Molecular Simulation Studies of the Diffusion of Methane, Ethane, Propane, and Propylene in ZIF-8

Abstract

ZIF-8 is a strong candidate for propane/propylene separation, which is regarded as one of the most industrially demanding. Molecular simulation of this separation must account for the flexibility of the structure, which enables the adsorption and diffusion of molecules with kinetic diameter larger than the apertures of the pores. Moreover, this simulation requires modeling subtle changes since the strong sieving effect upon the mixture depends on the very small differences between propane and propylene molecular sizes (∼0.2 Å). In this work, a new force-field for the ZIF-8 structure has been developed from DFT calculations in simplified structures. The new parameter set reproduces structural properties in very good agreement with the experimental measurements reported in literature. Molecular dynamics simulations and the Widom test particle insertion method were then employed for the calculation of diffusivities, activation energies and adsorption properties of propane and propylene. The results are in agreement with experiments and demonstrate that the sieving of such a mixture is a kinetic driven separation process.

New development in flow‐through pressure‐swing frequency response method for mass‐transfer study: Ethane in ZIF‐8

“…Corrected diffusivities from FR and IRM have similar trend, and show a slight minimum between 1 and 2 bar (∼50% coverage). The extracted corrected diffusivities at low coverage are in good agreement with the D 0 value of 8.8 × 10 −12 m 2 /s obtained using membrane permeation technique at 35°C and the self‐diffusivity of 1.0 × 10 −11 m 2 / s from molecular simulation . The transport diffusivities from IRM are more or less constant at low loadings, but have a strong increase in the range above 3.5 molecules per cage, corresponding to pressure above 1 bar.…”

Mechanical Control of the Kinetic Propylene/Propane Separation by Zeolitic Imidazolate Framework‐8

“…[8][9][10] In particular, the zeolitic imidazolate framework-8 (ZIF-8) was revealed as ap romising MOF for such ak inetic control of the separation. [11][12][13][14] Many approaches have recently been considered, such as linker exchange,metal substitution, and configuration optimization, to enhance the C 3 H 6 /C 3 H 8 diffusion selectivity achieved by ZIF-8 derivatives. [11,[15][16][17][18][19] Boosting the level of performance of ZIF-8 for such challenging separations is atopical area of research, since any improvement in terms of selectivity is expected to lead to asignificant advantage from an economic standpoint.…”

“…As afurther stage,the diffusivities of propylene and propane in ZIF-8 with the loading of five molecules/unit cell (corresponding roughly to the experimental half-saturation) were computed at 1bar. [13] Since these two hydrocarbons show slow diffusion rates,M Ds imulations were performed at high temperatures and their self-diffusivities (D s )a tr oom temperature were assessed from the Arrhenius representation (see Figure S7). [13] Ther esulting D s values for C 3 H 6 and C 3 H 8 are 1.97 10 À12 m 2 s À1 and 1.26 10 À13 m 2 s À1 , respectively,a t3 00 K. These simulated data are within the same range as available calculated values.…”

“…[13] Since these two hydrocarbons show slow diffusion rates,M Ds imulations were performed at high temperatures and their self-diffusivities (D s )a tr oom temperature were assessed from the Arrhenius representation (see Figure S7). [13] Ther esulting D s values for C 3 H 6 and C 3 H 8 are 1.97 10 À12 m 2 s À1 and 1.26 10 À13 m 2 s À1 , respectively,a t3 00 K. These simulated data are within the same range as available calculated values. [26] Moreover,o ur simulated D s values are in excellent agreement with the corresponding experimental data for C 3 H 6 ,a lthough they slightly deviate for C 3 H 8 .…”

Partial hessian fitting for determining force constant parameters in molecular mechanics

“…This problem was resolved by Seminario, who proposed determining force constants via diagonalization of partial QM Hessian matrices in Cartesian coordinates . Indeed, the Seminario method allows fast parameterization and has been implemented and used by many research groups, especially for deriving force constant parameters required for MM simulations of metalloproteins, nanosystems, a metal‐organic framework (MOF), an ionic liquid, and so on. Burger et al developed another efficient approach that makes use of a projected Hessian for a single internal coordinate to determine k b and k a , thereby circumventing the abovementioned ambiguity problem …”