ϒ suppression at forward rapidity in Pb–Pb collisions at √sNN=5.02 TeV

a r t i c l e i n f o a b s t ra c t

Articlehistory:

Received23May2018

Receivedinrevisedform28October2018 Accepted12November2018

Availableonline29December2018 Editor: M.Doser

Inclusive ϒ(1S) and ϒ(2S) production have been measuredin Pb–Pbcollisions at the centre-of-mass energypernucleon–nucleon pair√_s

NN=⁵.02 TeV,usingthe ALICEdetector atthe CERNLHC.The ϒ mesons are reconstructed in the centre-of-mass rapidity interval 2.5<y<4 and in the transverse- momentum range pT<15 GeV/c,via their decays to muon pairs. In thisLetter, we present results onthe inclusiveϒ(1S) nuclearmodificationfactor RAA as afunctionofcollisioncentrality, transverse momentumandrapidity.Theϒ(1S) andϒ(2S)RAA,integratedoverthecentralityrange0–90%,are0.37± 0.02(stat)±⁰.03(syst)and0.10±⁰.04(stat)±⁰.02(syst),respectively,leadingtoaratioR^ϒ(_AA^2S)/R^ϒ(_AA^1S)of 0.28±⁰.12(stat)±⁰.06(syst).Theobservedϒ(1S) suppressionincreaseswiththecentralityofthecollision andnosignificantvariationisobservedasafunctionoftransversemomentumandrapidity.

©^{2018PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense} (http://creativecommons.org/licenses/by/4.0/).FundedbySCOAP³.

1. Introduction

Adetailedstudyofthepropertiesof theQuark–Gluon Plasma (QGP) [1] is the main goal of heavy-ion experiments at ultra- relativisticenergies [2–6].Quarkonia,i.e.boundstatesofcharmor bottomquark–antiquarkpairs,aresensitiveprobesofcolordecon- finement,duetotheQuantum-ChromoDynamicsDebyescreening mechanism [7–9] leading to quarkonium suppression. Moreover, thevariousquarkoniumstateshavedifferentbindingenergiesand thereforedifferentdissociationtemperatures ina QGP, leading to sequentialsuppression[7,10]. Theoryestimates[11] indicate that bottomoniumformationmayoccurbeforeQGPthermalization[12]

becauseofthelargebottomquark mass.Inthissituation,aquan- titativedescriptionofthe influenceofthemedium on thebound states becomes challenging. While the dissociation temperatures varysignificantly between differentmodels [8,9], it is commonly acceptedthatthewidthsofthespectralfunctionsofthebottomo- nium statesincrease compared to the widthsin vacuum, dueto the high temperature of the surrounding medium [13]. Finally, taking into account that feed-down processes from higher-mass resonances(around40%fortheϒ(1S) and30% forthe ϒ(2S) [9]) arenot negligible,the evaluationof themedium temperaturevia bottomoniummeasurementsremainsacomplexendeavour.

Thefirst studies ofquarkonium productionin heavy-ioncolli- sions were devoted to charmonium states, anda suppression of theiryieldswas observedattheSPS [14–16],atRHIC [17,18] and

E-mailaddress:alice-publications@cern.ch.

attheLHC [19–22]. TheweakerJ/ψ suppression observedatLHC energies, where the centre-of-mass energy per nucleon–nucleon pair (√

s_NN) is one order of magnitude larger than at RHIC, can be explained by means of a competitive (re)generation mecha- nism, which occurs during the deconfined phase and/or at the hadronization stage [23–26]. Thisproduction mechanismstrongly dependson the(re)combination probability of deconfinedquarks presentinthemediumandthusontheinitialnumberofproduced cc pairs. Theeffect hasbeenfound tobe more importantatlow pT andinthemostcentralcollisions[22,20,27].

The high-energy collisions delivered by the LHC allow for a detailed studyof bottomonium states. For bottomonium produc- tion, perturbative calculations of production rates in elementary nucleon–nucleoncollisionsaremorereliablethanforcharmonium yieldsduetothehighermassofthebottomquarkwithrespectto charm.Sincethe numberofproducedbb pairsincentralheavy-ion collisions amountto afewpairsper eventattheLHC,theproba- bilityfor(re)generationofbottomoniathrough(re)combination is muchsmallerthaninthecaseofcharmonia.

The ϒ(1S) nuclearmodificationfactor R_AA isquantified asthe ratioof theϒ(1S) yieldinnucleus–nucleus collisions tothe pro- ductioncrosssection measuredinpp collisionsscaled bythenu- clearoverlapfunction^TAA^{.ThelatterisobtainedviatheGlauber} model [28,29]. Astrong suppression ofthe ϒ(1S) state inPb–Pb collisions has been observed at √

sNN=².76 TeV by ALICE [30]

andCMS[31,32] intherapidity ranges2.5<y<4 and|^y|<2.4, respectively. The suppression increaseswith the centralityof the collision, reaching about60% and 80% forthe mostcentral colli- sionsatmid [32] andforwardrapidity [30],respectively.Moreover, https://doi.org/10.1016/j.physletb.2018.11.067

0370-2693/©^{2018PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense}(http://creativecommons.org/licenses/by/4.0/).FundedbySCOAP³.

the ϒ(2S) suppressionreachesabout 90%and forϒ(3S) dataare compatiblewithacompletesuppression[32].Asafunctionof pT theϒ(1S) RAA,measured forpT<20 GeV/c byCMS [32],iscom- patiblewithaconstantvalue.Whenconsideringthe y-dependence resultingfromthecomparisonofALICEandCMSresults,thereis an indication for a stronger suppression at forward y. Transport models [26,33] aswell asan anisotropic hydro-dynamical model [34] fairlyreproducetheexperimental observationsofCMS,while theytendtooverestimatetheR_AAvaluesmeasuredbyALICE.

