Elements of Eoarchean life trapped in mineral inclusions

Journal name:
Nature
Volume:
548,
Pages:
78–81
Date published:
DOI:
doi:10.1038/nature23261
Received
Accepted
Published online

Metasedimentary rocks from Isua, West Greenland (over 3,700 million years old) contain 13C-depleted carbonaceous compounds, with isotopic ratios that are compatible with a biogenic origin1, 2, 3. Metamorphic garnet crystals in these rocks contain trails of carbonaceous inclusions that are contiguous with carbon-rich sedimentary beds in the host rock, where carbon is fully graphitized. Previous studies4, 5 have not been able to document other elements of life (mainly hydrogen, oxygen, nitrogen and phosphorus) structurally bound to this carbonaceous material. Here we study carbonaceous inclusions armoured within garnet porphyroblasts, by in situ infrared absorption on approximately 10−21 m3 domains within these inclusions. We show that the absorption spectra are consistent with carbon bonded to nitrogen and oxygen, and probably also to phosphate. The levels of C–H or O–H bonds were found to be low. These results are consistent with biogenic organic material isolated for billions of years and thermally matured at temperatures of around 500 °C. They therefore provide spatial characterization for potentially the oldest biogenic carbon relics in Earth’s geological record. The preservation of Eoarchean organic residues within sedimentary material corroborates earlier claims2, 6 for the biogenic origins of carbon in Isua metasediments.

At a glance

Figures

  1. Inclusions in garnet crystals from the Isua metasediments.
    Figure 1: Inclusions in garnet crystals from the Isua metasediments.

    a, Optical transmission picture of a polished thin section showing typical garnet crystals (red arrow) inside the 3,700-million-year-old metasedimentary rock. The lines of black inclusions within the garnet roughly follow the direction of the sedimentary layering. b, Backscattered electron scanning electron microscope (SEM) picture of a freshly fractured garnet surface showing a string of isolated inclusions with carbonaceous material (black patches, red arrow) that are roughly grouped into the banded structure across the garnet, as indicated by the arrow.

  2. Examples of inclusions.
    Figure 2: Examples of inclusions.

    a, An optical image of a dark irregular patch appearing on a freshly exposed garnet surface just after cracking the garnet open. b, The white outline in a that frames the patch marks the area covered by the AFM image mosaic (amplitude-tapping mode; it was not possible to do AFM-IR on this inclusion). c, An optical image of a freshly exposed surface with multiple inclusions. dg, The red box in c outlines the size of the AFM images in dg (deflection-contact mode), showing the structure of four different inclusions from the surface shown in c.

  3. AFM-IR inside an inclusion.
    Figure 3: AFM-IR inside an inclusion.

    a, AFM (deflection-contact mode) image mosaic across an inclusion (framed in blue). b, IR absorption spectra from 900 to 2,235 cm−1 at 10 different spots inside and 1 outside the inclusion (S1–S11). c, A map of the absorption intensity at 2,230 cm−1 (location framed in red in a). d, Magnified region of the absorption map (framed in white in a and c). e, Topography of the region. f, An overlay of the region in d with e, with 50% transparency to illustrate the correlation between 30–40-nm-wide lines of increased IR absorption and some edges in the sample.

  4. Geological context.
    Extended Data Fig. 1: Geological context.

    a, Geological map of the Isua Supracrustal Belt (ISB) in West Greenland. The samples were extracted from the area framed by the black rectangle. b, Bouma sequence alternating pelagic shales and turbiditic psamites. The iron-rich lithologies of this study have been sampled at approximately 4 m along strike from this outcrop from the bottom (left-hand) part of the sequence. c, MgO/FeO (wt%) ratio along a traverse across a garnet porphyroblast. This profile and the MnO profile in d are consistent with a single garnet growth event during prograde metamorphic heating. d, MnO (wt%) along a traverse across a garnet porphyroblast, showing a typical bell curve. The values along the x axis represent sequential automated EMP analytical points separated by 10 μm. Points intersecting inclusions and not representing pure garnet have been replaced by the average of the garnet values on either side. e, Table showing the stable isotopic composition of reduced carbon from inclusion bearing garnets (GNT 1–4) and two bulk rock samples of the garnet–biotite–epidote schist samples (030048 1–4 and 460531 1–4). The total reduced carbon (TOC) values found in the samples are given in relative numbers in column 2. The garnets and therefore all inclusions examined in this study were from the 030048 samples. Data obtained by Service d’Analyse des Roches et des Mineraux, CNRS, CRPG, Université de Lorraine.

