedhadrons. Theenergyofphotonsisdirectlyobtainedfrom the calibrated ECAL measurement. The energy of the electron is determined from a combination of its track momentum at the main interaction vertex, the corresponding ECAL cluster energy, andthe energy sumof all bremsstrahlungphotons associated to the track.The energy of a muon is obtained from its associated trackmomentum.Thechargedhadronenergyiscalculatedfroma combinationofthetrackmomentumandthecorrespondingECAL
andHCALenergies,correctedforzero-suppressioneffects,andcal- ibrated for the combined response function of the calorimeters.
Finally,theenergyofneutralhadronsisobtainedfromthecorre- spondingcorrectedECALandHCALenergies.Jetsarereconstructed from the particleflow “objects” using theanti-kT algorithm [24]
withadistanceparameterof0.5in y–φspace,where y isthera- pidity.
Jetenergyscalecorrections[25] areappliedtoaccountforthe combined response function of the calorimeters to hadrons. The correctionsarederivedfromMonteCarlo(MC)simulationandare confirmedwithinsitumeasurementsoftheenergybalanceofdi- jet and photon+jet events. In data, a small residual correction factor is included to account for differences in jet response be- tweendataandsimulation.Thetotalsizeoftheappliedcorrections isapproximately5–10%,andthecorrespondinguncertainties vary from 3 to 5%, depending on the measured jet pseudorapidity η
and pT. Toremove misidentified jets, which arise primarily from calorimeternoise, jet quality criteria[26] are applied. More than 99.8% ofall selected jets, in both dataandsignal eventsamples, satisfythesecriteria.
Toidentifyjetsproducedby bquark hadronization,theanaly- sis uses themedium selection ofthe combinedsecondary vertex b-tagging algorithm [27]. The algorithm employs a multivariate technique, which takes as input information fromthe transverse impactparameterwithrespecttotheprimaryvertexoftheassoci- atedtracksandfromcharacteristicsofthereconstructedsecondary vertices. The output of the algorithm is used to discriminate b quark jets from light-flavor and gluon jets, with typical values of b-tagging efficiency and misidentification probabilities of 72%
and 1.1%,respectively.
4. Generationofsimulatedevents
Both top squarkproduction andcoloronproduction are simu- latedusing the MadGraph 5.1.5.12 [28] eventgenerator withthe CTEQ6L1 parton distribution functions[29], andtheir decaysare simulated using the pythia 6.426 [30] MC program. Top squark signaleventsaregeneratedwithuptotwoadditionalinitial-state partons, and each top squark decays into two jets through the λUDD quark RPV coupling. Two scenarios are considered for this coupling.First,thecouplingλ312,wherethethreenumericalsub- scriptsrefertothequarkgenerationsofthecorrespondingquarks, issettoanon-zerovaluesuchthatthedecayofthetopsquarkto twolight-flavorjetsisallowed.Thesecondcaseinsteadsetsanon- zerovalueforλ323,resultingintopsquarkdecayintoonebjetand onelight-flavorjet.Inbothoftheabovecases,thebranchingfrac- tion of the top squark decay to two jets is set to 100%. For the generationofthissignal,all superpartnersexceptthetopsquarks are takento be decoupled [17–21] andnointermediate particles are produced inthetop squarkdecay. Topsquarks are generated with masses from 100 GeV to 1 TeV in 50 GeV steps for both coupling scenarios. The cross section estimates [31] are made at next-to-leading order(NLO) withnext-to-leading-logarithm (NLL) corrections [32–36], and assigned appropriate theoretical uncer- tainties[31].Forthecoloronsignalscenario,weconsiderthecase whereeachcolorondecaysintotwolight-flavorjetswithabranch- ing fraction of 100%. For this signal, masses are generated from 100 GeVto 2 TeV,andNLO crosssection estimatesare used.For boththetopsquarkandcoloronmodels,thenaturalwidthofthe signal resonanceistakentobe much smallerthantheresolution ofthedetector.BackgroundsfromSMmultijetprocessesaresimu- latedthroughmatchedtree-levelmatrixelementsfortwo- tofour- jet production using MadGraph, and these events are showered through pythia.Inall samples,theMLMmatchingprocedure[37]
is used, and simulation of the CMS detector is performed with Geant4[38].
