Comparison of CP-PC-SAFT and CS-SAFT-VR-Mie in Predicting Fluid Phase Behavior in Systems of Phenolic Compounds, Aromatic Amines, Acetophenone, and Benzaldehyde

Abstract

This study compares the capabilities of the Critical Point-based Revision of PC-SAFT (CP-PC-SAFT) and the Corresponding States-based SAFT of Variable Range and Mie Potential (CS-SAFT-VR-Mie) to simultaneously predict VLE, LLE and critical data available for 74 binary systems of phenol, m-cresol, 2-methoxyphenol, 2-phenylethanol, 2-ethylphenol, pyridine, aniline, benzylamine, o-toluidine, acetophenone, and benzaldehyde with nitrogen, carbon monoxide, methane, ethane, propane, n-alkanes, isoalkanes, and other hydrocarbons. For both models, the universal k 12 values adjusted to the VLE data of the system propane(1)−phenol(2) were applied. It was found that these models yield comparable and, in most cases, accurate estimations of vapor pressures and the pure compound saturated liquid densities. However, CP-PC-SAFT is superior in predicting the high-pressure density and sound velocity data. CS-SAFT-VR-Mie is able to predict the VLE data fairly accurately in most of the considered cases. However, it establishes an unrealistic transitional behavior from Type I to Type II via Type VI, which seriously affects its predictions of LLE. CP-PC-SAFT exhibited better overall reliability. In particular, in most cases. it is superior in estimating VLE and yields reasonably accurate predictions of the upper critical solution temperatures. According to the global analysis performed in this study, these results could probably be attributed to the obeying of the pure compound T c and P c . At the same time, CP-PC-SAFT systematically underestimates the widths of the LLE phase splits and, as a result, yields inaccurate predictions of the alkane-rich phases.

Application of CP-PC-SAFT and PC-SAFT for Simultaneous Prediction of LLE and High-Pressure VLE with Universal k12 Values in Systems of Gases and n-Alkanes with N-Methyl-2-pyrrolidone and Some Substituted Aromatic Compounds

“…Although such discrepancies make it difficult to quantitatively compare experimental data with predictions made by the models under consideration, a qualitative assessment can still be made. In particular, it can be seen that similar to the previously discussed systems of n -alkanes and substituted aromatic compounds, the simplest version of CP-PC-SAFT that neglects polarity and association accurately predicts the compositions of the NMP-rich phases and substantially overpredicts the NMP concentrations in the n -alkane-rich ones (see Figure A). It is noteworthy that very similar predictions are made by PC-SAFT with the parameters of Razavi et al (Figure C), which also does not consider polarity and association.…”

“…In the preceding study, CP-PC-SAFT was applied for predicting phase equilibrium data available for systems of pyridine with n -alkanes and gases. − Figures and compare the previously reported predictions of these data by CP-PC-SAFT with universal k 12 = 0.023 obtained for the system propane–phenol and PC-SAFT parametrized Van Niekerk et al and Razavi et al with universal k 12 values currently obtained at the UCST of n -hexane–NMP (0.011 and 0.017, respectively).…”

“…Figure compares the available experimental data on the solubilities of gases in NMP at high pressures and their predictions by the selected models with the universal k 12 values. As discussed elsewhere, the solubilities of gases are defined by their critical temperatures. In particular, the least soluble gas forming the most asymmetric system is nitrogen with the lowest T c = 126.1 K, argon with T c = 150.86 K is more soluble than nitrogen and then comes methane ( T c = 190.564 K), and the highest solubility is exhibited by ethane ( T c = 305.32 K).…”

“…Still, the predictions of CP-PC-SAFT for LLE in the systems of n -alkanes and NMP along with high-pressure VLE in its systems with gases are examined. The same universal value of k 12 = 0.023 initially obtained for propane–phenol is applied. Thus, the entire predictive performance of the model is considered.…”

“…At the same time, the ability of EoS models to accurately describe a wide range of similar systems with the same universal values of binary parameters may provide evidence of their inherent robustness and reliability. In a previous study, predictive capabilities of the simplest polarity- and association-neglecting versions of critical point-based revision of PC-SAFT (CP-PC-SAFT) and the corresponding state-based SAFT of variable range and Mie potential (CS-SAFT-VR-Mie) were examined. For each model, a binary parameter k 12 was adjusted to the VLE data of a randomly selected propane–phenol system.…”