Sample Preparation for Solid Petroleum-Based Matrices Based on Direct Matrix Introduction Oriented to Hydrocarbon Profiling

• Published in: Energy & fuels 2020-09, Vol.34 (9), p.10705-10712
• Main Authors: de Oliveira, Amilton Moreira, Castiblanco, Julián Eduardo Ballen, Fleming, Felipe Pereira, de Souza, Emanoel Nazareno Ramalho, Hantao, Leandro Wang
• Format: Article
• Language: English
• Subjects: Fossil Fuels
• ISSN: 0887-0624; 1520-5029
• DOI: 10.1021/acs.energyfuels.0c01613

Multicomponent mixtures may be considered complex because of the overwhelming number of sample constituents and their broad range of physical–chemical properties. Such features translate into challenges in resolving all components using separation techniques but also impact sample preparation, which is often underestimated in modern practices of oil industry. While the former may be effectively tackled using comprehensive two-dimensional gas chromatography (GC×GC), analyte mass discrimination remains overlooked. In the analysis of whole crude oils or heavy distillates, it is well known that pentatriacontane (C35) precipitates at room temperature in many solvents, such as carbon disulfide, n-hexane, and toluene. Therefore, solvent-based sample preparation methods are biased for analytes heavier than C35, which are found in unusual crude oils. Although, current methods using conventional thermal desorption are limited for C40 hydrocarbons, direct matrix introduction (DMI) may be used for sample introduction. In this work, we expanded the effective operating range of DMI for hydrocarbon analysis by GC. Two case studies were evaluated. First, an uncommonly heavy paraffinic oil fraction was characterized by DMI-GC to illustrate the effect of solubility-based discrimination during sample preparation. Second, we extended the benefits of DMI for the sample preparation of soil contaminated by oil spill, bypassing the need for solvent extraction. Relative standard deviations of 6.7–9.4% were observed for the measurement of n-alkanes, allowing for reliable qualitative analysis of weathered soils. Furthermore, calibration curves of n-alkanes from 6.25 μg g–1 to 100 μg mL–1 (Pearson coefficient, r 2 = 0.99) demonstrated the potential of the DMI-based method for quantitative analysis.