2024 Volume 58 Issue 4 Pages 138-154
Kerogen density fractions separated by the sequential density centrifugation of kerogens from Cretaceous Leenhardt and Miocene Onnagawa shales were analyzed using pyrolysis-GC/MS. The pyrolysates from both kerogens mainly comprised aliphatic compounds, such as n-alkane/n-alkene doublets and acyclic isoprenoid hydrocarbons, along with aromatic compounds including alkylbenzenes, alkylnaphthalenes, and alkylphenanthrenes. The sporomorph-concentrate obtained from the Onnagawa kerogen was mainly composed of sporopollenin geomacromolecules, consistent with the dominances of long-chain (C22–C25) n-alkane/n-alkene doublets. In the algae/sporomorph-abundant fraction from the Leenhardt kerogen, mid-chain-length (C16–C20) n-alkane/n-alkene dominances suggested a larger contribution of dinoflagellate-derived algaenan. Pyrolysis of the weakly fluorescent amorphous kerogen (WFA)-containing fractions of the Onnagawa kerogen generated C13–C20 acyclic isoprenoid hydrocarbons (except C17), as well as alkylbenzenes including 1,2,3,5- and 1,2,3,4-tetramethylbenzenes. The isoprenoid hydrocarbons and the alkylbenzenes were components of fossil geomacromolecules likely derived from marine microalgae (diatoms and cyanobacteria). Non-fluorescent amorphous kerogen (NFA), the major kerogen of the Leenhardt shale, contained terrigenous suberin geomacromolecules, consistent with the dominance of terrigenous organic matter in the Leenhardt kerogen and the abundance of long-chain even carbon-numbered (C22 and C24) n-alkanes. The distributions of the pyrolytic isoprenoid alkanes (C13, C16, and C18–C19) from the Leenhardt kerogen suggested their distinctive fossil geomacromolecular sources, such as freshwater/coastal microalgae and cyanobacteria.
Kerogen is the insoluble organic matter (OM) preserved in geological sedimentary rocks. In addition to serving as the precursor material of petroleum and gas (Tissot and Welte, 1984; Vandenbroucke and Largeau, 2007), kerogen is a substantial organic carbon reservoir and a carbon pool for biogeochemical systems, including carbon material cycling (Berner, 1989; Vandenbroucke and Largeau, 2007). Kerogen may be structural, consisting of biological organic remains such as plant debris, spores, resin, and algal resting cyst, or non-structural (i.e., amorphous kerogen). Structural kerogen was formed through the ‘selective preservation’ of the refractory components of biological organisms during diagenesis (Philp and Calvin, 1976; Hatcher et al., 1983), whereas amorphous kerogen is primarily the product of a series of recondensation and polymerization reactions of labile biological OM, such as the conversion of amino acids and sugar to melanoidin via the Maillard reaction (Tissot and Welte, 1984). Thus, the ‘geomacromolecule’ comprising amorphous kerogen is the refractory OM generated after the decomposition of labile biological OM in the water column and sediment (or even within an organism) during early diagenesis (e.g., Kohnen et al., 1991; Gupta et al., 2007). Within the decomposed OM of amorphous kerogen, the presence of thin, lamellar structures consistent with the outer walls of algae and bacteria (Flaviano et al., 1994) indicates the selective preservation of algal and bacterial refractory OM, such as algaenan and bacteran (Tegelaar et al., 1989; Largeau et al., 1990; Derenne et al., 1991). Variability in the relative proportions of the amorphous and structural OM in kerogens considerably influences the geochemical characteristics of bulk OM. Further studies are needed to allow accurate interpretations of geochemical signals with respect to the source rock; the biogeochemical processes that contribute to carbon reservoir formation; and the reconstruction of the paleoclimate, paleoenvironment, and paleoecosystems.
Individual kerogen constituents, including amorphous and structural OM, exhibit various densities because of differences in their biological origins, enabling their isolation by density separation. The separation of macerals and other coal constituents has been described in several studies (e.g., Dyrkacz and Horwitz, 1982; Karas et al., 1985; Crelling et al., 1992; Nip et al., 1992; Yan et al., 2019; Nag et al., 2021). Karas et al. (1985) separated coal macerals by density gradient centrifugation, although the sink-float technique has also been successfully utilized (e.g., Kinghorn and Rahman, 1983; Yan et al., 2019). However, few studies have examined the density separation of kerogens, especially type II kerogens, in marine sediments (Stankiewicz et al., 1994, 1996; Sawada and Akiyama, 1994; Hartgers et al., 1995; Mastalerz et al., 2012; Xie et al., 2020).
Knowledge of the chemical structure of kerogens can provide a better understanding of the biological source(s), depositional environment, and diagenetic process(es) of sedimentary OM. Pyrolysis (py)-gas chromatography (GC)/mass spectrometry (MS) has been used in the structural investigation of kerogen (e.g., Larter et al., 1978; Hartgers et al., 1994a; Gelin et al., 1997; Biller et al., 2015), but its applications have been limited to investigation of the presence (or absence) of the chemical building blocks that contribute to kerogen; the chemical structure of kerogen was not determined (Vandenbroucke and Largeau, 2007; Lis et al., 2008). The chemical building blocks of kerogen are mainly composed of fossil geomacromolecular organic substances derived from biotic sources and partly altered by diagenetic processes. However, py-GC/MS can also be used to obtain information regarding the molecular nature and characteristics of those substances.
In the present study, kerogens from Cretaceous and Miocene organic-rich marine shales were subjected to density separation followed by py-GC/MS analysis of the resulting fractions. The fossil geomacromolecular sources of the identified kerogens and thus the abilities of the pyrolytic compounds to serve as biomarkers were also explored.
