COSMO-SAC and QSPR combined models: A flexible and reliable strategy for screening the extraction efficiency of deep eutectic solvents

[Display omitted] •We propose a highly adaptable and dependable method for screening DESs.•The fundamental principles and operational steps of the screening method are elucidated.•The proposed method was applied for the extraction of oleocanthal from extra virgin olive oil.•Among the DESs screened, the choline chloride/phenol composition showed the most promising performance.•The established extraction method significantly shortens the extraction time and solvent consumption. At present, the composition of deep eutectic solvents (DESs) is usually screened by trial and error, which requires a lot of experimental time and reagent consumption. A hierarchical DESs screening method was proposed in this work and used to develop a fast, cost-effective, and pollution-free method to extract oleocanthal from extra virgin olive oil (EVOO). The extractive performance of 220 DESs were predicted by the thermodynamic model Conductor-like Screening Model for Segment Activity Coefficient (COSMO-SAC), while their melting point (Tm) and viscosity (η) were evaluated by the regressed solid–liquid equilibria (SLE) diagram and Quantitative Structure-Property Relationship (QSPR) models, respectively. Choline chloride (ChCl)/phenol was found to be the best DES candidate due to its high infinite dilution distribution ratio of oleocanthal (C∞=1.89g/g) and excellent physical properties (Tm=243.02 K and η=39.70cP). The screening results were confirmed by comparing the predicted thermophysical properties with the experimental ones and the extraction conditions were optimized. The extraction efficiency of oleocanthal was 59.54 μg/g, higher than the acetonitrile (19.90%) and conventional DES methods composed of ChCl, xylitol and water (21.94%), significantly reduced the solvent consumption and extraction duration. The flexible and reliable DESs screening models established in this paper can reduce the experimental workload, and provides a new idea for the efficient extraction of trace components, which has extensive reference significance.