JET QUENCHING AT LHC

Martin Spousta

Institute of Particle and Nuclear Physics,

Charles University in Prague, Czech Republic

martin.spousta@cern.ch

July 10, 2013

Abstract

We review up-to-date results on high-pT particles and jets in heavy ion collisions by threemajor LHC experiments, ALICE, ATLAS, and CMS. Results of analyses of 2010 and 2011Pb+Pb data at

√sNN = 2.76 TeV are discussed. We concentrate mainly on results by fully

reconstructed jets and discuss similarities and important differences in measurements amongexperiments. We point to the importance of understanding the results in a view of differencebetween quark-initiated and gluon-initiated jets.

1 Introduction

Collisions of heavy ions at ultra-relativistic energies can produce hot and dense colored mediumwhere relevant degrees of freedom are not complex hadrons but deconfined quarks and gluons [1].This consequence of asymptotic freedom of quantum chromodynamics (QCD) allows to investigatefundamental properties of strong interaction as originally suggested in Refs. [2, 3, 4]. High trans-verse momentum (pT) partonic interactions in perturbative quantum chromodynamics (pQCD)lead to a production of two highly virtual back-to-back partons (in the second order of pQCD)which subsequently evolve as parton showers, hadronize, and are experimentally observed as back-to-back di-jet events in the detector. If the partons traverse on their path a dense colored mediumthey can loose energy. The result of the energy loss can be detected as modifications of jet yieldsand jet properties. This phenomenon is commonly referred to as the jet quenching. The jetquenching was first proposed by Bjorken [5] as an experimental tool to investigate properties ofthe dense medium.

This review summarizes experimental results on the jet quenching from three major exper-iments at Large Hadron Collider (LHC), the ALICE [6], ATLAS [7], and CMS [8] experiment.The LHC heavy ion program follows up the successful research in the field of ultra-relativisticheavy ion collisions performed at Super Proton Synchrotron (SPS) and later at Relativistic HeavyIon Collider (RHIC). The results of four experiments at RHIC – BRAHMS, PHENIX, PHOBOS,and STAR have been summarized in Refs. [9, 10, 11, 12] and further discussed along with resultsfrom SPS e.g. in reviews [13, 14, 15, 16, 17, 18]. The increase of the collisional energy by anorder of magnitude at LHC compared to RHIC allowed collecting high statistics samples of fully

1

arX

iv:1

305.

6400

v2 [

hep-

ex]

9 J

ul 2

013

reconstructed jets, high-pT photons, and intermediate vector bosons W±, Z0. These hard probescan shad a new light on the jet quenching and properties of the medium.

There are handful of experimental results at low and intermediate transverse momentum1

(pT . 20 GeV) by all three LHC collaborations such as measurements of global event properties,particle correlations, and identified particle yields. Also, there are important measurements onquarkonia suppression provided by LHC experiments. We will not discuss these results sinceprimary focus of this review is the jet quenching. For extensive review of these topics at LHC seee.g. Refs. [19, 20].

The recent development in the theory of the jet quenching is summarized e.g. in Refs. [21, 22]or reviews [23, 24, 25]. No unified description of the jet quenching exists. There are differentphenomenological approaches [26, 27, 28, 29, 30] to the calculation of jet quenching, however theunderlying principle remains the same: hard partons can loose their energy via gluon radiation(an analogue of LMP effect in Quantum Electrodynamics [31, 32]) and/or by elastic (collisional)energy loss [5] which is of importance namely for the heavy quark energy loss [33]. Different modelsnot only use different techniques and approximations but also they characterize the medium interms of different primary model parameters. There is therefore a great importance in performingexperimental measurements of jets and other high-pT probes to help to constrain existing models.The first experimental evidence of the jet quenching at LHC has been provided by the asymmetrymeasurement discussed in the next section.

2 The first evidence of the jet quenching at LHC

The first experimental evidence of the jet quenching at LHC has been observed in the measurementof the di-jet asymmetry [37, 38]. The di-jet asymmetry has been defined as

AJ =ET,1 − ET,2

ET,1 + ET,2(1)

where ET,1 resp. ET,2 is the transverse energy of the leading resp. subleading jet in the event.Energy loss of parent partons in the created matter may reduce or “suppress” the rate for

producing jets at a given ET. This suppression is expected to increase with medium temperatureand with increasing path length of the parton in the medium. As a result, there should be moresuppression in central Pb+Pb collisions which have nearly complete overlap between incidentnuclei, and little or no suppression in peripheral events where the nuclei barely overlap. This wasindeed observed in the measurement of di-jet asymmetry where the jet suppression exhibits itselfby an increase of the number of events with larger jet asymmetry compared to Monte-Carlo (MC)reference.

This asymmetry has been accompanied by a balance in azimuth, that is jets in the di-jet systemremain “back-to-back” despite to a sizable modification of their energy. The original [37, 38] andupdated [39, 40] experimental observation was followed by theoretical work [41, 42, 43, 44, 45, 46]suggesting that the suppression can be explained as a consequence of the partonic energy loss inthe hot and dense QCD medium.

1Throughout this paper, natural units are applied with c = ~ = 1.

2

3 Jet reconstruction techniques

Before discussing further details of jet quenching measurements at LHC, we briefly summarize thejet reconstruction techniques of LHC experiments. All three LHC experiments use the anti-kT jetfinding algorithm [47] with the distance parameter R varying from 0.2−0.6. ALICE uses trackingplus π0s reconstructed in the electromagnetic calorimeter for the jet finding [49]. ATLAS typicallyuses the calorimeter information only to define jets [39]. CMS uses the particle-flow algorithmwhich combines the information from the tracking and calorimeter [48]. All three experimentsperform the subtraction of the underlying event (UE) present underneath the jet reconstructedin heavy ion collisions. The procedure of UE subtraction is based on, or similar to, techniquesdescribed in Refs. [50, 51, 52]. A typical transverse energy of the underlying event in the 0-10% most central heavy ion collisions per area of R = 0.4 jet exceeds 100 GeV [53, 54] whichis often above the measured jet energy. A careful evaluation of the jet energy scale (linearity),jet energy resolution, and jet reconstruction efficiency is therefore needed in order to verify theperformance of the jet reconstruction as stressed by theorists after the asymmetry observationmeasurement [55].

Several benchmarks summarizing the jet performance in heavy ion events of three LHC ex-periments are summarized in the table on Fig. 1 based on Refs. [53, 56, 57, 54, 49, 58]. Usualtechniques to determine the jet performance use the embedding of PYTHIA [59] jets into theminimum bias heavy ion background which is either modeled by HIJING [60] and HYDJET [61]MC generators or, optimally, taken directly from the data. The full detector simulation of theseevents is then performed. The full simulation of embedded events and data-driven tests of jetperformance discussed further represent a reliable tools to access the impact of large UE on thejet reconstruction2. Attempts to quantify the impact of UE subtraction on jet reconstruction us-ing toy models of detector response [55, 63] are fruitful but not always straightforward to connectwith specific measurements due to the over-simplification of the detector response.

