We review a subset of experimental results from the heavy-ion collisions at the Large Hadron Collider (LHC) facility at CERN. Excellent consistency is observed across all the experiments at the LHC (at center of mass energy  TeV) for the measurements such as charged particle multiplicity density, azimuthal anisotropy coefficients, and nuclear modification factor of charged hadrons. Comparison to similar measurements from the Relativistic Heavy Ion Collider (RHIC) at lower energy (  GeV) suggests that the system formed at LHC has a higher energy density and larger system size and lives for a longer time. These measurements are compared to model calculations to obtain physical insights on the properties of matter created at the RHIC and LHC.

1. Introduction

The main goal of the high energy heavy-ion collisions is to study the phase structure of the quantum chromodynamic (QCD) phase diagram [13]. One of the most interesting aspects of these collisions is the possibility of forming a phase of deconfined quarks and gluons, a system that is believed to have existed in a few microseconds-old universe. First principle QCD calculations suggest that it is possible to have such a state of matter if the temperatures attained can be of the order of the QCD scale ( 200 MeV) [46]. In laboratory, such temperatures could be attained by colliding heavy ions at relativistic energies. Furthermore, in very high energy collisions of heavy ions at the LHC and RHIC, the lifetime of the deconfined phase may be long enough to allow for the detailed study of the fundamental constituents (quarks and gluons) of the visible matter.

The results from heavy-ion collisions at RHIC have clearly demonstrated the formation of a deconfined system of quarks and gluons in collisions at  GeV [711]. The produced system exhibits copious production of strange hadrons, shows substantial collectivity developed in the partonic phase, and exhibits suppression in high transverse momentum ( ) hadron production relative to collisions and small fluidity as reflected by a small value of viscosity to entropy density ratio ( ). A factor of 14 increase in for collisions at LHC is expected to unravel the temperature dependence of various observables and to extend the kinematic reach in rapidity and of previous measurements at RHIC. On the other hand, the beam energy scan program at RHIC is expected to provide additional details of the QCD phase diagram not accessible at the LHC [12].

In this review paper, we discuss a subset of results that have come out from LHC collisions at  TeV. We have divided the discussion into three sections. In the second section, we discuss the consistency of various measurements among the three LHC experiments that have heavy-ion programs: ALICE, ATLAS, and CMS. Section 2.1 discusses the results on the charged particle multiplicity. Section 2.2 discusses the results on azimuthal anisotropy, and Section 2.3 discusses the results on the nuclear modification factor.

In the third section, we make a comparative study between similar observables measured at lower energy collisions at RHIC and those from LHC. In doing this, we highlight the additional information that heavy-ion collisions at LHC bring compared to RHIC. In Section 3.1, we discuss the bulk properties at freeze-out that include results on multiplicity, average transverse mass and Bjorken energy density, volume and decoupling time, kinetic freeze-out temperature and average flow velocity, and fluctuations. Section 3.2 is devoted on the results to azimuthal anisotropy, where we discuss the energy dependence of integrated , dependence of various azimuthal anisotropy coefficients on , and flow fluctuations. In Section 3.3, we discuss results for nuclear modification factor.

In the fourth section, we present a comparison of various model calculations to the corresponding measurements at LHC. We concentrate mainly on the results for charged particle multiplicity density and ratio in Section 4.1, azimuthal anisotropy in Section 4.2, and nuclear modification factor in Section 4.3.

Finally, we summarize our observations in the last section of the paper.

2. Consistency of Results among LHC Experiments

2.1. Charged Particle Multiplicity

One of the first measurements to come out of the heavy-ion collision program at LHC is the charged particle multiplicity per unit pseudorapidity in collisions at  TeV. Figure 1 shows the centrality (reflected by the number of participating nucleons, , obtained from a Glauber model calculation [13]) dependence of at midrapidity for collisions at  TeV from ALICE [14], CMS [15], and ATLAS [16] experiments. The error bars reflect statistical uncertainties. The ATLAS measurements of are obtained over using a minimum bias trigger with a central solenoid magnet off data set. The charged particles are reconstructed using two different algorithms using the information from pixel detectors covering 2.0. The values are obtained by comparing the summed transverse energy in the forward calorimeter over a pseudorapidity range to a Glauber model simulation. The CMS results for are from the barrel section of the pixel tracker covering 2.5. The minimum bias trigger data set was in the magnetic field off configuration so as to improve the acceptance of low particles. The centrality determinations as in the case of ATLAS experiment are done using information from hadron forward calorimeter ( ) and Glauber model simulations. The ALICE measurement uses a minimum bias data set from the silicon pixel detector ( ). The centrality selection is carried out using signals from VZERO detectors (2 arrays of 32 scintillator tiles) covering the regions and , along with the corresponding Glauber modeling of the data.

