Modeling of aqueous electrolyte solutions with an equation of state
Advancing the chemical engineering fundamentals
Thermodynamics: Developments with SAFT EOS (T2-1c)
Keywords: electrolyte systems, modeling, equation of state, Monte Carlo simulation
Modeling of electrolyte systems is important for many industrial processes, e.g. for sour gas treatment, extractive distillation and suspension crystallization. Currently most models for electrolyte systems are restricted to the addition of an ion-ion interaction term to a non-electrolyte model which handles the short range interactions. However, besides the long range ion-ion interactions, the molecular structure of an electrolyte system is mainly determined by the polar interactions between the solvent molecules and by the strong electrostatic interactions between ions and solvent molecules. Altogether, this is only accounted for in the so-called non-primitive models, for example in the Henderson-Blum-Tani perturbation theory and in integral equation theories, which explicitly account for the molecular structure of the solvent. One example for an integral equation theory which fulfills this requirement is the non-primitive mean spherical approximation (np MSA) which was used in a simplified form together with the PC-SAFT equation of state to model aqueous electrolyte systems. When solving the np MSA, it delivers three contributions in terms of the Helmholtz energy to account for the ion-ion, the ion-dipole and the dipole-dipole interaction. Therefore, this model does not require a temperature and ion concentration dependent dielectric constant.
Using our theory we describe mean ionic activity coefficient and osmotic coefficient data at 298 K, and system pressures and liquid densities between 293 and 373 K of several aqueous electrolyte systems over the entire solubility range. However, because a simplification of the np MSA is used, namely the semirestricted np MSA only one ion diameter can be specified. This mean ionic diameter was adjusted for every aqueous electrolyte system to mean ionic activity coefficient and osmotic coefficient data at 298 K. Furthermore, the theory underestimates the solvation of the ions which can be seen by calculating the Gibbs enthalpy of solvation at infinite ion concentration. To compensate for this, a further solvation contribution is introduced. So in total two parameters, the mean ionic diameter and the additional solvation energy are adjusted for each aqueous electrolyte system. The liquid density and pressure plots over molality show that the model including the two temperature-independent parameters can be extrapolated up to 373 K.
Our work on modeling of electrolyte systems using the np MSA indicates that the ion-dipole interaction is underestimated by the theory. To study the ion-dipole interaction more deeply we currently work on Monte Carlo simulations of a model ion-dipole system, consisting of Lennard-Jones molecules with point charges and point dipoles. Using these results we aim at evaluating the np MSA more deeply. In long term, it is planned to use our molecular simulation data to develop a better ion-dipole contribution and replace the one from the np MSA. Some results regarding the molecular simulations of the model system will be presented, too.
Presented Tuesday 18, 09:25 to 09:45, in session Thermodynamics: Developments with SAFT EOS (T2-1c).