Beside MC evaluation of the jet performance, the data driven tests have been performed. Sys-tematic uncertainty due to the jet energy scale and jet reconstruction efficiency can be constrainedor reduced by matching calorimeter jets to track jets [64]. The ratio of transverse momentum ofjets reconstructed in the calorimeter to the transverse momentum of jets reconstructed in thetracking system can be compared between data and MC which helps determining systematicuncertainties [53]. Systematic uncertainty due to jet energy resolution can be constrained byevaluating fluctuations in the UE in minimum bias events and comparing them to fluctuations ofenergy in reconstructed jets [54, 65, 53]. For calorimeter jets, jet energy resolution is given by adirect sum of stochastic and constant term which reflect the intrinsic properties of the detector,and by the noise term which reflects the fluctuations of UE [66]. This allows for a data-drivencross-check of jet energy resolution [67]. Similar data driven tests have been performed for trackjets [54]. These data-driven tests allow to determine systematic uncertainties and help to verify

2 The conclusions on jet performance are reliable up to a possible influence of jet reconstruction by stronglymodified fragmentation [62]. We shall see that the modifications of fragmentation are moderate and we can thereforeexpect no large impact on performance. It is worth to stress that in-vacuum fragmentation is not unique but consistsof many configurations of jet properties some of which may be close to the jet properties of medium modified jets.Thus, we could evaluate the performance as a function of “fragmentation” that is e.g. as a function of leadingparticle momentum fraction to make sure that the performance of in-medium jets does not suffer from biases.

3

ALICE ATLAS CMS(charged + π0) (calorimeter jets) (particle-flow jets)

Jet energy scale 19% correctionat 100 GeV < 4% uncert. < 2% < 3%

(pp jets, R = 0.4) (0-10%, R = 0.4) (0-5%, R = 0.3)

Jet energy resolution 18% 16% 18%at 100 GeV (pp jets, R = 0.4) (0-10%, R = 0.4) (0-5%, R = 0.3)

Jet energy resolution–

26% 24%at 50 GeV (0-10%, R = 0.4) (0-5%, R = 0.3)

pT threshold for jet recon-– 50 GeV 40 GeV

struction efficiency > 95%

UE ET or pT 138 GeV 105 GeV–density per R = 0.4 jet (0-10%, (0-1%)

pchargedT > 150 MeV)

11 GeV

12.5 GeV –Size of UE fluctuations (pchargedT > 150 MeV)per R = 0.4 jet (0-10%) 5 GeV

(pchargedT > 2 GeV)

Figure 1: Benchmarks for the jet reconstruction performance in terms of jet energy scale, jetenergy resolution and jet reconstruction efficiency. Characteristics of UE in terms of mean trans-verse energy or momentum density of UE per area of R = 0.4 jet or mean size of UE fluctuationsare also indicated. The centrality selection or choice of the jet distance parameter R is indicatedin brackets. Values are based on plots or explicit numbers given in Refs. [53, 56, 57, 54, 58, 49].

4

the detector response. It is desirable for the experimental community to continue investigatingsuch methods since they directly lead to better precision of the measurement.

4 Inclusive jet suppression

The measurement of di-jet asymmetry indicates that observed modifications of jets are due to thejet quenching. The di-jet asymmetry is an observation measurement. Alone, it cannot providedetails on the parton energy loss since it is insensitive to a configuration where two jets in a di-jetpair loose comparable amount of energy. The measurement of inclusive jet spectra is therefore thefirst step towards extracting detailed features of the jet suppression. The observable quantities ofinterest are nuclear modification factors, jet RCP and jet RAA, which characterize the suppressionof jet in a given centrality interval with respect to peripheral events (RCP) or with respect to p+pjet spectra (RAA),

RcentCP (pT) =

〈N60−80coll 〉〈N cent

coll 〉N cent(pT)

N60−80(pT), Rcent

AA (pT) =1

〈N centcoll 〉

N cent(pT)

Npp(pT), (2)

where 〈Ncoll〉 is the average number of nucleon-nucleon collisions in a given centrality interval [68]and N(pT) is a per event jet yield.

The jet RCP has been measured by ATLAS [53] over the pT range of 38 − 210 GeV. The jetRAA has been measured by CMS [56] over the pT range of 100 − 300 GeV and by ALICE [49]using jets reconstructed from charged particles plus π0s over the pT range of 30− 120 GeV. Thejet yields measured in central collisions are observed to be suppressed by a factor of about two athigh-pT. Both ATLAS and CMS see that the suppression is consistent with flat pT-dependence.For pT < 100 GeV ALICE preliminary result shows a trend of a decrease of RAA with decreasingpT whereas the RCP of ATLAS remains rather flat. A comparison of nuclear modification factorsmeasured by different experiments is shown on Fig. 2.

ATLAS has reported a significant dependence of the inclusive jet suppression on the jet sizefor jets with pT < 100 GeV. For jets with pT > 100 GeV, only jets reconstructed with thedistance parameter of R = 0.5 might be different compared to jets reconstructed using smallerdistance parameters. This is consistent with the result by CMS which reports no dependence ofthe suppression on the jet size for R = 0.2− 0.4 jets with pT > 100 GeV. It is important to notethat any comparison based on jet reconstructed using different size parameters might introducebiases since the same jets but reconstructed with smaller radii will tend to populate a lower pTregion [53]. E.g. a typical 100 GeV jet reconstructed in p+p collisions with R = 0.4 deposits15% of its energy in the region of R > 0.2 [73]. More details on features seen in the nuclearmodification factor of jets is discussed in Sec. 10 of this review.

5 Jet suppression via measurements of single particles

The first observation of jet quenching at RHIC was based on a suppression seen in the nuclearmodification factor of charged particles [34, 35] and based on di-hadron azimuthal correlations [36].Later, it was found that the nuclear modification factor of single charged particles alone is insuf-ficient to distinguish different scenarios of microscopical interaction of partons with the QCD

5

[GeV]charged

Tp

0 20 40 60 80 100

Nuc

lear

mod

ifica

tion

fact

or

0

0.2

0.4

0.6

0.8

1 |<1.0η CMS, 0-5%, |AAR

|<0.8η ALICE, 0-5%, |AAR

|<2.5η ATLAS Preliminary, 0-5%/60-80%, |CPR

partN0 50 100 150 200 250 300 350 400

b-jet

frac

tion

0

0.01

0.02

0.03

0.04

0.05

0.06CMS preliminary

= 2.76 TeVNNs

< 100 GeV/cT80 < Jet p

-1bμ L dt = 150 PbPb DataPythia+HydjetSyst. uncertainty

50-100% 20-50% 0-20%

0 [GeV]Tp

0 2 4 6 8 10 12 14

CPMu

on R

0

0.2

0.4

0.6

0.8

1

1.2

0-10% / 60-80%10-20% / 60-80%20-40% / 60-80%40-60% / 60-80%

Pb+Pb = 2.76 TeVNNs

| < 1.05η|

-1bμ L dt = 7 ∫ATLAS Preliminary

Jet R

CP

Jet pT (GeV/c)

Anti-kT algorithm R=0.3CMS Prelim. 0-10%/50-90% Particle Flow JetALICE Prelim. 0-10%/50-80% Track JetATLAS 0-10%/60-80% Calo Jet

Figure 2: Left: Nuclear modification factor of charged particles by ALICE [69], ATLAS [70], andCMS [71]. Right: Nuclear modification factor of jets at LHC from Ref. [72].

medium [74, 22]. At LHC, the full jet reconstruction with large statistics of jets with high trans-verse momenta is available. Nevertheless, measurements of single particles at high-pT is stillimportant for constraining the energy loss models at LHC energies [75]. Moreover, it is crucial forunderstanding the first measurements of jets: the nuclear modification factor of charged particlesat high-pT together with the nuclear modification factor of jets put constraints to a suppression ofjet fragments at high-pT (or vice versa). If the electroweak processes are excluded, each chargedparticle at high-pT has to be connected with a hard scattering and jet production. The nuclearmodification factor of inclusive charged particles measured by ALICE [69], CMS [71], and AT-LAS [70] is shown on Fig. 2. One can see that the nuclear modification factor of charged hadronsat high transverse momentum, pT & 40 GeV, reaches approximately 0.5 − 0.6 which are similarvalues to those measured in the nuclear modification factor of jets discussed in Sec. 4. This impliesthat there should be no substantial modification of jet fragmentation at high-pT. This was indeedobserved in the jet fragmentation measurement which is discussed in Sec. 9. Three independentmeasurements – nuclear modification factor of single hadrons, nuclear modification factor of jets,and jet fragmentation functions – are therefore in a good agreement with each other and also ina good agreement among experiments.