The kerogen samples examined in this study were separated from the Miocene Onnagawa Formation Shale On-7 (Sawada and Akiyama, 1994; Ogata et al., 2013) and from the Cretaceous Leenhardt Black Shale Leen-1 (Okano et al., 2008). Their total organic carbon contents and their carbon (C), hydrogen (H) and oxygen (O) ratios are presented in Table 1, along with maturity data based on the vitrinite reflectance (Ro) and the isomer ratio of C32 hopanes [22S/(22S + 22R)] (Sawada and Akiyama, 1994; Okano et al., 2008).
Data for TOC, CHO ratios, and maturity indices (Ro) and hopane isomer ratio (C32 hopane 22S/(22S + 22R)) in Onnagawa (On-7) and Leenhardt (Leen-1) shales
| Sample No. | Formation | Level | TOC (%) | Ro (%) | H/C | O/C | C32 Hopane S/(S + R) |
|---|---|---|---|---|---|---|---|
| On-7 | Onnagawa Formation | — | 2.77 | 0.45 | 1.22 | 0.11 | — |
| Leen-1 | Marnes Bleues Formation | Leenhardt Level | 3.60 | 0.58* | 0.98 | — | 0.56 |
*: Ro value estimated from hopane isomer ratio.
The On-7 sample was collected at 1794 m depth in the Yabase R-1 well, Akita, Japan (Sawada and Akiyama, 1994). The Onnagawa Formation, deposited in the Akita basin during the middle to late Miocene (e.g., Tada, 1991), consists of siliceous and diatomaceous sediments, as well as mudstone with a high OM content (Waseda et al., 1995; Asahina et al., 2022). According to the CHO ratios (H/C = 1.22, O/C = 0.11), the Onnagawa Formation Shale On-7 kerogen is a type II kerogen (Sawada and Akiyama, 1994). Its Ro is 0.45%, corresponding to the subbituminous coal in a coal rank (Sawada and Akiyama, 1994).
The Leen-1 sample was a laminated black shale from the Leenhardt Level in the Marnes Bleues Formation, deposited during the Cretaceous Albian Oceanic Anoxic Event (OAE) 1b. It was collected at the outcrops of Sauzeries, Southern France. Its C32 22S/(22S + 22R) hopane ratio is 0.56, and its Ro ca 0.58% correspond to the subbituminous to high-volatile bituminous coals in a coal rank (Okano et al., 2008).
Kerogen was separated using the classical method (Durand and Nicaise, 1980; Sawada and Akiyama, 1994). Briefly, 10 g of rock crushed to the size of rice grains (diameter: 2–5 mm) were extracted for 24 h in a Soxhlet apparatus with 100 ml of dichloromethane/methanol (3/1 v/v). The residues were then sequentially treated in a water bath shaker as follows: HCl 6 M (100 ml, 60°C, 12 h), HCl 12 M/HF 46% (1/1 v/v) (100 ml, 60°C, 24 h), and HCl 6 M (100 ml, 60°C, 4 h). After each treatment, centrifugation (3000 rpm, 10 min) was performed and the supernatant was removed. The residue, consisting of kerogen, was sequentially washed with HCl 6 M (2×) and distilled water (5×), and the recovered material was freeze-dried under vacuum.
Density separation of kerogensAfter the removal of framboidal pyrites with a solution of NaBH4 (Sawada and Akiyama, 1994), the kerogen samples were separated into their individual constituents via sequential density centrifugation (SDC), which combines the sink-float technique and sequential centrifugation (Sawada and Akiyama, 1994; Sawada, 2006). Briefly, ZnBr2 (~1.80 g/cm3) was used to prepare solutions with various specific gravity (S.G.) properties, obtained by mixing with dilute HCl (S.G. 1.20–1.80). Bulk kerogens (10–20 mg) were first separated into float and sink fractions using a solution with S.G. 1.80. The sink fraction was eliminated because it contained large amounts of heavy inorganic contaminants (e.g., FeS2) resistant to the acid treatment used in the initial kerogen preparation. The float fraction was further subdivided into several density fractions, then centrifuged for 20 min at 3000–4000 rpm in a high-speed refrigerated centrifuge (model 6000 KUBOTA Co.). The float fraction and the supernatant were transferred to centrifugal glass tubes, diluted, washed using dilute HCl, and recovered as density fractions. These density fractions were individually quantified based on their volumes using a laboratory-made centrifugal volumeter (Fig. S1). Petrographic examination of the kerogen compositions in the density fractions was carried out using transmitted-light and fluorescence microscopy, as described below.
Microscopic observation of kerogensThe kerogen composition (palynofacies analysis) was analyzed as described by Sawada (2006). Briefly, wet residual particles obtained after HCl/HF treatment were placed dropwise onto non-fluorescent slides, then mounted on permanent slides for microscopic examination. The OM composition of kerogen was determined by counting 500 points at each 100-μm interval under an Olympus BX41 reflected light fluorescence microscope equipped with an Olympus ULH100HG mercury lamp, a DM400 dichroic mirror containing a 330- to 385-nm excitation filter, and a 420-nm-long pass barrier filter. The emitted light was observed at ×200 magnification.
The various types of amorphous OM were distinguished by fluorescence light microscopy (Sawada and Akiyama, 1994; Sawada, 2006). Amorphous OM was further identified based on its morphology under transmitted-light microscopy, as reported by Tyson (1995), Tribovillard et al. (2001), Sawada (2006), Ercegovac and Kostic (2006), and Sawada et al. (2012). The morphological and fluorescence characteristics of the different organic particle types of kerogens, and the possible sources of those particles, are detailed in Table S1.