  5. Optical images of inclusions.
    Extended Data Fig. 2: Optical images of inclusions.

    a, These inclusions were imaged immediately after the garnets were opened and contain what appears to be liquid (white arrows indicate some examples). b, Inclusions containing more solid material.

  6. Carbon map.
    Extended Data Fig. 3: Carbon map.

    ac, Optical (a), SEM (b) and EDXS (c) intensity maps for carbon for the inclusion shown in Fig. 3. The white frames in b and c indicate the outline of AFM images shown in Fig. 3. EDXS is not surface-sensitive so thin layers of carbonaceous material will be dwarfed by other phases in the EDXS map. Nevertheless, the EDXS map confirms that the dark spots observed in the SEM image are carbon-rich material. In the inclusion studied (Fig. 3), we can confirm that we have carbon-rich material consistent with our AFM-IR results.

  7. AFM-IR results on a new inclusion.
    Extended Data Fig. 4: AFM-IR results on a new inclusion.

    (This is not the inclusion shown in Fig. 3.) a, b, Optical images of the freshly cleaved garnet surface with carbonaceous inclusions. The white frame in a marks the size and position of the image shown in b. The white frame in b indicates the approximate position and size of the AFM-IR deflection mode image (c) and the corresponding IR maps shown in e and f. d, IR absorption spectra from 900 to 2,235 cm−1 at 12 different spots where the positions of S1 to S9 are equally spaced along a line across the inclusion, while S10, S11 and S12 are inside the inclusion at random positions. e, f, A map of the intensity of absorption at 2,230 cm−1 (e) and a map of the intensity of absorption at 1,836 cm−1 (f) across the area shown in c. g, IR absorption spectra from 2,500 to 3,600 cm−1 at a central spot in the inclusion.

  8. Raman spectroscopy on inclusions.
    Extended Data Fig. 5: Raman spectroscopy on inclusions.

    Spectra on inclusion shown in Extended Data Fig. 4 and mapping across a neighbouring inclusion. a, The outline of the map is marked by the black frame. b, The peak position of the G-band across the inclusion. The G-band is usually found at around 1,575 cm−1, but in our experiments, there was a downward offset in the position of approximately 10 cm−1 attributed to laser heating35. The shift in the G-band peak position to higher wavenumbers is usually attributed to an increase in carrier concentration36; here, the higher value for the G-band on the rim suggests that the graphite contains other components, increasing in concentration towards the rim. We observe a 10 cm−1 shift between the centre and the rim of the inclusion, which is consistent with a 5–10 cm−1 upward shift in the G-band observed in graphite that has been functionalized with oxygen and nitrogen37. c, Map of the variation in the ratio between the D-band and G-band intensity. It varies between 0.13 (ratio for the peaks shown above) and 0.24. This corresponds to crystal size around 15 nm on average along the direction of the graphite plane11, and according to the RSCM thermometer method25 this corresponds to a peak metamorphic temperature of up to 600 °C. d, Raman spectroscopy of the inclusion studied by AFM-IR at the position marked by red arrow in the optical picture in a. The three distinct bands characteristic for graphite (G-band), D-band and 2D-band (arrows with values) are visible in the spectrum. A second spectrum was recorded on the garnet surface at the position indicated by the blue arrow next to the inclusions. For this spectrum, no specific bands for graphite were visible. e, Peak position of the G-band along the cross-section marked in the map shown in b by a black bar. f, Cross-section of the D-band and G-band corresponding to the black bar marked in c.