Below pT ≈ 20 GeV the nuclear modification factor exhibits similar behavior as seen atRHIC [9, 10, 11, 12]. The minimum of RAA at LHC is reached at pT = 6− 7 GeV with the valueof 0.13− 0.14 which is about 50% less than at RHIC. The rise of the RAA for pT above 8 GeV ispredicted by most of the jet quenching models [76], however the magnitude of the predicted slopevaries greatly among models. For a comparison of RAA with different jet quenching models seee.g. Refs. [77, 69, 71].

The consistency of results on suppression of inclusive jet yields is further checked by the

6

[GeV]T

p20 40 60 80 100 120 140 160 180 200

2 v

0

0.02

0.04

0.06

0.08

0.1

220-30% CMS particle v

2

30-40% CMS particle v

2

20-30% ATLAS Preliminary jet v

2

30-40% ATLAS Preliminary jet v

Figure 3: Comparison of v2 at high-pT for two centrality bins. Results on charged particle v2 athigh-pT of CMS [82] are shown along with results on v2 of jets by ATLAS [83].

measurement of particle-yield modifications of jet-like di-hadron azimuthal correlations done byALICE [78] and CMS [79]. The modifications of di-hadron azimuthal correlations is quantified bythe IAA modification factor defined as a ratio of integrated per-trigger-particle associated yieldmeasured in Pb+Pb collisions to the yield measured in p+p collision. Both ALICE and CMSreport a suppression of the away-side yield by a factor of 0.5 − 0.6 in central Pb+Pb collisionswhich does not show any strong pT dependence. This is again consistent with the inclusive jetsuppression discussed in Sec. 4. Another interesting feature of the data on IAA is the fact that thenear-side yield is not suppressed. At first glance, this is suggestive of a “corona jet production” [79]that is the predominant jet production at the surface of the medium and strong absorption in theinner core [80]. No suppression of nearside jet yield might also be explained as a consequence ofnon-trivial interplay between the modification of jet fragmentation and different suppression ofquark and gluon jets (discussed further in Sec. 9) both of which can lead to a bias in the selectionof the trigger particle in Pb+Pb collisions [81].

6 Azimuthal dependence of the jet suppression

The measurement of azimuthal dependence of charged particle production at high-pT along withthe measurement of nuclear modification factor can help distinguishing the microscopic mecha-nisms of parton energy loss through determining the path length (l) dependence of the energyloss, ∆E ≈ lα [84, 85, 86, 87, 88, 89, 90, 91]. Measurement of azimuthal anisotropy of π0 withpT = 7− 10 GeV at RHIC supported a scenario of α = 3 based on AdS/CFT gravity-gauge dualmodeling [92, 93, 86, 88], while solely perturbative QCD calculations (α = 1 for collisional energyloss and α = 2 for radiative energy loss) seemed to under-predict the data [94]. Recently, it hasbeen shown that it may be problematic to distinguish the weakly coupled pQCD-based medium

7

and strongly coupled AdS/CFT-modeled medium based only on the path length dependence oflight quark jets [95]. Nevertheless, the measurement of the path-length dependence remains essen-tial for understanding the parton energy loss. To get more insight into the path-length dependenceof the jet suppression the azimuthal anisotropy of high-pT particles is measured by ALICE [96],ATLAS [97], and CMS [82]. Further, the azimuthal anisotropy of jets is measured by ATLAS [83].

The azimuthal anisotropy is characterized by the second-order Fourier harmonic coefficient(v2) in azimuthal angle (φ) which is a measure of hadron emission relative to the reaction planeangle (ΨRP) [98, 97, 99],

dN

d(φ−ΨRP)≈ 1 + 2v2 cos(2(φ−ΨRP)). (3)

Such azimuthal anisotropy at low transverse momenta ensues because the hadron yield is moresuppressed along the long axis of the almond shaped fireball than the short axis. Thus themagnitude of the v2 is sensitive to the path length dependence of energy loss [100]. The measuredv2 of charged particles increases at low pT achieving a maximum at pT ≈ 3 GeV and then graduallydecreases towards zero but remains positive to at least 40 GeV. For charged particles with pT inthe interval of 14− 48 GeV the v2 lies approximately between 0.02 and 0.08. Even in the high-pTregion, the v2 has a characteristic centrality dependence with v2 being generally larger in lesscentral collisions where the eccentricity of initial overlap region is larger compared to the mostcentral collisions. This result is consistent among all there experiments though CMS provides themeasurement up to pT ≈ 60 GeV, whereas ALICE and ATLAS only up to pT = 20 GeV.

The results on the v2 of high-pT charged particles are qualitatively consistent with resultson jet v2 by ATLAS which measures the jet v2 for pT = 50 − 210 GeV. The non-zero jet v2is distinguishable up to pT = 160 GeV. The variation of the jet yield with the distance to thereaction plane, ∆φ, is also characterized by the ratio of jet yields between the most out-of-plane(3π8 < ∆φ < π

2 ) and most in-plane (0 < ∆φ < π8 ) jet directions. This ratio shows as much as

20% variation between the out-of-plane and in-plane jet yields. A comparison of measurements ispresented on Fig.3.

7 Photon-jet and Z0-jet correlations

The fact that the measured jet suppression at LHC is not an effect of initial state modifications wasproven by the measurement of inclusive yields of prompt photons (γ) [101, 102] and inclusive yieldsof W±, Z0 vector bosons [103, 104, 105] which do not exhibit any suppression in central heavyion collisions. Not only no suppression as a function of pT is seen, but also no modification of therapidity distribution of produced vector bosons is observed which is also sensitive to modificationsof nuclear parton distribution functions [106].

Since photons and vector bosons are insensitive to the colored medium they can provide, on astatistical average, constraints on the energy of the away-side parton shower [107]. The γ-jet andZ0-jet events has indeed been studied by ATLAS [108, 109] and CMS [110] and the energy losshas been evaluated in a similar way as for the original asymmetry measurement [37, 38]. Insteadof the asymmetry, a more simple quantity has been evaluated – the mean fractional energy carriedby a jet, xJγ = pjetT /pγT or xJZ0 = pjetT /pZ

0

T . The conclusion from these measurements was the same

8

as from the original measurement of the di-jet asymmetry: a significant increase of events withlarge imbalance of transverse momentum in the γ-jet or Z0-jet system has been observed, whilethe ∆φ distribution remains with no modification.

