Comparing QSAR Models of Ionic Liquid Toxicity for Different Microorganisms to Design Environmentally Friendly Ionic Liquids and Elucidate Mechanisms of Toxicity

Abstract

Ionic liquids (ILs) have attracted considerable attention as promising alternatives to conventional organic solvents because of their negligible vapor pressure, high thermal stability, and tunable physicochemical properties. However, their widespread implementation requires a comprehensive understanding of their potential environmental impacts. This work presents a comparative quantitative structure–activity relationship (QSAR) study aimed at predicting the toxicity of ionic liquids toward three representative aquatic organisms, Vibrio fischeri, Daphnia magna, and Pseudokirchneriella subcapitata. Robust predictive models were developed using molecular descriptors that describe hydrophobicity, molecular size, electronic distribution, and structural characteristics of both cations and anions. The comparison of the models reveals common molecular features governing ecotoxicity while identifying species-specific sensitivity patterns. Hydrophobicity and alkyl chain length were identified as the dominant factors controlling toxicity, supporting the hypothesis that membrane disruption is one of the principal mechanisms of toxic action. The proposed models provide excellent predictive capability while offering mechanistic insights that facilitate the rational design of safer ionic liquids. Beyond toxicity prediction, this approach establishes practical molecular design guidelines that enable the development of environmentally benign ionic liquids without compromising their functional performance. The integration of computational chemistry, artificial intelligence, and green chemistry principles represents an efficient strategy to accelerate the development of sustainable solvents, reducing experimental effort, minimizing costs, and promoting the implementation of Safe-and-Sustainable-by-Design (SSbD) concepts for future industrial applications.