Home | Sitemap | Contact us
this site all CERN
Physics Banner - nucleus, hadron, and QCD particle's artistic representation


Some results from the first two years of lead ion collisions

There have been two runs at the LHC with lead ion collisions at 2.76 TeV per nucleon pair : the first at the end of 2010 (November-December) and the second at the end of 2011 (November- December). In what follows there is a compilation of the most spectacular results from these runs. Often comparison is done with the corresponding results from RHIC, the Relativistic Heavy Ion Collider at Brookhaven, where the energy is ~ 14 times lower.

Calibrating at the LHC

The interactions between heavy ions are complex and for their interpretation the knowledge of the initial conditions of the fireball at the instant after the collision is essential. The multiplicity, that is, the total number of particles produced in a collision, tells us a lot about how the quarks and gluons of the incoming nuclei transform into particles (pions, kaons etc) observed in the detectors; also about the energy density reached within the collision and the temperature of the fireball. The number of generated particles is correlated with the distance between centres of the colliding nuclei (impact parameter). Head-on (central) collisions (small impact parameter), when the largest number of incoming protons and neutrons participate in the collision, generate most particles. The charged particle multiplicity per colliding nucleon pair measured by ALICE for the most central collisions is double that measured at RHIC, where the collision energy is factor 14 lower, fig.1. This shows that the system created at LHC has much higher energy density and is at least 30% hotter that at RHIC. Fig. 2 shows the charged particle multiplicity as a function of the number of participants.

Figure 1 (left). Charged particle multiplicity per colliding nucleon pair versus number of nucleons participating in the collision. Figure 2 (right). Charged particle multiplicity per colliding nucleon pair as a function of the collision energy.

A perfect liquid at the LHC

An interesting observable used for the study of the quark gluon plasma is flow: it provides information on the equation of state and the transport properties of matter created in heavy-ion collisions. Multiple interactions between the constituents of the created matter and initial asymmetries in the spatial geometry of non-central collisions result in an azimuthal anisotropy in particle production. The measured azimuthal distribution of particles in momentum space can be decomposed into Fourier coefficients. The second Fourier coefficient of this azimuthal asymmetry is known as elliptic flow. Its magnitude depends strongly on the friction in the created matter, characterized be the ratio η/s, where η is the shear viscosity and s the entropy. For a good fluid such as water the value of η/s is small. For a thick liquid such as honey η/s has large values. Measurements of the elliptic flow at RHIC had revealed that the hot matter created in heavy ion collisions flows like a fluid with little friction, with η/s close to the lower limit for a perfect fluid. At LHC this observation was confirmed, with values of the elliptic flow higher by 30% with respect to those at RHIC.

Elliptic Flow

The size of the fireball

A technique called Bose-Einstein or HBT Interferometry allows us to measure the size and lifetime of the fireball created in heavy-ion collisions. This method is devised by the pioneering work of Hanbury, Brown and Twiss in astronomy and looks at pairs of particles (two-pion correlation functions). In hadron and ion collisions, there is enhanced production of bosons close together in phase space, due to Bose Einstein statistics. This leads to an excess of pairs at low relative momentum. The width of the excess region is inversely proportional to the size of the system at decoupling, at the point when most particles stop interacting. The quark gluon plasma behaves as a fluid, with strong collective motions described well by hydrodynamic equations. The collective flow makes the size of the system appear smaller as the pair momentum increases. The radius of the pion source is measured in three dimensions : along the beam axis, Rlong; along the transverse momentum of the pair, Rout; and in a direction perpendicular to these two, Rside, show in in Fig. 1. The similarity between Rout and Rside indicates a short duration of the emission – explosive emission. The time when the emission reaches its maximum, with respect to the first encounter, is found to be 10-11 fm/c, significantly longer than at RHIC. The product of the three radii – an estimate of the homogeneity volume of the system at decoupling - is twice as large as at RHIC. The conclusion is that at LHC the fireball formed in heavy ion collisions is hotter, lives longer and expands to a larger size that at lower energies.

Partonic energy loss – Jet Quenching

When the fast partons (quarks and gluons) produced from heavy ion collisions propagate through the dense medium of the fireball, they lose energy via gluon radiation or elastic scattering. The amount of radiated energy depends on the density of the medium and distance travelled by the parton in the medium, as well as the flavour of the parton. These partons become observable as jets of hadrons when they hadronize and the energy loss becomes evident in a phenomenon known as “jet quenching”. Instead of two jets going back-to-back and having similar energies, a striking imbalance is observed, one jet being almost absorbed by the medium. This is demonstrated in the following figure.

