Ionic liquids are promising solvents for capturing CO2 and hydrogen sulfide from industrial gas streams, but engineers need accurate thermodynamic models to predict how much gas will dissolve under real conditions. A new open-access study in Discover Chemistry provides the largest curated database of binary interaction parameters for 22 CO2-ionic liquid and 15 H2S-ionic liquid systems, covering over 2,900 data points across wide temperature and pressure ranges. The researchers fitted Peng-Robinson and Soave-Redlich-Kwong equations of state with temperature-dependent interaction parameters, and developed two generalized estimators that can predict parameters for unmeasured ionic liquids. One uses property-based ridge regression, the other uses cation and anion fragments. Cross-validation shows the fragment model cuts solubility prediction errors significantly, especially for H2S, and the work includes machine-readable tables and a reproducible Python workflow. This gives process engineers a practical toolbox for screening ionic liquid solvents in carbon capture and acid gas removal, without waiting for new experimental data. The study also warns that ignoring the co-volume interaction term can multiply solubility errors by more than four, so the new parameters are a meaningful step toward commercial deployment.
