Unravelling the molecular significance of Rock-Eval® S2 pyrograms from soil and geological samples

Abstract

Rock-Eval® pyrolysis is dedicated to rapidly determining the quality and quantity of organic matter (OM) in environmental and geological samples. It has been proposed that detailed examination of S2 pyrograms acquired from a Flame Ionization Detector during the programmed pyrolysis of samples can provide complementary information to classical Rock-Eval® parameters. Previous mathematical deconvolution of the S2 pyrogram has been suggested to semi-quantify distinct thermal fractions that are assumed to be distinct in terms of chemical and/or biological lability. To date, there is no direct molecular support for such an assumption. This work proposes a methodological approach based on coupling a temperature-programmed pyrolyser to a standard mass spectrometer (Py-MS). A set of reference soil samples was analysed. The set was completed by dissolved OM, source rock and coal samples in order to test the relevance of this methodological approach to various OM types. Data analysis was based on Tpeak values defined at temperatures of maximum m/z fragment production during the temperature ramp. These Tpeak values were grouped into clusters of distinct m/z values for which a chemical attribution is proposed owing to an extensive literature survey. For the soil samples, all the pyrolytic fractions described in the literature from mathematical deconvolution of the S2 pyrogram were identified except for the most thermolabile fraction (280–320 °C). Fragments with Tpeak values between 330 and 390 °C were mainly attributed to proteins, lignin and carbohydrates. Tpeak values found in the 391–499 °C range corresponded to m/z fragments attributed to aliphatics, lignin and aromatics while fragments with Tpeak values higher than 480 °C were assigned solely to aromatics. In all samples, these fragments displayed a singular pattern of decreasing m/z with increasing temperature, the significance of which remains to be fully elucidated. This preliminary study provides key methodological guidelines for re-exploring Py-MS applications to disentangle the chemical nature of OM.