Prediction of Greenhouse Gas Solubility in Eutectic Solvents Using COSMO-RS

Abstract

Over the past few years, eutectic solvents (ESs) have drawn the scientific community's attention because they are usually more environmentally friendly than traditional organic solvents. One of the applications of ESs is in the gas capture field, where they are considered promising absorbers to replace amine- (MEA, DEA, or MDEA processes), methanol- (Rectisol process), dimethyl ethers of polyethylene glycol- (Selexol process), N-methyl-2-pyrrolidone- (Purisol process), propylene carbonate- (Fluor solvent process), or morpholine-based (Morphysorb process) solvents on CO₂ capture from the atmosphere. Although several studies have reported experimental gas solubility data in ESs, especially for CO₂, only a few existing options are covered. In fact, resorting to experimental methods to obtain the solubility data seems unfeasible considering the vast number of possible eutectic mixtures. Therewith, theoretical predictions of gas solubility in ESs are valuable for the fast pre-screening of prospective solvents. In this work, the ability of the thermodynamic model COSMO-RS to represent solubility data of CO₂, CH₄, and H₂S in 17 choline chloride-based (ChCl) ESs was evaluated. The experimental data were collected from the literature at different molar ratios, at 298.15 K or 313.15 K, and in the pressure range from 1 to 125 bar. COSMO-RS offers a qualitative description of these gases' solubility, which was expected due to the model's fully predictive character. To improve the CO₂ and CH₄ solubility data description, a temperature–pressure-dependent correction was applied to the COSMO-RS predictions for these gases, offering a global average relative deviation of 15%.