Automated and high-throughput reactivity analysis in homogeneous catalysis: The deactivation complexity of Mn(I) hydrogenation catalysts

In this thesis, I highlighted a number of projects aimed at developing and testing new simulation methods for studying complex reactive systems, with a particular emphasis on simulation strategies based on the concept of bonding graphs. These mathematical structures provide useful tools for a variety of algorithms developed over the last few decades. Through automated analysis of exhaustive exploration trajectories, I have been able to capture serendipities that could escape the expert heuristics or otherwise needed expertise in different disciplines to be interpreted correctly. Such discoveries could range from very obvious one-step reactions that were just not “normally” considered to multistep complex reactions that were not imaginable to the expert. With automated reactivity screenings on in silico catalyst libraries, I have taken a big step towards “rational”catalyst design.