  9. Raman spectroscopy of an unexposed micro-inclusion inside a garnet in a polished thin section.
    Extended Data Fig. 6: Raman spectroscopy of an unexposed micro-inclusion inside a garnet in a polished thin section.

    a, b, A reflected light image (a) and a transmitted light image (b) of a garnet in thin section. The bands of carbonaceous material can be clearly observed as dark bands in the transmitted light image shown in b, but are less obvious in the reflected light image (a). The Raman-equipped microscope was focused a few micrometres below the surface at the red arrow, so that the Raman signal came mainly from the carbonaceous material below the surface. c, The Raman spectra from the encapsulated carbonaceous material (position indicated by the red arrow) is shown as the red plot, and a Raman spectra from the pure garnet at the surface of the thin section (position indicated by the blue arrow) is shown as the blue plot. The three distinct bands characteristic for the G-band, D-band and 2D-band (arrows with values) are visible in the spectrum from the inclusions. The position of the bands along with the ratio between the D-band and G-band match those found for the exposed inclusion shown in Extended Data Fig. 5. The Raman spectra obtained from the uncleaned surface of this polished thin section display organic contamination with the broad signal from 1,200 cm−1 to 1,600 cm−1 and again from 2,800 cm−1 to 3,000 cm−1 (C–H).

  10. IR absorption spectra from 2,500 to 3,600 cm−1.
    Extended Data Fig. 7: IR absorption spectra from 2,500 to 3,600 cm−1.

    IR absorption spectra (in the range in which hydrogen-bond absorption would occur) from various spots in the inclusion shown in Fig. 3. a, The deflection mode AFM image (as in Fig. 3). b, AFM-IR absorption in the range from 2,500 to 3,600 cm−1 at four points inside the inclusion indicated in a with boxes of corresponding colour (red, green, purple and blue). C–H stretches have a strong adsorption between 2,850 and 3,100 cm−1, depending on whether the C–H bond is aliphatic (lower wavenumbers) or aromatic (higher wavenumbers) hydrocarbon. No such bands could be detected in our inclusions. Similarly, no evidence for O–H was observed. The pulsed laser power used here was the same as for the lower range 900–2,230 cm−1.

  11. AFM-IR on kerogen on a grain of sand from a core plug extracted from an oil reservoir.
    Extended Data Fig. 8: AFM-IR on kerogen on a grain of sand from a core plug extracted from an oil reservoir.

    a, An AFM topography image on part of the sand grain surface. In the right part of the image, a 1 × 1 μm2 square part of the surface was scraped clean of kerogen by applying high force (GPa) with the AFM tip while scanning the area. This causes a build-up of kerogen around the cleaned area. A height profile (the blue line in a) across the surface and into the scraped area is shown in b). Using the same settings for the AFM-IR and the same type of tip as used on the inclusions we recorded three spectra shown in c from 2,500 to 3,600 cm−1. The positions at which we recorded these spectra are marked with blue, red and green boxes in a, with the corresponding arrows in b. The position marked by green was inside the scraped area. The kerogen is only a few nanometres thick and the C–H absorption is therefore very low, but barely detectable. At the blue and red positions the kerogen is 20–40 nm thick and the absorption for the C–H and O–H stretch can be clearly detected. The laser power in the AFM-IR was the same as used for the inclusions.

  12. Loss of carbonaceous material in the SEM.
    Extended Data Fig. 9: Loss of carbonaceous material in the SEM.

    a, The AFM image (deflection mode) inside the inclusion a few hours after the inclusion was opened. b, The same area after analysis in the SEM. c, The height profile along the blue and red lines shown in a and b across the internal part of the inclusion. The difference between the blue and red lines shows up to 500 nm thick material was lost because of beam damage during the analysis in the SEM.

