Green Ultrasound-Assisted Extraction of Bioactive Phenolics From Quince (Cydonia oblonga) Leaves: Multi-Response Optimization, LC-MS/MS Profiling, and Comparison With Conventional Methods.

Quince (Cydonia oblonga Mill.) leaves are an abundant yet underutilized agricultural by-product that is rich in phenolic compounds and therefore represents a promising feedstock for value-added bioactive extracts. This study optimized the green ultrasound-assisted extraction (UAE) of bioactive compounds from quince leaves using a sequential statistical approach based on Plackett-Burman screening and Taguchi-based criteria importance through intercriteria correlation (CRITIC)-weighted grey relational analysis (GRA). Under the optimum conditions (solid-to-liquid ratio 3:50 g/mL, 50% (v/v) ethanol, and 75°C), the extract exhibited a total phenolic content (TPC) of 9210.44 ± 138.25 mg GAE L- 1, a total flavonoid content (TFC) of 6180.67 ± 94.25 mg CE L- 1 and a DPPH free radical-scavenging capacity of 76.53 ± 3.65 mmol TE L- 1. Triplicate confirmation experiments validated the model, with an overall prediction accuracy of 97.69%. Ethanol concentration exerted the greatest influence on the grey relational grade (36.28%). Under identical operating conditions, the optimized UAE provided the most balanced extraction performance. It yielded the highest TPC, antioxidant activity, anti-inflammatory activity, and α-amylase inhibition relative to maceration, magnetic stirring, hand-shaking, and reflux, while LC-MS/MS identified quinic and chlorogenic acids as the principal constituents. These results indicate that the UAE optimized through a Taguchi-based CRITIC-weighted GRA is an effective and sustainable strategy for producing bioactive-rich extracts from quince-leaf waste. PRACTICAL APPLICATIONS: This study provides a practical, statistically grounded route for converting quince (Cydonia oblonga Mill.) leaf waste, an abundant agro-industrial by-product, into phenolic-rich extracts using a green, food-grade ultrasound-assisted process. The sequential screening and multi-response optimization framework can be adopted by food, nutraceutical, and functional-ingredient manufacturers to maximize antioxidant, anti-inflammatory, and enzyme-inhibitory activities while minimizing solvent and energy use, supporting cost-effective and sustainable valorization of leaf biomass.