Extraction of phenolic compounds from cocoa shell: Modeling using response surface methodology and artificial neural networks

• Green extraction of phenolic compounds from the cocoa shell was modeled and optimized. Response surface methodology artificial neural networks showed comparable estimation abilities. Spectrophotometric results correlate with UPLC-measured individual compounds. Major phenolics in were protocatechuic acid mono- dimeric flavanols. An eco-friendly method using just water is presented to revalorize shell. This work’s objective model optimize a green as strategy this by-product, obtaining novel high-value products. According Box-Behnken design, 27 extractions carried out at different conditions temperature, time, acidity, solid-to-liquid ratio. Total compounds, flavonoids, flavanols, proanthocyanidins, acids, o -diphenols, vitro antioxidant capacity assessed each extract. (RSM) (ANN) used effect parameters on aqueous The obtained mathematical models fitted well for all responses. RSM ANN exhibited high capabilities. main factors affecting followed by ratio, acidity. optimal 100 °C, 90 min, 0% citric acid, 0.02 g mL ?1 water. Under these conditions, experimental values response variables matched those predicted, therefore, validating model. UPLC-ESI-MS/MS revealed presence 15 being procyanidin B2, (?)-epicatechin, (+)-catechin, major ones. significant correlation UPLC results, confirming their potential use screening optimization purposes. Aqueous extracts would have sustainable food-grade ingredients nutraceutical