Measurements of the jet suppression in the γ-jet or Z0-jet system provide a starting point fordetailed quantification of energy loss. In principal, by selecting a suitable range for the transversemomentum of the tagging photon accessible at different center-of-mass energies, the in-mediummodification of parton showers in the dense matter created at these different center-of-mass en-ergies can be studied. First theoretical study towards this goal were done in Ref. [111]. Besidestagging the energy of a quenched jet by the energy of γ or Z0, the γ-jet or Z0-jet system canprovide a tool to study the difference between quark and gluon energy loss. This is in principlepossible because the quark-to-gluon fraction of partons initiating the jets in these systems is onaverage different from the quark-to-gluon fraction of di-jets. Z0-jet and γ-jet systems are alsoconvenient from the experimental point of view since Z0 or γ is less sensitive to resolution effectscomparing to jets and therefore the unfolding of observables that depend on the kinematics ofboth, the jet and the vector boson, is less demanding than the unfolding of the di-jet observables.The clear disadvantage of these measurements are low yields due to smallness of electromagneticcoupling. Nevertheless, the Z0-jet and γ-jet measurements represent one of the golden channelsfor a long term heavy-ion program after the upgrade of LHC [112, 113].

8 Flavor dependence of the jet suppression

The quenching of jets in heavy-ion collisions is expected to depend on the flavor of the fragmentingparton. Jets initiated by heavy quarks are expected to radiate less than jets initiated by lightquarks. The gluon radiation of a heavy quark is suppressed at angles smaller than the ratio of theheavy quark mass to its energy. This is so-called dead-cone effect [114]. Consequently, the heavyflavor jets are expected to undergo a smaller suppression than the light quark jets. However,measurements of heavy quark production at RHIC via semi-leptonic decays to electrons showeda combined charm and bottom suppression in Au+Au collisions comparable to that observed forinclusive hadron production [115, 116, 117]. There is disagreement in the theoretical literatureregarding the interpretation of the RHIC heavy quark suppression measurements [87, 118, 119, 120]particularly regarding the role of non-perturbative effects [121, 122, 123].

To access the difference between the jet suppression of heavy quarks and light quarks CMSmeasures the b-tagged jets [124], ALICE and ATLAS measure the semi-leptonic decays of openheavy flavor hadrons into muons [125, 126]. The measurement of b-jets by CMS uses the secondaryvertex mass distribution to tag the jets initiated by a bottom quark. The main quantity extractedfrom the measurement is the bottom quark jet to inclusive jet ratio. This b-jet fraction is measuredin the range of 80 < pjetT < 200 GeV both in Pb+Pb and p+p data at

√sNN = 2.76 TeV. The

extracted b-jet fraction lies in the range of 2.9 − 3.5% and is comparable between Pb+Pb andp+p on one side and Pb+Pb and MC on the other side. The b-jet fraction as a function of pjetT

does not exhibit any centrality dependence though the systematic uncertainty is rather large. Themeasured b-jet fraction is compatible with approximately half a percent decrease with increasingpjetT predicted by PYTHIA for the pjetT range covered by the measurement. These observationsimply that the b-jet RAA (which can be defined as a product of b-jet fraction in Pb+Pb and p+p)

9

and the inclusive jet RAA exhibits the similar behavior as the inclusive jet RAA: suppression by afactor of two, possibly only a modest dependence on pjetT . Although, it needs to be stressed thatthe systematic and statistical uncertainties on that measurement are large.

The measurement of semi-leptonic decays of open heavy flavor hadrons by ATLAS [125] isperformed in the midrapidity (|η| < 1.05) and over the muon transverse momentum range 4 <pT < 14 GeV. The measurement by ALICE [126] is performed in forward rapidity (2.5 < |η| < 4)and over the muon transverse momentum range 4 < pT < 10 GeV. Over these pT ranges, muonproduction results predominantly from a combination of charm and bottom quark semi-leptonicdecays. ATLAS compares the differential yields of muons in a given centrality and pT bin to yieldsmeasured in peripheral collisions via central-to-peripheral ratio, RCP, whereas ALICE evaluatesthe difference in terms of RAA that is with respect to the p+p reference measurement. Bothexperiments see a smooth decrease of nuclear modification factor with increasing centrality. Thesuppression does not change with the muon pT while the size of the suppression changes by nearlya factor of two between the most central and the most peripheral collisions (ATLAS), resp. bya factor of 3 − 4 between the yields measured in the most central Pb+Pb collisions and yieldsmeasured in p+p (ALICE). The results of ATLAS measured in the midrapidity region can be alsocompared with results of CMS measurement of high-pT non-prompt J/ψ (pT = 6.5 − 30 GeV)produced in the decay of b-hadrons [127]. The result of CMS evaluated in terms of RAA in the0-20% collisions are consistent with the RCP of ATLAS. However, opposed to ATLAS, CMS doesnot report any centrality dependence of RAA of non-prompt J/ψ; the result in 20-100% centralitybin is consistent with the result in central 0-20

Both measurements, the b-jet fraction and semi-leptonic decays of open heavy flavor seem tosuggest that there is no dramatic difference between the suppression of inclusive jets and jetsinitiated by b-quarks. Although, as already stated above, more precision is needed on the directb-jet measurement to make this conclusion stronger.

9 Jet internal structure

One of the tools that allow precise comparison between the data and theoretical models of jetquenching is the measurement of jet fragmentation. The key question to address is how arethe parton showers modified by the medium. Generally, there is no agreement in the theoryupon the answer to this question, however most of the theoretical models predict a softeningof fragmentation functions – suppression at large momenta of fragments, enhancement at smallmomenta of fragments, and broadening of a jet [128, 129, 130, 131, 132, 133, 134]. The softeningof fragmentation was first indirectly observed by CMS in the measurement of the missing-pT [38].In that measurement, the projection of missing-pT of reconstructed charged particles onto the

leading jet axis was calculated as 6p ‖T = ΣipT,i cos(φi − φleading jet) for each event. The event

averaged missing-pT, 〈6p ‖T〉, was then evaluated as a function of the di-jet asymmetry in threeconfigurations – inside the leading and subleading jet cone, outside the leading and subleading

jet cones, and inclusively. An in-cone imbalance of 〈6p ‖T〉 ≈ 20 GeV was found for di-jet eventswith the largest asymmetry both in MC and data. This in-cone imbalance is balanced by the

out-of-cone imbalance of 〈6p ‖T〉 ≈ 20 GeV again both for MC and data. However, in the data theout-of-cone contribution is carried almost entirely by tracks with 0.5 < pT < 4 GeV whereas in MC

10

more than 50% of the balance is carried by tracks with pT > 4 GeV, with a negligible contributionfrom pT < 1 GeV. The MC events with large di-jet asymmetry are due to the semi-hard initialor final state radiation. This implies that the missing-pT in those MC events is balanced byrather hard-pT particles which are very often clustered into a third jet. On the contrary, theresults for large di-jet asymmetry in the data show that a large part of the momentum balanceis carried by soft particles radiated at large angles to the jet axes. Another important featureseen in the measurement is that the overall momentum balance in the event is recovered whencalculating the missing-pT using all particles with pT > 0.5 GeV. This confirms that the large di-jet asymmetry is not due to instrumental effects or e.g. due to neutrino production3. The physicspicture obtained from missing-pT measurement was further confirmed by the measurement of jet-like di-hadron correlations by ALICE [78] and CMS [79] which we have also discussed in Sec. 5.In those measurements, the yield of trigger-associated particles is observed to be enhanced in thesoft-pT region compared to the p+p reference.

