(211b) Molecular Simulation of Self- and Transport Diffusion of Mixtures in Metal-Organic Frameworks

Nanoporous materials are widely used as heterogeneous catalysts and in adsorption and membrane separations. Examples include zeolites and other molecular sieves, activated carbons, and clays. Recently, there has been tremendous activity to synthesize new nanoporous materials. In particular, one would like to design and synthesize materials having desired functional properties cite. A particularly interesting strategy that may ultimately allow materials design is to use mild synthesis conditions and well-defined, rigid building blocks that retain their integrity in the final material. These building blocks can be chosen so that they self-assemble into interesting porous structures, ranging from discrete molecular triangles, squares, and cages to extended amorphous or crystalline materials.

In many applications of nanoporous materials, the rate of molecular transport inside the pores plays a key role in the overall process. A membrane or adsorption separation process exploits differences in rates of diffusion and/or adsorbed-phase concentrations to differentiate between the different molecular species and separate them. For microporous materials (pores less than 20 Angstrom), a number of unusual diffusion effects are observed that have no counter-part in bulk phases. Because of this tight fit, diffusion properties of guest molecules in microporous materials can be quite sensitive to small differences between different host materials, and molecular-level modeling has, therefore, come to be a useful tool for gaining a better understanding of diffusion in nanoporous materials.

Molecular dynamics (MD) simulations in metal-organic frameworks (MOFs) were first performed by Sarkisov et al. and Skoulidas. Sarkisov et al. reported self-diffusion coefficients in MOF-5 for methane, n-pentane, n-hexane, n-heptane and cyclohexane. They were found to be of the same order of magnitude as in silicalite. Almost simultaneously, Skoulidas reported diffusion of argon in another MOF, Cu-BTC . This study probed the dependence of self- and transport diffusion over a wide range of pore loadings at room temperature. The two diffusivities differed by almost two orders of magnitude at high pore loadings. Skoulidas and Sholl continued with an extensive report on diffusion of light gases in several MOFs as a function of pore loading. They studied Ar, CH4, CO2, N2, and H2 diffusion in MOF-2, MOF-3, MOF-5, and Cu-BTC. Their results greatly expanded the range of MOFs for which data describing molecular diffusion is available. However, to date no simulation data is available on mixture diffusion in MOFs.

We present detailed molecular simulations of mixture diffusion in various MOF topologies: the IRMOFs of Yaghi and coworkers, Cu-BTC and some of the paddlewheel structures synthesized by the Hupp group. The self- and transport diffusivities are obtained from the Einstein equation using MD. We explore how confinement and topology influence the selectivity in a mixture. Adsorption isotherms and diffusion coefficients will be presented for binary mixtures of CO2/N2 as well as ternary mixtures of hexane/2-methylpentane/2,2-dimethylbutane. For reference we also computed the single component adsorption isotherms and diffusivities.