Doubly substituted isotopologues of methane hydrate (¹³CH₃D and ¹²CH₂D₂): Implications for methane clumped isotope effects, source apportionments and global hydrate reservoirs

Abstract

Recently developed methane clumped isotope techniques provides fresh and novel insights into methane biogeochemistry, which have been unobtainable through other techniques such as conventional stable isotope determinations and molecular composition measurements of hydrocarbons. Nonetheless, the governing processes and mechanisms which control the clumped isotope signatures (Δ¹³CH₃D and Δ¹²CH₂D₂) of natural methane samples remain an active area of investigation. Here, we present paired clumped isotope measurements in methane hydrate, which is a major methane reservoir widely distributed along the continental margins and which plays an important role in the global carbon cycle and climate system. Our study aims to shed new light into the fundamental processes of methane clumped isotope effects and their potential to answer fundamental questions regarding the source and migration of methane found in naturally occurring gas hydrate accumulations.

Gas hydrate samples were recovered from five shallow marine sediment sites on the eastern margin of the Japan Sea and most of them present Δ¹³CH₃D and Δ¹²CH₂D₂ temperatures ranging from ∼15 to ∼170 °C that apparently match expected methane formation temperatures. The distribution of these clumped isotope signatures along the equilibrium line is best explained by the mixing effect of equilibrated thermogenic methane formed at temperatures of 165 ± 15 °C and biogenic methane equilibrated at 1–2 °C, which may result from slow methanogenesis, anaerobic oxidation of methane (AOM), or their combination. The influences of gas migration/diffusion, hydrate formation and dissociation on Δ¹³CH₃D and Δ¹²CH₂D₂ values are insignificant. By combining clumped isotope results with other traditional approaches, a thermogenic and two microbial end-members as well as their isotopic compositions were identified and the relative contribution of each end-member was also quantified. The results not only demonstrate the applicability of methane clumped isotope data to identify potential end-members in natural methane samples, but also reveal that more conventional carbon isotope approaches may significantly underestimate the fraction of thermogenic methane present in global gas hydrate reservoirs. Improvements in the accuracy of source apportionment enable us to better understand the formation history and mechanisms of gas hydrate accumulation, as well as the role played by gas hydrate dissociation in past geological events. The estimated formation temperatures of thermogenic end-member can be further applied in reconstruction of the paleo geothermal gradient at the time when the thermogenic methane was formed at marine sedimentary environment.

Introduction

As a major energy source and significant greenhouse gas, methane (CH₄) plays an important role in the global carbon cycle. Quantifying the formation pathways of this gas will help us better constrain the global methane flux and its influence on global climate (Nisbet et al., 2014, Schaefer et al., 2016, Schwietzke et al., 2016). Given that methane is a crucial component in the early Earth’s ecosystems and climate (Ueno et al., 2006, Sauterey et al., 2020), a full understanding of the present-day methane cycle may enable us better decipher processes associated with early earth history.

Naturally occurring methane may be simply categorized as a mixture of gas from three principal sources (Etiope and Sherwood Lollar, 2013, Stolper et al., 2018): ‘thermogenic CH₄’, yielded from thermal-breakdown of large hydrocarbon molecules and other organic matter; ‘microbial CH₄’, produced by microbial methanogens; and ‘abiotic CH₄’, formed by chemical reactions in the absence of any organic matter and biotic activity. Methods for identifying CH₄ sources, such as the ‘δD vs. δ¹³C space’ (so called ‘Whiticar plot’; Whiticar, 1999) and ‘δ¹³C vs. C₁/(C₂ + C₃) space’ (‘Bernard plot’; Bernard et al., 1976), were developed and have been widely employed. However, obvious shortcomings exist in these methods, such as the partial overlap of isotopic signatures from different CH₄ sources, difficulties in constraining CH₄ generated from two or more sources, and the alteration of isotopic compositions via post-generation processes such as gas migration and microbial methanotrophy (Douglas et al., 2017, Milkov and Etiope, 2018). These limitations have provided an impetus for developing additional constraints which can improve on interpretations regarding methane sources and histories. The focus of this investigation is to demonstrate the potential application of methane clumped isotope techniques to refine our understanding of such processes.

