Ultrafast Realization of Ionic Liquids with Excellent CO2 Absorption: A trinity study of machine learning, synthesis, and precision measurement

Efficient CO2 capture is indispensable for achieving a carbon-neutral society while maintaining a high quality of life. Since the discovery that ionic liquids (ILs) can absorb CO2, various solvents composed of molecular ions have been developed and their CO2 solubility has been studied. However, it is challenging to optimize these materials to realize targeted properties as the number of candidate ion combinations for designing novel ILs is of the order of 1018. In this study, electronic- structure informatics was applied as an interdisciplinary approach to quantum chemistry calculations, and combined with machine learning to search 402,114 IL candidates to identify those with better CO2 solubility than known materials. Guided by the machine-learning results, trihexyl(tetradecyl)phosphonium perfluorooctanesulfonate was synthesized and it was experimentally confirmed that this IL has higher CO2 solubility than trihexyl(tetradecyl)phosphonium bis(trifluoromethanesulfonyl)amide, which is the previous best IL for CO2 absorption. The method developed in this study could be transferable to gas-absorbing liquids in general, such as deep eutectic solvents (hydrogen-bonded mixed organic solvents in a broad sense), which also have numerous practical applications. Therefore, we believe that our method for developing functional liquids will significantly contribute to the development of a carbon-neutral society.