Photon counting method for improvement of energy resolution in CANDLES experiment

The neutrino-less double beta decay (0νββ) is a powerful tool to probe the neutrino mass. The 0νββ is forbidden in Standard Model (SM) due to the violation of lepton number conservation. In experiments searching for 0νββ, the two-neutrino double beta decay (2νββ), which is an allowed process in SM, behaves as an irreducible background within the energy of interest. To discriminate the 0νββ from 2νββ, the energy resolution must be improved. The CANDLES experiment is searching for the 0νββ of 48Ca using CaF 2 (pure) scintillator as the detector and the source. Scintillation photons are collected by Photo Multiplier Tubes (PMTs) surrounding the crystals. A Flash Analog-to-Digital Converter records the waveform of each PMT. Because of the long decay time of CaF 2 (1 µsec), we make a signal integral of 4 µsec to calculate the obtained energy. Thus, the other fluctuation in the baseline is accumulated in the signal integration, and it makes the energy resolution worse. To reduce this fluctuation, we can count the number of photoelectrons in every PMT. Still, we have a significant overlapping probability of single photoelectron signals. This overlapping probability implies the photoelectrons lost, which causes terrible energy resolution. To improve the energy resolution, we introduce an alternative method named "partial photon counting" by dividing the waveform into two parts, then, perform signal integral in the prompt part and photon counting the second part. As a result, the energy resolution is reduced 4.0-3.5 % at 1460 keV (γ-ray of 40K) and 3-2.7 % at 2615 keV (γ-ray of 208Tl).