Machine Learning for Gas Capture in Ionic Liquids: Current Status and Future Trends

Ionic liquids, as green gas solubility media, have great potential for applications in carbon capture, industrial waste gas purification, and other fields. However, the massive combination of anions and cations makes their screening extremely difficult. Machine learning can break through the bottleneck of traditional experiments and simulations and achieve high-throughput prediction of gas solubility in ionic liquids. This article provides a systematic review of the research progress of machine learning in predicting the gas solubility performance of ionic liquids. The classification and modeling process of machine learning, the construction and performance of machine learning prediction models for the solubility of gases such as CO2, H2S, NH3, SO2, N2O and others in ionic liquids were analyzed and summarized. The progress and existing problems of machine learning application for gas capture in ionic liquids and the future development direction are discussed, in order to provide assistance and theoretical reference for the directional design and industrial application of ionic liquids.