The clumped isotope method, which refers to two or more rare isotopes in one isotopologue, was firstly applied in carbonate minerals as a paleo-thermometer (e.g. Ghosh et al., 2006). The first report of methane clumped isotopes in natural samples employed a prototype ultra high resolution isotope ratio mass spectrometry (Thermo Scientific 253 Ultra; hereafter ‘Ultra’) established in the California Institute of Technology (CalTech; Stolper et al., 2014), which can only detect the combined abundance of ¹³CH₃D and ¹²CH₂D₂ isotopologues. Meanwhile, another method for measuring the ¹³CH₃D isotopologue was developed by a research group in the Massachusetts Institute of Technology (MIT) which uses a tunable infrared laser direct absorption spectroscopy (TILDAS) (Ono et al., 2014). By 2016, a method for the independent determination of both ¹³CH₃D and ¹²CH₂D₂ isotopologues with satisfactory precision was achieved using a larger-radius high resolution isotope ratio mass spectrometry (Nu Instruments Panorama IRMS) at the University of California, Los Angeles (UCLA) (Young et al., 2016, Young et al., 2017). Recently, measurements of ¹³CH₃D and ¹²CH₂D₂ isotopologues in the same CH₄ sample have also been successfully carried out using updated versions of the Ultra (Eldridge et al., 2019, Dong et al., 2020, Dong et al., 2021, Thiagarajan et al., 2020) and TILDAS (Gonzalez et al., 2019) systems.

Unlike conventional isotopic determinations (δ¹³C and δD), methane clumped isotope signatures are defined as abundances of multiply substituted mass-18 isotopologues (¹³CH₃D and ¹²CH₂D₂) relative to the stochastic distribution of each isotope in the same sample, and the notations are Δ₁₈ (combination of ¹³CH₃D and ¹²CH₂D₂), Δ¹³CH₃D and Δ¹²CH₂D₂ (Young et al., 2016, Young et al., 2017, Douglas et al., 2017). Clumped isotope values may primarily record formation or equilibration temperatures if the samples which are analyzed present intra-species coincident temperatures (i.e., Δ¹³CH₃D and Δ¹²CH₂D₂ values located along the curve of thermodynamic equilibrium). Alternatively, disequilibrium clumped isotope signatures might be a proxy for identifying kinetic isotope processes during methane generation and other post-generation processes such as mixing, migration, oxidation, bond re-ordering (Wang et al., 2015, Young et al., 2017, Giunta et al., 2019, Giunta et al., 2021, Labidi et al., 2020, Warr et al., 2021). Thus, disequilibrium clumped isotope signatures can potentially provide additional information for interpreting the cycling of methane in natural systems.

Previous studies of the clumped isotope signatures of thermogenic gas have presented reasonable and apparently consistent temperatures in most cases (Douglas et al., 2017, Young et al., 2017, Stolper et al., 2018, Giunta et al., 2019, Giunta et al., 2021, Xie et al., 2021), indicating that they can record temperatures related to either methane formation or post-bond-re-ordering. Nonetheless, disequilibrium values have also been observed in some laboratory samples which were thermogenically produced both from the pyrolysis of shale and coal (Shuai et al., 2018), and hydrocarbons (Dong et al., 2021). The disequilibrium signal may be possibly derived from a statistical combinatorial effect and/or from a kinetic isotope effect (KIE) occurring during thermal-cracking in the laboratory and it is possible that natural hydrocarbon systems may approach equilibrium during more long-term maturation processes. Additional investigations of natural thermogenic samples are necessary to further improve these isotopic models (Xia and Gao, 2019, Dong et al., 2021, Xie et al., 2021).

Biogenic CH₄ is more complex and presents variable clumped isotope signatures. Natural biogenic CH₄ gases can be produced via various microbial methanogenesis pathways, such as hydrogenotrophic, acetoclastic and methylotrophic methanogenesis (Conrad, 2020, Kurth et al., 2020). To date, microbial CH₄ gases produced in laboratory cultivation experiments have consistently yielded clumped isotope compositions lower than the expected equilibrium values for the respective culture temperatures, regardless of their production pathways (Stolper et al., 2015, Wang et al., 2015, Young et al., 2017, Gruen et al., 2018, Giunta et al., 2019, Taenzer et al., 2020, Douglas et al., 2020). Dramatic depletion in ¹²CH₂D₂ (Δ¹²CH₂D₂: from −60‰ to −10‰) is mainly ascribed to the quantum tunneling and/or combinatorial effect (Young et al., 2017, Young, 2019, Cao et al., 2019, Taenzer et al., 2020), while relatively mild depletion in ¹³CH₃D (Δ¹³CH₃D: from −4‰ to +4‰) is probably produced by quantum tunneling and/or KIE which scales to the rate of methanogenesis (Wang et al., 2015, Young, 2019, Cao et al., 2019, Douglas et al., 2020). Additionally, Lloyd et al. (2021) suggested that even lower values associated with methylotrophic methanogens when compared to hydrogenotrophic methanogenesis may be partly inherited from methyl substrates depleted in both ¹³CH₂D and ¹²CHD₂.

