Preparation and Retention Behavior of Cyclodextrin-Bonded Stationary Phases on a Silica Support without an Unreacted Spacer Chain
Cyclodextrin-bonded stationary phases on a silica support without an unreacted spacer chain for liquid chromatography were prepared by the reaction of hydride-modified silica with monoallyloxyethylated /3-and y-cyclodextrins. Both cyclodextrins were coupled to 3-(2-tosyloxyethoxy)propyl silica to produce stationary phases with an unreacted spacer chain. The cyclodextrin units in these four stationary phases were connected to the silica surface with the same linkage. The enantiomeric separation
... iomeric separation ability of these cyclodextrin stationary phases was investigated using eleven dansylamino acids. Both /3-cyclodextrin stationary phases exhibited a comparable enantioselectivity, except for dansyl-DL-aspartic and -glutamic acids. However, the f3-cyclodextrin stationary phase without an unreacted spacer chain gave a higher column efficiency, compared with the /3-cyclodextrin stationary phase with an unreacted spacer chain. On the other hand, the y-cyclodextrin stationary phase without an unreacted spacer chain was more enantioselective than that with an unreacted spacer chain. The effect of silanization of unreacted silanol groups on the silica surface is discussed. Keywords Liquid chromatography, cyclodextrin dansylamino acid stationary phase, unreacted spacer chain, optical resolution, Cyclodextrins (CDs) have attracted much attention in various fields and have been increasingly used for enzyme models, drug and food formulations, separations and many other applications.1'2 In 1983 -1984, three CDbonded silica stationary phases for liquid chromatography were successively reported.3-5 To date, unmodified and chemically modified CD stationary phases have been marketed. On these stationary phases each CD unit is connected to a spacer previously bonded to the silica surface. The effect of different spacers on the chromatographic behavior of CD stationary phases has been investigated (e.g., for the column efficiency and retention of disubstituted benzene derivatives on /3-CD stationary phases6 and for chiral separations of racemic solutes on unmodified' and 3,5-dimethylphenyl car-bamoylated8 /3-CD stationary phases). The results suggest that the efficiency of the CD stationary phases is affected by the types of spacers, amounts of the CD units immobilized and degrees of substitution of the carbamate groups. As mentioned above, the CD units are coupled to a silica-gel derivative with an active terminal group previously bonded on the silica surface. All of the terminal groups cannot react with the CD units. In general, the proportion of terminal groups used for the immobilization of CD units is much less than those remaining unreacted. The existence of an unreacted spacer chain is unavoidable in this case, and may affect the selectivity of the CD stationary phases. Removing such unreacted spacer chains is expected to enhance the selectivity derived from the CD unit. In this paper we describe the preparation of /3-and y-CD stationary phases without unreacted spacer chains, and compare their separation abilities for dansylamino acid enantiomers with the abilities of the corresponding CD stationary phases with unreacted spacer chains. Experimental Apparatus The employed HPLC system comprised a Waters 6000A pump, a Rheodyne Model-7125 sample injector, a JAI UV-3702 detector operating at 220 nm and a Shimadzu Chromatopac C-R6A integrator. Each CD stationary phase was packed into a stainless-steel column (10 cmX4 mm i.d.). The flow rate of the eluent (triethylammonium acetate buffer-methanol or -acetonitrile) was 1.0 ml min 1. The buffer pH was adjusted by mixing 1 % triethylamine and 1 % acetic acid. The concentration of the sample solutes was 0.2 mM and a volume of less than 5 ml was injected.