A methodology to estimate transport diffusivities in 'single-file' permeation through zeolite membranes using molecular simulations

To investigate the influence of 'single-file' diffusion on nonequilibrium transport phenomena inside zeolite nanopores, we calculated permeate velocities of guest species in AFI-type AIPO(4)-5 and MFI-type silicalite membranes, using a grand canonical ensemble molecular dynamics (GCMD) simulation. We chose as examples He/CH4 and CH4/SF6 mixtures through AIPO(4)-5, and CH4/CF4 mixtures through silicalite. In AIPO(4)-5 systems, He permeates much faster than CH4 in an He/CH4 mixture, whereas CH4 moves at nearly the same speed as SF, in the CH4/SF6 mixture. In CH4/CF4 mixtures in silicalite, the permeate velocity of each component becomes close to each other, compared with that of pure gas. These results suggest that 'single-file' permeation occurs where one guest species cannot overtake the other inside nanopores under a concentration gradient, especially if at least one component has a molecular diameter similar to the host pore size. Moreover, the calculated separation factor for CH4/CF4 mixtures through silicalite showed a strong correlation with the permeate velocities of the guest species, suggesting the influence of 'single-file' permeation on the absolute values of the separation factor. Particularly, it is concluded that the decrease in CH4 permeate velocities directly leads to the decrease in the separation factors, compared with ideal separation factors. Finally, using these data from GCMD simulations and other molecular simulation techniques, we propose a methodology to predict transport diffusivities in binary mixtures. In this paper, the transport diffusivities of CH4/CF4 mixtures in silicalite were calculated based on Maxwell-Stefan (MS) theory. Further efforts to achieve the systematic estimation or these transport diffusivities will lead to the prediction of multicomponent separation factors using only single-component data.