Stony Brook Researchers Use Clean Electricity and AI to Find Best Solvents for CO2 Conversion

Researchers at Stony Brook University have developed a computational process that uses clean electricity to identify the most effective liquid solvents for dissolving carbon dioxide. The method combines physics simulations and machine learning to screen 1.3 million candidate molecules, narrowing the field to six promising solvents never tested for this purpose before. These solvents, including five cyclic ethers and one nitrile, could improve the efficiency of converting CO2 into valuable products like fuels and chemicals. The team published their findings in Cell Reports Physical Science and made the data freely available through an open-access database called COSMIC. This approach reduces the time needed to find effective solvents from decades to a much shorter timeframe, supporting the development of scalable carbon utilization technology. The work highlights how computational tools can accelerate progress in emissions reduction and clean energy innovation.