Predicting the outcome of a catalytic reaction is of relevance to high-throughput screening of chemical reactions for industrial applications. Here we show that machine learning can be used to accurately predict the outcomes of catalytic reactions of small molecules on the surface of gold. Our machine learning models exploit a chemical space representation of the molecules in a database of reactions. A data-driven study of catalytic reactions on Au in this chemical space may provide insights into their behavior. We explore extending this method to the study of reactions catalyzed by droplets. Our approach provides a framework for systematically exploring a high-dimensional space of properties of catalytic reactions and serves as a non-traditional means to gaining chemical insight were human intuition might fail.

How can we use machine learning to study droplet catalysis?

2019 Fall National Meeting

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