Pyrolysis-GC/MS of kerogensPy-GC/MS was performed as described by Ogata et al. (2013). The on-line system consisted of a Curie point pyrolyzer JCI-22 (Japan Analytical Industry Co.) with a Hewlett Packard 6890N capillary GC (60 m × 0.25 mm i.d. VF-200 ms column, Agilent J&W) directly coupled to a Hewlett Packard inert XL MSD (quadrupole mass spectrometer, electron voltage 70 eV, emission current 350 μA, mass range m/z 50–600 in 1.3 s). The pyrofoil (i.e., heat source for the Curie point pyrolyzer) has ferromagnetic properties (Japan Analytical Industry Co.). Five hundred micrograms of kerogen sample were placed on 9-mm-wide pyrofoils (F590; nickel (Ni)-cobalt (Co) alloy) and heated in the pyrolyzer for 15 s at 590°C. The generated compounds were transferred to the GC splitless injection system and heated at 320°C using helium gas as follows: 50°C for 4 min, 50–320°C at 5°C/min, and 320°C for 17 min (solvent delay: 6 min). Compounds were identified on the basis of their mass spectra and relative retention times after comparison with library data (NIST14) and the literature.
Figure 1 shows the compositions of various kerogens isolated from the bulk kerogen samples obtained from Onnagawa and Leenhardt shales, as determined by transmitted-light and fluorescence microscopy. Detailed descriptions are provided in Table S1. The Onnagawa kerogen was mainly composed of weakly fluorescent amorphous kerogen (WFA; amorphous OM originating from marine algae), a major OM type present in the Miocene pelagic sediments of the paleo-Japan Sea from the Akita and Niigata areas (Sawada and Akiyama, 1994; Omura and Hoyanagi, 2004). It is likely that this kerogen, and therefore the WFA, mainly originated from marine diatoms (Suzuki et al., 1993). Minor components in the Onnagawa kerogen were wood (woody and coaly kerogen), herbaceous OM such as sporomorphs (spore and pollen) and cuticle (from higher plants), and fluorescent amorphous OM (very fine debris from cuticles, spores, and pollen). The Leenhardt kerogen was mainly composed of non-fluorescent amorphous kerogen (NFA; originating from the woody tissue of higher plants and other terrestrial organisms) and WFA. A small portion consisted of wood and sporomorphs. The terrigenous OM, wood, and NFA together constituted >75% of the kerogens in the Leenhardt sample. This result is consistent with the palynofacies and biomarker analyses of sediments from the Vocontian basin deposited during the Cretaceous OAE1b, a period of high terrigenous input into the ocean (Tribovillard and Gorin, 1991; Heimhofer et al., 2006; Okano et al., 2008). Both the Onnagawa and the Leenhardt kerogens also contained very small amounts of algae (algal organic wall palynomorph) in the form of marine dinocysts.
Kerogen compositions (%) and volume percentages (%) of (a) the Onnagawa Formation Shale On-7 and (b) the Leenhardt Black Shale Leen-1. FA: fluorescent amorphous kerogen, NFA: non-fluorescent amorphous kerogen, WFA: weakly fluorescent amorphous kerogen. See Table S1.
The kerogen samples were separated into 6–11 density fractions by SDC, then observed by microscopy as described above (Fig. 1, Fig. S2). Sporomorph-concentrated fractions were present in the S.G. ranges of <1.25 and 1.25–1.30 g/cm3; high concentrations were evident in the Onnagawa sample at <1.25 g/cm3. Wood and NFA fractions were recovered at S.G. 1.45–1.50 g/cm3, consistent with a previous report (Sawada and Akiyama, 1994). The yields (volume percentages of the total yields) were maximal (46%) in the range of 1.37–1.40 g/cm3 (Fig. 1a).
Regarding the Leenhardt kerogen, NFA was dominant in all density fractions. Higher relative abundances of algal kerogen, originating from dinocysts and sporomorphs, were obtained in the S.G. 1.20–1.25 and 1.25–1.30 g/cm3 fractions (Fig. 1b). The relative abundances of wood were >20% at S.G. 1.40–1.45 and 1.45–1.50 g/cm3. Thus, both algae/sporomorph-abundant and wood-abundant fractions were recovered. The WFA in the Leenhardt samples was distributed over an upper limit of S.G. 1.60 g/cm3, and the NFA was predominant at >1.60 g/cm3. The volume percentages of the total yields were maximal (20%) at S.G. 1.35–1.40 g/cm3 (Fig. 1b).
n-Alkanes/n-alkenes in the pyrolysates from kerogen density fractionsThe pyrolysates from the Onnagawa kerogen principally consisted of aliphatic hydrocarbons, such as n-alkanes and n-alkenes (Figs. 2 and 3). C11–C35 n-alkane/n-alkene doublets were predominant, although in the <1.37 density fraction they were restricted to C11–C30 homologues. The <1.37 density fraction was pooled from the <1.25, 1.25–1.30, and 1.30–1.37 fractions because of the small amounts in each of these fractions. A unimodal distribution with a maximum abundance of C22–C25 n-alkanes was detected in the <1.37 density fraction of the sporomorph-concentrate, whereas shorter chain (<C17) n-alkanes had higher relative abundances in pyrolysates from the >1.37 and bulk kerogen fractions (Fig. 3). The unique carbon-number distributions in pyrolysates from the sporomorph-concentrated fraction resembled the distributions in pyrolysates from fossil microspores, as previously reported (van Bergen et al., 1993). The resistant components of sporomorphs, including pollen and spores (sporopollenin), presumably consist of biomacromolecules comprising aliphatic moieties (Guilford et al., 1988; Hayatsu et al., 1988), predominantly aromatic moieties (de Leeuw et al., 2006), or a mixture of both aliphatic and aromatic moieties (Wehling et al.,1989; Watson et al., 2012). The main macromolecular building blocks of sporopollenin include phenolics, such as p-coumaric acid and ferulic acid, and polyhydroxylated aliphatic analogues (mainly C16, C18, and C20 fatty acid units) covalently coupled by ether and ester bonds (Guilford et al., 1988; Watson et al., 2007, 2012; Quilichini et al., 2015; Nierop et al., 2019). However, in previous studies, the compounds detected in the pyrolysates of fossil spore/pollen clearly differed from the compounds in extant spore/pollen, as demonstrated by py-GC/MS analysis in which long-chain (C25–C30) n-alkane/n-alkene doublets were generated and the contents of fatty acids decreased (Hayatsu et al., 1988; van Bergen et al., 1993). These observations suggested that, during diagenesis, the sporopollenin biomolecules were substantially altered. Watson et al. (2012) reported that the simulated maturation of Lycopodium spores by heating at high temperatures (200–350°C) led to an increase in the polyalkyl hydrocarbon content, as determined by py-GC/MS analysis. The authors suggested that the highly aliphatic macromolecular moieties in mature spore/pollen fossils were derived from the in situ polymerization of the hydrolysable labile component of sporopollenin (specifically, the ester bound n-fatty acid moieties). The maturation-induced polymerization of alkyl units in fossil sporopollenin is supported by our analysis of the sporomorph-concentrated fraction. Higher relative abundances of cuticle were also present in the sporomorph-dominant fraction, although as a minor component. The cutin and cutan of fossil plant cuticle are mainly composed of long-chain alkyl units (Mösle et al., 1998; Almendros et al., 1999; Gupta et al., 2006, 2007). Therefore, the long-chain n-alkanes and n-alkenes in the pyrolysate of the sporomorph-concentrated fraction from the Onnagawa kerogen may have been partly derived from the aliphatic moieties of cuticular remains.
