Generalized window factor analysis for selective analysis of the target component in real samples with complex matrices.

In chromatographic analysis of multicomponent real samples, peak overlapping, high level of noise and background are frequently encountered, making the qualitative and quantitative analysis difficult or even impossible. In this work, an algorithm named as generalized window factor analysis (GWFA) was proposed for quantitative analysis of the target components in the samples with complex matrices by gas chromatography-mass spectrometry (GC-MS). The theory and calculation of GWFA are just similar with the conventional window factor analysis (WFA), but the "window" is defined as the selected channels (mass-to-charge ratios) in the mass spectral dimension of the data matrix, instead of a continuous region in chromatographic dimension along the retention time. Therefore, the generalized window for a target component can be easily determined with the help of the mass spectrum. Then, the calculated mass spectrum can be obtained with the window and quantitative determination can be achieved with the help of the standard. Both simulated and experimental data were investigated with the proposed method. Whether or not a peak shift occurs during the test, accurate results were obtained from the overlapping GC-MS signals with high level of noise and background.