Ab Initio Prediction of Adsorption Isotherms for Gas Mixtures by Grand Canonical Monte Carlo Simulations on a Lattice of Sites.

Gibbs free energies of adsorption on individual sites and the lateral (adsorbate-adsorbate) interaction energies are obtained from quantum chemical ab initio methods and molecular statistics. They define a Grand Canonical Monte Carlo (GCMC) Hamiltonian for simulations of gas mixtures on a lattice of adsorption sites. Coadsorption of CO2 and CH4 at Mg2+ sites in the pores of the metal-organic framework CPO-27-Mg (Mg-MOF-74) is studied as an example. Simulations with different approximations as made in widely used coadsorption models such as the ideal adsorbed solution theory (IAST) show their limitations in describing adsorption selectivities for binary mixtures.