Abstracts Chemistry

Second order chemometric methods and the analysis of complex data

by Cliona Mary Fleming

Samples that are analyzed in today's laboratories are becoming increasingly complex, in that they contain larger, more complex analytes than ever before, in addition to a huge number of such analytes. This has provided a challenge to analytical chemists, in that characterization and analysis of these samples is much more difficult than with the traditionally simple samples that were analyzed. The information content of these samples is much greater than before, and this has led to the use of "hyphenated" or "second order" instruments that produce more data that can correspondingly provide a greater information source. However, extracting the necessary information from this second order, "real world" data is a challenge that results from these analyses.The problem that exists with second order chemometric methods is that they are generally developed and tested on relatively simple problems that contain just two or three components. While the results that are produced are impressive, they do not always extend to more complex mixtures that contain more components. Since many "real world" analyses fit the latter description, second order analyses have not been applied in these cases before now. The aim of this research, therefore, was to develop methods that would enable second order chemometric techniques to be applied to complex "real world" data, so that the full power of hyphenated analytical techniques could be realized. In this presentation, some examples from the chemical and biotechnological industries will be described in order to show (a) how and why the samples are so complex, and (b) how chemometric method can be used to aid in their analysis. The types of data analyzed will be LC/MS and comprehensive two-dimensional GCxGC data.