From Molecular Simulations to Experimental Validation of Gas Permeance in Zeolite Membranes

Zeolite membranes are promising materials for gas separation applications due to their favorable permeability–selectivity balance. In this work, a multiscale approach is developed that integrates experimental measurements, mass transport modeling based on the Maxwell–Stefan formulation, and molecular simulations. Purely siliceous chabazite (Si-CHA) was selected as a model system owing to its high adsorption capacity and narrow pore openings (3.8 Å), which are well suited for separating CO 2 -containing mixtures. Permeation and separation performances were experimentally assessed for a series of industrially relevant gases, including CO 2, CH 4, N 2, O 2, H 2, Ar, and He. Adsorption isotherms were predicted using Grand Canonical Monte Carlo simulations, while diffusivities were computed via Molecular Dynamics. Notably, CH 4 exhibited anomalous diffusion behavior due to its kinetic diameter being comparable to the pore size of Si-CHA, leading to a pronounced molecular sieving effect. For the other gases, normal diffusion regimes were observed, allowing the incorporation of simulation-derived transport parameters into the Maxwell–Stefan model to predict membrane permeance and selectivity. Exploring the extent to which membrane properties such as permeance and selectivity can be predicted by molecular simulations and comparing these predictions with experimental data reveal the discrepancies between microscopic and macroscopic scales. Overall, this multiscale strategy provides critical insights into the transport mechanisms in zeolite membranes and emphasizes both the predictive potential and the current limitations of molecular simulations for the membrane-based gas separation process design.