A novel approach and mechanistic insight into the co-extraction of essential oil and pectin from Citrus aurantium L. var. amara Engl. based on machine learning and DFT calculations

A green and efficient approach for the co-extraction of essential oil and pectin from Citrus aurantium L. var. amara Engl. was achieved by combining deep eutectic solvents (DES) with steam distillation. The optimal technological parameters were precisely identified through single-factor, response surface, and machine learning, with the DES concentration of 11.2%, the liquid-solid ratio of 10.6 mL/g, and the distillation time of 56.0 min. Under these conditions, the essential oil yield was 1.33 ± 0.07%, and pectin yield was 14.50 ± 0.58%. This process significantly increased the essential oil yield, reduced the solvent dosage, and shortened the distillation time, demonstrating remarkable progressiveness. Meanwhile, the extraction mechanism was clarified by observing the changes in the surface appearance and crystallinity of the raw material during the co-extraction process, and by calculating the interactions at the molecular level via density functional theory. The results confirmed that the ability of DES to dissolve pectin was much greater than that of water, which facilitated the breakdown of cell walls and accelerated the outward diffusion of essential oil from within the plant. Furthermore, the compositions of essential oil and pectin were identified. This research presents an environmentally friendly and efficient strategy for extracting essential oil and pectin from plant flowers, promoting the sustainable growth of the plant extraction industry.