NuclearPhysicsA00(2011)1–4 Measurements of di-jets in p+p and Au+Au in the STAR experiment. ElenaBrunafortheSTARCollaboration PhysicsDepartment,YaleUniversity,272WhitneyAve.,NewHaven,06520-8120CT,USA. 1 1 0 2 Abstract n a Jetsareproducedfromhardscatteringsintheearlystagesofheavy-ioncollisions.Itisexpectedthatthesehigh-pT J partonstravelthroughthehotanddensemediumbeforefragmenting.Thereforetheyareexpectedtosufferenergyloss 7 intheQGPviagluonradiationorelasticcollisionsalongtheirpath.Measurementsfromfulljetreconstructionhelpin 2 theunderstandingofenergylossanditseffectonthejetstructureandenergyprofile. Adata-drivencharacterization of the background in Au+Au is needed in order to compare the results to p+p. The large coverage of the STAR ] x detectoralong withan onlinetrigger alsoallows forreconstructing di-jets. Suitableselection oftrigger jetsallows e for studying a subset of jets on the recoil side which are biased towards higher energy loss because of a larger, on - l average,in-mediumpathlengthtraversed.Triggerjetsarealsousedtostudyjet-hadroncorrelationsasanindependent c measurementtoassesstheeffectofenergylossontherecoilside.Wepresentmeasurementsofdi-jetsandjet-hadron u correlationsinthepresenceofreconstructeddi-jetsinAu+Auandp+pat200GeVintheSTARexperiment. n [ Keywords: Heavy-ioncollisions,fulljetreconstruction,dijets,jet-hadroncorrelations 1 v 3 1. Introduction 2 2 5 Fulljetreconstructionreducesthegeometricalbiases[1]inherentthestudyofjetquenchinginsingleparticleand . di-hadronanalyses,sinceitallowsamoreprecisedeterminationoftheenergyoftheinitialhardscatteringofincident 1 partons.Di-jetscanbeexploitedtoprobethehotanddensemediumproducedinhighenergynuclearcollisions.Such 0 1 measurementsarepossibleinSTARthankstoitslargeacceptancecoverage.Highmomentumtriggerjets,selectedby 1 theonlinetrigger,tendtobebiasedtowardspartonsthatscatternearthesurfaceofthemedium[1].Thecorresponding : recoiljetsontheaway-sidethatarenotemittedtangentially,areexpectedtotraveloveralargein-mediumpath-length v i duetothetriggerbias. Di-jetcoincidenceratesarecomparedinp+pandAu+Auinascenarioofbroadeningofjet X structure in the medium. First jet-hadron correlation studies in the presence of a reconstructed di-jet are presented r asatooltoassessbiasesinjet-findingandtoinvestigatethequenchingscenario. Modernjet-findingalgorithmsand a data-drivencorrectionschemesareusedinthisanalysis. 2. Datasetsandanalysis The STAR sub-detectors used for full jet reconstruction are the Time Projection Chamber (TPC) for charged particlesandtheBarrelElectromagneticCalorimeter(BEMC)fortheneutralenergy.BothTPCandBEMChavefull azimuthalcoverageandpseudo-rapidityacceptance η <1.Correctionsfordouble-countingofelectronsandcharged | | hadronicenergydepositionintheBEMCareapplied.Thisanalysisisbasedonp+pyear2006and0-20%mostcentral Au+Auyear2007events. BothdatasetswereselectedwithanonlineHigh-Tower(HT)triggerintheBEMCwhich requiresthetransverseenergyabove5.4GeVinatowerofsize0.05x0.05inη φ. − /NuclearPhysicsA00(2011)1–4 2 Thejet-findingutilizesrecombinationalgorithmsincludedintheFastJetpackage[2,3,4]. Weuse“anti-k ”to T reconstructthesignaljetsand“k ”toestimatebackgroundperunitarea. Thedi-jetsaredefinedbya“trigger”jet, T thatmatchestheonlinetriggeredtowerintheBEMC,anda“recoil”jet,reconstructedontheaway-sideofthetrigger jet(i.e. ∆φ >2.74).BothtriggerandrecoiljetsarereconstructedwitharesolutionparameterR=0.4andarerequired | | tobefullyincludedintheSTARacceptance(i.e. η <1 R). | jet| − Thebackgroundthatissubtractedtogetthereconstructedjetpjet iscalculatedbyFastJetasthemedian p per T T unit area (ρ) obtained from all the jets reconstructed with k . The trigger jets were reconstructed with a pcut = 2 T T GeV/contracksandneutraltowers. Thisrequirementtogetherwiththepresenceofthematchinghighcalorimeter towerprovidesaselectionbiasofhardfragmentingjetsproducedinproximityofthesurfaceofthemedium,and/or less interacting. This selects a similar jet energy scale in p+p and Au+Au. A precise quantification of the trigger bias is currently under study. The requirement of pjet > 20 GeV/c for the reconstructed trigger jets minimizes the T contaminationbybackgroundjets.Inordertostudythebiasintroducedbya pcutontherecoilside,therecoiljetsare T reconstructedwithtwodifferentchoicesofpcut,0.2and2GeV/c.Thebackgroundspectrumconsistingofuncorrelated T backgroundclustersandadditionalhardscatteringsisestimatedinthisdi-jetanalysisbylookingatthejetspectrumin awindowcenteredat π/2withrespecttothetriggerjetaxis. Thebackgroundspectrumisthensubtractedfromthe ± recoiljetspectrum. Region-to-regionfluctuationsareapproximatedbyaGaussiandistributionwithσ = 6.5GeV/c for pcut = 0.2GeV/c[5,6]. Therecoilspectrawith pcut = 2GeV/careunfoldedassumingaGaussianshapeforthe T T backgroundfluctuationwithwidthσ=1.5GeV/cforR=0.4. RecenteffortsarebeingmadeinSTAR[7]toderivea data-drivenformofthebackgroundfluctuationsthataccountsforthenon-Gaussiantails.Preliminarystudiessuggest thattheeffectofnon-Gaussiantailsissmallduetotheflatnessofthemeasuredrecoilspectrum. Thereconstructedjetmomentaarecorrectedforthedifferenttrackingefficienciesinp+pandAu+Au(thediffer- enceis 10%). ∼ 3. Di-jetcoincidencerate Theratioofdi-jetspectrainAu+Aurelativetop+pisshowninFig.1(left),fortwodifferentvaluesof pcut on T tracks and neutral towers, pcut = 0.2 and 2 GeV/c. A similar suppression is observed for the two different values T of pcut. Acrucialstepinthisanalysisisthecontrolofthetriggerenergyinthetwodifferentcollisionsystems. As T mentionedinSec.2,atightpcutisappliedonthetriggerjetstoallowsimilarenergyscales.However,thereconstructed T trigger jet momentum might still be affected by upward background fluctuations of the order of 1 2 GeV/c that artificially enhance the measured pjet compared to the same jet in p+p. On the other hand, given t−hat trigger jets T in Au+Au may actually interact in the medium (as observed from the near side of jet-hadron correlations [8]), the measured pjet couldcorrespondtoahighermomentumjetinp+p. Boththeseeffectsarequantifiedbyvaryingthe T triggerjetmomentuminp+pby 2GeV/c,asindicatedbythecoloredbandsinFig.1(left). ± Theobservedsuppressioncanbeexplainedwithabroadeningoftherecoiljetenergyprofile,duetoenergyloss tolargeangles. Thiswouldmovepartofthejetenergyoutofthejetarea,shiftingthereconstructedjetspectrumto lower p . AssumingthatbroadeningisresponsibleforaconstantshiftoftheAu+Aurecoilspectrumtowardslower T p relativetop+p,weestimatedthatashiftof7 8GeV/cforrecoiljetswith pcut =0.2GeV/candof6 7GeVfor T − T − pcut = 2GeV/creproducesthemeasuredratiosinleftpanelofFig.1. Thesepreliminaryresultsareconsistentwith T jet-hadroncorrelationstudies(seeSec.4)andmaysuggestlargeangleradiation(i.e. duetobroadening)oftherecoil jetsexposedtoamaximumin-mediumpath. Asaconsequenceofthehigh-p hadronsuppressionrelatedtothesurfacebiasofhigh-p particles[1],onecould T T thinkthattherecoiljetsreconstructedwiththetightrequirementofpcut =2GeV/ccouldbebiasedtowardsthesurface T ofthemediuminAu+Au,hencebeinglessinteractingandlesssubjecttobroadening. Thisissuecanbeinvestigated viajet-hadroncorrelationsstudiedinpresenceofbothtriggerandrecoiljets,asreportedinthefollowingsection. 