Recovery of phenolic compounds from cashew apple pomace using green solvents

The Brazilian cashew-processing industry generates significant quantities of cashew apple pomace (CAP), an agro-industrial by-product rich in nutrients and bioactive compounds, particularly phenolic compounds. The recovery of these compounds may contribute to the development of value-added ingredients and promote sustainable waste valorization. Natural deep eutectic solvents (NADES) represent promising alternatives to conventional extraction media due to their tunable physicochemical properties. However, extraction efficiency is influenced by solvent composition and processing conditions, making optimization essential. Therefore, this study aimed to optimize ultrasound-assisted extraction of phenolic compounds from CAP using betaine-based NADES and compare their performance with a conventional hydroethanolic system. A three-factor, three-level Box–Behnken design comprising 15 experimental runs per extraction system was implemented to maximize total phenolic content (TPC). The extraction systems consisted of hydroethanolic mixtures with water contents of 30%, 50%, and 70% (w/w), betaine: glycerol (BG, 1:2 molar ratio), and betaine: malic acid (BM, 1:1 molar ratio). Extractions were conducted in an ultrasonic bath, varying solvent water content (30–70%, w/w), extraction time (10–40 min), and sample-to-solvent mass ratio (0.05–0.15 g/g). TPC was determined using the Folin–Ciocalteu assay, while antioxidant capacity was evaluated using DPPH. Second-order polynomial models were fitted by ordinary least squares regression and reduced by excluding non-significant terms while preserving model hierarchy (p<0.10). The reduced models were statistically significant for all extraction systems, with BM (R² = 0.954; adjusted R² = 0.936) and BG (R² = 0.889; adjusted R² = 0.844) fitting better than than for the hydroethanolic system (R² = 0.686; adjusted R² = 0.511). The sample-to-solvent ratio showed significant positive linear effects across all systems, whereas water content showed significant negative quadratic effects, with additional negative linear effects in both NADES systems. Extraction time had limited influence, with a significant quadratic effect observed only in the hydroethanolic system. Model-based optimization predicted optimal water contents of approximately 52%, 45%, and 44% for the hydroethanolic, BG, and BM systems, respectively, at a sample-to-solvent ratio of 0.15 g/g. Under evaluated conditions, BM exhibited the highest TPC (3,350.59±164.06 µg GAE/mL), followed by BG (2,461.59±26.25 µg GAE/mL) and the hydroethanolic system (419.41±20.63 µg GAE/mL). Corresponding DPPH capacities were 2,606.95±248.52, 2,369.62±119.64, and 677.68±20.59 µg TE/mL, respectively. These findings highlight the potential of betaine-based NADES for producing phenolic-rich CAP extracts with antioxidant properties. Independent validation of the predicted optimal conditions and chromatographic characterization is still required. This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - Brasil (CAPES) - Finance Code 001. The authors also acknowledge CAPES for funding this research through scholarships