Liquid Organic Hydrogen Carriers: High-throughput Screening of Homogeneous Catalysts

In the times of an ever-increasing rate of global warming and rapidly depleting fossil fuels, renewable sources of energy are attracting vast attention and numerous efforts have been directed towards achieving a hydrogen-based economy over the past few years. However, the biggest technical challenge so far has been the development of materials and the required infrastructure for efficient storage and transportation of hydrogen. To this end, liquid organic hydrogen carriers (LOHCs) have been extensively studied as they provide a safer alternative to storing high-purity hydrogen using the existing fuel infrastructure. However, commercial applications of LOHCs are only feasible when expensive, noble metal catalysts are substituted with inexpensive but equally efficient alternatives. In this work, we employ our group's cyberinfrastructure for the data-driven discovery and design of novel catalysts for LOHCs. We screen a library of homogeneous Ir-based pincer catalysts for the dehydrogenation of perhydro-N-ethyl carbazole. We develop a computational protocol to evaluate these catalysts based on thermodynamic parameters calculated using Density Functional Theory. Next, we use this data to train machine learning models for predicting the Gibbs free energies of the reactions and analyze hidden structure-property relationships in these systems.