Projection of multidimensional GC data into alternative dimensions—exploiting sample dimensionality and structured retention patterns

Abstract

Comprehensive multidimensional gas chromatography (GC×GC) is a powerful separation technique. One of the features of this technique is that it offers separations with more apparent structure than that offered by conventional one-dimensional GC (1-D GC). While some previous studies have alluded to this structure, and used structured retention patterns for some simple classifications, the topic of structured retention in GC×GC has not been studied in any great detail. Using the separation of fatty acid methyl esters (FAME) on both nonpolar/polar and polar/nonpolar column sets, the interaction between the separation dimensions and the sample dimensions is explored here. The GC×GC separation of a series of compounds is presented as a projection of the sample from sample space, a p-dimensional space with dimensions defined by the dimensionality of the sample, into separation space: for GC×GC, a two-dimensional plane passing through the sample space in an orientation defined by the separation conditions. Using this conceptual model and some a priori knowledge of the sample, it is shown how the image of the sample in the separation space can be used to construct an image of the sample in alternate dimensions, such as second dimension retention factor (²k) vs. chain length in the case of FAME. These projections into alternate dimensions should facilitate the interpretation of the complex patterns found within the GC×GC chromatogram for the identification and classification of compounds.