2023 年 118 巻 ANTARCTICA 号 論文ID: 230401
Organic carbon and carbonate carbon are two important reservoirs that control the carbon geodynamic cycle at convergent margins during plate subduction, arc magmatism, and continent building processes. The movement of carbon through different reservoirs in the Earth relating to the global tectonic activities is key in understanding the carbon geodynamic cycle. In this contribution, a comprehensive synthesis on the different types of occurrences of graphite, the purest form of carbon in continental crust, in the Lützow-Holm Complex (LHC), East Antarctica is carried out and carbon isotopic composition is used as a proxy to identify the movement of carbon during orogenesis. Graphite is an important reservoir of carbon in continental crust and occurs in a variety of rock types in the LHC. Based on the mode of occurrence they were classified into several types, disseminated flakes in gneissic rocks, coarse aggregates in leucosomes, graphite concentration in lithological contacts and as monomineralic graphite veins. Disseminated graphite in pelitic gneisses record the lowest carbon isotopic composition (δ13CVPDB values between −25 to −15‰), suggesting biogenic signatures, however those in metacarbonate rocks have equilibrated with carbonate carbon during high temperature metamorphism to show heavier values (δ13CVPDB values between −3 to −1‰). The carbon isotopic composition of disseminated graphite is modified during prograde metamorphism by devolatilization and also exchange of carbon isotopes with carbonate minerals. Coarse-grained graphite is observed in leucosomes in the migmatized metapelitic rocks. During the high-temperature metamorphism and partial melting of graphite-bearing rocks, graphite decomposes to form COH fluids, part of which, especially the lighter isotope-bearing fluids have escaped the system causing a shift toward heavier values (δ13CVPDB values in the range between −18 to −10‰). Based on the field, textural and carbon isotope evidence a model is suggested, where biotite dehydration melting of graphite-bearing rocks caused the dissolution of pre-existing graphite formed from organic materials, and graphite was reprecipitated as coarse aggregates in leucosomes during melt crystallization and cooling. This resulted in the carbon remobilization and isotopic reorganization. Carbon isotopic composition of graphite concentrations in lithological contacts (δ13CVPDB values ranging between −1.8 to −5.7‰) and monomineralic veins (δ13CVPDB values between −3.5 and −6.0‰) suggest that they were precipitated from CO2 fluids locally released through decarbonation reactions. The presence of large volume of skarn mineralization in the contact between carbonate and silicate rocks and similarities of carbon isotopic composition of graphite in contact zones and veins support a local source for CO2 fluids rather than a mantle derived carbon-bearing fluid for vein type graphite. Thus, carbon is recycled and retained as graphite in the continental crust during high-grade metamorphism and anatexis, though its isotopic composition can be considerably modified during orogenesis. In summary, a comprehensive study of carbon isotopic composition of graphite occurrences in the LHC, East Antarctica has thus revealed that prograde metamorphism, anatexis and interaction between carbonate lithologies with silicate rocks can modify carbon isotopic composition of graphite in the continental crust. Recycling of carbon within the continental crust during orogenesis where graphite act as ‘long-term sinks’ of carbon has to be considered for envisaging realistic models on Earth’s carbon cycle.
Carbon and its geodynamic cycle in tectonically active convergent margins play a significant role in controlling the Earth’s carbon budget (e.g., Dasgupta, 2013; Kelemen and Manning, 2015; Plank and Manning, 2019; Muller et al., 2022). Organic carbon and carbonate carbon are two key reservoirs that acts as source or sink for carbon in convergent margins during plate subduction, arc magmatism, and continent building processes (Mills et al., 2019; Wong et al., 2019; Satish-Kumar et al., 2021a; Li et al., 2022). Decarbonation and devolatilization reactions are important in subduction zones, since they control the recycling mechanism of surface carbon to interior of the Earth (e.g., Kerrick and Connolly, 2001; Dasgupta, 2013; Muller et al., 2022). Recently, there is increased focus on how carbonation reactions act as carbon sinks in subduction zone and the mantle wedge above, since carbon released from subducting slabs are partially trapped as carbonate minerals in the mantle wedge (e.g., Okamoto et al., 2021). Therefore, it is important to understand the movement of carbon between various reservoirs at different tectonic settings, in particular, at convergent margins.
Graphite is an important reservoir of carbon in the crust, however its role in continental crust has been underestimated in many geodynamic models on carbon cycle. In a recent study, Carvalho et al. (2023a) considered C-O-H fluid-melt-rock interaction in graphitic granulites and signified the uncertainties in quantifying the carbon budget in the lower continental crust. Despite the occurrence of extensive deposits of graphite in orogenic belts, their role as carbon sinks or source were less considered except in some recent studies (e.g., Parnell and Brolly, 2021; Parnell et al., 2021; Li et al., 2022). Carbon isotope geochemistry is a widely applied tool in deciphering the movement of carbon through different reservoirs, because it records systematic variations in δ13CVPDB values due to fractionations between different carbon-bearing phases. Bulk silicate earth’s carbon isotope value is considered to be controlled by the mantle values of around −5‰, and carbon cycle between different reservoirs having different isotopic composition and different time scales of cycling process, as documented in Figure 1 (after Des Marais, 2001), although recent high-pressure high-temperature experimental studies also reveals a role for carbon in the Earth’s core in controlling the bulk silicate Earth carbon isotopic composition (Satish-Kumar et al., 2011a). The lighter 12C carbon is preferably fixed in organic materials, whereas the heavier 13C is in higher proportions in carbonate carbon in the Earth’s surface and crustal conditions, controlled by biologic and metamorphic processes (Fig. 1). The carbon isotope distribution in the Earth’s surface and crustal conditions vary in the range of δ13C(VPDB) values from ∼ −30 to ∼ +20‰ (Fig. 1), largely controlled by the fractionation between carbonate carbon and organic carbon. The peculiarity of this distribution is that the balance between organic and inorganic reservoirs has remained steady throughout the Earth’s history for billions of years (Eichmann and Schidlowski, 1975; Schidlowski, 1988). During metamorphism, the organic material trapped in the sediments are converted into crystalline graphite through graphitization process, which involves some re-organization of isotopes by devolatilization process (e.g., Wada et al., 1995; Luque et al., 1998; Nakamura et al., 2017, 2020). Previous studies on graphite occurrence in regional metamorphic terrains in the continental collision zones mostly focused on the origin of carbon and its role in controlling the fluid composition and stability of mineral assemblages. The source and movement of carbon, the carbon isotope fractionation between carbonate minerals, graphite and carbon-bearing fluids (CO2 or CH4) during metamorphism were considered (Farquhar and Chacko, 1991; Farquhar et al., 1999; Satish-Kumar, 2000; Satish-Kumar et al., 2002; Santosh et al., 2003; Satish-Kumar, 2005). However, the role of graphite in carbon geodynamic cycle in continental collision zones is yet to be explored.
