The bulk hadron production data reveal major aspects in which central nuclear collisions deviate from elementary pp, pA, or collisions. Most markedly, in heavy ion collisions the abundance of strange particles are enhanced by a factor of two or more, depending on particle type, and the slopes of the momentum spectra change significantly and in a systematic way for different hadron species. The question whether these modifications are consistent with the formation of hadronic matter in thermodynamic equilibrium can be addressed by present experimental data at least for the late freeze-out stage when particles seize to interact strongly.
In a purely thermal system of hadrons, the momentum distributions, when expressed as a function of the transverse mass mT ( ), will be independent of the particle mass with a slope inversely proportional to the temperature T. In a transversely expanding system, an additional collective flow component can develop which blue-shifts the momentum spectra with a common transverse velocity leading to a mass dependent component. Similarly, longitudinal expansion can be characterised in terms of a longitudinal velocity or a rapidity interval y over which thermal sources are distributed. The abundance of particle species in equilibrium hadronic matter is given by two independent parameters, i.e. the temperature T and a baryochemical potential (which reflects the baryon asymmetry in the initial state). A hadronic system in both 'thermal' (momentum) and 'chemical' (particle abundances) equilibrium is therefore fully determined by only a few independent parameters: T, , and the expansion velocity profile. Such a simple prescription seems to be indeed borne out by the data. This is illustrated in Fig. , which shows a comparison of measured particle ratios with predictions based on chemical equilibrium. Within the experimental accuracy, these ratios are in rather good agreement for a narrow temperature range of 160 to 180 MeV. Taking into account the yields of all available hadron ratios simultaneously, best agreement is reached for T=16510 MeV and = 175 10 MeV. A similar analysis for the AGS data yields comparably good agreement with a somewhat lower temperature and a higher baryochemical potential. The resulting freeze-out points in the plane for both AGS and SPS are shown in Figure . The corresponding baryon densities are 2/3 and 1/3 of normal nuclear matter density, respectively.
shows the transverse mass spectra of and
in central Pb+Pb collisions. The spectra appear near-exponential but the
inverse slopes -- which would naively be identified with emission temperatures
in an expanding hadron gas -- increase with hadron mass and reach up to
about 300 MeV for baryons. While the slopes in pp reactions are independent
of particle type, this so called mT scaling is not observed
in Pb+Pb collisions and, moreover, the inverse slope parameters far exceed
the Hagedorn limit. However, the momentum spectra of all different particle
species in Pb+Pb reactions are well described with a single temperature
value of 165 MeV, consistent with the one derived from the analysis of
particle ratios, if, in addition, a common flow velocity of
0.3 c is introduced. Recent data from experiment NA49 for two-pion Bose
correlation functions confirm the existence of an ordered transverse, and
in fact also longitudinal expansion velocity pattern. This provides an
independent experimental tool to disentangle disordered (thermal) and ordered
(expansion) motion. The temperature and flow velocity determined from momentum
spectra at the AGS and SPS have been included in the systematics of Figure .
A large set of independent hadronic observables, i.e. momentum spectra,
particle ratios and HBT correlation results seems to be consistent with
a surprisingly simple picture of the late stages of heavy ion reactions:
a dense hadronic system expanding in almost complete thermodynamical equilibrium,
until a rapid freeze-out fixes momentum spectra and particle ratios to
the finally observed values. In addition, the location of this freeze-out
point in the temperature-density plane is located very close to the phase
boundary. Model calculations that include a phase transition and follow
abundances first of quarks and then of hadrons in the QGP phase, the mixed
phase and the hadron phase via rate equations indicate in fact that the
hadron yields are characteristic for the point when the system has completely
converted into the hadronic phase. There is in fact the possibility that
the spectral distributions are frozen in at a somewhat lower temperature
than the particle yields. The spectra would then require an accordingly
somewhat higher flow velocity.
Production: A Signal for Deconfinement?
