Content uploaded by Yuji Sano
Author content
All content in this area was uploaded by Yuji Sano on Oct 07, 2017
Content may be subject to copyright.
Early trace of life from 3.95 Ga sedimentary rocks in Labrador, Canada 1
2
Takayuki Tashiro1, Akizumi Ishida2,3, Masako Hori2,4, Motoko Igisu5, Mizuho Koike2, 3
Pauline Méjean2, Naoto Takahata2, Yuji Sano2*, Tsuyoshi Komiya1*
4
5
1 Department of Earth Science and Astronomy, Graduate School of Arts and Sciences, 6
The University of Tokyo, Tokyo 153-8902, Japan 7
2 Atmosphere and Ocean Research Institute, The University of Tokyo, Chiba 277-8564, 8
Japan 9
3 WiscSIMS lab., NASA Astrobiology Institute, Department of Geoscience, University of 10
Wisconsin-Madison, Wisconsin 53706, USA 11
4 Department of Arts and Sciences, Osaka Kyoiku University, Osaka 582-8582, Japan 12
5 Laboratory of Ocean-Earth Life Evolution Research, Japan Agency for Marine-Earth 13
Science and Technology (JAMSTEC), Kanagawa 237-0061, Japan 14
15
Corresponding authors 16
Tsuyoshi Komiya, e-mail address: komiya@ea.c.u-tokyo.ac.jp, Tel: +81 3-5454-6609, 17
Fax: +81 3-5465-8244 18
Yuji Sano, e-mail address: ysano@aori.u-tokyo.ac.jp, Tel:+81 4-7136-6100, Fax: +81 19
4-7136-6067. 20
21
22
Submitted to Nature, 2017/2/14 23
Revised, 2017/05/12 24
The vestiges of life in Eoarchean rocks have the potential to elucidate the 25
origin of life. However, gathering evidence from many terrains is often not always 26
possible1-3, thus biogenic graphite has thus far been found only in the 3.7-3.8 Ga 27
Isua supracrustal belt4-7. Here we show total organic carbon contents and carbon 28
isotope values of graphite (δ13Corg) and carbonate (δ13Ccarb) in the oldest 29
metasedimentary rocks from northern Labrador8,9. Some pelitic rocks have rather 30
low δ13Corg values of -28.2; comparable to the lowest value in younger rocks. The 31
consistency between crystallization temperatures of the graphite and metamorphic 32
temperature of the host rocks establishes that the graphite does not originate from 33
later contamination. A clear correlation between the δ13Corg values and 34
metamorphic grade indicates that variations in the δ13Corg values are due to 35
metamorphism, and that the pre-metamorphic value was lower than the minimum 36
value. We concluded that the large fractionation between the δ13Ccarb and δ13Corg 37
values, up to 25‰, indicates the oldest evidence of organisms >3.95 Ga. The 38
discovery of the biogenic graphite enables geochemical study of the biogenic 39
materials themselves, and will open the door to elucidating the early life not only in 40
the earth but also in other planets. 41
The presence of life on early Earth is still controversial due to the scarcity and 42
poor preservation of the Eoarchean records. Isotopic compositions of graphite in the 43
Eoarchean sedimentary rocks in the Isua supracrustal belt (ISB) suggest that the graphite 44
grains have biogenic origins because of the enrichment of light carbon isotope4-7. 45
However, biogenic graphite has not been discovered in the 3.83 Ga Akilia association and 46
3.75 Ga Nuvvuagittuq supracrustal belt1-3. Recent reassessment of U-Pb dating and 47
cathodoluminescence observation of zircons from the Uivak Gneiss in Saglek Block (Fig. 48
1), northern Labrador, Canada indicated the presence of the oldest supracrustal rock in 49
the world, intruded by the >3.95 Ga Uivak-Iqaluk Gneiss8,9. This paper, is the first to 50
report on the occurrence and geochemical characteristics of the oldest graphite. 51
We found graphite from the oldest metasedimentary rocks, including pelitic 52
rocks, conglomerates, carbonate rocks, and chert nodules in the carbonate rocks from five 53
areas of different metamorphic grades in the Saglek Block (Fig. 1). We obtained 54
concentrations and isotopic compositions of the graphite using a graphite combustion 55
method with an elemental analyzer (vario MICRO cube, Elementar) connected to a mass 56
spectrometer (IsoPrime100, Isoprime)10. We also performed in-situ analyses of their 57
carbon isotope ratios using a NanoSIMS 50 instrument. The results are presented as δ13C 58
