Design of countercurrent separation of Ginkgo biloba terpene lactones by nuclear magnetic resonance.

2012 
Terpene lactones such as bilobalide, ginkgolides A, B, C, and J are major bioactive compounds of Ginkgo biloba L. Purification of these compounds is tedious due to their similar chemical properties. For the purpose of developing an effective and efficient method for both analytical and preparative separation of terpene lactones in G. biloba, an innovative orthogonality-enhanced high-speed countercurrent chromatography (HSCCC) method was established. Taking advantage of quantitative 1H NMR (qHNMR) methodology, partition coefficients (K) of individual terpene lactones were calculated directly from crude G. biloba leaf extract, using their H-12 signals as distinguishing feature. The partitioning experiment assisted the design of a two dimensional (2D) HSCCC procedure using a pair of orthogonal HSCCC solvent systems (SSs), ChMWat +4 and HEMSoWat +3/0.05%. It was surprising that the resolution of ginkgolides A and B was improved by 25% in the HEMWat +3 SS modified with 0.5% DMSO. Consequently, all five terpene lactones could be well separated with qHNMR purity > 95% from G. biloba leaf extract. The separation was further evaluated by offline qHNMR analysis of HSCCC fractions associated with Gaussian curve fitting. The results showed less than 2% error in HSCCC retention predicted from the partitioning experiment. This compelling consistency demonstrates that qHNMR-derived K determination (“K-by-NMR”) can be used to predict CCC fractionation and target purification of analytes from complex mixtures. Furthermore, Gaussian curve fitting enabled an accurate prediction of less than 2% impurity in the CCC fraction, which demonstrates its potential as a powerful tool to study the presence of minor constituents, especially when they are beyond the detection limit of conventional spectroscopic detectors.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    26
    References
    22
    Citations
    NaN
    KQI
    []