Combined Machine Learning and Computational Chemistry Guided Discovery of Chevrel Phases for Electrocatalytic CO2 Reduction

点击翻译按钮获取中文摘要

Emission of carbon dioxide into the atmosphere is the major driver of climate change, and the path to a sustainable future will rely heavily on removing carbon dioxide from the air and either storing it or converting it into fuels or other valuable chemicals. One promising route for accomplishing this is to use electricity to drive the reaction of carbon dioxide with water to produce new chemicals. This reaction can occur on the surfaces of various materials with appropriate catalytic properties. Recently, an interesting family of materials known as Chevrels were shown to convert carbon dioxide to fuels. However, despite these promising initial results, this family of materials remains relatively unstudied and the efficiency of this reaction still needs substantial improvement to become economical. The objective of this work is to identify new Chevrel materials of the vast number of possible Chevrels that are capable of effectively converting carbon dioxide into valuable products. Identification of superior materials for this reaction could provide a major step towards reducing the level of carbon dioxide in the atmosphere and transitioning towards a sustainable future.Electrocatalytic production of methanol and C1+ products (reduction products with 1 carbon atom) remains a significant materials discovery challenge due to the poor selectivity and/or high overpotentials of existing electrochemical CO2 reduction (eCO2R) catalysts. Intercalated Chevrels (MyMo6X8, M = metal, X = S, Se, Te) are a promising but relatively unexplored class of materials that, like perovskites, provide a highly tunable framework for materials design and discovery with a broad range of potential applications. Furthermore, they were recently demonstrated to produce methanol selectively from CO2, suggesting that intercalated Chevrel phase materials may also be a relatively unexplored class of promising electrocatalysts that can be tuned for catalytic performance. The objective of this project is to computationally analyze and guide the design and accelerated discovery of new Chevrel phase electrocatalysts for efficient and selective CO2 conversions to valuable products. The strategy for accomplishing this goal is to 1) use state-of-the art computational quantum modeling tools to determine the mechanism of eCO2R on Chevrel surfaces in solvent and under an applied bias and 2) develop machine learned descriptors of catalyst stability, selectivity, and activity that enable the rational, high-throughput discovery of new high-performance Chevrel electrocatalysts that employ earth-abundant elements for economically-competitive CO2 conversions to valuable products. This research aligns closely with the topic areas of interest to this program, including renewable energy related catalysis, electrocatalysis, closing the carbon cycle, conversion of CO2, new catalyst designs and materials, basic understanding of catalyst materials and mechanisms and advances in tools for computational catalysis.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.