One way of studying nuclear effects of the medium is by looking at the nuclear modification factor, RAA. This is the ratio of the charged particle pT spectrum in lead-lead collisions, normalized to the number of binary collisions , divided by the corresponding spectrum from proton-proton collisions.

Direct photons and measurement of the QGP temperature

One of the classic signals expected for a quark–gluon plasma (QGP) is the radiation of "thermal photons", with a spectrum reflecting the temperature of the system. With a mean-free path much larger than nuclear scales, these photons leave the reaction zone created in a nucleus–nucleus collision unscathed. So, unlike hadrons, they provide a direct means to examine the early hot phase of the collision. However, thermal photons are produced throughout the entire evolution of the reaction and also after the transition of the QGP to a hot gas of hadrons. In the PbPb collisions at the LHC, thermal photons are expected to be a significant source of photons at low energies (transverse momenta, pT, less than around 5 GeV/c). The experimental challenge in detecting them comes from the huge background of photons from hadron decays, predominantly from the two-photon decays of neutral pions and ? mesons.

Direct photons are defined as photons not coming from decays of hadrons, so photons from initial hard parton-scatterings (prompt photons and photons produced in the fragmentation of jets) – i.e. processes already present in proton–proton collisions – contribute to the signal. Indeed, for pT greater than around 4 GeV/c, the measured spectrum agrees with that for photons from initial hard scattering obtained in a next-to-leading-order perturbative QCD calculation. For lower pT, however, the spectrum has an exponential shape and lies significantly above the expectation for hard scattering, as the figure shows. The inverse slope parameter measured by ALICE, TLHC = 304 ± 51 (stat.+syst.) MeV, is larger than the value observed in gold–gold collisions at √s = 0.2 TeV at Brookhaven’s Relativistic Heavy-Ion Collider (RHIC), TRHIC = 221 ± 19 (stat.) ± 19 (syst.) MeV. In typical hydrodynamic models, this parameter corresponds to an effective temperature averaged over the time evolution of the reaction. The measured values suggest initial temperatures well above the critical temperature of 150–160 MeV (approx. 1.8 × 1012 K) at which the transition between ordinary hadronic matter and the QGP occurs. The ALICE measurement also indicates that the LHC has produced the hottest piece of matter ever formed in a laboratory.

Strangeness enhancement

Ordinary matter around us is made of protons and neutrons, which in turn are composed of up (u) and down (d) quarks. The next quark that can be liberated from the sea of quark-anti-quark pairs that populate the vacuum is the strange quark (s-quark). It is heavier than u and d, yet close enough in mass to undergo production and modification processes in similar manner. That, and the relative abundance of the strange quark in high-energy interactions, make the s-quark a very useful study tool for proton-proton and heavy nucleus collisions. Strange particles, K-mesons (Kaons, made up of a strange and a non-strange quark pair), Lambda (uds), Xi (dss), and Omega (sss) baryons have an appreciable lifetime before they decay into ordinary matter. These decays have a characteristic geometrical configuration, which allows an effective reconstruction of strange particles. The strangeness data obtained in pp collisions are particularly important to improve the overall modeling of those collisions. PYTHIA is a software package that is able to generate events from a model. The model parameterizes the low-momentum processes taking place in elementary collisions, and calculates the higher energy processes up to the next to leading order perturbative term in the perturbative QCD expansion. The latest PYTHIA versions describe the general properties of real collisions fairly closely, but significantly underestimate the yields of strange particles: the more the strangeness content of the particle, the worse is the discrepancy. One of the recent PYTHIA versions, PYTHIA Perugia-2011, made significant modifications to its s-quark cross-section, and as a result came very close to reproducing the yields of baryons with multiple strange quarks, especially at higher transverse momentum. ALICE is the only experiment at the LHC to have measured Omega baryon yields in pp collisions. In addition, the pp measurement serves as a baseline for the measurements in Pb-Pb collisions, where we expect to produce the Quark-Gluon-Plasma (QGP). An enhancement in the production of particles with strange quarks has long been thought to be a signature of extra degrees of freedom available in the QGP. Indeed, this enhancement has been seen at lower energies as well: the larger the volume of the collision, the more the number of Lambda, Xi and Omega baryons increases with respect to the baseline (a pp or a Be-Be collision). This is also observed at 2.76 TeV Pb-Pb collisions, however, with a caveat: the enhancement is smaller than that at lower energies! We think this is due to the complexity of our baseline: at these high energies, pp collisions to which we compare are more complex and produce much more strangeness than events at lower energies.

