Thesis

Enhanced event discrimination in liquid scintillators through the use of advanced pattern recognition techniques

• Call:

  IDPASC Portugal - PHD Programme 2019

• Academic Year:

  2019 / 2020

• Domains:

  Experimental Particle Physics | Astroparticle Physics

• Supervisor:

  Fernando Barao

• Co-Supervisor:

  Nuno Barros

• Institution:

  Instituto Superior Técnico

• Host Institution:

  Laboratório de Instrumentação e Física Experimental de Partículas

• Abstract:

  The SNO+ experiment aims to observe the rare neutrinoless double beta decay by doping a a large volume of liquid scintillator with natural tellurium, having the largest sensitivity of the experiments currently built. By using a large volume of liquid scintillator, SNO+ is able to obtain a competitive sensitivity to this decay, and display a portfolio with a rich physics program such as solar neutrinos, reactor and geo-antineutrinos and also be sensitive to the neutrino signal of a supernova. One of the main challenges of a liquid scintillator experiment consists in the discrimination of different physics channels and involving different particles. For instance, solar neutrinos analysis could benefit from directionality separation from backgrounds, which are not correlated with sun direction. Conversely, solar neutrinos are a major background for double beta decay search and its impact can be strongly reduced using directionality reconstruction. The main signal emitted by a charged particle in SNO+ comes from scintilation light, that is uniformly emitted over all directions. Nevertheless, a smaller light component is present due to the Cerenkov radiation, which is highly directional. This thesis proposes to explore the use of advanced methods of pattern recog- nition to enhance the discrimination capabilities of the experiment to different particle types and event topologies. The work will include the development of methods able to separate the Cerenkov and scintilation components. In addition, identification of different particle types (electron, positron, proton, alpha) can be explored via time and signal charge profiles. This will allow not only to gain a better understanding of the backgrounds, but also to improve the purity of the signal data sample and the physics sensitivity of the experiment. In a later phase the implemented techniques will be applied to perform a search for solar neutrinos and double beta decay in the SNO+ datasets.