Heavy-ion results from the LHC

0.47 MB·English

by  Francesco Prino

XXXI PHYSICS IN COLLISION, Vancouver, BC Canada, August 28 - September 1, 2011
Heavy-ion results from the LHC
F. Prino
INFN, Sezione di Torino, Italy
I. INTRODUCTION II. COLLISION GEOMETRY AND SYSTEM
2 EVOLUTION
1
0
2 Nuclear collisions are characterized by the impact
n In November 2010, the first heavy-ion run began parameter between the colliding nuclei, i.e. the dis-
a at the LHC. Pb nuclei were collided at the centre- tance between the centers of the nuclei in the trans-
√
J of-mass energy sNN = 2.76 TeV, about 14 times verse plane. The impact parameter defines the cen-
1 higher than that achieved at RHIC. The integrated trality of the collision: in head-on (“central”) colli-
1 luminosity delivered by the LHC during the 5 weeks sions a large number of nucleons participates in the
of running was 10 µb−1. Three experiments collected interactionandmanybinarynucleon-nucleoncollision
]
x data with Pb–Pb collisions, namely ALICE, ATLAS occur, leading to a large energy deposit in the reac-
e and CMS. tion volume. Conversely, collisions with large impact
- parameter(“peripheral”)havelowernumberofpartic-
l
c ipant nucleons and binary collisions and consequently
u
Heavy-ion collisions at relativistic energies are lower energy densities are attained. Experimentally,
n
[ aimed at studying nuclear matter at extreme condi- thedefinitionofthecollisioncentralityisbasedonthe
tionsoftemperatureandenergydensity,whereLattice assumption that the impact parameter is monotoni-
1 QCDpredictsthemattertobeinastatewherequarks cally related to the particle multiplicity or the energy
v
and gluons are deconfined over volumes much larger produced in the collision. It is typically expressed in
8
4 than the size of a hadron (see e.g. [1]). Such a state terms of fractions (percentiles) of the total hadronic
3 is called Quark Gluon Plasma (QGP). The goal of cross section. Starting from the centrality-related ex-
2 heavy-ion collision experiments is to collect evidence perimental observables, the collision geometry can be
. for the existence of this new state of matter and to reconstructed utilizing the Glauber model [2], which
1
0 study its properties. describes a nuclear collision in terms of multiple-
2 scatteringofnucleonsinnucleartargets,togetherwith
1 amodelforparticleproduction. Thisapproachallows
: one to compute the average numbers of participant
v Pb–Pb collisions at the LHC are expected to gen-
nucleons((cid:104)N (cid:105)),ofelementarynucleon-nucleoncol-
i erate a medium that has a higher initial temperature part
X lisions ((cid:104)N (cid:105)), and other geometrical quantities for
andenergydensitythatthatgeneratedatlowervalues coll
r √ each centrality interval [3].
a of s. Experimentalmeasurementsinthisnewenergy
regimeareakeybenchmarkformodelsthatreproduce The main difference between a nucleus-nucleus col-
the features observed at lower collision energy. Fur- lision and a p–p collision is the larger collision vol-
thermore, the input from the LHC is also crucial to ume where a dense medium of strongly interacting
shed light on some issues that are not completely un- matter (fireball) is formed. The evolution of the pro-
derstoodfromtheSPSandRHICresults(e.g. theJ/ψ duced medium goes through various stages which are
√
suppression). Also, the steep increase with s of the schematically described in the following (see e.g. [4]
cross-sectionforQCDscatteringswithhighvirtuality for more details). The fireball first goes through an
results in copious production of hard partons at the earlynon-equilibriumstage. Re-scatteringsamongits
LHC, thus allowing one to reach higher precision on constituents can establish thermal equilibrium. If the
the experimental observables related to high momen- system thermalizes quickly enough and at sufficiently
tum and heavy flavoured particles. Finally, at LHC large energy density, it is expected to be in the QGP
energies, new (higher mass) probes, such as W and phase. Experimental results indicate that, for cen-
Z0 bosons, become available and provide new tools to tralAu-AucollisionsatRHIC,thermalequilibriumis
study the properties of the medium. reached at a time τ = 0.6 fm/c after the collision
eq
20
XXXI PHYSICS IN COLLISION, Vancouver, BC Canada, August 28 - September 1, 2011
when the temperature of the fireball is ≈ 360 MeV,
well above the critical values for deconfinement [4].
The thermalized medium undergoes a collective hy-
drodynamic expansion driven by the pressure gradi-
entsgeneratedbythehightemperaturesanddensities
that characterize these stages of the collision evolu-
tion[5]. Asaconsequence,thefireballcoolsdownand
its energy density decreases. When the energy den-
sity reaches the critical threshold for deconfinement
((cid:15) ∼ 1 GeV/fm3), the partons convert to hadrons
(hadronization). After this phase transition, hadrons
keepre-scatteringoffeachother,thusbuildingupfur-
thercollectiveflow,untilthesystembecomessodilute
that all hadronic interactions cease and the particles
decouple(“freeze-out”). Thefreeze-outisactuallyan-
ticipated to occur in two stages [4]. First (at higher
temperatures) inelastic processes stop, thus freezing
the abundances of the various hadron species (chem-
ical freeze-out). Afterwards (at lower temperatures),
alsoelasticcollisionamonghadronsceaseandthepar-
ticle momenta get frozen (thermal freeze-out). At the FIG. 1: Top: Collision energy dependence of dNch/dη
per colliding nucleon pair in p–p (open symbols) and AA
thermal freeze-out, the bulk of hadrons which under-
(closedsymbols)collisions. Bottom: (dN /dη)/(N /2)
went the thermalization and the successive collective ch part
vs. centralityatLHCcomparedtoRHICresultsscaledby
expansion have an approximately exponential trans-
a factor 2.15. Taken from [9].
