Macroporous monolithic chiral stationary phases for capillary electrochromatography: New chiral monomer derived from cinchona alkaloid with enhanced enantioselectivity

Abstract A new chiral monomer derived from cinchona alkaloid, namely O ‐9‐( tert ‐butylcarbamoyl)‐11‐[2‐(methacryloyloxy)ethylthio]‐10,11‐dihydroquinine 1 , was employed for the preparation of enantioselective monolithic capillary columns by an in situ copolymerization with 2‐hydroxyethyl methacrylate 2 (HEMA), ethylene dimethacrylate 3 (EDMA) in the presence of cyclohexanol and 1‐dodecanol as porogens (UV or thermal initiation of azobisisobutyronitrile (AIBN) as radical initiator). The porous properties and the electrochromatographic behavior of the new chiral monoliths were comparatively evaluated with previously described analogs obtained from O ‐9‐[2‐(methacryloyloxy)ethylcarbamoyl]‐10,11‐dihydroquinidine 4 as chiral monomer. Despite close structural and physicochemical similarities of the both chiral monomers, the pore distribution profiles of the resulting monoliths were shifted typically towards larger pore diameters with the new monomer 1 . Once more, it was confirmed that a low cross‐linking (10 wt% related to total monomers) and a pore diameter of about 1 μm in the dry state provides the best electrochromatographic efficiency as a result of lower resistance to mass transfer (smaller C‐term contribution to peak broadening) and more homogeneous flow profile (smaller A‐term). Most importantly, as expected the new poly( 1 ‐ co ‐HEMA‐ co ‐EDMA) monoliths showed enhanced enantioselectivities and in addition faster separations as compared to poly( 4 ‐ co ‐HEMA‐ co ‐EDMA) analogs, which represents a significant improvement. Further, the elution order was reversed owing to the pseudoenantiomeric behavior of quinine‐ and quinidine‐derived monomers. Fluorescence‐labeled 9‐fluorenylmethoxycarbonyl (FMOC), dansyl (DNS), 7‐dimethylaminosulfonyl‐1,3,2‐benzoxadiazol‐4‐yl (DBD), carbazole‐9‐carbonyl (CC) amino acids could be separated with resolution values between 2 and 4 (with efficiencies typically between 100 000 and 200 000 plates/m) and fluorescence detection (variable wavelength fluorescence detector in‐line with UV) yielding routinely a gain in detection sensitivities up to two orders of magnitude without specific optimization of the conditions with regards to fluorescence efficiency.