Improved Prediction of Phase Behaviors of Ionic Liquid Solutions with the Consideration of Directional Hydrogen Bonding Interactions

Abstract

The consideration of directional hydrogen bonding in the COSMO-SAC model based on the minima in the molecular electrostatic potential has been shown to significantly improve the description of solvation properties of a wide variety of associating fluids containing nonionic species. In this work, we report the use of this method for the prediction of thermodynamic properties and phase equilibria of ionic liquid solutions, including infinite dilution activity coefficients (IDAC), osmotic coefficients (ϕ), liquid–liquid equilibria (LLE), and vapor–liquid equilibria (VLE). When compared to the previous version of the COSMO-SAC model without the consideration of directional hydrogen bonding, the new method, which has a tendency giving more polar σ-profiles, improves the prediction of osmotic coefficients and VLE by 10% (26% AARD, 1372 data points) and 5% (32% AARD, 1645 data points), respectively. Comparable results are observed from the two methods in LLE [0.2 and 0.1 root mean square (rms) error for ionic liquid-rich and solvent-rich regions, respectively, with 2876 data points] and IDAC (0.71 rms error with 3555 data points) predictions.

Impact of extended long-range electrostatics on the correlation of liquid-liquid equilibria in aqueous ionic liquid systems

“…As experiments may be time-consuming and costly, one proper tool to evaluate properties of different compounds in mixtures is predictive thermodynamic modeling. When considering IL-based systems, their behavior has been accurately modeled using equations of state (EoS) [12][13][14][15][16][17][18][19][20], such as the Perturbed Chain based on Statistical Associating Fluid Theory (PC-SAFT) and Cubic Plus Association (CPA), or excess Gibbs free energy (G E ) models [21][22][23][24][25][26][27][28][29][30][31][32][33][34], such as the Non-Random Two-Liquid (NRTL) [31,[35][36][37][38], UNIversal QUAsi-Chemical (UNIQUAC) and UNIversal Function group Activity Coefficient (UNIFAC) [24,26,[39][40][41][42][43][44] or even the predictive Conductor-like Screening Model for Realistic Solvents (COSMO-RS) [29][30][31][32][33][45][46][47]…”

“…Recently, interest in modeling more complex, concentrated, non-aqueous electrolytes has been growing and many approaches stick to Debye-Hückel theory, showing its applicability [7,[17][18][19]27,31,32,34,44]. However, application in these systems, particularly when highly concentrated, requires consideration of the changing mixture properties, such as density, molar mass and, most importantly [55], the relative permittivity.…”

“…However, application in these systems, particularly when highly concentrated, requires consideration of the changing mixture properties, such as density, molar mass and, most importantly [55], the relative permittivity. Based on this, recent works have provided good arguments for the inclusion of a concentration dependent relative permittivity in electrolyte models [18,20,31,32,37,[56][57][58][59][60][61][62][63][64][65].…”

“…This extension of the conventional PDH equation has been combined with other short-ranged interaction models by several authors. Chang and Lin [31,32] presented its full derivation and coupled it with the COSMO-SAC model tailoring the long-range term to fit ILs in organic solvents. This approach was afterwards tested by Ganguly et al [41] with an eUNIQUAC model.…”

Computer-Aided Molecular Design of Double-Salt Ionic Liquid Solvents for Extractive Distillation with the COSMO-SAC and Genetic Algorithm

“…COSMO-SAC model is another method to calculate the activity coefficients of mixtures. The COSMO-SAC model was proposed by Lin et al , and developed by many authors. − The COSMO-based models (COSMO-SAC and COSMO-RS) are not as accurate as UNIFAC models, , but they can predict the activity coefficients of components in any solutions with known molecular structures. COSMO-SAC can be utilized in mixtures containing ILs.…”