Improvement studies on neutron-gamma separation in HPGe detectors by using neural networks

TL;DR

This paper explores the use of neural networks to enhance neutron-gamma separation in HPGe detectors, aiming to improve background rejection in nuclear reaction gamma-ray spectra.

Contribution

It introduces an ANN-based method that improves neutron rejection and reduces gamma-ray loss compared to previous criteria.

Findings

Neutron rejection ratio increased by 27%

Gamma-ray loss decreased by 50%

ANN method enhances spectral clarity

Abstract

The neutrons emitted in heavy-ion fusion-evaporation (HIFE) reactions together with the gamma-rays cause unwanted backgrounds in gamma-ray spectra. Especially in the nuclear reactions, where relativistic ion beams (RIBs) are used, these neutrons are serious problem. They have to be rejected in order to obtain clearer gamma-ray peaks. In this study, the radiation energy and three criteria which were previously determined for separation between neutron and gamma-rays in the HPGe detectors have been used in artificial neural network (ANN) for improving of the decomposition power. According to the preliminary results obtained from ANN method, the ratio of neutron rejection has been improved by a factor of 1.27 and the ratio of the lost in gamma-rays has been decreased by a factor of 0.50.