versemomentumspectrumreflectingthetemperature
of the fireball at that point, blue-shifted by the av-
erage transverse collective flow [4]. The evolution of
cle production. All three experiments have measured
thefireballfromthemomentwhenlocalthermalequi-
dN /dη at mid-rapidity as a function of the collision
ch
librium is reached until the thermal freeze-out can be
centrality [7–10]. The multiplicity in the most cen-
described with hydrodynamics [5].
tral collisions at the LHC is larger by a factor ≈2.1
The bulk of particles produced in heavy-ion colli-
with respect to central collisions at top RHIC energy,
√
sionsarelow-momentum(soft)hadronsthathaveun-
in disagreement with the scaling with log s observed
dergonethecollectiveexpansionandhaveabundances
at lower energies (see top panel of Fig. 1). The cen-
andmomentathatarefixedatthestagesofthechem-
tralitydependenceof(dN /dη)/(N /2)hasasim-
ch part
icalandthermalfreeze-out. Experimentalobservables
ilar shape to that observed at RHIC (Fig. 1, bottom
for low-momentum particles produced in Au-Au col-
panel) and is reasonably reproduced both by models
lisions at RHIC are well described by hydrodynamics
based on gluon saturation in the initial state and by
with a QGP equation of state and very low shear vis-
two-component Monte Carlo models [8, 10].
cosity (see e.g. [5, 6]). The results from the LHC in
The produced transverse energy E was estimated
t
the soft particle sector are described in Section III.
by ALICE by measuring the charged hadronic en-
Ontopofthebulkofsoftparticles,alsohardparti-
ergy with the tracking system and adding the con-
cles, with either a large mass or large transverse mo-
tribution of neutral particles. The measured E per
t
menta (p (cid:29) 1 GeV/c), are created. Hard particle
t (pseudo)rapidity unit can be used to estimate the en-
production occurs in the early stages of the collision,
ergy density with the Bjorken formula [11]:
before the bulk of quanta have time to re-scatter and
tthraevremrsaelizteh.e mOendcieumprocdouncsteidt,utheadrdbypathrteicbleuslkhoavfesotfot εBj = A1τ ddEyt(cid:12)(cid:12)(cid:12)(cid:12) (1)
particles and can therefore be used as probes for the y=0
medium properties as discussed in Section IV. where A is the transverse overlapping area in the
collision of the two nuclei and τ is the formation
time. For the most central (0–5%) collisions at the
III. GLOBAL EVENT CHARACTERISTICS LHC, the resulting value is ε τ ≈ 16 GeV/(fm2c),
Bj
about a factor 3 larger than the corresponding value
The measurement of the multiplicity of produced at RHIC [12, 13].
particles, quantified by the charged particle density ThesystemsizeismeasuredfromtheHBTradiiex-
per unit of rapidity (dN /dη), is a global observ- tracted from the study of two-pion correlations. For
ch
able that provides insight into the density of gluons central collisions, it is found to be larger by a factor
in the initial stages and on the mechanisms of parti- two with respect to the one observed in central colli-
20
XXXI PHYSICS IN COLLISION, Vancouver, BC Canada, August 28 - September 1, 2011
sions at the top RHIC energy [14].
v2 ALICE preliminary, Pb Pb events at sNN = 2.76 TeV
Atypicalfeatureofthemediumproducedinheavy- centrality 40% 50%
0.3
ioncollisionsisthepresenceofcollectivemotionsaris- π±, v{SP, |∆η|>1}
K±, v2{SP, |∆η|>1}
ing from the large pressure gradients generated by p, v{2SP, |∆η|>1}
2
compressingandheatingthenuclearmatter. Thefirst 0.2
type of collective motion is called radial flow and it
stemsfromtheisotropicexpansionofthefireball. Itis
accessed experimentally by measuring the transverse 0.1
hydro LHC
momentum (p ) spectra of identified hadrons which (CGC initial conditions)
t (η/s=0.2)
at low p show a thermal (Boltzmann) distribution
t
0
blue-shifted by the collective velocity of the system 0 0.5 1 1.5 2 2.5 3 3.5
p (GeV/c)
expansion. The p spectra of identified hadrons (π, t
t
K and p) have been measured by ALICE in a wide
momentum range [15]. The spectra are seen to be FIG. 2: Measured elliptic flow of identified particles for
centrality bin 40%-50%, compared to hydro predictions
harder (i.e. characterized by a less steep distribu-
from [22]
tion and a larger (cid:104)p (cid:105)) than those observed at RHIC
√ t
at s =200 GeV. This is a first indication for a
NN
stronger radial flow at the LHC. Deeper insight is ob-
tained by fitting the π, K and p spectra with a blast- Elliptic flow has been studied by all three exper-
wave function [16], which allows one to extract an iments [19–21]. The pt integrated elliptic flow of
estimate of the parameters of the system at the ther- charged particles is found to incre√ase by about 30%
malfreeze-out,namelythefreeze-outtemperatureand from the highest RHIC energy of s = 200 GeV to
the radial flow velocity. The radial flow velocity from LHC energy [19].
