The need for efficient discovery of new materials and catalysts mandates that we identify smart ways to map out and explore transition metal chemical space. Computational high throughput screening with density functional theory (DFT) calculations has emerged as a powerful tool for the discovery of new materials. Within heterogeneous catalysis on bulk metal surfaces, linear free energy or scaling relationships further help simplify the computational prediction of new catalytic materials through relating kinetics to a single thermodynamic descriptor. However, challenges remain in applying these tools to single-site, open-shell transition metal catalysts. I will describe our development of a general purpose open source toolkit for inorganic molecular discovery, molSimplify, that utilizes a divide-and-conquer approach to generate precise geometries of new inorganic complexes for automated simulation with DFT or our trained neural networks. I will describe our recent extension to transition state structure generation for further acceleration of catalyst screening. These tools enable us to identify when scaling relations hold and when they break down, allowing us to discover new catalyst design rules for challenging reactions (e.g., direct methane-to-methanol conversion). I will describe our generalization of this concept as a step towards understanding when a descriptor-based approach is sufficient in single-site catalysis versus when characterization of the transition state is essential. Time permitting, I will describe how pairing these tools with machine learning models enable us to accelerate multi-objective catalyst design.

Discovering and breaking design rules in single-site catalysis with new computational tools

2019 Spring National Meeting

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