Ionic liquid screening for solvent design of herbal phytochemical extraction

Abstract

Ionic liquids (ILs) have been used in many applications and currently have been a favourable solvent in separation technology due to their solvation power compared to organic solvent. A development of IL solvent design approach is necessary in order to apply the most optimal solvents in herbal phytochemical extraction. In this study, solvent design utilised a systematic approach combined with property predictive model rather than trial-And-error experimental approach to reduce the amount of solvent waste and extraction time. This work focus on the screening of ILs as phytochemical extraction solvents of phenolic acids (e.g. gallic acid and caffeic acid) where we used property models of solubility and toxicity as part of solvent design. The methodology consisted of several stages. Stage 1 specifies the user needs of an extraction solvent, problems and constraints of new solvent design. Stage 2 involved in the development of a comprehensive Excel-based database of ionic liquid properties (e.g. solubility, heat capacity etc.) and factors that affect phytochemical extraction (e.g. extraction time, particle size etc.). In Stage 3, property library was assembled by collecting property models relevance for ILs from other studies to identify the most suitable models and estimate property values for solvent design. In Stage 4, ILs available in the database were first screened based on four factors: Toxicity, heat capacity, density and viscosity (properties which represent characteristics of solvent and which affect the extraction). Only those ILs which have acceptable value range of each properties were selected. Finally, the ILs candidates were further screened down based on their solvation performance by using a solubility parameter-solid-liquid equilibrium approach involving UNIFAC-IL models to select most optimal solvent that can extract highest amount of phytochemicals. From the screening process, 16 best IL solvent candidates for the phenolic acid extraction were obtained from a database of 880 imidazolium-based IL.