A natural question arises – what pT region of jet fragments feeds up the soft region? This ques-tion can be addressed by a direct measurement of fragmentation functions. The first measurementsof the fragmentation functions [135, 39] were consistent with no-modification due to the large sys-tematic and statistical uncertainties. However, the latest preliminary results by ATLAS [137] andCMS [136] show a significant change in the structure of fragmentation functions between centraland peripheral collisions. Both experiments evaluate the fragmentation for jets having the sameminimum-pT threshold of 100 GeV, although otherwise there are important differences betweenthese two measurements:

• ATLAS uses radial distance of ∆R = 0.4 to match particles to jets that were reconstructedwith different radii (R = 0.2, 0.3, 0.4); CMS uses ∆R = 0.3 for jets reconstructed with theradius of R = 0.3.

• ATLAS evaluates the fragmentation functions in terms of variable z, z = (pparticleT /pjetT ) cos ∆R

while CMS uses variable ξ, ξ = ln(pjetT /pparticleT ) ≈ ln(1/z).

• ATLAS uses jets in the pseudorapidity interval of |η| < 2.1 while CMS uses interval of0.3 < |η| < 2.

• ATLAS uses minimum pT threshold for charged particles of 2 GeV while CMS uses 1 GeV.

Despite to these differences the two experiments provide qualitatively the same result. Bothexperiments evaluate the modification of fragmentation in terms of the ratio of fragmentationfunctions in central collisions to the fragmentation functions in peripheral (ATLAS) or p+p col-lisions (CMS). In the case of no modification this ratio should be unity. The results for the mostcentral collisions are shown on upper plots of Fig. 4. The lower left plot of Fig. 4 compares theresult by ATLAS with the ξ-to-z recalculated result by CMS. The recalculated CMS result obvi-ously has to be taken as an approximate, but one can directly see a good qualitative agreementin these two measurements. The enhanced yield of low-z / low-pT / high-ξ particles in central

3There is no a priory reason why the balance of the event should occur for the threshold of pT = 0.5 GeV. Itwould clearly be of interest for the theory to extract this threshold with higher precision, if experimentally feasible

11

z-110 1

D(z

)R

0.8

0.9

1

1.1

1.2

1.3

1.4

1.5

1.6

0-10%/60-80%

>100 GeVjet

Tp

R=0.4Tanti-k

ATLAS Preliminary=2.76 TeVNNsPb+Pb -1=0.14 nbintL

= ln(1/z)ξ0 1 2 3 4 5

PbP

b/pp

0

0.5

1

1.5

2

2.5

3CMS Preliminary -1bµ = 150 IntL

> 100GeV/cT

Jet p

2010, 0-30%, Leading jet2011, 0-10%, Inclusive jet2011, 10-30%,Inclusive jet

z-210 -110 1

D(z

) R

0.6

0.8

1

1.2

1.4

1.6

1.8

0-10%/60-80%ATLAS Preliminary

0-10%/pp"recalculated"CMS Preliminary

z-210 -110 1

D(z

)R

0.95

1

1.05

1.1

1.15

1.2

1.25

1.3

1.35

>100 GeVjet

Tp

Quark to quark+gluon jets

Figure 4: Upper left: Central to peripheral ratio of fragmentation functions measured by AT-LAS [137]. Fragmentation functions are evaluated in terms of momentum fraction, z. Upperright: Ratio of fragmentation functions in central Pb+Pb collisions to p+p collisions measured byCMS [136]. Fragmentation functions are evaluated in terms of ξ ≈ ln(1/z). Lower left: CMS resultrecalculated in terms of z and compared to the result by ATLAS. Error bars represent combinedstatistical and systematic uncertainties. For differences between ATLAS and CMS measurementssee the text. Lower right: Ratio of fragmentation functions in light-quark (u, d, s) jets to inclu-sive jets (light-quark + gluon jets). Jets with pT > 100 GeV are considered. Simulation of p+pcollisions at

√sNN = 2.76 TeV by PYTHIA 6.4.

12

collisions is accompanied by a reduction in the yield of intermediate-z / ξ / pT particles4. Noreduction of yields at high-z is seen. On the contrary, rather an indication of enhancement canbe seen. This may be puzzling in a view of the theory that commonly predicted a reduction athigh z [129, 131, 130, 133, 134]. It is important to note that experiments evaluate naturally thefragmentation functions with respect to the jet energy which is different from the energy of theoriginal parton due to the quenching. On the other hand, theoretical studies commonly eval-uate the fragmentation functions with respect to the energy of parton prior to the quenching.This makes any quantitative comparison between measured data and predicted modifications offragmentation functions difficult.

Recently, the observed results of modified fragmentation were rather successfully reproduced bytheoretical calculations [138, 139]. Both references use the quenched jet energy to define the frag-mentation functions. Reference [138] uses a full simulation of events based on PYTHIA+JEWELMC generators. In this framework, PYTHIA is used to generate the hard-scattering and virtual-ity ordered parton showers. Then, JEWEL generates the final state parton shower modified bythe medium. Finally, PYTHIA takes over and does the hadronization. In the reference [139],authors use an effective 1+1 dimensional quasi-Abelian model to describe non-perturbatively thedynamics of fragmentation of quark jets in the medium. Both models are successful in describingthe trends seen in experimental data even though being based on very different modeling of thejet quenching.

10 Discussion

Throughout previous sections we could see a multitude of experimental observations of featuresconnected with the jet quenching. Let us first discuss more closely the features seen in thefragmentation measurement. Without considering any details on the jet quenching models, it isremarkable that the shape of the central-to-peripheral ratio of fragmentation functions resemblessome features of the ratio of quark jet fragmentation to quark+gluon jet fragmentation in thevacuum. This ratio modeled by PYTHIA is shown on the lower right plot of Fig. 4. The charac-teristic difference between the quark and gluon jet fragmentation in the vacuum was measured indetails at LEP [140, 141]. The difference reflects the fact that quark jets are more collimated andcontain more high-z fragments whereas gluon jets are generally broader and softer [142, 143, 144].In p+p collisions at the center-of-mass energy of 2.76 TeV, the quark jets should constitute abouttwo thirds of all jets at 100 GeV whereas gluon jets should constitute about a third of all jets. Thiscan be seen from the PYTHIA estimate of quark-initiated vs. gluon-initiated jet fractions in di-jetevents shown on Fig. 5. In the radiative energy loss scheme, gluon jets should be quenched morestrongly than light quark jets due to the larger color factor for gluon emission from gluons thanfrom quarks. Thus, the quark-to-gluon jet fraction has to be larger in central compared to periph-eral collisions. It is therefore inevitable that any change in the pattern of fragmentation functionsin central compared to peripheral collisions (or compared to p+p reference) will be modulated by adifference from different quark-to-gluon jet fractions. The ratio of fragmentation functions will be

4It is worth to note that the estimate of the magnitude of low-pT enhancement might be biased throughout thenecessary subtraction of the UE which might also be modified by the enhancement of soft particles as seen in themissing-pT study. The sensitivity to these effects might be tested by varying the distance between the jet and thearea used for the determination of the UE.

