Abstract

The KOTO experiment at J-PARC aims to search for physics beyond the Standard Model through the observation of the rare neutral kaon decay . Within the Standard Model, its branching ratio is predicted to be extremely small, approximately , while benefiting from an exceptionally low theoretical uncertainty. Since this decay originates from short-distance interactions, it is highly sensitive to contributions from new physics. However, the current experimental sensitivity of KOTO has not yet reached the Standard Model prediction. To achieve a statistically significant observation, a next-generation experiment, KOTO-Ⅱ, which requires more than two orders of magnitude improvement in sensitivity, is necessary. As experimental sensitivity improves, enhanced background rejection capability is also required. Accordingly, the development of a new detector capable of measuring previously inaccessible photon incident angles is underway. The detector, named PAScal (Photon Angle Sampling calorimeter), aims to reconstruct the photon incident angle by measuring the spatial distribution of electromagnetic showers. The detector consists of narrow and elongated modules composed of alternating layers of tungsten (W) and scintillating fibers, arranged in both horizontal and vertical directions to achieve three-dimensional sampling of electromagnetic showers.​ In this study, we investigated the light yield, one of the fundamental performance parameters of the calorimeter, to improve the performance of the PAScal module. Scintillation light produced in the scintillating fibers by shower particles is transmitted through the fiber bundle cross section fixed with optical cement and converted into electrical signals by a photon sensor, an MPPC (Multi-Pixel Photon Counter). The number of photons reaching the MPPC depends on the surface condition of the fiber bundle cross section. We quantitatively investigated the dependence of the light yield on the surface condition of the PAScal module. The fiber bundle cross section was polished step by step using sandpapers with grit numbers ranging from #120 to #4700, and the light yield was measured at each polishing stage. As the grit number increased, the surface roughness decreased, which led to an improvement of the light yield by up to approximately 46%. Furthermore, we confirmed that additional improvement in light yield can be achieved by optimizing the optical coupling between the fiber bundle cross section and the MPPC.