Due to limited mechanistic insights and the complexity of catalyst-substrate interactions, catalyst design and optimization in transition metal catalysis are often based on chemical intuition and experimental trial-and-error. I will discuss our recent efforts towards a streamlined computational approach to understand and predict the reactivity and selectivity of various transition metal catalyzed reactions, including the functionalization of C-H bonds and olefins, as well as olefin polymerization reactions.<br/>Our computational studies aim to address two long-standing challenges in mechanistic studies — reducing the time to identify reaction mechanisms and predicting how different factors, such as steric, electronic, dispersion, and strain effects, affect reactivity and selectivity. We employed several new computational approaches, including the use of energy decomposition analysis (EDA) methods to analyze different types of non-covalent interactions between the transition metal catalyst and the substrate, modern statistical tools and machine learning algorithms for data analysis and catalyst prediction, conformational sampling techniques to model catalysts with flexible ligands, and <i>ab initio</i> molecular dynamics (AIMD) simulations to study reactions involving strong interactions with solvent molecules. Together, these effects are leading towards a robust catalyst design and prediction platform that is based on understanding of reaction mechanisms and origin of reactivity and selectivity. In collaborations with synthetic organic chemistry groups, we guided the catalyst development for the Cu-catalyzed olefin hydroamination and the Rh-catalyzed asymmetric allenic Pauson-Khand reactions.

Towards mechanistically guided design of transition metal catalyzed organic reactions

ACS Spring 2022

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.