Counter-Intuitive Molecular Sieving via Interaction-Regulated Gas Admission in Molecular Trapdoor Adsorbents for Advanced Gas Separation

The gas separation industry accounts for approximately 10–15% of global energy consumption. Developing energy-efficient adsorbents capable of distinguishing similar-sized molecules remains a grand challenge in chemical engineering. This project aims to establish a new paradigm of molecular sieving by revealing that the extent of door-keeping cation displacement—co-governed by the size and interaction strength of both guest molecules and cations—dictates selective gas admission in molecular trapdoor adsorbents. Building on PI Shang’s discovery of the molecular trapdoor mechanism, in which gas entry is regulated by interaction-driven cation motion rather than rigid pore size, this work transforms the mechanism into a predictive framework integrating size, interaction, and framework–cation environment as mechanistically synergistic design handles. Zeolites are selected as the model platform because their well-defined topologies, atomic-level cation positions, and exceptional thermal and structural stability enable rigorous quantification of cation displacement and adsorption energetics—conditions essential for establishing predictive understanding. Once validated, the mechanistic framework and AI-guided methodology developed here will be readily extendable to hybrid porous materials such as metal–organic frameworks (MOFs) and covalent–organic frameworks (COFs). Through four interlinked aims, the project will: (i) elucidate how molecular size and interaction strength jointly control selective admission in model zeolites (Ag-LTA and Na-LTA), revealing counterintuitive invertible and size-inverse sieving; (ii) identify adjacent cations as remote regulators of the activation barrier and threshold admission temperature; (iii) establish adsorption configuration as an orthogonal handle enabling same-size sieving in flexible zeolites such as PHI; and (iv) integrate these mechanistic insights into an AI-assisted predictive design framework, enabling the synthesis and validation of next-generation trapdoor zeolites for benchmark ‘hard’ separations—CO2/C2H2, C2H4/C2H6, C3H6/C3H8, N2/CH4, and Xe/Kr—where achieving sieving-level selectivity remains challenging. Combining in situ neutron and synchrotron diffraction, X-ray absorption spectroscopy, operando FTIR spectroscopy, molecular simulations, and AI-assisted data integration, the project will map how cation displacement, guest configuration, and energetics jointly determine selective gas admission. The outcomes will establish a unifying mechanistic foundation for counter-intuitive molecular sieving, enabling rational design of next-generation, energy-efficient trapdoor adsorbents. The resulting fundamental insights will not only advance adsorption-based gas separations but also inform related fields—gas storage, sensing, and catalysis—contributing to sustainable chemical manufacturing and global decarbonization.