Total ion chromatograms of pyrolysates from the kerogen density fractions obtained from the Onnagawa Formation Shale On-7. Solid circles, triangles, and diamonds are n-alkane/n-alkene doublets, branched and isoprenoid alkanes, and aromatic compounds, respectively. Numbers are the carbon numbers of n-alkanes/n-alkenes; x: contaminants.
Mass fragmentograms of m/z 55 and m/z 57 of pyrolysates from the kerogen density fractions in the Onnagawa Formation Shale On-7. Solid circles and triangles are n-alkane/n-alkene doublets, and branched and isoprenoid alkanes, respectively. Numbers are the carbon numbers of n-alkanes/n-alkenes; x: contaminants.
Short-chain homologues were dominant among the n-alkane/n-alkene doublets (maximal peaks of C11) in the 1.37–1.40 (WFA-concentrated), 1.40–1.45 (WFA and wood-concentrated), and 1.45–1.50 (WFA, NFA, and wood-concentrated) density fractions of the Onnagawa kerogen. Within the pyrolysates, the absence of the preferential accumulation of n-alkanes and n-alkenes with odd carbon numbers indicates that these pyrolytic compounds did not directly originate from biological precursors, such as plant wax; instead, they were generated from polymethylene structures in the geomacromolecules via thermal cracking of C–C bonds. Aromatic hydrocarbons were minor components in pyrolysates from the bulk fraction and from all density fractions of the Onnagawa kerogen. The predominance of aliphatic compounds is a common characteristic of type II kerogen (Kralert et al., 1995; del Río et al., 1995; González-Vila et al., 2001). It is likely that the WFA in the Onnagawa kerogen mainly originated from marine diatoms, as noted above (Sawada and Akiyama, 1994).
In the density fractions of the Leenhardt kerogen, n-alkane/n-alkene doublets in the pyrolysates were mainly in the range of C11–C30 (Figs. 4 and 5). In the <1.30 density fraction (algae/sporomorph-abundant fraction), which combined the <1.20, 1.20–1.25, and 1.25–1.30 fractions, mid-length (C16–C20) n-alkane/n-alkene doublets were dominant and long-chain (C22–C25) homologues were less abundant, in contrast to the distribution in the sporomorph-concentrated fraction of the Onnagawa kerogen (Fig. 3). This result suggests that fossil sporopollenin constituted a small portion of the algae/sporomorph-abundant fraction. Instead, the n-alkane/n-alkene doublets may have been related to algaenan, consisting of resistant biomacromolecules (and/or geomacromolecules). In their analysis of pyrolysates from spongeous dinoflagellate remains (i.e., dinocasts) isolated from Eocene sediment, Versteegh et al. (2004) detected a pronounced peak of C16 n-alkane/n-alkene doublets and a pronounced peak of C16 and C18 fatty acids, implying that the fossil dinoflagellate algaenan was mainly composed of aliphatic C16 and C18 fatty acid-based geomacromolecules. However, the cyst walls of extant dinoflagellate are highly aromatic and contain tocopherol as a major monomeric building block (Kokinos et al., 1998; Versteegh et al., 2012). Dinocysts were mainly present as algal kerogen in the algae/sporomorph-abundant fraction of the Leenhardt kerogen, as noted above. Although the highly abundant C17 alkane/alkene homologues were presumably generated via decarboxylation of the C18 fatty acid moiety of dinoflagellate-derived algaenan, the C15 alkane/alkene homologues, generated from the C16 fatty acid moiety, were only present at low levels. Accordingly, the generation of pyrolytic alkyl products via decarboxylation of fatty acid moieties cannot be uniformly assumed. C19–C20 homologues in the pyrolysate may have been generated by the in situ polymerization of labile components in the algaenan of the algae-derived remains, although hypothesis this requires confirmation.
Total ion chromatograms of pyrolysates from the kerogen density fractions in the Leenhardt Black Shale Leen-1. Solid circles, triangles, diamonds, and squares are n-alkane/n-alkene doublets, isoprenoid alkanes, (alkyl) naphthalene, and (alkyl) phenanthrenes, respectively. Numbers are the carbon numbers of n-alkanes/n-alkenes; x: contaminants.
