New algorithms for data analysis were coupled with new techniques for event-by-event data handling. These were tested using data collected with an AmBe fast neutron source and compared to simulated data. Using the measured fast neutron background and estimates of the measurement uncertainties from stationary data runs, simulations involving relative motion between source and detector show promising results for this technology.

A rich number of astronomical and cosmological observations suggest the existence of a massive, non-luminous, and non-relativistic, matter component in the universe which is five times more abundant than baryonic matter and is commonly referred as to dark matter (DM). Although so far eluding from detection, one class of promising DM candidates are weakly interacting massive particles (WIMPs) which arise naturally from many beyond the Standard Model (BSM) theories. The XENON Dark Matter Project aims to directly detect WIMPs, and other kinds of rare event signals, by utilizing large-scale liquid xenon (LXe) dual-phase time projection chambers (TPCs). The newest generation of experiment, called XENONnT, utilizes a TPC with a total sensitive LXe mass of 5.9 t, and was designed as a fast upgrade of its predecessor XENON1T. In addition to its larger TPC, XENONnT was augmented with the world's first water Cherenkov neutron veto (NV), which was mounted inside the already existing water Cherenkov muon veto water tank of XENON1T. Neutrons emitted by detector materials can undergo a single back-scatter inside the TPC producing a signal which is indistinguishable from WIMPs. The NV has the task to mitigate this potential threat for the scientific reach of the experiment by tagging these escaping neutrons through their delayed neutron capture on hydrogen. In the presented work, the results of the first weakly interacting massive particle (WIMP) search science run, called SR0, are discussed. SR0 features a blind analysis between 3.3 keV and 60.5 keV nuclear recoils energies with a total exposure of about 1.1 tonne-year, utilizing the lowest ever achieved electronic recoil background of (15.8 ± 1.3) events/(t · y · keV) in a LXe. No significant excess was found in the data, setting the lowest upper limit of 2.58 · 10-47 cm2 for spin-independent (SI) interactions of 28 GeV/c2 WIMPs at a 90% confidence level. These results have also been published in [Apr+23b] as part of the presented work. To obtain these results, this thesis discusses the commissioning of the XENONnT neutron veto (NV), and the calibration of its neutron tagging efficiency. The tagging efficiency was found to be (53.1±2.8)% which is the highest efficiency ever measured in a water Cherenkov detector. The efficiency of the NV, as well as the nuclear recoil (NR) response of the time projection chamber (TPC), were calibrated using tagged neutrons from an Americium-Beryllium (AmBe) neutron source. This technique was deployed for the first time in a liquid xenon (LXe) TPC. It enables a calibration of the NR response with high purity and a remaining pollution of less than 0.1 %. Further, the same calibration data was used to determine the thermal neutron capture cross section of hydrogen which was found to be 336.7 ± 0.4 (stat.)+2.0 -0.0 (sys.) mb. All these analyses are based on the data provided by XENONnT's new processing framework called STRAXEN. As part of the lead developing team, the entire processing chain for the two veto systems of XENONnT was developed, and many additional tools have been implemented. Finally, to enhance the neutron tagging efficiency of the NV even further, the water inside the water tank is going to be doped with Gd-sulfate. As part of the presented work, different Gd-salt samples of the manufacturer Treibacher were analyzed regarding their suitability for the experiment.

The work presented in this book is a major step towards understanding and eventually suppressing background in the direct search for dark matter particles scattering off germanium detectors. Although the flux of cosmic muons is reduced by many orders of magnitude in underground laboratories, the remaining energetic muons induce neutrons through various processes, neutrons that can potentially mimic a dark matter signal. This thesis describes the measurement of muon-induced neutrons over more than 3 years in the Modane underground laboratory. The data are complemented by a thorough modeling of the neutron signal using the GEANT4 simulation package, demonstrating the appropriateness of this tool to model these rare processes. As a result, a precise neutron production yield can be presented. Thus, future underground experiments will be able to reliably model the expected rate of muon-induced neutrons, making it possible to develop the necessary shielding concept to suppress this background component.

This book discusses the upgrade of the Super-Kamiokande (SK) detector, which consists in the addition of a salt of gadolinium into the detector’s water, the goal being to endow it with a very high-efficiency ability to detect neutrons: the SuperK-Gd project. This will substantially improve the scientific value of the SK detector because, among others, neutron production is related to the matter–antimatter character of the interacting neutrino. In this book the authors develop several procedures for maximizing the impact of neutron tagging in various physics analyses involving a broad range of neutrino energy. They thoroughly study the impact of new backgrounds introduced by Gd in key physics analyses, most remarkably including the search for the Diffuse Supernova Neutrino Background. At GeV energies, the neutron tagging improvements are evaluated by performing a complete neutrino oscillation sensitivity study using atmospheric and long baseline neutrinos, with a focus on the neutrino mass hierarchy and the leptonic CP violation. In order to prove the relevance of neutron tagging with the available data, the authors apply the neutron-tagging tools developed here to the 4th phase of the SK detector, which is already capable of detecting a low fraction of the neutrons produced through hydrogen-neutron captures. A global oscillation analysis of the SK’s atmospheric neutrino data is also conducted.