Study of rare FCNC Lambda_b baryon decays.

The rare FCNC decays of hadrons containing b quark provide highly sensitive tests of new physics beyond the standard model. In particular semileptonic decays of B mesons to strange meson and pair of oppositely charged leptons like B0 --> K* l+ l- received significant attention in recent years. The Lambda_b baryon decays can provide information complementary to the meson decays. Up to now only Lambda_b --> Lambda mu+ mu- decay was studied in the details. In addition, LHCb experiment performed a search for CP violation in the decay Lambda_b --> p K- mu+ mu-. In this project we aim to perform a more detailed experimental study of the Lambda_b --> p K- mu+ mu- decay. At the start, the differential branching fraction will be measured as a function of the dimuon invariant mass squared. One of the difficulties for interpretation of the measurements performed for the Lambda_b --> p K- mu+ mu- decay comes from the fact that at this stage there is no theoretical prediction or experimental information about the hadronic structure of the pK- system. The second goal of the project will be to exploit angular distributions of the decay with the aim to first contribute to the understanding of the pK- structure of this decay and possibly measure at least some of the angular observables which are sensitive to the new physics.

The important aspect of the proposed measurements of Lambda_b --> p K- mu+ mu- decay is that it heavily relies on the positive particle identification of protons and kaons. In the current LHCb detector, this is possible for kaon momentum over 10 GeV/c and proton momentum from about 15 GeV/c. These are relatively high thresholds and significant improvement in the sensitivity could be achieved if the particle identification at lower momenta can be achieved. Such improvement would not only benefit study in this project but large number of measurements in LHCb would benefit from such improvement, with one particular area to point out being flavour tagging. One proposed solution is to install a TORCH detector during LS3 to provide particle identification capability for low momentum tracks. The idea of TORCH detector is to exploit time-of-flight measurement for charged particles by precise timing of the Cherenkov photons emitted in TORCH quartz radiator bar. Preliminary studies suggest that it can provide significant improvement in LHCb detector capabilities. In this project, we will attack question of how to reconstruct information from TORCH detector into high-level quantities on particle identification at the track level. Due to the geometry of the TORCH detector, resulting Cherenkov image seen by photodetectors is complex and currently available reconstruction needs significant speedup for practical use. We will exploit alternative ideas based on machine learning, which could possibly speed up reconstruction to the point that it could be usable at the trigger level.