Search

A new silylating reagent that forms cyanosyl (cyanoethyldimethylsilyl) derivatives with monohydroxysteroids is introduced and its usefulness for the gas chromatographic analysis of these compounds evaluated. The reactivity of the new reagent, N-methyl-N-cyanoethyldimethylsilyltrifluoroacetamide, has been studied using a series of monohydroxysteroids and compared to that of other reagents forming cyanosyl or alkylsilyl derivatives. The results obtained indicate that at room temperature, in polar solvents, the reagent reacts rapidly and selectively with most hydroxyl groups found in steroids. The cyanosyl derivatives formed have retention properties that are comparable to those of the corresponding alkylsilyl derivatives. However, the former offer advantageous properties for detection since they can be analyzed with a nitrogen-phosphorus detector, and they exhibit mass spectral features that are suitable for identification and quantitation at low levels using selected ion monitoring mass spectrometry. The detection limits for all the monohydroxysteroids studied in this work, as cyanosyl derivatives, are typically on the order of 30-50 pg and the relative response factors are unity within a range of 25 percent.

Qualitative and quantitative effects of classical reactions on steroids observed by gas-liquid chromatography (GLC) under standardized conditions, including the double internal standard technique are reported. Simple procedures applicable to nanogram amounts of reactants are described. Reactions studied include the conversion of keto groups to hydroxyl groups by NaBH4, and to dioxolone derivatives by 1,2-diethanol; 17 alpha-hydroxylation of C20-ketosteroids; the conversion of hydroxyl groups to trimethylsilyl (TMS) ethers by hexamethyldisilazane; the hydrolysis of dioxolone and TMS derivatives by H+. Effects of Wolff-Kishner reagents, and CrO3 were also studied. GLC chromatograms of reaction mixtures of single- and multistep reactions readily provide information on effects on functional groups at positions 3, 17, 20, and 21 in the pregnane series, and the retention times of many steroids unavailable from commercial and other sources. GLC data analysis provides relationships between steroid structure and retention time from which methods for the computation of retention times and for steroid identification are designed. The accuracy of the computation methods is demonstrated.