Jet studies with STAR at RHIC: jet algorithms, jet shapes, jets in AA
aa r X i v : . [ nu c l - e x ] N ov Jet studies with STAR at RHIC: jet algorithms, jetshapes, jets in AA
J Kapit´an (for the STAR Collaboration)
Na Truhlarce 38/64, Praha 8, Czech RepublicE-mail: [email protected]
Abstract.
Hard scattered partons are predicted to be well calibrated probes of thehot and dense medium produced in heavy ion collisions. Interactions of these partonswith the medium will result in modifications of internal jet structure in Au+Au eventscompared to that observed in the p+p/d+Au reference. Full jet reconstruction isa promising tool to measure these effects without the significant biases present inmeasurements with high- p T hadrons.One of the most significant challenges for jet reconstruction in the heavy ionenvironment comes from the correct characterization of the background fluctuations.The jet momentum irresolution due to background fluctuations has to be understoodin order to recover the correct jet spectrum. Recent progress in jet reconstructionmethodology is discussed, as well as recent measurements from p+p, d+Au and Au+Aucollisions at √ s NN = 200 GeV.
1. Introduction
Jets are remnants of hard-scattered partons, which are the fundamental objects ofperturbative QCD. At Relativistic Heavy Ion Collider (RHIC), they can be used as aprobe of the hot and dense matter created in heavy ion collisions. Interaction and energyloss of energetic partons in the medium lead to jet quenching in heavy ion collisions.Until recently, jet quenching was studied indirectly using single particle spectra anddi-hadron correlations [1]. These measurements are however limited in the sensitivityto probe partonic energy loss mechanisms due to biases toward hard fragmentation andsmall energy loss [2].Developments in theory (for example [3, 4]) and experiment (detector upgrades,increased RHIC luminosity) finally enabled full jet reconstruction in heavy ioncollisions [5]. Full jet reconstruction reduces the biases of indirect measurements andenables access to qualitatively new observables such as energy flow and fragmentationfunctions. As a baseline measurement for heavy ion jet studies, p+p collisions atthe same energy are used. To isolate initial state effects from medium modification,measurements in d+Au are essential.We present current jet analyses at STAR, starting with recent results on initialstate effects (d+Au). Status of jet spectra analysis in Au+Au follows, including studies et studies with STAR at RHIC
2. Jet reconstruction
The present analysis is based on √ s NN = 200 GeV data from the STAR experiment,recorded during 2006-2008. The Barrel Electromagnetic Calorimeter (BEMC) detectoris used to measure the neutral component of jets, and the Time Projection Chamber(TPC) detector is used to measure the charged particle component of jets. In thecase of a TPC track pointing to a BEMC tower, its momentum is subtracted from thetower energy to avoid double counting (electrons, MIP and possible hadron showers inthe BEMC). Pseudorapidity acceptance for jets is | η | < . | η | < .
55 in the case of d+Au collisions.Recombination jet algorithms k T and anti − k T , part of the FastJet package [3], areused for jet reconstruction. To subtract the background, a method based on active jetareas [4] is applied event-wise: p RecT = p CandidateT − ρ · A , with ρ estimating the backgrounddensity per event and A being the jet active area.An important aspect of underlying event background are its fluctuations. Wediscuss data-driven methods used to correct the jet observables for these fluctuations.
3. Initial state: d+Au
This analysis is based on minimum bias triggered √ s NN = 200 GeV data from theSTAR experiment, recorded during RHIC run 8 (2007-2008). The Beam Beam Counterdetector, located in the Au nucleus fragmentation region, was used to select the 20%highest multiplicity events in d+Au collisions. 10M events after event cuts were used forjet finding (anti − k T algorithm) with a resolution parameter R = 0 . p T > . /c cut was applied to tracks and towers.PYTHIA 6.410 and GEANT detector simulations (adjusted to match the realisticTPC tracking efficiency in d+Au run 8 running) were used for jet corrections to hadronlevel. Embedding into real d+Au events at level of reconstructed tracks and towers wasused to correct for background fluctuations. A bin-by-bin correction was applied to thejet spectrum [6].To compare the per event jet yield in d+Au to jet cross section measurements inp+p collisions, MC Glauber studies were utilized: h N bin i = 14 . ± . σ inel , pp = 42 mb. These factors were used to scale thep+p jet cross section measured previously by the STAR collaboration [7] using a MidPoint Cone (MPC) jet algorithm with a cone radius of R = 0 .
4. The resulting d+Au jet p T spectrum is shown in Figure 1 together with the scaled p+p jet spectrum. Withinthe systematic uncertainties, the d+Au jet spectrum scales with h N bin i .The leading systematic uncertainty is the Jet Energy Scale (JES) that is driven et studies with STAR at RHIC R AA and R CP for jets, respectively.
