Stable hydrogen isotope ratios of individual hydrocarbons in sediments and petroleum

Abstract

Early research into the stable hydrogen isotopic compositions (δD) of petroleum involved bulk deuterium/hydrogen (D/H) measurements which, while providing some useful information, had to contend with the analysis of complex mixtures of hydrocarbons, and alteration resulting from the rapid exchange of nitrogen-, oxygen- and sulphur-bound hydrogen. The use of gas chromatography-isotope ratio mass spectrometry (GC-irMS) overcomes these problems by allowing the analysis of individual compounds containing only the most isotopically conservative aliphatic carbon-bound (C-bound) hydrogen. This project investigates the geochemical utility and reliability of compound-specific δD values, with the aim to better understand and exploit this analytical capability. To demonstrate the source diagnostic potential of compound-specific δD values, normal and branched alkanes extracted from series of immature bog-head coals (torbanites) were analysed. The torbanites contain immature organic matter predominantly from a single, freshwater algal source, i.e. Botryococcus braunii (B. braunii). The δD values of n-alkanes reflect the climate regime at the time of deposition of the torbanites, and vary mainly in response to the δD values of the source meteoric waters in their depositional environments. n-Alkanes from torbanites deposited at high latitude in a glacial climate are depleted in D by up to 70% relative to those from a torbanite deposited at low latitude under a tropical climate regime. Torbanites deposited in a mid-latitude region under cool-temperate conditions contain n-alkanes with δD values falling in between those of n-alkanes from tropical and glacial torbanites.The δD values of the n-alkanes also reflect their multiple source inputs. For example, a saw-toothed profile of n-alkane δD values in Australian torbanites is attributed to a dual-source system: a predominant B. braunii input, with a minor terrestrial plant input to odd-carbon-numbered n-alkanes in the range n-C20 [subscript] to n-C29 [subscript]. The δD values of n-alkanes and isoprenoids (pristane and phytane) differ significantly in two Permian torbanites from Australia, thought to be reflective of the offset between the δD values of their precursors in extant organisms. The torbanite data indicate that a biological δD signal has been preserved for at least 260–280 million years, extending the utility of δD values for palaeoclimate studies. To elucidate the effect of sedimentary processes on the δD values of petroleum hydrocarbons, three sedimentary sequences have been studied. These comprise one from the Perth Basin (Western Australia) and two from the Vulcan Sub-basin (northern Australia) covering a wide range of maturities, i.e. 0.53–1.6% vitrinite reflectance (Ro). The δD values of n-alkanes extracted from immature-early mature sediments (marine shales/siltstones and mudstones) are consistent with that expected of marine-derived n-alkyl lipids. The hydrocarbons become enriched in D with increasing maturity. The large (ca. 115%) biologically-derived offset between the δD values of n-alkanes and acyclic isoprenoids from immature sediments gradually decreases with increasing maturity, as the isoprenoids become enriched in D more rapidly than the n-alkanes. The D-enrichment in isoprenoids correlates strongly with Ro and traditional molecular maturity parameters.This suggests that H/D exchange during maturation occurs via a mechanism involving carbocation-like intermediates, which proceeds more rapidly with compounds containing tertiary carbon centres. Significant epimerisation of pristane and phytane coincides with their D-enrichment, suggesting that hydrogen exchange occurs at their tertiary carbons. A mechanism is proposed which can account for both H/D exchange and the epimerisation of pristane and phytane in the sedimentary environment. Pristane and phytane extracted from a post-mature sediment from the Paqualin-1 sequence are significantly enriched in D (ca. 40%) relative to the n-alkanes, indicating that D-enrichment persists at very high maturity, and is more pronounced for the regular isoprenoids than the n-alkanes. This supports the notion that H/D exchange causes the observed shift in δD values, rather than free-radical hydrogen transfer. The differences between the δD values of pristane and phytane show opposite trends in the Perth Basin and Vulcan Sub-basin sediments. In the Perth Basin, phytane is enriched in D relative to pristane, likely due to a dominant algal source. In the Vulcan Sub-basin, pristane is enriched in D relative to phytane, and thus is attributed to a lower relative input of algal organic matter. The variance of the δD values of pristane and phytane is generally consistent throughout the maturity range and provides evidence that pristane and phytane exchange hydrogen at similar rates. δD analysis of crude oils and condensates reservoired in the Perth Basin and Vulcan Sub-basin has been carried out to evaluate potential applications in oil-source correlation.The n-alkanes from crude oils and condensates are often more enriched in D than n-alkanes extracted from their supposed source rocks, and the oils also show relatively small differences between the δD values of n-alkanes and isoprenoids. These results suggest significant H/D exchange has occurred, implying that the liquids were generated from mature source rocks. A Perth Basin crude oil (Gage Roads-1) thought to be derived from a lacustrine/terrestrial source contains hydrocarbons that are significantly depleted in D relative to Perth Basin oils derived from a marine source, attributed to variability in the isotopic composition of marine and terrestrial source waters. δD values of n-alkanes from Vulcan Sub-basin crude oils and condensates are largely consistent with their prior classification into two groups: Group A, having a marine source affinity; and Group B, having a terrigenous source affinity. Some oils and condensates are suggested to be mixtures of Group A and Group B hydrocarbons, or Group A hydrocarbons and other as yet unknown sources. An exception is a former Group A oil (Tenacious-1) containing n-alkanes that are enriched in D relative to those from other Group A oils and condensates, attributed to mixing with another source of more mature hydrocarbons. The n-alkane δD profile appears to be indicative of source and sedimentary processes. One Perth Basin crude oil (Dongara-14) contains lower-molecular-weight n-alkanes that are depleted in D relative to higher-molecular-weight n-alkanes, attributed to a mixed marine/terrestrial source.Group A crude oils and condensates from the Vulcan Sub-basin display a ‘bowl-shaped’ profile of n-alkane δD values. An upward inflection in the n-alkane δD profile from n-C11 [subscript] to n-C15 [subscript] is suggested to represent the addition of D-enriched lower-molecular-weight n-alkanes from a more mature wet gas/condensate to an initial charge of lower maturity oil. Ultimately, this project has demonstrated that the δD values of individual petroleum hydrocarbons can be used to elucidate the nature of source organic matter and depositional environments. The preservation potential of lipid δD values is greater than previously thought, although it is clear that H/D exchange accompanying maturation can have a significant effect on the δD values of certain hydrocarbons. Thus, great care must be taken when interpreting δD values of individual hydrocarbons, particularly those derived from sediments of high thermal maturity.