The bottomoniumsuppression due tothe QGPshould be dis- entangledfromthesuppressionduetoColdNuclearMatter(CNM) effects, such as the nuclear modification of the parton distribu- tionfunctionsduetoshadowing[35,36],aswellaspartonenergy loss[37].Theseeffectsonthebottomoniumproductionwerestud- iedinp–PbcollisionsbyALICE[38] andLHCb[39],who reported for the ϒ(1S) a nuclear modification factor slightly lower than unityatforward rapidityandcompatiblewithunity atbackward rapidity, although with significant uncertainties. Recently, ATLAS resultsindicateasignificantsuppressionoftheϒ(1S) for pT <40 GeV/c aroundmid-rapidity [40]. Additional measurements at for- ward/backwardrapiditywithhigherstatistics,areneededtofully constrain the models and perform a meaningful extrapolation of CNMeffectstoPb–Pbcollisions.

In this Letter we present the first results on the ϒ(1S) and ϒ(2S) R_AA measured by the ALICE Collaboration in Pb–Pb colli- sions at √

s_NN=⁵.02 TeV. The pp reference cross sections used inthe RAA calculationshavebeendetermined byaninterpolation procedurebasedonvariousALICE[41,42] andLHCb[43,44] results atdifferentenergies.Thenuclearmodificationfactorfortheϒ(1S) ispresentedasafunctionofthecentralityofthecollisionandalso differentially in pT andrapidity. Forthe ϒ(2S), an RAA value in- tegratedover thecentralityofthe collisionisquoted. Finally,the resultsarecomparedtotheoreticalcalculations.

2. Experimentalapparatusanddatasample

An extensivedescriptionof theALICEapparatus canbe found in[45,46].TheanalysispresentedinthisLetterisbasedonmuons detected atforwardrapidity (2.5<y<4)¹ withthemuon spec- trometer [47].ThedetectorsrelevantforϒmeasurementsinPb–Pb collisionsaredescribedbelow.

TheSiliconPixelDetector,correspondingtothetwo innermost layers ofthe Inner TrackingSystem [48], is usedforthe primary vertexdetermination.Theinnerandouterlayercoverthepseudo- rapidityranges|

η

_|<2 and|

η

_|<1.4,respectively.

TheV0scintillatorhodoscopes [49] providethecentralityesti- mate.They are made oftwo arraysof scintillatorsplaced in the pseudo-rapidity ranges 2.8<

η

<5.1 and −³.7<

η

<−¹.7. The logical AND of the signals from the two hodoscopes constitutes theMinimumBias(MB)trigger.TheMBtriggerisfullyefficientfor thestudied0–90%mostcentralcollisions.

TheZeroDegreeCalorimeters (ZDC)areinstalled at±¹¹².5 m fromthe nominalinteraction point along the beam line. Eachof thetwoZDCsiscomposedoftwosamplingcalorimetersdesigned for detecting spectator protons, neutrons andnuclear fragments.

Theevaluationofthesignalamplitude oftheZDCsallows forthe rejectionofeventscorrespondingtoanelectromagneticinteraction ofthecollidingPbnuclei[50].

Themuonspectrometercovers thepseudorapidityrange−⁴<

η

<−².5.Itiscomposedofafrontabsorber, whichfiltersmuons

1 IntheALICEreferenceframe,themuonspectrometercoversanegativeηrange andconsequentlyanegativeyrange.Wehavechosentopresentourresultswitha positiveynotation.

upstream ofthemuon tracker,consistingoffivetracking stations with two planes of cathode-pad chambers each, andof a dipole magnetprovidinga 3 T·^{mintegratedmagneticfield.Downstream} of the tracking system, a 1.2 m thick iron wall stops efficiently the punch-through hadrons. The muon trigger systemis located downstream of the iron wall and consists of two stations, each one equippedwithtwoplanesofResistive PlateChambers (RPC), withanefficiencyhigherthan 95% [51].Themuon-triggersystem isabletodeliversingleanddimuontriggersselectingmuonswith p_T largerthana programmablethreshold,viaan algorithmbased ontheRPCspatialinformation [52].Throughoutitsentirelength,a conicalabsorbershieldsthemuonspectrometeragainstsecondary particles produced by the interaction of primary particles in the beampipe.

