A HPLC-assisted mathematical prediction method of ternary solvent systems to develop an intelligent online selection...

Countercurrent chromatography (CCC) is an efficient technique for purifying bioactive natural compounds, but selecting the solvent system can be a time-consuming and crucial process for successful separation. This paper discussed the HPLC-assisted mathematical prediction method for the n-hexane/alcohol solvents/water (HAWat) and ethyl acetate/n-butanol/water (EBuWat) systems and designed an intelligent online selection system to simplify the separation process. First, the applicable rage of HAWat and EBuWat solvent systems were quantified by the methanol concentration at the column inlet when template molecules peak in a HPLC analysis (B%). Then, we analyzed the relationship between the logarithmic partition coefficient (log K) of template molecules and the alcohol solvent ratio by using different brands of chromatography columns. Moreover, the prediction function was developed by correlating the best solvent composition of CCC solvent system with the B% of template molecules. The results indicated that the HAWat system was suitable for separating compounds with B% > 85 %, while the EBuWat system was effective for compounds with 40 % < B% < 55 %. The quadratic function for HAWat systems and one universal EBuWat system (8.2:1.8:10) were determined and verified to develop the HPLC-assisted mathematical prediction function method. The method could select suitable solvent systems for separfigation easily and quickly. And a step of result correction was required for the occasional discrete value. Finally, an intelligent CCC online selection system that integrates established mathematical functions was designed to enable real-time solvent selection, on-demand solvent configuration, and automated sample separation. It eliminates the iterative process of shake-flask experiments and facilitates intelligent upgrading of CCC instrumentation.