(664c) An Improved Spectroscopic Method for Determination of Association/Solvation Parameters Used in Process Models

Process modeling of systems continues to be challenging when association and solvation are present. Accurate modeling of mixtures with such species will become increasingly important because renewable chemicals, fuel intermediates and biopolymers typically associate via hydrogen bonding. Equations of state such as SAFT, CPA and ESD have gained popularity by explicitly representing hydrogen bonding. In a previous study, we have developed an activity-coefficient thermodynamic model which incorporates Wertheim’s perturbation theory (TPT-1). However, association parameters used in literature differ significantly depending on the type of data used to derive them. For example, the parameter for the self-association of methanol can vary by up to a factor of four.

This work focuses on the determination of the Wertheim association parameters using spectroscopic measurements such as FTIR. While this approach is not new, it is complicated by the challenges involved in quantifying the associating clusters such as alcohol monomers and dimers. Issues persist in curve fitting for evaluating the spectra where absorbances are highly concentration and temperature dependent. More importantly, the significant variation in the IR molar extinction coefficient across the broad hydroxyl band has been ignored in most studies and thus hindered accurate quantification of association. To address these challenges, we have developed a method based on quantum calculations for scaling experimental FTIR spectra such that more meaningful extraction of model-relevant thermodynamic constants is possible. The parameters acquired from our methods are compared to those obtained from other studies and used to model the phase equilibria of binary alcohol-alkane mixtures. The results demonstrate the capabilities of the activity coefficient model and further prove the importance of accurately capturing the hydrogen bonding phenomena in thermodynamic modeling.