Charmonium production in heavy-ion collisions: status and prespectives

XLVIII International Winter meeting on Nuclear Physics,Bormio (Italy) 25-29 January 2010

in Memoriam of Ileana Iori

E. ScomparinINFN Torino (Italy)

Charmonia suppression in AA collisions: a 25 year-long story

SPS RHIC LHC

17 GeV/c 200 GeV/c 5.5 TeV/c√s

year ~1990 ~2000 ~20101986

Last year, new high precision data (HERA-B, NA60, PHENIX/STAR) have become available

significant improvements in the overall understanding of the charmonium behavior in the hot medium

Outline

Study of charmonium production/suppression in pp, pA and AA collisions

AA collisions

• Charmonia suppression by color screening has been proposed, more than 20 years ago, as a signature of QGP formation

• Sequential suppression of the resonances is a thermometer of the temperature reached in the collisions

T/TC

J/(1S)

c(1P)

’(2S)

Physics motivation: AA collisions

understand the J/ behaviour in the cold nuclear medium (CNM) complicate issue, because of many competing mechanisms:

pp collisions

provide information on production models (CSM, NRQCD, CEM…)

reference for the study of charmonia dissociation in a hot medium approach followed at SPS and also at RHIC (with dAu data)

pA collisions

provide a reference for nuclear collisions results

Final state: cc dissociation in the medium, final energy loss

p

μ

μJ/

Initial state: shadowing, parton energy loss, intrinsic charm

(not covered by this talk)

Physics motivation: pp,pA collisions

Why CNM effects are so relevant ?• The cold nuclear matter effects present in pA collisions are of course present also in AA and can mask genuine QGP effects• Most of them (in particular final state interaction) scale with L, the mean thickness of nuclear matter crossed by the J/

L

J//N

coll

L

J//N

coll/

nu

cl.

Ab

s.

1

Anomalous suppression!

pA

AA

• It is very important to measure cold nuclear matter effects before any claim of an “anomalous” suppression in AA collisions

• Final state break-up is very important (expected to scale with √sN )....

But....... there are many nuclear effect at play

• Initial state

• Shadowing

• Various parameterizations (EKS98, EPS08,EPS09, nDS, HKN,...) with significant uncertainties

• Parton energy loss

• Shifts back the x1 of the incoming parton

• Model-dependent • Parameters can be tuned e.g. on Drell-Yan data q ~ 0.1 GeV2/fm

F. Arleo, JHEP 0211 (2002) 044

c=0.5qL2

• Enhancement at SPS energy• Depletion at LHC energy

Reduces the effective √s of the interaction producing the cc pair

Fixed target experiments

AA collisions

NA38 S-U 200 GeV/nucleon, 0<y<1 (M.C. Abreu et al., PLB449(1999)128)

NA50 Pb-Pb 158 GeV/nucleon, 0<y<, pT<5 GeV (B. Alessandro et al., EPJC39 (2005)335)

NA60 In-In 158 GeV/nucleon, 0<y<1, pT<5 GeV (R. Arnaldi et al., PRL99(2007) 132302, Nucl. Phys. A 830 (2009) 345)

pA collisions

HERAB p-Cu (Ti) 920 GeV,-0.34<xF<0.14,pT<5 GeV

(I. Abt et al., arXiv:0812.0734) E866 p-Be,Fe,W 800 GeV,-0.10<xF<0.93,pT<4 GeV

(M. Leitch et al., PRL84(2000) 3256) NA50 p-Be,Al,Cu,Ag,W,Pb,400/450 GeV,-0.1<xF<0.1,pT<5 GeV (B. Alessandro et al., EPJC48(2006) 329)

NA3 p-p p-Pt, 200 GeV, 0<xF<0.6, pT<5 GeV (J. Badier et al., ZPC20 (1983) 101)

