Slobodan Sasić1. 1. Analytical R&D, Pfizer Global Research and Development, Ramsgate Road, Sandwich, CT13 9NJ, UK. slsasic@yahoo.com
Abstract
PURPOSE: Raman chemical imaging of API was carried out to provide information on whether there is any structural difference between the commercial Xanax and Alprazolam tablets (the API content less than 1% w/w in both cases) in the batches with low and good recovery. MATERIALS AND METHODS: Raman mapping spectra were collected from the flattened surfaces of six tablets. The mapping spectra were analyzed by principal component analysis because the Raman signal of the API (Alprazolam) was not reliably detected from the raw spectra. The complexity of the obtained grey-scale score images was such that no information about the domain sizes of the API could be obtained and thus binarization was applied to simplify these images. RESULTS: It was found that reliable detection of the Raman signal of the API was only achieved after principal component analysis was employed with the mapping being facilitated by a surprising similarity between a high principal component loading and the spectrum of the API. The binarization was successful only if the outlying pixels in the score images are eliminated. SUMMARY: The final chemical images represent quantitative characterization of the domains of the API in the tablets in contrast to chemical images of tablets that have been reported so far in the literature which have usually been descriptive only. The abundance of Alprazolam in all six tablets of Xanax and Alprazolam, respectively, was very similar. The domain sizes were found to be below 75 mum in diameter for all the tablets analyzed.
PURPOSE: Raman chemical imaging of API was carried out to provide information on whether there is any structural difference between the commercial Xanax and Alprazolam tablets (the API content less than 1% w/w in both cases) in the batches with low and good recovery. MATERIALS AND METHODS: Raman mapping spectra were collected from the flattened surfaces of six tablets. The mapping spectra were analyzed by principal component analysis because the Raman signal of the API (Alprazolam) was not reliably detected from the raw spectra. The complexity of the obtained grey-scale score images was such that no information about the domain sizes of the API could be obtained and thus binarization was applied to simplify these images. RESULTS: It was found that reliable detection of the Raman signal of the API was only achieved after principal component analysis was employed with the mapping being facilitated by a surprising similarity between a high principal component loading and the spectrum of the API. The binarization was successful only if the outlying pixels in the score images are eliminated. SUMMARY: The final chemical images represent quantitative characterization of the domains of the API in the tablets in contrast to chemical images of tablets that have been reported so far in the literature which have usually been descriptive only. The abundance of Alprazolam in all six tablets of Xanax and Alprazolam, respectively, was very similar. The domain sizes were found to be below 75 mum in diameter for all the tablets analyzed.