CHEMOMETRICS AND DSI-GC-MS METHOD FOR PETROLEUM SOURCE-ROCKS ANALYSIS

Pyrolysis coupled with gas chromatography/mass spectrometry (Py-GC/MS) has been utilized in various fields. In the petroleum exploration field, it is possible to simulate the conversion of kerogen into bitumen and the subsequent generation of gas and oil—a crucial process for studying the maturation of hydrocarbon source rocks. The small-scale simulation enables precise experimental control of variables, such as temperature and pressure, which is crucial for obtaining reproducible data. Moreover, it enables the real-time analysis of the composition of generated products. The source rock composition and geochemical indices are essential for studying the maturation level and the potential of petroleum generation. Direct sample introduction (DSI) is a potential approach for analyzing solid samples without solvent extraction, utilizing a Py-GC-MS system. However, there are few studies aimed at optimizing the methods currently used, and the need for improvement is evident. This work aimed to evaluate different temperatures and their influence on the volatilization and pyrolysis of the organic matter of a source rock sample in the DSI-GC-MS and the application of chemometric methods to process the data. The source rock was subjected to different temperatures (100°C-500°C) with an unheated/original sample serving as a control. Approximately 40 mg of each grounded (80 mesh) sample was weighed into a deactivated stainless-steel cup. Analyses were performed using a Frontier-Lab EGA/PY-3030D pyrolyzer, coupled to a Shimadzu GC/MS (GC-2010Plus/MS-QP2020). An HP-PONA column (50 m x 0.20 mm x 0.5 µm) was used. The oven was programmed from 30°C (5 min) to 320°C at 15°C min^-1 (40 min hold), using a 1:25 split and helium at 1.05 mL min^-1. The MS transfer line and ion source were kept at 310°C and 250°C, respectively, with data acquired with a scan of m/z 35-500, 70eV. Results showed that increasing temperature led to greater expulsion of overall hydrocarbons (HC) from the rock. However, the classes of HC that increased varied. Classes such as cyclic HC showed a decrease, indicating that higher temperatures led to their cracking prior to their volatilization from the rock. Some parameters were analyzed using a single ion monitor (SIM) method and it was possible to determine geochemical indices Pristane/n-C₁₇, Phytane/C₁₈, Pristane/Phytane, C₁₉/C₂₃, C₂₂/C₂₁, C₂₄/C₂₃, C₂₆/C₂₅, C₂₄/C₂₆ for the temperatures of 200, 250, 300, 450°C. The temperature of 300°C appears to be optimal for an analytical approach for volatilization phenomena, while temperatures above could lead to the cracking of some HC classes. However, temperature and matrix effects appeared to affect peak resolution. Towards that, a graphical interface was developed in a python environment using chemometrics to assist the user on the data treatment and deconvolution of peaks. This tool optimizes analytical workflows by including features such as automatic peak identification, area integration with adjustable parameters, and peak deconvolution using the Multivariate Curve Resolution (MCR-ALS) method. In general, the DSI-GC-MS method demonstrated good repeatability in analyzing hydrocarbons and assessing the different phenomena related to the organic matter within the rock sample, optimizing information for a systematic method development and further analysis of different rocks with different COT, kerogen and maturity in the future.