NA60 p-Be,Al,Cu,In,W,Pb,U 158/400 GeV,-0.1<xF<0.35,pT<3 GeV (E. Scomparin et al., Nucl. Phys. A 830 (2009) 239)

(Relatively) large amount of fixed-target data (SPS, FNAL, HERA)

Data sets from fixed target experiments

Anomalous J/ suppression in AA is evaluated wrt to a reference obtained extrapolating, from pA to AA, the CNM effects affecting the J/

In-InPb-Pb

absJ/ = 4.2±0.5 mb,

(J//DY)pp =57.5±0.8 (Glauber analysis)

Observed suppression in AA exceeds nuclear absorption

AA collisions

• extrapolated to AA assuming

•obtained from pA at 400/450 GeV (NA50)

In the NA50 approach:all initial/final CNM effects are described through an effective abs. cross section abs

J/

absJ/ (158 GeV) = abs

J/ (400/450 GeV)

(J//DY)pp rescaled from 450/400 to 158 GeV

• Onset of the suppression at Npart 80• Good overlap between Pb and In

pA collisions

~e−ρLσabs

(R. Arnaldi et al., PRL99(2007) 132302)

Fixed target results (before 2009)

I. Abt et al., arXiv:0812.0734

ApppA

• E866 vs HERAB (similar √s) agreement in the common xF range

• E866/HERAB vs NA50

These effects are quantified, in pA collisions, in two ways:

decreases when decreasing √s

Satisfactory theoretical description still unavailable!

Strong xF dependence of

(R. Vogt, Phys. Rev. C61(2000)035203, K.G.Boreskov A.B.Kaidalov JETP Lett. D77(2003)599)

Because of the dependence on xF and energy the reference for the AA suppression must be obtained under the same kinematic/energy domain as the AA data

To understand the J/ dissociation in the hot matter created in AA collisions, cold nuclear matter effects have to be under control

absLpppA Ae ~

pA collisions: new HERA-B data

NA60 has collected pA data (using 7 different targets):

158 GeV: no data available up to now. First pA data at the same energy as AA collisions400 GeV: already investigated by NA50 (cross check)

A-dependence of the relative cross sections is fitted using the Glauber model and abs is extracted

shadowing neglected, as usual (but not correct!) at fixed target

abs J/ (158 GeV) = 7.6 ± 0.7 ± 0.6 mbabs J/ (400 GeV) = 4.3 ± 0.8 ± 0.6 mb

Using

(158 GeV) = 0.882 ± 0.009 ± 0.008 (400 GeV) = 0.927 ± 0.013 ± 0.009

ApppA

Very good agreement with the NA50 value

E. Scomparin et al., Nucl. Phys. A 830 (2009) 227

New pA data from NA60

NA60 pA results can be compared with values from other experiments

In the region close to xF=0, increase of with √s

NA60 158 GeV: smaller , hints of a decrease

towards high xF ?

NA60 400 GeV very good agreement with

NA50

Systematic error on for the new NA60 points ~0.01

Comparison between experiments: vs xF

pattern vs x1 at lower energies resembles HERA-B+E866 but systematically lower

shadowing effects and nuclear absorption scale with x2 (V. Tram and F. Arleo, arXiv:0612043) clearly other effects are present

yT esmx /1

yT esmx /2

Comparison between experiments: vs x1,2

2

21~

x

xms JNJ

need to disentangle the different contributions

Size of shadowing-related effects may be large and should be taken into account when comparing results at different energies

• anti-shadowing (with large uncertainties on gluon densities!)• final state absorption…

158 GeV free proton pdf158 GeV free proton pdfEKS98

Interpretation of results not easy many competing effects affect J/ production/propagation in nuclei

with antishadowing (EKS) = 9.3± 0.7± 0.7 mbwithout antishadowing: 7.6± 0.7± 0.6 mb

abs J/ (158 GeV)

Significantly higher than the “effective” value

C. Lourenco et al., arXiv:09013054

Kinematic dependence of nuclear effects

Apart from shadowing, other effects not very well known, as parton energy loss, intrinsic charm may complicate the picture even more