13

T,jetp

100 150 200 250 300

ratio

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

|<2.8jet

ηPYTHIA, |Gluon to quark+gluon jetsQuark to quark+gluon jets

jetη

-2 -1.5 -1 -0.5 0 0.5 1 1.5 2

ratio

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1>100 GeV

T,jetPYTHIA, p

Gluon to quark+gluon jets

Quark to quark+gluon jets

Figure 5: Ratio of multiplicities of jets initiated by gluon (blue) or light-quark (green) to multi-plicities of inclusive jets (light-quark + gluon jets). Ratio is evaluated as a function of jet transversemomentum (left) and jet pseudorapidity (right). Simulation of p+p collisions at

√sNN = 2.76 TeV

by PYTHIA 6.4.

given by a convolution of modification of fragmentation due enhanced quark-to-gluon fraction anda modification of fragmentation due to the energy loss itself. If we convolute the original predic-tions on medium modified fragmentation (predicting generally a significant depletion of fragmentsat high-z and an increase of fragments at low-z) with the characteristic pattern of fragmentationdue enhanced quark-to-gluon fraction (implying increase of fragments at high-z and decrease offragments at low-z) we might get a pattern measured by ATLAS and CMS.

The argument of influence of measured fragmentation by the difference in quark-to-gluon jetfractions may also provide a simple explanation for an enhanced yield of fragments at high-pT thatwas seen in the measurement by ATLAS5 : the decrease in the yields of fragments at high-pT can becompensated by an effective increase due to different quark-to-gluon jet ratio in central comparedto peripheral collisions. Further, the quark-to-gluon jet fractions differ at different pseudorapidi-ties as illustrated in the right plot of Fig. 5. This implies that any measurement performed atdifferent pseudorapidities will differ. This might be a reason for the difference between the overallmagnitude of RD(z) of ATLAS and CMS seen in the lower left plot of Fig. 46. It is clear that these

5Note that ATLAS evaluates the ratio of pT distribution whereas CMS evaluates the difference. The differenceof pT distribution is not sensitive to any modifications at high-pT due to small values of the distribution at high-pT(the order of 10−2 − 10−3). The difference might also have a different position of minimum than the ratio.

6It is also worth to stress that the ratio of fragmentation functions is sensitive to the overall normalization.Fragmentation functions should be normalized by the total number of all jets even if they do not have any associatedparticle with pT above a certain threshold (2 GeV in case of ATLAS or 1 GeV in case of CMS). Normalizing onlyby jets that have particles with pT above the threshold leads to a bias since there is a larger probability for a softjet to exceed the threshold in central compared to peripheral collisions due to the presence of UE. Even if the UE

14

arguments are qualitative, based on the observation of trends in various experimental results. Therole of the quark-to-gluon ratio in the measured modification of fragmentation functions mightbe further explored by measurement of fragmentation at different pseudorapidities, measurementof fragmentation in γ-jet events as well as by a measurement of particle multiplicity inside jets.The multiplicity is clearly sensitive to the jet quenching [128] but it is also powerful to distinguishbetween quark and gluon jets fragmenting in the vacuum [145].

One of striking observations we saw in the experimental results is the flatness of the jet nuclearmodification factor at high-pT reported by ATLAS and CMS. We have hinted at an importanceof appropriate treatment of quark-to-gluon jet fractions to disentangle the features seen in thejet fragmentation. Different fractions of quark and gluon jets at different pT and η have to beclearly taken into account also when trying to understand the result of flat jet nuclear modificationfactor. The clear pT and η dependence of quark versus gluon fractions of partons initiating p+pjets might in fact help us to reveal some information about the difference in the quark and gluonjet quenching as we will illustrate. The jet spectra can be generally described by the power-lawdistribution (dσ/dpT ∼ pTk). Let us imagine the quenching as an effective shift of the spectra by∆pT at a given pT value. Then, the RAA can be expressed as

RAA =Nquenched

quark−jet(pT) +Nquenchedgluon−jet(pT)

Nvacuumquark−jet(pT) +Nvacuum

gluon−jet(pT)=

(pT + ∆pT)kfq + (pT + κ∆pT)k(1− fq)pTk

(4)

where κ is the effective color factor, fq = fq(pT, η) is the fraction of quark-to-inclusive jets, andexponent k = −5 for LHC energies [53]. If we plug in standard QCD κ = 9/4, fq from Fig.5, andRAA of 0.5, we get for 100 GeV jet a reasonable ∆pT of 11 GeV. It is clear that Eqn.4 representsan over-simplification of the problem – the whole dynamics of the quenching resides in the ∆pT.However, it allows us to see that the jet RAA carries an important information which may berigorously accessed if it is supplied by an information from another quark-to-gluon jet sensitivemeasurement such as appear to be the fragmentation7.

This discussion illustrates that the kinematic coverage of the LHC data and the multitude ofmeasurements summarized in this review give a very good discriminating power and ultimatelyshould lead to a better understanding of the jet quenching.

11 Summary and conclusions

In this short review we have summarized the up-to-date experimental results on jet suppressionwith high-pT observables obtained from ALICE, ATLAS, and CMS. The diversity of experimentalmeasurements at high-pT involving different objects (jets, charged particles, γ, Z0), differentobservables (nuclear modification factors, fragmentation functions, high-pT v2, hadron-jet, γ-jet,or Z0-jet correlations), and different kinematic selections provide clear possibilities to constrain themodels of the jet quenching. The complexity of the physics problem (e.g. appropriate modelingof the space-time evolution of the medium along with the modeling of the jet quenching) and thecomplexity of the measurements (e.g. UE subtraction, fragmentation evaluated with respect to

subtraction is correctly performed, the bias in the normalization is not removed.7Especially the value in the minimum of RD(ξ) or RD(z) may be valuable for this purpose.

15

the quenched jet, possible influence of the UE by fragmentation) suggest a strong need of publiclyavailable fully simulated models of the jet quenching. The full simulation allows for a validationof a given model by a precise comparison to variety of different experimental results in the samemanner as different MC generators are confronted with the data in elementary collisions.

Experimental results on jets in heavy ion collisions at LHC clearly provide a rich set of datawhich has a great potential to shad light on the jet quenching, properties of the deconfined QCDmedium, and fundamentals of the strong interaction. These results are expected to be followedby precision measurements after the upgrade of LHC.

Disclosure statement: The author is not aware of any affiliation, membership, funding, orfinancial holdings that might be perceived as affecting the objectivity of this review.

12 Acknowledgment

I’m grateful for stimulating discussions with Jirı Dolejsı and Brain Cole. I’d like to thank toAaron Angerami, Marco van Leeuwen, Alexander Milov, Mateusz Ploskon, and Christof Rolandfor a help with selecting appropriate data for the summary plots. This work was supported bygrants MSM0021620859, UNCE 204020/2012, LG13009.

References

[1] F. Karsch, E. Laermann, Quark-Gluon Plasma III, ed. R. Hwa, arXiv:hep-lat/0305025.

[2] J. C. Collins, M. J. Perry, Phys. Rev. Lett. 34 (1975) 1353.

[3] N. Cabbibo, G. Parisi, Phys. Lett. B59 (1975) 67.

[4] E. V. Shuryak, Phys. Lett. B78 (1978) 150.

[5] J. D. Bjorken, FERMILAB-PUB-82-059-THY (1982).

[6] K. Aamodt et al. [ALICE Collaboration], JINST 3 (2008) S08002.

[7] G. Aad et al. [ATLAS Collaboration], JINST 3 (2008) S08003.

[8] R. Adolphi et al. [CMS Collaboration], JINST 3 (2008) S08004.

[9] I. Arsene et al. [BRAHMS Collaboration], Nucl. Phys. A757 (2005) 1.

[10] K. Adcox et al. [PHENIX Collaboration], Nucl. Phys. A757 (2005) 184-283.

