Abstract
Deep eutectic solvents (DESs) are increasingly recognized as viable environmentally friendly alternatives to traditional organic solvents because of their low toxicity and exceptional dissolving capabilities. In this research, 12 DESs were synthesized and systematically integrated with ultrasonic-assisted extraction (UAE) to extract flavonoids from Scutellaria Radix. Comprehensive screening revealed that the optimal DES outperformed conventional solvents (water and 70 % ethanol) in extraction efficiency. Through statistical optimization, including single-factor experiments, response surface methodology, and an artificial neural network-genetic algorithm, we achieved remarkable extraction yields for six target flavonoids: scutellarin (3.88 ± 0.37), baicalin (137.67 ± 8.61), wogonoside (13.40 ± 0.58), baicalein (27.95 ± 1.12), wogonin (15.95 ± 1.48), and chrysin (1.81 ± 0.21 mg/g). These values represented significant improvements over conventional extraction methods. Subsequent enrichment processes yielded flavonoid recovery rates of 76.80-85.13 %, and demonstrated excellent DES recyclability over four consecutive extraction cycles. In sum, this work establishes a sustainable, high-efficiency platform for phytochemical extraction, offering substantial advances in sustainable natural product processing through intelligent solvent design and process optimization.