A Predictive Model for the Selective Separation of Aromatic/Non Aromatic Hydrocarbon Mixtures by Pervaporation Using Dense Homogenous Polymer Membranes

Separation of aromatic/non-aromatic mixtures by membrane based pervaportion process is usually analyzed by the solution-diffusion model which assumes that a permeating component is first dissolved in the membrane and then diffuses through the membrane due to its driving force. Therefore, the separate evaluation of sorption equilibria is necessary in order to determine the controlling step in overall mass transport in the membrane. In the present study, the pervaporative separation behavior of theses mixtures through dense homogeneous polymer membranes was modeled by a combination of FloryHuggins and Maxwell-Stefan model. The systems chosen to validate the model were the separation of benzene/n-hexane and benzene/cyclohexane mixtures in dense polyurethane (PU). The sorption of aromatic/aliphatic hydrocarbon mixtures was described by the classical Flory-Huggins thermodynamic model. The binary (polymer-component and component-component) interaction parameters were calculated using pure component uptake data. Using these parameters, the multi-components sorption performance was then predicted by the extended Flory-Huggins model. The results revealed that the aromatic compounds are preferentially adsorbed into polyurethane over the entire range of composition. Diffusion step was modeled by Maxwell-Stefan model with concentration dependent diffusion coefficients. The results showed a close agreement between the experimental data and the predicted values of the component fluxes and selectivity.