A simple unimodal distribution, with maxima of C16–C19 n-alkane/n-alkene doublets and no preferential abundances of n-alkanes with odd or even carbon numbers, indicated that the pyrolysate of the 1.35–1.40 density fraction of the Leenhardt kerogen comprised NFA, wood, and WFA fractions. This distribution resembled the distribution of the WFA-rich fractions of the Onnagawa kerogen, indicating that the n-alkanes/n-alkenes were generated from polymethylene structures in the geomacromolecules associated with thermal cracking of the C–C bonds in WFA. In the 1.40–1.45 density fraction, woody kerogen was the most abundant. In this fraction, C11, C16, C19, and C22 homologues had significantly higher relative abundances and followed a simple unimodal distribution, revealing the strong influence of parent biomacromolecular substances and polymethylene from geomacromolecules. In the n-alkane/n-alkene distributions of higher-density 1.50–1.55 S.G. fractions (NFA and wood), C11 and C16 homologues were dominant; the abundances of C12–C22 homologues were nearly similar. The relative abundances of long-chain n-alkanes with even carbon numbers, such as C22 and C24, were obviously increased in the higher-density 1.50–1.55 and 1.55–1.60 fractions. These alkyl compounds, including n-alkanols, n-alkanoic acids (fatty acids), ω-hydroxy acids, and α,ω-alkanedioic acids, can be attributed to suberin monomers in periderm tissue (Holloway, 1983; Pollard et al., 2008). C22 and C24 n-alkanols are mainly released from the geomacromolecules of plant fossils and their carbon-number distributions are well-preserved, as demonstrated in our previous study (Sawada et al., 2008). Thus, suberin may have been a source of those compounds in the pyrolysates from NFA and woody kerogens in the higher-density fractions. In the highest-density fractions (S.G. 1.60–1.65) the n-alkane/n-alkene doublets had a unimodal distribution, with a maximal peak of C17 and decreased relative abundances of C22 and C24 homologues. The C17 n-alkane/n-alkene doublet may have originated from C18 fatty acid and/or ω-hydroxy acid moieties, both of which constitute major components of suberin (Pollard et al., 2008), via decarboxylation during pyrolysis or during diagenesis of the NFA and woody kerogens. Therefore, the main components of the geomacromolecules of NFA kerogens in the highest-density fractions (1.60–1.65) differed from the main components in the 1.50–1.55 and 1.55–1.60 density fractions and may have been suberin-derived geomacromolecules of different plant tissues.
Aromatic hydrocarbons in the pyrolysatesAromatic hydrocarbons, such as C2–C3 methyl indene (Fig. 2B, C and Table 2), C0–C2 naphthalene, and C0–C1 phenanthrene, represented only a minor proportion of the compounds in pyrolysates from the bulk, WFA, and WFA/wood fractions of the Onnagawa kerogen (Fig. 2). These lower-molecular-weight aromatic hydrocarbons may have been generated from marine algal OM via aromatization during diagenesis and maturation, rather than from terrestrial plants. In pyrolysates from the Leenhardt kerogen, aromatic C0–C4 naphthalene and C0–C2 phenanthrene were present as major components, in addition to the aliphatic compounds (Fig. 4). The distribution patterns of naphthalene and phenanthrene were nearly identical among all density fractions of the Leenhardt kerogen, although they were less abundant in the <1.30 (algae/sporopollenin-abundant) fraction. This consistency suggests that NFA and woody kerogen served as the source of these compounds. Because NFA likely represents terrigenous OM, naphthalene and phenanthrene in the Leenhardt kerogen pyrolysates were presumably generated from the aromatic parts of terrestrial plant-derived geomacromolecules. Woody and coaly kerogens are type III kerogens, and their geomacromolecular structures are mostly aromatic. However, upon pyrolysis, lacustrine and marine algal kerogens (i.e., type I and type II kerogens) generate higher absolute yields of aliphatic and aromatic hydrocarbons than type III kerogens, despite the higher aromaticity of type III kerogens (Larter and Senftle, 1985; Larter and Horsfield, 1993). The naphthalene and phenanthrene in kerogen pyrolysates are presumably generated by β-cleavage of the alkyl aromatic structures (Larter and Horsfield, 1993). Alkylbenzenes generally originate from lignin in terrestrial plants through the preservation and/or alkylation of benzene structures in biomacromolecules, but they are also generated by diagenetic processes that act on the aliphatic lipids in kerogen, as discussed below. Thus, in addition to terrigenous sources, the presence of naphthalene and phenanthrene in the kerogen pyrolysates may be attributable to the acyclic components of geomacromolecules in kerogen.
Assignment of the pyrolysates labelled in Fig. 2 to 6.
