Fluid Phase Equilibria, Vol.472, 117-127, 2018
A classical density functional theory for vapor-liquid interfaces consistent with the heterosegmented group-contribution perturbed-chain polar statistical associating fluid theory
A Helmholtz energy functional consistent with the heterosegmented group-contribution perturbed chain polar statistical associating fluid theory (GC-PCP-SAFT) is developed in this study and is applied to predict interfacial properties of planar vapor-liquid interfaces. Predicted surface tensions of pure substances are in very good agreement with experimental data for systems of non-polar as well as polar compounds including biodiesel systems where the group-contribution approach proves especially advantageous. The average deviation of the proposed model from experimental data is only 5% for these substances. Short hydrogen-bonding compounds, such as methanol and ethanol, are not predicted convincingly well. For longer hydrogen-bonding molecules the results are satisfying and deviations decrease to values of 3-10% for the entire liquid-vapor region. The orientation of hydrogen-bonding molecules at the vapor-liquid interface is reproduced qualitatively by the proposed model. Surface tensions for mixtures (including mixtures with hydrogen-bonding components) are predicted surprisingly well, with average deviations from experimental data of 5%. For mixtures, transferable group-group interaction parameters are adjusted to binary vapor-liquid equilibria data. The description of vapor-liquid phase equilibria is thereby significantly improved, however, the improvement of predicted surface tensions is mild. Further, in order to compensate for a weakness of group-contribution equations of state, we propose a concept for individualizing the group-contribution approach for substances that are well characterized by experimental data. We introduce a component-specific parameter. This concept improves the description mostly of vapor pressures, especially for short and multi-functional molecules. The effect of this measure on predicted surface tensions, however, is very modest. (C) 2018 Published by Elsevier B.V.