In this study, the mode of occurrence of graphite in the regionally metamorphosed Lützow-Holm Complex (LHC) in East Antarctica is reported. Carbon isotopic composition of different types of graphite occurrences were characterized. Based on the field occurrence and isotopic characteristics, mobility and sequestration process of carbon in continental crust are discussed. Furthermore, the role of graphite as ‘long-term sinks’ of carbon during orogenesis is evaluated from a carbon isotope perspective.
The LHC, is a Neoproterozoic-Cambrian metamorphic terrain located along the Prince Olav, Prince Herald, and Soya coasts (∼ 400 km), in East Antarctica (Fig. 2), comprises of a thick pile of metasedimentary and metaigneous rocks. The basement rocks in this region have developed in a continental marginal setting during the Gondwana amalgamation (Shiraishi et al., 1994, 2008; Satish-Kumar et al., 2008, 2013). The region has experienced a progressive metamorphism from amphibolite facies at the Prince Olav coast to granulite facies at the Soya coast (Hiroi et al., 1983; Shiraishi et al., 1989; Hiroi et al., 1991), and attained ultrahigh-temperature conditions at Rundvågshetta, where a maximum peak metamorphic conditions of around 1000 °C and 1.1 GPa are reported (Motoyoshi and Ishikawa, 1997; Durgalakshmi et al., 2021; Carvalho et al., 2023b). The presence of kyanite and staurolite inclusions within garnet porphyroblasts and plagioclase along with decompression corona textures suggests that the LHC has experienced a ‘clockwise’ metamorphic P-T-t evolution (Motoyoshi et al., 1989; Hiroi et al., 1991). The age of peak metamorphism in the LHC is between 553 ± 6 and 521 ± 9 Ma, as inferred from SHRIMP U-Pb zircon data (Shiraishi et al., 1994, 2008; Mori et al., 2023). However, electron microprobe monazite geochronology and LA-ICP-MS U-Pb zircon dating also provided evidence for an older population of ages around ∼ 650-580 Ma (Hokada and Motoyoshi, 2006; Kawakami et al., 2016), whereas recent studies also report sporadic occurrences of Tonian metamorphic event without Cambrian overprints in LHC (Baba et al., 2022; Kitano et al., 2023). In a recent compilation of U-Pb detrital and protolith age data from the LHC, Dunkley et al. (2020) proposed a new geological subdivision based on protolith ages: the Innhovde Suite (1070-1040 Ma) composed mainly of felsic orthogneiss; the Rundvågshetta Suite (2520-2470 Ma), mostly felsic orthogneiss with minor mafic and metasedimentary gneisses; the Skallevikshalsen Suite (1830-1790 Ma), felsic to mafic orthogneiss with abundant dolomitic marbles, calc-silicate rocks and other metasediments; the Langhovde Suite (1100-1050 Ma), mostly felsic orthogneiss with minor mafic and calc-silicate gneisses; the East Ongul Suite (630 Ma), with various orthogneisses and metasediments; and the Akarui Suite (970-800 Ma) with diverse orthogneisses and paragneisses.
Fifty-six samples containing graphite from key localities of Skallevikshallsen, Rundvågshetta, Skarvsnes, and Langhovde, and sporadic occurrences of graphite from Byobu rock and East Ongul Island in the LHC were considered in this study (Fig. 2). These localities comprise various kinds of gneisses, migmatites, quartzite, calc-silicate rocks and marbles, and pegmatites. Samples used in this study were collected during the 46th Japanese Antarctic Research Expedition (2004-2005). Field occurrence of graphite in different rock types were carefully documented and petrography of the samples were carried out. Based on the mode of occurrence, four types of graphite were identified; disseminated fine to medium flakes (mm scale), coarse aggregates (cm scale), graphite concentrations in lithological contacts, and monomineralic graphite veins. Graphite occurrences in representative localities are described in detail below and illustrated in Figures 3 and 4. Typical examples of graphite under a microscope are shown in Figure 5.
Skallevikshalsen is located in the Lützow-Holm Bay coast, about 70 km southwest of Syowa Base (Fig. 2). Different types of metasedimentary and meta-igneous rocks occur in this area, which include garnet-sillimanite gneiss, quartzofeldspathic gneiss, metacarbonate rocks (marbles and calc-silicate boudins), pyroxene gneiss, granitic gneiss, and metabasite (Yoshida et al., 1976; Yoshida, 1977; Satish-Kumar et al., 2006a; Mizuochi et al., 2010). Pyroxene gneisses are the most widely distributed rock type, which are intruded by granitic pegmatites (Yoshida et al., 1976). Thick skarn zones occur in the contact between marble layers and leucogneisses. Metabasites occur as lenses and layers within orthogneiss and paragneiss. All lithologies, except for the granitic pegmatites, experienced strong deformation with a regional NE-SW strike with a gentle dip toward SE (<40°). Four deformational events were identified by Kawakami and Ikeda (2004) at Skallevikshalsen, where the metamorphic P-T conditions reached a peak of 780-960 °C and 0.6-1.1 GPa (Yoshimura et al., 2004). Mizuochi et al. (2010) reported accurate metamorphic temperature estimates of 850-870 °C for metacarbonate rocks using calcite-dolomite solvus thermometry combined with calcite-graphite carbon isotope thermometry. Recent estimates using the Ti-in-zircon thermometer indicate peak temperatures of 820-850 °C (Kawakami et al., 2016). Suzuki and Kawakami (2019) recently reported a peak P-T conditions of 834 ± 4 °C and ∼ 1.05 GPa from the pelitic gneiss. The earlier reports on high-temperature peak metamorphism at Skallevikshalsen is supported by the occurrence of spinel + quartz assemblage in the core of garnet in garnet - sillimanite leucogneiss (Kawakami and Motoyoshi, 2004).
Different types of graphite occurrences are observed at Skallevikshalsen. The most common occurrence is as disseminations in pelitic, psammitic, and carbonate rocks. The disseminated graphite are mostly aligned to the regional foliation in metapelitic and metapsammitic rocks (Figs. 3a and 5b), but are randomly distributed in metacarbonate rocks (Fig. 5c). The occurrence of graphite flakes as disseminations in the metapelitic and metapsammitic rocks are similar to those observed in metasedimentary rocks in southern Indian Kerala Khondalite Belt (Radhika et al., 1995) and the Highland Complex in Sri Lanka (Binu-Lal et al., 2003). Generally, the size of the flakes ranges from <1 mm to several mm, and some large flakes longer than 2 cm are observed. The disseminated graphite gains in the metacarbonate rocks are randomly distributed (Fig. 5c). The basal planes of disseminated graphite flakes are parallel to the regional foliation in the metapelitic, metapsammitic rocks and quartzite (Fig. 5d).