Enhancement in the production of particles with strange quarks

ALICE and the charm of heavy-ion collisions

The ALICE collaboration has measured the production of the charmed mesons D0 and D+ in lead–lead collisions at the LHC. In central (head-on) collisions they find a large suppression with respect to expectations at large transverse momentum, pt, indicating that charm quarks undergo a strong energy loss in the hot and dense state of QCD matter formed at the LHC. This is the first time that D meson suppression has been measured directly in central nucleus–nucleus collisions.

Heavy-flavour particles are recognized as effective probes of the highly excited system (medium) formed in nucleus–nucleus collisions; they are expected to be sensitive to its energy density, through the mechanism of in-medium energy loss. The nuclear modification factor RAA – the ratio of the yield measured in nucleus–nucleus collisions to that expected from proton–proton collisions – is well established as a sensitive observable for the study of the interaction of hard partons with the medium. Because of the QCD nature of parton energy-loss, quarks are predicted to lose less energy than gluons (which have a higher colour charge); in addition, the so-called "dead-cone" effect and other mechanisms are expected to reduce the energy loss of heavy partons with respect to light ones. Therefore, there a pattern of gradually decreasing RAA suppression should emerge when going from the mostly gluon-originated light-flavour hadrons (e.g. pions) to the heavier D and B mesons: RAA( ) < RAA(D) < RAA(B). The measurement and comparison of these different probes provides, therefore, a unique test of the colour-charge and mass dependence of parton energy-loss.

Experiments at the Relativistic Heavy Ion Collider at Brookhaven measured the suppression of heavy flavour hadrons indirectly in gold–gold collisions at 200 GeV through the RAA of the inclusive decay electrons. Using data from the first lead–lead run at the LHC (?sNN = 2.76 TeV), the ALICE collaboration has measured the production of prompt D mesons via the reconstruction of the decay vertex in the channels D0?K–p+ and D+?K–p+p+. The results show a suppression of a factor 4–5, as large as for charged pions, above 5 GeV/c (see figure). At lower momenta, there is an indication of smaller suppression for D than for mesons. Data with higher statistics, expected from the 2011 lead–lead run, will allow the collaboration to study this region with more precision and address this intriguing mass-dependence in QCD energy-loss. The result implies a strong in-medium energy loss for heavy quarks, as also suggested by the suppression measured by the ALICE collaboration for electrons and muons from heavy flavour decays, and by the CMS collaboration for J/ψ particles from B meson decays.

The mystery of the J/Psi

The J/ψ is composed of a heavy quark–antiquark pair with the two objects orbiting at a relative distance of about 0.5 fm, held together by the strong colour interaction. However, if such a state were to be placed inside a QGP, it turns out that its binding could be screened by the huge number of colour charges (quarks and gluons) that make up the QGP freely roaming around it. This causes the binding of the quark and antiquark in the J/ψ to become weaker so that ultimately the pair disintegrates and the J/ψ disappears – i.e. it is "suppressed". Theory has shown that the probability of dissociation depends on the temperature of the QGP, so that the observation of a suppression of the J/ψ can be seen as a way to place a "thermometer" in the medium itself.

Such a screening of the colour interaction, and the consequent J/ψ suppression, was first predicted by Helmut Satz and Tetsuo Matsui in 1986 and was thoroughly investigated over the following years in experiments with heavy-ion collisions. In particular, Pb–Pb interactions were studied at CERN’s Super Proton Synchrotron (SPS) at a centre-of-mass energy, √s, of around 17 GeV per nucleon pair and then Au–Au collisions were studied at √s=200 GeV at Brookhaven’s Relativistic Heavy-Ion Collider (RHIC).