  13. Modelling the formation of nitrile and the loss of hydrogen.
    Extended Data Fig. 10: Modelling the formation of nitrile and the loss of hydrogen.

    ac, Molecular modelling of the free energy for transforming the dipeptide dileucine (a) into an aromatic amine (blue) (b) and an aromatic nitrile (red) (c). d, The modelling predicts that the free energy for the reactions favour formation of aromatic groups and loss of hydrogen and water above approximately 250 °C. On the basis of the calculated reaction free energy, the nitrile is favoured above 700 °C, while the amine is favoured between 250 and 700 °C. If hydrogen is removed from the inclusion via diffusion into the surrounding garnet, the shift in equilibrium can still lead to nitrile formation instead of amines, even at temperatures below 700 °C, as discussed in the main text and Supplementary Information.

References

  1. Rosing, M. T. 13C-Depleted carbon microparticles in >3700-Ma sea-floor sedimentary rocks from west greenland. Science 283, 674676 (1999)
  2. Ohtomo, Y., Kakegawa, T., Ishida, A., Nagase, T. & Rosing, M. T. Evidence for biogenic graphite in early Archaean Isua metasedimentary rocks. Nat. Geosci. 7, 2528 (2014)
  3. Bell, E. A., Boehnke, P., Harrison, T. M. & Mao, W. L. Potentially biogenic carbon preserved in a 4.1 billion-year-old zircon. Proc. Natl Acad. Sci. USA 112, 1451814521 (2015)
  4. Pinti, D. L., Hashizume, K. & Matsuda, J. Nitrogen and argon signatures in 3.8 to 2.8 Ga metasediments: clues on the chemical state of the Archean ocean and the deep biosphere. Geochim. Cosmochim. Acta 65, 23012315 (2001)
  5. van Zuilen, M. A. et al. Nitrogen and argon isotopic signatures in graphite from the 3.8-Ga-old Isua Supracrustal Belt, Southern West Greenland. Geochim. Cosmochim. Acta 69, 12411252 (2005)
  6. Nutman, A. P., Bennett, V. C., Friend, C. R. L., Van Kranendonk, M. J. & Chivas, A. R. Rapid emergence of life shown by discovery of 3,700-million-year-old microbial structures. Nature 537, 535538 (2016)
  7. Boak, J. L. & Dymek, R. F. Metamorphism of the ca-3800 Ma supracrustal rocks at Isua, West Greenland—implications for early Archean crustal evolution. Earth Planet. Sci. Lett. 59, 155176 (1982)
  8. Nutman, A. P., Friend, C. R. L., Bennett, V. C. & McGregor, V. R. Dating of the Ameralik dyke swarms of the Nuuk district, southern West Greenland: mafic intrusion events starting from c. 3510 Ma. J. Geol. Soc. Lond. 161, 421430 (2004)
  9. Dazzi, A. et al. AFM-IR: combining atomic force microscopy and infrared spectroscopy for nanoscale chemical characterization. Appl. Spectrosc. 66, 13651384 (2012)
  10. Stuart, B. Infrared Spectroscopy: Fundamentals and Applications (Wiley & Sons, 2004)
  11. Tuinstra, F. & Koenig, J. L. Raman spectrum of graphite. J. Chem. Phys. 53, 1126 (1970)
  12. Hofmeister, A. M. & Chopelas, A. Vibrational spectroscopy of end-member silicate garnets. Phys. Chem. Miner. 17, 503526 (1991)
  13. Hofmeister, A. M., Fagan, T. J., Campbell, K. M. & Schaal, R. B. Single-crystal IR spectroscopy of pyrope-almandine garnets with minor amounts of Mn and Ca. Am. Mineral. 81, 418428 (1996)
  14. Farmer, V. C. The Infrared Spectra of Minerals Vol. 4 (Mineralogical Society, 1974)
  15. Ruhela, R. et al. N,N,N-Trimethyl benzyl ammonium bis-(2-ethylhexyl)-phosphonate grafted polymer—a solid supported ionic liquid for the separation of uranium from aqueous processing streams. Green Chem. 17, 827830 (2015)
