Formation of continental crust has shaped the surface and interior of our planet and generated the land and mineral resources on which we rely. However, how the early continental crust of Earth formed is still debated1,2,3,4,5,6,7. Modern continental crust is largely formed from wet and oxidizing arc magmas at subduction zones, in which oceanic lithosphere and water recycle into the mantle8,9,10. The magmatic H2O content and redox state of ancient rocks that constitute the early continental crust, however, are difficult to quantify owing to ubiquitous metamorphism. Here we combine two zircon oxybarometers11,12 to simultaneously determine magmatic oxygen fugacity (fO2) and H2O content of Archaean (4.0–2.5 billion years ago) granitoids that dominate the early continental crust. We show that most Archaean granitoid magmas were ≥1 log unit more oxidizing than Archaean ambient mantle-derived magmas13,14 and had high magmatic H2O contents (6–10 wt%) and high H2O/Ce ratios (>1,000), similar to modern arc magmas. We find that magmatic fO2, H2O contents and H2O/Ce ratios of Archaean granitoids positively correlate with depth of magma formation, requiring transport of large amounts of H2O to the lower crust and mantle. These observations can be readily explained by subduction but are difficult to reconcile with non-subduction models of crustal formation3,4,5,6,7. We note an increase in magmatic fO2 and H2O content between 4.0 and 3.6 billion years ago, probably indicating the onset of subduction during this period.
This is a preview of subscription content, access via your institution
Access Nature and 54 other Nature Portfolio journals
Get Nature+, our best-value online-access subscription
$29.99 / 30 days
cancel any time
Subscribe to this journal
Receive 51 print issues and online access
$199.00 per year
only $3.90 per issue
Rent or buy this article
Prices vary by article type
Prices may be subject to local taxes which are calculated during checkout
New and compiled zircon trace-elemental data are presented in Supplementary Table 1. Compiled whole-rock elemental data are presented in Supplementary Table 2. Calculated median zircon and whole-rock/melt compositions, as well as the calculated magmatic fO2 values and H2O contents, are presented, along with their estimated uncertainties, in Supplementary Table 3. The results of thermodynamic–geochemical modelling are presented in Supplementary Table 4. All the supplementary data and the Excel spreadsheet used to calculate magmatic H2O contents are available to download from the EarthChem library (https://doi.org/10.26022/IEDA/112745). Source data are provided with this paper.
Martin, H. Effect of steeper Archean geothermal gradient on geochemistry of subduction-zone magmas. Geology 14, 753–756 (1986).
Rapp, R. P., Shimizu, N. & Norman, M. D. Growth of early continental crust by partial melting of eclogite. Nature 425, 605–609 (2003).
Moore, W. B. & Webb, A. A. G. Heat-pipe Earth. Nature 501, 501–505 (2013).
Rozel, A. B., Golabek, G. J., Jain, C., Tackley, P. J. & Gerya, T. Continental crust formation on early Earth controlled by intrusive magmatism. Nature 545, 332–335 (2017).
Johnson, T. E., Brown, M., Gardiner, N. J., Kirkland, C. L. & Smithies, R. H. Earth’s first stable continents did not form by subduction. Nature 543, 239–242 (2017).
Smithies, R. H. et al. Oxygen isotopes trace the origins of Earth’s earliest continental crust. Nature 592, 70–75 (2021).
Johnson, T. E. et al. Giant impacts and the origin and evolution of continents. Nature 608, 330–335 (2022).
Kelley, K. A. & Cottrell, E. Water and the oxidation state of subduction zone magmas. Science 325, 605–607 (2009).
Zimmer, M. M. et al. The role of water in generating the calc-alkaline trend: new volatile data for Aleutian magmas and a new tholeiitic index. J. Petrol. 51, 2411–2444 (2010).
Wallace, P. J., Plank, T., Edmonds, M. & Hauri, E. H. in The Encyclopedia of Volcanoes 2nd edn (eds Sigurdsson, H. et al.) 163–183 (Elsevier, 2015).
Smythe, D. J. & Brenan, J. M. Magmatic oxygen fugacity estimated using zircon-melt partitioning of cerium. Earth Planet. Sci. Lett. 453, 260–266 (2016).
Loucks, R. R., Fiorentini, M. L. & Henríquez, G. J. New magmatic oxybarometer using trace elements in zircon. J. Petrol. 61, egaa34 (2020).
Aulbach, S. & Stagno, V. Evidence for a reducing Archean ambient mantle and its effects on the carbon cycle. Geology 44, 751–754 (2016).
