Multivariate data analysis for Raman imaging of a model pharmaceutical tablet |
| |
Authors: | Lin Zhang Mark J Henson S Sonja Sekulic |
| |
Institution: | Analytical Research and Development, Pfizer Global R&D, Groton, CT 06340, USA |
| |
Abstract: | Spectroscopic imaging techniques provide spatial and spectral information about a sample simultaneously and are finding ever-increasing application in the pharmaceutical industry. Effective extraction of chemical information from imaging data sets is a crucial step during the application of imaging techniques. Multivariate imaging data analysis methods have been reported but few applications of these methods for pharmaceutical samples have been demonstrated. In this study, a bilayer model tablet consisting of avicel, lactose, sodium benzoate, magnesium stearate and red dye was prepared using custom press tooling, and Raman mapping data were collected from a 400 μm × 400 μm area of the tablet surface. Several representative multivariate methods were selected and used in the analysis of the data. Multivariate data analysis methods investigated include principal component analysis (PCA), cluster analysis, direct classical least squares (DCLS) and multivariate curve resolution (MCR). The relative merits and drawbacks of each technique for this application were evaluated. In addition, some practical issues associated with the use of these methods were addressed including data preprocessing, determination of the optimal number of clusters in cluster analysis and the optimization of window size in second derivative calculation. |
| |
Keywords: | Raman imaging PCA Cluster analysis DCLS Multivariate curve resolution Pharmaceutical tablet |
本文献已被 ScienceDirect 等数据库收录! |
|