Monomineralic centimeter thick veins of graphite also occur adjacent to the lithological contact of metacarbonate rock and pyroxene gneiss (Fig. 3b). The vein cut across the regional foliation and is at high angles to the contact plane. This occurrence of pure graphite vein resembles with vein graphite occurrences in Sri Lanka (Touzain et al., 2010; Hewathilake et al., 2018; Touret et al., 2019). Within the vein graphite crystals are oriented perpendicular to the fracture suggesting precipitation of graphite in an open fracture.
A peculiar occurrence of graphite concentration is observed in the contact of marble and pyroxene gneiss, where centimeter scale graphite-rich zone and scapolite + clinopyroxene + graphite intergrowth is observed (Figs. 4d and 4e). Graphite concentration increases toward the contact where it is intergrown with scapolite (Fig. 4d). Graphite also occurs in calc-silicate boudins within the pyroxene gneiss and garnet-biotite gneiss (Figs. 4e and 4f).
Graphite-bearing leucosomes at RundvågshettaThe exposure at Rundvågshetta comprises of compositionally diverse high-temperature granulites including the metapelitic rocks with orthopyroxene-sillimanite-bearing and sapphirine-bearing assemblages (Ishikawa et al., 1994; Motoyoshi and Ishikawa, 1997). Earlier studies have revealed peak metamorphic condition of around 1000 °C at ∼ 1.1 GPa (Motoyoshi and Ishikawa, 1997). Further evidence for UHT conditions were reported by Yoshimura et al. (2008) based on the finding of co-existing sapphirine + quartz assemblage. The peak metamorphism occurred during the Pan-African Orogeny (517 ± 9 Ma; Fraser et al., 2000). Ishikawa et al. (1994) identified early intense deformation in the metamorphic rocks of Rundvågshetta. Based on detailed structural analyses of the region, they suggested that the dominant isoclinal folds and ductile boudinage structures are coeval and represent the regional ductile deformation during the peak metamorphism. In a recent study, Carvalho et al. (2023b) reported textural and geochemical characteristics of melt inclusions in a metapelitic UHT granulite from Rundvågshetta with compositional affinities of A-type granites, thereby proposing a link between UHT anatexis of residual metasedimentary crust and granite generation.
Graphite occurs in metapelitic rocks as disseminations. Quartzo-feldspathic layers alternating with garnet, sillimanite and biotite-rich layers are peculiar for the metapelitic rocks. Graphite, in subordinate quantities, is commonly associated with biotite and sillimanite and are aligned parallel to the regional foliation. Grain size varies between 0.1 and 0.5 mm. Occasionally graphite grains are also observed as inclusions within garnet.
Garnet-biotite gneiss at Rundvågshetta has undergone a high degree of partial melting and are intensely deformed to form stromatic migmatites. The rock unit comprises of restitic layers of garnet + biotite + feldspar + quartz ± graphite that are interlayered with decimeter scale semi-concordant bands and pods of garnet-bearing leucosomes. Aggregates of graphite crystals that are larger than disseminated graphite occur in feldspar + quartz + garnet-rich leucosomes (Fig. 4a). At places they show radiating crystal growth patterns, suggesting crystal growth in a strain free environment. Large garnet crystals in metapelitic rocks (diameter up to 10 cm) also contain graphite inclusions. Coarse graphite grains are considered to be in textural equilibrium with the melt, as evidenced by either intergrowth of graphite and K-feldspar + quartz representing contemporaneous crystallization (Fig. 5d) or radial growth of graphite crystals (Fig. 4b). Graphite-rich fractures are also found cutting across the regional foliation in metapelitic rocks (Fig. 4c).
Graphite occurrences in LanghovdeLanghovde area is one of the largest outcrops in the LHC, where pyroxene gneiss and migmatitic garnet-biotite gneiss are the dominant lithologic units. Garnet-bearing granitic gneiss and metabasite along with minor occurrences late-stage granite and pegmatites are also found (Ishikawa, 1976). The southern part of the Langhovde area is characterized by monoclinal structure trending in the NNE-SSW direction dipping toward east. Coarse grains of graphite occur in leucocratic pods and veins within migmatites (Fig. 4d). Representative samples were collected from two localities which contain coarse graphite grains up to 10 mm in longer dimension.
Graphite occurrences in SkarvsnesSkarvsnes, situated about 40 km south of Syowa Base, is one of the largest ice-free outcrop in the LHC. This region is characterized by mostly gneissose rocks, including pyroxene gneiss, garnet-biotite gneiss, and hornblende gneiss (Ishikawa et al., 1977). Subordinate amounts of metasedimentary rocks and metabasic rocks, migmatites, granitic rocks, and pegmatites are also found in this region. Several generations of fold structures, fractures, and thrusts were identified (Ishikawa et al., 1977). Skarvsnes is also characterized by the occurrence of calc-silicate boudins (Hiroi et al., 1986; Satish-Kumar et al., 2006a).
Graphite is observed in several rock units. The most common occurrence is in garnet-biotite gneiss and in calc-silicate boudins. In both these occurrences, graphite occur as medium to fine grains (Fig. 4e). Coarse graphite is also observed in leucocratic portions of migmatites. In addition, a sillimanite-rich quartz vein also contain abundant graphite crystals.
Graphite occurrences in other localities in LHCGraphite is observed in some other outcrops as well. A graphite-bearing garnet-biotite-hornblende layer is observed at East Ongul island. Here as well, coarse grains are observed in leucocratic layers (Fig. 4f), compared to fine grained occurrences in the melanocratic portions. At the Byobu Rock outcrop (Satish-Kumar et al., 2006b), metapelitic rocks ubiquitously contain graphite whereas garnet-biotite quartzo-feldspathic gneiss have disseminated graphite. The metapelitic rocks are banded with garnet + biotite + sillimanite-rich domains have higher proportions of graphite compared to the rare occurrence in the quartzo-feldspathic domains. Gneissic foliation is defined by the planar arrangement of biotite, sillimanite, and graphite.
Single graphite crystals were hand picked under a binocular microscope from the residue of silicate rocks treated with concentrated hydrofluoric acid, whereas metacarbonate rocks were treated with concentrated hydrochloric acid. The crystals were introduced to a quartz glass tube and appropriate amount of vanadium pentoxide added as an oxidizing agent. The quartz glass tubes were preheated at 1100 °C for 10 hours, before sample introduction, to remove any contamination from the tubes (cf. Wada et al., 1984). The graphite samples with vanadium pentoxide were again preheated at 500 °C for 30 min. to remove any surficial contamination in the sample/reagent. The tubes were then sealed in vacuum and then allowed to combust for 2 hours at 1100 °C. The glass tube was cracked under vacuum in the inlet system and the released gas was purified using an n-pentane slush and used for measuring the carbon isotope ratios using MAT 250 mass spectrometer at Shizuoka University. Additional measurements were carried out using ThermoFischer MAT 253 mass spectrometer at Niigata University following the procedures described in Satish-Kumar et al. (2021b). The carbon isotope results are reported in the conventional δ notation with respect to VPDB. Repeated measurement of laboratory standard CO2 gas gave reproducibility better than 0.03‰ for δ13C (Wada and Ito, 1990; Satish-Kumar et al., 2021b).
