Development of Aqueous Two-Phase System for Eco-Friendly and Decolorized Triterpenoids Extraction of Centella asiatica Optimized by Response Surface Methodology

Centella asiatica (L.) Urban extracts are widely used in food, cosmetic, and pharmaceutical products. Conventional solvent extraction requires organic solvents and often produces dark green extracts due to chlorophyll co-extraction, limiting their industrial applications. This study developed an environmentally friendly aqueous two-phase system (ATPS) for the simultaneous extraction of triterpenoids of Centella asiatica. Response surface methodology with a central composite design was employed to optimize ethanol concentration, ammonium sulfate concentration, and temperature. A rapid and validated HPLC method for quantification of four triterpenoids—madecassoside, asiaticoside, madecassic acid, and asiatic acid—was developed and validated on a Poroshell 120 SB-C18 column (3.0 × 150 mm, 2.7 microns) at 40 °C. The developed method for all four triterpenoids was rapid (14 min run time) with acceptable system suitability, specificity, linearity (R > 0.9995), accuracy (98.42–101.52%), precision (0.16–1.36%), LOD (0.04–0.64 µg/mL), and LOQ (0.13–1.93 µg/mL). The ANOVA data from the central composite design of response surface methodology for total triterpenoid content were fitted with quadratic models, yielding acceptable R2 (>0.97), adjusted R2 (>0.95), and predicted R2 (>0.80). The optimum extraction conditions were 18.29% w/w ethanol, 33.93% w/w ammonium sulfate, and 60.04 °C, yielding a total triterpenoid content of 30.53 mg g−1 dry weight. The optimized ATPS produced higher yields of madecassoside (13.15 mg g−1), asiaticoside (9.02 mg g−1), madecassic acid (5.26 mg g−1), and asiatic acid (3.10 mg g−1) than conventional ethanol extraction while simultaneously reducing chlorophyll-derived coloration. This ATPS extraction strategy provides a sustainable approach for producing suitable Centella asiatica extracts for cosmeceutical and pharmaceutical applications.