Neutral meson production measurements with the ALICE at the LHC

Identified hadron spectra are considered to be sensitive to the transport properties of strongly interacting matter produced in high-energy nucleus-nucleus collisions. π0 and η mesons in ALICE are identified via their two-photon decays by using calorimeters and the central tracking system. In the latter, photons are measured via their conversion to electron-positron pairs in the material of the inner ALICE barrel tracking detectors. The measured production spectra in pp, p−Pb and Pb−Pb collisions at mid−rapidity and over a wide pT range will be presented in the available Large Hadron Collider (LHC) energies of Run I. The resulting nuclear modification factor RAA at different centrality classes shows a clear pattern of strong suppression in the hot QCD medium with respect to pp collisions. Comparison of the ALICE results on neutral mesons with lower-energy experiments is also discussed.


Introduction
Light mesons such as π 0 and η provide a tool to test perturbative QCD (pQCD) with pp collisions, but also to study the properties of the QGP, which is created in ultra-relativistic heavy-ion collisions. These neutral mesons in pp collisions are produced from the fragmentation of partons created in hard processes during the early stages of the collision. Thus, the study of hadron production can provide constraints for the parton fragmentation functions (FF) and the parton distribution functions (PDF).
In heavy-ion collisions, the dense environment allows the observation of the modification of the FF or PDF via linear and non-linear recombination effects [1]. The Quark-Gluon Plasma formed in central heavy-ion collisions causes collisional energy loss and radiative parton energy loss induced by multiple scattering, that lead to modifications of parton fragmentation. The first observable of such phenomena, the yield suppression of inclusive hadrons in heavy-ion collisions compared to pp collisions, is arguably the most thoroughly studied quantity and provides a tool to extract the properties of the hot medium created in nucleus-nucleus collisions by comparing theoretical calculations with experimental measurements.
Measured neutral mesons give different information depending on their p T range. At p T 3 − 5 GeV/c, they can give insight about collective effects via the modification of their p T spectra. At the LHC energies, neutral mesons come mostly from gluon fragmentation. Due to the different color factor, gluons suffer a larger energy loss in the medium than quarks. This effect together with the different relative contribution of quarks and gluons to the π 0 and η production, lead to differences in the suppression pattern of the two particles [2,3]. On the other hand, at p T 5 GeV/c hadron production results from the hadronization of partons created in the initial hard scattering. These partons lose energy in the dense medium via radiation and collisions and this energy loss can be quantified via the measured neutral mesons spectra at high p T .

Neutral meson reconstruction in the ALICE experiment
Neutral mesons are reconstructed in ALICE [4] by using four different detectors, the electromagnetic calorimeters PHOS [5] and EMCal [6] and the two tracking detectors, the ITS [7] (Inner Tracking System) and the TPC [8] (Time Projection Chamber).
In the calorimeters, the electron and photon reconstruction are made by measuring the electromagnetic showers produced by the incident particles. PHOS is a high granularity and high energy resolution calorimeter installed at a radial distance of 4.6 m from the interaction point and is composed of lead-tungstate crystals, PbWO 4 , of 2.2 × 2.2 × 18 cm 3 dimensions and 0.89 cm radiation length. The pseudorapidity range covered is 0.26 units and the azimuthal coverage is 60 • . The energy resolution is parametrized by the equation: The Electromagnetic Calorimeter (EMCal) is a Shashlik type Pb-scintillator sampling calorimeter built from alternating lead and scintillator segments of 1.44 mm and 1.76 mm thickness, respectively, together with longitudinal wavelength-shifting fibres for light collection. The cell (tower) size of the EMCal is approximately 0.014 × 0.014 rad in ∆ϕ × ∆η (6 cm) with a depth corresponding to 20.1 X 0 . It is installed at a radial distance of 4.28 m from the interaction point and covers a pseudorapidity range of 1.4 units with 100 • azimuthal coverage. The energy resolution of EMCal is described by the equation: The two tracking devices are used to reconstruct the tracks and the particle identification (to identify electrons and positrons) is made by measuring the energy loss of the tracks in those detectors. After identification, electrons and positrons are paired to reconstruct conversion photons. This is the Photon Conversion Method (PCM). The conversion point can be reconstructed with a z and ϕ resolution of 1.5 cm and 7 mrad, respectively. The photon conversion probability is rather small, ∼ 8.5%, but it is compensated by the wider acceptance of TPC: full azimuthal coverage and a pseudorapidity range of 1.8 units. The neutral mesons are reconstructed in the di-photon channel by using the invariant mass technique.

