The effect of silica surface chemistry in reverse-phase liquid chromatography of steroid compounds

Four steroid compounds, ethinyl estradiol (EE), levonorgestrel, gestodene and delta 6,7-dehydrogestodene were chromatographically studied using various batches of Spherisorb ODS1 and ODS2, 5 mum particle size HPLC column packing.;The chromatographic results from the four component steroid test solution highlight the repeatable chromatography achieved on the Spherisorb ODS1 phase and the non-repeatable chromatography expressed on the Spherisorb ODS2 phase, with respect to retention, resolution and the notable decrease in analyte selectivity of the keto steroids measured against EE. The physical and chemical parameters of both materials suggest that endcapping is the critical difference (ODS2 is endcapped, ODS1 is not) and the source of observed behavior. Analyte chemistry is also reviewed to ultimately determine a solute-stationary phase interaction mechanism responsible for the chromatographic effect shown, noting that simultaneous interactions, such as hydrogen bonding, other than hydrophobic interactions could be occurring.;The evaluation of various batches of both Spherisorb ODS1 and ODS2 was performed to characterize the source of irreproducible chromatographic behavior of the four component steroid solution on ODS2. An initial chromatographic study characterizes and relates the hydrophobic property and cation exchange potential of the ODS2 columns to the observed chromatographic behavior of the four component steroid solution. The hydrophobic subtraction model of reverse-phase column selectivity was also utilized to characterize the selectivity differences between the ODS1 and ODS2 phases, where the five column parameters: H (hydrophobicity), S (steric selectivity), A (hydrogen-bond acidity), B (hydrogen-bond basicity) and C (cation exchange capacity) were evaluated to provide additional insight to further define the nature of different solute-stationary phase interactions, specifically focusing on the hydrogen bonding capacities of the columns. The test mixtures analyzed include an evaluation of acids, bases and neutral solutes. The experimental investigation also includes the determination of whether the ODS1/ODS2 metal rich silica material contributes to a chromatographic chelating effect on either the EE or keto steroid solutes with the silica surface.;The overall results show that simultaneous solute-stationary interactions are occurring and that the dominant interaction is the hydrophobic interaction (H). The variability in retention and resolution, and the decrease in selectivity from ODS1 to ODS2 of the delta 6,7-dehydrogestodene, gestodene and levonorgestrel solutes, with respect to EE, in the four component steroid solution were directly impacted by the significant decrease in column property A, due to endcapping. EE was shown to be the least variable of the other steroids, with respect to retention variability from ODS1 to ODS2. A decrease in the solute-stationary phase hydrogen bonding basicity potential was also found, as represented by column property B, but to a lesser extent than the effect of column property A. The metals contained in the silica matrix had no direct chromatographic effect to the irreproducible behavior of the four component steroid solution on ODS2.