The trigger condition used for data taking is a dimuon- MinimumBias(

μμ

-MB)triggerformedby thelogicalANDofthe MBtriggerandanunlike-signdimuontriggerwitha pT threshold of1 GeV/cforeachofthetwomuons.

ThecentralityestimationisperformedusingaGlauberfittothe sumofthesignalamplitudesoftheV0scintillators [53–55].Cen- tralityrangesaregivenaspercentagesofthetotalhadronicPb–Pb cross section.In additionto thecentrality,the Glauber modelal- lows an estimate of the average number of participantnucleons ^Npart^{,of the averagenumber ofbinary collisions N}coll ^andof thenuclearoverlapfunction^TAA^{,foreachcentralityinterval[56].}

In the present analysis, the data sample corresponds to an inte- gratedluminosity Lint≈^{225 μb−1 inthecentralityinterval0–90%}

that hasbeendividedintofourcentralityclasses:0–10%,10–30%, 30–50%and50–90%.

3. Dataanalysis

The evaluationof RAA is performedthroughthe followingex- pression:

R_AA

=

_BR ^Nϒ

ϒ→μ⁺μ⁻·(A×ε)_ϒ→μ⁺μ⁻·^Nμμ-MB·^Fnorm·σpp^ϒ· ^TAA , (1) where N^ϒ is the numberof detected resonance decays to muon pairs, while BR_ϒ→μ⁺μ⁻=(2.48±⁰.05)% and(1.93±⁰.17)% are the branching ratios for the dimuon decay of ϒ(1S) and ϒ(2S), respectively [57].The (A×

ε

)_ϒ→μ+μ⁻ factoristheproductofac- ceptanceanddetectionefficiencyfortheϒstateunderstudy.The normalization factor Nμμ-MB·^Fnorm is the product of the num- ber of analyzed

μμ

-MB events and the inverse of the proba- bility to obtain an unlike-sign dimuon trigger in a MB-triggered event [22].Adatasetof1.5·^{109 minimumbiasequivalentevents,} Nμμ-MB·^Fnorm,hasbeenusedforbottomoniummeasurements.Fi- nally,

σ

_pp^ϒ is the referencepp cross section and ^TAA ^represents thenuclearoverlapfunction [55].

Thesignalyieldsareevaluatedbyperformingfitstothe

μ

⁺

μ

⁻

invariant massdistributions.Inordertoimprovethepurityofthe dimuonsampleasetofselectioncriteria [30] hasbeenappliedon the muon tracks, includingthe requestof the matchingbetween the tracks reconstructed in the trigger and tracking detectors of themuonspectrometerandacutonthetracktransversemomen- tum(pT>2 GeV/c).Thelattercuthasasmalleffect(∼^2%)onthe number of detected resonances. The raw ϒ yields are extracted using thesumof threeextended Crystal Ball(CB)functions [58], one foreach of ϒ(1S), ϒ(2S) and ϒ(3S). The extended CB func- tion consistsof a Gaussian corewith non-Gaussian tails onboth sidestotakeintoaccounttheradiativecontributionsoftheϒpro- duction andtheabsorber effects ofmuon energylossin thelow masstail,whereas thehighmasstailisattributedtothemultiple

Fig. 1.Redandmagentasolidlinescorrespondtoϒ(1S) andϒ(2S) signalfunctions,respectively.Thecontributionfromϒ(3S) yieldiscompatiblewithzero.Dottedbluelines representthebackground(left)andresidualbackground(right),respectively.Thesumofthevariousfunctionsisalsoshownasasolidblueline.

Coulombscatteringinthefrontabsorberandthemomentumreso- lutionofthetrackingchambers.Thebackgroundisfittedwiththe sumoftwo exponentialfunctions(see left panel ofFig.1). Since the signal-to-background (S/B)ratio is low in the tail regions of theextendedCBfunctions, thetailparameters arefixedtovalues obtainedfrom the Monte Carlo(MC) simulation. The mass posi- tionand thewidth parameters of theϒ(1S) are left free forthe integratedspectrum(i.e.centralityclass0–90%,pT<15 GeV/c and 2.5<y<4). Whereas for the signal extraction asa function of centrality,themasspositionandwidth(160±^{15 MeV/c2)ofthe} ϒ(1S) arefixedto thevaluesobtainedinthefittothecentrality- integrated(0–90%)massspectrum.Finally,forstudiesasafunction of pT and y, the mass position and the width obtained for the centrality-integratedmass spectrum are scaled accordingto their evolution observedin the MC. Due to the poorS/B ratio forthe highermassstates,thevaluesofthemassoftheϒ(2S) andϒ(3S) arefixed to thePDG massdifferenceswithrespect tothe ϒ(1S), and the ratio of ϒ(2S) (ϒ(3S)) to ϒ(1S) widths is fixed to val- ues fromthe MC simulation, i.e.1.03 (1.06). In the fit shownin Fig.1onlysignalscorresponding totheϒ(1S) and ϒ(2S) arevis- ible,sincethe ϒ(3S) contribution iscompatiblewithzero events.

