Optimization of Zanthoxylum bungeanum essential oil extraction based on response surface methodology and machine learning with activity evaluation.

This study extracted Zanthoxylum bungeanum essential oil via steam distillation, with the extraction process optimized using Response Surface Methodology (RSM) and machine learning techniques. The results indicated that the optimal conditions predicted by RSM were a distillation time of 92 min, an ultrasonic time of 25 min, and a liquid-solid ratio of 11.6 mL/g, yielding a predicted extraction rate of 2.77 % and an actual rate of 2.92 %. The Artificial Neural Network (ANN) model predicted optimal conditions as a distillation time of 117 min, an ultrasonic time of 30 min, and a liquid-solid ratio of 1:8, giving a predicted yield of 3.01 % and an actual yield of 3.06 %. The Random Forest model suggested conditions of 75 min, 30 min, and a ratio of 1:10, corresponding to a predicted yield of 2.99 % and an actual yield of 2.74 %. The chemical composition of the essential oil obtained under the optimal process was analyzed by GC-MS, identifying γ-terpinene (14.7294 %), 3-carene (14.1696 %), and linalool (13.574 %) as the major components. Antimicrobial assays demonstrated that the essential oil exhibited significant inhibitory effects against S.aureus and E.coli, indicating its potential as a natural antibacterial agent. Furthermore, antioxidant activity tests showed that at a concentration of 40 mg/mL, the ABTS⁺ radical scavenging rate reached 92.81 %. At 80 mg/mL, the scavenging rates for DPPH and hydroxyl radicals were 94.57 % and 68.33 %, respectively, demonstrating strong radical scavenging capacity. The essential oil also displayed considerable ferric ion reducing power.