Automated data processing of neutron depth profiling spectra using an Artificial Neural Network

Abstract

This work examines the possibility of using an Artificial Neural Network (ANN) to interpret Neutron Depth Profiling (NDP) spectra. AMonte Carlo of N-Particle (MCNP6) radiation transport model was used to simulate the alpha energy spectrum of NIST standard SRM 2137 during a 10B(n, α)7Li NDP measurement. Simulations of 300 randomly generated specimens were also performed and used to train an Artificial Neural Network (ANN). The depth profile of boron in the SRM2137 NIST standard was obtained by processing the MCNP pulse height light tally with the trained ANN. This was compared to the results of traditional analysis using stopping tables. The traditional analysis and ANN results both agree well with the reference SRM 2137 boron profile.