High-throughput Screening of Catalysts for CO2 Reduction Reaction

Catalysis underpins modern society by enabling critical processes in agriculture, energy, and environmental protection. The discovery and rational design of advanced catalysts are therefore essential for accelerating scientific and technological progress. The integration of artificial intelligence (AI) is transforming catalyst research by enabling rapid materials discovery, property prediction, and the identification of structure–performance relationships. This project aims to develop a large language model (LLM)-driven framework for the high-throughput discovery of catalysts for the electrochemical carbon dioxide reduction reaction (CO₂RR). By integrating scientific literature, computational and experimental databases, and density functional theory (DFT) datasets, the LLM will extract chemical knowledge, identify structure–activity relationships, generate and rank promising catalyst candidates, and recommend optimal compositions and surface structures for targeted computational validation. By coupling the automated DFT workflows and machine learning models, the framework will accelerate catalyst screening by predicting activity, selectivity, stability, and reaction pathways, significantly reducing the time and computational cost required to identify efficient catalysts for the sustainable conversion of CO₂ into value-added fuels and chemicals.