  16. Slawin, A. M. Z., Woollins, J. D. & Zhang, Q. Z. Novel chiral phosphine ligands and complexes from amino acid esters. J. Chem. Soc., Dalton Trans. 5, 621632 (2001)
  17. Coates, J. Interpretation of Infrared Spectra, A Practical Approach (John Wiley & Sons, 2006)
  18. Kitson, R. E. & Griffith, N. E. Infrared absorption band due to nitrile stretching vibration. Anal. Chem. 24, 334337 (1952)
  19. Anderson, L. A. On the hydrogen and oxygen-content of marine-phytoplankton. Deep Sea Res. Part I Oceanogr. Res. Pap. 42, 16751680 (1995)
  20. Beyssac, O., Rouzaud, J. N., Goffe, B., Brunet, F. & Chopin, C. Graphitization in a high-pressure, low-temperature metamorphic gradient: a Raman microspectroscopy and HRTEM study. Contrib. Mineral. Petrol. 143, 1931 (2002)
  21. Landis, C. A. Graphitization of dispersed carbonaceous material in metamorphic rocks. Contrib. Mineral. Petr. 30, 3445 (1971)
  22. De Gregorio, B. T., Sharp, T. G., Flynn, G. J., Wirick, S. & Hervig, R. L. Biogenic origin for Earth’s oldest putative microfossils. Geology 37, 631634 (2009)
  23. Ferry, J. M. & Spear, F. S. Experimental calibration of partitioning of Fe and Mg between biotite and garnet. Contrib. Mineral. Petrol. 66, 113117 (1978)
  24. Spear, F. S. On the interpretation of peak metamorphic temperatures in light of garnet diffusion during cooling. J. Metamorph. Geol. 9, 379388 (1991)
  25. Beyssac, O., Goffe, B., Chopin, C. & Rouzaud, J. N. Raman spectra of carbonaceous material in metasediments: a new geothermometer. J. Metamorph. Geol. 20, 859871 (2002)
  26. Klamt, A., Eckert, F. & Arlt, W. COSMO-RS: an alternative to simulation for calculating thermodynamic properties of liquid mixtures. Annu. Rev. Chem. Biomol. Eng. 1, 101122 (2010)
  27. Wang, L. P., Zhang, Y. X. & Essene, E. J. Diffusion of the hydrous component in pyrope. Am. Mineral. 81, 706718 (1996)
  28. Bauschlicher, C. W., Peeters, E. & Allamandola, L. J. The infrared spectra of very large, compact, highly symmetric, polycyclic aromatic hydrocarbons (PAHs). Astrophys. J. 678, 316327 (2008)
  29. Bauschlicher, C. W. et al. The Nasa Ames polycyclic aromatic hydrocarbon infrared spectroscopic database: the computed spectra. Astrophys. J. Suppl. Ser. 189, 341351 (2010)
  30. Ahlrichs, R., Bar, M., Haser, M., Horn, H. & Kolmel, C. Electronic-structure calculations on workstation computers — the program system TURBOMOLE. Chem. Phys. Lett. 162, 165169 (1989)
  31. Becke, A. D. Density-functional exchange-energy approximation with correct asymptotic behavior. Phys. Rev. A 38, 30983100 (1988)
  32. Perdew, J. P. Density-functional approximation for the correlation energy of the inhomogeneous electron gas. Phys. Rev. B Condens. Matter 33, 88228824 (1986)
  33. Schäfer, A., Horn, H. & Ahlrichs, R. Fully optimized contracted Gaussian-basis sets for atoms Li to Kr. J. Chem. Phys. 97, 25712577 (1992)
  34. Klamt, A. & Schüürmann, G. COSMO: a new approach to dielectric screening in solvents with explicit expressions for the screening energy and its gradient. J. Chem. Soc. Perkin Trans. 2 5, 799805 (1993)
  35. Kagi, H. et al. Proper understanding of down-shifted Raman-spectra of natural graphite — direct estimation of laser-induced rise in sample temperature. Geochim. Cosmochim. Acta 58, 35273530 (1994)
  36. Ferrari, A. C. Raman spectroscopy of graphene and graphite: disorder, electron-phonon coupling, doping and nonadiabatic effects. Solid State Commun. 143, 4757 (2007)
  37. Khai, T. V. et al. Influence of N-doping on the structural and photoluminescence properties of graphene oxide films. Carbon 50, 37993806 (2012)

