The enantiomers of 5-dimethylamino-1-naphthalene sulfonyl (DNS)-derivatives of selected amino acids were successfully separated using capillary electrophoresis (CE) employing cyclodextrins (CD) as enantio-selective running buffer additives. A previously described model for retention and chiral recognition in CD-modified CE is shown to adapt well in this application. Resolution of the isomers is strongly influenced by the type and concentration of cyclodextrin employed, as predicted by the model. Although data indicates differences in the electrophoretic mobilities for some of the completely complexed enantiomer pairs, selectivity generally requires exploiting differences in the amino acid-CD complexation constants for enantiomer pairs. In this work, the D-enantiomers exhibit larger formation constants and are complexed to a greater degree (elute first) at moderate CD concentration. When mixtures of amino acids are analyzed, the effects of separation conditions on general elution behavior must be considered or separated enantiomer pairs will co-elute with other enantiomers. Preliminary results aimed at predicting the strength of DNS-amino acid enantiomer-CD interactions based on molecular modeling studies are presented. A statistical mechanical approach to treating computationally derived enantiomer-CD interaction energies is shown to provide reasonable correlation with separation performance.

A spectrophotometric method has been developed for the selective determination of sulfate in the 0-10 and 0-100 microgram (ppm in water) ranges. The aqueous samples are treated with a mixture of hydriodic acid, acetic anhydride, and sodium hypophosphite and heated in a modified countercurrent reaction apparatus to evolve hydrogen sulfide. This gas is swept by nitrogen into a buffered solution of ferric ion and 1,lO-phenanthroline, where it reduces ferric ion to ferrous ion. The bright orange tris(1,lO-phenanthroline)iron(ll) complex is formed and is measured spectrophotometrically at 510 nm. Of over 20 common ions, only nitrite and those ions capable of yielding hydrogen sulfide under the same conditions interfered. Sulfonated surfactants did not interfere. SULFATE REDUCTION to hydrogen sulfide is a very attractive basis for a simple, selective, and sensitive method for the determination of this extremely important anion. Use of this one step eliminates the shortcomings of barium sulfate and related precipitation methods. Hydrogen sulfide is a gas, so its quantitative transfer presents no problems. It is highly reactive, so its measurement is also easy. Determination of small amounts of sulfate thus need be no more difficult than the determination of small amounts of nitrogen by the Kjel-dah1 method.Sulfate reduction, however, is a problem because the reaction is highly irreversible. Thermodynamically, sulfate should be reduced in a Jones reductor. In reality, few reagents are known which can satisfactorily accomplish the task, and reduction has best been carried out by the use of hydriodic acid under drastic conditions. It has been found in this work that a mixture of hydriodic acid and acetic anhydride containing a small amount of sodium hypophos-

The enantiomers of 1, 1'-bi-Bnaphthol, 1, 1'-binaphthyl diyl hydrogen phosphate, and 1 , l '-binaphthyldicarboxylic acid are separated using capillary electrophoresis with cyclodextrins added to the running buffer. It is demonstrated that the type and concentration of cyclodextrin employed are critical for maximum enantiomeric resolution. A modified version of a previously described model of enantiomeric separations in capluary electrophoresis is shown to support the observed separation behavior. Molecular modeling is employed to calculate interaction energies between the various enantiomers and cyclodextrins. A reasonable correlation between these computationally derived interaction energies and separation behavior resulted from a statistical mechanical treatment of the molecular modeling data. The importance of hydrogen bonding in inclusion complex formation was probed and the effects of minimization and solvation in molecular modeling calculations are also discussed.0 1995 Wdey-Liss, Inc.

No abstract