Potential of group contribution methods for the prediction of phase equilibria and excess properties of complex mixtures*

Abstract: Reliable knowledge of the thermophysical properties of pure compounds and their mixtures in the whole composition and a wide temperature and pressure range is a vital prerequisite for computer-aided synthesis, design, and optimization of chemical processes. Knowledge of the various phase equilibria is most important for the development of thermal separation processes (but also for other applications, such as the design of multiphase reactors, the prediction of the fate of a chemical in the environment, etc.). Whereas 25 years ago, the main interest was directed to the development of predictive tools for vapor–liquid equilibria of subcritical compounds of similar size (ASOG, UNIFAC), 15 years later a proper description of the temperature dependence (excess enthalpies), the activity coefficients at infinite dilution, and solid–liquid equilibria of eutectic mixtures (including strong asymmetric systems) was achieved. After the combination with cubic equations of state [Soave–Redlich–Kwong (SRK), Peng–Robinson (PR)], the group contribution concept was extended to supercritical compounds [predictive SRK (PSRK)]. With the development of an adequate electrolyte model (LIFAC), the equation-of-state approach can even be used for systems with strong electrolytes. With the revision of the group interaction parameters, the extension of the parameter matrix (introduction of new structural groups, filling of parameter gaps), and the help of a large database (Dortmund Data Bank), the predicted results of group contribution methods were significantly improved and the range of applicability greatly extended. Furthermore, still-existing problems with the group contribution approach (proximity effects, etc.) were reduced. With the help of a volume-translated PR equation of state and application of temperature-dependent and improved mixing rules, the remaining weaknesses of group contribution equations of state (such as poor results for liquid densities, excess enthalpies, and the problems with asymmetric systems) were minimized.