values relative to a VPDB (Vienna Pee Dee Belemnite) standard. 59
60
Pelitic rocks occur ubiquitously in the Saglek block and often have preserved 61
bedding planes (Extended Data Fig. 1-2). The graphite grains occur as aggregates and 62
elongated shapes (up to a few tens to hundreds µm long) mostly along grain boundaries 63
of other minerals, parallel to the bedding planes (Extended Data Fig. 2ef). Some are 64
enclosed within quartz, garnet and biotite to form small globular shapes. The occurrence 65
of graphite is analogous to the Phanerozoic organic matter with laminated morphology11. 66
The pelitic rocks have large variations between 0.02 and 0.62 wt.% in TOC (Total organic 67
carbon) contents and between -28.2 and -11.0‰ in δ13Corg values, respectively (Extended 68
Data Table 2), and show a negative correlation between them (Fig. 2). Graphite grains in 69
a pelitic rock (LAF491) at St. John’s Harbour South (SJHS) showed δ13Corg values 70
between -19.3 and -30.8‰ (Extended Data Table 3), which were consistent with the 71
whole-rock carbon isotope value (Extended Data Fig. 3a). 72
Conglomerates at St. John’s Harbour East (SJHE) contain pebble- to 73
boulder-sized quartzite clasts (Extended Data Fig. 2b). Most of the graphite grains are 74
elongated (up to a few tens µm long) or form aggregates along the bedding planes. Some 75
graphite grains also occur as small globular inclusions of 1-2 µm, in the quartz, garnet, 76
biotite, plagioclase, and amphibole grains. The TOC contents of the matrices are 77
relatively low (~0.07 wt.%) and the δ13Corg values range between -27.6 and -20.8‰ (Fig. 78
2). 79
Carbonate rocks are also found in SJHE, and have some chert nodules 80
(Extended Data Fig. 2c). Graphite grains form elongated shapes or aggregates of globules. 81
The carbonate rocks have distinctive positive La, Eu, and Y anomalies on the 82
shale-normalized rare earth element (REE) patterns, diagnostic of chemical sediments, 83
precipitated from seawater mixed with hydrothermal fluid (Extended Data Figure 3b). 84
The TOC contents of carbonate rocks range between 0.09 and 0.16 wt.%, and the δ13Corg 85
and δ13Ccarb values are between -6.9 and -9.9‰ and between -3.8 and -2.6‰, respectively 86
(Extended Data Table 2, Fig. 2). It is well known that the δ13Ccarb values decreases with 87
later alteration7 so that a δ13C value of marine inorganic carbon in the Eoarchean is 88
estimated to be higher than the maximum value (-2.6‰). The graphite grains in the chert 89
nodules form globular shapes, ranging between 1 and 100 µm across, and occur along 90
grain boundaries or as inclusions within quartz grains (Extended Data Fig. 2h). The 91
graphite grains range between -26.1 and -33.6‰ in δ13Corg values (Extended Data Table 92
3) whereas the TOC contents and δ13Corg values of the host rock are ca. 0.02 wt.% and 93
-10‰, respectively (Fig. 2). No graphite was recognized in the Pangertok Inlet and cherts 94
in all the areas. In the SJHE area, the δ13Corg values are apparently dependent on the 95
lithology, and increase in order from conglomerate through pelitic rocks to carbonate 96
rocks and chert nodules (Fig. 2). The lithology-dependent variation of the δ13Corg values 97
suggests that the organic matter is autochthonous. 98
99
Metamorphic temperatures of host rocks with graphite grains were estimated 100
via two methods: conventional mineral parageneses of ambient metabasites and 101
geothermometry of garnet-biotite pairs in pelitic rocks. The metabasites in SJHS, SJHE, 102
and Big Island have mineral parageneses of hornblende (Hbl), plagioclase (Pl) and 103
titanite (Ttn), whereas those in Shuldham Island have a mineral paragenesis of 104
clinopyroxene (Cpx)+Hbl+Pl. The metamorphic temperatures of SJHS, SJHE, Big Island, 105
and Shuldham Island were estimated from chemical compositions of garnet and biotite as 106
653 ± 16 (1σ) ºC, 691 ± 26 ºC, 585 ± 45 ºC and 700 to 800 ºC, respectively (see Extended 107
Data). 108
The study also estimated crystallization temperatures of graphite based on 109
Raman spectra of graphite12,13 with confocal laser Raman microspectroscopy. It was 110
