Experimental design and on-line stacking capillary electrophoresis for the determination of abused drugs and some major metabolites in urine

• Main Authors: Yu-Tzu Hsiao, 蕭郁慈
• Other Authors: Shou-Mei Wu
• Format: Others
• Language: zh-TW
• Published: 2007
• Online Access: http://ndltd.ncl.edu.tw/handle/26412444531739000327

碩士 === 高雄醫學大學 === 臨床藥學研究所碩士班 === 95 === A capillary electrophoresis method was established for the determination of some common abused drugs in urine samples, including alprazolam, α-hydroxyalprazolam, clonazepam, 7-aminoclonazepam, diazepam, nordiazepam, flunitrazepam, 7-aminoflunitrazepam, norflunitrazepam, oxazepam, codeine, morphine, methamphetamine and ketamine . The method was developed using chemometric experimental design in a three-step procedure, including preliminary experiments, fractional factorial design (fFD) and central composite design (CCD). To find the significant variables, six factors were selected from the preliminary experiments, including the concentration and the pH of the phosphate buffer, the concentration of surfactant in the sweeping buffer, the percentage of organic solvent in the separation and the sweeping buffer, and the concentration of reconstitution solution. Then fFD was performed. Four out of these six factors were found to be siginificant, and were further investigated in the final step, CCD. In our study, samples were pre-treated with liquid-liquid extraction, then analyzed by using stacking with reverse migrating micelles as the on-line concentration method. The optimized CE conditions were as follow: First, phosphate buffer (75 mM, pH 2.5) containing 30% methanol was filled into capillary, then pressure injection (1 psi, 200 s) was used to load the samples. Phosphate buffer (75 mM, pH 2.5) containing 10% methanol and 65 mM SDS was used as the sweeping buffer to separate the analytes at −15 kV. During method validation, calibration plots were linear (r >0.992) over a range of 0.1-3 μg/mL for codeine, ketamine and methamphetamine and 0.15-3 μg/mL for morphine, 0.1-1 μg/mL for alprazolam and oxazepam, and 0.1-1.2 μg/mL for other analytes. During intra- and inter-day analysis, relative standard deviations (RSD) and relative errors (RE) were all less than 14%. The analytes could be simultaneously analyzed and have a detection limit down to 20-50 ng/mL. The results showed good coincidence with GC–MS or LC-ESI-MS. This method was feasible for application to detect trace levels of abused drugs.