In spite of the difference in operating conditions and measurement techniques, the versus results for collisions at = 2.76 TeV show a remarkable consistency across the three experiments. The results show that the charged particle multiplicity per unit pseudorapidity per nucleon pair increases from peripheral to central collisions. This gradual increase in per participating nucleon pair indicates that in central head-on collisions, where the number of participating nucleons is more, the charged particle production is different compared to that in peripheral collisions.

2.2. Azimuthal Anisotropy

Azimuthal anisotropy has been studied in great detail in heavy-ion collision experiments. It can provide information about initial stages of heavy-ion collisions. Figure 2 (top panels) shows the azimuthal anisotropy of produced charged particles ( ) as a function of for 30–40% collisions at  TeV from the three different experiments: ATLAS, ALICE, and CMS. Here, is the azimuthal angle of the produced particles, and is the order reaction plane angle measured in the experiments. The left panel in the figure corresponds to , the middle panel corresponds to , and the right panel corresponds to , respectively. Bottom panels show the ratio of the experimental data to a polynomial fit to the ALICE data.

In the CMS experiment [1720], the measurements use the information from the silicon tracker in the region with a track momentum resolution of 1% at = 100 GeV/ kept within a magnetic field of 3.8 Tesla. The event plane angle ( ) is obtained using the information on the energy deposited in the hadron forward calorimeter. A minimum gap of 3 units is kept between the particles used for obtaining and . This ensures suppression of nonflow correlations which could arise, for example, from dijets. The event plane resolution obtained using three subevents technique varies from 0.55 to 0.84, depending on the collision centrality. The ATLAS experiment [2124] measured using the inner detectors in the , kept inside a 2 Tesla field of superconducting solenoid magnet. The event planes are obtained using forward calorimeter information, with a resolution varying from 0.2 to 0.85, depending on collision centrality. The ALICE experiment [25] measured using charged tracks reconstructed from the Time Projection Chamber ( ); the event plane was obtained using information from VZERO detectors kept at a large rapidity gap from the TPC. The momentum resolution of the tracks is better than 5%.

A very nice agreement for , , and versus is found between all the experiments to a level of within 10% for most of the ranges presented. The results show an increase of , , and values with for the low and a decrease for above 3 GeV/ . The hydrodynamical evolution of the system affects most of the low particles and hence the increasing at low .

2.3. Nuclear Modification Factor

One of the established signatures of the QGP at top RHIC energy is the suppression of high transverse momentum ( ) particles in heavy-ion collisions compared to corresponding data from the binary collisions scaled collisions. It has been interpreted in terms of energy loss of partons in QGP. This phenomenon is referred to as the jet quenching in a dense partonic matter. The corresponding measurement is called the nuclear modification factor ( ).

Figure 3 shows the nuclear modification factor for inclusive charged hadrons measured at midrapidity in LHC experiments for collisions at = 2.76 TeV. The nuclear modification factor is defined as . Here, the overlap integral with being the number of binary collisions commonly estimated from Glauber model calculation and is the cross section of charged hadron production in collisions at = 2.76 TeV.

The ALICE experiment [26] uses the inner tracking system (ITS) and the time projection chamber (TPC) for vertex finding and tracking in a minimum bias data set. The CMS experiment [27] reconstructs charged particles based on hits in the silicon pixel and strip detectors. In order to extend the statistical reach of the spectra in the highly prescaled minimum bias data recorded in 2011, it uses unprescaled single-jet triggers. Both experiments take the value of  mb. The result shows that the charged particle production at high in LHC is suppressed in heavy-ion collisions relative to nucleon-nucleon collisions. The suppression value reaches to a minimum at 6-7 GeV/ and then gradually increases to attain an almost constant value at 40 GeV/ . This can be understood in terms of energy loss mechanism differences in intermediate and higher regions. The rise in the above 6-7 GeV/ may imply the dominance of the constant fractional energy loss which is the consequence of flattening of the unquenched nucleon-nucleon spectrum. An excellent agreement for versus for charged hadrons in 0–5% central collisions at  TeV is observed between the two experiments.

Having discussed the consistency of these first measurements in collisions among different experiments, the major detectors used, acceptances, and ways to determine centrality and event plane, we now discuss the comparison between measurements at RHIC and LHC heavy-ion collisions.

