Measurements of jet structure and fragmentation from full jet reconstruction in heavy ion collisions at RHIC
aa r X i v : . [ nu c l - e x ] O c t Measurements of jet structure and fragmentation from full jetreconstruction in heavy ion collisions at RHIC
Elena Bruna for the STAR Collaboration
Physics Department, Yale University, New Haven, CT 06520, U.S.A
Abstract
Measurements of inclusive hadron suppression and di-hadron azimuthal correlations have pro-vided important insights into jet quenching in hot QCD matter. However, they do not provideaccess to the energy of the hard scattering and are limited in their sensitivity since they can bea ff ected by biases toward hard fragmentation and small energy loss. Full jet reconstruction inheavy-ion collisions enables a complete study of the modification of jet structure due to energyloss, but is challenging due to the high-multiplicity environment. We present measurements offully reconstructed di-jets at mid-rapidity in 200 GeV p + p and central Au + Au collisions per-formed with the STAR detector. We compare fragmentation functions measured in 200 GeVp + p and central Au + Au collisions and assess the systematic uncertainties of their ratio.
1. Introduction
There has been recent significant progress in high-p T physics at RHIC with the success offull jet reconstruction in heavy-ion collisions [1, 2]. Unbiased reconstruction of jets providesexperimental access to the initial hard scattering, independent of the presence of the nuclearmedium. Therefore, a binary scaling of jet production from p + p to Au + Au is expected. Insightinto the jet’s structure is provided by studying fragmentation functions (FF). In-medium soft-ening of the FF with respect to p + p reference measurements, i.e. a distortion of the so called“hump-backed” plateau [3, 4], should be observable in Au + Au with an unbiased jet population.The large background and its fluctuations make full jet reconstruction a challenge in the high-multiplicity environment at RHIC. We performed jet measurements in STAR with the modernjet-finding techniques and utilizing data-driven correction schemes.
2. Experimental techniques and analysis
The STAR sub-detectors used for jet reconstruction are the Time Projection Chamber (TPC)for charged particles and the Barrel Electromagnetic Calorimeter (BEMC) for the neutral energy.Both TPC and BEMC have full azimuthal coverage and pseudo-rapidity acceptance | η | < + p year 2006 and 0-20% mostcentral Au + Au year 2007 events. Both data sets were selected with an online High-Tower (HT)trigger in the BEMC which requires the transverse energy in a tower to be E T > . η − φ plane. The analysis reported in these proceedings utilizes the “anti- k t ” algorithm [5], Preprint submitted to Nuclear Physics A May 30, 2018 hich is a recombination algorithm and part of the FastJet package [6]. Like all recombinationalgorithms, “anti- k t ” is collinear and infrared safe. “Anti- k t ” is used in this analysis because it isexpected to be less sensitive to background e ff ects in heavy-ion collisions.The di-jet analysis presented here is performed on events with a “trigger” jet, which matchesthe online triggered tower in the BEMC, and a “recoil” jet on the away side of the trigger jet(i.e. ∆ φ ∼ π ). A critical issue in jet analyses in Au + Au is the background. FastJet providesan estimate of the background p t per unit area, that is subtracted to get the jet component p jett , rec .The background energy is of the order of 45 GeV in a cone radius of R = R = p ∆ φ + ∆ η ).The background exhibits significant fluctuations in a central Au + Au event. We parameterize theupwards fluctuations of the background by a Gaussian with width σ of the order of 6-7 GeV [1,7]. A resolution parameter R = . ∼
80% of the jet energy lies within R = + p jets [8]).The background fluctuations can be suppressed by requiring a minimum transverse momentum( p cutt ) for a particle to be included in the jet. For the trigger jet a p cutt = / c was applied inorder to have the same energy scale in p + p and Au + Au. A p cutt introduces a strong bias in the jetpopulation, therefore the recoil jets are reconstructed with a minimal p cutt = .
15 GeV / c due tothe TPC acceptance. The requirement of p jett , rec ( trigger ) >
10 GeV / c for the reconstructed triggerjet minimizes the contribution of “fake” jets. Other sources of background in di-jet analysesare (a) “fake” jets and (b) hard scatterings uncorrelated to the di-jet. Additional hard scatteringsuncorrelated to the one that produced the di-jet pair frequently occur in heavy-ion collisions. Thebackground di-jet coincidence rate is estimated from the spectrum of associated jets located at ∼ π/ + p and Au + Auevents is reported in [7].
