Comparisons of equation of state models for electrolytes: e-CPA and e-PPC-SAFT

Many different models have been published with the aim of describing electrolyte thermodynamics. In this study two successful Equations of State (EoS) are compared. The models investigated in this work are the electrolyte Cubic Plus Association (e-CPA) EoS and the electrolyte Polar Pertubed Chain Statistical Associating Fluid Theory (e-PPC-SAFT). The model terms are not changed from previous works, but they are reparameterized with various models for the relative static permittivity. The models are parameterized as similarly and fairly as possible by identifying key parameters in the models that are intended to describe the same physics. Several different sets of comparable parameters are investigated to test the models with the inclusion of various physics, and to avoid favoring one model due to the choice of adjustable parameters. The parameters of the models are optimized with an objective function containing the properties mean ionic activity coefficients (MIAC), osmotic coefficients and density in a wide temperature range of 273.15–473.15 K and they are tested for their ability to quantitatively describe these properties. The difference between e-CPA and e-PPC-SAFT is generally found to be relatively minor and the permittivity model and considered adjustable parameters are generally more important for the overall model performance. It is found that the permittivity models that only predict part of experimental decrease tend to perform the best for the properties MIAC, osmotic coefficients and density. It is also found that including both association parameters and dispersion/attractive parameters for ions is better than neglecting one type of parameters, even when the same number of adjustable parameters is included. To be able to capture the temperature trend of MIAC it is found to be important to include an additional temperature dependency, which in this work is a temperature dependent interaction parameter for the physical term. It is also found that it is important to include sufficiently strong ion association to be able to obtain the correct qualitative trend for individual ion activity coefficients.