Polymorphism is an important property in the quality control of pharmaceutical products. In this regard, partial least squares regression and the net analytical signal were used to build and validate a multivariate calibration model using diffuse reflectance infrared spectroscopy in the region of 900-1100 cm(-1) for the determination of the polymorphic purity of carbamazepine. Physical mixtures of the polymorphs were made by weight, from 80 to 100% (w/w) form III mixed with form I. Figures of merit, such as sensitivity, analytical sensitivity, selectivity, confidence limits, precision (mean, repeatability, intermediate), accuracy, and signal-to-noise ratio were calculated. Feasible results were obtained with maximum absolute error of 2% and an average error of 0.53%, indicating that the proposed methodology can be used by the pharmaceutical industry as an alternative to the X-ray diffraction (United States Pharmacopoeiamethod).
Copyright 2004 Wiley-Liss, Inc. and the American Pharmacists Association J Pharm Sci 93:2124-2134, 2004