Modeling vapor-liquid-liquid-solid equilibrium for acetone-water-salt system

Abstract

A compilation of available experimental data for acetone-water mixtures with the reciprocal salt system Na+, K+ || Cl−, SO42− is presented. Significant inconsistencies among experimental data are pointed out. New freezing point measurements are reported for the binary acetone-water system at 12 different compositions. UNIQUAC parameters are determined on the basis of the available data from literature. Modeling results are presented. Vapor-liquid, liquid-liquid, and solid-liquid equilibria together with thermal properties are reproduced well by the model using only 14 parameters. The major drawback of the model is that the calculated liquid-liquid equilibrium regions of systems with KCl and NaCl are larger than the experimentally determined regions. The model is valid in the temperature range from −16 to 100 °C.

Comparison of models for the relative static permittivity with the e-CPA equation of state

“…𝑚 𝑖 𝑚 𝑚𝑎𝑥 (19) where 𝑁 is the number of experimental points of the property, 𝛾 ± is the mean ionic activity coefficients (MIAC), 𝛷 is the osmotic coefficient and 𝜌 is the density. The superscripts 𝑒𝑥𝑝 and 𝑐𝑎𝑙𝑐 refer to experimental and calculated values respectively.…”

“…The literature regarding the dependencies of the relative static permittivity for e-EoS's or electrolyte activity coefficient models is divided, as models have been presented that include a salt concentration dependency [12][13][14][15] and others do not include such a dependency [16][17][18][19]. With both methods the models are capable of describing some properties well under some conditions.…”

Mixed Solvent Electrolyte Solutions: A Review and Calculations with the eSAFT-VR Mie Equation of State

“…The most recent version of the DTU approach is the extended UNIQUAC of Thomsen and co-workers. This version of the extended UNIQUAC model has been applied to numerous solvents and types of phase behavior (VLE, LLE, SLE, but not MIAC) in a series of studies. − In this model, water has been assumed to be the “only” solvent, permitting LLE and SLE calculations. This approach means that the activity coefficient is rationalized using the infinite dilution value in pure water, rather than in the mixed-solvent solution.…”

“…A linear, salt-free, volume-based mixing rule was used for RSP. The same standard state assumption is used, as in the work of Thomsen and collaborators in extended UNIQUAC, − where water is the only solvent, which makes LLE calculations straightforward. Good vapor pressure/VLE was obtained for water and other solvents, including mixed solvent VLE, while acceptable results were obtained for LLE (where the salt concentration in the organic phase is very small and hard to represent accurately).…”

“…The VLE correlations by Sander et al and Macedo et al are of similar accuracy, showing deviations below 0.06 bar and 0.02 in the vapor phase mole fractions. Slightly higher are the deviations for the electrolyte UNIFAC implementation of the Macedo et al model by Kikic et al One of the most successful models, the extended UNIQUAC by Thomsen and coworkers − (UNIQUAC + DH + pure water RSP even in mixed solvents) yields VLE with very high accuracy. The AAD in vapor phase mole fractions is 0.021 and 1690 Pa in bubble point pressure for ethanol and higher alcohols and 0.014 and 0.043 bar for methanol, respectively.…”