Systematic error of counting efficiency estimation due to spectral conformity assessment in Geant4 based liquid scintillation counter simulation

Abstract

Liquid scintillation counter (LSC) is the most powerful tool for a quantitative measurement of radioactivity in the field of low level and low energy beta decay radionuclide. LSC counts the number of flushes of scintillation lights generated by the charged particles of decay products, e.g. electrons in C-14 and H-3 decay. However, due to the influence of quench, part of the nuclear decay energy and the generated scintillation photons may escape the detection. Therefore, the compensation of the count rate loss is essential to determine the sample's activity. We previously investigated the performance of Geant4 Monte Carlo simulation for estimating counting efficiencies of C-14 and H-3 samples, respectively. In the simulation, the counting efficiency of each quench sample was estimated by adjusting the scintillation photon yield to reproduce the measured LSC spectrum. Since the conformity assessment of LSC spectra between a simulation and a measurement directly influences the counting efficiency, the assessment protocol need to be verified. In this paper, the conformity of simulated and measured LSC spectra was examined by introducing three feature points, which were defined at 50 %, 20 % and 5 % of the peak counts in the spectrum. The deviation of pulse height at each feature point and the detection efficiency were obtained by varying the scintillation photon yield in simulations. Then, the relations among the pulse height deviations, detection efficiencies and scintillation photon yields were analyzed. As a result, the most feasible feature point was the 5 % of the peak counts in the spectrum for both C-14 and H-3, respectively. By requiring that the simulated spectrum coincides with the measured spectrum at the 5 % feature point, the differences of counting efficiencies between simulations and measurements became within the measurements' uncertainty.