blast-wave fits results, for the most central collisions The pt-differential elliptic flow is sensitive to the
at the LHC, about 10% higher than what observed in evolution and freeze-out conditions of the expanding
central collisions at top RHIC energy [15]. medium. The measured v2(pt) at the LHC is found
The build-up of a collective motion is also sig- to be compatible with that observed at RHIC. The
naled by the presence of anisotropic flow patterns 30% increase in the integrated elliptic flow is there-
in the transverse plane due to an initial geometri- fore coming from an increase in the average trans-
cal anisotropy in the spatial distribution of the nu- verse momentum, du√e to the increase of the radial
cleons participating in the collision [17]. Rescatter- flow with increasing s. The larger radial flow leads
ings among the produced particles convert this initial also to a more pronounced mass dependence of the
geometrical anisotropy into an observable momentum elliptic flow. In Fig. 2 the measured v2(pt) for iden-
anisotropy. For non-central collisions, the geometri- tified pions, kaons and protons is shown for semi-
cal overlap region of the colliding nuclei present an peripheral (40–50%) collisions. Hydrodynamic model
almond-like shape and the impact parameter defines predictions, based on the assumption that the QGP
a preferred direction in the transverse plane. The shear viscosity over entropy ratio does not change
anisotropy of produced particles is characterized by from RHIC to LHC [22], are shown in the figure and
the Fourier coefficients vn = (cid:104)cos[n(ϕ−Ψn)](cid:105), where provideagooddescriptionofv2(pt)forthethreepar-
n is the order of the harmonic, ϕ is the azimuthal an- ticle species. This is not the case in more central
gle of the particle and Ψ is the angle of the initial (10–20%) reactions where the hydrodynamic predic-
n
state spatial plane of symmetry. If the matter dis- tion does not provide a reasonably good description
tribution in the colliding nuclei varied smoothly, the of anti-protons [23]. This mismatch is also observed
plane of symmetry would coincide with the reaction forthept spectraofidentifiedprotons[15]anditmay
plane (i.e. the plane defined by the impact parameter be due to a larger radial flow in the data. At RHIC
and the beam direction) and odd Fourier coefficients energies, a better description of the antiproton flow
would be zero by symmetry. However, due to fluctu- was obtained by introducing a hadronic cascade af-
ations in the spatial distribution of the participating terthehydrodynamicevolution. Theresultsobtained
nucleonsinthecollidingnuclei,theplaneofsymmetry forellipticflowindicatethatthehotanddensematter
Ψ fluctuates event-by-event relative to the reaction createdinheavy-ioncollisionsattheLHCstillbehaves
n
plane. This gives rise to the presence of odd harmon- likeastronglyinteractingfluidwithexceptionallylow
ics,suchasv andv ,intheparticleazimuthaldistri- viscosity, as that observed at RHIC [6, 22].
3 5
butions. Since the impact parameter vector and the Higher Fourier harmonics have been measured by
spatial planes of symmetry are not directly measur- the three experiments [24–26]. In the most central
able, various experimental techniques have been de- events, dominated by fluctuations in the initial geo-
veloped to estimate the anisotropic flow coefficients metricalconfiguration,theelliptic(v2)andtriangular
from measured correlations among the observed par- (v3) flow are found to have similar magnitude.
ticles [18]. The presence of higher harmonics provides a natu-
20
XXXI PHYSICS IN COLLISION, Vancouver, BC Canada, August 28 - September 1, 2011
ral explanation of the structures observed in the two- TheR observableisthereforesensitivetotheprop-
AA
particle azimuthal correlations at low p , namely the erties of the medium created in the collision. It has
t
“ridge”at∆ϕ≈0andlarge∆ηandthedouble-bump however to be considered that other effects related to
intheawayside(∆ϕ∼π). Thesestructureswerefirst the presence of nuclei in the initial state (e.g. nuclear
observed at RHIC in Au-Au collisions [27] and their modifications of the PDFs, Cronin enhancement) can
interpretationisalongstandingpuzzle: modelsbased break the expected binary scaling.
on jet-medium interactions were at first developed
(e.g. [28]), while more recently an explanation based
ontriangularflowwasproposed[29]. Thesestructures A. Quarkonia
in the two-particle correlations have been studied in
detailbyperformingaFourieranalysisofthe∆η−∆ϕ As mentioned above, quarkonium states are ex-
correlations with large ∆η gap [25, 30, 31]. The data pected to be suppressed (R < 1) in the QGP, due
AA
arefoundtobewelldescribedbythefirstfivetermsof to the color screening of the force which binds the cc¯
the Fourier series. At low pt (i.e. pt∼<3−4 GeV/c), (or b¯b) state [32]. The suppression is predicted to oc-
the Fourier components extracted from two-particle cur both for the charmonium (J/ψ, Ψ(cid:48), χ , ...) and
c
correlations are found to factorize into single-particle bottomoniumfamilies(Υ(1S,2S,3S))above(orclose
harmoniccoefficientswhichareinagreementwiththe to) the critical temperature for the phase transition.
