Integration of an automated identification-quantification pipeline and statistical techniques for pyrolysis GC/MS tracking of the molecular fingerprints of natural organic matter

Pyrolysis gas chromatography/mass spectrometry is a potent tool for studying the molecular fingerprints of natural organic matter (NOM). With advances in analytical techniques, a pyrogram generally consists over a hundred pyrolysates, which increases the difficulty of interpreting the associated data. Here, we propose a systematic approach that includes an automatic peak identification and quantification pipeline and statistical techniques for the analysis of NOM. White oak (Quercus alba) and forest floor litter samples from a 48-week field decomposition study including sites along a soil moisture gradient were used to evaluate the applicability. An analysis of variance of the chemical classes indicated that the composition differed among sites, although a trend following the moisture gradient was not observed. Factor analysis of the pyrolysates clearly identified two decomposition stages in both types of decomposition. For the oak litter, 2,6-dimethoxy-phenol originating from syringyl lignin was dominant in the early stage, whereas for the forest floor litter, 2-methoxy-4-vinylphenol was enriched in the early stage, while 4-ethyl-2-methoxyphenol and 3-allyl-6-methoxyphenol were dominant in the later stage. These compounds originated from guaiacyl lignin, which suggested that guaiacyl lignin was relatively constant. The proposed approach provides a convenient and effective way to study the chemical composition of NOM.