found that the crystallization temperatures of graphite are over 563 ± 50 ºC for the pelitic 111
rocks, conglomerates, and carbonate rocks, and between 536 ± 50 and 622 ± 50 ºC for the 112
chert nodules (Extended Data Table 4). The estimated crystallization temperatures of 113
graphite are mostly consistent with the metamorphic temperature of the host rocks, 114
except for chert nodules (Extended Data Fig. 4). The metamorphic history of the Saglek 115
Block is complex, and some major metamorphisms were estimated through dating of the 116
zircons of the Uivak-Iqaluk Gneisses9. The first metamorphic episode was estimated to 117
have occurred at ca. 3.89 Ga due to the intrusion of a younger suite of Uivak Gneisses. 118
The second episode occurred at 3.6 Ga due to the intrusion of Uivak II gneiss14, and the 119
third at 2.7 Ga from U-Pb dating of zircon overgrowths9,15 and secondary isochrons of the 120
Uivak Gneiss16. The discontinuous intrusions of many generations of granitoid at least ca. 121
3.95, 3.87, 3.6 and 3.3 Ga9,14,17,18 and lack of basal conglomerates indicate that the 122
supracrustal rocks have settled throughout in a deep crust so that later contamination of 123
sedimentary graphite grains would have been impossible from 3.9 to 3.3 Ga. The lines of 124
evidence of the occurrence of graphite parallel to the bedding planes, consistency 125
between the Raman spectra and metamorphic temperature, lithology-dependent δ13Corg 126
variation and discontinuous metamorphic and magmatic ages suggest that the graphite 127
has a sedimentary origin that predates the first metamorphic episode. 128
The maximum δ13Corg values of pelitic rocks increase between -18.1 and 129
-11.0‰ with an increasing metamorphic grade from amphibolite facies to granulite facies 130
(Fig. 3). The good correlation indicates that the variations of δ13Corg values were due to 131
later metamorphism so that the pre-metamorphic δ13Corg value was lower than the 132
minimum δ13Corg values of -28.2‰. The low δ13Corg values of pelitic rocks and 133
conglomerates are consistent with the biotic origin of graphite. However, it is well known 134
that 13C-depleted graphite could be produced by abiotic processes including high 135
temperature disproportionation of siderite19,20, low temperature Fischer-Tropsch-type 136
(FTT) synthesis21 and incorporation of meteoritic organic matter7. The disproportionation 137
of the siderite can be represented by the following reaction19: 6FeCO3 → 2Fe3O4 + 5CO2 138
+ C, which occurs at temperature above 450 ºC for pure siderite. However, the graphite in 139
clastic sedimentary rocks is inconsistent with siderite decomposition because they contain 140
neither siderite nor magnetite. On the other hand, the siderite decomposition pathway 141
cannot be excluded for the carbonate rocks because they contain magnetite, and the 142
isotopic differences, 4 to 6‰, between graphite and carbonate (δ13Ccarb -δ13Corg) in the 143
carbonate rocks are consistent with equilibrium isotopic fractionation during graphite 144
formation from siderite at ~700 ºC20. The FTT synthesis requires an appropriate catalyst 145
such as Ni-Fe metal and magnetite, and a source of H2 and CO, and the reaction is 146
operated between 200 and 350 ºC21,22. Although the H2-rich reducing conditions can be 147
produced by hydrothermal alternation of ultramafic rocks21,22, the highly 13C-depleted 148
graphite grains are present only in the clastic sedimentary rocks without ultramafic 149
rock-derived chromite, hence the contribution of the FTT synthesis is insignificant. 150
151
We conclude that the graphite from clastic sedimentary rocks in the Saglek 152
Block has a biogenic origin and the primary δ13Corg and δ13Ccarb values were estimated to 153
be <-28.2 and >-2.6‰, respectively. As a result, the isotopic fractionation between 154
graphite and carbonate (δ13Ccarb-δ13Corg) reached -25.6‰ more than those in 155
turbidite-derived sedimentary rocks of the Isua supracrustal belt6. The large fractionation 156
provides the oldest evidence for autotrophs, utilizing reductive acetyl-CoA pathway or 157
the Calvin cycle, over 3.95 Ga (Extended Data Fig. 5). 158
159
160