4. Inclusive jet spectra and background fluctuations in Au+Au
Preliminary results on jet p T spectrum in Au+Au collisions at √ s NN = 200 GeVwere reported in [8]. In this analysis, the background fluctuations were estimated bygenerating PYTHIA jets and embedding them into real central Au+Au events. Theresulting spectrum distortion was parametrized by Gaussian, for R = 0 . σ = 6 . ± R AA is shown in Figure 2. The systematic uncertainties prevent us from preciselyquantifying the suppression for R = 0 . R = 0 . R AA ≈ . R = 0 . R = 0 . (GeV/c) T jet p
12 14 16 18 20 22 24 26 28 30 32 ] - [( G e V / c ) T dp h d N d ev t N p -7 -6 -5 > uncertainty bin 200 GeVSTAR Preliminary Figure 1. Jet p T spectrum fromd+Au collisions [6] compared to h N bin i scaled p+p spectrum [7]. Figure 2. R AA of jets incentral Au+Au collisions for k T and anti − k T algorithmsand R = 0 . , . Precise characterization of underlying event background fluctuations is essentialto reduce systematic uncertainties in jet measurements. These are hence a subjectof intense study, both theoreticlly [9] and experimentally. We summarize here recentresults of STAR studies of background fluctuations [10].To quantify the background fluctuation, a method is used where a probe “jet”(single particle, PYTHIA jet, QPYTHIA jet) is embedded into real central Au+Au et studies with STAR at RHIC p T = p embedT and apply jet reconstruction on the hybrid event(anti − k T algorithm with R = 0 . p T to the probe jet and quantify the response of the hybrid system to theembedded jet via: δp T = p recoT − ρ · A reco − p embedT , (1)where A reco is the area of the matched reconstructed jet and ρ is determined priorto the embedding step. This definition is identical to Eq. (1) in [9]. The normalizeddistribution of δp T is the probability distribution to find jet energy (after event-wisebackground correction) p corr T = p true T + δp T . If there were no background fluctuations, δp T would be a delta function at zero. For very low p T probes, areas of anti − k T jets getvery small, so a cut A reco > . δp T distribution turns outto be largely independent of p embedT [10].We have investigated dependence of δp T on jet fragmentation pattern. Figure 4shows the overlay of multiple δp T distributions for single particle jets and for jets withboth low and high p T generated by PYTHIA and Q-PYTHIA (ˆ q = 5 GeV / fm). Inorder to compare their shapes directly, the distributions were aligned horizontally byfitting a Gaussian function to δp T < δp T distribution is to a largeextent universal, within a factor ∼ δp T = 30 GeV, especially in region δp T > 5. Jet triggered correlations A highly biased jet population was used as trigger in di-jet and jet-hadron correlations.Trigger jets are required to contain a BEMC tower with E T > . p T > /c . A 2 GeV systematic uncertainty on trigger jet energy was used toaccount for any remaining effect.Recoil jet p T spectrum was measured in p+p and Au+Au (0-20% most central)collisions [11]. A Gaussian model of background fluctuations was used to unfold theAu+Au spectrum with systematic uncertainty ± p T spectrum in Au+Au compared to p+p for R = 0 . 4, whichsuggests jet broadening beyond R = 0 . 4. Considering also the observation from inclusivejet analysis (suggestive of jet broadening from R = 0 . R = 0 . 4) there appearsto be a smooth jet broadening trend. Note that the recoil jet p T spectrum is much et studies with STAR at RHIC Figure 3. Ratio of R =0 . /R = 0 . p T spectrain p+p and Au+Au colli-sions [8]. Figure 4. Quantifying the back-ground fluctuations and their depen-dence on probe p T and fragmenta-tion [10]. flatter (harder) than the inclusive one, therefore impact of uncertainties on backgroundfluctuations is much reduced.Full jet reconstruction is not feasible for R > . v modulated background (withfixed v values). The uncertainties in the (a priori unknown) jet v value were chosen tocover the extreme cases of no v and 50% higher than v { } at p T = 6 GeV /c (defaultis v { } at p T = 6 GeV /c ). The associated track v values and uncertainties follow theanalysis in [13]. Due to ambiguities of ZYAM for broad jet structures, the backgroundlevel was determined by the fit. For comparison ZYAM was applied (as expected forbroad structures it leads to an underestimation of the correlated