Download references

Author information

Affiliations

  1. Nano-Science Center, Department of Chemistry, University of Copenhagen, Copenhagen 2100, Denmark

    • T. Hassenkam,
    • M. P. Andersson &
    • K. N. Dalby
  2. Center for Nanostructured Graphene (CNG), Department of Micro- and Nanotechnology, Technical University of Denmark, Kongens Lyngby 2800, Denmark

    • D. M. A. Mackenzie
  3. Natural History Museum of Denmark, University of Copenhagen, Copenhagen 1350, Denmark

    • M. T. Rosing

Contributions

T.H. and M.T.R. conceived the idea for this study. M.T.R. collected, dated and isolated the individual garnet crystals and discovered the inclusions. M.T.R. made the optical transmission images of the garnets in their bedding. T.H. developed the method for accessing the inclusions with the AFM-IR. T.H. did the AFM-IR measurements and the assignment of IR absorption peaks to chemical bonds. The interpretation of the data where done mostly by M.T.R. and T.H. M.P.A. and K.N.D. confirmed the assignment of the IR absorption peaks. K.N.D. made and interpreted the SEM pictures and the EDXS maps. M.P.A. performed the modelling. D.M.A.M. did the Raman mapping. T.H. and D.M.A.M. analysed the result. T.H. wrote the paper with input from all authors. All authors contributed to scientific discussions.

Competing financial interests

The authors declare no competing financial interests.

Corresponding author

Correspondence to:

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Author details

Extended data figures and tables

Extended Data Figures

  1. Extended Data Figure 1: Geological context. (218 KB)

    a, Geological map of the Isua Supracrustal Belt (ISB) in West Greenland. The samples were extracted from the area framed by the black rectangle. b, Bouma sequence alternating pelagic shales and turbiditic psamites. The iron-rich lithologies of this study have been sampled at approximately 4 m along strike from this outcrop from the bottom (left-hand) part of the sequence. c, MgO/FeO (wt%) ratio along a traverse across a garnet porphyroblast. This profile and the MnO profile in d are consistent with a single garnet growth event during prograde metamorphic heating. d, MnO (wt%) along a traverse across a garnet porphyroblast, showing a typical bell curve. The values along the x axis represent sequential automated EMP analytical points separated by 10 μm. Points intersecting inclusions and not representing pure garnet have been replaced by the average of the garnet values on either side. e, Table showing the stable isotopic composition of reduced carbon from inclusion bearing garnets (GNT 1–4) and two bulk rock samples of the garnet–biotite–epidote schist samples (030048 1–4 and 460531 1–4). The total reduced carbon (TOC) values found in the samples are given in relative numbers in column 2. The garnets and therefore all inclusions examined in this study were from the 030048 samples. Data obtained by Service d’Analyse des Roches et des Mineraux, CNRS, CRPG, Université de Lorraine.

  2. Extended Data Figure 2: Optical images of inclusions. (213 KB)

    a, These inclusions were imaged immediately after the garnets were opened and contain what appears to be liquid (white arrows indicate some examples). b, Inclusions containing more solid material.