| Peak | Compound name | Formula | MW |
|---|---|---|---|
| a | 2,6-dimethyl undecane | C13H28 | 184 |
| b | 2,6-dimethyl dodecane | C14H30 | 198 |
| c | monomethyl tridecene | C14H28 | 196 |
| d | 2,6,10-trimethyl dodecane (farnesane) | C15H32 | 212 |
| e | 2,6,10-trimethyl tridecane | C16H34 | 226 |
| f | 2,6,10-trimethyl pentadecane (norpristane) | C18H38 | 254 |
| g | 2,6,10,14-tetramethyl pentadecene (pristene) | C19H38 | 266 |
| h | 2,6,10,14-tetramethyl hexadecene (phytene) | C20H40 | 280 |
| A | tetramethyl thiophene | C8H12S | 140 |
| B | 1,3-dimethylindene | C11H12 | 144 |
| C | 1,2,3-trimethylindene | C12H14 | 158 |
| D | dihydronaphthalene | C10H10 | 130 |
| E | naphthalene | C10H8 | 128 |
| F | monomethyl naphthalene | C11H10 | 142 |
| G | dimethyl naphthalene | C12H12 | 156 |
| H | trimethyl naphthalene | C13H14 | 170 |
| I | tetramethyl naphthalene | C14H16 | 184 |
| J | phenanthlene | C14H10 | 178 |
| K | monomethyl phenanthlene | C15H12 | 192 |
| L1 | trimethylbenzene | C9H12 | 120 |
| L2 | trimethylbenzene | C9H12 | 120 |
| M1 | 1-methyl-3 or 4-propyl benzene | C10H14 | 134 |
| M2 | dimethylethylbenzene | C10H14 | 134 |
| M3 | dimethylethylbenzene | C10H14 | 134 |
| M4 | dimethylethylbenzene | C10H14 | 134 |
| M5 | dimethylethylbenzene | C10H14 | 134 |
| M6 | 1,2,3,5- or 1,2,4,5-tetramethylbenzene | C10H14 | 134 |
| M7 | 1,2,3,4-tetramethylbenzene | C10H14 | 134 |
C13–C20 acyclic isoprenoid hydrocarbons were identified as the major components, although C17 was consistently absent, in the pyrolysates of all density fractions of the Onnagawa kerogen (Figs. 2 and 3). C13–C16, C18, and C19–C20 isoprenoid alkenes (prist-1-ene and phyt-1-ene) were specifically identified. Among the isoprenoid alkanes and alkenes of all kerogen density fractions, the C19 isoprenoid alkene (prist-1-ene) had the highest abundance. In previous investigations of kerogen pyrolysates (e.g., Larter and Horsfield, 1993; Zhang et al., 2016), prist-1-ene was often identified as the main isoprenoid hydrocarbon. The main precursors of prist-1-ene are chlorophyll and phytol (or dihydrophytol), as well as tocopherols (vitamin E) (Goossens et al., 1984). Pyrolytic experiments conducted using chlorophylls and phytol indicated that phytenes and phytadienes, rather than pristenes, were the major products (van de Meent et al., 1980; Ishiwatari et al., 1991). A previous chemical degradation experiment using Onnagawa kerogen revealed that tocopherols were bound in the geomacromolecules via ether bonds (Sawada, 2003). Tocopherols serve as antioxidants and are biosynthesized by diverse organisms such as higher plants, algae, cyanobacteria, and common bacteria (Newton et al., 1977; Velasco et al., 2000); they are especially synthesized by marine microalgae, in which they constitute major lipids (Goossens et al., 1984; Brown et al., 1999). Thus, it is likely that geomacromolecule-bound tocopherols were the main source of the large amount of prist-1-ene in our Onnagawa kerogen pyrolysates.
The shorter-chain (C13–C16 and C18) isoprenoid alkanes were most likely the pyrolytic products of longer-chain isoprenoid homologues (e.g., pristene and phytene) bound in the kerogens, as previously reported (Zhang et al., 2016). Within this group, the most abundant compound was the C13 isoprenoid alkane (2,6-dimethylundecane). These isoprenoid alkane units may have originated from the resistant cyanobacterial biomacromolecule algaenan. In pyrolytic experiments, Biller et al. (2015) found that the biomacromolecules of cyanobacteria, such as Chlorogloeopsis, included C13 (2,6-dimethylundecane) and C20–C21 isoprenoid alkane moieties. Therefore, the 2,6-dimethylundecane detected in pyrolysates of the bulk and WFA-containing fractions of the Onnagawa kerogen also may have originated from cyanobacteria. C14–C16 isoprenoid alkanes and monomethyl tridecene were identified in the pyrolysates of the bulk and the density fractions of kerogen, with the exception of the sporomorph-concentrated fraction (Fig. 3). These components might have been WFA kerogens of algal or bacterial origin.
The pyrolysates of the density fractions of the Leenhardt kerogen contained C13, C16, C18, and C19 isoprenoid hydrocarbons (Fig. 5), with a predominance of the C18 isoprenoid alkane (norpristane) rather than the C19 isoprenoid alkene (prist-1-ene). The relative abundances of norpristane tended to increase in the lower-density fractions (algae/sporomorph, NFA/wood/WFA, and wood-abundant fractions). The limited presence of C16 and C18 isoprenoid alkanes (and absence of C14–C15) suggests that they were derived from a specific source organism(s). A similar distribution of isoprenoid hydrocarbons was identified in the isoprenoid algaenan produced by the freshwater green alga Botryococcus braunii L race (Derenne et al., 1989; Salmon et al., 2009). Although the black shale of the Leenhardt level was deposited in a hemipelagic environment, Botryococcus-like freshwater microalgae or unknown freshwater/coastal microalgae may have been the main sources of isoprenoid alkyl moieties in the geomacromolecule. The presence of the C13 isoprenoid alkane (2,6-dimethylundecane) as a major component in the pyrolysates of all density fractions except the algae/sporomorph-abundant fraction also suggested that it originated from cyanobacterial algaenan (Biller et al., 2015) and highlighted the strong contribution of cyanobacteria to marine primary production during depositions at the Leenhardt level.
Mass fragmentograms of m/z 55 and m/z 57 of pyrolysates from the kerogen density fractions in the Leenhardt Black Shale Leen-1. Solid circles and triangles are n-alkane/n-alkene doublets and isoprenoid alkanes, respectively. Numbers are the carbon numbers of n-alkanes/n-alkenes.