Disequilibrium clumped isotope signatures have also been widely observed in natural biogenic CH₄ samples, such as those collected from cow rumen and freshwater ecosystems (Stolper et al., 2015, Wang et al., 2015, Douglas et al., 2016, Young et al., 2017, Douglas et al., 2020). It is however interesting that clumped isotopes of methane recovered from marine sediments, which are supposedly of biogenic origin, typically exhibited reasonable formation temperatures which are reasonably consistent with the local environment (Wang et al., 2015, Inagaki et al., 2015, Young et al., 2017, Ijiri et al., 2018). Mechanisms underlying the equilibrium of methane clumped isotopes naturally occurring in marine sediments are still debatable. Early studies argued that equilibrium clumped isotope signals in biogenic CH₄ are the product of low methanogenesis rates combined with high enzymatic reversibility, according to theoretical models of KIE at different degrees of reversibility (Wang et al., 2015), and this has been supported in part by the cultivation experiments and natural observations at much greater rates of methanogenesis (Stolper et al., 2015, Wang et al., 2015, Douglas et al., 2016, Douglas et al., 2020). However, these predictions remain to be tested (by field-based measurements for example) since the direct observations of clumped isotope signatures related to sufficiently low rates of methanogenesis in the laboratory seem rather difficult at this moment. Alternatively, microbial methanotrophy including aerobic and anaerobic oxidation of methane, can alter initial clumped isotope signatures of methane as well (Wang et al., 2015, Ash et al., 2019, Young, 2019, Ono et al., 2021). In particular, reversible bond re-ordering during anaerobic oxidation of methane (AOM) is proposed to be the key process leading to near-equilibrium values based on observations in natural samples (Ash et al., 2019). Preliminary laboratory culture experiments (Young, 2019, Ono et al., 2021) qualitatively support the role of AOM in altering methane clumped isotope signatures, but some of these experimental results do not quantitatively agree with observations in natural samples. For instance, Ono et al. (2021) observed Δ¹³CH₃D values in residual CH₄ to be as much as 3.1‰ higher than the expected equilibrium values which may be related to KIE derived from fast AOM rates during AOM incubation, while Young (2019) reported different bond re-ordering trends in the Δ¹³CH₃D vs. Δ¹²CH₂D₂ space at different sulfate concentrations. In any case, the detailed mechanism of AOM catalyzed bond re-ordering needs deeper investigation.

The application of the methane clumped isotope approach to the accurate source identification and temperature reconstruction of natural samples requires further resolution of the aforementioned uncertainties and inconsistencies. To improve our knowledge of this novel method, especially in the marine sedimentary system, further laboratory and field studies are required. Here we suggest that gas hydrates recovered from marine sediments, which consist primarily of methane-water clathrate, are ideal materials to potentially provide new insights into the fundamental mechanisms involved in clumped isotope effects. The formation and preservation history of gas hydrates can be fairly complex and a number of processes which potentially influence methane clumped isotope compositions can be evaluated, such as: the mixing of various methane sources (Snyder et al., 2020b), the migration from deep sources to the surface reservoir (Hachikubo et al., 2015), the diffusion or migration between shallow sediment reservoirs (Ginsburg and Soloviev, 1997), the repeated cycling between gas hydrate formation and dissociation (Snyder et al., 2020b), and the potential for bond re-ordering to be catalyzed by AOM (e.g. Ash et al., 2019) and/or by clay minerals (e.g. Giunta et al., 2021). Several preliminary Δ¹³CH₃D results for microbial hydrates recovered from Northern Cascadia Margin have been reported by Wang et al. (2015), presenting equilibrium temperatures which appear to be comparable to the sampling environment, yet the mechanisms leading to equilibrium are still under debate. In this study, we further extend the investigations using the paired clumped isotope approach to include gas hydrates recovered from shallow sediments in the Japan Sea. The common occurrence of AOM in the Japan Sea shallow marine sediment (Matsumoto et al., 2009, Hiruta et al., 2009, Tomaru et al., 2012), existence of both thermogenic and microbial methane sources (Hachikubo et al., 2015, Snyder et al., 2020b), and upward/advective flow of gas-saturated fluid in the gas chimneys and surrounding sediments render this region a natural laboratory for carrying out the above-mentioned study as many ambient parameters have been well documented (Monzawa et al., 2006, Snyder et al., 2007, Snyder et al., 2020a, Snyder et al., 2020b, Okui et al., 2008, Watanabe et al., 2008, Hiruta et al., 2009, Machiyama et al., 2009, Nguyen et al., 2016, Aoki et al., 2018, Zhang et al., 2019).

The paper is organized as follows. 2 Geological and geochemical context of sampling sites, 3 Materials and methods describe details in sampling sites and analytical methods, respectively. In Section 4, isotopic compositions of methane hydrates are reported. The underlying mechanisms for the observed isotopic compositions are discussed in 5.1 Bulk stable isotopes and molecular composition of gas hydrates, 5.2 Quasi-equilibrium signatures of methane clumped isotopes in gas hydrate samples. Using the observed results, we estimate the isotopic compositions of thermogenic and biogenic end-members and their contributions to our methane hydrate samples (Section 5.3). We then briefly discuss the implications for methane clumped isotope systematics and their future applications in methane biogeochemistry in Section 6.