I’ll present the results from lower energy running of CERES experiment at SPS. The data taking period of 1999 was the first run of CERES after the TPC upgrade. So my talk is divided into two parts, one is the analysis of e+e- production and the other is the investigation of hadronic observables at 40 GeV.

The CERES experiment is dedicated to the measurement of low-mass dileptons in heavy-ion collisions. Dileptons have long been predicted to be sensitive probes of the hot and dense nuclear medium due to their large mean free path. They can leave the interaction region without experiencing the final state interactions and carry information about the very early stage of the collision where temperature and density are the highest. Their measures yield and spectra reflect the full space-time evolution of the system, including the production of e+e- pairs via quark-antiquark annihilation and quark-gluon Compton scattering in case of quark-gluon plasma formation,  pi+pi- annihilation in the hot and dense hadron gas and decays of neutral mesons at freeze-out. 

The previous CERES results have generated a lot of attention. It was shown that in proton-beryllium collisions the invariant mass distribution can be well described by the hadronic decays of neutral mesons. On the other hand, in lead-gold collisions at the top SPS energy we observed an enhancement of dilepton yield in the mass region of 0.25 to 0.7 GeV/c2. The amount of excess yield may be accounted for via thermal radiation from a hadronic fireball. The spectral shape, however, is difficult to explain and the best description of the data has been obtained by the models which require introducing the in-medium modifications of the vector meson properties. A lower energy run of CERES allows to probe higher baryon densities, thus providing additional constraints for the theoretical models.   

So, how do me measure dileptons in CERES? The heart of the CERES spectrometer are two Ring Imaging Cherenkov detectors which provide electron identification, I’ll illustrate this on the next slide. The beam comes from the left and hits the segmented gold target. The two silicon drift detectors are used for vertex reconstruction and charged particle multiplicity measurements.  The radial drift TPC is located behind the original CERES spectrometer and provides the momentum and dE/dx measurement for all charged particles.

The event display of the central led-gold event at 40 GeV as it is seen by the RICHes is shown here. The RICHes are operated at high gamma-threshold such that they are almost insensitive to the charged hadrons. Electrons, on the other hand, produce ring of assymptotic radius which are recognized and matched to the external detectors. In 1999 the CERES operated without magnetic field between the two RICHes, such that the rings in two detectors would completely overlap. This allowed to improve efficiency in ring finding from 81 to 94%.

The measurement of dilepton spectrum is quite challenging as the dominant sources of e+e—pairs are of trivial origin, namely pi-zero Dalitz decays and gamma-conversions. These sources are characterized by small mass and opening angle. Two such pairs in an event, with only one leg reconstructed in the spectrometer, can mimic a high-mass open pair, resulting in a combinatorial background. The effective rejection of dalitzes and conversion is crucial for extracting the open pair signal. The most effective rejection tool is a pt-cut of 200 MeV/c, which rejects 85% of the low-mass pairs while keeping 97% of the high mass pairs. The pairs with opening angles smaller than 10 mrad cannot be recognized in the RICHes as two separate rings. We can reject such pairs, however, using the energy loss information in the two silicon drift detectors. This is illustrated on this plot, where we show the dE/dx measured in one silicon detector versus the one in the other. The isolated tracks show single de/dx in both silicon detectors. e+e- pairs with small opening angle deposit double energy in the silicon detectors and can be clearly rejected. In addition to this, we also required that electron track candidates have a proper de/dx measured by the TPC 

The results of the analysis at 40 GeV are presented here. From the total event sample of 8.7 M events we obtained 180 pairs with masses above 200 MeV. The signal-to-background achieved in the analysis is 1 to 6, compared to 1 to 13 in the 160 GeV analysis. The measured invariant mass distribution, similar to the measurement at the top SPS energy, shows an excess of dilepton yield compared to hadronic decays of neutral mesons.

Here we compare the experimental data with theoretical calculations done by Ralf Rapp. The hadronic decay cocktail with ro-meson taken out is shown by the dashed curve. Addition of  ro contribution assuming a vacuum rho spectral function is illustrated by the red curve. This is so-called hadronic scenario without involving any in-medium modifications of the rho properties. Also shown are two calculations with modifications of the rho spectral function: one assuming the broadening of the rho due to interactions with the surrounding hadrons in black, and the other assuming a reduction of the rho-mass as a precursor of chiral symmetry restoration. Similar to 158 GeV, these two calculations provide an adequate description of the measured spectra.  

Let me now switch to hadron observables measured by CERES. The addition of TPC allows to investigate the whole spectrum of hadron physics, including particle spectra and yields, ratios, HBT, flow, event-by-event fluctuations. As an example, I show invariant mass distributions of lambda’s, k-zero shorts, and phi’s. Some of the analyses are still in progress, today I’ll report on some selected results obtained from 40 GeV run.

The spectra of negative hadrons and proton-like positive net charges are shown here for the top 15% of the geometrical cross-section. The extracted slopes are similar to those observed at the AGS and the top SPS energies. We also show the rapidity densities measured at mid-rapidity number of participants. The midrapidity yield of negative hadrons rises significantly stronger than linear with centrality. Such non-linear rise has been observed at the AGS, while close to linear dependence has been measured at the top SPS energy.

The measured Lambda pt-spectra are shown on the left plot for three different centralities. The extracted slopes are plotted versus number of participants and compared to those measured at the AGS, top SPS, and RHIC energies. For the highest centrality, the slopes are quite similar, also there’s an indication of some energy dependence as 40 GeV slopes are systematically lower than those at 160, which are in turn lower than those at RHIC. 

The beam energy dependence of  lambda yields at midrapidity is shown on the left plot. Our data seem to favor smooth rise of the mid-rapidity lambda yield. We have also been able to extract lambda-bar to lambda ratio which also smoothly follows the energy dependence. 

We have also performed a three-dimensional correlation analysis of h-h- and h+h+ pairs at midrapidity. The extracted radius parameters, R-long and R-side are shown as function of the pair transverse momentum for different event centralities. Consistent with the picture of a boost-invariant longitudinal expansion, R-long shows a strong kt dependance in all centrality bins. Using the following parameterization, the derived duration of the expansion is 6-7 fm/c for the freeze-out temperature of 120 MeV. Kt-dependence of R-side, on the other hand, becomes more pronounced for more central collisions, suggesting an increase with centrality of radial flow. An extracted transverse expansion velocity is close to the velocity of sound in nuclear medium and is similar to finding at top SPS and AGS energies. 

Using azimuthal hit distributions in the silicon drift chambers, we have measured the strength of elliptic flow which is plotted versus beam energy. The CERES measurements at 40, 80 and 160 GeV show in-plane elliptic flow with the amplitude increasing with energy. The measurements follow a general trend with energy,  that is an increase in initial pressure achieved in the collision.

Finally, let me summarize. The dilepton spectra at 40 GeV show similar enhancement of dilepton yield observed at 158 GeV.  The data, although of low statistics, provide useful constraint to theory. The hadron data at 40 GeV help filling the gap in between the AGS and top SPS energies. More very interesting results will come from the analysis of high-statistics data taken by CERES in 2000. These data should provide a long-awaited high-resolution measurement of dileptons in rho/omega/phi mass region. And, of course, more results on hadron production is in the works, including the comparison within the same experiment the phi-yields in the leptonic and hadronic channels.