Predicted Optimal Bifunctional Electrocatalysts for Both HER and OER Using Chalcogenide Heterostructures Based on Machine Learning Analysis of In Silico Quantum Mechanics Based High Throughput Screening.

Two-dimensional van der Waals heterostructure materials, particularly Transition Metal Dichalcogenides (TMDC), have proved to be excellent photoabsorbers for solar radiation, but performance for such electrocatalysis processes as water splitting to form H2 and O2 is not adequate. We propose that dramatically improved performance may be achieved by combining two independent TMDC while optimizing such descriptors as rotational angle, bond length, distance between layers, and the ratio of the bandgaps of two component materials can provide. In this paper we apply the Least Absolute Shrinkage and Selection Operator (LASSO) process of Artificial Intelligence incorporating these descriptors together with Quantum Mechanics (density functional theory) to predict novel structures with predicted superior performance. Our predicted best system is MoTe2/WTe2 with rotation of 300°, which is predicted to have an overpotential of 0.03 V for HER and 0.17 V for OER, dramatically improved over current electrocatalysts for water splitting.