Thermodynamic Prediction of SO ₂ Absorption in Deep Eutectic Solvents Using the UNIFAC Model

Deep eutectic solvents (DESs) have emerged as promising candidates for SO2 capture due to their tunable structures and favorable physicochemical properties. In this study, the UNIFAC model is applied to describe SO2 absorption in DESs through regression of missing group interaction parameters using experimental data and COSMO-RS calculations. A total of 329 experimental solubility data points are collected to compute SO2 activity coefficients and regress UNIFAC parameters. To fill data gaps, COSMO-RS is employed to generate additional SO2 activity coefficient data. Following systematic evaluation, COSMO-RS predictions for 2-aminopyridine- and triethylene glycol-based DESs are calibrated and incorporated into the parameter regression. In total, 88 new group interaction parameters are obtained. The resulting UNIFAC parameter set enables reliable description of SO2 absorption behavior in different DESs, as validated against both experimental and calibrated COSMO-RS data. This work provides a practical thermodynamic framework for modeling SO2 solubility in complex solvent systems.