3. Comparison of LHC and RHIC Results

In the first subsection, we discuss the energy dependence of basic measurements made in heavy-ion collisions. These include , ( ; here, represents mass of hadron), Bjorken energy density ( ), life time of the hadronic phase ( ), system volume at the freeze-out, kinetic and chemical freeze-out conditions, and finally, the fluctuations in net-charge distributions. In the next subsection, we discuss the energy dependence of integrated , versus , and flow fluctuations at RHIC and LHC. In the final subsection, we compare the nuclear modification factor for hadrons produced in heavy-ion collisions at RHIC and LHC.

3.1. Bulk Properties at Freeze-Out
3.1.1. Multiplicity

Figure 4(a) shows the charged particle multiplicity density at midrapidity ( ) per participating nucleon pair produced in central heavy-ion collisions versus . We observe that the charged particle production increases by a factor 2 as the energy increases from RHIC to LHC. The energy dependence seems to rule out a logarithmic dependence of particle production with and supports a power law type of dependence on . The red solid curve seems to describe the full energy range. More detailed discussions on the energy dependence of these measurements can be found in [28].

Figure 4(b) shows the excess of / in collisions [15, 16, 2937] over corresponding yields in ( ) [3847] and collisions [29, 48, 49]. This observation also seen at RHIC persists at LHC but is proportionately larger at the higher energy collisions at the LHC. A power law fit to the collision charged particle multiplicity density leads to a dependence ~ , while those for collisions go as ~ . There is no scaling observed in the charged particle multiplicity density per participating nucleon, when compared between elementary collisions like and heavy-ion collisions. This is a clear indication that collisions at RHIC and LHC are not a simple superposition of several collisions, whereas the collisions scale with the collisions.

3.1.2. Average Transverse Mass and Bjorken Energy Density

Figure 5(a) shows the values for pions in central heavy-ion collisions as a function of . The value increases with at lower AGS energies [50, 51], stays independent of for the SPS energies [52, 53], and then tends to rise further with increasing at the higher beam energies of LHC. About 25% increase in is observed from RHIC [41, 54] to LHC [55]. For a thermodynamic system, can be an approximate representation of the temperature of the system, and may represent its entropy [56]. In such a scenario, the observations could reflect the characteristic signature of a phase transition, as proposed by Van Hove [57]. Then, the constant value of versus has one possible interpretation in terms of formation of a mixed phase of a QGP and hadrons during the evolution of the heavy-ion system. The energy domains accessed at RHIC and LHC will then correspond to partonic phase, while those at AGS would reflect hadronic phase. However, there could be several other effects to which is sensitive, which also need to be understood for proper interpretation of the data [56].

Figure 5(b) shows the product of the estimated Bjorken energy density ( ; [58] is the transverse overlap area of the nuclei, and is the transverse energy) and formation time ( ) as a function of [5964]. The product of energy density and the formation time at LHC seem to be a factor of 3 larger compared to those attained at RHIC. If we assume the same value of   (=1 fm/c) for LHC and RHIC, the Bjorken energy density is about a factor of 3 larger at the LHC compared to that at RHIC in central collisions.

3.1.3. Volume and Decoupling Time

The top panel of Figure 6 shows the energy dependence of the product of the three radii ( , , and ) obtained from pion HBT or Bose-Einstein correlation analysis. Here, the “out” corresponds to the axis pointing along the pair transverse momentum, the “side” to the axis perpendicular to it in the transverse plane, and the “long” corresponds to the axis along the beam (Bertsch-Pratt convention [65, 66]). The product of the radii is connected to the volume of the homogeneity region at the last interaction. The product of the three radii shows a linear dependence on the charged-particle pseudorapidity density. The data indicates that the volume of homogeneity region is two times larger at the LHC than at RHIC.

Furthermore, within a hydrodynamic picture, the decoupling time for hadrons ( ) at midrapidity can be estimated from the magnitude of radii as follows: = , with , where is the mass of the pion, is the kinetic freeze-out temperature, and and are the integer-order modified Bessel functions [67]. For the estimation of , the average value of the kinetic freeze-out temperature is taken to be 120 MeV from AGS to LHC energies. However, the energy dependence of kinetic freeze-out temperature, as discussed in the next subsection, would provide a more accurate description of the values. The extracted values for central heavy-ion collisions at midrapidity at AGS [68], SPS [69, 70], RHIC [71, 72], and LHC [73] energies are shown as a function of cube root of in the bottom panel of Figure 6. We observe that scales linearly with and is about 10 fm/ at LHC energies. This value is about 40% larger than at RHIC. It may be noted that the above expression ignores transverse expansion of the system and finite chemical potential for pions. Also there are uncertainties associated with freeze-out temperature that could lead to variations in the extracted values.

