Normalization and filling optimization of the τSPECT experiment

Abstract

The free neutron decay is one of the simplest forms of β-decay, and the experimentally measured neutron mean lifetime is a relevant parameter in astrophysics, cosmology, and particle physics. It can be used in unitarity tests of the Cabibbo-Kobayashi-Maskawa (CKM) matrix and is an essential input to Big Bang Nucleosynthesis (BBN), where it determines the theoretical primordial helium abundance. To date, two main methods have been established to measure the neutron lifetime: 1. Beam experiments, which determine the neutron decay rate by detecting the decay products of a cold neutron beam and 2. bottle experiments, which confine neutrons with very low kinetic energies, called ultra- cold neutrons (UCNs), and count the number of surviving neutrons after a known storage period. However, the results of those two methods deviate by more than 4σ, a discrepancy known as the “neutron lifetime puzzle”. In order to eliminate systematic effects related to interactions with material walls, recent bottle-type experiments are using magneto-gravitational traps, to confine neutrons by utilizing gravity and magnetic fields. The τ SPECT experiment aims to deliver the first competitive neutron lifetime measurement based on purely magnetic storage, with a target uncertainty of 0.3 s. The experiment was commissioned at the research reactor TRIGA Mainz in Germany and was relocated to the Paul Scherrer Institute (PSI) in Switzerland in 2023. τSPECT confines neutrons in a magnetic field which is a superposition of a radial field of a Halbach permanent-magnet octupole and a longitudinal field generated by superconducting coils. The trap is filled by manipulating the spin of incoming neutrons using rf magnetic fields. Filling of τSPECT can be realized with two different methods: the single spin- flip (sSF) method, and the double spin-flip (dSF) method. In this thesis, the dSF method was refined from prototype stage to an established procedure. This improvement was achieved by an upgrade to the rf circuitry and the implementation of an absolute magnetic-field reference. With the improved dSF method, the number of filled neutrons increased by a factor of 1.94(5) compared to the previously used sSF method. A second major contribution of this thesis is the development and characterization of an in-beamline neutron detector to measure the UCN flux at the experiment entrance. The detector uses a coincidence scheme for event discrimination, to suppress background caused by dark counts and γ-radiation. In the characterization the monitor detector shows good agreement with a commercial UCN detector, and it can be considered effectively background-free for operation with τSPECT. Once the detector was integrated into the experiment, it became possible to monitor the UCN source output. This knowledge is key, since extracting the neutron lifetime from τSPECT measurements requires correcting for fluctuations in the number of initially trapped neutrons. Using data of the first scientific data run of the experiment in winter 2024, I investigated whether, and how well, the τSPECT measurements can be normalized using the monitor detector measurements. A promising approach to normalization has been found and is presented in this thesis.