The mass fragmentograms of m/z 92, m/z 106, m/z 119, and m/z 134 indicated the presence of trimethyl- and tetramethylbenzenes in pyrolysates from all density fractions of the Onnagawa kerogen (Fig. 6). Based on the mass spectra and retention times previously reported (Hartgers et al., 1992, 1994b; Lis et al., 2008), 1-methyl-3 or 4-propylbenzene (M1), 1,2,3,5-tetramethylbenzene (M6), and 1,2,3,4-tetramethylbenzene (M7) were tentatively identified, although the GC column used in our study differed from the columns used in previous studies. The methyl positions and/or side chain lengths of trimethyl (L1–L2) could not be determined (Table S2). Dimethyl-ethylbenzenes (M2–M5) were approximately identified based on the higher peaks of m/z 119 and m/z 134, as well as the retention times (Hartgers et al., 1992). The presence of alkylbenzenes in the pyrolysates reflects β-cleavage between the α and β carbon atoms of an aliphatic chain attached to an aromatic ring, although these compounds can also arise from γ- and δ-cleavages (Hartgers et al., 1994b; Lis et al., 2008). Aromatic carotenoids derived from anaerobic photosynthetic sulfur bacteria belonging to Chlorobiaceae (Summons and Powell, 1987; Koopmans et al., 1996; Pedentchouk et al., 2004) and from microalgae (Hoefs et al., 1995; Carmo et al., 1997; Pedentchouk et al., 2004) likely explain the detection of 1,2,3,4-tetramethylbenzene. Chlorobiaceae precursor compounds may also have contributed to the presence of 1,2,3,5-tetramethylbenzene (Requejo et al., 1992; Hartgers et al., 1994a). Chlorobiaceae thrive under strictly anoxic and euxinic conditions; they require both light and hydrogen sulfide. Their presence generally indicates that euxinic waters reached the photic zone and thus imply photic zone euxinia (PZE) (Trüper and Genovese, 1968; Summons and Powell, 1987; Grice et al., 2005). However, it is unlikely that PZE conditions occurred in the surface layers of the paleo-Japan Sea during deposition of the Onnagawa Formation. Instead, the source of the 1,2,3,4-tetramethylbenzene in Onnagawa kerogen pyrolysates may have been marine microalgae. The transformation of algal carotenoids such as β-carotene into aromatic carotenoids through aromatization was demonstrated in an artificial maturation experiment (Koopmans et al., 1996). Accordingly, 1,2,3,4-tetramethylbenzene might have been generated by the pyrolysis of diagenetic products of algal carotenoids. The distribution of alkylbenzenes in the pyrolysates across the density fractions, including the sporomorph-concentrate of the Onnagawa kerogen, was relatively constant (Fig. 6) and supports the conclusion that all pyrolytic alkylbenzenes originated from marine microalgae as sources of WFA.
Partial mass fragmentograms of m/z 57, m/z 92, m/z 106, m/z 119, and m/z 134 of (a) the Onnagawa Formation Shale On-7 and (b) the Leenhardt Black Shale Leen-1.
In the pyrolysates of all density fractions of the Leenhardt kerogen, only 1-methyl-3 or 4-propylbenzene (M1) and dimethyl-ethylbenzenes (M2–M5) were detected as minor components (Fig. 6). The distributions of pyrolytic alkylbenzenes were nearly constant among density fractions <1.55 (<1.30 to 1.50–1.55), whereas the relative abundances of these compounds, especially 1-methyl-3 or 4-propylbenzene, significantly decreased in density fractions >1.55 (1.55–1.60 and 1.60–1.65) (Fig. 6). As in the Onnagawa kerogen, these results suggested that the pyrolytic alkylbenzenes were derived from the WFA of marine microalgal origin. In contrast, 1,2,3,4- and 1,2,3,5-tetramethylbenzenes were not identified in the Leenhardt kerogen pyrolysates. Lis et al. (2008) showed that the amounts of pyrolytic 1,2,3,4- and 1,2,3,5-tetramethylbenzenes decreased with increasing kerogen maturity and are almost lost in more mature (>1.0% Ro) kerogen. However, alkylbenzene abundances in our study were not affected by diagenesis and maturation because the Leenhardt shale is less mature (~0.6% Ro), as estimated from the C31 hopane 22S/22R ratio (Table 1). These results imply the absence of low levels of source organisms, such as anaerobic photosynthetic sulfur bacteria and microalgae, during deposition of the Leenhardt level of the OAE1b.
Fossil geomacromolecular sources in kerogen-concentrated fractionsSpecific kerogen concentrates could be isolated from the Onnagawa and Leenhardt kerogens. A sporomorph-concentrate mainly composed of sporopollenin geomacromolecules was obtained from the Onnagawa kerogen; an algae/sporomorph-abundant fraction containing dinoflagellate-derived algaenan was obtained from the Leenhardt kerogen. In both kerogens, the dominance of long-chain n-alkane/n-alkene doublets suggested the recondensation and (in situ) polymerization of labile compounds as another main component of their geomacromolecules. Woody kerogen concentrates could not be obtained from either kerogen. Moreover, in the wood-abundant fractions of the Onnagawa (S.G. 1.45–1.50) and Leenhardt (S.G. 1.40–1.45) kerogens, the dominance of wood-specific components, such as lignin phenols and their derivatives, was rare. Thus, the woody kerogens in marine sediments might mainly consist of aliphatic alkylated components, approximately similar to the alkylated homologues in marine producer-derived amorphous kerogen, rather than aromatic compounds. Further studies of pyrolysates from terrigenous kerogen in marine sediment are necessary.