3. Di-jet coincidence rate and recoil fragmentation functions
The data shown in these proceedings were not corrected for the trigger e ffi ciency and forthe di ff erence in tracking e ffi ciency between p + p and Au + Au. An absolute correction to getthe parton energy will be performed in the future. A data-driven unfolding of the backgroundfluctuations from the Au + Au di-jet spectra and FF of recoil jets was applied to allow a directcomparison of p + p and Au + Au. The unfolding procedure removes the artificial hardening ofthe spectra due to the convolution of the background fluctuations with the steeply falling jetspectrum.The recoil jet spectra in p + p and Au + Au normalized to the number of trigger jets are shownin Fig. 1 (left). The ratio of the di-jet spectra of Au + Au to p + p is reported in Fig. 1 (right), indi-cating a strong suppression of recoil jets in Au + Au with respect to p + p at a given reconstructedjet energy. This is in constrast with the expected value of unity for unbiased jet reconstruction.We also report measurements of the fragmentation functions for recoil jets and compare theresults in p + p and Au + Au. In contrast to trigger jets, recoil jets are not a ff ected by the triggerbias that artificially enhances the neutral energy component in the jet firing the trigger. In orderto include the soft jet fragments, the FF are measured from charged hadrons in a radius of 0.7around the jet axis, while the reconstructed jet energy is from R = + Au is estimated on an event-by-event basis from the charged particle spectrumin the area outside the two jets with the highest reconstructed energy.2 recoil) [GeV/c] t,rec p ( r ec o il ) t , r ec ( T r i gg e r ) d N / dp j e t / N -6 -5 -4 -3 -2 -1 p+p recoilAuAu fake corrAuAu unfolded STAR Preliminary
Au+Au (0-20%) 200 GeVp+p 200 GeVAntiKt R=0.4>10 GeV/c trigt p (recoil) [GeV/c] t,rec p
10 15 20 25 30 35 40 A u A u ( - % ) / pp >10 GeV/c trigt AntiKt R=0.4, p 1 GeV ± bkg σ Background uncertainty Trigger jet energy uncertainity
STAR Preliminary
Au+Au (0-20%) 200 GeVp+p 200 GeV
Figure 1: Left plot: p t spectra of recoil jets (for p trigt >
10 GeV / c) in p + p (close circles) and in Au + Au after the correctionfor fake jets (open triangles) and unfolding of background fluctuations (close triangles). The spectra are normalized tothe number of trigger jets. No corrections for trigger and tracking e ffi ciency were applied. Right plot: ratio of p t spectraof recoil jets in Au + Au (corrected for fake jets and background fluctuations) to p + p. The curves indicate the systematicuncertainties in the estimation of the background fluctuations (solid) and of the p t of the trigger jet assuming a jet- p t resolution of 25% (dotted) [8]. The FF of reconstructed recoil jets are shown for p recoilt , rec ( AuAu ) >
25 GeV / c (Fig. 2). The p + pjets corresponding to p recoilt , rec ( AuAu ) >
25 GeV / c are selected taking the background fluctuationsinto account. The mean jet p t of the selected p + p jets is p recoilt , rec ( pp ) ≃
25 GeV / c. The resultsof the unfolding procedure applied to the z distributions ( z = p hadront / p recoilt , rec ) are shown in Fig. 2(left). The FF are harder since the artificial hard contribution in the jet spectrum due to thebackground fluctuations has been removed after the unfolding. The ratio of the z distributionsof Au + Au to p + p is shown in Fig. 2 (right). No significant modification of the fragmentationfunctions of recoil jets is observed with p recoilt , rec ( AuAu ) >
25 GeV / c for z > ∼ .
2, in contrast to theexpectation of a softening of the FF for an unbiased jet population and compared to the measuredsuppression of high-p t hadrons at RHIC ( R AA ∼ . z part of the fragmentationfunction ( z < ∼ .
1) is still under investigation since it is dominated by the underlying backgroundin Au + Au and therefore might be a ff ected by a large uncertainty due to background subtraction.