First attempts of a systematic study recently appeared (C. Lourenco, R. Vogt and H.Woehri, JHEP 0902:014,2009, INT Seattle workshop 2009, F. Arleo and Vi-Nham Tram Eur.Phys.J.C55:449-461,2008, arXiv:0907.0043 )

No coherent picture from the data no obvious scaling of or abs with any kinematical variable

Clear tendency towardsstronger absorption at low √s

Kinematic dependence of nuclear effects(2)

• Cold nuclear matter effects on J/ in AA collisions can be determined by means of an extrapolation of pA results

abs shows an energy/kinematical

dependencereference now obtained from 158 GeV pA data (same energy/kinematical range as the AA data, contrarily to what was done in the past)

B. Alessandro et al., EPJC39 (2005) 335R. Arnaldi et al., Nucl. Phys. A (2009) 345

AA collisions shadowing affects not only the target, but also the projectile

proj. and target antishadowing taken into account in the reference determination

R. Arnaldi, P. Cortese, E. Scomparin Phys. Rev. C 81 (2010), 014903

Using the new reference:

• Central Pb-Pb: still anomalously suppressed• In-In: almost no anomalous suppression?

In-In 158 GeV (NA60)Pb-Pb 158 GeV (NA50)

What about anomalous suppression ?

Collider experiments: RHIC

Experiments

PHENIX J/e+e- |y|<0.35 & J/+- |y| [1.2,2.2] STAR J/e+e- |y|<1

pp, dA collisions

pp 200 GeV/nucleon PHENIX, PRL 98, 232002 (2007) STAR, Phys. Rev. C 101 041902 (2009) dAu 200 GeV/nucleon PHENIX, Phys.Rev.C 77 024912 (2008) Nucl.Phys.A 830 (2009) 227

All data have been collected at the same collision energy (√s = 200 GeV) and (for each experiment) in the same kinematic domain

AA collisions

Au-Au 200 GeV/nucleon PHENIX, PRL 98 232301 (2007) Nucl.Phys.A 830 (2009) 331Cu-Cu 200 GeV/nucleon PHENIX, PRL 101 122301 (2008) STAR, Phys. Rev. C 101 041902 (2009)

Data sets from RHIC

RHIC J/ results are usually provided as in terms of nuclear modification factor

The pp reference, used up to now, is based on Run 5 improvement expected from new Run 6 high statistics data

pp results essential to • understand the J/ production mechanism• provide a reference for AA collisions (RAA)

arXiv:0904.0439

ppJcoll

AAJ

AA dNN

dNR

C.L. da Silva, Nucl. Phys. A 830 (2009) 227

pp results

Similar Npart dependence of RAA for CuCu and AuAu

PRL 101, 122301 (2008)

J/ suppression is stronger at forward rapidity wrt. to midrapidity

Phys. Rev. Lett 98, 232301 (2007)

AuAu

AA results

How can we intepret the RAA results ?

Several theoretical models have been proposed in the past, starting from the following observations

• RAA at forward y is smaller than at midrapidity• RAA at RHIC and SPS are similar, in spite of the very different √s

Different approaches proposed:

SPS RHIC LHC

s (GeV) 17.2 200 5500

Ncc ≈ 0.2 ≈10 ≈100-200

1) Only J/ from ’ and c decays are suppressed at SPS and RHIC

The 2 effects may balance: suppression similar to SPS

2) Also direct J/ are suppressed at RHIC but cc multiplicity high

J/ regeneration ( Ncc2) contributes to the J/ yield

same suppression at SPS and RHIC results do not show evidence for the sequential suppression

Interpretation of the results

• Models including J/ regeneration from heavy quark recombination qualitatively describe the RAA data

