Aqueous Choline Chloride-Urea DES-Promoted MAE Extraction ofBioactive Compounds from Otostegia limbata Twigs and ProcessOptimization Based on RSM and ANN

Introduction/Objective: The growing demand for environmentally responsible and resource-efficient processing has accelerated interest in green extraction technologies. This study aimed to develop and optimize a sustainable microwave-assisted extraction (MAE) method using natural deep eutectic solvents (NaDES) for the efficient recovery of antioxidant and enzymeinhibitory phytochemicals from twigs of Otostegia limbata, a medicinal plant of notable ethnopharmacological importance. Methods: A choline chloride–urea-based aqueous NaDES was employed as the extraction solvent. Key MAE parameters, including microwave power, solid–solvent ratio, extraction time, and NaDES concentration, were optimized using response surface methodology (RSM). Total phenolic content (TPC), total flavonoid content (TFC), DPPH radical scavenging activity (RSA), ironchelating activity (ICA), and α-amylase inhibitory activity (AIA) were used as response variables. Artificial neural network (ANN) modeling was applied to evaluate nonlinear relationships, and GC–MS analysis was conducted for phytochemical profiling. Results: Optimal extraction conditions were identified as 770 W microwave power, 30 mL/g solid–solvent ratio, 60-second extraction time, and 30% aqueous NaDES. Under these conditions, TPC (11.44 mg GAE/g DW), TFC (3.93 mg RE/g DW), RSA (2.23 mg AAE/g DW), ICA (52.69 mg EDTAE/g DW), and AIA (53.18%) were achieved. The experimental results closely matched the predicted values, confirming the model's reliability. ANN demonstrated slightly superior predictive accuracy compared to RSM. GC–MS analysis revealed key bioactive compounds, including hardwickiic acid, hardwickiic acid methyl ester, camphor, and thymol. Discussion: The combined MAE–NaDES system effectively enhanced extraction efficiency while reducing solvent toxicity and processing time. The improved predictive performance of ANN highlights its usefulness in modeling complex extraction systems. Conclusion: This study demonstrates that MAE coupled with NaDES is a rapid, efficient, and green extraction strategy for O. limbata, supporting its sustainable utilization in food, nutraceutical, and bioactive ingredient applications.