应用近红外光谱的新光谱分析法

会议poster

New Approach for Spectroscopic Analysis Applied to Infrared Spectroscopy

Abstract

Methods such as Principal Component Analysis (PCA) to perform multivariate analyses on spectral and chromatographic data have been a mainstay of chemometrics for years. This poster describes a new method that combines cheminformatics tools with chemometrics tools for PCA in an intuitive environment for performing such analyses. A new patent-pending technology—Overlap Density Heatmaps (ODH)—now allows the comparative visualization of heretofore unheard of numbers of spectra or chromatograms. Overlap Density Heatmaps are used for visual data mining and analysis to assess the similarities and dissimilarities in large amounts of
spectral, chromatographic, and other graphical data. This new approach for spectroscopic analysis will be examined in specific case studies as applied to IR and Raman data. We will demonstrate the successful use of PCA and ODH to analyze a query, and the hit list resulting from an IR spectral search, as well as an overall analysis of a database.

Conclusions

Principal Component Analysis appears to be a valuable tool to analyze the results of standard spectral searches-a spectral query and hit list-providing useful insights into the nature of the compounds in the hit list relative to the query. Overlap Density (OD) Heatmaps not only confirm the value of the technique, but are also a useful complement to the multivariate processing capabilities afforded by PCA.  This technique is an excellent tool to identify components in mixtures and can be used effectively in the polymer industry.