As predicted by the theory, a suppression of the J/ψ yield was observed with respect to what would be expected from a mere superposition of production from elementary nucleon–nucleon collisions. However, the experiments also made some puzzling observations. In particular, the size of the suppression (about 60–70% for central, i.e. head-on nucleus–nucleus collisions) was found to be approximately the same at the SPS and RHIC, despite the jump in the centre-of-mass energy of more than one order of magnitude, which would suggest higher QGP temperatures at RHIC. Ingenious explanations were suggested but a clear-cut explanation of this puzzle proved impossible.

At the LHC, however, extremely interesting developments are expected. In particular, a much higher number of charm–anticharm pairs are produced in the nuclear interaction, thanks to the unprecedented centre-of-mass energies. As a consequence, even a suppression of the J/ψ yield in the hot QGP phase could be more than counter-balanced by a statistical combination of charm–anticharm pairs happening when the system, after expansion and cooling, finally crosses the temperature boundary between the QGP and a hot gas of particles. If the density of heavy quark pairs is large enough, this regeneration process may even lead to an enhancement of the J/ψ yield – or at least to a much weaker suppression with respect to the experiments at lower energies. The observation of the fate of the J/ψ in nuclear collisions at the LHC constitutes one of the goals of the ALICE experiment and was among its main priorities during the first run of the LHC with lead beams in November/December 2010.

The ALICE experiment is particularly suited to observing a J/ψ regeneration process. For simple kinematic reasons, regeneration can be more easily observed for charm quarks with low transverse-momentum. Contrary to the other LHC experiments, both detector systems where the J/ψ detection takes place – the central barrel (where the J/ψ→e+e– decay is studied) and the forward muon spectrometer (for J/ψ→μ+μ–) – can detect J/ψ particles down to zero transverse momentum.

As the luminosity of the LHC was still low during its first nucleus–nucleus run, the overall J/ψ statistics collected in 2010 were not huge, of the order of 2000 signal events. Nevertheless, it was possible to study the J/ψ yield as a function of the centrality of the collisions in five intervals from peripheral (grazing) to central (head-on) interactions.

Clearly, suppression or enhancement of a signal must be established with respect to a reference process. And for such a study, the most appropriate reference is the J/ψ yield in elementary proton–proton collisions at the same energy as in the nucleus–nucleus data-taking. However, in the first proton run of the LHC the centre-of-mass energy of 7 TeV was more than twice the energy of 2.76 TeV per nucleon–nucleon collision in the Pb–Pb run. To provide an unbiased reference, the LHC was therefore run for a few days at the beginning of 2011 with lower-energy protons and J/ψ production was studied at the same centre-of-mass energy of Pb–Pb interactions.

The Pb–Pb and p–p results are compared using a standard quantity, the nuclear modification factor RAA. This is basically a ratio between the J/ψ yield in Pb–Pb collisions, normalized to the average number of nucleon–nucleon collisions that take place in the interaction of the two nuclei and the proton–proton yield. Values smaller than 1 for RAA therefore indicate a suppression of the J/ψ yield, while values larger than 1 represent an enhancement.

Fig. 1. A comparison of the J/ψ suppression between RHIC (PHENIX) and the LHC (ALICE). The ALICE results show a strikingly smaller suppression, in particular for head-on collisions (large Npart), despite the much larger centre-of-mass energy.

The results from the first ALICE run are rather striking, when compared with the observations from lower energies (figure 1). While a similar suppression is observed at LHC energies for peripheral collisions, when moving towards more head-on collisions – as quantified by the increasing number of nucleons in the lead nuclei participating in the interaction – the suppression no longer increases. Therefore, despite the higher temperatures attained in the nuclear collisions at the LHC, more J/ψ mesons are detected by the ALICE experiment in Pb–Pb with respect to p–p. Such an effect is likely to be related to a regeneration process occurring at the temperature boundary between the QGP and a hot gas of hadrons (T≈160 MeV).

The picture arises from these observations is consistent with the formation, in Pb–Pb collisions at the LHC, of a deconfined system (QGP) that can suppress the J/ψ meson, followed by a hadronic system in which a fraction of the charm–anticharm pairs coalesce and ultimately give a J/ψ yield larger than that observed at lower energies. This picture should be clarified by the Pb–Pb data that were collected in autumn 2011. Thanks to an integrated luminosity for such studies that was 20 times larger than in 2010, a final answer on the fate of the J/ψ inside the hot QGP produced at the LHC seems to be within reach.