J/$\psi$ suppression at forward rapidity in Pb-Pb collisions at $\sqrt{s_{\rm NN}}$ = 2.76 TeV

The ALICE experiment has measured the inclusive J/$\psi$ production in Pb-Pb collisions at $\sqrt{s_{\mathrm{NN}} } = 2.76$ TeV down to zero transverse momentum in the rapidity range $2.5<y<4$. A suppression of the inclusive J/$\psi$ yield in Pb-Pb is observed with respect to the one measured in pp collisions scaled by the number of binary nucleon-nucleon collisions. The nuclear modification factor, integrated over the 0-80% most central collisions, is $0.545 \pm 0.032 \rm{(stat.)} \pm 0.083 \rm{(syst.)}$ and does not exhibit a significant dependence on the collision centrality. These features appear significantly different from measurements at lower collision energies. Models including J/$\psi$ production from charm quarks in a deconfined partonic phase can describe our data.

C, providing a high triggering efficiency for hadronic interactions. The beam induced background was further reduced by timing cuts on the signals from the VZERO and from the zero degree calorimeters (ZDC). The contribution from electromagnetic processes was removed with a cut on the energy deposited in the neutron ZDCs. The centrality determination is based on a fit of the VZERO amplitude distribution as described in [14]. A cut corresponding to the most central 80% of the nuclear cross section was applied; for these events the MB trigger is fully efficient. A data sample of 17.7 × 10 6 Pb-Pb collisions collected in 2010 satisfying all the above conditions is used in the following analysis. It corresponds to an integrated luminosity L int ≈ 2.9 µb −1 . This data sample was further divided into five centrality classes from 0%-10% (central collisions) to 50%-80% (peripheral collisions). J/ψ candidates are formed by combining pairs of opposite-sign (OS) tracks reconstructed in the geometrical acceptance of the muon spectrometer. To reduce the combinatorial background, the reconstructed tracks in the muon tracking chambers are required to match a track segment in the muon trigger system. The resulting invariant mass distribution of OS muon pairs for the 0%-80% most central Pb-Pb collisions is shown in Fig. 1, where a J/ψ signal above the combinatorial background is clearly visible. The J/ψ raw yield was extracted by using two different methods. The OS dimuon invariant mass distribution was fitted with a Crystal Ball (CB) function to reproduce the J/ψ line shape, and a sum of two exponentials to describe the underlying continuum. The CB function connects a Gaussian core with a power-law tail [15] at low mass to account for energy loss fluctuations and radiative decays. At high transverse momenta (p t ≥ 3 GeV/c), the sum of two exponentials does not describe correctly the underlying continuum; it was replaced by a third order polynomial. Alternatively, the combinatorial background was subtracted using an event-mixing technique. The resulting mass distribution was fitted with a CB function and an exponential or a first order polynomial to describe the remaining background. The event mixing was preferred to the like-sign subtraction technique since it is less sensitive to correlated signal pairs present in the like-sign spectra and gives better statistical precision. The ψ(2S) was not included in the signal line shape since its contribution is negligible. The width of the J/ψ mass peak depends on the resolution of the spectrometer which can be affected by the detector occupancy that increases with centrality. This effect, evaluated by embedding simulated J/ψ → µ + µ − decays into real events, was found to be less than 2%. This conclusion was confirmed by a direct measurement of the tracking chamber resolution versus centrality using reconstructed tracks. Therefore, the same CB line shape can be used for all centrality classes. The parameters of the CB tails were fixed to the values obtained either in simulations or in pp collisions where the signal to background ratio is much higher. For each of these choices, the mean and width of the CB Gaussian part were fixed to the value obtained by fitting the mass distribution in the centrality range 0%-80%. In addition, a variation of the width of ± 1 standard deviation was applied to account for uncertainties (varying the mean has turned out to have a negligible effect in comparison). The raw J/ψ yield in each centrality class was determined as the average of the results obtained with the two fitting approaches and the various CB parametrizations, while the corresponding systematic uncertainties were defined as the RMS of these results. It was also checked that every individual result differs from the mean value by less than three RMS. The raw J/ψ yield in our Pb-Pb sample is 2350 ± 139(stat.) ± 189(syst.). The invariant mass resolution is around 78 MeV/c 2 , in very good agreement with the embedded J/ψ simulations. The signal to background ratio integrated over ± 3 σ of the mass resolution varies from 0.1 for central collisions to 1.5 for peripheral collisions.