Alternatively, the combinatorial background is modeled with the event-mixingmethod.Inthisapproach,aninvariant massdimuon spectrum is constructed by pairing muonsfrom different events with similar multiplicities as described in [22]. The combinato- rialbackgroundisthensubtractedfromtherawdimuonspectrum (rightpanel ofFig. 1)andtheresultingdistributionisfittedwith thesumofthreeextended CBandanexponential functiontoac- countfortheresidualbackground.Finally,thenumberofdetected ϒ resonances, N^ϒ,is obtained as the average [58] of the fitting methodsdescribedabove(andalsobelowinthediscussiononsig- nalsystematics),leadingtoN^ϒ(1S)=¹¹²⁶±^53(stat)±47(syst) and N^ϒ(2S)=⁷⁷±^33(stat)±^17(syst).

Themeasured ϒyields, N^ϒ,arecorrected forthe detectorac- ceptanceandefficiencyusingMCsimulations.Sincetheoccupancy ofthedetectorvarieswiththecentralityofthecollisions,thegen- eratedϒdecaysareembeddedintorealMBeventstosimulatethe variousparticlemultiplicityscenariosasindata.The pTandydis- tributionsofthegeneratedϒareobtainedfromexistingpp mea- surements [59–61] using the interpolation procedure described in [62]. The EKS98 nuclear shadowing parameterization [35] is used to includean estimate of CNM effects. Since available data favora small ornull polarization for ϒ(1S) [63–66], an unpolar- izedproductionisassumed. Thevariations oftheperformance of

thetrackingandtriggeringsystemsthroughoutthedata-takingpe- riodaswellastheresidualmisalignmentofthetrackingchambers aretakenintoaccountinthesimulation.The A×

ε

values,forthe range pT<15 GeV/c, 2.5<y<4 and the0–90% centralityclass are0.263 and0.264 fortheϒ(1S) and ϒ(2S),respectively,witha negligible statisticaluncertainty. A decrease of 2% is observed in A×

ε

forthe0–10%centralcollisionswithrespecttothe50–90%

sample dueto the higher occupancy in the most central events.

The A×

ε

ishigherby20%in3<y<3.5 comparedtothevalues at2.5<y<3 and3.5<y<4 mainlyduetothegeometricaccep- tanceofthedetector,whereasithasnovariationasafunctionof pT.ThesystematicuncertaintyonA×

ε

isdiscussedbelow.

The systematic uncertainty on the signal extraction is evalu- atedusingvarious functionsformodellingthebackgroundshape, as well as adopting two fitting ranges, i.e. (7–14) GeV/c² and (7.5–14.5)GeV/c².Thetailparametersofthesignalfunctionshave been varied using estimates provided by two MC particle trans- port models:GEANT4 [67] andGEANT3 [68].Inthecentrality, p_T or y differential studies, the mass position and width are also varied by amounts,whichcorrespond totheuncertainties on the masspositionandthewidthreturnedbythefittothecentrality- integratedinvariantmassspectrum. Theratioofϒ(2S) (ϒ(3S))to ϒ(1S) widthsisvariedfrom1(1)to1.06(1.12).ThevaluesofN^ϒ andtheirstatisticaluncertaintiesareobtainedbytakingtheaver- ageof N^ϒ andofthecorresponding statisticaluncertainties from the variousfits. Thisprocedure isapplied to bothfits ofthe raw and combinatorial-background subtracted spectra. The systematic uncertainties areestimatedastherootmeansquare ofthedistri- bution of N^ϒ obtained from the various fits. The effect induced by the pT>2 GeV/c cutonsingle muonsonthe A×

ε

-corrected ϒyields was estimatedby varyingthatcut by±^{0.2GeV/c inthe} MC.A±^{2%maximumvariationon Nϒ}/(A×

ε

)wasobservedand includedinthesystematicuncertainties.

Various sources contribute to the systematic uncertainties of A×

ε

,suchasthe p_T and y shapesoftheinput distributionsfor theMCsimulations,thetriggerefficiency,thetrackreconstruction efficiencyandfinallythematchingefficiencybetweentracksinthe muon tracking and triggering chambers. Various sets of simula- tions areproducedwithdifferentϒinput pT and y distributions, obtainedfromempiricalparameterizationsand/orextrapolationsof availabledatasetsatdifferentenergies.Themaximumrelativedif- ferenceof A×

ε

forthevariousshapesistakenasthesystematic uncertaintyduetotheinputMC.Inordertocalculatethesystem- aticuncertaintyontriggerefficiency,thetriggerresponsefunction

Table 1

SummaryofthesystematicuncertaintiesfortheRAAcalculation.TypeI(II)referstocorrelated(uncorrelated)systematicuncertainties.

