(253r) Automating Workflows for Surface Science and Catalysis

Density functional theory is rapidly becoming integral to the investigation of materials properties at the atomic scale. In chemical engineering, DFT has been particularly successful in the simulation of chemical surfaces for catalysis and adsorption, having aided in explaining various phenomena related to catalytic activity and even predicted new catalysts which have been confirmed in experiment. While DFT-based tools have become significantly more advanced in recent years, most scientific workflows for determining catalytic activity using electronic structure calculations still require significant effort by users to manage individual DFT simulations, parse output, and analyze the resultant data.

In this work, we present recent progress in the development of a high-throughput workflow to yield adsorption energies of low-index facets of a given material. Using the Materials Project infrastructure and approach, we demonstrate that simple analytical tools for the generation of surfaces and placement of adsorbates can enable more efficient automation of DFT calculations. In addition, our workflow automates collection of results from electronic structure simulations and performs a preliminary analysis in order to construct a database of properties that can be compared to experiment and filtered to identify promising candidate materials for further study. Perhaps most importantly, our workflow is designed to process materials and adsorbates in such a way as to minimize the necessity for user intervention, even for complex materials, and thus enable a more simple approach to comprehensive analysis. Lastly, we demonstrate how data collected using our workflow can be leveraged into insights into applications in nanoparticle synthesis and shape prediction.