Solubility of Electrolytes in Organic Solvents: Solvent-Specific Effects and Ion-Specific Effects

Salt solubility in organic solvents is of particular interest in industry, for example, for carbon capture and storage or utilization processes, battery technology, or biotechnology. Electrolyte thermodynamic models have been developed to reduce the experimental effort for the design of an electrolyte toward desired properties, for example, high solubility in an organic solvent. In this work, the solubility of nine salts (NaCl, KCl, CsCl, NaHCO3, KHCO3, CsHCO3, Na2CO3, K2CO3, and Cs2CO3) in the organic solvents methanol, ethanol, and N-methyl-2-pyrrolidone (NMP) was studied at temperatures between 288.15 and 348.15 K. These systems were chosen since the largest amount of experimental data points were available in order to ensure a broad set of data for high modeling accuracy. Experimental solubility data were collected from literature, and missing data were measured in this work by both ion-chromatography analysis and the all-gravimetric method. The thermodynamic solubility product KSP of the salts was determined at 298.15 K and 1 bar. These KSP values do NOT depend on the solvent; that is, once known, they can be used to predict the solubility in any solvent or solvent mixture. KSP requires that the solid form of the precipitating salt be the same in different organic solvents. Therefore, powder X-ray diffractometer measurements were carried out to investigate possible hydrate formation or solvate formation, and Karl Fischer measurements were used to validate the absence of water. The equation of state ePC-SAFT was applied to model salt solubility in organic solvents by accounting for concentration-dependent dielectric constants within Debye–Hückel theory and Born theory. The required KSP values of the salts were determined using experimental literature data on the salt solubility in water together with the corresponding mean ionic activity coefficients (MIACs) at saturation. The availability of KSP and the predicted MIACs allowed modeling of the salt solubility in organic solvents in excellent agreement with experimental data. Furthermore, solvent-specific analysis and ion-specific analysis revealed a non-intuitive behavior of salt solubility in the organic solvents. One example is that ion-specific effects of salt solubility in alcohols are not valid in other organic solvents such as NMP. Furthermore, in contrast to aqueous solutions, salt solubility in organic solvents does not depend linearly on the cation size. Thus, experimental rules of thumb cannot be applied, and the experimental effort to screen salt solubility in organic solvents can only be significantly reduced by theoretical approaches such as ePC-SAFT.