Data-Driven Optimization of Polyphenol Recovery and Antioxidant Capacity from Medicinal Herbs Using Chemometrics and HPLC Profiling for Functional Food Applications

The optimization of bioactive compound extraction from medicinal herbs is critical for developing functional food ingredients with substantiated health benefits. This study employed response surface methodology (RSM) and partial least squares (PLS) regression to maximize polyphenol recovery and antioxidant capacity from five medicinal herbs (Helichrysum stoechas, Chelidonium majus, Mentha pulegium, Artemisia absinthium, and Adiantum capillus-veneris). A custom experimental design assessed the effects of herb identity, extraction technique, and solvent-to-solid ratio on total polyphenolic content (TPC), total flavonoid content (TFC), ferric reducing antioxidant power (FRAP), and DPPH radical scavenging activity. The PLS compromise optimum was identified for M. pulegium using 60% ethanol at 55 mL/g, yielding 37.54 ± 2.10 mg GAE/g dw TPC, 21.62 ± 1.15 mg RtE/g dw TFC, 334.38 ± 12.37 µmol AAE/g dw FRAP, and 262.67 ± 9.46 µmol AAE/g dw DPPH. HPLC-DAD profiling revealed 18 polyphenolic compounds (10.22 ± 0.34 mg/g dw), dominated by kaempferol-3-O-β-rutinoside, protocatechuic acid, and luteolin-7-O-glucoside. These compounds contribute complementary mechanisms: protocatechuic acid modulates oxidative and antioxidant pathways, kaempferol-3-O-β-rutinoside exerts cardioprotective and anti-inflammatory effects via VEGF-C binding, and luteolin-7-O-glucoside suppresses NF-κB-mediated inflammatory signaling. Principal component analysis (PCA) explained 84.8% of variance, clearly separating optimized from non-optimized extracts, while PLS confirmed strong correlations between specific phenolics and antioxidant indices. Overall, this integrated chemometric approach demonstrates that data-driven optimization can deliver phenolic-rich herbal extracts with robust and balanced antioxidant potential for functional food applications.