Sources ϒ(1S) ϒ(2S)

Centrality y pT Integrated Integrated

Signal extraction 4.3–6.1%(II) 4.2–6.8%(II) 5.2–8.7%(II) 4.1% 21.7%

MuonpTcut 0.3–2.4%(II) 0.1–1.2%(II) 0.1–2.4%(II) 0.7% 0.7%

Input MC 0.9%(I) 0.6–2.6%(II) 1–1.4%(II) 0.9% 0.9%

Tracker efficiency 3%(I) and 0–1%(II) 1%(I) and 3%(II) 1%(I) and 3%(II) 3% 3%

Trigger efficiency 3%(I) 1.4–3.7%(II) 0.4–2.6%(II) 3% 3%

Matching efficiency 1%(I) 1%(II) 1%(II) 1% 1%

Centrality 0.2–2.4%(II) – – – –

Fnorm 0.5%(I) 0.5%(I) 0.5%(I) 0.5% 0.5%

^TAA 3.1–5.3%(II) 3.2%(I) 3.2%(I) 3.2% 3.2%

BR_ϒ→μ+μ⁻·σ_ϒ^pp 6.3%(I) 6.6–11.3%(II) 5.5–11.5%(II) 6.3% 7.5%

for single muons is evaluated using eitherMC or data. The two response functions are then separately applied to simulations of an ϒ sample and the difference obtained for the ϒ reconstruc- tion efficiencyis taken assystematic uncertainty. The systematic uncertainty on the tracking efficiency is obtained starting from an evaluation of the single muon tracking efficiency in MC and data.Thisevaluationisperformedviaaprocedure,detailedin[22], basedontheredundancyofthetrackingchamberinformation.The dimuontrackingefficiencyisthenobtainedbycombiningthesin- gle muon efficiencies and the systematicuncertainty is taken as thedifferenceofthevaluesobtainedwiththeprocedurebasedon MCanddata.Themuon tracksfordataanalysisarechosenbased on a selection on the

χ

² ofthe matching betweena trackseg- mentinthetriggersystemwithatrackinthetrackingchambers.

The matching systematics are obtainedby varying the

χ

² selec- tioncut indata andMC andcomparingtheeffects onthe muon reconstructionefficiency [22].

The systematic uncertainty on the centrality measurement is evaluated by varying the V0 signal amplitude by ±⁰.5% corre- spondingto90%ofthehadroniccrosssectioninPb–Pbcollisions, usedasanchorpointtodefine thecentralityclasses.The system- atic uncertainty on the evaluationof

σ

_pp^ϒ is detailedin the next section.Finally,thesystematicuncertaintyevaluationofFnormand ^TAA ^{are described} in [22] and [53], respectively. The different systematicuncertaintysourcesonthe RAAcalculationaresumma- rizedinTable1.Iftheabovementionedsystematicuncertaintyis correlatedasa functionofcentrality, pT or y,itisquotedascor- related(type I) systematic uncertainty, otherwise it istreated as uncorrelated(typeII).

4. Proton–protonreferencecrosssections

Theppreferencecrosssectionforϒ(1S) and ϒ(2S) production arecomputedbymeansofaninterpolationprocedureasdescribed forϒ(1S) in [69].Theenergyinterpolationfortheϒcrosssection, asa function of rapidity andfor the pT and y integratedresult, usesthemeasurements ofϒ productioncrosssectionsinppcol- lisionsat√

s=^{7 and8TeV byALICE}[41,42] andat√

s=².76,7 and8TeVby LHCb[43,44].Theinterpolationisperformedbyus- ing variousempiricalfunctionsand, inaddition,the shapeofthe energydependenceof thebottomonium crosssectionscalculated usingtwo theoreticalmodels,i.e.theLeadingOrderColour Evap- orationModel(LO-CEM)[70] andtheFixedOrderNext-to-Leading Logs(FONLL)model [71].The lattergivescross sectionsforopen beauty,whichishereusedasaproxytostudytheevolutionofthe bottomoniumcross section [69]. Theenergy interpolationfor the ϒ(1S) crosssectionasafunctionof pTisbasedonLHCbmeasure- ments only, since the pT coverage of the results of this analysis (pT<15 GeV/c) is more extended than that of the correspond- ing ALICEpp data (pT<12 GeV/c). The result of the interpola- tion procedure gives BR_ϒ(1S)→μ⁺μ⁻·

σ

pp^ϒ(1S)=¹²²¹±77(syst) pb

Table 2

Theinterpolatedbranchingratiotimescrosssectionofϒ(1S) forthepTandybins understudy.Thequoteduncertaintiesaresystematic.

pT(GeV/c) y BRϒ(1S)→μ⁺μ⁻·σpp^ϒ(1S)(pb) [0–2]

[2.5–4]

226±²⁶

[2–4] 361±²⁰

[4–6] 288±24

[6–15] 311±23

[0–15]

[2.5–3] 506±⁵⁷

[3–3.5] 415±²⁸

[3.5–4] 288±²⁴

andBR_ϒ(2S)→μ⁺μ⁻·

σ

pp^ϒ(2S)=³⁰²±23(syst) pb assumingunpolar- ized quarkonia and integrating over the ranges 2.5<y<4 and pT<15 GeV/c.Theuncertaintiescorrespondtothequadraticsum of two terms. The first term dominates the total uncertainty on the interpolated value andreflects the statistical and systematic uncertainties on the data points used in the interpolation pro- cedure. The second term is relatedto the spreadamong the in- terpolated cross sections obtained by using either the empirical functionsorthe energydependenceestimatedfromthetheoreti- cal modelsmentionedabove.Thenumericalvaluesobtainedfrom theinterpolationprocedurearesummarizedinTable2forthevar- iouskinematicrangesusedintheanalysis.