A total of 113 carbon isotope compositions of graphite were obtained in this study. 30 additional data are also compiled from two previous studies (Satish-Kumar and Wada, 2000; Mizuochi et al., 2010), where the authors reported carbon isotopic composition graphite in marbles for the purpose of estimating metamorphic temperature conditions using carbon isotope thermometry. The results of the carbon isotope composition of graphite in various metamorphic rocks from the LHC are compiled in Table 1 and shows as a histogram in Figure 6. The δ13CVPDB values display a large variation, the lowest value of −25.1‰ is observed in a calc-silicate rock boudin and the highest value of −1.8‰ was found for a graphite grain in metacarbonate rock.
| Locality name and sample number (coordinates and reference) |
Analysis No. | Rock type | Mineralogical/textural description |
δ13C(V-PDB) ‰ |
| Rundvågshetta | ||||
| ST050112-4 | 12-4-1-gr1 | Metapelite | Disseminated graphite in garnet-sillimanite graphite gneiss | −15.36 |
| (S69°54.53′; E39°1.658′) | 12-4-1-gr2 | −16.58 | ||
| 12-4-2-gr1 | Leucosome | Graphite-rich leucocratic portion | −16.42 | |
| 12-4-2-gr2 | −17.21 | |||
| 12-4-2-gr3 | −16.95 | |||
| ST050113-2 | 13-2-1A-gr1 | Leucosome | Coarse graphite-rich felsic portion | −15.70 |
| 13-2-1A-gr2 | −16.26 | |||
| 13-2-1B-gr1 | Leucosome | Coarse graphite-rich felsic portion | −16.17 | |
| 13-2-1C-gr1 | Leucosome | −15.91 | ||
| 13-2-2-gr1 | Metapelite | Graphite in fracture filling | −15.92 | |
| 13-2-2-gr2 | −15.89 | |||
| 13-2-2-gr3 | −17.22 | |||
| 13-2-3-gr-1 | Felsic gneiss | Graphite-bearing gneiss | −15.91 | |
| 13-2-3-gr-2 | −16.26 | |||
| 13-2-3-gr-3 | −16.19 | |||
| 13-2-4-gr1 | Felsic gneiss | Graphite-bearing gneiss | −15.97 | |
| 13-2-4-gr2 | −14.71 | |||
| 13-2-5-gr1 | Pegmatite | Coarse graphite associated with K-feldspar | −16.00 | |
| 13-2-5-gr2 | −16.06 | |||
| 13-2-5-gr3 | −13.28 | |||
| 13-2-6-gr1 | Metapelite | Disseminated graphite in garnet-sillimanite graphite gneiss | −20.16 | |
| 13-2-6-gr2 | −20.03 | |||
| 13-2-6-gr3 | −19.99 | |||
| 13-2-7-gr1 | Leucosome | Graphite in melt seggregation | −18.42 | |
| 13-2-7-gr2 | −18.40 | |||
| 13-2-7-gr3 | −18.57 | |||
| 13-2-8-gr1 | Leucosome | Coarse graphite in melt seggregation with host gneiss | −15.35 | |
| 13-2-8-gr2 | −14.63 | |||
| 13-2-8-gr3 | Melanosome | Fine graphite in restite | −19.87 | |
| 13-2-8-gr4 | −20.55 | |||
| 13-2-8-gr5 | −20.31 | |||
| 13-2-9-gr1 | Felsic gneiss | Graphite-enriched portion | −17.62 | |
| 13-2-9-gr2 | −17.83 | |||
| 13-2-9-gr3 | −17.81 | |||
| 13-2-10-gr1 | Pegmatite | Graphite-bearing pegmatite | −19.75 | |
| 13-2-10-gr2 | −17.37 | |||
| 13-2-10-gr3 | −14.14 | |||
| Skallevikshalsen | ||||
| ST050118-1 | 18-1-1-gr1 | Metacarbonate rock | Disseminated graphite flakes in calcite matrix | −3.62 |
| (S69°41.815; E39°16.638) | 18-1-1-gr2 | −3.47 | ||
| 18-1-1-gr3 | −3.63 | |||
| 18-1-1-gr4 | −3.48 | |||
| ST050118-2 | 18-2-12-gr1 | Quartzite | Coarse graphite grains in quartzite | −4.47 |
| (S69°41.92; E39°16.161) | 18-2-12-gr2 | −4.57 | ||
| ST050119-1 | 19-1-6-gr1 | Metacarbonate rock | Disseminated graphite flakes in calcite matrix | −3.15 |
| (Mizuochi et al., 2010) | 19-1-6-gr2 | −2.47 | ||
| 19-1-6-gr3 | −2.35 | |||
| ST050119-2 | 19-2-2-gr-1 | Felsic gneiss | Fine grained disseminated graphite | −13.75 |
| (S69°41.697; E39°17.285) | 19-2-2-gr-2 | −14.14 | ||
| 19-2-2-gr-3 | −14.18 | |||
| 19-2-2-gr-4 | −15.12 | |||
| 19-2-2-gr-5 | −13.87 | |||
| 19-2-2-gr-6 | −14.14 | |||
| 19-2-5A-gr-1 | Graphite vein | Vein graphite cutting across felsic graphite-bearing gneiss | −5.41 | |
| 19-2-5A-gr-2 | −5.44 | |||
| 19-2-5A-gr-3 | −6.50 | |||
| 19-2-5A-gr-4 | −5.92 | |||
| 19-2-5A-gr-5 | −5.98 | |||
| 19-2-5A-gr-6 | −5.73 | |||
| 19-2-5A-gr-7 | −5.46 | |||
| 19-2-5A-gr-8 | −5.32 | |||
| ST050120-1 | 20-1-1-gr-1 | Metacarbonate rock | Disseminated graphite in calcite/dolomite matrix | −1.35 |
| (Data from Mizuochi et al., 2010) | 20-1-1-gr-2 | −1.50 | ||
| 20-1-1-gr-3 | −1.58 | |||
| 20-1-4A-2-gr-1 | Calc-silicate rock | Scapolite-graphite-rich layer | −3.64 | |
| 20-1-4A-2-gr-2 | −4.55 | |||
| 20-1-4A-2-gr-3 | −3.58 | |||
| 20-1-4B-1-gr-1 | Calc-silicate rock | Scapolite-graphite layer | −4.32 | |
| 20-1-4B-1-gr-2 | −3.98 | |||
| ST050120-2 | 20-2-1-1-gr-1 | Quartzite | Graphite-bearing quartzite in the upper contact | −2.68 |
| (S69°41.696; E39°18.552) | 20-2-1-1-gr-2 | −6.07 | ||
| 20-2-1-1-gr-3 | −3.66 | |||
| 20-2-2-1-gr-1 | Quartzite | Graphite-bearing quartzite in the upper contact | −4.07 | |
| 20-2-2-1-gr-2 | −6.44 | |||
| 20-2-2-1-gr-3 | −6.21 | |||
| 20-2-3-gr-1 | Pyroxene gneiss | Fine grained disseminated graphite | −13.83 | |
| 20-2-3-gr-2 | −13.46 | |||
| 20-2-3-gr-3 | −14.06 | |||
| 20-2-4-gr-1 | Pyroxene gneiss | Fine grained disseminated graphite | −14.41 | |
| 20-2-4-gr-2 | −9.74 | |||
| 20-2-4-gr-3 | −17.18 | |||
| 20-2-10-3-gr-1 | Graphite vein | Vein-type graphite - edge of the vein | −4.55 | |
| 20-2-10-3-gr-2 | Vein-type graphite - 1 mm from the contact | −4.36 | ||
