ESTRO 2025 - Abstract Book

S2880

Physics - Dose prediction, optimisation and applications of photon and electron planning

ESTRO 2025

damage induced by ionizing radiation. This study investigates the effect of refining the diffusion dynamics of radiolytic products, by inserting interaction potentials in the equations of motion, on the temporal evolution of the radiochemical yields. Material/Methods: Radiolytic products interact with each other through potentials, which represent intermolecular forces, such as electrostatic or Lennard-Jones forces. A path integral approach can be obtained by exploiting the stochastic nature of the underlying motion of Brownian particles. By propagating the particles along the path where the action takes the least possible value, the action can generate the ensemble of trajectories in the force field as well as their probabilities. The consideration of intermolecular interactions will consequently lead to a change in the reaction probabilities, which are necessary for the evaluation of the reactions within the reaction model. By a local evaluation of the stochastic action in the force field, the reaction probability can be sampled numerically. A C++ module has been developed to implement the proposed concept. Using a water-filled volume, the diffusion dynamics modelled using the presented method were compared to established Monte Carlo simulations using Geant4-DNA [1]. For this purpose, the G-values (number of molecules / 100 eV of deposited energy) of different radiolytic products were compared time-resolved for an irradiation of 1 MeV electrons. Results: The time-dependent prediction of G-values between Geant4-DNA and the model presented in this study is compatible when potentials are disregarded in our approach as it is the case in Geant4-DNA. When accounting for the interaction potentials, differences can only be seen for later times (> 1 ns) and result in differences of 8% at maximum (see fig. 1).

Figure 1: Time evolution of the G-value of radiolytic products for 1 MeV incident electrons, depositing a total energy of 1 keV in the water phantom, as simulated by Geant4-DNA and the code developed in this work.

Conclusion: In this study, a C++ code was developed that is able to model the interactions of radiolytic products, incorporating potentials that are neglected in the diffusion dynamics of Geant4-DNA.

Keywords: radiation chemistry, radiolysis, Monte Carlo

References: [1] Karamitros, M. et al. “Diffusion-controlled reactions modeling in Geant4-DNA”, Journal of Computational Physics . 2014; 274:841-882.