Detection of nuclear recoils at the 100 eV scale : characterization of the response of bolometers and application to coherent scattering of reactor neutrinos with the nucleus experiment

Abstract

This thesis is structured around two main axes: on the one hand, my contribution to theNUCLEUS experiment, which aims to detect nuclearrecoils (NR) induced by the coherent elastic scatteringof reactor neutrinos off nuclei (CEνNS); on theother hand, the development of the CRAB method (Calibrated Recoil for Accurate Bolometry) for the energy calibration of NR in cryogenic detectors. TheNUCLEUS experiment will soon be installed near the Chooz nuclear power plant (Ardennes, France). At this site, the main sources of background are ambient γ radiation, atmospheric neutrons, and cosmic muons. Background reduction is a central challenge for NUCLEUS, and a prerequisite for the observation of the CEνNS signal. To address this, the experiment relies on a cryogenic target detector, passive shielding, and two veto systems, designed to identify and reject various background contributions. Prior to its installation at Chooz, the experiment was deployed in an underground laboratory in Munich for a preparatory phase, including a blank assembly followed by a First Commissioning Run (FCR). In this thesis, I present the commissioning of the muon veto system. The response and performance of all modules were successfully validated. I also discuss the analysis of FCR data, which confirmed the proper functioning of all NUCLEUS detectors. Particular effort was devoted to the synchronization of the different data acquisition systems to validate the efficiency of the veto systems in rejecting background signals. In particular, the muon veto was shown to be highly effectivein rejecting cosmic muon-induced background.The second part of this thesis presents my contributions to the development and validation of the CRAB method. This method enables the calibration of nuclear recoils in the sub-keV energy range, representing a key advancement for future neutrino and dark matter experiments. The method is based on the radiative neutron capture by a target nucleus. During the de-excitation of the nucleus via γ emission, it undergoes a recoil with energy determined by two-body kinematics, thereby generating a calibration peak in the region of interest. I conducted simulations to validate this approach for several target materials (CaWO4, Al2O3, Ge, Si), focusing in particular on the impact of the γ-NR coincidence on the signal-to-background ratio. I also took part in the experimental validation of the method on an Al2O3 target. As part of my joint PhD program with the Technical University of Vienna (Austria), I spent one year on site contributing to the instrumental development of Phase II of the CRAB project,carried out at the TRIGA-Mark II research reactor at the Atominstitut, which provides a purely thermal neutron beam. This thesis presents the setup and commissioning of the γ detectors at this facility. Finally, the analysis of the first data collected in Vienna yielded promising results, in good agreement with simulations, and confirmed the γ-NR coincidence.