Monte Carlo Simulation of Binary Mixtures Adsorbed on Heterogeneous Surfaces

Most of the solids used in adsorption processes are heterogeneous and most isotherm models are not derived to use information about the solid structure and energy heterogeneity. A possible alternative is to use molecular simulation techniques, which turn the thermodynamic model more microscopic. Then, starting from information on the solid structure, it is possible to develop appropriate solid models for each type of real system. Therefore, the objective of this work is to use the grand canonical Monte Carlo technique of molecular simulation for correlating pure-component data and predicting adsorption of multicomponent mixtures. To decrease the computational effort, we use a simplified solid model that is represented by a twodimensional square lattice. The solid heterogeneity is represented by a random distribution of different kinds of sites, characterized by distinct energy levels. The methodology developed by Cabral et al. (2003a) is used here to correlate pure components and for predicting binary mixtures of ethane, ethylene, isobutane, and carbon dioxide on zeolite 13X with good results.