EFFICIENT RECOVERY OF PHENOLICS FROM GARLIC PEEL USING PROBE ULTRASOUND: PROCESS OPTIMIZATION BY RESPONSE SURFACE METHODOLOGY

In Barranca (Lima, Peru), the agro-export of peeled garlic (Allium sativum) generates low-value peel residues that may be a source of phenolic compounds. Ultrasound-assisted extraction (UAE) can recover them rapidly, but its yield depends on interacting variables; therefore, this study aimed to optimize probe-based UAE of total phenolics from garlic peel using a Box-Behnken design. The peel, dried (50 °C, 24 h) in a POL-EKO SLW 400 SMART oven and ground in an IKA A11 basic mill, was extracted with methanol–water (80 mL) using a U.S. Solid FS1200 probe sonicator (tip depth 3 cm) in an ice bath, and the extracts were centrifuged (Sirena AFI-C060R-E; 4200 rpm, 15 min), filtered, and stored at 4 °C protected from light. The design (29 runs, five center points) varied amplitude (40–80%), time (5–25 min), methanol concentration (40–80%, v/v), and solid-to-liquid ratio (1:10–1:30 g/mL). Total phenolic content (TPC) was determined by the Folin–Ciocalteu method (Agilent BioTek Synergy H1 microplate reader; 725 nm) with a gallic acid calibration curve and expressed as mg gallic acid equivalents (GAE)/g of dry peel. Data were fitted in Minitab 22.5.2 to a quadratic model with backward elimination (ANOVA, R², adjusted R², predicted R², lack of fit; α = 0.05), and the optimum was obtained with the response optimizer. The reduced model was significant (p < 0.0001; R² = 64.0%, adjusted R² = 58.0%, predicted R² = 45.8%) with non-significant lack of fit, showing that methanol concentration, solid-to-liquid ratio, and time significantly affected TPC (p ≤ 0.045), whereas amplitude did not (p > 0.05). Accordingly, the highest values were 2.71 mg GAE/g (60% amplitude, 15 min, 1:30, 40% methanol) and 2.53 mg GAE/g (60% amplitude, 25 min, 1:30, 60% methanol), while the four lowest (0.85–1.25 mg GAE/g) all used 80% methanol, which suggests that water-rich mixtures of intermediate polarity favor the solubilization of polar phenolics. Likewise, the higher yield per gram at 1:30 is consistent with a larger solvent volume per sample mass, favoring mass transfer, and the time effect agrees with longer cavitation exposure, whereas the absence of an amplitude effect suggests that cavitation was already sufficient in the tested range. The optimizer predicted 2.52 mg GAE/g at 25 min, 1:30, and 43.6% methanol (desirability = 0.90); as the time and ratio optima lie at the upper limits studied, higher yields may be possible beyond them. In conclusion, methanol concentration, solid-to-liquid ratio, and time were the key variables for phenolic recovery from garlic peel, and amplitude can be kept low to save energy; although the predicted optimum still requires validation, the results support garlic peel as a promising phenolic source and a baseline for greener solvents.