The measured number of J/ψ (N i J/ψ ) was normalized to the number of events in the corresponding centrality class (N i events ) and further corrected for the branching ratio (BR) of the dimuon decay channel, the acceptance A and the efficiency ε i of the detector. The inclusive J/ψ yield in each centrality class i for our measured p t and y ranges (∆p t , ∆y) is then given by: The product Aε was determined from Monte Carlo simulations. The generated J/ψ p t and y distributions were extrapolated from existing measurements [16], including shadowing effects from EKS98 calculations [17]. As the measured J/ψ polarization in pp collisions at √ s = 7 TeV is compatible with zero [18], and J/ψ mesons produced from charm quarks in the medium are expected to be unpolarized, we presume J/ψ production is unpolarized. For the tracking chambers, the time-dependent status of each electronic channel during the data taking period was taken into account as well as the residual misalignment of the detection elements. The efficiencies of the muon trigger chambers were determined from data and were then applied in the simulations [19]. Finally, the dependence of the efficiency on the detector occupancy was included using the embedding technique. For J/ψ mesons emitted at 2.5 < y < 4 and p t ≥ 0, a runaveraged value of Aε = 0.176 with a 8% relative systematic uncertainty was obtained. A 8% ± 2%(syst.) relative decrease of the efficiency was observed when going from peripheral to central collisions.
The J/ψ yield measured in Pb-Pb collisions in centrality class i is combined with the inclusive J/ψ cross section measured in pp collisions at the same energy to form the nuclear modification factor R AA defined as: The inclusive J/ψ cross section in pp collisions σ pp J/ψ (∆p t , ∆y) was measured using the same apparatus and analysis technique within the corresponding p t and y range [12]. The reference value σ pp J/ψ used for the calculation of R AA integrated over p t and y is 3.34±0.13(stat.) ±0.24(syst.) ±0.12(lumi.) +0.53 −1.07 (pol.) µb. The centrality intervals used in this analysis, the average number of participating nucleons N part and Table 1: The average number of participating nucleons N part without and with N coll weighting, the mid-rapidity charged-particle density dN w ch /dη| η=0 with N coll weighting and the average value of the nuclear overlap function T AA for the centrality classes expressed in percentages of the nuclear cross section [14]. average value of the nuclear overlap function T AA derived from a Glauber model calculation [14] are summarized in Table 1. Since our most peripheral bin is rather large, the variables N w part and the chargedparticle density measured at mid-rapidity dN w ch /dη| η=0 were weighted by the number of binary collisions N coll . Indeed in absence of nuclear matter effects, the J/ψ production cross section in nucleus-nucleus is expected to scale with N coll . The weighted values are given in Table 1 and are used for the ALICE data points in the following figures. All systematic uncertainties entering the R AA calculation are listed in Table 2. In the figures below, the point to point uncorrelated systematic uncertainties are represented as boxes at the position of the data points while the statistical ones are indicated by vertical bars. Correlated systematic uncertainties are quoted directly on the figures.
The inclusive J/ψ R AA measured by ALICE at √ s NN = 2.76 TeV in the range 2.5 < y < 4 and p t ≥ 0 is shown in Fig. 2 as a function of dN ch /dη| η=0 (left) and N part (right). The charged-particle density closely relates to the energy density of the created medium whereas the number of participants reflects the collision geometry. The centrality integrated J/ψ R AA is R 0%−80% AA = 0.545 ± 0.032(stat.) ± 0.083(syst.), indicating a clear J/ψ suppression. The contribution from beauty hadron feed-down to the inclusive J/ψ yield in our y and p t domain was measured by the LHCb collaboration to be about 10% in pp collisions at √ s = 7 TeV [21]. Therefore, the difference between the prompt J/ψ R AA and our inclusive measurement is expected not to exceed 11% if N coll scaling of beauty production is assumed and shadowing effects are neglected. All R AA results are presented assuming unpolarized J/ψ production in pp and Pb-Pb collisions. The comparison with the PHENIX measurements 2 at √ s NN = 200 GeV at forward rapidity 1.2 < |y| < 2.2 [5,20] shows that our inclusive J/ψ R AA is almost a factor of three larger for dN ch /dη| η=0 600 (N part 180). In addition, our results do not exhibit a significant centrality dependence.   Table 1.  data. The two models [22,23] predict the R AA due only to shadowing effects for nDSg (shaded areas) and EPS09 (lines) nPDF respectively.
