Precision of Radiation Chemistry Networks: Playing Jenga with Kinetic Models for Liquid-Phase Electron Microscopy.

Liquid-phase transmission electron microscopy (LP-TEM) is a powerful tool to gain unique insights into dynamics at the nanoscale. The electron probe, however, can induce significant beam effects that often alter observed phenomena such as radiolysis of the aqueous phase. The magnitude of beam-induced radiolysis can be assessed by means of radiation chemistry simulations potentially enabling quantitative application of LP-TEM. Unfortunately, the computational cost of these simulations scales with the amount of reactants regarded. To minimize the computational cost, while maintaining accurate predictions, we optimize the parameter space for the solution chemistry of aqueous systems in general and for diluted HAuCl solutions in particular. Our results indicate that sparsened kinetic models can accurately describe steady-state formation during LP-TEM and provide a handy prerequisite for efficient multidimensional modeling. We emphasize that the demonstrated workflow can be easily generalized to any kinetic model involving multiple reaction pathways.