Abstract:
On the role of outside vs. inside interactions for the enantiorecognition by cyclodextrins: conventional mechanism and novel insights
Native and derivatized cyclodextrins represent versatile chiral selectors for the separation of enantiomers by chromatographic and electromigratory methods. However, the underlying mechanisms of enantiorecognition are still poorly understood. While the presence of a cavity and inclusion therein is often thought to be of prime importance for enantioseparation of racemic analytes it is conceivable that this effect can be complemented, or even overridden, by interaction of analytes with the external surface of cyclodextrins. To probe the importance of the molecular inclusion mechanism, linear dextrin derivatives - devoid of a true cavity - have been synthesized and compared with their cyclic analogues with regard to their properties as chiral selectors for enantioselective gas chromatography.
Via the comparison of cyclic and acyclic dextrins (cyclodextrins versus 'acyclodextrins') it is demonstrated that the existence of a cavity is not a prerequisite to enantiorecognition. Thus, the well known gas-chromatographic selector heptakis(2,3-di-O-acetyl-6-O-tert-butyldimethylsilyl)-beta-cyclodextrin has been compared with the corresponding linear derivative heptakis[(2,3-di-O,4''-O)-acetyl-(1'-O,6-O)-tert-butyldimethylsilyl]-maltoheptaose and a complementary enantioseparation ability for alpha-amino acid derivatives (N-trifluoroacetyl-O-methyl/ethyl-esters) and racemic halogenated compounds was observed. In several cases, the acyclic chiral stationary phase (CSP) was even more enantioselective than the cyclic CSP. While it is conceivable that the maltoheptaose derivative could still form a "pseudo cavity" which is responsible for the observed enantiorecognition, this is certainly not the case with the maltotriose and glucose derivatives, these molecules being too small to be able to form a helical conformation.
Surprisingly, even on the single building block comprising cyclodextrins and linear dextrins, i.e., (2,3,4-tri-O-acetyl-1,6-di-O-tert-butyldimethylsilyl)-D-glucose, the enantiomers of some alpha-amino acid derivatives and racemic halogenated compounds could be separated by gas chromatography. Whereas cyclodextrins are available only in the D form, the linear counterparts can readily be synthesized in the D and L form, which is of interest for peak inversion scenarios in the chromatographic analysis of non-racemic mixtures.
These unexpected enantioseparations prove the existence of enantioselective interactions with the outer surface of the selector.
Enantiorecognition mechanism by NMR spectroscopic studies
The controversial role of the inclusion phenomenon by cyclodextrins has also been verified by NMR studies in solution, where investigations were focused on the enantiorecognition process and on the conformation of the derivatized chiral selector. In this respect, derivatized cyclodextrins differ strongly from the native macrocycles. Due to the distortion of the glucose units and their tilting across the glycosidic linkages, the conformation of modified alpha-, beta- and gamma-cyclodextrin selectors are often significantly different from a classical torus shaped structure and the diameter of the cavity is markedly affected by derivatization.
Thus, the functional groups on the two rims of cyclodextrins play an important role in the complexation step as demonstrated in this work for the enantiodiscrimination of 'compound B' by heptakis(2,3-di-O-acetyl-6-O-tert-butyldimethylsilyl)-beta-cyclodextrin.
The yet unchallenged role of the cavity in the enantioseparation process of racemic hydrocarbons
In the case of apolar racemic analytes such as saturated aliphatic hydrocarbons, the inclusion process still appears to be the driving force for enantiorecognition on the well established selector Chirasil-beta-Dex, since no enantioseparation is observed on the acyclic counterpart of linear permethylated maltoheptaose (both as the diluted and supported CSP).
However, in many cases, e.g. for the per-O-methylated beta-cyclodextrin, the role of the cavity is not very well understood. Thus, the application of linear selectors as novel CSP represents a comparable tool to determine whether an inclusion phenomenon is present or absent with the established cyclic selectors.
Furthermore, the size of the cavity cannot be the sole parameter in rationalizing the enantiorecognition because the nature of the functional groups on the two rims plays a key role for the overall structure by modifying the cavity size and by directing the enantiorecognition towards the external surface of the macrocycles.