WFA and NFA were the major amorphous OM types in the Onnagawa and Leenhardt kerogens, respectively. Both are suspected to arise from recondensation and polymerization reactions, such as reactions resulting in the formation of melanoidin from labile sugars and amino acids via the Maillard reaction (Tissot and Welte, 1984). Additionally, selectively preserved algal and bacterial refractory components, such as algaenan and bacteran (Tegelaar et al., 1989; Derenne et al., 1991), may have served as partial or major building blocks in the main structural skeletons and networks of amorphous kerogen. Indeed, pyrolysis of the fossil geomacromolecules in the WFA fractions of the Onnagawa kerogen generated distinctive isoprenoid hydrocarbons and alkylbenzenes, in addition to n-alkane/n-alkene doublets from polymethylene structures. The source organisms of the fossil geomacromolecules were evaluated based on these results and previous paleoenvironmental and paleoecological interpretations of the Onnagawa Formation. This formation is characterized by siliceous and diatomaceous shales with a high OM content (Suzuki et al., 1993), indicating that the Onnagawa WFA kerogen mainly originated from marine diatoms. Because diatoms do not synthesize resistant biomacromolecules (Versteegh and Blokker, 2004), the isoprenoid hydrocarbons and alkylbenzenes in WFA pyrolysates can be explained by geomacromolecules formed via the recondensation and polymerization of diatom-derived labile compounds. The WFA pyrolysates also contained the pyrolytic C13 isoprenoid alkane (2,6-dimethylundecane), which likely originated from cyanobacterial algaenan, as a major isoprenoid alkane. This result is consistent with sedimentary porphyrin-specific carbon and nitrogen isotopic records, indicating substantial contribution of diazotrophic cyanobacteria to the OM in sediments of the Onnagawa Formation (Kashiyama et al., 2008). It is also possible that a cyanobacterial species lived symbiotically with marine diatoms, such that although the main structure in the WFA fraction originated from diatoms, the geomacromolecule was derived from cyanobacteria.
The NFA pyrolysate of the Leenhardt kerogen contained a high abundance of long-chain even carbon-numbered (C22 and C24) n-alkanes (Fig. 5), presumably suberin monomers from periderm tissue (Holloway, 1983; Pollard et al., 2008). This result provided clear evidence of the dominance of terrigenous OM in the Leenhardt kerogen. Previous biomarker and palynofacies analyses showed that the sediments deposited during OAE1b, including at the Leenhardt level, contained abundant terrigenous organic compounds, indicative of the transport of a large amount of terrigenous matter by freshwater runoff to the Vocontian Basin during this event (Erbacher et al., 1998; Heimhofer et al., 2006; Okano et al., 2008). Thus, the NFA likely originated from the large terrigenous OM inputs into the marine environment of the Tethys Ocean (Vocontian Basin). Moreover, the limited presence of C16 and C18 isoprenoid alkanes suggested a distinctive fossil geomacromolecular source for the isoprenoid algaenan, synthesized by freshwater green algae such as Botryococcus or unknown freshwater/coastal microalgae that were transported in the extensive freshwater runoff during deposition of the Leenhardt level. Pyrolysates of the Leenhardt kerogen also contained a C13 isoprenoid alkane unit attributable to cyanobacterial algaenan, suggesting a strong contribution by cyanobacteria under anoxic conditions during deposition of the Leenhardt level, as previously reported (Kuypers et al., 2004). Kuypers et al. (2002) identified distinctive isoprenoid alkanes, such as 2,6,10,15,19-pentamethylicosane (PMI) and 2,6,15,19-tetramethylicosane (TMI), as major components in the pyrolysates of kerogen from the Paquier level, the acme of the OAE1b. PMI and TMI may have originated from marine archaea; both were also detected as major free biomarkers in the Paquier and Kilian levels of the OAE1b (Okano et al., 2008). However, neither compound was detected in the pyrolysates of Leenhardt kerogen, ruling out contributions from archaeal geomacromolecules.
Our study showed that reconstruction of the paleoenvironment and paleoecosystem is possible based on the geochemical characteristics of fossil geomacromolecular sources elucidated in py-GC/MS analyses of amorphous and structural kerogens. When used in combination with data regarding free lipid biomarkers, the pyrolytic compounds released from the different kerogen density fractions can be powerful tools for paleoenvironmental geochemistry applications.
In this study, density fractions generated from kerogens isolated from the Cretaceous Leenhardt and Miocene Onnagawa shales were subjected to py-GC/MS analysis. The pyrolysates from both kerogens mainly contained aliphatic compounds, such as n-alkane/n-alkene doublets and acyclic isoprenoid hydrocarbons, as well as aromatic compounds including alkylbenzenes, alkylnaphthalenes, and alkylphenanthrenes. The sporomorph-concentrate obtained from the Onnagawa kerogen mainly consisted of sporopollenin geomacromolecules, evidenced by the dominance of long-chain n-alkane/n-alkene doublets. In the algae/sporomorph-abundant fraction from the Leenhardt kerogen, there was a strong contribution of dinoflagellate-derived algaenan. Pyrolysis of the WFA-containing fractions of the Onnagawa kerogen yielded distinctive isoprenoid hydrocarbons and alkylbenzenes, in addition to n-alkane/n-alkene doublets. The isoprenoid hydrocarbons and alkylbenzenes in the WFA pyrolysates were attributed to fossil geomacromolecules produced at the surface of the paleo-Japan Sea by marine microalgae such as diatoms and cyanobacteria. In contrast, NFA, the major kerogen of the Leenhardt shale, mainly consisted of terrigenous suberin geomacromolecules, evidenced by the dominance of long-chain n-alkanes with even carbon numbers. This result supports the dominance of terrigenous OM in the Leenhardt kerogen, consistent with previous biomarker and palynofacies analyses of the Vocontian basin during OAE1b. The distributions of pyrolytic isoprenoid alkanes in the Leenhardt kerogen suggested a distinctive fossil geomacromolecular source, such as freshwater/coastal microalgae and cyanobacteria. The application of SDC in combination with pyrolytic analyses to other sediments will allow identification of the fossil geomacromolecular sources of the OM that contributed to kerogen formation.
We thank Mr. R. Yokoyama of Hokkaido University for helping the separation and microscopic observation of the kerogens. Prof. Dr. N. Suzuki and Dr. Hideto Nakamura of Hokkaido University for their discussion to conduct laboratory work. We are grateful to two anonymous reviewers for constructive comments. This study was supported in part by JSPS KAKENHI Grant Numbers 21H04503, 18KK0091 (to Reishi Takashima), 23540542, 18H01322, and 22H01339 (to Ken Sawada).
