Central composite design process optimization of ultrasound-assisted extraction of phenolic compounds from Manihot esculenta leaves

Abstract

Cassava (Manihot esculenta) leaves are an underutilized biomass source of phenolic compounds. This study optimized ultrasound-assisted extraction using a central composite design with extraction time and solid-to-liquid ratio as independent variables and total phenolic content (TPC) as the response. Across 13 runs, TPC ranged from 1.37 to 37.92 mg gallic acid equivalents (GAE)/g dry weight (DW). The quadratic model was significant (F = 21.63, p = 0.0004), explained 93.92% of response variability (R² = 0.9392), and showed a non-significant lack of fit (p = 0.2927). Solid-to-liquid ratio was the dominant factor (p < 0.0001), whereas extraction time was not significant within the studied domain (p = 0.5121). At the selected conditions (113.68 min and 1:15 w/v), three independent validation extractions produced a mean TPC of 35.82 ± 0.26 mg GAE/g DW (n = 3; CV = 0.73%), close to the predicted 34.63 mg GAE/g DW (3.43% relative error). The three independent optimized extracts contained 191.56 ± 0.56 mg quercetin equivalents/g DW (n = 3; CV = 0.29%) and showed 5.02 ± 0.08% DPPH inhibition (n = 3; CV = 1.50%) at the single extract concentration tested. SEM revealed irregular, aggregated freeze-dried particles. The results identify solvent-to-solid ratio as the principal variable governing TPC recovery; however, concentration-response antioxidant testing, and quantitative particle characterization are required before functional material performance can be inferred. Not applicable