A hybrid machine-learning model for accurate prediction of formation energies

Interfacial Structure and Electrolyte Effects in Electrochemical Catalysis

color:#212121">Abstract: Conversion of solar energy into storable fuels is essential to meet future global energy demands. Efficient, robust materials that are exclusively made of non-precious elements are imperative for a sustainable energy economy. I rationally designed first-row transition metal (hydr)oxide water oxidation nanocatalysts and realized them by pulsed-laser in liquids synthesis. My approach enabled atomistic level understanding and concomitant optimization of highly active, robust nickel–iron layered double hydroxide nanocatalysts for water oxidation in base. Structural analysis of these mixed-metal nanocatalysts provided evidence that water oxidation occurred at edge-site iron centers. We discovered that interlayer anions played key roles during turnover, as incorporating anions with different basicities tuned the catalytic performance of these materials. Our nanocatalysts were regenerated and most active in alkaline electrolyte in ambient air, where ubiquitous carbonate rapidly replaced other interlayer anions. And we gained structural and mechanistic insights from in-situ spectroscopy data: we identified a cis-dioxo iron(VI) reactive intermediate as the lowest-energy species before O–O bond formation during water oxidation catalysis.

Accelerating Material Discovery using a Machine Learning Surrogate Model