References 161
1 Fedo, C. M. & Whitehouse, M. J. Metasomatic origin of quartz-pyroxene rock, 162
Akilia, Greenland, and implications for Earth's earliest life. Science 296, 163
1448-1452 (2002). 164
2 Papineau, D. et al. Young poorly crystalline graphite in the >3.8-Gyr-old 165
Nuvvuagittuq banded iron formation. Nature Geoscience 4, 376-379 (2011). 166
3 Sano, Y., Terada, K., Takahashi, Y. & Nutman, A. P. Origin of life from apatite 167
dating? Nature 400, 127 (1999). 168
4 Hayes, J. M., Kaplan, I. R. & Wedeking, K. W. in Earth’s Earliest Biosphere: Its 169
Origin and Evolution (ed J.W. Schopf), 93-134 (Princeton University Press, 170
1983). 171
5 Ohtomo, Y., Kakegawa, T., Ishida, A., Nagase, T. & Rosing, M. T. Evidence for 172
biogenic graphite in early Archaean Isua metasedimentary rocks. Nature 173
Geoscience 7, 25-28 (2014). 174
6 Rosing, M. T. 13C-depleted carbon microparticles in > 3700-Ma sea-floor 175
sedimentary rocks from West Greenland. Science 283, 674-676 (1999). 176
7 Schidlowski, M., Appel, P. W. U., Eichmann, R. & Junge, C. E. Carbon isotope 177
geochemistry of the 3.7 x 109-yr-old Isua sediments, West Greenland: implications 178
for the Archaean carbon and oxygen cycles. Geochimica et Cosmochimica Acta 179
43, 189-199 (1979). 180
8 Komiya, T. et al. Geology of the Eoarchean, >3.95 Ga, Nulliak supracrustal rocks 181
in the Saglek Block, northern Labrador, Canada: The oldest geological evidence 182
for plate tectonics. Tectonophysics 662, 40-66 (2015). 183
9 Shimojo, M. et al. Occurrence and geochronology of the Eoarchean, ∼3.9 Ga, 184
Iqaluk Gneiss in the Saglek Block, northern Labrador, Canada: Evidence for the 185
oldest supracrustal rocks in the world. Precambrian Research 278, 218-243 186
(2016). 187
10 Vandenbroucke, M. Kerogen: from types to models of chemical structure. Oil & 188
Gas Science and Technology - Rev. IFP 58, 243-269 (2003). 189
11 Vandenbroucke, M. & Largeau, C. Kerogen origin, evolution and structure. 190
Organic Geochemistry 38, 719-833 (2007). 191
12 Beyssac, O., Goffé, B., Chopin, C. & Rouzaud, J. N. Raman spectra of 192
carbonaceous material in metasediments: a new geothermometer. Journal of 193
Metamorphic Geology 20, 859-871 (2002). 194
13 Igisu, M. et al. FTIR microspectroscopy of Ediacaran phosphatized microfossils 195
from the Doushantuo Formation, Weng'an, South China. Gondwana Research 25, 196
1120-1138 (2014). 197
14 Collerson, K. D. & Bridgwater, D. in Trondhjemites, dacites, and related rock (ed 198
F. Barker), (Elsevier, 1979). 199
15 Nutman, A. P. & Collerson, K. D. Very early Archean crustal-accretion complexes 200
preserved in the North Atlantic craton. Geology 19, 791-794 (1991). 201
16 Collerson, K. D. The Archean gneiss complex of northern Labrador. 2. Mineral 202
ages, secondary isochrons, and diffusion of strontium during polymetamorphism 203
of the Uivak gneisses. Canadian Journal of Earth Sciences 20, 707-718 (1983a). 204
17 Schiøtte, L., Compston, W. & Bridgwater, D. U-Th-Pb ages of single zircons in 205
Archaean supracrustals from Nain Province, Labrador, Canada. Canadian Journal 206
of Earth Science 26, 2636-2644 (1989). 207
18 Komiya, T. et al. A prolonged granitoid formation in Saglek Block, Labrador: 208
Zonal growth and crustal reworking of continental crust in the Eoarchean. 209
Geoscience Frontiers in press (2016). 210
19 Perry, E. C. & Ahmad, S. N. Carbon isotope composition of graphite and 211
carbonate minerals from 3.8-AE metamorphosed sediments, Isukasia, Greenland. 212
Earth and Planetary Science Letters 36, 280-284 (1977). 213
20 van Zuilen, M. A. et al. Graphite and carbonates in the 3.8 Ga old Isua 214
Supracrustal Belt, southern West Greenland. Precambrian Research 126, 331-348 215
(2003). 216
21 Horita, J. Some perspectives on isotope biosignatures for early life. Chemical 217
Geology 218, 171-186 (2005). 218
22 De Gregorio, B., Sharp, T., Rushdi, A. & Simoneit, B. T. in Earliest life on Earth: 219
Habitats, environments and methods of detection (eds Suzanne D. Golding & 220
Miryam Glikson), 239-289 (Springer, 2011). 221
222
FIGURE LEGENDS 223
224
Figure 1. Geological maps and sample localities in Saglek Block. (A) Distribution of the 225