anisotropic flow coefficients. This suggests that the Forthermore, the quarkonium suppression is antici-
features observed in two-particle correlations at low pated to occur sequentially according to the binding
pt are consistent with the collective response of the energy, i.e. strongly bound states (such as J/ψ and
system to the initial state geometrical anisotropy. Υ(1S))shouldmeltathighertemperaturesrelativeto
morelooselyboundstates. Ithasalsotobeconsidered
√
that with increasing s, due to the more abundant
IV. CHARACTERIZATION OF THE production of charm in the initial state, charmonium
MEDIUM WITH HARD PROBES regenerationfromcandc¯recombinationathadroniza-
tion time can lead to an enhancement in the number
Particles with large transverse momentum and/or of observed J/ψ [35].
mass, which are produced in large-virtuality parton At the LHC, the J/ψ yield has been measured by
scatterings, are powerful tools to probe the medium all the three experiments. ATLAS measured J/ψ at
created in heavy-ion collisions. The production of mid-rapidity and high pt (80% of J/ψ’s measured
such“hardprobes”innuclearcollisionsisexpectedto by ATLAS have pt > 6.5 GeV/c) and reported a
scalewiththenumberofnucleon–nucleoncollisionsin suppression which increases with increasing central-
thenucleus–nucleuscollision(binaryscaling). Theex- ity [36]. CMS measured the nuclear modification fac-
perimentalobservableusedtoverifythebinaryscaling tor of prompt and secondary J/ψ’s at mid-rapidity,
is the nuclear modification factor: finding a suppression which increases with central-
ity [37]. Also in this case, high p (>6.5 GeV/c) J/ψ
t
Yield in AA aremeasured. TheR measuredattheLHCatmid-
R = (2) AA
AA (cid:104)N (cid:105)·Yield in pp rapidityshowsmoresuppressionthanobservedbythe
coll
PHENIX and STAR experiments at RHIC at central
where (cid:104)N (cid:105) is the average number of binary rapidity. ItshouldbenoticedthatPHENIXmeasured
coll
(nucleon-nucleon)collisionsfortheconsideredcentral- J/ψ mesons down to pt =0, while thepreliminary re-
ityclass. Ifnonucleareffectsarepresent,RAA should sults from STAR [38] are for high-pt (> 5 GeV/c)
be equal to 1 by construction. It is anticipated that J/ψ’sandshowasystematicallyhigherRAA thanob-
the medium created in the collision affects the abun- servedbyPHENIXatlow-pt. Ithasalsotobeconsid-
dances and spectra of the originally produced hard eredthattheCMSresultsareforpromptJ/ψ athigh
probes, resulting in a value of RAA different from 1. pt,whilePHENIXandSTARmeasuredinclusiveJ/ψ
In particular: (i.e. including those from B feed-down). CMS also
reported the R for non-prompt J/ψ’s from B feed-
AA
• QuarkoniaareexpectedtomeltintheQGPdue down which are found to be less suppressed than the
to color charge screening which would lead to a prompt ones. In this case, the mechanism that gives
suppression of the measured yield of charmonia risetotheobservedsuppressionisnotthequarkonium
and bottomonia [32]. melting: Bmesonsdecaywelloutsidethemedium,but
arise from the fragmentation of a b quark which suf-
• Partons are expected to lose energy while fers energy loss when traversing the colored medium.
traversing the strongly-interacting medium, via Additional information about J/ψ suppression at the
gluon radiation and elastic collisions with the LHC is presented by the ALICE measurement [39]
partonicconstituents(seee.g.[33,34]forrecent which is performed at forward rapidity and down to
reviews). p = 0 for inclusive J/ψ. The resulting R shows
t AA
20
XXXI PHYSICS IN COLLISION, Vancouver, BC Canada, August 28 - September 1, 2011
a suppression almost independent of centrality and
smaller than that observed by PHENIX at forward 2 SPS 17.3 GeV (PbPb) GLV: dNg/dy = 400
p0 WA98 (0-7%) GLV: dNg/dy = 1400
rapidity. In summary, the LHC results on J/ψ sup- RHIC 200 GeV (AuAu) GLV: dNg/dy = 2000-4000
pression provide hints that the J/ψ suppression is pt p0 PHENIX (0-10%) YaJEM-D
dependent and that regeneration may play an impor- 1.5 h– STAR (0-5%) elastic, small Pesc
elastic, large P
tant role at low pt. A deeper understanding requires SPS LHC 2.76 TeV (PbPb) YaJEM esc
studies of initial state effects by measuring J/ψ pro- h– CMS (0-5%) ASW
duction in p–A collisions at LHC energy. RAA 1 h– ALICE (0-5%) PQM: <q> = 30 - 80 GeV2/fm
CMS also measured Υ states in both p–p and
√
Pb–Pb collisions at s=2.76 TeV, resolving the 1S, CMS preliminary
2S and 3S states. The strongly-bound Υ(1S) state is
0.5
foundtobesuppressed: thepreliminarymeasurement
RHIC
ofthenuclearmodificationfactorintegratedovercen-
trality is R =0.62±0.11±0.10 [37]. The suppres-
AA
sion of the higher-mass states is measured relative to 0
1 2 34 10 20 100 200
the ground state, by building a double ratio between
p (GeV/c)
Υ(2S+3S) and Υ(1S) yields in Pb–Pb and p–p. The T
result is:
FIG.3: Chargedhadronnuclearmodificationfactoratthe
[Υ(2S+3S)/Υ(1S)]
PbPb =0.31+0.19(stat)±0.03(syst) LHCcomparedtomeasurementsatlowerenergies(RHIC
[Υ(2S+3S)/Υ(1S)]pp −0.15 and SPS). Taken from [43].