  3. Extended Data Figure 3: Carbon map. (191 KB)

    ac, Optical (a), SEM (b) and EDXS (c) intensity maps for carbon for the inclusion shown in Fig. 3. The white frames in b and c indicate the outline of AFM images shown in Fig. 3. EDXS is not surface-sensitive so thin layers of carbonaceous material will be dwarfed by other phases in the EDXS map. Nevertheless, the EDXS map confirms that the dark spots observed in the SEM image are carbon-rich material. In the inclusion studied (Fig. 3), we can confirm that we have carbon-rich material consistent with our AFM-IR results.

  4. Extended Data Figure 4: AFM-IR results on a new inclusion. (513 KB)

    (This is not the inclusion shown in Fig. 3.) a, b, Optical images of the freshly cleaved garnet surface with carbonaceous inclusions. The white frame in a marks the size and position of the image shown in b. The white frame in b indicates the approximate position and size of the AFM-IR deflection mode image (c) and the corresponding IR maps shown in e and f. d, IR absorption spectra from 900 to 2,235 cm−1 at 12 different spots where the positions of S1 to S9 are equally spaced along a line across the inclusion, while S10, S11 and S12 are inside the inclusion at random positions. e, f, A map of the intensity of absorption at 2,230 cm−1 (e) and a map of the intensity of absorption at 1,836 cm−1 (f) across the area shown in c. g, IR absorption spectra from 2,500 to 3,600 cm−1 at a central spot in the inclusion.

  5. Extended Data Figure 5: Raman spectroscopy on inclusions. (161 KB)

    Spectra on inclusion shown in Extended Data Fig. 4 and mapping across a neighbouring inclusion. a, The outline of the map is marked by the black frame. b, The peak position of the G-band across the inclusion. The G-band is usually found at around 1,575 cm−1, but in our experiments, there was a downward offset in the position of approximately 10 cm−1 attributed to laser heating35. The shift in the G-band peak position to higher wavenumbers is usually attributed to an increase in carrier concentration36; here, the higher value for the G-band on the rim suggests that the graphite contains other components, increasing in concentration towards the rim. We observe a 10 cm−1 shift between the centre and the rim of the inclusion, which is consistent with a 5–10 cm−1 upward shift in the G-band observed in graphite that has been functionalized with oxygen and nitrogen37. c, Map of the variation in the ratio between the D-band and G-band intensity. It varies between 0.13 (ratio for the peaks shown above) and 0.24. This corresponds to crystal size around 15 nm on average along the direction of the graphite plane11, and according to the RSCM thermometer method25 this corresponds to a peak metamorphic temperature of up to 600 °C. d, Raman spectroscopy of the inclusion studied by AFM-IR at the position marked by red arrow in the optical picture in a. The three distinct bands characteristic for graphite (G-band), D-band and 2D-band (arrows with values) are visible in the spectrum. A second spectrum was recorded on the garnet surface at the position indicated by the blue arrow next to the inclusions. For this spectrum, no specific bands for graphite were visible. e, Peak position of the G-band along the cross-section marked in the map shown in b by a black bar. f, Cross-section of the D-band and G-band corresponding to the black bar marked in c.

  6. Extended Data Figure 6: Raman spectroscopy of an unexposed micro-inclusion inside a garnet in a polished thin section. (436 KB)

    a, b, A reflected light image (a) and a transmitted light image (b) of a garnet in thin section. The bands of carbonaceous material can be clearly observed as dark bands in the transmitted light image shown in b, but are less obvious in the reflected light image (a). The Raman-equipped microscope was focused a few micrometres below the surface at the red arrow, so that the Raman signal came mainly from the carbonaceous material below the surface. c, The Raman spectra from the encapsulated carbonaceous material (position indicated by the red arrow) is shown as the red plot, and a Raman spectra from the pure garnet at the surface of the thin section (position indicated by the blue arrow) is shown as the blue plot. The three distinct bands characteristic for the G-band, D-band and 2D-band (arrows with values) are visible in the spectrum from the inclusions. The position of the bands along with the ratio between the D-band and G-band match those found for the exposed inclusion shown in Extended Data Fig. 5. The Raman spectra obtained from the uncleaned surface of this polished thin section display organic contamination with the broad signal from 1,200 cm−1 to 1,600 cm−1 and again from 2,800 cm−1 to 3,000 cm−1 (C–H).

