Comparing qsar models of ionic liquid toxicity for different microorganisms to design environmentally friendly ionic liquids and elucidate mechanisms of toxicity

The increasing interest in ionic liquids (ILs) as alternative solvents for biocatalytic and industrial applications is limited by concerns regarding their potential toxicity. In this study, we compiled one of the most comprehensive ecotoxicity datasets to date for Vibrio fischeri, Daphnia magna, and Pseudokirchneriella subcapitata. Using Partial Least Squares (PLS)-based Quantitative Structure-Activity Relationship (QSAR) models, we established predictive correlations between IL structural features and ecotoxicity. The models demonstrated high robustness and predictive power, enabling identification of key molecular descriptors driving toxicity. Results consistently indicate that longer alkyl side chains in cations, higher hydrophobicity, and aromatic ring structures are associated with increased toxicity, whereas the incorporation of heteroatoms (O, OH, CN) or non-aromatic rings reduces toxicity. Anions generally exert a lower influence compared to cations, although certain highly fluorinated species substantially increase toxicity. The comparative analysis across microorganisms highlights both shared and species-specific structural determinants. These findings provide valuable insights into IL toxicity mechanisms, particularly the role of membrane accumulation and disruption, and establish design rules for the development of safer, environmentally benign ionic liquids suitable for green chemistry and biotechnological processes.