| There is a need for the development of methods that can accurately and precisely quantify the components of solid-state mixtures without resorting to dissolution of the samples. One important area where there is such a need is in the pharmaceutical industry where the compounds of interest are polymorphs or optical isomers. Diffuse Reflectance Fourier Transform Infrared (DRIFT) Spectroscopy has been tested as a tool to achieve this objective. Solid-state mixtures employed in this project were prepared from the samples of cimetidine polymorphs and isomers of ephedrine hydrochloride. The work described here shows that very good results can be obtained using DRIFT when experimental conditions such as particle size and sample homogeneity are carefully controlled.; Quantitative analysis of these mixtures in the solid-state is a must because polymorphs or optical isomers have identical spectra when in solution. However, in the solid-state there are differences in the IR spectra due to the differences in intermolecular interactions, rather than to the differences in the molecules. In general, the differences in the IR spectra of these polymorphic and isomers are only due to a few bands shifting positions or perhaps a few bands changing shapes. Thus any quantitative method must be able to use what sometimes are subtle changes in the spectra data, for the basis of the determination.; The results presented here show that multivariate calibration techniques, in particular the partial least squares algorithm, can produce good calibration models from such data. Data were collected both in the mid and near IR regions of spectrum using diffuse reflectance sampling. The results indicate that the use of mid infrared data allow for the development of a more accurate/precise model than do near infrared data collected from the same sample set. (Abstract shortened by UMI.)... |
