Sterile neutrino search with KATRIN - modeling and design-criteria of a novel detector system

Abstract

A fundamental phenomenon in particle physics is the absence of massive objects in our universe: Dark Matter. A promising candidate that could explain these observations are sterile neutrinos with a mass of several $\mathrm{keV}/c^2$. While it is presumed that sterile neutrinos do not interact via the weak force, they, due to their mass, still partake in neutrino oscillation. Consequently, it is experimentally possible to investigate their imprint in beta-decay experiments, such as the Karlsruhe tritium neutrino experiment (KATRIN). A dedicated search for sterile neutrinos however ensues a steep increase in the electron rate and thus requires the development of a new detector system, the TRISTAN detector. In addition, as the imprint of sterile neutrinos is presumably $<10^{-7}$, systematic uncertainties have to be understood and modeled with high precision. In this thesis systematics prevalent at the detector and spectrometer section of KATRIN will be discussed and their impact to a sterile neutrino sensitivity illuminated. The derived model is compared with data of the current KATRIN detector and with characterization measurements of the first TRISTAN prototype detectors, seven pixel silicon drift detectors. It is shown that the final TRISTAN detector requires a sophisticated redesign of the KATRIN detector section. Moreover, the combined impact of the back-scattering and electron charge-sharing systematic lead to an optimal detector magnetic field of $B_\mathrm{det}=0.7\dots0.8\,\mathrm{T}$, which translates to a pixel radius of $r_\mathrm{px}=1.5\dots1.6\,\mathrm{mm}$. The sensitivity analysis discusses individual effects as well as the combined impact of systematic uncertainties. It is demonstrated that the individual effects can be largely mitigated by shifting the tritium \bd energy spectrum above the \bd endpoint. In contrast, their combined impact to the sensitivity leads to an overall degradation and only mixing amplitudes of $\sin^2\theta_4<3\cdot10^{-6}$ would be reachable, even in an optimized case with very low and homogeneous detection deadlayer $z_\mathrm{dl}=20\pm1\,\mathrm{nm}$. Assessing sterile neutrino mixing amplitudes of $\sin^2\theta_4<10^{-7}$ thus requires disentangling of systematic effects. In a future measurement this could be for example achieved by vetoing detector events with large signal rise-times and small inter-event times.