Monte Carlo simulation for the analysis of various solid samples using handheld X-ray fluorescence spectrometer and evaluation of the effect by environmental interferences

Handheld X-ray fluorescence (HH-XRF) has expanded its utilization areas according to recent technological developments. Most current applications, though, are still concentrated in traditional areas including mineral resource analysis and environmental regulation rather than forensic science for the purpose of investigating a nuclear security event involving nuclear material out of regulatory control. To apply HH-XRF to nuclear material analysis, it is necessary to first obtain calibration data using standard reference materials. Considering the difficulty in obtaining such standard reference materials as well as the high costs involved, one well-known alternative method is to use Monte Carlo simulation code. This study investigated the feasibility of employing Monte Carlo N-Particle transport 6 (MCNP6) simulation to provide calibration data through comparison with experimental measurements of pure solid samples of graphite, copper, SiO2, and UO2 using HH-XRF. The results showed that the MCNP6 simulation results were entirely consistent with the measurement spectra, except for environmental interferences stemming from interactions with the mechanical components below 10 keV which varied slightly according to sample type. To quantitatively evaluate the effect of these environmental interferences on the whole spectrum, the coefficient of determination (R2) was used. In the case of graphite, the effect of the environmental interferences was evaluated to be about 20% on the conformity of the measured and simulated results, while those for copper, SiO2, and UO2 were about 1%, 3%, and less than 1%, respectively. These results indicate that samples having elements with higher rates of photoelectric absorption followed by fluorescence compared to scattering tend to decrease the effect of the environmental interferences over the entire spectrum. The origin of the environmental interferences was estimated to be interference with the detector shield and/or X-ray tube collimator, which are particular design features of the device used. Their effect on contributing to the environmental interferences was evaluated by experiment for the detector shield and simulation for the X-ray tube collimator. As the detector shield was found to only contribute to a decrease in overall spectrum intensity, the major contributor to the environmental interferences was determined to be the collimator. It is believed that the results of this study will help to confirm that Monte Carlo sim