| 20-2-10-3-gr-3 | Vein-type graphite - 3 mm from the contact | −4.49 | ||
| 20-2-10-3-gr-4 | Vein-type graphite - middle port of the vein | −6.19 | ||
| 20-2-10-3-gr-5 | Vein-type graphite - middle port of the vein | −5.20 | ||
| ST050121-1 | 21-1-2-gr-1 | Metapelite | Disseminated graphite | −10.18 |
| 21-1-2-gr-2 | −4.61 | |||
| 21-1-6-gr-1 | Calc-silicate rock | Scapolite-graphite boudin core | −1.79 | |
| 21-1-6-gr-2 | −3.23 | |||
| 21-1-7-gr-1 | Calc-silicate rock | Scapolite-graphite boudin | −1.97 | |
| 21-1-7-gr-2 | −2.11 | |||
| 21-1-7-gr-3 | −2.59 | |||
| 21-1-11-gr-1 | Tremolite-graphite intergrowth at the contact between marble and scapolite boudin | −2.71 | ||
| 21-1-11-gr-2 | −2.30 | |||
| 21-1-11-gr-3 | −2.54 | |||
| ST050121-2 | 21-2-11-gr-1 | Graphite vein in the contact between calc-silicate rock and pyroxene gneiss | −5.52 | |
| 21-2-11-gr-2 | −5.73 | |||
| 21-2-12-gr-1 | Graphite enrichment in felsic portion in pyroxene gneiss | −5.14 | ||
| 21-2-12-gr-2 | −5.35 | |||
| Skarvsnes | ||||
| ST050125-1 | 25-1-1A-gr-1 | Calc-silicate rock | Graphite-bearing calc-silicate boudin | −25.14 |
| (S69°26.945; E39°35.954) | 25-1-5A-gr-1 | Vein | Sillimanite vein with graphite and quartz | −11.60 |
| 25-1-6-gr-1 | Metapelite | Garnet-biotite-graphite gneiss | −14.53 | |
| ST050130-2 | 30-2-1-gr-1 | Leucosome | Garnet-biotite gneiss with melt pod having coarse graphite | −13.06 |
| (S69°27.001; E39°35.938) | 30-2-7-gr-1 | Calc-silicate rock | Zonned calc-silicate boudin with graphite | −20.70 |
| ST050130-3 | 30-3-2-gr-1 | Calc-silicate rock | Zonned calc-silicate boudin with graphite | −22.58 |
| (S69°27.481; E39°35.486) | 30-3-4-gr-1 | Leucosome | Coarse melt pod with graphite | −12.85 |
| Langhovde | ||||
| ST050204-2 | 04-2-3-gr-1 | Garnet-biotite gneiss | Caorse graphite in leucosome in gt-bt gneiss | −5.39 |
| (S69°14.553; E39°42.678) | 04-2-3-gr-2 | −9.76 | ||
| ST050207-2 | 07-2-1-gr-1 | Garnet-biotite gneiss | Caorse graphite in leucosome in gt-bt gneiss | −17.52 |
| (S69°15.072; E39°43.765) | 07-2-2-gr-1 | Caorse graphite in leucosome in gt-bt gneiss | −9.76 | |
| 07-2-4-gr-1 | Caorse graphite in leucosome in gt-bt gneiss | −9.14 | ||
| 07-2-5-gr-1 | Caorse graphite in leucosome in gt-bt gneiss | −9.76 | ||
| ST050207-3 | 07-3-1-gr-1 | Garnet-biotite gneiss | Caorse graphite in leucosome in gt-bt gneiss | −7.39 |
| (S69°15.094; E39°43.707) | 07-3-2-gr-1 | Caorse graphite in leucosome in gt-bt gneiss | −7.95 | |
| Byobu Rock | ||||
| ST050103-2 | 03-2-1-gr-1 | Metapelite | Graphite associated with biotite and sillimanite | −16.83 |
| (S69°24.252; E41°59.578) | 03-2-2-gr-1 | Metapelite | Graphite associated with biotite and sillimanite | −16.14 |
| 03-2-3-gr-1 | Metapelite | Graphite associated with biotite and sillimanite | −16.92 | |
| West Ongul Island | ||||
| ST050126-1 | 26-1-1-gr-1 | Garnet biotite migmatitiic gneiss | Disseminated graphite | −12.53 |
| 26-1-2-gr-1 | Leucosome | Coarse graphtie in leucosome | −11.37 | |
| Skallen (Data from Satish-Kumar and Wada, 2000) | ||||
| 602b-g1 | Marble | Disseminated graphite in calcite/dolomite matrix | −2.62 | |
| 602b-g2 | −2.70 | |||
| 602b-g3 | −2.79 | |||
| 602b-g4 | −2.91 | |||
| 602c-g1 | Marble | Disseminated graphite in calcite/dolomite matrix | −2.70 | |
| 602c-g2 | −2.90 | |||
| 602c-g3 | −3.36 | |||
| 602c-g4 | −3.27 | |||
| 602d-g1 | Marble | Disseminated graphite in calcite/dolomite matrix | −2.68 | |
| 602d-g2 | −2.23 | |||
| 602d-g3 | −2.27 | |||
| 602d-g4 | −2.13 | |||
| 602d-g5 | −2.10 | |||
| 602d-g6 | −3.18 | |||
| 602d-g7 | −1.68 | |||
| 602d-g8 | −2.59 | |||
| 602d-g9 | −2.03 | |||
| 602d-g10 | −2.70 | |||
| 602e-g1 | Marble | Disseminated graphite in calcite/dolomite matrix | −2.20 | |
| 602e-g2 | −2.92 | |||
| 602e-g3 | −1.74 | |||
| 602e-g4 | −2.65 | |||
| 602e2-g1 | Marble | Disseminated graphite in calcite/dolomite matrix | −2.49 | |
| 602e2-g2 | −2.66 | |||
A total of 62 carbon isotopic measurements were carried out from various metamorphic rocks samples collected from Skallevikshalsen. The δ13CVPDB values of graphite are widely distributed in the range between −1.4 and −17.2‰. The heaviest isotopic composition in graphite are observed in metacarbonate rocks. Calc-silicate rocks yielded δ13CVPDB values in the range of −1.8 to −4.6‰, whereas graphite grains within marbles have values between −1.4 and −3.4‰. Vein-type graphite yielded δ13CVPDB values in a narrow range of −3.5 to −6.0‰, with no major differences within a single vein (Table 1). Fine-grained disseminated graphite from a felsic gneiss yielded δ13CVPDB values in a narrow range of −13.8 to −15.1‰. Similarly, fine-grained graphite in pyroxene gneiss yielded δ13CVPDB values in the range of −9.7 to −17.2‰. Graphite in a quartzite layer have δ13CVPDB values in the range of −2.7 to −6.4‰. Graphite concentrations observed in contact reaction zones between metacarbonate rocks and pyroxene gneisses have 13C enriched values (δ13CVPDB values ranging between −1.8 to −5.7‰).