4. Jet-hadroncorrelationswithreconstructeddi-jets Results reported above show that recoil jets reconstructed with pcut = 2 GeV/c are significantly suppressed, T suggestingstrong interactionsofaway-side jetswiththe medium. Thequestionwe wanttoaddressis whetherthe /NuclearPhysicsA00(2011)1–4 3 AuAu/pp000...7891 R±R ==200 G..44e ppVTT/,,cccuu ttT==r20ig .G2g eeGrV eu/Vcn /cpc.T p(tTr(itgr)ig>)2>02 0G eGVe/Vc/c φ∆ dN/d1/Ndi-jets 456 0S.T5A0R< pprTe,alsimsoicn<a1r.y00 GeV/c ApAp++uupp++ AAapuull rapelTlc, rr>oeec1iolc 0iljo>e iG1tl s0jee VGts/ecV/c T,recoil 0.6 3 0.5 0.4 2 0.3 1 0.2 STAR preliminary 0.1 0 0 15 20 25 30 35 -1 0 1 2 3 4 5 p [GeV/c] ∆φ T,jet Figure1:Left:Ratioofdi-jetcoincidenceratesinAu+Aurelativetop+pfortwovaluesofpcutontherecoilside,0.2and2GeV/c.Thecolored T bandsrepresenttheuncertaintyonthetriggerjetenergy.ThetriggerjetisselectedwithpT,trigger >20GeV/c.Right:Jet-hadroncorrelationsfor 0.5<pT,assoc<1GeV/cinAu+Au(solidlines)andp+p(dashed),withnorequirementsonrecoiljets(black)andwithrecoiljetsselectedabove pT,recoil>10GeV/c(red).Sameselectionoftriggerjetsasleftplot. reconstructedrecoiljetsarebiasedtowardsthesurfaceofthemedium,aswecouldexpectbecauseofthebiasgiven bythe pcut. T Thisanalysisisafirstlookatcorrelationsofhadronswithrespecttothetriggerjetsineventswheredi-jetsare reconstructed. In this context, jet-hadron correlations are expected to show possible biases in the jet-finding via a direct measurement of the recoil side structure. Trigger and recoil jets are reconstructed with pcut = 2 GeV/c and T R=0.4,bothinp+pandAu+Au.Thetriggerjetsarerequiredtohave pjet >20GeV/c. T Figure1(right)showstheazimuthaljet-hadroncorrelationsfor0.5< pT,assoc <1GeV/c,inAu+Au(solidlines) andp+p(dashed),forthecasewherenorequirementisappliedontherecoiljetp (black)andforrecoiljetsselected T above pT,recoil > 10 GeV/c (red). The background level is estimated via a fit with a two-Gaussian function plus a v contribution. Thedefaultjetv valueisthehadronv 2 at6GeV/c. Theassociatedhadronv valuesaretaken fr2om[9]. Theresultsarenotcorr2ectedforbackground2je{ts}. However, about1/30oftherecoilje2tswith pjet > 10 T GeV/creconstructedwith pcut =2GeV/careexpectedtobebackgroundjets(thisisestimatedviathejetspectrumat T π/2withrespecttothetriggerjetaxis,seeSec.2). A broadening of the the away-side and an enhancement of low-p particles in Au+Au relative to p+p can be T observedinFig.1(right). Thiseffectisobservedinbothcaseswithandwithoutrequirementsonthe p oftherecoil T jet.Theaway-sideGaussianwidthisreportedinFig.2(left)asafunctionofpT,assocforthecasewhererecoiljetsare selectedwith pTjet > 10GeV/cand pcTut = 2GeV/c. Athigh pT,assoc theGaussianwidthoftheaway-sideissimilar inAu+Auandp+p,indicatingonlyaminorbroadeningofthehardsectorofthejetfragmentation. Thesystematic uncertainties represent the 2 GeV/c uncertainty on the energy scale for both trigger (red shaded area) and recoil ± jets(bluedashedlines)thatisexpectedduetoenergylossandbackgroundfluctuationsabove pcut = 2GeV/c. The T uncertaintybandduetothechoiceofthejetv isalsoreported(redsolidlines). Thelowerlimitisthecasewhere 2 no jet v is assumed. The upper limit isthe case where the jet v is 50% higherthan the