Property model prediction of the boiling point for pure and mixture solvents applied in herbal extraction

Abstract

Studies on the effect of different solvent used in herbal extraction on the extraction yield have been commonly investigated. However, the impact of solvent property, such as boiling point attracts less attention. Besides, the measurement of the solvent boiling point requires much time and effort. Therefore, this study aims to use property models in predicting the boiling point of single and mixture solvents. Three pure property models were applied for several solvents while two mixture property models were utilised for the selected mixture solvents. The prediction data then were compared with the experimental data obtained from the previously published result. The mean relative error was computed to choose the best property model that can be applied to calculate the boiling point. In this study, the Marrero and Gani’s method was selected to be the best model in predicting boiling point for pure solvent. In contrast, Klein and Wu’s method was chosen as the best model for the boiling point prediction of mixture solvents. The selection of the best property model was made by choosing the lowest mean relative error.