Archean cratons, and Saglek Block is the western part of the North Atlantic 226
Craton. (B) Distribution of the Archean terrains in the western part of the North 227
Atlantic Craton. The Saglek Block is equivalent to the Akulleq terrane in West 228
Greenland. (C) Northeastern part of the Saglek Block, showing our five studied 229
areas. (D) Localities of the graphite-bearing samples in the studied areas. 230
231
Figure 2. Correlation between total organic carbon (TOC) contents and δ13Corg values. 232
We analyzed the whole rock TOC contents and carbon isotopes of 20 pelitic 233
rocks, four conglomerates, three carbonate rocks and two chert nodules in 234
carbonate rocks. The carbonate rocks and silica nodules in the carbonate rocks 235
have relatively higher δ13Corg values whereas the δ13Corg values of matrices of the 236
conglomerates are much lower. The TOC contents and the δ13Corg values of the 237
pelitic rocks have large variations and display a negative correlation. 238
239
Figure 3. Comparison between the δ13Corg values of pelitic rocks and metamorphic 240
grades. Metamorphic grades are estimated from mineral parageneses of the 241
surrounding metabasalts and garnet-biotite geothermometry of the 242
graphite-bearing pelitic rocks. The 13Corg values are positively correlated with 243
the metamorphic grades. 244
245
FIGURE LEGENDS in EXTENDED DATA 246
Extended Data Figure 1. Detailed geological maps of four areas in the Saglek Block. 247
(a) A geological map of St. John’s Harbour South area (SJHS). The area is 248
composed of the supracrustal rocks, Iqaluk-Uivak Gneisses, Saglek dykes, young 249
granite intrusion and the Proterozoic mafic dikes. The supracrustal rocks form a 250
NS-trending belt, and are intruded by ca. 3.95 Ga Iqaluk-Uivak Gneisses. The 251
pelitic rocks are predominant in the supracrustal rocks. (b) A geological map of 252
Big Island area. The area is subdivided into two parts by a NS-trending fault. The 253
eastern side is composed of the supracrustal rocks, Iqaluk-Uivak Gneisses, 254
Saglek dykes, young granite intrusion and the Proterozoic mafic dykes. The 255
western side is predominant in pelitic rocks, and contains ultramafic and mafic 256
rocks, and carbonate rocks. (c) A geological map of a small point of the western 257
coast of the Shuldham Island. The area is characterized by ultramafic rocks with 258
large olivine-needle structures. The ultramafic rock-bearing body consists of 259
harzburgitic ultramafic rocks, olivine-clinopyroxene rocks, 260
clinopyroxene-hornblendite, gabbroic rocks, fine-grained amphibolite and pelitic 261
rocks, in ascending order. (d) A geological map of St. John’s Harbour East area 262
(SJHE). A supracrustal belt is composed of some fault-bounded blocks from 263
ultramafic rocks through mafic rocks to sedimentary rocks of pelitic rocks, 264
carbonate rocks and cherts in ascending order. 265
266
Extended Data Figure 2. Photos of outcrops and thin sections of the metasedimentary 267
rocks. (a) An outcrop of pelitic rocks (LAA269, LAA270) at St. John’s Harbour 268
South. (b) An outcrop of a conglomerate (LAD849A, LAD849B, LAD849C, 269
LAD852) at St. John’s Harbour East. The photo was taken from the east. A large 270
siliceous clast in the conglomerate displays north to south extension. (c) A 271
carbonate rock (LAA742) with chert nodules at St John’s Harbour East. (d) An 272
outcrop of the chert at St. John’s Harbour East. (e) A representative microscopic 273
image of a pelitic rock (LAF492), containing biotite (Bt), garnet (Grt), quartz 274
(Qtz), pyrrhotite (Po) and graphite. The most graphite grains have elongated 275
shapes and occur along biotite grains and within garnet grains. (f) Another 276
representative microscopic image of a pelitic rock (LAF491). The graphite 277
occurs along the biotite grains, forming bedding planes, or along the cleavages of 278
the biotites. (g) A microscopic image of a carbonate rock (LAA766). The 279
needle-like mineral is serpentine (Srp), and sparry carbonate consists of calcite 280
(Cal) and dolomite (Dol). Magnetite (Mgt)-rich rings are present in the 281
