Computer simulation of the light yield nonlinearity of inorganic scintillators

To probe the nature of the physical processes responsible for the nonlinear scintillation light yield of inorganic scintillators, we have combined an ab initio based Monte Carlo code for calculating the microscopic spatial distributions of electron-hole pairs with an atomistic kinetic Monte Carlo (KMC) model of energy-transfer processes. In the present study, we focus on evaluating the contribution of an annihilation mechanism between self-trapped excitons (STE) to the scintillation response of pure CsI and Ce-doped LaBr3. A KMC model of scintillation mechanisms in pure CsI was developed previously and we introduce in this publication a similar model for Ce-doped LaBr3. We show that the KMC scintillation model is able to reproduce both the kinetics and efficiency of the scintillation process in Ce-doped LaBr3. Relative light output curves were generated at several temperatures for both scintillators from simulations carried out at incident gamma-ray energies of 2, 5, 10, 20, 100, and 400 keV. These simulations suggest that STE-STE annihilation can account for the initial rise in relative light yield with increasing incident energy for both types of materials. This is due to the fact that the proportion of high-density regions decreases as the incident energy increases, thus reducing the likelihood for STE-STE encounter. In addition, the simulations clearly show a lack of temperature dependence of the relative light output, in agreement with a majority of experimental work on the temperature dependence of nonlinearity in inorganic scintillators. The collective modeling tool is a fundamental advance over phenomenological modeling approaches because it has its foundation in first-principles physics of scintillation. While the KMC simulations here are parametrized largely by empirically derived rate constants, this study suggests that combining ab initio based electron-hole pair distributions with ab initio derived rate constants for a select set of energy transfer processes is in principle sufficient to detach this tool from experiment entirely, yielding a holistic predictive simulation framework useful for exploring a wide range of scintillator performance characteristics. (C) 2009 American Institute of Physics. [DOI: 10.1063/1.3143786]