  7. Extended Data Figure 7: IR absorption spectra from 2,500 to 3,600 cm−1. (315 KB)

    IR absorption spectra (in the range in which hydrogen-bond absorption would occur) from various spots in the inclusion shown in Fig. 3. a, The deflection mode AFM image (as in Fig. 3). b, AFM-IR absorption in the range from 2,500 to 3,600 cm−1 at four points inside the inclusion indicated in a with boxes of corresponding colour (red, green, purple and blue). C–H stretches have a strong adsorption between 2,850 and 3,100 cm−1, depending on whether the C–H bond is aliphatic (lower wavenumbers) or aromatic (higher wavenumbers) hydrocarbon. No such bands could be detected in our inclusions. Similarly, no evidence for O–H was observed. The pulsed laser power used here was the same as for the lower range 900–2,230 cm−1.

  8. Extended Data Figure 8: AFM-IR on kerogen on a grain of sand from a core plug extracted from an oil reservoir. (360 KB)

    a, An AFM topography image on part of the sand grain surface. In the right part of the image, a 1 × 1 μm2 square part of the surface was scraped clean of kerogen by applying high force (GPa) with the AFM tip while scanning the area. This causes a build-up of kerogen around the cleaned area. A height profile (the blue line in a) across the surface and into the scraped area is shown in b). Using the same settings for the AFM-IR and the same type of tip as used on the inclusions we recorded three spectra shown in c from 2,500 to 3,600 cm−1. The positions at which we recorded these spectra are marked with blue, red and green boxes in a, with the corresponding arrows in b. The position marked by green was inside the scraped area. The kerogen is only a few nanometres thick and the C–H absorption is therefore very low, but barely detectable. At the blue and red positions the kerogen is 20–40 nm thick and the absorption for the C–H and O–H stretch can be clearly detected. The laser power in the AFM-IR was the same as used for the inclusions.

  9. Extended Data Figure 9: Loss of carbonaceous material in the SEM. (390 KB)

    a, The AFM image (deflection mode) inside the inclusion a few hours after the inclusion was opened. b, The same area after analysis in the SEM. c, The height profile along the blue and red lines shown in a and b across the internal part of the inclusion. The difference between the blue and red lines shows up to 500 nm thick material was lost because of beam damage during the analysis in the SEM.

  10. Extended Data Figure 10: Modelling the formation of nitrile and the loss of hydrogen. (95 KB)

    ac, Molecular modelling of the free energy for transforming the dipeptide dileucine (a) into an aromatic amine (blue) (b) and an aromatic nitrile (red) (c). d, The modelling predicts that the free energy for the reactions favour formation of aromatic groups and loss of hydrogen and water above approximately 250 °C. On the basis of the calculated reaction free energy, the nitrile is favoured above 700 °C, while the amine is favoured between 250 and 700 °C. If hydrogen is removed from the inclusion via diffusion into the surrounding garnet, the shift in equilibrium can still lead to nitrile formation instead of amines, even at temperatures below 700 °C, as discussed in the main text and Supplementary Information.

Supplementary information

PDF files

  1. Supplementary Information (236 KB)

    This file contains information on geological context broken down into protolith, age, metamorphism, and deformation and also includes an analysis of the molecular modelling. The file also contains supplementary table 3 which shows the IR absorption assignment.

Excel files

  1. Supplementary Table (33 KB)

    This file contains supplementary table 1 which shows the major elements in the metasedimentary rock. The inclusions studied are from 30048.

  2. Supplementary Table (48 KB)

    This file contains supplementary table 2 which shows composition data in biotite and garnets at various positions close to the inclusions studied. Also shown is the estimated temperature using a calibration by Ferry and Spear.

Additional data