The increased availability of open-source software, the rapid development of large datasets, and advances in machine learning methods have revolutionized the field of computational catalysis. While significant advances have been made for automating workflows solid catalysts (e.g., metals, alloys), similar developments within the zeolite community have been lacking. In this work, we will describe the new zeolite-focused software toolkit - Multiscale Atomistic Zeolite Simulation Environment (MAZE). This new interface builds on top of the Atomic Simulation Environment and facilitates zeolite calculations by greatly simplifying the steps needed to perform common zeolite workflows. The API usability is assessed with the criteria knowability, operability, and robustness and by demonstrating example workflows that include (1) exhaustive sampling of Al sites, (2) extraction, modification, and re-insertion of clusters, and (3) systematic investigation of Cu-speciation. Preliminary results related to (1) the development of a comprehensive zeolite catalyst database, and (2) machine learning-assisted screening of Cu zeolites for selective methane oxidation reactions will be discussed.

Simplifying computational catalysis workflows with the multiscale atomic zeolite simulation environment (MAZE)

ACS Spring 2022

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