Thirty nine graphite grains were analyzed from Rundvågshetta. The δ13CVPDB values lie in a narrow range compared to other localities −13.3 to −20.6‰ (Table 1). The heaviest carbon isotope values are observed in a pegmatite vein. In general, disseminated graphite grains in metapelitic rocks and melanosomes in migmatites are slightly enriched in 12C, whereas the leucosomes are enriched in 13C.
The coarse-grained graphite in leucosomes within garnet-biotite gneiss at Langhovde yielded δ13CVPDB values between −5.4 and −17.5‰. At Skarvsnes, a large range of carbon isotope values are observed. The graphite grains in the calc-silicate rocks yielded δ13CVPDB values from −20.7 to −25.1‰, whereas the graphite in the leucosomes have δ13CVPDB values from −11.6 to −14.5‰. At Byobu Rock outcrop, 3 graphite samples from a metapelitic layer have δ13CVPDB values −16.1 to −16.9‰ (Table 1).
Based on carbon isotopic composition, graphite in the continental crust is generally considered to form from two different sources of carbon; graphite recrystallized from organic material and graphite precipitated from fluids. Hence, carbon isotopic composition usually serve as a valuable criteria in characterizing the origin of carbon (Schidlowski, 1988; Farquhar and Chacko, 1991; Luque et al., 2012). These two end members are usually distinguished by a distinct carbon isotopic composition, shown to be inert to isotope diffusion even at temperature conditions prevailing at the lower crust or upper mantle (Thrower and Mayor, 1978), and hence graphite preserve intragranular isotopic zonation (e.g., Santosh and Wada, 1993; Satish-Kumar et al., 2011b). The isotopic data of graphite from LHC suggests a broad distribution of δ13CVPDB values, with a major group of values between −2 and −5‰. A second group with δ13CVPDB values between −15 and −20‰ can also be recognized and the rest of the samples scattered in between (Fig. 6). Considering the geological setting of LHC, where large expanses of metasedimentary rocks are exposed, recrystallization from organic matter is the most plausible origin of carbon. However, the results of carbon isotope composition of graphite from LHC are relatively enriched in 13C than the normal ‘biogenic’ values (∼ −25‰, Schidlowski, 1988). The metapelitic rocks in the LHC have suffered from high to ultra-high temperature metamorphism and field evidence shows that most of the rocks show records of anatexis. We link the first group of enriched values from deposition of enriched CO2 released from decarbonation of carbonate lithologies (cf. Santosh et al., 2003). In contrast, the enrichment of carbon isotopic composition in metapelitic and metapsammitic rocks, where primary organic material were present, is related to a process of volatile escape during anatexis, whereby the lighter carbon isotopes escaped during melting and recrystallization of graphite (cf. Satish-Kumar et al., 2011b).
Origin and isotopic evolution of graphite associated with metacarbonate rocksGraphite grains in marbles are considered to have exchanged carbon isotopes with associated carbonate minerals during prograde metamorphism. They have preserved equilibrium carbon isotope fractionation with carbonates that correspond to peak metamorphic temperature conditions (e.g., Satish-Kumar and Wada, 2000; Mizuochi et al., 2010), even at ultra-high temperature metamorphic conditions (Satish-Kumar, 2000). Field evidence at Skallevikshalsen shows that graphite is also found to be concentrated at the contact zones between metacarbonate rocks and surrounding silicate rocks (Figs. 3c and 3d). Furthermore, monomineralic graphite veins (Fig. 3b) and graphite enriched fractures (Fig. 4c) at high angles to the regional layering are observed in the vicinity of metacarbonate rocks. The δ13CVPDB values of graphite in the contact zones mentioned above are 13C enriched, indicating their origin from fluids released through the decarbonation reactions. Formation of thick skarn layers at the contact between pelitic/psammitic layers with marbles at Skallevikshalsen support the origin of CO2-rich fluids, although precipitation of graphite from CO2-rich fluids requires extremely reducing conditions or mixing of fluid species (Farquhar and Chacko, 1991; Cesare, 1995; Luque et al., 2012). In addition, as described in Satish-Kumar et al. (2006c), the LHC is characterized by a P-T-fluid evolution where a prominent shift in fluid composition from moderately dense CO2 during peak to high density CO2 in the early retrograde and shifting to low density CO2 fluids in the last phase of metamorphism (Fig. 7). The lowering of CO2 contents in the fluid regime might be a consequence of graphite precipitation.
Graphite is a common accessory mineral in high-grade metasedimentary rocks. Santosh and Wada (1993) and Radhika et al. (1995) identified a variety of graphite associations in the Kerala Khondalite belt in the southern Indian granulite facies terrain. These include: (1) disseminations and strata-bound or podiform concentrations in aluminous metapelites and carbonate-rich lithologies; (2) coarse flakes and flaky aggregates in veins, pegmatites, and mesoscopic shear zones; and (3) coarse crystals associated with patchy cordierite-rich zones and within veins and lenses of ‘incipient charnockites’. From carbon isotopic measurements of the different graphite types, these authors inferred graphite formation have resulted from two major processes, namely: (1) the conversion of organic matter trapped in sediments during metamorphism, and (2) by precipitation from CO2-rich fluids which infiltrated through fractures. Organic materials trapped within sediments undergo graphitization during high-grade metamorphism resulting in the formation of fully crystalline flakes of graphite. On the other hand, coarse graphite flakes in veins and pegmatites possess high δ13CVPDB values, ranging from −10 to −15‰. Graphite associated with incipient charnockites (−10.4 to −13.4%; Santosh and Wada, 1993) and shear planes (−8.2 to −12.4‰; Radhika and Santosh, 1996) show further heavier values, possibly deposited from CO2-rich fluids. The carbon isotope data on graphite thus reveal the influx of CO2-rich fluids at high-grade conditions resulting in mineralogic as well as isotopic alteration. The isotopic composition of coarse graphite flakes is consistent with a fluid source from sub-lithospheric magmas.