which indicates that the excited states are signifi-
cantly more suppressed than the ground state. The
quarks, thus providing a further benchmark for theo-
probability to measure such a value, or a lower one,
retical models. Radiative energy loss models predict
if the true double ratio is equal to one, is less than
that quarks lose less energy than gluons (that have a
1% [40].
largercolourcharge)andthattheamountofradiated
energy decreases with increasing quark mass. Hence,
a hierarchy in the values of the nuclear modification
B. Parton energy loss in the QCD medium
factor is anticipated, namely the R of B mesons
AA
shouldbelargerthanthatofDmesonsthatshouldin
As mentioned above, hard partons, produced at
turn be larger than that of light-flavour hadrons (e.g.
the initial stage of the collision, traverse the medium
pions), which mostly originate from gluon fragmenta-
and are expected to be sensitive to its energy density,
tion.
throughthemechanismofin-mediumpartonicenergy
TheR ofunidentifiedchargedparticleshasbeen
loss. The amount of energy lost is sensitive to the AA
measuredbyALICE[42]andCMS[43]. Inparticular,
medium properties (density) and depends also on the
the CMS R (which extends up to p =100 GeV/c)
path-length of the parton in the deconfined matter as AA t
is shown in Fig. 3 for the 0–5% centrality class com-
well as on the properties of the parton probing the
paredtotheALICEmeasurementandtoresultsfrom
medium. The nuclear modification factor as a func-
experiments at lower energies. The R is smaller
tion of p is therefore considered: AA
t at the LHC, suggesting a larger energy loss relative
1 dN /dp 1 dN /dp to RHIC energies and indicating that the density of
RAA(pt)= (cid:104)N (cid:105) dNAA/dpt = (cid:104)T (cid:105) dσAA/dpt themedium√createdinthecollisionincreaseswiththe
coll pp t AA pp t
increase of s. It should also be considered that
where (cid:104)T (cid:105) is the average nuclear overlap function the fraction of hadrons originating from gluon jets
AA
computed with the Glauber model for the considered increases with increasing centre-of-mass energy and,
centrality class [3]. The parton energy loss manifests in radiative energy loss models, this is expected to
itself as an RAA value lower than 1 at large pt. Al- give rise to a lower RAA because gluons lose more en-
ready at RHIC energies, the production of high pt ergythanquarkswhiletraversingtheQGP.TheRAA
(in the range 5-10 GeV/c) hadrons was found to be presents a minimum at pt ≈ 6−7 GeV/c and then
stronglysuppressed(byafactorfive)inthemostcen- increases slowly up to about 40 GeV/c. For pt above
tral collisions [41]. 40 GeV/c the RAA seems to level off at a value of
The higher centre-of-mass energy attained at the approximately 0.5. Also at momenta of 100 GeV/c a
LHCallowsextensionofthep rangewheretheR is significant suppression of charged hadron yield is still
t AA
measured,thusprovidingimportantconstraintstothe present.
particleenergylossmodels. Moreover,themoreabun- Medium-blind probes, such as photons and Z0
dant production of charm and beauty enables higher bosons, can be used to test that the initial produc-
precision measurements of the energy loss for heavy tion of hard probes follows the expected binary scal-
20
XXXI PHYSICS IN COLLISION, Vancouver, BC Canada, August 28 - September 1, 2011
ing, so as to separate initial and final state effects
ianrethperoodbusceerdveddirRecAtAly.fHroimghtehneehrgayrd(pscroamtteprti)ngphoofttownos RAA1.28 Pb Pb sNN=2.76 TeV D0 RAA 0 20% CC
partons. In nuclear collisions, they traverse the pro- 1.6 D+ R 0 20% CC
AA
duced medium without suffering from strong interac- 1.4 π± RAA 0 20% CC
ALICE Preliminary
tion,thusprovidingadirecttestofpQCDproduction 1.2
and initial state effects (such as nuclear modification 1
of the parton distribution functions). From the ex- 0.8
perimentalpointofview, themeasurementofprompt 0.6
photonsischallengingbecauseofthehugebackground 0.4
from π0 and η decays. CMS has performed a mea- 0.2
surementofisolatedphotonproductioninPb–Pbcol- 0
0 2 4 6 8 10 12 14
lisions as a function of centrality [44]. The measured p [GeV/c]
t
E spectra of isolated photons in Pb–Pb are com-
t
pared with a theoretical (NLO pQCD) p–p reference FIG. 4: Nuclear modification factor for prompt D0 and
from JETPHOX 1.2.2 [45] that provides a good de- D+ mesons in the 20% most central collisions compared
with the charged pions. Statistical (bars), systematic
scription of the photon cross section in p–p collisions
√ (empty boxes), and normalization (full box) uncertainties
at s = 7 TeV. The resulting R is found to be
AA are shown. Taken from [47].
