The capacity fade in organic redox flow batteries (ORFBs) is mainly caused by the degradation of their organic components. Predicting how these molecules evolve is fundamental to understanding their suitability for energy storage applications. In this project I aim to develop a modeling and simulation tool for the screening of degradation mechanisms of redox-active molecules.

The idea is to create a tool capable of exploring all the possible reaction pathways that transform the components of an electrolyte cell
(redox-active molecules, solvent, and occasionally salt) into other species. Key aspects will be the determination of the possible reactive sites on the molecules and Transition State search. It is pivotal to automate the processes and eliminate the trial and error approach in favor of an automated system while maintaining chemical intuition. This makes it possible to identify the reaction network without missing critical products and pathways.

For this purpose, I will employ a single-ended Transition State search method to explore the potential energy surface (PES), through an algorithm implemented in the Amsterdam Simulation Suite (AMS). The idea is to use semi-empirical methods (such as DFTB) to efficiently evaluate reaction pathways and then use more accurate methods (DFT) to refine the results. As the dataset expands, machine learning techniques may be integrated to accelerate predictions, enhancing the efficiency of degradation analyses in RFBs