Article Details

This research investigated the application of heat-assisted extraction (HAE) for the recovery of phenolic-rich antioxidants from Achyranthes aspera Linn leaves, employing Grey Wolf Optimizer (GWO) for multi-objective optimization and Response Surface Methodology (RSM) alongside Gaussian Process Regression (GPR) for predictive modelling. The experimental design was based on the central composite design (CCD) approach available in RSM. Key extraction parameters, including operating temperature (OT), solid-to-liquid ratio (S/L), and extraction time (ET), were optimized to maximize total polyphenol content (TPC), extract yield (EY), and antioxidant activity (AA). Both CCD-RSM and GPR with a Rational Quadratic  kernel (GPR-RQ) demonstrated high predictive accuracy for the extraction outcomes, with CCD-RSM exhibiting slightly superior performance within the defined experimental domain and was therefore integrated with GWO for optimization purposes. Validationexperiments under optimal conditions achieved a TPC of 36.2023 mg GAE/g db, EY of 18.2291%, and AA of 6.2178 µM AAE/g dw, with percentage relative standard deviations (% RSD) of 7.64%, 1.13%, and 7.46%, respectively, confirming excellent agreement between predicted and experimental results. The findings emphasize the effectiveness and simplicity of HAE as scalable extraction method while showcasing the robustness of GWO, CCD-RSM and GPR as advanced tools for optimization and modelling in bioactive compound extraction. This research provides key insights into the development of eco-friendly industrial methods for extracting phenolic antioxidants, with promising applications in the food, pharmaceutical, and cosmetic sectors.