[11] B. B. Back et al. [PHOBOS Collaboration], Nucl. Phys. A757 (2005) 28.

[12] J. Adams et al. [STAR Collaboration], Nucl. Phys. A757 (2005) 102-183.

[13] P. Braun-Munzinger, J. Wambach, Rev. Mod. Phys. 81 (2009) 1031-1050.

[14] B. Muller, J. L. Nagle, Annu. Rev. Nucl. and Part. Phys. 56 (2006) 93-135.

16

[15] P. Jacobs, X. N. Wang, Prog. Part. Nucl. Phys. 54 (2005) 443-534.

[16] P. Huovinen, P. V. Ruuskanen, Ann. Rev. Nucl. Part. Sci. 56 (2006) 163-206.

[17] E. V. Shuryak, Nucl. Phys. A750 (2005) 64-83.

[18] A. Accardi, F. Arleo, W. K. Brooks, D. d’Enterria, V. Muccifora, Riv.Nuovo Cim. 32 (2010)439-553.

[19] B. Muller, J. Schukraft, B. Wyslouch, Annu. Rev. Nucl. Part. Sci. 62 (2012) 361-386.

[20] R. Singh, L. Kumar, P. Kumar Netrakanti, B. Mohanty, arXiv:1304.2969 (unpublished).

[21] N. Armesto, B. Cole, C. Gale et al., Phys. Rev. C86 (2012) 064904.

[22] S. A. Bass, C. Gale, A. Majumder, C. Nonaka, G.-Y. Qin, T. Renk, J. Ruppert, Phys. Rev.C79 (2009) 024901.

[23] A. Majumder, M. Van Leeuwen, Prog. Part. Nucl. Phys. 66 (2011) 41-92.

[24] U. A. Wiedemann, Landolt-Boernstein Handbook of Physics, ed. R. Stock, arXiv:0908.2306.

[25] Y. Mehtar-Tani, J. G. Milhano, K. Tywoniuk, Int. J. of Mod. Phys. A28 (2013) 1340013.

[26] R. Baier, Yu-L. Dokshitzer, A. H. Mueller, S. Peigne, D. Schiff, Nucl. Phys. B483 (1997)291-300.

[27] M. Gyulassy, P. Levai, I. Vitev, Phys. Rev. Lett. 85 (2000) 5535.

[28] X.-N. Wang, X. Guo, Nucl. Phys. A696 (2001) 788-832.

[29] B. G. Zakharov, JETP Lett. 80 (2004) 617-622.

[30] P. B. Arnold, G. D. Moore, L. G. Yaffe, JHEP 0112 (2001) 009.

[31] M. Gyulassy, Wang. X.-N, Nucl. Phys. B420 (1994) 583.

[32] X.-N. Wang, M. Gyulassy, M. Pl”umer, Phys. Rev. D51 (1995) 3436.

[33] M. G. Mustafa, Phys. Rev. C72 (2005) 014905.

[34] C. Adler et al. [STAR Collaboration], Phys. Rev. Lett. 89 (2002) 202301.

[35] K. Adcox et al. [PHENIX Collaboration], Phys. Rev. Lett. 88 (2002) 022301.

[36] C. Adler et al. [STAR Collaboration], Phys. Rev. Lett. 90 (2003) 082302.

[37] G. Aad et al. [ATLAS Collaboration], Phys. Rev. Lett. 105 (2010) 252303.

[38] S. Chatrchyan [CMS Collaboration], Phys. Rev. C84 (2011) 024906.

[39] ATLAS Collaboration, ATLAS-CONF-2011-075 (2011).

17

[40] S. Chatrchyan et al. [CMS Collaboration], Phys. Lett. B B712 (2012) 176-197.

[41] G. Y. Qin, B. Muller, Phys.Rev.Lett. 106 (162302) 2011.

[42] C. Young, B. Schenke, S. Jeon, C. Gale, Phys. Rev. C84 (2011) 024907.

[43] T. Renk, Phys. Rev. C86 (2012) 061901.

[44] Y. He, I. Vitev, B.-W. Zhang, Phys. Lett. B713 (2012) 224-232.

[45] J. Casalderrey-Solana, J. G. Milhano, U. A. Wiedemann, J. Phys. G38 (2011) 035006.

[46] I. P. Lokhtin, A. V. Belyaev, A. M. Snigirev, Eur. Phys. J. C71 (2011) 1650.

[47] M. Cacciari, G. P. Salam, G. Soyez, JHEP 0804 (063) 2008.

[48] CMS Collaboration, CMS-PAS-PFT-09-001 (2009).

[49] R. Reed et al. [ALICE Collaboration], arXiv:1304.5945 (unpublished).

[50] M. Cacciari, G. P. Salam, G. Soyez, JHEP 0804 (2008) 005.

[51] M. Cacciari, G. P. Salam, Phys. Lett. B659 (2008) 119.

[52] M. Cacciari, J. Rojo, G. P. Salam, G. Soyez, Eur. Phys. J. C71 (2011) 1539.

[53] G. Aad et al. [ATLAS Collaboration], Phys. Lett. B719 (2013) 220-241.

[54] B. Abelev et al. [ALICE Collaboration], JHEP 03 (2012) 053.

[55] M. Cacciari, G. P. Salam, G. Soyez, Eur. Phys. J. C71 (2011) 1692.

[56] CMS Collaboration, PAS-HIN-12-004 (2012).

[57] M. Nguyen et al. [CMS Collaboration], J. Phys. G38 (2011) 124151.

[58] B. Abelev et al. [ALICE Collaboration], Phys. Lett. B722 (2013) 262.

[59] T. Sjostrand, S. Mrenna, P. Skands, JHEP 05 (2006) 026.

[60] X. N. Wang, M. Gyulassy, Comput. Phys. Commun. 83 (1994) 307.

[61] I. P. Lokhtin, A. M. Snigirev, Eur. Phys. J. C45 (2006) 211.

[62] P. M. Jacobs et al. [STAR Collaboration], arXiv:1012.2406 (unpublished).

[63] L. Apolinario, N. Armesto, L. Cunqueiro, JHEP 1302 (2013) 022.

[64] CMS Collaboration, CMS-PAS-PFT-10-002 (2010).

[65] ATLAS Collaboration, ATLAS-CONF-2012-045 (2012).

[66] W. C. Fabjan, F. Gianotti, Rev. Mod. Phys. 75 (2003) 1243-1286.

18

[67] M. Spousta et al. [ATLAS Collaboration], J. Phys. Conf. Ser. 270 270 (2011) 012013.

[68] M. L. Miller, K. Reygers, S. J. Sanders, P. Steinberg, Ann. Rev. Nucl. Part. Sci. 57 (2007)205.

[69] B. Abelev et al. [ALICE Collaboration], Phys. Lett. B720 (2013) 52-62.

[70] ATLAS Collaboration, ATLAS-CONF-2011-079 (2011).

[71] S. Chatrchyan et al. [CMS Collaboration], Eur. Phys. J. C72 (2012) 1945.

[72] J. Casalderrey-Solana, A. Milov, arXiv:1210.8271 (unpublished).

[73] ATLAS Collaboration, Phys. Rev. D83 (2011) 052003.

[74] B. A. Cole, Nucl. Phys. A774 (2006) 225-236.

[75] T. Renk, arXiv:1302.3710 (unpublished).

[76] S. Abreu, S. V. Akkelin, J. Alam et al., J. Phys. G35 (2008) 054001.