3.1.4. Freeze-Out Temperature and Radial Flow Velocity

The hadron yields and spectra reflect the properties of the bulk matter at chemical and kinetic freeze-out, respectively. Generally, the point at which the inelastic collisions cease is called the chemical freeze-out, and the point where even the elastic collisions stop is called the kinetic freeze-out.

The transverse momentum distribution of different particles contains two components: one random and the other collective. The random component can be identified with the temperature of the system at kinetic freeze-out ( ). The collective component, which could arise from the matter density gradient from the center to the boundary of the fireball created in high energy nuclear collisions, is called collective flow in transverse direction ( ). Using the assumption that the system attains thermal equilibrium, the blast wave formulation can be used to extract and . These two quantities are shown in Figure 7 versus [41, 55, 7477]. For beam energies at AGS and above, one observes a decrease in with . This indicates that the higher the beam energy is, the longer interactions are among the constituents of the expanding system and the lower the temperature. From RHIC top energy to LHC, there seems to be, however, a saturation in the value of . In contrast to the temperature, the collective flow increases with the increase in beam energy, rapidly, reaching a value close to 0.6 times the speed of light at the LHC energy.

Figure 8 shows the chemical freeze-out temperature ( ) versus the baryon chemical potential ( ) in central heavy-ion collisions [41, 55, 7885]. These quantities are obtained by fitting the particle yields to a statistical model assuming thermal equilibrium within the framework of a Grand Canonical ensemble. There are two values of temperature quoted for LHC energies. A value of about 164 MeV and fixed value of 1 MeV seem to reproduce the multistrange ratios (involving and ) quite well but were observed to miss the data for and . On the other hand, the statistical thermal model prediction with  MeV and fixed fits the measured and ratios better but misses the ratios involving multistrange hadrons [86]. This issue is not yet resolved, being possibly related to hadronic final state interactions [87]. The curve corresponds to generalization of the energy dependence of using statistical thermal model calculations [78, 79]. The model works within the framework of a Grand Canonical ensemble and takes as input the produced particle yields from experiments to extract the freeze-out parameters such as and .

3.1.5. Fluctuations

One of the proposed signatures to search for the phase transition from hadronic to partonic medium is to study the net-charge fluctuations in heavy-ion collisions. The partonic phase has constituents with fractional charges, while the hadronic phase has constituents with integral units of charge; hence, the measure of the fluctuations in the net-charge particle production is expected to be different in these two cases. Specifically, net-charge fluctuations are expected to be smaller if the system underwent a phase transition. However, it is important to address how these fluctuations may or may not survive the evolution of the system in the heavy-ion collisions. An experimental measure of net-charge fluctuations is defined as , where and are average negative and positive charged particle multiplicity, respectively [88].

Figure 9 shows the product of and (average number of charged particles) as a function of [8991]. We find that this observable fluctuation rapidly decreases with and approaches expectation for a simple QGP-like scenario [92] as we move from RHIC to LHC energies. Given that several other observables already indicate that a hot and dense medium of color charges has been formed at RHIC and LHC, the net-charge fluctuation result may indicate that the observable is not sensitive enough to QGP physics or the process of hadronization washes out the QGP signal for this observable. It may be also noted that the model’s results do not incorporate the acceptance effects and do not consider any dynamic evolution of the system like, for example, the dilution of the signals in the hadronization process.

3.2. Azimuthal Anisotropy
3.2.1. Energy Dependence of Integrated

Figure 10 shows the integrated close to midrapidity of charged particles for collision centralities around 20–30% as a function of center of mass energy. We observe that there is an increase in magnitude of by about 30% from top RHIC energy ( = 200 GeV) to LHC energy (  TeV). This needs to be viewed within the context of a similar magnitude of increase in of pions from RHIC to LHC energies. The increase of beyond beam energy of 10 GeV is logarithmic in . This is expected to be determined by the pressure gradient-driven expansion of the almond-shape fireball produced in the initial stages of a noncentral heavy-ion collision [93] while for measured at lower beam energies, the dependences observed are due to interplay of passing time of spectators and time scale of expansion of the system. A preference for an inplane emission versus out-of-plane (“squeeze-out”) pattern of particles as a function of beam energy is observed. The experimental data used are from FOPI [94, 95], EOS, E895 [96], E877 [97], CERES [98], NA49 [99], STAR [100], PHOBOS [101], PHENIX [102], ALICE [25], ATLAS [103], and CMS [1720] experiments. Charged particles are used for LHC, RHIC, CERES, and E877 experiments, pion data is used from NA49 experiment, protons’ results are from EOS and E895 experiments, and FOPI results are for all particles with .

