Predicting the adsorption of gas mixtures: adsorbed solution theory versus classical density functional theory

Accurate prediction of the adsorption properties of fluid mixtures in equilibrium with surfaces and/or nanoporous structures is of considerable scientific and practical importance. Often, while the pure fluid adsorption isotherms are known for each component, those for the mixture are not. Using data from Monte Carlo simulations of model mixtures (including hydrogen and carbon dioxide) adsorbed in graphitic slit pores, for a range of pressures to 1000 bar, we compare theories for mixed adsorption which require pure fluid isotherm data as input. In particular, we develop and evaluate methods based on adsorbed solution theory (AST) and classical density functional theory (DFT). We find that a novel approximate DFT-based model is generally more accurate than AST methods in predicting the adsorption isotherms of mixtures of simple gases.