fine-grained carbonate (Cal). (h) A microscopic image of a chert nodule 282
(LAA760) in the carbonate rock (c). The graphite grains have globular shapes, 283
and form an aggregate. 284
285
Extended Data Figure 3. Carbon isotope values of individual graphite grains and 286
REE+Y patterns of carbonate rocks. (a) Carbon isotope values of individual 287
graphite grains in a pelitic rock (LAF491) and a chert nodule of carbonate rock 288
(LAF760). The graphite grains in the LAF491 range from -19.3 to -30.8‰ in 289
δ13Corg values whereas those in the LAF760 vary from -26.1 to 33.6‰ in δ13Corg 290
values. The formers are consistent with the Whole-rock carbon isotope ratio 291
(-28.2‰) but the latter is much lower than the whole-rock value (-10.3‰). (b) 292
Post-Archean Australian shale (PAAS)-normalized REE + Y diagrams of 293
carbonate rocks with low Y and Zr contents. The carbonate rocks show 294
diagnostic Eu and Y anomalies in the SJHS (A), Big Island (B), SJHE (C) and 295
Pangertok Inlet (D). 296
297
Extended Data Figure 4. The comparison between metamorphic temperature and 298
crystallization temperature of graphite. The metamorphic temperatures were 299
estimated from mineral parageneses of metabasaltic rocks and compositions of 300
garnet and biotite in pelitic rocks, whereas the crystallization temperatures were 301
estimated from Raman spectra of graphite. In the case of absence of D1 bands, 302
the estimated crystallization temperature is over 650 ± 50 ºC. The estimated 303
crystallization temperatures of graphite are consistent with the metamorphic 304
temperatures except for those from a chert nodule. 305
306
Extended Data Figure 5. The distribution of the δ13Corg values in Saglek Block and 307
Isua supracrustal belt. The δ13Corg values in Isua supracrustal belt range 308
between -28 and -6‰35-42. The lower column shows variations of carbon isotope 309
fractionation in four different carbon fixation pathways by modern autotrophic 310
bacteria43. 311
312
313
Methods 314
No statistical methods were used to predetermine sample size. The experiments were not 315
randomized and the investigators were not blinded to allocation during experiments and 316
outcome assessment. 317
Extraction and isotope analysis of organic carbon. We collected 156 sedimentary 318
rocks from the Saglek area, and found graphite in 54 rock samples. We selected 28 319
samples with a larger amount of the graphite to cover all of lithologies and studied areas 320
for geochemical works. The powdered rock samples were prepared by crushing rock 321
chips. They (1-3 g) were decarbonated by 6 N HCl at 70 °C for 12 hours. The HCl treated 322
samples were further treated with mixed acid of HCl, HF and H2O (1:4:4 v/v/v in 10 N 323
HCl) at 60 °C for 3 days and repeated 3-4 times, followed by 6 N HCl at 70 °C for 12 324
hours in order to eliminate the remaining elements which are able to form complex 325
fluorides such as ralstonite (NaxMgxAl2–x(FOH)6H2O) upon drying10. All acid reactions 326
were performed in shaking bath to facilitate acid attacks. The acid-treated samples were 327
finally washed with pure water several times and freeze-dried. The HCl-HF-treated 328
sample powders were placed in an Sn capsule in the range of 1-40 mg and 30-500 µg, 329
respectively, and were combusted with oxygen under He career flow at 1,100 °C in an 330
elemental analyzer (vario MICRO cube, Elementar) connected to mass spectrometer 331
(IsoPrime100, Isoprime), housed at Atmosphere and Ocean Research Institute, The 332
University of Tokyo in order to measure carbon concentration and isotope composition, 333
respectively. Carbon, nitrogen, and sulfur in samples were converted into CO2, N2, and 334
SO2, respectively. Carbon concentrations and isotope compositions of samples were 335
calibrated against an in-house standard material (sulfanilamide), whose carbon 336
concentration and isotope composition were known (41.81 wt.%C, δ13C = -26.6‰). 337
Based on the replicate analyses of the in-house standard material, analytical 338
reproducibility is within ±0.5‰ (2σ). Results were reported as δ13C values relative to a 339