away-side yields forlower associated p T ).Figure 6 shows the awayside Gaussian width of JH in p+p and 0 − 20% mostcentral Au+Au collisions. There is a significant broadening (Au+Au w.r.t. p+p) for p assoc T < /c , while no broadening at higher p assoc T is observed. I AA , the ratio ofper-trigger associated yields, is plotted in Figure 7. There is a significant suppressionof high p T particles on the away side accompanied by an enhancement at low p assoc T . Inorder to quantify the energy redistribution on the away side, it’s better to instead of I AA use D AA : D AA ( p assoc T ) = p assoc T · ( Y AA ( p assoc T ) − Y pp ( p assoc T )) , (2)where Y AA , pp are per-trigger associated yields in AA,pp. Away side D AA for JH isshown in Figure 8. In fact, the energy “lost” at high p T is approximately compensatedby low p T enhancement [14]: jet quenching in action. et studies with STAR at RHIC p assoc T < /c , no broadening for high p assoc T and I AA shape independent of p assoc T at high p assoc T , one can speculate that the originalparton loses energy by emission of soft radiation (and therefore the original jet directionchanges little: no broadening is observed at high p assoc T ). These soft fragments traversethe medium, receive transverse kicks and therefore appear at large angles with respectto the original parton direction. The energy loss is followed by a possibly vacuum-likefragmentation of a parton with reduced energy. [GeV/c] T,jet p15 20 25 30 35 A u A u / pp (trig)>20 GeV T =0.2 GeV p T,cut R=0.4 p (trig)>20 GeV T =2 GeV p T,cut R=0.4 p 2 GeV Trigger unc. – STAR preliminary Figure 5. Ratio of recoil jet p T spectra in Au+Au/p+p [11]. Figure 6. Away side Gaussianwidth in JH correlations [12]. Figure 7. Away side I AA in JHcorrelations [12]. Figure 8. Away side D AA in JHcorrelations [12]. 6. Summary We have presented STAR results on full jet reconstruction in d+Au collisions. Withincurrent systematics there appears to be binary collision scaling compared to p+pcollisions, but the final measurements (jet R AA and R CP ) with reduced systematicuncertainties are yet to be completed. The study of background fluctuations δp T inAu+Au collisions suggests its independence of p embedT and the probe jet fragmentationpattern. The shape of δp T will be used to unfold the measured jet p T spectrum to obtainthe final result with decrased systematic uncertainties. The hints of jet broadeningobtained first in inclusive jet p T spectrum and di-jet correlations were further studiedusing jet-hadron correlations, where significant broadening and enhancement at low et studies with STAR at RHIC p assoc T was accompanied by suppression (and no broadening) at high p assoc T . This isconsistent with the picture of pQCD-like energy loss to soft fragments followed by(possibly vacuum-like) fragmentation of a parton with reduced energy. Acknowledgement This work was supported in part by grants LC07048 and LA09013 of the Ministry ofEducation of the Czech Republic. References [1] I. Arsene et al , Nucl. Phys. A757 , 1-27 (2005). B. B. Back et al , Nucl. Phys. A757 , 28-101 (2005).J. Adams et al , Nucl. Phys. A757 , 102-183 (2005). K. Adcox et al , Nucl. Phys. A757 , 184-283(2005).[2] T. Renk, Phys. Rev. C78 , 034904 (2008).[3] M. Cacciari, G. Salam, Phys. Lett. B641 , 57 (2006). M. Cacciari, G. Salam, G. Soyez, JHEP , 063 (2008).[4] M. Cacciari, G. Salam, Phys. Lett. B659 , 119 (2008). M. Cacciari, G. Salam, G. Soyez, JHEP , 005 (2008).[5] J. Putschke (for the STAR Collaboration), Eur. Phys. J. C61 , 629-635 (2009). S. Salur (forthe STAR Collaboration), Eur. Phys. J. C61 , 761-767 (2009). M. Ploskon (for the STARCollaboration), Nucl. Phys. A830 , 255c-258c (2009). E. Bruna (for the STAR Collaboration),Nucl. Phys. A830 , 267c-270c (2009).[6] J. Kapitan (for the STAR Collaboration), Nucl. Phys. A855 , 412-415 (2011).[7] B. I. Abelev et al , Phys. Rev. Lett. , 252001 (2006).[8] M. Ploskon (for the STAR Collaboration), Nucl. Phys. A830 , 255c-258c (2009).[9] M. Cacciari, J. Rojo, G. Salam, G. Soyez, Eur. Phys. J. C71 , 1539 (2011).[10] P. Jacobs (for the STAR Collaboration), Nucl. Phys. A855 , 299-302 (2011).[11] E. Bruna (for the STAR Collaboration), arXiv:1010.3184.[12] J. Putschke (for the STAR Collaboration), Nucl. Phys. A855 , 83-91 (2011).[13] B. I. Abelev et al , Phys. Rev.