Using nonrandom two-liquid model for solvent system selection in counter-current chromatography

• A semi-empirical predictive method was established by using the NRTL model.
• The established method can be used for predicting partition coefficient.
• The method was tested and validated with model solutes and a real life separation.
• The experimental efforts and cost involved in solvent system selection are reduced.
• Two major compounds were successfully separated from Malus hupehensis.

Selection of an appropriate solvent system is of great importance for a successful counter-current chromatography separation. In this work, the nonrandom two-liquid (NRTL) model, a thermodynamic method, was used for predicting the partition coefficient based on a few measured partition coefficients. The NRTL method provides quite satisfactory results for model solutes in first correlating measured partition coefficient in a few representative biphasic liquid systems and then successfully predicting partition coefficient in other two-phase liquid systems. According to the predicted partition coefficient, a suitable solvent system can be screened. Assisted with the NRTL method, the solvent system composed of hexane/ethyl acetate/methanol/water (1:4:1:4, v/v) was rapidly screened for the successful separation of two major compounds with high purity from Malus hupehensis leaves. The results demonstrated that the NRTL model can offer a simple and practical strategy to estimate partition coefficients in support of CCC solvent system selection, which will significantly minimize the experimental efforts and cost involved in solvent system selection.