VPDB standard. 340
Carbon isotopes of carbonate. Powdered samples were prepared from several parts of 341
fresh-cut surfaces of rock samples using a micro-drill with a 3 mm-diameter bit. They 342
were analyzed with a Finnigan MAT Delta Plus mass spectrometer interfaced with a Gas 343
Bench II, housed at Atmosphere and Ocean Research Institute, The University of Tokyo. 344
The samples were reacted with purified H3PO4 at 70°C in a glass vial preliminary filled 345
with He gas. The results were reported in ‰ relative to VPDB using the NBS18 standard 346
(δ13C = -5.014‰). The analytical reproducibility is within ±0.2‰ (2σ). 347
In-situ carbon isotope analyses of individual graphite grains. We conducted in-situ 348
analyses of carbon isotope values of individual graphite grains with a NanoSIMS 50 349
instrument installed at the Atmosphere and Ocean Research Institute. Prior to the carbon 350
isotope analyses, carbon coating on thin sections was completely removed by 351
re-polishing. The samples were then gold coated and baked at ca. 100 ºC in the 352
NanoSIMS air-lock for a week. A ~2 pA Cs+ primary beam with a beam diameter of less 353
than 0.6 µm was rastered over 5 × 5 µm2 square areas of the graphite gains in the sections. 354
Each analysed area was pre-sputtered with a 200 pA primary beam over the larger raster 355
area (10 × 10 µm) for 240 seconds. We used two magnetic fields with the NanoSIMS 356
multi-collection system: In the magnetic field 1 (B1), secondary ions of 12C– and 12C12C– 357
were detected by electron multiplier EM4 and EM5, with counting time of 3 seconds. 358
Then, the magnetic field was cycled to the second mass (B2). Secondary ions of 12C–, 359
13C–, and 12C14N– were detected simultaneously by EM3, EM4, and EM5, respectively. 360
The counting time for B2 was 20s. Total 25 cycles are required for 1 measurement, 361
corresponding to the total counting time of 75s for B1 and 500s for B2, respectively. The 362
carbon isotopic ratios (13C/12C)EM4 were calculated from 12C– in B1 and 13C– in B2, using 363
the single detector (EM4). The multi-collection (13C/12C)multi ratios, which used the 364
simultaneous collection of 2 detectors (EM3 and EM4 in B2), were also calculated and 365
compared to the (13C/12C)EM4 ratios, to check the stability of the measurement. During the 366
analyses, 12C12C–/12C– and 12C14N–/12C– ratios were monitored to check potential 367
contamination. The carbon isotope compositions were calibrated against an in-house 368
standard material (artificial pure graphite). Its carbon isotope composition was 369
determined as δ13C = -26.6‰, using the conventional method with the IsoPrime 100 370
analytical system. Based on the replicate analyses of the standard, analytical 371
reproducibility of NanoSIMS 50 is within ±3‰ (1σ). 372
Major and trace element analysis of carbonate rocks. We analyzed major element and 373
rare earth element compositions of carbonate rocks to determine their origin. The major 374
element compositions were analyzed with X-ray fluorescence spectrometry (XRF: 375
RIGAKU RIX-2100) at the Tokyo Institute of Technology using fused glass beads. The 376
rare earth elements (REEs) were analyzed with an inductively coupled plasma mass 377
spectrometry (ICP-MS: Agilent 7500s) at Komaba, the University of Tokyo. The 378
analytical methods were described elsewhere23. 379
Mineral compositions and garnet-biotite geothermometry. We analyzed mineral 380
compositions of garnet and biotite in the pelitic rocks and employed garnet-biotite 381
geothermometry24 to estimate metamorphic temperature. Back-scattered electron images 382
of all the thin sections of pelitic rocks and chemical compositions of minerals were 383
obtained with an electron probe microanalyzer (JEOL-JXA-8800) at The University of 384
Tokyo. All analyses were performed with an accelerating voltage of 15 kV, 12 nA beam 385
current and a counting time of 10-40s. The oxide ZAF correction method was applied. 386
The analyses were performed on adjacent grains and across the grain boundaries in order 387
to check compositional zonation for each mineral. 388
