Accelerating Discovery of High-Performance Electrocatalysts for Ammonia Oxidation Via Machine Learning

Electrooxidation of ammonia to dinitrogen plays a crucial role within the nitrogen transformation cycle. While platinum (Pt) based catalysts, specifically terminated with (100) facets, have shown promising activity. They still suffer from a large overpotential (~0.5 V) and surface deactivation¹. Various alloys such as PtIr, PtRh and NiCu have offered an approach to tackle these issues. However due to vast combination of metals, the material space consisting of bimetallic alloys can quickly become overwhelming to navigate through. Additionally finding good electrocatalysts which not only have a lower overpotential but are also resistant to surface poisoning can be very difficult in this material space. Building upon our previous work², machine learning models are built in order to screen through the bimetallic alloy material space.

Our recent work showed that the optimization of *NHx and *OH adsorption energies is key to find thermodynamically efficient electrocatalysts. Additionally it was also shown that *N adatoms could be a precursor to surface poisoning. Using both these insights we have developed machine learning models to search for materials with optimal *NHx and *OH adsorption energies. This allows us to suggest potentially efficient electrocatalysts for ammonia electrooxidation³.

(1) Katsounaros, I.; Figueiredo, M. C.; Calle-Vallejo, F.; Li, H.; Gewirth, A. A.; Markovic, N. M.; Koper, M. T. M. On the Mechanism of the Electrochemical Conversion of Ammonia to Dinitrogen on Pt(1 0 0) in Alkaline Environment. J. Catal. 2018, 359, 82–91.
(2) 1. H. S. Pillai, H. Xin, New Insights into Electrochemical Ammonia Oxidation on Pt(100) from First Principles. Ind. Eng. Chem. Res. 58, 10819–10828 (2019).

(3) Y. Li, X. Li, H. S. Pillai, J. Lattimer, N. Mohd Adli, S. Karakalos, M. Chen, L. Guo, H. Xu, J. Yang, D. Su, H. Xin, G. Wu, Ternary PtIrNi Catalysts for Efficient Electrochemical Ammonia Oxidation. ACS Catal. 10, 3945–3957 (2020).