Abstract :
[en] The successful analysis of complex mixtures requires the combination of highly efficient separation techniques with advanced detection technologies to provide individual structural information. The coupling of gas chromatography and mass spectrometry was one of the first steps in hyphenated method development. Nowadays, due to continuous hardware improvement, the term hyphenation has been extended to multidimensional separations and/or multiple detectors.
The objective of the thesis is to develop hyphenated methods to extend the range of applications and propose solutions to improve the characterization of specifically selected complex samples. The main hyphenated technique that was used is comprehensive two-dimensional gas chromatography coupled to time-of-flight mass spectrometry (GC×GC-TOFMS). Three thematic areas, presenting different levels and types of complexity were investigated: Forensic geotaphonomy with grave soil samples, lignocellulose biomass valorization with termite gut microbiota samples, and human monitoring for emerging organohalogen contaminants with blood samples. Data processing strategies were also developed in order to extract the relevant information from the large quantities of data produced. The different strategies included retention time alignment, comparative data processing, multivariate analysis using unsupervised learning algorithm such as principal component analysis (PCA), univariate analysis of variance (one-way ANOVA), and scripting for automated filtering of mass spectra.
All developed methods allowed to successfully improve the characterization of samples originating from the different areas. For the first time, and because of the high level of specificity of GC×GC-TOFMS, the presence of methyl-branched alkane compounds was highlighted in grave soil samples, allowing to develop a screening method that has the capacity to differentiate between soils at proximity of buried decaying pig carcasses and control soils. This strategy could complement other approaches and contribute to aid forensic geotaphonomy investigators. In the biomass valorization area, more than 300 compounds were isolated and partly identified in the 1µL fluid volume available from the termite gut. A comparative study demonstrated that the adaptation of the termite system to non-optimal carbon sources is reflected in the metabolite profile. These results demonstrated the potential interest to investigate metabolite profiling with state-of-the-art separation science tools, to contribute to a better understanding of how termites efficiently degrades lignocellulose. In the human monitoring area, the analysis of serum samples using GC×GC-TOFMS and scripting allowed the efficient reduction of the data matrix from thousands of detected signals to a few important dozens, including the environmental contaminants of interest present in human blood plasma. Such an approach reduced data processing and reviewing time by several orders of magnitude while maintaining a high degree of identification power. This exploratory approach could result in the identification of emerging toxicants and help regulation bodies to gather better information before they rule on emerging issues. GC with high resolution mass spectrometry (HRMS) was also used, using a selective approach to evaluate human exposure to dechloranes, a family of emerging halogenated flame retardants recently reported as significantly present in the environment. A comprehensive approach was also developed using GC×GC coupled to high resolution TOFMS (HRTOFMS). The developed methodology will facilitate the monitoring of these emerging dechlorane contaminants and analogues in future exploratory studies.
