Machine learning-driven design of CO2 capture materials: A focus on progress and challenges for MOFs and ILs

Numerous studies have shown that the continuous rise in atmospheric CO 2 concentration is one of the main drivers of climate warming. Therefore, controlling CO 2 emissions has become a core issue that needs to be urgently addressed in global environmental governance. As a core technology in the greenhouse gas emission reduction pathway, carbon capture puts forward higher requirements for improving material performance and optimizing process efficiency. In recent years, with the rapid development of artificial intelligence technology, machine learning has been gradually applied to the screening and design of carbon capture materials, which has significantly accelerated the research and development process of new and efficient materials. This paper reviews the application of machine learning to two typical CO 2 capture materials, namely metal-organic frameworks (MOFs) and ionic liquids (ILs), with a focus on performance prediction and material design. The review covers ILs, MOFs, MOF membranes, and IL@MOF composites. The reviewed studies show that descriptor selection, data quality, and model choice strongly affect prediction accuracy and design efficiency, while hybrid and deep learning models often perform better for complex nonlinear problems. Finally, the article discusses current challenges, including insufficient descriptor standardization and limited model interpretability, and outlines future research directions.