Abstract
To improve radiotherapy, especially that with ion beams such as proton and carbon ion beams, the mechanisms of interactions induced by ionizing radiation must be understood. MPEXS2.1-DNA is a Monte Carlo simulation code developed for water radiolysis studies and DNA damage simulations that uses GPU devices for fast computation. However, the original chemistry model in MPEXS2.1-DNA did not include detailed chemical reactions for reactive oxygen species (ROS), e.g., O(•-), O(2), O(2)(•-), HO(2)(•), HO(2)(-). In the present study, drawing the former work on the step-by-step (SBS) model for the RITRACKS code, we implemented an alternative SBS model into MPEXS2.1-DNA to increase the capabilities and computational speed of water radiolysis simulations under ion irradiation. This model is based on the theory of Green's function of the diffusion equation (GFDE-SBS). Also, we implemented multiple ionization processes which enhance ROS generation under high-LET irradiation. We compared the simulation results obtained by GFDE-SBS with experimental data from previous studies. The validation results demonstrated that the GFDE-SBS model accurately reproduced the measured radiation chemical yields of major species, such as hydroxyl radicals and hydrogen peroxide. Furthermore, the computational speed of GFDE-SBS was increased approximately ten times faster than the original model due to the changes in time stepping. Additionally, simulations using a Fricke dosimeter confirmed that this model is reliable for long-term simulations over seconds. These improvements enable simulations of radiation interactions and can help in the study of DNA damage mechanisms.