The role of quantum-chemical descriptors and sigma-profile overlapping of different co-solvents on tuning ethanolic extraction of soybean oil

Abstract

Co-solvents can significantly improve ethanol efficiency in vegetable oil extraction, making their use a viable, cost-effective, and sustainable alternative to conventional solvents. The challenge, however, is to understand the complex interactions between the solvents and how this can affect the yield and quality of the extracted oil. In this study, the COnductor-like Screening Model with Segment Activity Coefficient and Hansen solubility parameters were combined and used to optimize soybean oil extraction using ethanol along with seven different co-solvents. In parallel, an experimental approach was carried out to verify the accuracy of the models. Both theoretical and empirical results indicated that the oil extraction yield decreased as the co-solvent polarity increased. The best performance in solubilizing non-polar lipids was observed for the mixture of 10 wt% of p-cymene, a non-polar solvent, with 90 wt% ethanol, reaching a yield of 24.35 g/100 g. According to the thermodynamic analysis, the effects of polarity weigh heavily on the standard entropy, which rules the driving force of mass transfer during the extraction. The interaction mechanisms were carefully investigated employing quantum mechanical calculations, revealing how the differences in chemical potential (μ), hardness (η), and electrophilicity (ω) underlie the macroscopic results and rule the efficiency obtained for different solvent ratios and combinations.