A decade ago the suggestion was made that suppression of charmonium production should be a signature of deconfinement. The heavy pair which, at the modest SPS energy, can only be created in a hard parton collision in the initial phase of a nucleus-nucleus reaction serves as a probe of the surrounding high energy density matter. It was pointed out that the evolution of an initial pair towards its final J/or ' hadronic state could be blocked if it is embedded in a state of deconfined quarks and gluons. In such a medium, Debye screening renders colour interactions short-ranged, thus breaking up the co-travelling pairs (like any other hadronic bound states), to end up in open charm mesons. A suppression of the eventually observed J/or ' in central nuclear collisions was thus expected in a dynamical evolution proceeding via a deconfinement phase. Consequently one of the SPS experiments has concentrated on production of the J/ or ' vector mesons.
Subsequently acquired data on J/
production in p+p, p+A and S+A collisions indeed exhibited an increasing
suppression as shown in Figure .
The J/ yield of the various collision
systems is plotted as its ratio to the continuum, the latter yield being
independent of the course of the reaction dynamics. The horizontal scale
represents the average path length L traversed by the
pair after its creation inside the target and projectile nuclei. This length
thus also increases with the collision centrality. The normalised J/
yield drops exponentially, with the system size, up to and including central
S-U collisions, finally to depart from this attenuation law in the new
data gathered for Pb+Pb. The exponential attenuation, consistent with an
absorption cross-section of about 6 mb, is seen today as resulting from
the interaction between the nuclear medium and a pre-resonance state, a
configuration which evolves only later (and outside the nucleus) into the
physical, color neutral J/or '
hadron. The final drop of the normalised J/
yield in Pb+Pb indicates a further suppression mechanism, setting in rather
abruptly at a core energy density slightly above the one reached in central
S-U. The tentative conclusion, debated intensely at present, is that we
witness in these data the effect of a partonic medium which blocks the
development of the pairs
Low Mass Lepton Pairs: A Signal for Chiral Symmetry Restoration?
Weakly interacting electromagnetic probes (photons or leptons) are a direct means of gaining information on the early dense and hot stages of the collision, as they leave the interaction volume without being altered by final state effects. In fact, since electromagnetic radiation is emitted throughout the evolution of the system it contains information from all stages of the dynamics cumulatively. While, so far, only upper limits exist for direct (thermal) photon production, recent data on lepton pairs show an unexpectedly large yield at low masses, below the meson mass. Figure shows the electron pair mass spectrum observed in central S+Au collisions by NA45. The upper part summarises model calculations which include contributions from hadronic decays (shaded area) and from in-medium pion annihilation and Bremsstrahlung. An excess at 0.2 0.6 remains unexplained. The lower panel exhibits perfect agreement with the data obtained in models which include in addition an in-medium reduction of the and masses, driven by the high baryon density. A similar excess, consistent with the same model calculations, has been found in the mass spectrum by NA34/3. Possible indications of in-medium modifications of kaons at high baryon density were already discussed in section in connection with results from the GSI SIS. In-medium modification of vector mesons, if experimentally confirmed by better statistics and resolution data, could be a direct consequence of the chiral symmetry transition at the phase boundary between hadronic matter and the QGP. The rapidly varying quark condensate (see Figure ) should lead to changes in the properties of hadrons (masses, width) in the vicinity of the phase transition, which will be observable in the lepton mass spectrum for mesons decaying in the dense transition regime. This would indeed be a spectacular verification of the concept underlying the generation of light hadron masses in QCD. To establish this interpretation of the data a systematic variation of essential variables, baryon density and temperature, appears necessary. This will imply on the one hand running of the SPS at the lowest possible beam momentum of 30-40 GeV/c per nucleon. On the other hand, as discussed in the previous section, this physics will be addressed by the Hades experiment at the GSI SIS with heavy ions in the 2-3 GeV/c per nucleon range. It clearly would be desirable to fill the gap in between. But also the studies with nucleon and pion beams at the SIS will provide important information about in-medium modifications at nuclear matter density.