3.2.2. Azimuthal Anisotropy Coefficients versus Transverse Momentum

Figure 11(a) shows the comparison of , , and for 30–40% collision centrality at RHIC (PHENIX experiment [104]) and LHC (ALICE [105]) at midrapidity in and collisions, respectively. The bottom panel of this figure shows the ratio of LHC and RHIC results to a polynomial fit to the LHC data. The measurement techniques are similar at RHIC and LHC energies. One observes that at lower ( 2 GeV/ ), the and are about 10–20% smaller at RHIC compared to the corresponding LHC results. However, at higher , the results are quite similar. The seems higher at RHIC compared to that at LHC.

One of the most striking observations to come out from RHIC is the number of constituent quark ( ) scaling of for identified hadrons. The basis of such a scaling is the splitting of between baryons and mesons at intermediate (2–6 GeV/ ). This is shown in the bottom panels of Figure 11(b). Such a splitting between baryon and meson is also observed at intermediate at LHC energies (seen in the top panels of Figure 11(b)). However, the degree to which scaling holds could be different at RHIC [106] and LHC [107] energies. The scaling is much more closely followed at RHIC compared to LHC. It may be noted that there are several factors which could dilute such scaling, which include energy dependence of radial flow, an admixture of higher Fock states, and consideration of a realistic momentum distribution of quarks inside a hadron [108, 109]. The observation of the baryon-meson splitting is commonly interpreted as due to substantial amount of collectivity being generated in the deconfined phase. Another important feature is that at both RHIC and LHC energies, a clear hydrodynamic feature of mass dependence of is observed at low ( 2 GeV/ ).

Figure 12 shows the charged hadron for 30–40% collision centrality in collisions at = 200 GeV and collisions at  TeV for [1720]. This figure demonstrates the kinematic reach for higher energy collisions at LHC relative to RHIC. LHC data allows us to study the in the range never measured before in heavy-ion collisions. The for  GeV/ might suggest that those particles must have been emitted very early in the interactions when the collective effects had not set in. These high transverse momentum data are useful to understand the effects of the initial geometry or path-length dependence of various properties associated with parton modification inside the hot QCD medium. In addition, it also provides significantly improved precision measurement of for  GeV/ .

3.2.3. Flow Fluctuations

Fluctuations in azimuthal anisotropy coefficient have gained quite an attention in recent times. In particular, the measurement of event-by-event fluctuations can pose new constraints on the models of the initial state of the collision and their subsequent hydrodynamic evolution. In extracting event-by-event fluctuations, one needs to separate nonflow effects, and so far, there is no direct method to decouple fluctuations and nonflow effects in a model independent from the experimental measurements. However, several techniques exist where the nonflow effects can be minimized; for example, flow and non-flow contributions can be possibly separated to a great extent with a detailed study of two particle correlation function in and its dependence on and . Here, we discuss another technique to extract and compare the fluctuations at RHIC and LHC. We assume that the difference between (two-particle cumulant) and (four-particle cumulant) is dominated by fluctuations, and nonflow effect is negligible for . Then, the ratio can be considered as an estimate for fluctuations in the data. Figure 13 shows the as a function of collision centrality and for RHIC [110] and LHC [107] energies. The centrality dependence of at RHIC or LHC as seen in Figure 13 could be an interplay of residual nonflow effects which increases for central collisions and multiplicity fluctuations which dominate smaller systems. It is striking to see that when presented as a function of % cross section is similar at RHIC and LHC, suggesting it reflects features associated with initial state of the collisions, for example, the event-by-event fluctuations in the eccentricity of the system. But when presented as a function of , it tends to suggest a different behavior for most central collisions at RHIC.

Recently, a great interest has been generated on extracting initial condition and flow fluctuation information from the measurement of the probability distribution of at LHC. The probability density of can be expressed as a Gaussian function in transverse plane [111] as or as one dimensional Bessel-Gaussian function [112, 113] as , where is the modified Bessel function of the first kind and is the fluctuation in , with for ( measured with respect to reaction plane).

Figure 14 shows the and values extracted from the distributions as a function of by fitting to the above probability functions [114]. They are compared with values of and obtained directly from the distributions. The value is always smaller than the value for , and it decreases to zero in the 0–2% centrality interval. The value of is close to , except in the most central collisions. This leads to a value of larger than over the full centrality range as shown in Figure 14(c). The value of decreases with and reaches a minimum at but then increases for more central collisions. Thus, the event-by-event distribution brings additional insight for the understanding of fluctuations.

