Chemometric optimization of cation-selective exhaustive injection sweeping micellar electrokinetic chromatography for quantification of ractopamine in porcine meat.

An online stacking capillary electrophoresis (CE) method, cation-selective exhaustive injection sweeping micellar electrokinetic chromatography (CSEI-sweep-MEKC), is developed and optimized for analysis of ractopamine (RP) and its homologue dehydroxyractopamine (DRP) in porcine meat. Chemometric experimental design was used to achieve the best possible optimization and reduce the number of trials and errors. The CSEI-sweep-MEKC method enables nanogram per gram level analysis, with limits of detection (LODs) in meat of 5 ng/g for RP and 3 ng/g for DRP (S/N = 3). A higher conductivity buffer (HCB) zone was injected into the capillary, allowing for the analytes to be electrokinetically injected at a voltage of 9 kV for 12 min. Using 125 mM sodium dodecyl sulfate and 15% methanol in the sweeping buffer, RP and DRP were well-separated. The method was validated with a linear calibration curve of 10-300 ng/g (r > 0.994). In comparison to the normal capillary zone electrophoresis method (1 psi for 10 s), this stacking strategy resulted in 900 times sensitivity enhancement. This technique was further applied for analyzing seven kinds of commercial meats, and the residual RP was detected in one (5.76 ng/g of RP). The data were corresponding to the data analyzed by the commercial testing kit and mass spectrometry spectra. This method was successfully used on real samples and is considered feasible for serving as a tool for routine examination in markets.

On-line sample stacking and short-end injection CE for the determination of fluoxetine and norfluoxetine in plasma: Method development and validation using experimental designs.

A short-end injection CE method combining field-amplified sample stacking (FASS) is presented for the analysis of fluoxetine (FL) and norfluoxetine in plasma. In this study, FASS enhanced the sensitivity about 1100-fold, while short-end injection reduced the analysis time to less than 4 min. Parameters involved in the separations were investigated using a central composite design (CCD) and response surface methodology to optimize the separation conditions in a total of only 32 runs. Samples injected into the capillary for 99.9 s at a voltage of -5 kV were stacked in a water plug (0.5 psi, 9 s). Baseline resolution of FL and its major metabolite was achieved using a BGE formulation consisting of phosphate-triethanolamine at low pH, and a separation voltage of -10 kV. Five percent methanol was added as organic modifier to enhance selectivity and resolution. The linear range was between 10 and 500 ng/mL (r >0.9946), covering the expected plasma therapeutic ranges. The LOD in plasma were 4 ng/mL (S/N = 3), a value comparable to that obtained using LC-MS, showing the success of the on-line stacking technique. Our method was also successfully validated in quantification and pharmacokinetic studies with three volunteer plasma samples and could be applied to pharmacogenetic studies.