Quantifying Impact of Intrinsic Flexibility on Molecular Adsorption in Zeolites

Abstract

The majority of molecular simulations of adsorption in crystalline nanoporous materials such as zeolites and metal–organic frameworks (MOFs) are performed using a rigid framework. This assumption is made for convenience, as simulations with flexible frameworks require dedicated force fields and more computational power. Several recent studies of molecular adsorption in metal–organic frameworks have hinted that computed adsorption properties may, in many cases, be quite sensitive to the inclusion of framework flexibility associated with thermal vibration of framework atoms. It is unclear whether this conclusion is also appropriate for molecular adsorption in zeolites. To investigate this issue, we have used molecular simulation methods to quantify the impact of intrinsic framework flexibility on molecular adsorption in zeolites without volume change (ΔV = 0) and ignoring adsorbate-induced deformations. We investigated the adsorption properties (loading, Henry’s constant, and selectivity) of four molecules (CO2, CH4, N2, and butane) and two binary mixtures (CO2/N2 and butane/CH4) in different pure silica zeolite frameworks. We also performed a limited set of simulations for a cationic zeolite with Si:Al = 1. We find that in zeolites, the influence of framework flexibility on adsorption is small, particularly when compared with MOFs.

A Transferable Force Field for Predicting Adsorption and Diffusion of Hydrocarbons and Small Molecules in Silica Zeolites with Coupled-Cluster Accuracy

“…GCMC simulations were performed in rigid zeolite frameworks. Although including framework vibrations can be important in making accurate adsorption predictions in some MOFs, , these effects are small for zeolites. − Sodalite cages in LTA (ITQ-29) and other regions inaccessible to adsorbates were blocked. , Electrostatics were calculated using Ewald summation with a precision of 10 –6 and dispersion interactions were computed using a 12 Å cutoff for a truncated potential with a tail correction. If the lattice parameter was smaller than 24 Å in any direction, the framework was expanded until the cell was large enough to satisfy the minimum image convention.…”

Computational Screening of MOFs and Zeolites for Direct Air Capture of Carbon Dioxide under Humid Conditions

“…All Monte Carlo simulations were performed in a rigid framework. Use of a rigid framework has been shown to be accurate for zeolites, although in some cases, framework flexibility can play a role in making quantitative predictions for adsorption in MOFs. , A 12 Å cutoff was used for van der Waals interactions. Whenever a unit cell had a lattice parameter smaller than 24 Å, the cell was expanded enough to satisfy the minimum image convention.…”

Performance-Based Screening of Porous Materials for Carbon Capture

“…This type of flexibility occurs without any change in unit cell volume and is associated with effects such as thermal vibrations at equilibrium or presence of adsorbed molecules inside pores. Recent studies have shown that the importance of intrinsic flexibility arising from thermal vibrations varies considerably for MOFs depending on the adsorption property of interest. , For example, it was shown that intrinsic flexibility can be more important where pore sizes are comparable to the kinetic diameter of adsorbate molecules. , For zeolites, however, the effect of thermal framework vibrations on molecular adsorption was shown to be negligible and the rigid framework assumption can be reliably used in GCMC simulation for these materials . For simulation of flexible materials that undergo significant changes in unit cell volume, more advanced simulation methods such as the osmotic ensemble and Gibbs ensemble Monte Carlo should be used …”

“…206,208 For zeolites, however, the effect of thermal framework vibrations on molecular adsorption was shown to be negligible and the rigid framework assumption can be reliably used in GCMC simulation for these materials. 209 For simulation of flexible materials that undergo significant changes in unit cell volume, more advanced simulation methods such as the osmotic ensemble and Gibbs ensemble Monte Carlo should be used. 204 A schematic diagram of the GCMC workflow is shown in Figure 12.…”