3.3. Nuclear Modification Factor

Figure 15 shows the of various particles produced in heavy-ion collisions at RHIC and LHC. In Figure 15(a), we observe that the shape of the versus of charged hadrons at RHIC and LHC [26, 27] is very similar for the common range of measurements. The values at RHIC are higher compared to those at LHC energies up to  GeV/ . The higher kinematic reach of LHC in allows us to see the full evolution of in high energy heavy-ion collisions. All these measurements suggest that the energy loss of partons in the medium formed in heavy-ion collisions at LHC energies is perhaps larger compared to that at RHIC. In Figure 15(b), we observe that the nuclear modification factors for collisions at  GeV [115] and collisions at  TeV [116] are greater than unity for the  GeV/ . The values for RHIC are slightly larger compared to those for LHC. A value greater than unity for the nuclear modification factor in collisions is generally interpreted as due to Cronin effect [117, 118]. However, several other physics effects could influence the magnitude of the nuclear modification factor in collisions such as nuclear shadowing and gluon saturation effects. But the results that the nuclear modification factors in collisions are not below unity strengthen the argument (from experimental point of view) that a hot and dense medium of color charges is formed in collisions at RHIC and LHC. In Figure 15(c), we show the of particles that do not participate in strong interactions, and some of them are most likely formed in the very early stages of the collisions. These particles (photon [119, 120], [121], and [122] bosons) have an 1, indicating that the , observed for charged hadrons in collisions, is due to the strong interactions in a dense medium consisting of color charges.

4. Comparison to Model Calculations

In this section, we compare some of the experimental observables discussed above with corresponding model calculations. This helps us to interpret the data at both RHIC and LHC energies. We restrict our discussion on the comparison of the models with the experimental data for charged particle production, ratio of kaon to pion yields as a function of beam energy, dependence of , and for charged particles and pions. For the charged particle production, we compare the experimental data with models inspired by the perturbative QCD-based calculations (HIJING, DPMJET) with macroscopic models (statistical and hydrodynamical), microscopic models (string, transport, cascade, etc.), and calculations which are derived by the different parametrizations of the nucleon-nucleon and nucleus-nucleus lower energy data. The ratio of kaon to pion yields for different beam energies is compared with the statistical and thermal models. The transverse momentum dependence of is compared with models incorporating the calculations based on hydrodynamic and transport approaches. Finally, the results are compared with the perturbative QCD-based calculations with different mechanism for the parton energy loss in the presence of colored medium.

4.1. Charged Particle Multiplicity Density and Particle Ratio

Figure 16 compares the measured charged particle pseudorapidity density at RHIC (0.2 TeV) and LHC (2.76 TeV) energies to various model calculations.

Empirical extrapolation from lower energy data (named “Busza” in the figure) [123] significantly under-predicts the measurement at LHC energies. A simple power-law growth of charged-particle multiplicities near midrapidity in central collisions seems to be followed up to RHIC energies (named as “Barshay and Kreyerhoff” in the figure) [124]. Perturbative QCD-inspired Monte Carlo event generators, the HIJING model without jet quenching [125], the Dual Parton Model [126] (named “DPMJET III” in the figure), and the Ultrarelativistic Quantum Molecular Dynamics model [127] (named “UrQMD” in the figure) are consistent with the measurement. The HIJING model results without jet quenching were also consistent with the RHIC measurements. The semimicroscopic models like LEXUS are successful in explaining the observed multiplicity at RHIC (named as “Jeon and Kapusta” in the figure) [128]. Models based on initial-state gluon density saturation have a range of predictions depending on the specifics of the implementation [129133]. The best agreement with LHC data happens for model as described in (named as “Kharzeev et al.” and “Armesto et al.” in the figure) [131, 133]. Conclusions for RHIC energy from these models are similar. The prediction of a hybrid model based on hydrodynamics and saturation of final-state phase space of scattered partons (named as “Eskola et al.” in the figure) [134] is slightly on a higher side compared to the measurement at LHC. But such a model seems to do a reasonable job for RHIC energies [135]. Another hydrodynamic model in which multiplicity is scaled from collisions overpredicts the measurement (named as “Bozek et al.” in the figure) [136]. Models incorporating constituent quark scaling and Landau hydrodynamics (named as “Sarkisyan and Sakharov” in the figure) [137, 138] and based on modified PYTHIA and hadronic re-scattering (named as “Humanic” in the figure) [139] underpredict the measurement at LHC energy. At RHIC energies, models considering minijet production in ultrarelativistic heavy-ion collisions by taking semihard parton rescatterings explicitly into account underpredict the multiplicities (named as “Accardi” in the figure) [140]. It is also seen at RHIC energies that models based on string fusion [141] and dual string model [142] seem to work well, whereas those based on heavy-ion cascade LUCIFER model [143] underpredict the data.

