A Unified Track-Structure–Informed Radiochemistry Simulation Framework for Continuous Chemical-Stage Modeling Applicable to Radiotherapy Delivery
Abstract
Purpose
Radiochemistry modeling is pivotal for elucidating the spatiotemporal kinetics of reactive oxygen species and biological effects, particularly for emerging modalities like FLASH radiotherapy. However, conventional hybrid radiochemistry models are constrained by the limited phantom volume and often introduce artificial temporal artifacts and restrictive periodic boundary conditions (PBCs). To address these challenges, this study develops a unified track-structure-informed framework to model water radiolysis continuously from the non-homogeneous to the homogeneous chemical stage.
Methods
Specifically, we introduce a novel concentration-dependent correction term into the reaction-diffusion equations (RDEs), dynamically rectifying reaction rates in high-concentration regions where classical kinetics fail. To justify the concentration-based kinetics at the microscopic scale, a Minimum Reaction Unit is defined to balance accuracy and cost. Furthermore, a Dual-grid strategy based on domain decomposition and buffer zone extension is implemented. This approach eliminates the reliance on PBCs and decouples the computational cost from incident particle numbers, ensuring both accuracy and computational efficiency via parallel processing.
Results
Sensitivity analysis confirms the robustness of the Dual-grid structure, where varying grid configurations yield consistent concentration profiles independent of domain decomposition, ensuring computational flexibility without compromising accuracy. Validation against Geant4-DNA simulations reveals that the framework successfully reconciles the kinetic discrepancies typical of classical RDEs in high-density track regions. Through simulations extending to 1 μs, the framework reproduces both the spatial diffusion and the temporal evolution of radicals, showing excellent consistency with Geant4-DNA.
Conclusion
This study provides a practical solution to the limitations of existing radiochemistry approaches, specifically the model discontinuity and boundary artifacts. By unifying the modeling of the chemical stage, our framework offers a unique capability to resolve the spatiotemporal interactions of radiolytic species generated by multiple consecutive pulses. This makes it an indispensable tool for simulating realistic radiotherapy delivery sequences and for investigating the mechanisms of the FLASH effect.