4. Discussion of the results
Due to the strong trigger bias in the events used in this analysis, a particular class of di-jetswas selected. Indeed, the trigger jets seem to be unmodified and therefore most likely come fromthe surface of the medium, which would bias the recoil jets towards a maximum in-medium pathlength and to a maximum energy loss e ff ect.Given this “extreme” selection of recoil jets, we reported a significant suppression of di-jetcoincidence rates in central Au + Au with respect to p + p at a given reconstructed jet energy. Inaddition, we observed the absence of strong modification of the FF of recoil jets in 0-20% Au + Auwith p recoilt , rec ( AuAu ) >
25 GeV / c with respect to p + p.The above observations can be explained via a scenario where the jet is broadened and its en-ergy is not fully recovered in R = . + p with the current jet-finding algorithms.In this case a shift of the recoil spectrum towards smaller energies would be expected (and wouldexplain the results in Fig. 1). A softening of the FF is expected in Au + Au to account for the mea-sured high-p t hadron suppression. In a jet broadening scenario, the observed absence of a strongmodification of the measured FF in Au + Au (Fig. 2, right) could be due to an underestimation3 recoil) t,rec /p hadront =p rec z r ec d N / d z j e t / N -3 -2 -1 STAR Preliminary p+p recoil0-20% Au+Au recoil0-20% Au+Au recoil unfolded
Au+Au (0-20%) 200 GeVp+p 200 GeVAntiKt R=0.4>10 GeV/c trigt p (AuAu)>25 GeV/c recoilt p (recoil) t,rec /p hadront =p rec z A u A u ( - % ) / pp -1 AntiKt R=0.4 1 GeV ± bkg σ Background uncertainty Trigger jet energy uncertainty 20% ± Fake uncertainity
STAR Preliminary
Au+Au (0-20%) 200 GeVp+p 200 GeV(AuAu)>25 GeV/c recoilt p >10 GeV/c trigt p Figure 2: Left plot: FF for recoil jets (normalized to the number of recoil jets) in p + p, Au + Au (open triangles), Au + Auafter unfolding for p recoilt , rec ( AuAu ) >
25 GeV / c and p trigt , rec >
10 GeV / c. Right plot: ratio of the z distributions measuredin 0-20% central Au + Au events to p + p collisions for p recoilt , rec ( AuAu ) >
25 GeV / c. The curves indicate the systematicuncertainties in the estimation of the background fluctuations (solid) and of the trigger jet energy (dotted). The uncertaintydue to the estimation of fake jets (small dotted) was assumed to only a ff ect the normalization of the z distributions. of the energy in R = = + Au would be those of the surviving recoil jets, i.e. the ones that only mini-mally interact without being absorbed, such as the ones emitted tangentially to the surface of themedium. To test this scenario, the di-jet coincidence rates for di ff erent radii will be measuredand their ratio should be the same in p + p and Au + Au.The corrections and systematic error bands reported do not yet extensively assess the system-atic uncertainties in the di-jet analysis. Furthermore, detailed systematic studies (i.e. trackinge ffi ciency, calorimeter calibration, missing neutral energy, etc.) are being conducted in order toaccess the kinematics of the hard scattering. The jet-energy profile will also be investigated inorder to improve our understanding of the mechanism of jet broadening in the medium. References [1] J. Putschke for the STAR Collaboration, Hard Probes 2008 Proceedings, arXiv:0809.1419.[2] S. Salur for the STAR Collaboration, Hard Probes 2008 Proceedings, arXiv:0809.1609.[3] S. Sapeta and U. A. Wiedemann, Eur. Phys. J. C 55, 293 (2008), arXiv:0707.3494.[4] N. Borghini and U. A. Wiedemann, arXiv:hep-ph /0506218v1.[5] M. Cacciari, G. Salam, G. Soyez, arXiv:0802.1189.[6] M. Cacciari, G. Salam, G. Soyez, arXiv:0802.1188.[7] M. Ploskon for the STAR Collaboration, Quark Matter 2009 Proceedings, this volume.[8] H. Caines for the STAR Collaboration, Quark Matter 2009 Proceedings, this volume.[9] J Adams et al., STAR Collaboration, Nucl. Phys. A 757 (2005) 102.