Graphite, once crystallized, is isotopically inert to further alteration, and this character makes it a potential candidate for detecting fluid flow patterns and changing fluid regimes. The isotopically inert character and sluggish growth rates of graphite would allow preexisting graphite to preserve their isotopic composition attained at the time of crystallization even in situations where there is evidence subsequent influx of highly oxidizing fluids. Changing fluid regimes in the crust has been investigated from graphite crystals in association with orthopyroxene and feldspars in charnockites adjacent to a calc-silicate horizon at Korani, southern India (Santosh et al., 2003). Such graphite exhibit δ13CVPDB values in the range of −4.1 to −9.7‰, suggesting precipitation by reduction from CO2-rich fluids. Microscale traverses in two domains of a single crystal show extreme 13C enrichment within the core portion, with majority of the δ13CVPDB values lying between −4 to −5‰. A systematic carbon isotopic variation is observed with δ13CVPDB values becoming progressively lighter (up to −8.3‰) towards the rims suggesting changing fluid regimes during graphite growth. A two-stage model for precipitation of graphite is developed whereby initial graphite growth was largely controlled by CO2 derived from decarbonation reactions in the adjacent carbonate layers resulting in the formation of 13C enriched cores. These graphite cores acted as effective oxygen buffer in precipitating mantles of 13C depleted graphite from a subsequent regime of CO2-rich fluids which extensively infiltrated from external (magmatic) sources. The CO2 also stabilized the anhydrous mineralogy of the charnockite, promoting the growth of orthopyroxene preserved in pristine state with total absence of biotite. The graphite concentration at calc-silicate-pyroxene gneiss interface at Skallevikshalsen is in harmony with such a model, where we have observed very similar δ13CVPDB values for graphite. The pooling of CO2 and oversaturation at structurally controlled contact zones, where carbonate rocks often act as impermeable layer for CO2-rich fluids (cf. Harley and Santosh, 1995).
The stability of graphite is controlled directly by the oxygen and hydrogen fugacity. Therefore, under favorable conditions of fluid-rock interaction with low fluid/rock ratio, carbon will be locked up as graphite in the continental crust for extremely long duration (in the order of 108 years). A plausible mechanism of returning carbon stored as crustal graphitic carbon to the mainstream carbon cycle is by consumption of graphite to generate a mobile graphite-saturated C-O-H fluid. This normally happens during progressive devolatilization of graphitic metasediments (Connolly and Cesare, 1993). Hence, it is essential to evaluate the processes that occur in graphite-bearing rocks in the continental crust in order to clarify whether graphite acts as a sink or source for carbon in the short-term carbon cycle. In particular, the whereabouts of carbon during anatexis is ambiguous and poorly characterized.
Carbon isotope reorganization in graphite during anatexisThe examples of graphite-bearing leucosome can be considered as typical of anatexis of carbonaceous metapelitic rocks in the continental crust. Two scenarios are considered here. Previous studies on partially melted xenoliths from El Hoyazo, where melting occurred in the presence of a graphite-saturated C-O-H (‘GCOH’; Connolly and Cesare, 1993) fluid (Cesare and Maineri, 1999), suggested that the progress of melting will result in preferential partition of the H2O component of the fluid into the melt. This fluid re-speciation and graphite precipitation is interpreted with the model reaction of mineral assemblage A + GCOH = assemblage B + melt + graphite (Cesare, 1995). Graphite precipitates until either CH4 or CO2 is consumed and the fluid is converted to a binary CO2 + H2O or CH4 + H2O fluid. At this point graphite precipitation would be ceased. In such a case of closed system partial melting, we observe homogenous carbon isotope composition, irrespective of the textural association of graphite, within the scale of our investigation.
Another possible scenario is the widely recognized model for crustal anatexis, namely biotite dehydration melting during regional metamorphism. Prior to dehydration melting carbon isotopes evolve through a process of graphitization associated with an early loss of hydrocarbons, resulting in a limited enrichment of 13C values of graphite (Wada et al., 1995). This explains the higher δ13C values of graphite in high-grade metamorphic rocks and restites in all three terrains examined in this study. However, unlike in the fluid-present scenario in which melting induces graphite precipitation, biotite dehydration melting of graphitic rocks is accompanied by the reduction of Fe3+ in biotite and oxidation of graphite (Cesare et al., 2005). The dissolution of graphite will result in the formation of CO2 and minor amounts of CO. These species will be mostly partitioned into the fluid phase coexisting with the melt, only minor amounts being dissolved into the melt. Release of CO to the fluid phase will result in the enrichment of 13C in the co-existing CO2. Fluid mobilization and graphite precipitation will commence during melt crystallization as the fluid becomes oversaturated in C (Cesare, 1995; Satish-Kumar, 2005).
Partial melting and migmatization of graphite-bearing lithologies are important in understanding how carbon behaves during high-temperature metamorphism. Satish-Kumar et al. (2011b) documented carbon isotopic zonation in coarse graphite grains in a leucosome in a metapelitic granulite from the Kerala Khondalite Belt, in southern India. Partitioning of carbon isotope during reductive deposition of graphite from a COH fluid pooled in fractures and melt rich domains during anatexis can thus explain the observed carbon isotopic variation in graphite from the metapelitic migmatites. A compilation of available carbon isotope data in the KKB metapelites shows a spread from −4 to −34‰ (Binu-Lal et al., 2003). This large spread of carbon isotope values for graphite has also been reported for hydrothermal graphite from New Hampshire, where mixing of oxidizing and reducing fluids resulted in the intermediate isotope values of graphite (Rumble and Hoering, 1986). The large spread of the carbon isotope data observed in the leucosomes in LHC, especially the presence of enrichment in 13C, indicate that the carbon isotopic composition of graphite evolved from an initial biogenic value by a process involving the oxidation and/or re-speciation and/or reduction during high-grade metamorphism and anatexis in the deep crust. As a consequence, the anatexis of carbon-bearing rocks in the continental collision zones will result in a reorganization of carbon isotopic composition of graphite.
The carbon isotopic data presented here are consistent with an initial enrichment of 13C during melt generation and a subsequent depletion in 13C during melt crystallization, which is reverse to what observed for graphite precipitating from an evolving externally derived infiltrating CO2-rich fluid (Farquhar and Chacko, 1991; Farquhar et al., 1999). A two stage isotope enrichment process is proposed for the carbon isotope composition of graphite with an initial biogenic value. In the first stage, during prograde metamorphism, volatile loss during graphitization process will enrich the graphite (cf. Ohmoto and Kerrick, 1977; Nakamura et al., 2017, 2020). The second enrichment is by a dissolution-reprecipitation mechanism during anatexis and melt crystallization (cf. Satish-Kumar et al., 2011b). The enrichment of 13C in the continental crust has important implications for not only understanding the carbon isotope geochemistry but also in modeling the movement of carbon in deep crust using carbon isotopes.