Fig. 1. ProbabilitydensitydistributionsofthefourthhighestjetpT(top)andRmin (bottom)foreventsfromdata,thesimulatedSMmultijetsample,anda400 GeV topsquarksignal.Statisticaluncertaintiesareshownforthetopsquarksignalas verticalbarsandfor dataasarrows.Eventscontainatleastfourjets,eachwith pT>120 GeV and|η|<2.5,andalldistributionshaveanareanormalizedtounity.
5. Eventselection
Eventsrecordedwiththefour-jettriggersare requiredto have a well-reconstructed primary eventvertex[39].Events must also contain at least four jets, each with |η|<2.5 and reconstructed pT greater than 80 GeV forthe low-pT trigger and120 GeV for thehigher-pTtrigger.Withtheaboverequirements,theofflineef- ficiencyisabove99%forallselectedevents.
The leading four jets, ordered in pT, are used to createthree unique combinationsof dijetpairs per event. A distancevariable isimplemented to selectthejet pairing that bestcorresponds to thetworesonancedecays,R=
(η)2+ (φ)2,whereηand
φ are the differences in η andφ ofbetween two the jets, re- spectively.Thisvariable[40] exploitsthesmallerrelative distance between daughter jets from the same top squark parent decays compared to that between uncorrelated jets. For each dijet pair configurationthevalueofRdijetiscalculated:
Rdijet=
i=1,2
|Ri−1|, (2)
where Ri represents the separation between two jets in dijet pair i. An offset of 1 has been chosen since this maintains a maximal signal efficiency while minimizing the selection of di- jetsystemscomposedofresolvedjetsfromradiatedgluonspaired withtheir parent jet.The configurationthat minimizesthe value
Rdijetisselected,withRminrepresentingtheminimumRdijet fortheevent.Fig. 1showstheprobability densitydistributions of thefourthhighestjet pT andthe Rmin variablefordataevents,
Fig. 2. Probabilitydensitydistributionsofm/mav (top)andηdijet (bottom)for eventsfromdata,the simulatedSMmultijetsample,and a400 GeVtopsquark signal.Statisticaluncertaintiesareshownforthetopsquarksignalasverticalbars andfordataasarrows.Eventscontainatleastfourjets,eachwith pT>120 GeV and|η|<2.5,andalldistributionshaveanareanormalizedtounity.
thoseofa simulatedSMmultijetsample,andthoseof400 GeVtop squarksignalsample.
Onceadijetpairconfigurationischosen,two additionalquan- titiesare usedtorejectthe backgroundsfromSMmultijetevents and incorrect signal pairings: the pseudorapidity difference be- tween the two dijet systems ηdijet, and the absolute value of the fractional mass difference m/mav, where m is the differ- encebetweenthetwodijetmassesandmav istheiraveragevalue.
Insignal eventswherethecorrectpairingischosen,the m/mav quantityispeakedatzerowithamuchnarrowerdistributionthan thatforSMmultijetbackgroundorincorrectlypairedsignalevents.
Thus, thesensitivityofthesearchbenefits fromimposing amax- imum value on m/mav. Similarly, it is advantageous to require thatηdijetbesmall.Fig. 2showstheprobabilitydensitydistribu- tionsofthem/mav andηdijetvariablesfordataevents,thoseof asimulatedSMmultijetsample,andthoseof400 GeVtopsquark signalsample.Anadditionalkinematicvariableiscalculatedfor eachdijetsystem:
=
⎛
⎝
i=1,2
|piT|
⎞
⎠ −mav, (3)
where the pT sum is over the two jets in the dijet configura- tion. Thistype ofvariable hasbeen usedextensively inhadronic resonance searchesatboth theTevatronandthe LHC[16,41–44].
Requiringaminimumvalueofresultsinaloweringofthepeak positionvalueofthemav distributionfrombackgroundSMmultijet events.Withthisselectionthemodelingofthebackgroundshape
Fig. 3. DistributionsofthefittosimulatedbackgroundSMmultijetevents(solid redline)anda500 GeVtopsquark(dashedblueline),normalizedtoafactorof tentimesitscrosssection,areshownforthehigh-massoptimizationscenario.The dottedverticallinesrepresentthe integrationwindowusedbythe optimization procedure.
canbeextendedtolower valuesofmav,makingawider rangeof topsquarkandcoloronmassesaccessibletothesearch.
