The solubility of fluorinated refrigerants (FRs) in ionic liquids (ILs) plays a critical role in IL-based absorption separation processes for the mitigation of greenhouse gases. In this work, the thermodynamic performance of the fully predictive COSMO-RS model is comprehensively assessed using an extensive dataset comprising 3853 experimental FR-in-IL solubility data points covering hydrofluorocarbons (HFCs), hydrofluoroolefins (HFOs), and other FR types. The results indicate that COSMO-RS generally captures qualitative solubility trends across diverse refrigerant-IL combinations, though notable quantitative deviations exist for specific systems. An empirical linear correction strategy is applied to reduce systematic biases, achieving significant accuracy improvements particularly for systems exhibiting approximately linear prediction deviations. Based on the evaluated prediction capability, a modeling-driven screening of ILs for the capture of R-134a, a high-GWP refrigerant, is conducted based on absorption capacity (AC), desorption ease (DE), and critical physical properties. Three acetate-based ILs, namely P 4442 Ac, P 4441 Ac, and pmpyAc, are identified as promising candidates from a predictive perspective. Mechanistic insights derived from molecular simulations, including σ-profile and interaction energy analyses, reduced density gradient (RDG) visualization, and molecular dynamics (MD) calculations, reveal that the absorption behavior of R-134a in ILs arises from the combined effects of anion–refrigerant hydrogen-bonding interactions and cation-dependent van der Waals interactions. Overall, this study highlights the potential of predictive thermodynamic modeling combined with systematic screening as a useful framework for guiding IL selection in FR capture and greenhouse gas mitigation. • A comprehensive COSMO-RS evaluation is conducted with 3853 FR-in-IL solubility data • Simple linear correction is applied to greatly improve COSMO-RS predictive accuracy • Systematic screening is performed to identify top promising ILs for R-134a capture • Absorption mechanism is revealed by quantum chemistry calculation and MD simulation
Wáng et al. (Wed,) studied this question.