5. Results

The nuclearmodificationfactorsforinclusiveϒ(1S) andϒ(2S) productioninPb–Pb collisionsat√

sNN=⁵.02 TeV fortheranges p_T <15 GeV/c, 2.5< y < 4 and the 0–90% centrality class are R^ϒ(_AA^1S)=⁰.37±⁰.02(stat)±⁰.03(syst) and R^ϒ(_AA^2S)=⁰.10± 0.04(stat)±⁰.02(syst), respectively. The ratio R^ϒ(_AA^2S)/R^ϒ(_AA^1S) is 0.28±⁰.12(stat)±⁰.06(syst). Since the decay kinematics of the two ϒ states is very similar, mostof the systematic uncertainty sources entering the ratiocancel out except those on the signal extraction and on the pp crosssection, which are the dominant contributions to the total systematic uncertainty. The measure- ments show a strong suppression for both bottomonium states with themore weakly bound state being significantly moresup- pressed. The ratio between the ϒ(1S) RAA at √

sNN=⁵.02 TeV and2.76TeV [30] is 1.23±⁰.21(stat)±⁰.19(syst).The sourcesof systematic uncertainties entering the calculation of the ratio are considered uncorrelated, except forthe ^TAA ^{component, whose} uncertaintycancelsout.Theratioiscompatiblewithunitywithin uncertainties.

Thecentrality, pT andy dependencesoftheϒ(1S) RAA atfor- wardrapidityat√

s_NN=⁵.02 TeVareshowninFig.2.Adecrease of RAA with increasing centrality is observed down to R^ϒ(_AA^1S)= 0.34±⁰.03(stat)±⁰.02(syst) forthe0–10%mostcentralcollisions.

No significant p_T-dependence is observed up to p_T=^{15 GeV/c}

Fig. 2.Inclusiveϒ(1S)RAAasafunctionofcentrality(top),pT(left)andy(right)atforwardrapidityat√

sNN=5.02 TeV.ALICEresultsat√

sNN=2.76 TeVasafunction ofcentralityand yareshownforcomparison [30].Theverticalerrorbarsandtheboxesrepresentthestatisticalanduncorrelatedsystematicuncertainties,respectively.

Therelativecorrelateduncertaintyisshownasboxesatunity.ALICEϒ(1S)RAAmeasurementsat√_s

NN=⁵.02 TeVarecomparedtopredictionsfromtwotransportmodels [33,72] andonehydro-dynamicalmodel[34] asafunctionofcentrality(top),pT(left)andy(right).Seetextfordetailsonthemodels.

withinuncertainties.Thenuclearmodificationfactorshowsnosig- nificantdependence on rapidity.The ϒ(1S) RAA asa function of centralityandrapiditymeasuredbyALICEat√

sNN=².76 TeV [30]

are alsoshownin Fig. 2.Similar trends can be observedat both collisionenergies.

Theinclusiveϒ(1S) RAA measurementsarecompared inFig.2 to several calculations: two transport models (TM) [33,72] and onehydro-dynamicalmodel[34].Todescribethequarkoniummo- tioninthemedium,bothtransportcodes usearate-equationap- proach which accounts for both suppression and (re)generation mechanisms in the QGP. In the TM1 model [33] the evolution ofthe thermal medium is based ona thermal-fireball expansion while the TM2 model [72] uses a 2+1 dimensional version of the idealhydrodynamic equations.The two models usedifferent rateequationsandbothmodels includeafeed-downcontribution fromhigher-mass bottomonia to the ϒ(1S). In TM2,two sets of feed-down fractions are assumed. Finally, the ϒ(1S) production cross section in pp collisions at √

s=⁵.02 TeV in the rapidity range2.5<y<4 istaken asd

σ

_pp^ϒ(1S)/dy=²⁸.8 nb inTM1 and d

σ

pp^ϒ(1S)/dy=^{30 nb in TM2. Those values deviate by about 2}

σ

(TM1)and1.4

σ

(TM2)fromtheresultobtainedusingtheppinter- polationmethodreportedintheprevioussection.TM1predictions areshownasbandsaccountingforshadowingeffectsascalculated in [36]. The upper limit shownin Fig. 2 corresponds to the ex- treme case of the absence of shadowing while the lower limit

reflects a reduction of 30% due to shadowing. The TM1 model implements the feed-down fractions reported in[9]. In the TM2 model, the shadowing parameterization is based on EKS98 [35]

andthebandedgescorrespondtotwodifferentsetsoffeed-down fractions (27%from

χ

b; 11% fromϒ(2S+^3S) and37% from

χ

b; 12%fromϒ(2S+^3S))adoptedbytheauthors.Inthethirdmodel [34], a thermal suppression of the bottomonium states is calcu- latedusinga complex-valued heavy-quarkpotential parametrized bymeansoflatticeQCDandembeddedinamediumevolving ac- cording to 3+1d anisotropic hydrodynamics.In this recent study, the RAA shows no sensitivity to the plasma shear viscosity-to- entropy density ratio (4

