Machine learning-based exploration of the effect of promoters on transition metal carbide catalysts for hydrogen evolution reaction

Abstract

The development of cost-effective electrocatalysts to substitute expensive platinum group metals (PGMs) in catalyzing the hydrogen evolution reaction (HER) holds immense promise for advancing sustainable energy solutions. In recent years, machine learning (ML) techniques have opened up new possibilities for smart screening and prediction of efficient heterogeneous catalysts. ML methods offer powerful tools that can analyze large datasets, extract meaningful patterns and correlations, and make accurate predictions based on the learned information. We utilized combined density functional theory (DFT) with supervised machine learning methods to discover earth abundant, durable, low-cost, efficient, transition metal carbides (TMCs) based active heterogeneous catalyst for HER. The study utilizes a Kernel Ridge Regression (KRR) model that has been optimized to identify the catalyst that is most active towards HER. The study demonstrates that this approach enables the efficient screening of a vast array of catalysts, identifying the most active HER catalysts. Hence, our methodology offers an effective means to find heterogeneous catalysts suitable for diverse electrochemical reactions.