Finally,asthepresenceofheavy-flavorfinal statejetsisanat- uralextension oftheRPV topsquarkscenarios, theuse ofbtag- gingis exploitedtofurther increase signal sensitivityby increas- ing backgroundrejection.We consider two scenarios:the heavy- flavorsearch, whichuses btagging toincrease thesensitivityfor topsquarkdecaysintoheavy-flavor jets,andtheinclusivesearch, whichfocusesinsteadondecaysintolight-flavorjets.
Theoptimizationforthesignalselectionisperformedasafunc- tion of the three kinematic variables described above: m/mav,
ηdijet, , aswell as the fourth jet pT. Because the number of expectedbackgroundeventsislarge,we use S/√
B asthemetric forsignal optimization, where S and B are thenumberof signal andbackgroundevents, respectively,and B isdetermined by us- ing the mav ofsimulated SM events.The valuesof S and B are settothenumberofeventswithinawindowofwidth±10%cen- teredatthegeneratedtopsquarkmass,wherethevalueof10%is roughlytwicetheexpectedresolutionforsignalmasses.Westudy thismetricbyevaluating S andB basedoneventspassinganum- ber of thresholds of each kinematic variable and obtain several four-dimensional tables, in which a value of S/√
B is found for everycombinationofthefourvariables.Thesetablesareproduced in the low- and high-mass search regions, and for the inclusive andheavy-flavor analyses separately.An exampleof thisisgiven inFig. 3,wherethedistributionfora500 GeVtopsquarkandfor afitto thesimulatedSM multijetdistributionare shownforone operatingpoint.Thesignalshapeisbimodalowingtoasmallfrac- tionofeventswithincorrectsignalpairings,andtheGaussianpeak centeredatthegeneratedmassisthepartofthedistributionused in the optimization. The threshold values of the four kinematic variables,correspondingtomaximumvaluesofS/√
B intheseta- bles,aretakenasaworkingpoint.Becauseofsimilarresultsinthis optimization,theinclusiveandheavy-flavorsearchesusecommon workingpoints,withtheexceptionoftheheavy-flavoranalysisre- quirementofb tagging.Asummary ofthe requirementsis listed inTable 1forboth thelow- andhigh-masssearches.Anexample oftheηdijet variableisshowninFig. 4.Thecorrelation between thepseudorapidityvaluesforthe twodijetsystemsis plottedfor both400 GeVtopsquarkandsimulatedSMsamples,withthere- gion of allowed values of the ηdijet variable indicated. For the heavy-flavorsearch, we repeatthe optimizationprocedureby us- ingselections basedon fivedifferentb-tagged jet configurations:
atleastoneb-taggedjetintheevent,atleastoneb-taggedjetin
Fig. 4. Theηvalueforthehigher-pTreconstructeddijetsystemversusthatofthe lower-pTdijetsystemintheselectedpair.Thisdistributionisshownfor400 GeV topsquark(top)andsimulatedSMmultijetsamples(bottom),withtherighthand scaleindicatingtheexpectednumberofeventsperbin.Thediagonallinesindi- catetheoptimizedregionofallowedηdijetvalues,andeventswithvaluesfalling betweenthetwolinespassthisrequirement.
Table 1
Summaryofthelow- andhigh-massselectioncriteriaforboththeinclusiveand heavy-flavoranalyses.Fortheheavy-flavoranalysis,inadditiontotherequirements below,atleasttwoofthefourhighestpTjetsmustbeb-tagged.
Low-mass search High-mass search
Mass range 200–300 GeV >300 GeV
Integrated luminosity 12.4 fb−1 19.4 fb−1
m/mav <0.15 <0.15
ηdijet <1.0 <1.0
>70 GeV >100 GeV
Fourth jet pT >80 GeV >120 GeV
the four highest pT jets, at leasttwo b-taggedjets inthe event, atleasttwo b-taggedjetsinthefourhighest pT jets, andatleast oneb-taggedjetineachofthetwochosendijetsystems.Wefind that theoptimal selectionis therequirementthat eventscontain atleasttwob-taggedjetsamongthefourhighestpTjets.