Figure 17 shows the versus for collisions at  TeV [14]. Also shown are the corresponding RHIC results scaled up by a factor 2.15. Remarkable similarity is observed in the shape of the distributions at RHIC and LHC energies. Particle production based on saturation model explains the trends nicely (named as “ALbacete and Dumitru” in the figure) [144] (published after the most central value [25] was known). simple fit to the data using a power law form for the also explains the measurements. In addition, a functional form inspired by the detailed shape of pseudorapidity distribution of charged particle multiplicity distributions at RHIC [47] explains the centrality trends nicely.

Strangeness production in heavy-ion collisions is a classic signature for formation of QGP [145]. The particle yield ratio could reflect the strangeness enhancement in heavy-ion collisions with respect to the elementary collisions. Figure 18 shows the energy dependence of ratio for central collisions at midrapidity. It will be interesting to see which model explains such an impressive collection of systematic data on ratio. Figure 18 also shows the energy dependence of ratio from various theoretical model calculations. The energy dependence of ratio has been interpreted using the Statistical Model of Early Stage (SMES) [146]. The model predicts first-order phase transition and the existence of mixed phase around beam energy of 7-8 GeV. The SHM or statistical hadronization model [147] assumes that the strong interactions saturate the particle production matrix elements. This means that the yield of particles is controlled predominantly by the magnitude of the accessible phase space. The system is in chemical nonequilibrium for  GeV, while for higher energies, the oversaturation of chemical occupancies is observed. The statistical model [148] assumes that the ratio of entropy to as a function of collision energy increases for mesons and decreases for baryons. Thus, a rapid change is expected at the crossing of the two curves, as the hadronic gas undergoes a transition from a baryon-dominated to a meson-dominated gas. The transition point is characterized by  MeV,  MeV, and  GeV. In the thermal model [79], the energy dependence of is studied by including -meson, which is neglected in most of the models, and many higher mass resonances (  GeV/ ) into the resonance spectrum employed in the statistical model calculations. The hadronic nonequilibrium kinetic model [149] assumes that the surplus of strange particles is produced in secondary reactions of hadrons generated in nuclear collisions. Then, the two important aspects are the available energy density and the lifetime of the fireball. It is suggested that these two aspects combine in such a way so as to show a sharp peak for the strangeness-to-entropy or ratio as a function of beam energy. In the hadron resonance gas and hagedorn model [150], all hadrons as given in PDG with masses up to 2 GeV/ are included. The unknown hadron resonances in this model are included through Hagedorn’s formula for the density of states. The model assumes that the strangeness in the baryon sector decays to strange baryons and does not contribute to the kaon production. The energy dependence of ratio seems to be best explained using model.

This systematic measurement of ratio reveals two interesting pieces of information. (a) The ratio shows a peak around  GeV, while the ratio increases monotonically; the peak indicates the role of the maximum baryon density at freeze-out around this collision energy. (b) For  GeV, pair production becomes the dominant mechanism for production, so both the ratios and approach the value of 0.16. Taking into account the different masses between pions and kaons, this asymptotic value corresponds to a temperature of the order of 160 MeV.

4.2. Azimuthal Anisotropy

The azimuthal anisotropy parameter , measured at RHIC and LHC, provides a unique opportunity to study the transport properties of the fundamental constituents of any visible matter, a system of quarks and gluons. Furthermore, it provides an opportunity to understand whether the underlying dynamics of the evolution of the system formed in the collisions are governed by macroscopic hydrodynamics [151153] or by microscopic transport approach [154]. Figure 19 shows the versus for 30–40% collision centrality and collisions at midrapidity for  GeV and 2.76 TeV, respectively. The measurements are compared to a set of model calculations based on hydrodynamic approach (including THERMINATOR [155, 156]) and another set of calculations based on transport approach. It is observed that hydrodynamic-based models explain the measurements both at RHIC and LHC energies. Transport-based models including partonic interactions (like AMPT [154]) also explain the measurements. However, those transport models which do not incorporate partonic interactions like UrQMD [157, 158] fail to explain the data. The model comparison also reveals that the data favors a high degree of fluidity reflected by a small value of shear viscosity to entropy density ratio . A more detailed comparison of the model calculations with various order azimuthal anisotropy parameters would in the near future give us a more quantitative picture of the temperature (or energy) dependence of transport coefficients of the system formed in the heavy-ion collisions.

