Identification Of Photons In Ultra-Relativistic Heavy-Ion Collisions

ALICE (A Large Ion Collider Experiment) is an experiment at the LHC (Large Hadron Collider), in construction at CERN, and optimized for the study of heavy-ion collisions at a center-of-mass energy of 5.5 A TeV. The prime aim of the experiment is to study the production of the quark-gluon plasma at high temperatures and high energy densities.PHOS (PHOton Spectrometer) is the electromagnetic calorimeter of high granularity integrated in the ALICE experiment. One of the major goal of PHOS is to detect and identify direct photons, represent only about 5 to 20% of the total photon yield. The photon background mainly comes from π0 and η decays and misidentified hadrons. The identification of particles is provided in two ways. On one hand a charged particle detector is attached to PHOS to discriminate charged particles against neutral particles. On the other hand particle identification is provided during offline analysis by algorithms that examine the electromagnetic shower shape in PHOS to discriminate electromagnetic particles against hadrons.In the present work we have studied theses two complementary methods to identify photons in the electromagnetic spectrometer PHOS.We have first studied a new concept, PPSD (PHOS Pre-Shower Detector) , for the charged particle detector. PPSD (PHOS Pre-Shower Detector) (Fig.3.1) consists of a sandwich of charged particle detectors: Charged Particle Veto detector (CPV), Passive Converter and Photon Conversion detector (PC). The charged particle detectors. CPV and PC, are the same gaseous detectors based on the MICROMEGAS technology. The detector is an asymmetric two-stage parallel plate detector. The first stage, a 6mm thick conversion gap, is separated from the second one, a 200 μm thick amplification gap, by a micro-mesh frame resting on insulating spacers. The gas mixture flows through thedetector with a pressure slightly above atmospheric pressure. Prototypes of PPSD have been tested at CERN PS (Proton Synchrotron) with beams of 1 or 2 GeV/c in momentum. The gains, efficiencies and probabilities of spark have been calculated.We have then tested the offline algorithm with data simulated with ALIROOT which is a offline framework for simulation, reconstruction and analysis of the data coming from the real interactions in the ALICE experiment. ALIROOT uses C++ as an implementation language and is based on ROOT. In particular, we have developped an identification method based on PCA (Principal Components Analysis). This method has the advantage to reduce the number of necessary parameters by using the correlations between many parameters that characterize the shower developping in the calorimeter following the impact of an ionising particle. Photon identification in the momentum range from 0.5 to 100 GeV by using PCA method is presented and compared with the results attained by traditional shower topology method.