Unlocking the potential of ionic liquids in Anion-Pillared MOFs for enhanced He/H2 separation Performance: A combined computational screening and Machine learning study

Abstract

Efficient separation of helium (He) from hydrogen (H₂) remains a significant challenge in membrane-based separation processes. In this study, we constructed a comprehensive database of fluorine-rich ionic liquid@anion-pillared metal–organic frameworks (IL@APMOFs) and performed high-throughput computational screening (HTCS) to identify promising IL@APMOF membranes for He/H₂ separation. CatBoost was identified as the optimal machine learning (ML) algorithms, and using this model, we revealed that IL content (IL%) is the key factor governing the separation performance of these membranes. Based on this insight, we designed and optimized IL@APMOF membranes by fine-tuning the IL content. The results validated the ML-driven findings and demonstrated that this strategy produces IL@APMOF structures with significantly enhanced He/H₂ separation efficiency. This work not only provides a rational design strategy for the development of IL@APMOF membranes but also underscores the critical role of IL modification in advancing the discovery of high-performance MOF-based membrane materials.