compatiblewithunitywithinuncertainties. Thiscon-
firms that the initial production of hard probes fol-
lows the expected scaling of the production rate from
p–p to AA by the number of nucleon-nucleon colli- C. Jet quenching
sions. Z bosons have been measured in Pb–Pb colli-
sionsbyATLAS[36]andCMS[46]viatheirdecayinto In-mediumQCDenergy lossisalsopredicted toaf-
µ+µ−. The measured Z0 yield scaled by the number fect fully reconstructed jets, due to “jet quenching”.
of nucleon-nucleon collisions is found to be indepen- Thestudyofjetmodificationinheavyioncollisionsis
dent of centrality within uncertainties. Furthermore, predictedtobeapowerfultooltoprobetheproperties
theZ0 yieldinPb–Pbresultstobecompatiblewithin of the QGP. In particular, in-medium parton energy
uncertainties with the theoretical next-to-leading or- losscansignificantlyaltertheobservedenergybalance
der pQCD p–p cross sections scaled with the number between the most energetic (“leading”) and the sec-
of binary collisions. ond most energetic (“sub-leading”) jets in the event.
Asmentionedabove, toprovidefurtherinsightinto It is also interesting to measure the azimuthal angle
the energy loss mechanisms, it is interesting to mea- betweenthetwojetsandverifyifsignificantdeviations
surethenuclearmodificationfactorsforidentifiedpar- with respect to back-to-back emission (∆ϕdijet ≈ π)
ticles, in particular for heavy-flavoured hadrons. The are observed.
suppression of open charm and open beauty has been Jet reconstruction in the high-multiplicity environ-
measured by ALICE with three different techniques: mentofhighenergyheavy-ioncollisionsisachalleng-
exclusive reconstruction of D0 and D+ hadronic de- ing task that was pioneered in Au-Au reactions at
caysatmid-rapidity,singleelectronsaftersubtraction RHIC.Inparticular,jetmeasurementshavebeencar-
of a cocktail of background sources at mid-rapidity, ried on in STAR: inclusive jet cross-section, di-jet co-
and single muons at forward rapidity [47]. The mea- incidence rate and hadron-jet correlations. They all
surement of prompt D0 and D+ R is shown in the convergetowardapictureofbroadeningandsoftening
AA
left panel in Fig. 4. A strong suppression is observed, of the jet fragmentation [48].
reachingafactor4-5forpt >5GeV/c. Athighpt the AttheLHC,ATLASandCMSreportedresultswith
suppression is similar to the one observed for charged fully reconstructed jets, using the energies measured
pions, while at low pt there seems to be an indication in η,ϕ cells with the calorimeters and analyzed with
for RAA(D) > RAA(π±). differentalgorithms(anti-kt,iterativecone)forjetre-
As mentioned in section IVA, CMS has measured construction as well as with different techniques for
the R of displaced J/ψ, i.e. coming from B meson the subtraction of the background from the underly-
AA
decays, and reported a value of R = 0.36±0.08± ing event.
AA
0.03 for the 20% more central collisions. Also in this ATLAS [49] reported shortly after the Pb–Pb run
case, J/ψ mesons are measured for pt > 6.5 GeV/c. the observation of a significant imbalance between
The physical mechanism behind this suppression is themeasuredtransverseenergiesofdi-jetsinopposite
the b quark energy loss. The larger RAA of J/ψ from hemispheres. The jet energy imbalance is quantified
bwithrespecttothatmeasuredforpromptDmesons by the asymmetry A defined as:
J
is a first indication for the b quark to lose less energy
that the c quark, as anticipated by radiative energy E −E
A = T1 T2 (3)
loss models. J E +E
T1 T2
20
XXXI PHYSICS IN COLLISION, Vancouver, BC Canada, August 28 - September 1, 2011
where E is the transverse energy of the leading jet leading and sub-leading jets. The results show that a
T1
and E the one of the most energetic jet in the op- large part of the energy needed for balancing the jet
T2
posite hemisphere. A is positive by construction. It momenta is carried by soft particles (p < 2 GeV/c)
J t
is required that E > 100 GeV and E > 25 GeV radiated at large angles with respect to the jet axes
T1 T2
and therefore di-jets in which the sub-leading jet is (∆R>0.8) [50].
below the 25 GeV threshold are not included in the
Theobservedlargefractionofhighlyimbalanceddi-
A calculation.
