Ocena wpływu przygotowania próbek na uzyskiwane wyniki analiz pirolitycznych

Abstract

The aim of the work, the results of which are presented in the article, was to evaluate the influence of the method of sample preparation on the results of pyrolytic analyses: RockEval, Py-GC-FID (gas chromatography with FID detector) and Py-GC-IRMS (isotope mass spectrometry with gas chromatography). Mortars and ball mills are most often used to grind the samples. Three rock samples with a higher organic carbon content (shales) and five with a lower organic carbon content (including two samples of dolomites and three of anhydrites) were prepared. The rocks were homogenized and divided into three parts. Each part was ground: by hand in a mortar, in a ball mill for 5 minutes, and in a ball mill for 15 minutes. A total of 24 RockEval analyses, 48 Py-GC-FID analyses (two pyrolysis temperatures corresponding to RockEval pyrolysis conditions) and 24 Py-GC-IRMS analyses were performed. The grinding of the samples for the RockEval analyses is insignificant. The differences in the results seem to indicate the influence of the analysis error and the nature of the sample. For the Py-GC-FID methodology, the influence of milling on the results of desorption carried out at the temperature of 300ºC for most of the tested samples can be considered insignificant and negligible. At the temperature of 500ºC, various types of samples show some correlations, but they are insufficient to reject one of the methods of sample preparation. Therefore, it is important to analyze the results comprehensively, taking into account both the group composition and the distribution of pyrolysis products. In the case of the isotopic composition analyses (Py-GC-IRMS), also no differences in δ13C values related to different types of grinding samples can be found. Summarizing, all the results show a negligible influence of the method of grinding the samples on the results of pyrolysis analyses. Nevertheless, it is important that only one method of sample preparation is used for a separate sample series.