Optimization of Microwave-Assisted Polyphenol Extraction From Ash Gourd Peel Waste Using RSM and ANN–GRA Techniques

Abstract

Ash gourd peel waste represents a valuable but underutilized source of polyphenolic compounds with potent antioxidant properties. In the present study, a two-stage optimization strategy was developed to maximize polyphenol recovery from ash gourd peel powder through microwave-assisted extraction (MAE). In the first stage, the solvent mixture was optimized using a simplex lattice design. The optimum solvent proportion was identified as 29.4% ethanol, 36.4% water, and 34.2% acetone. In the second stage, the optimized solvent was utilized for MAE, and experimental design and optimization were performed using the response surface methodology (RSM). The RSM model predicted optimum extraction conditions at 600 W microwave power, 39.70 mL/g solvent-to-solid ratio, and 3.2 min extraction time, yielding 24.60 mg GAE/g TPC and 74.58% AA. Additionally, the RSM technique revealed that microwave power was the most significant factor affecting the extraction process. In contrast, ANN architectures (3–15–1 for TPC and 3–16–1 for AA) exhibited superior prediction performance compared to RSM, as evidenced by higher coefficients of determination and lower prediction errors. The ANN–GRA approach identified slightly modified optimum conditions of 600 W microwave power, 40 mL/g solvent-to-solid ratio, and 4 min extraction time, resulting in enhanced TPC (26.18 mg GAE/g) and AA (80.23%). These findings confirm the effectiveness of ANN–GRA as a robust tool for modeling and optimizing polyphenol extraction from ash gourd peel waste.