Ammonia pollution currently wreaks devastating effects on watersheds and coasts, causing dead zones and toxic blooms that poison water sources and aquatic life. To combat this, we propose an automated methodology for catalyst design to skirt the time consumption and cost of current techniques. This method will then design catalysts for conversion of ammonia to nitrogen gas to remove the pollutants from watersheds. Using inverse design to scan energy and free energy values along oxidation and dehydrogenation pathways, lowest energy catalytic pathways are determined and used to select the best performing catalysts of a given structure. Machine learning is used to predict these energies from the training data to span a larger ligand combination space, and Hammett parameters are used to simplify the complex space of molecule description. We propose that this methodology is generalizable to similar catalytic design problems.

Artificial intelligence for catalyst design: Removal of ammonia pollution from water

2019 Spring National Meeting

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