J jetsincentralPb–Pbcollisionsandthecorresponding
The measured distributions of A in peripheral
J softening and widening of the fragmentation pattern
Pb–Pb collisions is similar to that observed in p–p
ofthesub-leadingjetareconsistentwithahighdegree
events as well as to that expected from Monte Carlo
of jet quenching in the produced medium.
simulations without in-medium energy loss. For more
First studies of fragmentation functions in Pb–Pb
central collisions, the A distribution broadens, does
J
collisions have been performed by ATLAS [51] and
not show a peak at A = 0 and its mean is shifted
J
CMS[52]. The resultsshowthatthehardcomponent
to higher values. In particular, for the most central
of jet fragmentation functions in heavy-ion collisions
events a peak is visible at A ≈ 0.5. The different
J
resemblethoseofPYTHIAdi-jetevents(withpartons
characteristicsoftheA distributionincentralPb–Pb
J
fragmentinginthevacuum)forbothleadingandsub-
collisions with respect to p–p indicate an increased
leading jets, in different centrality bins and different
rateofhighlyasymmetricdi-jetevents. Thisobserva-
di-jet imbalance (A ) bins.
tionisconsistentwithadegradationofthepartonen- J
ergywhiletraversingthemediumproducedinPb–Pb
collisions. The azimuthal separation between the two
jetshasalsobeenstudiedandshowsthatthetwojets
are essentially back-to-back in all centrality bins with
V. SUMMARY AND CONCLUSIONS
no additional azimuthal broadening relative to that
expected from QCD effects in vacuum.
Di-jet imbalance studies have also been performed The first Pb–Pb run at the LHC enabled the study
by the CMS collaboration [50] in a different energy of heavy ion physics at a center of mas√s energy about
range(p thresholdsat120and50GeV/cforthelead- 14 times higher than at the largest s attained at
t
ing and sub-leading jets respectively) and lead to the RHIC. ALICE, ATLAS and CMS experiments col-
same conclusion about the presence of a significant lected data during this first heavy ion run demon-
excess of imbalanced di-jets in central collisions. This stratingexcellentperformanceandcomplementaryca-
modification of the jet momentum balance implies a pabilities, that allowed (together with the impres-
corresponding modification in the distribution of the sive performance of the accelerator) a large variety
jet fragmentation products which can be addressed ofhighqualityresultsforalltheexperimentalobserv-
by studying track-jet correlations. In particular, the ablesthathavebeenstudied. Theoverallpicturethat
missingjetenergycanbeeithertransportedoutofthe emerges from the “soft” (low pt) sector is that the
cone area used to define the jet, or it can be carried medium produced in Pb–Pb collisions at the LHC is
bylow-momentumparticleswhicharenotmeasuredin strongly interacting, has exceptionally low viscosity
the calorimeter jets. The overall momentum balance and can be well described by hydrodynamics. In gen-
in the di-jet events was studied using the projection eral, the results from soft physics observables show
ofthemissingp ofreconstructedchargedtracksonto a smooth evolution from RHIC to LHC. By compar-
t
the leading jet axis: ingthehighprecisionmeasurementsattheLHCwith
model predictions and with the results obtained at
(cid:54)p(cid:107) =(cid:88)−pi cos(φ −φ ), (4) lower energies at the SPS and at RHIC, it will be
T T i Leading Jet
possible to obtain a deeper insight into the proper-
i
ties of the QGP. Furthermore, with the first heavy
The results of this analysis show that, for charged- ion run, the LHC experiments started to exploit the
particle jets, if tracks down to p = 0.5 GeV/c are abundance of high p and large mass probes which is
t t
considered, the momentum balance is indeed recov- a unique and novel aspect of heavy-ion collisions at
eredwithinuncertaintiesalsoforthemostcentralcol- LHC energies. From the hard physics sector we ex-
lisions. In particular, a large negative contribution to pect to find answers to some of the issues that have
(cid:104)p(cid:54) (cid:107)(cid:105) (i.e. in the direction of the leading jet) comes not been addressed completely by the experiments at
T
from particles with p >8 GeV/c that is balanced by lower collision energies, such as the J/ψ (and quarko-
t
the contribution from lower p particles, with a large nia) suppression, the prediction of a different energy
t
fractionofthebalancingmomentumcarriedbytracks loss of beauty, charm and light hadrons and the in-
with p < 2 GeV/c. Further insight is obtained by medium modification of jet fragmentation properties.
t
studying (cid:104)p(cid:54) (cid:107)(cid:105) separately for tracks inside and outside
T
(cid:112)
cones of size ∆R = ∆ϕ2+∆η2 = 0.8 around the
20
XXXI PHYSICS IN COLLISION, Vancouver, BC Canada, August 28 - September 1, 2011
Acknowledgments M. Masera and L. Ramello for useful discussions and
suggestions during the preparation of the conference
The author wishes to thank R. Arnaldi, E. Bruna, talk and of these proceedings.
A. Dainese, P. Giubellino, R. Granier de Cassagnac,
[1] F. Karsch, J. Phys. Conf. Ser. 46 (2006) 122. etal.,PHOBOSCollaboration,J.Phys.G35,104080
[2] R.J.GlauberinLecturesinTheoreticalPhysics,NY, (2008).
1959, Vol. 1, 315. [28] J. Casalderrey-Solana, E. V. Shuryak, D. Teaney, J.
[3] M. Miller et al., Ann.Rev.Nucl.Part.Sci. 57 (2007) Phys. Conf. Ser. 27 (2005) 22.
205. [29] B. Alver and G. Roland, Phys. Rev. C 81 (2010)
[4] U. W. Heinz, arXiv:hep-ph/0407360v1 054905.
[5] P. F. Kolb, U. W. Heinz in Hwa, R.C. (ed.) et al.: [30] K. Aamodt et al., ALICE Collaboration,
Quark gluon plasma, 634-714 [nucl-th/0305084]. arXiv:1109.2501 [nucl-ex].
