A generalized procedure for the prediction of multicomponent adsorption equilibria

Prediction of multicomponent adsorption equilibria has been investigated for several decades. While there are theories available to predict the adsorption behavior of ideal mixtures, there are few purely predictive theories to account for nonidealities in real systems. Most models available for dealing with nonidealities contain interaction parameters that must be obtained through correlation with binary-mixture data. However, as the number of components in a system grows, the number of parameters needed to be obtained increases exponentially. Here, a generalized procedure is proposed, as an extension of the predictive real adsorbed solution theory, for determining the parameters of any activity model, for any number of components, without correlation. This procedure is then combined with the adsorbed solution theory to predict the adsorption behavior of mixtures. As this method can be applied to any isotherm model and any activity model, it is referred to as the generalized predictive adsorbed solution theory.