Validation and investigation of reactive species yields of Geant4‐ DNA chemistry models
Abstract
Purpose: Indirect biological damage due to reactive species produced in water radiolysis reactions is
responsible for the majority of biological effect for low linear energy transfer (LET) radiation. Modeling
water radiolysis and the subsequent interactions of reactive species, as well as track structures, is
essential to model radiobiology on the microscale. Recently, chemistry models have been developed
for Geant4-DNA to be used in combination with the comprehensive existing physics models. In the
current work, the first detailed, independent, in silico validation of all species yields with published
experimental observations and comparison with other radiobiological simulations is presented. Additionally,
the effect of LET of protons and heavier ions on reactive species yield in the model was
examined, as well as the completeness of the chemical reactions following the radiolysis within the
time after physical interactions simulated in the model.
Methods: Yields over time of reactive species were simulated for water radiolysis by incident electrons,
protons, alpha particles, and ions with various LETs using Geant4 and RITRACKS simulation
tools. Water dissociation and recombination was simulated using Geant4 to determine the completeness
of chemical reactions at the end of the simulation. Yield validation was performed by comparing
yields simulated using Geant4 with experimental observations and other simulations. Validation was
performed for all species for low LET radiation and the solvated electron and hydroxyl radical for
high LET ions.
Results: It was found that the Geant4-DNA chemistry yields were generally in good agreement with
experimental observations and other simulations. However, the Geant4-DNA yields for the hydroxyl
radical and hydrogen peroxide at the end of the chemistry stage were found to be respectively considerably
higher and lower than the experimentally observed yields. Increasing the LET of incident
hadrons increased the yield of secondary species and decreased the yield of primary species. The
effect of LET on the yield of the hydroxyl radical at 100 ns simulated with Geant4 was in good agreement
with experimental measurements. Additionally, by the end of the simulation only 40% of dissociated
water molecules had been recombined and the rate of recombination was slowing.
Conclusions: The yields simulated using Geant4 are within reasonable agreement with experimental
observations. Higher LET radiation corresponds with increased yields of secondary species and
decreased yields of primary species. These trends combined with the LET having similar effects on
the 100 ns hydroxyl radical yield for Geant4 and experimental measurements indicate that Geant4
accurately models the effect of LET on radiolysis yields. The limited recombination within the modeled
chemistry stage and the slowing rate of recombination at the end of the stage indicate potential