Particle identification by nuclear emulsions in the OPERA experiment

The OPERA experiment will try to confirm the νμ→ντ oscillations by the appearance of the ν τ in a pure νμbeam. Indeed, a neutrino beam almost pure is produced at CERN (CNGS Beam) and sent to the OPERA detector. The detector is composed of two muons spectrometers and a target formed by walls of bricks. Each brick is an alternation of lead plates and emulsions. This modular structure allows to reconstruct the kink topology of the τ lepton decay with a high spatial resolution. The great challenge of the OPERA experiment is to detect the ντinteractions with the less uncertainty. To reduce this uncertainty it is essential to identify with the greatest efficiency any background event not including a tau particle. My work permits to reduce background. My principal contribution concerns the selection development, the reconstruction and the muons identification at low energy. This work is based on the setting of variables related to the deposit energy and the multiple scattering. Previously, only deposit energy was used in the analyses of pion/muon separation. This study allows doubling the muon identification efficiency at low energy. This leads to increase the background events rejection in OPERA and to decrease the contamination by 30 pour cent. I also studied the nuclear emulsions capacity to identify charged particles through the analysis of a test beam carried out by the Nagoya group. This test contains protons and pions with different energies. My work proves that the European scan system gives comparable results with those obtained by the Japanese scan system. (Author)