ABSTRACT
Spectrophotometry was used with multivariate calibration to simultaneously determine compounds in mixtures. Two antidepressant mixtures were investigated: imipramine hydrochloride and chlordiazepoxide and nortriptyline hydrochloride and fluphenazine hydrochloride. Considerable spectral overlap and large differences in component concentrations were challenges. Since this type of analysis is often performed using complex algorithms, a simple strategy was used here for the simultaneous determination of both mixture components by classical least squares, principal component regression, and partial least squares. Experimental design was used to select the optimum parameters including the wavelength range, sampling interval, software, and derivative order. Accuracy was enhanced by proper wavelength selection. In addition, derivatives of the raw spectra improved the selectivity. The standard deviation, deviation of mean recovery from 100%, and prediction ability of the models were used as the responses. In respect to these terms, first-order derivatization of the spectra and a sampling interval of 1 nm provided the best results. In particular, the low concentration compounds in the mixtures (chlordiazepoxide and fluphenazine) were determined more accurately with precision lower than 3%. The strategy was used for the quality control of pharmaceuticals containing the mixtures without chemical pretreatment.
Acknowledgments
M. F. M. would like to thank the NODCAR, Egypt for the joint mission scholarship award. The authors declare no conflicts of interest.