On-line application of the orthogonal projection approach (OPA) and the soft independent modelling of class analogy approach (SIMCA) for the detection of the end point of a polymorph conversion reaction by near infrared spectroscopy (NIR)

The dissolution of the polymorph P1 in an organic solvent, results in the conversion of P1 into P2. This process was performed at a constant temperature and followed by measuring near infrared (NIR) spectra on-line in the reaction vessel. The successive recorded spectra are ordered in function of the reaction time into a data matrix. Newly measured spectra are classified in this data set according to their reaction time. In this way an increasing data matrix is obtained. The objective was to follow the reaction progress and to detect the end point of the conversion process on-line. The application of the orthogonal projection approach (OPA) on the increasing NIR data set seems to achieve the stated aim. Replacing the mean spectrum by the spectrum of the starting product P1, to plot the dissimilarities with the newly recorded spectra, might result in a more sensitive method because the reference vector is not influenced anymore by newly recorded spectra. Nevertheless the same conclusions were obtained in both cases. Cumulative dissimilarity plots seems also useful to determine the end point of the conversion. The soft independent modelling of class analogy approach (SIMCA) succeeds in the detection of the start and the end point of the reaction if only the wavelength region is considered in which the spectra have significant differences in the shape of the absorption bands. (C) 1999 Elsevier Science B.V. All rights reserved.