Raman microspectroscopy. A Laser Raman micro-spectrometer (JASCO NRS-2000) 389
was used to estimate metamorphic temperature of the graphite. The thin sections were 390
twice exposed to an Ar laser (514.5 nm) for 60 to 80 seconds at a laser power of about 5 391
mW at the sample surface to obtain Raman spectra in the range of 1,800 to 1,100 cm−1 at 1 392
cm−1 resolution. A 100× (NA=0.84) and a 50× (NA=0.80) objective lenses were used, so 393
the spatial resolution of the Raman analysis was about 1-2 µm. We only analyzed the 394
graphite embedded within the rocks below the surface of the thin-section in order to 395
avoid the effect of polishing, which can induce deformation of carbonaceous matter 396
during sample preparation and thus possibly induce artificial modification of the Raman 397
spectroscopic feature25. Details of the analytical methods are described elsewhere12,13. A 398
Raman spectrum of graphite is composed of first-order and second-order regions, and the 399
first-order region from 1,100 to 1,800 cm-1 is often considered26,27. In this region, a 400
graphite band (G band) occurs at ~1,580 cm-1 whereas poorly ordered carbon displays D1 401
(~1,350 cm-1), D2 (~1,620 cm-1) and D3 bands (~1,500 cm-1), respectively. The 402
crystallization temperature can be estimated using following spectral parameters12: 403
T (°C) = -455 ×D1/(G + D1 + D2) band area ratio + 641 404
This equation works in the temperature range of 330 to 650 ºC and the error is estimated 405
±50 ºC12. Peak position, band-area (i.e. integrated area) and band-width (i.e. full width at 406
half maximum, FWHM) were determined using a computer program PeakFit 4.12 407
(SeaSolve Software Inc.). 408
409
23 Koshida, K., Ishikawa, A., Iwamori, H. & Komiya, T. Petrology and geochemistry 410
of mafic rocks in the Acasta Gneiss Complex: Implications for the oldest mafic 411
rocks and their origin. Precambrian Research 283, 190-207 (2016). 412
24 Ferry, J. M. & Spear, F. S. Experimental calibration of the partitioning of Fe and 413
Mg between biotite and garnet. Contributions to Mineralogy and Petrology 66, 414
113-117 (1978). 415
25 Pasteris, J. D. In situ analysis in geological thin-sections by laser raman 416
microprobe spectroscopy: A cautionary note. Applied Spectroscopy 43, 567-570 417
(1989). 418
26 Nemanich, R. J. & Solin, S. A. First- and second-order Raman scattering from 419
finite-size crystals of graphite. Physical Review B 20, 392-401 (1979). 420
27 Tuinstra, F. & Koenig, J. L. Raman spectrum of graphite. The Journal of 421
Chemical Physics 53, 1126-1130 (1970). 422
423
424
425
Author Information is available in the online version of the paper. 426
427
Acknowledgments This research was supported by the Ministry of Education, Culture, 428
Sports, Science and Technology, Japan (grant numbers: 23253007, 26220713) and the 429
Mitsubishi Foundation. We thank Prof. Kenneth D. Collerson and Dr. Bruce Ryan for 430
sharing their geological information. We are grateful to Mr. Wayne Broomfield, Parks 431
Canada, Labrador Inuit Development Corporation (LIDC) and many bear monitors who 432
assisted with our geological fieldwork at the Saglek Block. 433
434
Author contributions T.K. designed the study and Y.S. designed the geochemical study. 435
T.K. T.T., A.I., M.H., M.I., M.K., P.M., N. T., and Y.S. conducted geochemical analyses. 436
T.K. and T.T. collected samples in the field. T.K. wrote the manuscript with important 437
contributions from all co-authors. 438
439
Author Information Reprint and permissions information is available at 440
www.nature.com/reprints. The authors declare no competing financial interests. All of our 441
studied samples are stored at Department of Earth Science and Astronomy, The University 442
of Tokyo, Komaba. Correspondence and requests for the materials and data should be 443
addressed to T.K. (komiya@ea.c.u-tokyo.ac.jp). 444
445
Reviewer Information Nature thanks three anonymous reviewers for their contribution to 446
the peer review of this work. 447
448
A preview of this full-text is provided by Springer Nature.
Content available from Nature
This content is subject to copyright. Terms and conditions apply.