π η

/s) parameter of the hydro evolution, whichisthereforesetto4

π η

/s=2 consistentwithparticlespec- trafits.Thebandofthemodelquantifiestheheavy-quarkpotential uncertainty,which hasbeenestimatedbyincludinga ±^{15% vari-} ation of theDebye mass ofthe QCD medium that is tuned by a fit to the real-part of the lattice in-medium heavy-quark poten- tial.Furthermore,thepredictionsshownarereferringtotheinitial momentum-spaceanisotropyparameterξ0=^0,whichcorresponds to a perfectly isotropic QGP at the starting point of the hydro- dynamicalevolutionat

τ

0=⁰.3 fm/c.Finally,thismodelaccounts forfeed-down contributions butit includes neither a (re)genera- tionmechanismnorCNMeffects.Thecentralitydependenceofthe ϒ(1S) RAAisfairlyreproducedbythemodelcalculationsinthetop panelofFig.2.ThedataarebestdescribedbyTM1when(re)gen- eration isincludedandby TM2when (re)generationisnot taken

into account. The hydro-dynamical model describesthe trend of thedata, thefact that thedatalie on theupper edgeofthe un- certaintybandforNpart>70 couldindicateasmallerDebyemass andthus astrongerheavy-quarkpotential.Thedataasafunction of p_T (bottomleftpanel ofFig.2)canbedescribed withorwith- out the(re)generation scenarioofthe TM1modelwhile showing agreementwiththehydro-dynamicalmodelfortheupperedgeof theuncertaintyband.Finally,the y-dependenceoftheϒ(1S) RAA isdescribed, within uncertainties, by the hydro-dynamicalmodel inthe bottom rightpanel ofFig. 2despite the possibly different trendbetweendataandcalculations.

Thelowϒ(1S) RAA reportedinthisLetterraisestheimportant question whetherdirectϒ(1S) aresuppressed atLHCenergies or only the feed-down contribution from higher mass states. How- ever,thelargeuncertaintiesofthecurrentmeasurements ofCNM effects[38–40] preventafirmconclusion.

6. Summary

The nuclearmodification factors ofinclusive ϒ(1S) and ϒ(2S) productionatforwardrapidity(2.5<y<4)andforpT<15 GeV/c inPb–Pbcollisionsat√

s_NN=⁵.02 TeVhavebeenmeasuredusing theALICEdetector.Theobservedϒ(1S) suppressionincreaseswith the centrality of the collision andno significant variation is ob- servedasafunctionoftransversemomentumorrapidity.Alarger suppression of the ϒ(2S) bound state compared to the ground stateisalsoreported.Transportanddynamicalmodelcalculations reproducequalitativelythecentralityandkinematicdependenceof theϒ(1S) nuclearmodificationfactor.

Acknowledgements

The ALICE Collaboration would like to thank all its engineers andtechnicians fortheir invaluablecontributionstotheconstruc- tionoftheexperimentandtheCERNacceleratorteamsfortheout- standingperformanceoftheLHCcomplex.TheALICECollaboration gratefully acknowledges the resources and support provided by all Gridcentres andtheWorldwide LHCComputing Grid(WLCG) collaboration. The ALICE Collaboration acknowledges the follow- ing funding agencies for their support in building and running the ALICE detector: A. I.Alikhanyan National Science Laboratory (YerevanPhysicsInstitute)Foundation(ANSL),StateCommittee of Science andWorld Federationof Scientists (WFS), Armenia; Aus- trian Academy of Sciences and Österreichische Nationalstiftung für Forschung, Technologie undEntwicklung, Austria; Ministry of CommunicationsandHighTechnologies,NationalNuclearResearch Center,Azerbaijan;Conselho NacionaldeDesenvolvimentoCientí- ficoeTecnológico(CNPq), UniversidadeFederaldo RioGrande do Sul(UFRGS), Financiadorade Estudose Projetos(Finep)andFun- dação de Amparo à Pesquisa do Estado de São Paulo (FAPESP), Brazil; Ministry of Science & Technology of China (MSTC), Na- tionalNaturalScienceFoundationofChina(NSFC)andMinistryof EducationofChina(MOEC),China;MinistryofScienceandEduca- tion,Croatia;MinistryofEducation,YouthandSportsoftheCzech Republic,CzechRepublic;TheDanishCouncilforIndependentRe- search — Natural Sciences, the Carlsberg Foundation and Danish NationalResearchFoundation(DNRF), Denmark;HelsinkiInstitute ofPhysics(HIP),Finland;Commissariatàl’EnergieAtomique(CEA) andInstitutNationaldePhysiqueNucléaireetdePhysiquedesPar- ticules (IN2P3) andCentre National de la Recherche Scientifique (CNRS), France; Bundesministerium für Bildung, Wissenschaft, Forschung und Technologie (BMBF) and GSI Helmholtzzentrum für Schwerionenforschung GmbH, Germany; General Secretariat forResearchandTechnology,MinistryofEducation,Researchand