[6] M. Luzum and P. Romatschke, Phys. Rev. Lett. 103 [31] S. Chatrchyan et al, CMS Collaboration, JHEP 07
(2009) 262302. (2011) 076.
[7] K. Aamodt et al., ALICE Collaboration, Phys. Rev. [32] T. Matsui, H. Satz, Phys. Lett. B178 (1986) 416.
Lett. 105 (2010) 252301. [33] A.Majumder,M.VanLeeuwen,Prog.Part.Nucl.Phys
[8] K. Aamodt et al., ALICE Collaboration, Phys. Rev. 66 (2011) 41.
Lett. 106 (2011) 032301. [34] D. d’Enterria, Springer Verlag. Landolt-Boernstein
[9] G. Aad et al, ATLAS Collaboration, Vol. 1-23A, arXiv:0902.2011 [nucl-ex].
arXiv:1108.6027v1 [hep-ex], submitted to PLB. [35] P. Braun-Munzinger, J. Stachel, arXiv:0901.2500
[10] S.Chatrchyanet al,CMSCollaboration,J.HighEn- [nucl-th].
ergy Phys. 08 (2011) 141. [36] G. Aad et al., ATLAS Collaboration, Phys. Lett. B
[11] J.D. Bjorken, Phys. Rev. D 27 (1983) 140. 697 (2011) 294.
[12] A. Toia for the ALICE Collaboration, QM2011 pro- [37] CMS Collaboration, CMS PAS HIN-10-006,
ceedings, arXiv:1107.1973 [nucl-ex]. http://cdsweb.cern.ch/record/1353586.
[13] CMS Collaboration, CMS-PAS-HIN-11-003, [38] Z. Tang for the STAR Collaboration, QM2011 pro-
http://cdsweb.cern.ch/record/1354215. ceedings, arXiv:1107.0532 [hep-ex]
[14] K. Aamodt et al., ALICE Collaboration, Phys.Lett. [39] G. Martinez Garcia for the ALICE Collaboration,
B 696 (2011) 328. QM2011 proceedings, arXiv:1106.5889 [nucl-ex]
[15] M.FlorisfortheALICECollaboration,QM2011pro- [40] S. Chatrchyan et al, CMS Collaboration, Phys. Rev.
ceedings, arXiv:1108.3257 [hep-ex]. Lett. 107 (2011) 052302.
[16] E. Schnedermann, J. Sollfrank, and U. Heinz, Phys. [41] Nuclear Physics A, Volume 757 (2005) 1-283.
Rev. C 48 (1993) 2462. [42] K. Aamodt et al., ALICE Collaboration, Phys.Lett.
[17] J. Y. Ollitrault, Phys. Rev. D 46 (1992) 229. B 696 (2011) 30.
[18] S. A. Voloshin, A. M. Poskanzer and R. Snellings, [43] CMS Collaboration, CMS-PAS-HIN-10-005,
arXiv:0809.2949 [nucl-ex]. http://cdsweb.cern.ch/record/1352777.
[19] K. Aamodt et al., ALICE Collaboration, Phys. Rev. [44] CMS Collaboration, CMS-PAS-HIN-11-002,
Lett. 105 (2010) 252302. http://cdsweb.cern.ch/record/1352779.
[20] G. Aad et al, ATLAS Collaboration, [45] S. Catani et al., JHEP 05 (2002) 028.
arXiv:1108.6018v1 [hep-ex], submitted to PLB. [46] S. Chatrchyan et al, CMS Collaboration, Phys. Rev.
[21] CMS Collaboration, CMS-PAS-HIN-10-002, Lett. 106 (2011) 212301.
http://cdsweb.cern.ch/record/1347788. [47] A. Dainese for the ALICE Collaboration, QM2011
[22] C. Shen, U. W. Heinz, P. Huovinen, H. Song, proceedings, arXiv:1106.4042 [nucl-ex].
arXiv:1105.3226 [nucl-th]. [48] M. Ploskon for the STAR collaboration, Nucl.Phys.
[23] R. Snellings for the ALICE Collaboration, QM2011 A830(2009)255c,E.BrunafortheSTARcollabora-
proceedings, arXiv:1106.6284 [nucl-ex]. tion, arXiv:1110.2795 [nucl-ex].
[24] K. Aamodt et al., ALICE Collaboration, Phys. Rev. [49] G.Aadetal,ATLASCollaboration,Phys.Rev.Lett.
Lett. 107 (2011) 032301. 105 (2010) 252303.
[25] ATLAS Collaboration, ATLAS-CONF-2011-074, [50] S. Chatrchyan et al, CMS Collaboration, Phys. Rev.
http://cdsweb.cern.ch/record/1352458. C 84 (2011) 024906.
[26] CMS Collaboration, CMS-PAS-HIN-11-005, [51] ATLAS Collaboration, ATLAS-CONF-2011-075,
http://cdsweb.cern.ch/record/1361385. http://cdsweb.cern.ch/record/1353220.
[27] J. Adams et al., STAR Collaboration, Phys. Rev. C [52] CMS Collaboration, CMS-PAS-HIN-11-004,
73 (2006) 064907, A. Adare et al., PHENIX Collab- http://cdsweb.cern.ch/record/1354531.
oration, Phys. Rev. C 78 (2008) 014901, B. Alver
20