The second research field aims the examination of the dynamics of charge transfer processes using quantum molecular dynamics simulations. We model the prototype models of both of the most fundamental charge transfer reactions, electron transfer and proton transfer processes. Our model system for studying the basic molecular aspects of electron transfer is the hydrated electron, while it is protonated water clusters for proton transfer.

My first research topic is of methodical nature. To model the temporal, dynamical behavior of atoms and molecules precisely, it is of fundamental importance to describe the interactions between these particles. We employ machine learning techniques for this task, in particular potential energy modeling using artificial neural networks. We perform high quality ab initio computations on the investigated systems and these computations are reproduced via neural networks. The big challenge on the field is to generate and fit such a potential energy surface that is independent of the system size.