Abstract

This study explores the selective decaffeination of green tea using supercritical (SC) CO2 modified with ethanol and/or water as cosolvents based on experimental data in eight manuscripts in literature. We analyzed the solubility trends of caffeine (data from three literature sources) and epigallocatechin-3-gallate (EGCg) (data from a single source) using the Hansen solubility parameter (HSP) model to evaluate the solute-solvent interactions through the distance parameter R. We identified that the highest selectivity (favoring caffeine removal and EGCg retention) occurred at 70 °C and ?25 MPa using pure CO2 and hypothesized that the addition of water facilitated substrate swelling and improved inner mass transfer. Ethanol showed limited benefits at higher pressures. Data for operational solubility were more in line with actual solubility than apparent solubility, although both were still dependent on substrate-solute interactions. Solubility measurements revealed a significant scatter in the data, particularly for caffeine in water- and/or ethanol-modified CO2, underscoring the need for better experimentally derived data. Despite these challenges, HSP-based approaches effectively identified conditions that maximized selective decaffeination while preserving EGCg content. © 2025