The above discussion implies that carbon isotopic composition of graphite can be used as a monitor for fluid-rock interaction in continental collision zones. Conventional isotopic analyses can characterize the general biogenic or abiogenic nature of graphite. Bulk analyses of minerals in conventional stable isotopic studies provide only the average isotopic composition at the grain scale. High-precision stable isotopic measurements at the micrometer scale within the domain of individual grains can tap potential information on isotopic alteration and fluid behavior with respect to changing pressure-temperature conditions including pattern of fluid flow, extent of fluid-rock interaction and changing fluid regimes (Satish-Kumar et al., 1998). Furthermore, careful textural observations of graphite crystals can also guide us to understand the crystal growth in graphite from fluids as documented in a case study in Naxos marbles (Satish-Kumar et al., 2011c). The carbon isotopic composition of graphite in the LHC metamorphic rocks were modified by volatilization and graphitization during early prograde metamorphism and also during partial melting of graphite-bearing metapelite. In addition, mixing of CO2 derived from decarbonation reactions may have also caused in the enrichment of 13C. At the present scale of our investigation, we were unable to find any direct evidence for CO2 fluids derived from deeper levels of the crust or mantle, as invoked by several previous studies (e.g., Farquhar and Chacko, 1991; Santosh and Wada, 1993), instead the field evidence suggests local re-equilibration of carbon isotopes. Indeed, the vein type graphite was deposited from fluids, but the close proximity of vein to the marble horizon indicates a direct genetic relation with CO2 released through decarbonation reactions. The data presented in this study, thus, clearly signify the importance of graphite in tracing the carbon budget in the continental crust.
Graphitization, anatexis and reprecipitation of graphite in continental crustFigure 7 illustrates a plausible model on the evolution of graphite in the continental crust during orogenesis and regional high-grade metamorphism. The evolution of carbon (as graphite) in the continental crust can be considered in four major processes: 1) Devolatilization of organic material and graphitization, 2) Exchange with carbonate carbon; 3) Partial melting of graphite-bearing lithologies, and 4) Re-deposition of graphite during cooling and exhumation of the crust.
Several previous studies have considered the graphitization process from a graphite crystallinity perspective. Only few studies have addressed isotopic evolution during prograde metamorphism, (e.g., Wada et al., 1995; Ueno et al., 2002; Nakamura and Akai, 2013; Kiran et al., 2021), where they found that graphitization is associated with a concomitant isotopic enrichment in graphite. This is consistent with our observations in high-grade metamorphosed silicate rocks in LHC in this study. As discussed above, metamorphism of organic materials trapped in sediments in the presence of carbonate minerals will lead to temperature dependent isotopic exchange and metamorphic graphite will have isotopic compositions very close to carbonate carbon, as in the case of granulite facies to UHT metamorphism in several orgogenic belts across the globe (e.g., Kitchen and Valley, 1995; Satish-Kumar et al., 2002). Thus, devolatilization of organic materials and exchange with carbonate carbon (Stage-I, Fig. 7) are major processes involved in the isotopic evolution of graphite in continental crust during prograde metamorphism associated with orogenesis.
During granulite facies metamorphism, nearing to the peak temperature conditions, graphite-bearing metamorphic rocks are subjected to partial melting. Depending on the water content of the lithologies the temperature conditions of melting vary and furthermore the stability of graphite is also controlled by the prevailing redox conditions. As discussed in Carvalho et al. (2023a), the fluid composition and fate of graphite depends on the oxygen fugacity conditions and melting favors oxidation of graphite. This results in the production of graphite saturated COH fluids. The presence of coarse-grained graphite crystals in leucosomes in anatectic metapelitic rocks in LHC is clearly pointing out to a reprecipitation of graphite during melt crystallization (Stage-II, Fig. 7). Carbon isotope enrichment in this stage suggest that a fraction of 12C enriched carbon-bearing residual fluid is being expelled from the system or being trapped as fluid inclusions. Further studies on carbon isotopic composition in fluid inclusion in leucosomes are required to testify this interpretation.
The mobility of carbon-bearing fluids in the continental crust and deposition of graphite in favorable conditions have been well documented in several previous studies. The presence of decimeter to meter thick monomineralic graphite veins, such as those observed in Sri Lanka are classic examples (e.g., Touret et al., 2019 and references therein). However, the source of such large volume of carbon saturated fluids is still debated. The results presented here suggest to local decarbonation and graphite deposition has occurred in the upper crustal conditions during retrograde P-T conditions (Stage-III, Fig. 7). Carbon isotopic composition of graphite in veins are supporting this assertion.
Implications for carbon geodynamic cycle in continental crustBased on the carbon isotopic composition of high-grade metamorphic rocks in continental crust, it is clear that orogenesis results in the re-organization of carbon isotopes. The isotope distribution is controlled by metamorphism and partial melting in the crust, whereas melting and recycling can play a significant role in deep crust and upper mantle conditions. The wide range of carbon isotopic distribution in graphite (Fig. 6) envisages that a combination of devolatilization, decarbonation, and fluid rock/melt interaction is active in continental crust. In addition to the crustal carbon inventories, carbon existing in the deep mantle and core also can impact the carbon geodynamic cycle. Core-mantle interaction can also play a role in the carbon isotopic composition of the bulk silicate earth, as shown in recent experimental studies (Satish-Kumar et al., 2011a; Dasgupta, 2013).
It is significant to consider the quantity of carbon that can be recycled and stored in the continental crust in the form of graphite. The present-day carbon imbalance in the subduction zones is estimated to be approximately 2.5-3.5 × 1012 Moles of C yr−1 (Kerrick and Connolly, 2001). Rough estimate of recoverable graphite reserves is in the order of 1014 M. However, if we take into account the estimated total organic carbon reservoir in the crust (e.g., 1.2 × 1021 Moles of carbon prior to the Cambrian organic explosion; Des Marais et al., 1992) the carbon imbalance in the global carbon cycle can be resolved. The fact that the carbon residing in the continental crust in the form of graphite will have a crustal residence time of hundreds of millions of years provide evidence for the long-term storage of carbon. Furthermore, the results of our study preclude the possibility of carbon release during partial melting process in the continental crust. We therefore suggest that the continental crust acts as a major carbon sink in the long-term carbon cycle in the Earth.
I would like to express my sincere gratitude to Hideki Wada and M. Santosh for introducing the world of carbon isotope studies in graphite and for valuable discussions at various stages of this work. Sampling of graphite-bearing rocks from the Lützow Holm Complex was carried out with the logistics and field support of JARE-46 (2004-2005) expedition. I thank the expedition members for their constant support in the field. So Hayato and S. Kiran are thanked for the support in the carbon isotope analyses of graphite. This work was supported by MEXT KAKENHI Grant Numbers JP15H05831, JP20KK0081, and JP22H04932 to MS-K. Finally, I thank Bruna Carvalho and two other reviewers for their constructive comments and Tetsuo Kawakami for his efficient handling of this manuscript.