The rapidity dependence of the J/ψ R AA is presented in Fig. 3 for two p t domains, p t ≥ 0 and p t ≥ 3 GeV/c. The J/ψ reference cross sections in pp collisions 3 and the R AA total systematic uncertainties, indicated as open boxes in the figure, were evaluated in the same kinematic range. Our results are shown together with a measurement from CMS [11] of the inclusive J/ψ R AA in the rapidity range 1.6 < |y| < 2.4 with p t ≥ 3 GeV/c. No significant rapidity dependence can be seen in the J/ψ R AA for p t ≥ 0. For p t ≥ 3 GeV/c, a decrease of R AA is observed with increasing rapidity reaching a value of 0.289 ± 0.061(stat.) ± 0.078(syst.) for 3.25 < y < 4. At LHC energies, J/ψ nuclear absorption is likely to be negligible and the modification of the gluon distribution function is dominated by shadowing effects [24]. An estimate of shadowing effects is shown in Fig. 3 within the Color Singlet Model at Leading Order [22] and the Color Evaporation Model at Next to Leading Order [23]. The shadowing is respectively calculated with the nDSg and the EPS09 parametrizations [23] of the nuclear Parton Distribution Function (nPDF). For nDSg (EPS09) the upper and lower limits correspond to the uncertainty in the factorization scale (uncertainty of the nPDF). The effect of shadowing shows no dependence with rapidity and its overall amount is reduced by the addition of a transverse momentum cut. At most, shadowing effects are expected to lower the R AA from 1 to 0.7. Recent Color Glass Condensate (CGC) calculations for LHC energies may indicate a larger initial state suppression (R AA ≈ 0.5) [25]. However, any J/ψ suppression due to initial state effects, CGC or shadowing, will be stronger at lower p t contrary to the 3 We report here σ pp J/ψ (p t ≥ 3GeV/c, 2.5 < y ≤ 3.25) = 0.34 ± 0.03(stat.) ± 0.03(syst.) ± 0.02(lumi.) µb and σ pp J/ψ (p t ≥ 3GeV/c, 3.25 < y < 4) = 0.50 ± 0.04(stat.) ± 0.04(syst.) ± 0.02(lumi.) µb that can not directly be extracted from [12].  Table 1. data behavior.
In Fig. 4, our measurement is compared with theoretical models that include a J/ψ regeneration component from deconfined charm quarks in the medium. The Statistical Hadronization Model [6,26] assumes deconfinement and a thermal equilibration of the bulk of the cc pairs. Then charmonium production occurs only at phase boundary by statistical hadronization of charm quarks. The prediction is given for two values of dσ cc /dy in absence of a measurement for Pb-Pb collisions. The two transport model results [27,28] presented in the same figure differ mostly in the rate equation controlling the J/ψ dissociation and regeneration. Both are shown as a band which connects the results obtained with (lower limit) and without (higher limit) shadowing. The width of the band can be interpreted as the uncertainty of the prediction. In both transport models, the amount of regenerated J/ψ in the most central collisions contributes to about 50% of the production yield, the rest being from initial production.
In summary, we have presented the first measurement of inclusive J/ψ nuclear modification factor down to p t = 0 at forward rapidity in Pb-Pb collisions at √ s NN = 2.76 TeV. The J/ψ R AA is larger than the one measured at the SPS and at RHIC for most central collisions and does not exhibit a significant centrality dependence. Statistical hadronization and transport models which respectively feature a full and a partial J/ψ production from charm quarks in the QGP phase can describe the data. Towards a definitive conclusion about the role of J/ψ production from deconfined charm quarks in a partonic phase, the amount of shadowing needs to be measured precisely in pPb collisions. In this context, the measurement of open charm and J/ψ elliptic flow will also help to determine the degree of thermalization for charm quarks.

Acknowledgements
The ALICE collaboration would like to thank all its engineers and technicians for their invaluable contributions to the construction of the experiment and the CERN accelerator teams for the outstanding performance of the LHC complex.