hadron v 2 at 6 GeV/c. 2 2 2{ } ThesofteningofrecoiljetsisclearlyvisibleinFig.2(right),wheretheratiooftheaway-sidepeakinAu+Aurelative top+p(IAA)isshownfordifferent pT,assoc bins. Thedecreaseofhigh-pT associatedparticlesandthecorresponding enhancementofthelow-p particlesinAu+Auwithrespecttop+psuggeststhatalsorecoiljetsreconstructedwitha T largebiasduetothetight pcutaresubjecttotheeffectsofenergyloss. T Preliminaryresultsonjet-hadroncorrelations[8]showthattheaverageenergyoutsideR=0.4isoftheorderof 6-9GeV.Thiscouldindicate,inagreementwithSec.3,thatthemeasuredsuppressionofthedi-jetcoincidencecan beexplainedintermsofabroadeningscenarioduetoenergylossinAu+Authatleadstoanunderestimationofthe jetenergy. /NuclearPhysicsA00(2011)1–4 4 Away-side Gaussian width00000.....23456 STAR tpArvpre2+iurgc eps+ ojlyAeiilsm tujt e.u i (tnun0 uanc-2n cr±y0c. %2± )G2 eGVe/cV/c : AuAu(0-20%)/ppSide Yield IAA 1 STAR prvtrere2ilgc ims ojyeiilsn tjt ae.ur tuny uncn c±c. 2± G2 eGVe/cV/c y- wa 0.1 A 0 0 2 4 6 8 10 0 2 4 6 8 10 p [GeV/c] p [GeV/c] T,assoc T,assoc Figure2:Left:Gaussianwidthoftheaway-sidepeakofjet-hadroncorrelationsasafunctionofpT,assoc,forAu+Au(red)andp+p(black),inthe presenceofarecoiljethavingpjet>10GeV/candpcut=2GeV/c.Thesystematicerrorsaredescribedinthetext.Thetriggerjetisselectedwith T T pT,trigger >20GeV/c. Right: IAAoftheaway-sidepeakofjet-hadroncorrelationsasafunctionofpT,assoc,inthepresenceofarecoiljethaving pjet>10GeV/candpcut=2GeV/c.Thesystematicerrorsaredescribedinthetext.Sameselectionoftriggerjetsasleftplot. T T 5. Summary Themeasuredsuppressionofthedi-jetcoincidencerateinAu+Aurelativetop+pisinqualitativeagreementwith ascenarioofjetbroadeningoutsideofR=0.4.Thisobservationcanbeinterpretedastheeffectofthelargein-medium path-lenghtthatnon-tangentiallyemittedrecoiljetsmusttraverse,giventhetightrequirementonthetriggerjetsthat bias them towards the surface. Jet-hadron correlations performed in the presence of reconstructed di-jets are used bothtoaddresspossiblebiasesinthejet-findingandtoquantifythemissingout-of-conejetenergyduetoquenching effects. The observed broadening and softening of jet fragmentation for the highly biased population of recoil jets reconstructedwith pcut = 2GeV/csuggeststhatthejetfindingisnotbiasedtowardsonlysurfaceornon-interacting T jets. Completionofthecorrections,assessmentofthetriggerbiasandoftheeffectofv inthebackgroundsubtraction 2 willprovideaquantitativeestimateoftheout-of-coneenergylossviajet-hadroncorrelations. Theuseofquenching modelswillbeusefultocomparethemeasuredout-of-coneenergywiththetheoreticalpredictions. 6. Acknowledgments TheauthorwishestothanktheBulldogcomputingfacilityatYaleUniversity. References [1] T.Renk,K.Eskola,Phys.Rev.C75(2007)054910,arXiv:hep-ph/0610059. [2] M.Cacciari,G.Salam,Phys.Lett.B64,57(2006),arXiv:hep-ph/0512210v2. [3] M.Cacciari,G.Salam,G.Soyez,JHEP0804,005(2008),arXiv:0802.1188. [4] M.Cacciari,G.Salam,G.Soyez,JHEP0804,063(2008),arXiv:0802.1189v2. [5] M.PloskonfortheSTARCollaboration,QuarkMatter2009Proceedings,Nucl.Phys.A830(2009)255c,arXiv:0908.1799. [6] E.BrunafortheSTARCollaboration,QuarkMatter2009Proceedings,Nucl.Phys.A830(2009)267c,arXiv:0907.4788. [7] P.JacobsfortheSTARCollaboration,theseproceedings. [8] J.PutschkefortheSTARCollaboration,theseproceedings. [9] I.Abelevetal.(STAR),Phys.Rev.C80,064912(2009),arXiv:0909.0191 NuclearPhysicsA00(2011)1–1 Abstract 1 1 Keywords: 0 2 n 1. a J 7 2 ] x e - l c u n [ 1 v 3 2 2 5 . 1 0 1 1 : v i X r a