A novel 9 × 9 map-based solvent selection strategy for targeted counter-current chromatography isolation of natural products

• A 9 × 9 map-based solvent selection strategy for targeted CCC isolation was introduced.
• To select a suitable CCC solvent is only to screen two to four random solvent systems in the map.
• After simple screening and calculation, the sweet systems with suitable K values were obtained.
• These sweet systems were found to distribute on the same line, but have different sample capacities.
• Thus the solvent system with sweet K and higher sample capacity was selected for CCC isolation.

Counter-current chromatography (CCC) is an efficient liquid–liquid chromatography technique for separation and purification of complex mixtures like natural products extracts and synthetic chemicals. However, CCC is still a challenging process requiring some special technical knowledge especially in the selection of appropriated solvent systems. In this work, we introduced a new 9 × 9 map-based solvent selection strategy for CCC isolation of targets, which permit more than 60 hexane–ethyl acetate–methanol–water (HEMWat) solvent systems as the start candidates for the selection of solvent systems. Among these solvent systems, there are clear linear correlations between partition coefficient (K) and the system numbers. Thus, an appropriate CCC solvent system (i.e., sweet spot for K = 1) may be hit by measurement of k values of the target only in two random solvent systems. Besides this, surprisingly, we found that through two sweet spots, we could get a line (“Sweet line”) where there are infinite sweet solvent systems being suitable for CCC separation. In these sweet solvent systems, the target has the same partition coefficient (K) but different solubilities. Thus, the better sweet solvent system with higher sample solubility can be obtained for high capacity CCC preparation. Furthermore, we found that there is a zone (“Sweet zone”) where all solvent systems have their own sweet partition coefficients values for the target in range of 0.4 < K < 2.5 or extended range of 0.25 < K < 16. All results were validated by using 14 pure GUESSmix mimic natural products as standards and further confirmed by isolation of several targets including honokiol and magnolol from the extracts of Magnolia officinalis Rehd. Et Wils and tanshinone IIA from Salvia miltiorrhiza Bunge. In practice, it is much easier to get a suitable solvent system only by making a simple screening two to four HEMWat two-phase solvent systems to obtain the sweet line or sweet zone without special knowledge or comprehensive standards as references. This is an important advancement for solvent system selection and also will be very useful for isolation of current natural products including Traditional Chinese Medicines.