[77] T. Renk, arXiv:1207.4885 (unpublished).

[78] K. Aamond et al. [ALICE Collaboration], Phys. Rev. Lett. 108 (2012) 092301.

[79] CMS Collaboration, PAS-HIN-12-010 (2012).

[80] V. S. Pantuev, JETP Lett. 85 (2007) 104-108.

[81] T. Renk, K. Eskola, Phys.Rev. C75 (2007) 054910.

[82] S. Chatrchyan et al. [CMS Collaboration], Phys. Rev. Lett. 109 (2012) 022301.

[83] ATLAS Collaboration, ATLAS-CONF-2012-116 (2012).

[84] R. Baier, Y. L. Dokshitzer, A. H. Mueller, D. Schiff, Phys. Rev. C58 (1998) 1706-1713.

[85] S. Peigne, A. V. Smilga, Phys. Usp. 52 (2009) 659.

[86] S. S. Gubser, D. R. Gulotta, S. P. Silviu, F. D. Rocha, JHEP 0810 (2008) 052.

[87] S. Wicks, W. Horowitz, M. Djordjevic, M. Gyulassy, Nucl. Phys. A784 (2007) 426-442.

[88] C. Marquet, T. Renk, Phys. Lett. B685 (2010) 270-276.

[89] T. Renk, Phys. Rev. C83 (2011) 024908.

[90] B. Betz, M. Gyulassy, G. Torrieri, Phys. Rev. C84 (2011) 024913.

[91] J. Jia, W. A. Horowitz, J. Liao, Phys. Rev. C84 (2011) 034904.

[92] H. Liu, K. Rajagopal, U. A. Wiedemann, Phys. Rev. Lett. 97 (2006) 182301.

19

[93] S. S. Gubser, A. Karch, Ann. Rev. Nucl. Part. Sci. 59 (2009) 145-168.

[94] A. Adare et al. [PHENIX Collaboration], Phys. Rev. Lett. 105 (2010) 142301.

[95] B. Betz, Eur. Phys. J. A48 (2012) 164.

[96] B. Abelev et al. [ALICE Collaboration], Phys. Lett. B719 (2013) 18-28.

[97] G. Aad et al. [ATLAS Collaboration], Phys. Lett. B707 (2012) 330-348.

[98] K. Aamodt et al. [ALICE Collaboration], Phys. Rev. Lett. 105 (2010) 252302.

[99] S. Chatrchyan et al. [CMS Collaboration], Phys. Rev. C 87 (2013) 014902 C87 (2013) 014902.

[100] X. N. Wang, Phys. Lett. B579 (2004) 299-308.

[101] ATLAS Collaboration, ATLAS-CONF-2012-052 (2012).

[102] S. Chatrchyan et al. [CMS Collaboration], Phys. Lett. B710 (2012) 256-277.

[103] G. Aad et al. [ATLAS Collaboration], Phys. Rev. Lett. 110 (2013) 022301.

[104] ATLAS Collaboration, ATLAS-CONF-2011-078 (2011).

[105] S. Chatrchyan et al. [CMS Collaboration], Phys. Rev. Lett. 106 (2011) 212301.

[106] H. Paukkunen, C. A. Salgado, JHEP 1103 (2011) 071.

[107] X.-N. Wang, Z. Huang, I. Sarcevic, Phys. Rev. Lett. 77 (1996) 231-234.

[108] ATLAS Collaboration, ATLAS-CONF-2012-121 (2012).

[109] ATLAS Collaboration, ATLAS-CONF-2012-119 (2012).

[110] S. Chatrchyan et al. [CMS Collaboration], Phys. Lett. B718 (2013) 773-794.

[111] W. Dai, I. Vitev, B.-W. Zhang, Phys. Rev. Lett. 110 (2013) 142001.

[112] ATLAS Collaboration, ATL-PHYS-PUB-2012-002 (2012).

[113] CMS Collaboration, Future Heavy-Ion Programme at CMS (Contribution to the Update ofthe European Strategy for Particle Physics).

[114] Yu. L. Dokshitzer, D. E. Kharzeev, Phys. Lett. B519 (2001) 199-206.

[115] A. Adare et al. [PHENIX Collaboration], Phys. Rev. Lett. 98 (2007) 172301.

[116] B. I. Abelev et al. [STAR Collaboration], Phys. Rev. Lett. 98 (2007) 192301.

[117] A. Adare et al. [PHENIX Collaboration], Phys. Rev. C84 (2011) 044905.

[118] M. Djordjevic, Phys. Rev. C85 (2012) 034904.

20

[119] P. B. Gossiaux, J. Aicheli, Phys. Rev. C78 (2008) 014904.

[120] J. Uphoff, O. Fochler, Z. Xu, C. Greiner, Phys. Rev. C84 (2011) 024908.

[121] W. A. Horowitz, M. Gyulassy, J. Phys. G35 (2008) 104152.

[122] H. Van Hees, M. Mannarelli, V. Greco, R. Rapp, Phys. Rev. Lett. 100 (2008) 192301.

[123] A. Adil, I. Vitev, Phys. Lett. B649 (2007) 139-146.

[124] CMS Collaboration, PAS-HIN-12-003 (2012).

[125] ATLAS Collaboration, ATLAS-CONF-2012-050 (2012).

[126] B. Abelev et al. [ALICE Collaboration], Phys. Rev. Lett. 109 (2012) 112301.

[127] S. Chatrchyan et al. [CMS Collaboration], JHEP 1205 (2012) 063.

[128] C. A. Salgado, U. A. Wiedemann, Phys. Rev. Lett. 93 (2004) 042301.

[129] I. P. Lokhtin, A. M. Snigirev, Phys.Lett. B567 (2003) 39-45 B567 (2003) 39-45.

[130] N. Borghini, U. A. Wiedemann, arXiv:hep-ph/0506218 (unpublished).

[131] N. Armesto, L. Cunqueiro, C. A. Salgado, W.-C. Xiang, JHEP 0802 (2008) 048.

[132] I. Vitev, S. Wicks, B.-W. Zhang, JHEP 11 (2008) 093.

[133] N. Armesto, L. Cunqueiro, C. A. Salgado, Eur. Phys. J. C63 (2009) 679-690.

[134] T. Renk, Phys. Rev. C79 (2009) 054906.

[135] S. Chatrchyan et al. [CMS Collaboration], JHEP 10 (2012) 087.

[136] CMS Collaboration, PAS-HIN-12-013 (2012).

[137] ATLAS Collaboration, ATLAS-CONF-2012-115 (2012).

[138] K. C. Zapp, F. Krauss, U. A. Wiedemann, JHEP 1303 (2013) 080.

[139] D. E. Kharzeev, F. Loshaj, arXiv:1212.5857 (unpublished).

[140] D. Buskulic et al. [ALEPH Collaboration], Phys. Lett. B384 (1996) 353-364.

[141] G. Alexander et al. [OPAL Collaboration], Z. Phys. C69 C69 (1996) 543-560.

[142] D. De Florian, R. Sassot, M. Stratmann, Phys. Rev. D76 (2007) 074033.

[143] D. Acosta et al. [CDF Collaboration], Phys. Rev. D71 (2005) 112002.

[144] R. Barate et al. [ALEPH Collaboration], Eur. Phys. J. C17 (2008) 1-18.

[145] J. Gallicchio, M. D. Schwartz, JHEP 1304 (2013) 090.

21