Drug identification in biological matrices using capillary electrophoresis and chemometric software

Drugs and their metabolites in biological specimens are analyzed by a variety of techniques. Capillary electrophoresis could provide another useful approach because of its unique selectivity and high resolving power. For routine use, however, rugged methods must be developed and combined with detection that confirms peak purity and identity in difficult sample matrices, such as, urine. In this study, capillary electrophoresis is used with diode array detection, and chemometric software is employed for spectral analysis. The software includes a series of chemometric tools. Principle Component Analysis and Iterative Target Transform Factor Analysis are used to inspect each electropherogram for spectral homogeneity of the peaks and to deconvolute comigrations. These algorithms are used to confirm the assay results. This approach is tested and demonstrated for the analysis of amphetamine and common interferences in human urine.