In particular the J/ should inherit the positive heavy quark elliptic flow

• Indirect way kinematic distributions and

elliptic flow should be affected by regeneration

• A direct way for quantitative estimate goes through cc cross section

No accurate measurement available

Recombination

X. Zhao, R. Rapp arXiv:0810.4566, Z.Qu et al. Nucl. Phys. A 830 (2009) 335

(and in particular the larger suppression observed at forward rapidity)

J/ production by statistical hadronization of charm quarks (Andronic, BraunMunzinger, Redlich and Stachel, PLB 659 (2008) 149)

• charm quarks produced in primary hard collisions• survive and thermalize in QGP • charmed hadrons formed at chemical freeze-out (statistical laws)• no J/ survival in QGP

yA. Andronic et al. arXiv:0805.4781

Agreement between data and model

Recombination should be tested on LHC data!

Statistical hadronization

BackwardMid Forward

Similarly to SPS, CNM effects are obtained from dAu data

RdAu is fitted with a theoretical calculation assuming

y

Phys. Rev. C 77, 024912 (2008)

RHIC data explore different x2 regions corresponding to shadowing (forward and midrapidity) anti-shadowing (backward rapidity)

• nuclear modifications of the PDFs• breakup as a free parameter

The result is then extrapolated to AA

results from dAu Run 3 do not allow to draw conclusions on AA results, because of the large error on breakup

dAu, first estimates of CNM effects

High statistics dAu data (Run8 ~ 30x Run3) are now available

a single value of break-up cannot reproduce the RCP ratios

Fit RCP separately for each rapidity bin, look for the y-dependence of the break-up cross section

EKS98: 0,1,…4,…mb

The Run8 dAu data

(T. Frawley ECT*,INT quarkonium,Joint Cathie-TECHQM workshop)

Peripheral ------------------------------------------------------------ Central

trend at high y is similar to the one observed by E866

suppression beyond CNM effects is found to be similar at y=0 and at y=1.7

Extrapolate to AA and comparewith data

(T. Frawley Joint Cathie-TECHQM workshop)

midrapidity

backward y forward y

breakup shows a strong rapidity dependence

RAA/RAA (CNM)

Is the highest suppression at forward rapidity a CNM effect ?

SPS results on anomalous suppression can be compared with RHIC RAA results normalized to RAA(CNM)

For central collisions more important suppression in Au-Au (RHIC) with respect to Pb-Pb (SPS)

still some model dependence also in this approach: Cu results are fitted using dAu, since dCu data do not exist

Comparison with SPS vs Npart

Effect related to thehigher energy density reached at RHIC ?

Results can be shown as a function of the multiplicity of charged particles (~energy density, assuming SPS~RHIC)

Comparison can also be done in terms of * Bjorken energy density

energy density evaluation is based on several assumptions

A

dydET

0

dET/d from WA98 data for SPS data no dET/d for CuCu, so AuAu data at the same NPart are used

comparing results from different experiments is not easy, significant systematic errors

Comparison with SPS

nice scaling btw SPS and RHIC!

Perspectives for the LHC

High charm quark multiplicity (NCC~100)

J/ regeneration (not yet firmly established at RHIC) might become dominant

New scenarios will open up, thanks to the high beam energy

Pb ion beams (√s=5.5 TeV)p-p collisions will be also studied (√s=7 – 14 TeV)

Quarkonium physics at the LHC

Factor 10 (100) increase in charmonium (bottomonium) cross section with respect to RHIC

Bottomonium physics will be accessible

Charmonium measurements will be carried out by all the LHC experiments, in different kinematical regions

LHCbCMSATLASALICE

Some features relative to J/ measurement in central PbPb collisions

Acc

(M)

S/BpT

ALICE(+-) ALICE(e+e-) ATLAS(+-) CMS(+-)

2.5<<4 -0.9<<0.9 -2.7<<2.7 -2.4<<2.4

70 MeV 30 MeV

0.13 (7)

>0 GeV/cindirect id.

