Multiscale Modeling of Molecular Sieve Membranes

Abstract

Gases permeating through nanopores experience selective adsorption with sub-diffusive dynamics. In this work we investigate how the physicochemical properties of nanoporous membranes influence adsorption dynamics of gases, employing multiscale molecular simulations: molecular dynamics (MD), grand canonical Monte Carlo (GCMC), and kinetic Monte Carlo (KMC) simulations. We report the existence of an optimum chain length of highest curvature, leading to low density, high porosity, and high gas uptake without compromising the membrane’s sorption selectivity [1]. The trajectory extending kinetic Monte Carlo (TEKMC) technique efficiently extends molecular trajectories from nanosecond to microsecond timescales and is utilized to evaluate the mixture diffusivities in binary mixtures to determine the permselectivity of gases in mixtures, to reliably evaluate the performance of nanoporous membranes for gas separations for industrial applications [2]. An open-source software package is built and made available for easy computation of long-time diffusion of small penetrants in nanoporous materials. The simulation data is adjoined with existing databases to train ML architectures allowing in silico prediction of gas adsorption isotherm with remarkable accuracy. The novel use of single and binary gas adsorption isotherms, inside molecular sieving membrane and metal organic frameworks, opens a new avenue for predicting complex adsorption processes for gas mixtures for future materials [3]. The work also provides insights into interfacial polymerization of thin-film composite microstructures [4] and explain gas permeability trends observed in experiments.

References:

[1] Dasgupta, S., Rajasekaran, M., Roy, P.K., Thakkar, F.M., Pathak, A.D., Ayappa, K.G. and Maiti, P.K., 2022. Influence of chain length on structural properties of carbon molecular sieving membranes and their effects on CO2, CH4 and N2 adsorption: A molecular simulation study. Journal of Membrane Science, 664, p.121044.

[2] Dasgupta, S., KS, A., Ayappa, K.G. and Maiti, P.K., 2023. Trajectory-Extending Kinetic Monte Carlo Simulations to Evaluate Pure and Gas Mixture Diffusivities through a Dense Polymeric Membrane. The Journal of Physical Chemistry B, 127(45), pp.9841-9849.

[3] Dasgupta, S., RS, A. and Maiti, P.K., “Unifying Mixed Gas Adsorption in Polymers and MOFs using Machine Learning”. Separation and Purification Technology 353 (2025): 128477.

[4] Song, W., Park, J., Dasgupta, S., Yao, C., Maroli, N., Behera, H., Yin, X., Acharya, D.P., Zhang, X., Doherty, C.M. and Maiti, P.K., 2022. Scalable pillar [5] arene-integrated poly (arylate-amide) molecular sieve membranes to separate light gases. Chemistry of Materials, 34(14), pp.6559-6567.