4.3. Nuclear Modification Factor

The nuclear modification factor ( ) is an observable used to study the structure of strongly interacting dense matter formed in heavy-ion collisions. Here, we discuss the observation of at high seen at RHIC and LHC by comparing two models within perturbative QCD- (pQCD-) based formalisms. In this picture, the high hadrons are expected to originate from the fragmentation of hard partons (hard scattering scales larger than QCD scales of 200 MeV). The hard partons lose energy through interactions with the hot and dense mediums, which get reflected in the observed values of . The processes by which they could lose energy includes radiative energy loss and elastic energy loss. For a more elaborate discussion on these models, we refer the reader to the review article [159].

In Figure 20, we show a comparison between experimentally measured versus at LHC and RHIC energies and corresponding pQCD-based model calculations. All theoretical formalisms require a microscopic model of the medium to set the input parameters for the energy loss calculation. These parameters, for example, are denoted as , the transport coefficient of the medium or the gluon number density per unit rapidity. The parameter , on the other hand, reflects the strength of elastic energy loss put in the model calculations. Without going into deeper theoretical discussions of each model, we refer the readers to the following related publications: PQM [160], GLV [161], ASW [162], YaJEM [163], WHDG [164], and ZOWW [165]. However, for completeness and to elucidate the approach taken in the model calculations, we briefly mention two formalisms as examples: the GLV approach named after their authors Gyulassy, Levai, and Vitev and ASW approach named after the corresponding authors Armesto, Salgado and Wiedemann, where the medium is defined as separated heavy static scattering centers with color screened potentials, where as in some other formalism, a more precise definition of the medium is considered as being composed of quark gluon quasiparticles with dispersion relations and interactions given by the hard thermal loop effective theory.

We observe that most models predict the dependence of well for collisions both at RHIC and LHC energies. The models specially capture the generally rising behavior of that is observed in the data at high for the LHC energies. The magnitude of the predicted slope of versus varies between models, depending on the assumptions for the jet-quenching mechanism. The models shown do not need larger values of medium density in the calculation to explain the for  GeV/ at RHIC and LHC for the common kinematic range. They however, require a high medium density at LHC energy to explain the values of for  GeV/ .

5. Summary

In summary, the results on multiplicity density in pseudorapidity, HBT, azimuthal anisotropy, and nuclear modification factor from LHC experiments indicate that the fireball produced in these nuclear collisions is hotter, lives longer, and expands to a larger size at freeze-out compared to lower energies. These results also confirm the formation of a deconfined state of quarks and gluons at RHIC energies. The measurements at LHC provide a unique kinematic access to study in detail the properties (such as transport coefficients) of this system of quarks and gluons.

In this review, we showed that the first set of measurements made by the three LHC experiments within the heavy-ion programs, ALICE, ATLAS, and CMS, show a high degree of consistency. These measurements include centrality dependence of charged particle multiplicity, azimuthal anisotropy, and nuclear modification factor versus transverse momentum. Next, we discussed the comparison of various measurements made at RHIC and LHC energies. LHC measurements of clearly demonstrated the power law dependence of charged particle multiplicity on the beam energy. They also reconfirmed the observation at RHIC that particle production mechanism is not a simple superposition of several collisions. The values of , , freeze-out volume, decoupling time for hadrons, and and are larger at LHC energies compared to those at RHIC energies, even though the freeze-out temperatures are comparable. The value of the net-charge fluctuation measure is observed to rapidly approach towards a simple model-based calculation for QGP state. However, the sensitivity of this observable for a heavy-ion system as well as the lack of proper modeling of the heavy-ion system theoretically for such an observable needs careful consideration. The fluctuations as a function of centrality fraction have a similar value at both RHIC and LHC. This reflects their sensitivity to initial state effects. Just like at RHIC, the and direct photon measurements experimentally demonstrated that the observed for charged hadrons is a final state effect; also at LHC, the , direct photon, and and measurements showed that the observed is indeed due to formation of a dense medium of colored charges in central heavy-ion collisions. All these conclusions were further validated by the comparison of several observables to corresponding model calculations. Further, it was found that the fluid at LHC shows a comparable degree of fluidity as that at RHIC. This is reflected by a small value of shear viscosity to entropy density ratio.

Measurements-related heavy quark production [166168], dilepton production, jet-hadron correlations [169, 170], and higher-order azimuthal anisotropy [171, 172] which are now coming out of both RHIC and LHC experiments will provide a much more detailed characterization of the properties of the QCD matter formed in heavy-ion collisions.


The authors would like to thank F. Antinori, S. Gupta, D. Keane, A. K. Mohanty, Y. P. Viyogi, and N. Xu for reading the paper and for their helpful discussions and comments. Bedangadas Mohanty is supported by the DAE-SRC project fellowship for this work. Lokesh Kumar is partly supported by the DOE Grant DE-FG02-89ER40531 for carrying out this work.