Religions, Greece; National Research Development and Innova- tion Office, Hungary; Department ofAtomic Energy, Government of India (DAE), Department of Science and Technology, Govern- ment of India (DST), University Grants Commission, Government of India (UGC) and Council of Scientific and Industrial Research (CSIR), India; Indonesian Institute of Sciences, Indonesia; Centro Fermi - MuseoStorico della Fisica e Centro Studi e Ricerche En- rico Fermi andIstituto Nazionale diFisica Nucleare (INFN), Italy;

Institute forInnovativeScienceandTechnology,NagasakiInstitute of AppliedScience (IIST), Japan Societyforthe Promotion ofSci- ence(JSPS)KAKENHIandJapaneseMinistryofEducation,Culture, Sports, Science and Technology (MEXT), Japan; Consejo Nacional deCiencia(CONACYT)yTecnología,throughFondodeCooperación Internacional en Ciencia y Tecnología (FONCICYT) and Dirección General de Asuntos del Personal Academico (DGAPA), Mexico;

NederlandseOrganisatievoorWetenschappelijkOnderzoek(NWO), Netherlands; The Research Council of Norway, Norway; Commis- sion on Science and Technology for Sustainable Development in theSouth(COMSATS),Pakistan;PontificiaUniversidadCatólicadel Perú,Peru;MinistryofScienceandHigherEducationandNational Science Centre, Poland; Korea Institute of Science and Technol- ogyInformationandNationalResearchFoundationofKorea(NRF), Republic of Korea;Ministry of Educationand Scientific Research, Institute of Atomic Physics and Romanian National Agency for Science, Technology and Innovation, Romania; Joint Institute for Nuclear Research(JINR), MinistryofEducationandScienceofthe Russian FederationandNationalResearch CentreKurchatov Insti- tute, Russia; Ministry of Education, Science, Research and Sport of the Slovak Republic, Slovakia; National Research Foundation of South Africa, South Africa; Centro de Aplicaciones Tecnológi- cas yDesarrolloNuclear(CEADEN),Cubaenergía,Cuba andCentro de Investigaciones Energéticas, Medioambientales y Tecnológicas (CIEMAT),Spain;SwedishResearchCouncil(VR) andKnut&Alice WallenbergFoundation(KAW),Sweden;EuropeanOrganizationfor Nuclear Research, Switzerland; National Science and Technology Development Agency (NSDTA), Suranaree University of Technol- ogy (SUT) andOffice of theHigher EducationCommission under NRU projectofThailand,Thailand;TurkishAtomicEnergy Agency (TAEK),Turkey;NationalAcademyofSciencesofUkraine,Ukraine;

ScienceandTechnologyFacilitiesCouncil(STFC),UnitedKingdom;

NationalScienceFoundationoftheUnitedStatesofAmerica(NSF) andUnitedStatesDepartmentofEnergy,OfficeofNuclear Physics (DOENP),UnitedStatesofAmerica.

References

[1]E.V.Shuryak,Quark–gluonplasmaandhadronicproductionofleptons,photons andpsions,Phys.Lett.B78(1978)150,Yad.Fiz.28(1978)796.

[2]BRAHMSCollaboration, I.Arsene,etal.,Quarkgluonplasmaandcolorglass condensateatRHIC?ThePerspectivefromtheBRAHMSexperiment,Nucl.Phys.

A757(2005)1–27,arXiv:nucl-ex/0410020 [nucl-ex].

[3]PHOBOSCollaboration,B.B.Back,etal.,ThePHOBOSperspectiveondiscoveries atRHIC,Nucl.Phys.A757(2005)28–101,arXiv:nucl-ex/0410022 [nucl-ex].

[4]STARCollaboration,J.Adams,etal.,Experimentalandtheoreticalchallengesin thesearchforthequarkgluonplasma:theSTARCollaboration’scriticalassess- mentoftheevidencefromRHICcollisions,Nucl.Phys.A757(2005)102–183, arXiv:nucl-ex/0501009 [nucl-ex].

[5]PHENIXCollaboration, K. Adcox,et al., Formation ofdensepartonicmatter inrelativisticnucleus–nucleuscollisionsatRHIC:experimentalevaluationby thePHENIXcollaboration,Nucl. Phys.A757(2005)184–283,arXiv:nucl-ex/ 0410003 [nucl-ex].

[6]B.Muller,J.Schukraft,B.Wyslouch,FirstresultsfromPb+^{Pbcollisionsatthe} LHC,Annu.Rev.Nucl.Part.Sci.62(2012)361–386,arXiv:1202.3233 [hep-ex].

[7]T.Matsui,H. Satz,J/ψ suppressionbyquark–gluonplasmaformation,Phys.

Lett.B178(1986)416–422.

[8]N.Brambilla,etal.,Heavyquarkonium:progress,puzzles,andopportunities, Eur.Phys.J.C71(2011)1534,arXiv:1010.5827 [hep-ph].

[9]A.Andronic,etal.,Heavy-flavourandquarkoniumproductionintheLHCera:

fromproton–protontoheavy-ioncollisions,Eur.Phys.J.C76 (3)(2016),arXiv:

1506.03981 [nucl-ex].