Extension of CPA and SAFT to New Systems: Worked Examples and Guidelines

Abstract

Application of CPA and other association equations of state (EoS) to new associating systems (for which parameters are not available in the literature) includes the following three (often interconnected) stages:1. Choice of pure component parameters. 2. Establishment of the number of possible association sites and choice of an appropriate association scheme. 3. Cross-association and solvation schemes.Stage 2 is at first based on Tables 8.11 and 8.12 (Chapter 8). However, various association schemes are possible for a specific compound and, moreover, various sets of parameters of CPA/SAFT models can be obtained which equally well correlate the available vapor pressure and liquid density data. Experience has shown that, when possible, it is best to select the most appropriate parameter set for an associating compound using, in addition to pure compound properties, LLE data of the compound under investigation with an inert compound, e.g. n-alkane. For example, water-hexane LLE data can be used to determine the optimum parameters for water, among the various sets which represent the vapor pressure and liquid density data of water equally well.In this way, reliable parameters were estimated in the cases of water and glycols as many LLE data are available for these highly immiscible systems with alkanes. LLE data for partially miscible systems like methanol or ethanol with alkanes (for which VLE data are also available) are useful for the same purpose. There are families of compounds such as amines where LLE data are not available for mixtures with alkanes. Then, optimum sets of parameters can be estimated using, in addition to pure compound properties, other 'sensitive' data, e.g. VLE at low temperatures. Second, virial coefficients can also be used but their accuracy, especially at low temperatures, should be evaluated carefully.In this chapter, we will illustrate how association models like CPA and SAFT can be extended to 'new' compounds and systems, not previously studied, where the association sites/schemes and types of crossassociation/solvation are not known a priori. A working approach will be illustrated first for the case study of sulfolane, while other systems (aniline and phenols) will be briefly considered next. The underlying purpose of the chapter is to provide 'mechanisms and working tools' for extending association EoS to new compounds and mixtures without significant changes to the framework, e.g. without introducing new terms in the models for explicit accounting of the polarity and other effects beyond those already considered in the models (physical, chain and association terms).