1.2 (5)

>0 GeV/c

yes yes? yes?

>2 GeV/c >2 GeV/c

35 MeV

1.2

70 MeV

0.15

prompt/displ.

Measurements at the LHC

(LHCb plans still not finalized)

ALICE is the LHC experiment dedicated to nucleus-nucleus collisions

Central Barrel:-0.9<<0.9e+e- decay channel

Forward Muon Arm2.5<<4+- decay channel

Quarkonium production will be measured in both the central barrel and in the forward muon spectrometer in p-p and Pb-Pb collisions

In proton-proton collisions:

Measurement of differential distributions (y,pT) and polarization to constrain production models to provide a reference for AA

ALICE

Central rapidity Forward rapidity

• e- identification in TPC+TRD• integrated J/ acceptance ~29%

• identified in a Muon Spectrometer• integrated J/ acceptance ~35%

J/ (*)

N. 200 103 103

M MeV/c2 30 80

S/B 1.2 1.1

S/√(S+B) 245 21

J/ (2S)

N. 130 103 3.7 103 1.3 103

M MeV/c2 70 70 100

S/B 0.2 0.01 1.7

S/√(S+B) 150 7 29(*) requires Level-1 trigger on e-

significance still rather high smaller statistics compensated by background reductionWorst situation for the ’ statistics , but much larger background

Quarkonium in central Pb-Pb collisions (106 s running time, L=51026cm-2 s-1)

Quarkonium in ALICE (central PbPb)

Simulations with dNch/dy~3000 Simulations with dNch/dy~8000

With the expected 1–year statistics: J/ suppression can be studied as a function of centrality and pT (up to ~10 GeV/c) J/ polarization study will be performed as a function of pT

A fraction of the J/ produced at LHC comes from B-hadron decays useful to evaluate the beauty production cross section need to be disentangled to study prompt J/ production

At midrapidity prompt and secondary J/ can be discriminated thanks to the vertexing capabilitiesAt forward y J/ from B can be determined only indirectly

Higher charmonia states (’, c) can be measured cleaner signal for theory feasible in pp, more complicate in Pb-Pb (higher background, smaller significance)

Charmonium in Pb-Pb: physics studies

First dimuons have been seen in ALICE in pp collisions at √s=900GeV, even if … not yet a J/!

First dimuons in ALICE!

• J/ suppression is a good observable for QGP studies

J/ behaviour in cold nuclear matter is already a complicate issue: many competing initial/final state effects

Many steps forward thanks to new high precision data

but for a correct evaluation of anomalous effects, cold nuclear matter effects have to be under control

Important to understand J/ behaviour from lower to higher energy in a coherent scenario

• New data at LHC energy will soon be available!

They will help to discriminate among the different processes (suppression, regeneration…) affecting the J/

• In the future, the “J/ picture” will be further sharpened by the results from CBM, exploring high baryon-density matter, and (hopefully) also from an NA60-like experiment filling the gap between FAIR and top SPS energy

Conclusions

• An anomalous J/ suppression has been observed at both SPS and RHIC

Thanks !!!

Furthermore CNM effects may depend on the assumed J/ production mechanisms (E. Ferreiro et al. arXiv:0809.4684)

intrinsic (gg J/) extrinsic (gg J/ + g) (emission of a hard gluon)

J/ produced through different partonic processes involve gluons in different x2 region different shadowing corrections

Extrinsic vs intrinsic production

RCuCu up to pT = 9 GeV/c suppression looks roughly constant up to high pT

PHENIX (minimum bias) STAR (centrality 0-20% & 0-60%)

RCuCu =1.4±0.4±0.2 (pT>5GeV/c) RAA increases from low to high pT

Difference between high pT results, but strong conclusions limited by poor statistics

Both results in contradiction with AdS/CFT+Hydro

Increase at high pT already seen at SPS

NA50: Pb-Pb

High-pT J/ in Cu-Cu