ALICE performance in the reconstruction of neutral mesons in Run I and
Run II   4 π 0 and η spectra in pp collisions ALICE has measured the p T invariant differential production cross section of neutral mesons in pp collisions at all available centre-of-mass energies of Run I, 0.9, 2.76, 7 [9] (PHOS and PCM combined measurements) and 8 TeV (PHOS only preliminary result, [11]). In Fig. 4 (left panel), these π 0 spectra are presented fitted by a Tsallis function and compared to PYTHIA8 generator and to the next-toleading order pQCD (NLO pQCD) calculations for two renormalization and factorization scales and with the MSTW PDF and the DSS14 FF [12]. The π 0 spectra are better described by these calculations compared to previous ones [9], however, an increased discrepancy is observed between the data and the calculations with increasing p T and centre-of-mass energy. In Fig. 4 (right panel), the η spectra in pp collisions at √ s = 0.9, 2.76 and 7 TeV are shown. The spectra at √ s = 0.9 and 2.76 TeV are obtained from PCM only while the √ s = 7 TeV η spectrum is the combined result from PCM and PHOS. The spectra are compared to the NLO pQCD calculations with CTEQ6M5 PDF and the AES fragmentation function [9,13]. This comparison reveals a bigger discrepancy compared to the π 0 case. The η/π 0 ratio is presented in Fig. 5 for √ s = 2.76 and 7 TeV, respectively, and compared to the NLO pQCD calculations using CTEQ6M5 PDF for both particles and DSS FF for the π 0 and AES FF for the η [9]. Although this choice for the PDF and fragmentation functions does not reproduce the individual spectra, in the ratio their influence seems to be reduced. The ratio reaches above p T = 4 GeV/c a constant value around four, compatible with the available world measurements.  Fig. 6 [10]. This is the combined measurement from PCM and PHOS. The modification of the hadron yields for different p T intervals in heavy-ion (A-A) collisions with respect to pp collisions can be quantified with the nuclear modification factor:  where the nuclear overlap function < T AA > is related to the average number of inelastic nucleonnucleon collisions as < T AA > = < N coll >/σ pp inel . The combined R AA from PCM and PHOS is shown in Fig. 7 (left panel), for three centrality classes: central (0-5%), semi-central (20-40%) and peripheral (60-80%). In all centrality classes the measured R AA exhibits a maximum around p T ≈ 1 − 2 GeV/c. At high p T , the observed suppression is attributed to an increasing energy loss in the medium with centrality. Agreement of R AA between the neutral pions and the charged pions, kaons and protons is observed at high p T [14].
The energy dependence of the R AA has also been studied by comparing the ALICE results with those from previous [10] experiments (Fig. 7 right panel), WA98 at the SPS measurements in central   Pb−Pb events and PHENIX measurements in central Au−Au events at two energies. The decrease of the R AA observed at the LHC energies compared to the lower centre-of-mass energies indicates that the higher initial energy densities created at larger √ s NN have a stronger influence in the R AA compared to the harder initial parton p T spectra. Considering the data for all shown energies one observes that the value of p T with the maximum R AA value appears to shift towards lower p T with increasing √ s NN .
The analysis of the 2011 Pb−Pb data collected by ALICE (10 times the integrated luminosity with respect to the results presented previously), allowed the extension of the p T reach for π 0 up to 20 GeV/c, and the measurement of η. In Fig. 8, the combined measurement from EMCal and PCM for two centrality classes is given for π 0 and η (left and right panel, respectively). The differential invariant yields are compared to the NLO pQCD predictions (with MSTW PDF and DSS14 (π 0 ), DSS07 (η) FF) scaled by N coll , confirming the suppression of the neutral mesons yields in the Pb−Pb data.
In Fig. 9, the η/π 0 ratio is compared to the corresponding measurement in pp collisions at √ s = 7 TeV (left panel), to the pQCD predictions (middle panel, [2]) and to the ratio K ± /π ± at the same centre-of-mass energy (right panel, [14]). A similar magnitude for the ratio is observed in both pp and Pb−Pb collisions, consistent with pQCD within the current experimental uncertainties [11]. The agreement of the η/π 0 ratio with the ratio K ± /π ± implies no visible effects due to the strange quark content of the η and K ± mesons. 6 π 0 and η spectra in p−Pb collisions at √ s NN = 5.02 TeV Parton energy loss depends on the transport properties of the medium and their space-time evolution, the initial parton energy, and the parton type [10]. The nuclear modification factor, R AA , is also affected by the slope of the initial parton transverse momentum spectrum prior to any interaction with the medium and initial-state effects like the modifications of the parton distributions in nuclei. An important constraint for modelling these effects comes from the study of p-A collisions [15]. ALICE has measured the neutral mesons differential invariant yields in p−Pb collisions at √ s NN = 5.02 TeV presented in Fig. 10 (left panel), fitted by a Tsallis function. The π 0 spectrum is the combined measurement from PCM, Dalitz, PHOS and EMCal, and the η spectrum from PCM and EMCal. The η/π 0 ratio is calculated and compared to the result from pp collisions at √ s = 7 TeV in Fig. 10 (right, up panel). As in the case of Pb−Pb data, the η/π 0 ratio in p−Pb collisions agrees with that obtained from pp collisions within the systematic and statistical uncertainties. The calculated π 0 R p−Pb is consistent with unity above p T 2.5 GeV/c as shown in Fig. 10 (right, bottom panel). This measurement indicates that the strong suppression of hadron production at high p T observed in Pb−Pb  collisions is not due to an initial-state effect, but is the fingerprint of jet quenching in hot QCD matter. The experimental result is also consistent with model calculations such as EPS calculations assuming three different fragmentation functions [16] and the Color Glass Condensate model [1].

Conclusions
The π 0 and η invariant cross sections are measured in pp, Pb−Pb and p−Pb collisions over a wide p T range by using different and complementary detection methods. The pp results are compatible with NLO pQCD calculations at the lower p T and collision energies, however, there is a growing discrepancy with increasing p T and √ s. In Pb−Pb collisions, the π 0 invariant yields in different centrality classes and the measured R AA show an increasing suppression with increasing centrality and centre-of-mass energy. The shape of R AA is comparable between RHIC and LHC energies. However, at the LHC energies there is a stronger suppression observed resulting from the higher initial energy densities.
ALICE has also measured the π 0 and η invariant cross sections in p−Pb collisions to disentangle initial state effects from parton energy loss in the medium. Indeed, the R p−Pb is found to be consistent with unity above p T 2.5 GeV/c, demonstrating that the strong suppression observed in central Pb−Pb collisions at the LHC is not due to an initial-state effect.
The η/π 0 ratio in the three systems, pp, Pb−Pb and p−Pb is consistent within the current statistical and systematic uncertainties and reaches a constant value for p T > 4 GeV/c.