Abstract:

Most of the studies on cyclodextrin (CD)-based chiral stationary phase (CSP) have focused on the functional derivatization of CD or the bridging arms to introduce more interaction sites and thus improve the chiral resolution ability. At present, there are only a few reports on CSP that can reflect the intrinsic recognition ability of natural CD. In this study, a mono(6-mercapto-6-deoxy)-β -CD CSP (CSP1) with a clear and controllable structure was synthesized by the "thiol-ene" click reaction. CSP1 retained the intrinsic structure of natural CD to the maximum extent, and the bridge arm had no recognition site. The results of ¹³ C solid-state nuclear magnetic resonance (SSNMR) and Fourier transform infrared (FTIR) analyses confirmed the successful preparation of CSP1. Elemental analysis results showed that compared with double-bond functionalized silica, the percentages of C, H, and N in CSP1 increased, and the calculated CD loading of CSP1 was 0.82 μmol/m² . Reversed-phase high performance liquid chromatography was performed for the chiral resolution of more than 50 chiral enantiomers, including isoxazoline, chiral lactide, chiral ketone, flavone, and dansyl amino acid. This fully demonstrated the intrinsic chiral recognition ability of natural CD, and the results showed that the intrinsic recognition ability of cyclodextrin was more conducive to the separation of Ph-Ph samples containing two hydrophobic benzene ring groups in the isoxazoline samples. For the Ph-Py and Ph-OPr samples, the separation effect was not satisfactory. The separation results for the Ph-Py samples were not ideal because the outer hydroxyl group of cyclodextrin could form a hydrogen bond with the pyridine nitrogen, thus hindering the inclusion and the separation effect. This eventually led to poor separation of the enantiomers. While the pyrrolidone group in the Ph-OPr sample could also form a good inclusion complex with cyclodextrin, its higher polarity weakened the inclusion effect compared to that for benzene rings, thus leading to poor chirality separation results. For chiral lactides, the intrinsic recognition ability of CD was good only for the separation of some samples. In the separation of chiral ketones, large steric hindrance effect inhibited the intrinsic recognition ability of CD, and the separation effect of such samples on CSP1 was not ideal. External functional groups were required in some cases to further regulate the chiral recognition performance. The molecular structure of dansyl amino acids played an important role in the separation effect, in addition to the intrinsic recognition ability of CD. The number of side chains in the substituent also affected the quality of separation. Lengthening the side chain or increasing the hydrophobicity could effectively improve the separation efficiency. The separation effect of flavanone samples on CSP1 was ordinary. The substituent positions also affected the separation effect. In order to further explore the intrinsic recognition ability of CD, the functional triazole-bridged CD-CSP (CSP2) and imidazole-bridged CD-CSP (CSP3) (the surface CD loadings of CSP2 and CSP3 were 0.51 μmol/m² and 0.46 μmol/m² , respectively) prepared earlier were selected and compared under the same chromatographic conditions. The results showed that the separation of the sample was related not only to the structure of the chiral medium but also to the structure of the sample molecules. Functional modification of the bridge arm could improve the selectivity of some enantiomers, but would also cause partial loss of the intrinsic chiral recognition ability of CD. For samples with the intrinsic recognition ability of CD to facilitate separation, no functional group had to be added to the bridge arm when designing a chiral medium. This study provides a useful reference for the design of CD-based CSPs.

Key words: cyclodextrin, click chemistry, high performance liquid chromatography (HPLC), chiral recognition