Gao, L. et al. Oxidation of Archean upper mantle caused by crustal recycling. Nat. Commun. 13, 3283 (2022).
Moyen, J. The composite Archaean grey gneisses: petrological significance, and evidence for a non-unique tectonic setting for Archaean crustal growth. Lithos 123, 21–36 (2011).
Hoffmann, J. E., Zhang, C. & Nagel, T. in Earth’s Oldest Rocks 2nd edn (eds Van Kranendonk, M. J. et al.) 133–168 (Elsevier, 2015).
Waters, L. E. & Lange, R. A. An updated calibration of the plagioclase-liquid hygrometer-thermometer applicable to basalts through rhyolites. Am. Mineral. 100, 2172–2184 (2015).
Sobolev, A. V. et al. Komatiites reveal a hydrous Archaean deep-mantle reservoir. Nature 531, 628–632 (2016).
Sobolev, A. V. et al. Deep hydrous mantle reservoir provides evidence for crustal recycling before 3.3 billion years ago. Nature 571, 555–559 (2019).
Smythe, D. J. & Brenan, J. M. Cerium oxidation state in silicate melts: combined fO2, temperature and compositional effects. Geochim. Cosmochim. Acta 170, 173–187 (2015).
Meng, X. et al. Variable modes of formation for tonalite–trondhjemite–granodiorite–diorite (TTG)-related porphyry-type Cu ± Au deposits in the Neoarchean southern Abitibi subprovince (Canada): evidence from petrochronology and oxybarometry. J. Petrol. 62, egab079 (2021).
Lu, Y. et al. Zircon Fingerprinting of Magmatic-Hydrothermal Systems in the Archean Yilgarn Craton (Geological Survey of Western Australia Report 197, 2019).
Lu, Y.-J. et al. in Tectonics and Metallogeny of the Tethyan Orogenic Belt Vol. 19 (ed. Richards, J. P.) 329–347 (Society of Economic Geologists, 2016).
Lu, Y., Loucks, R. R., Fiorentini, M. L., Yang, Z. & Hou, Z. Fluid flux melting generated postcollisional high Sr/Y copper ore–forming water-rich magmas in Tibet. Geology 43, 583–586 (2015).
Bell, E. A., Boehnke, P. & Harrison, T. M. Recovering the primary geochemistry of Jack Hills zircons through quantitative estimates of chemical alteration. Geochim. Cosmochim. Acta 191, 187–202 (2016).
Bouvier, A. et al. Li isotopes and trace elements as a petrogenetic tracer in zircon: insights from Archean TTGs and sanukitoids. Contrib. Mineral. Petrol. 163, 745–768 (2012).
Laurent, O., Moyen, J., Wotzlaw, J., Björnsen, J. & Bachmann, O. Early Earth zircons formed in residual granitic melts produced by tonalite differentiation. Geology 50, 437–441 (2021).
Reimink, J. R., Davies, J. H. F. L., Bauer, A. M. & Chacko, T. A. A comparison between zircons from the Acasta Gneiss Complex and the Jack Hills region. Earth Planet. Sci. Lett. 531, 115975 (2020).
Reimink, J. R., Chacko, T., Stern, R. A. & Heaman, L. M. The birth of a cratonic nucleus: lithogeochemical evolution of the 4.02–2.94 Ga Acasta Gneiss Complex. Precambrian Res. 281, 453–472 (2016).
Ge, R., Wilde, S. A., Zhu, W., Zhou, T. & Si, Y. Formation and evolution of Archean continental crust: a thermodynamic–geochemical perspective of granitoids from the Tarim Craton, NW China. Earth-Sci. Rev. 234, 104219 (2022).
Ge, R., Zhu, W., Wilde, S. A. & Wu, H. Remnants of Eoarchean continental crust derived from a subducted proto-arc. Sci. Adv. 4, aao3159 (2018).
Rasmussen, D. J., Plank, T. A., Roman, D. C. & Zimmer, M. M. Magmatic water content controls the pre-eruptive depth of arc magmas. Science. 375, 1169–1172 (2022).
Tang, M., Erdman, M., Eldridge, G. & Lee, C. A. The redox “filter” beneath magmatic orogens and the formation of continental crust. Sci. Adv. 4, eaar4444 (2018).
Collins, W. J., Murphy, J. B., Johnson, T. E. & Huang, H. Critical role of water in the formation of continental crust. Nat. Geosci. 13, 331–338 (2020).
Roman, A. & Arndt, N. Differentiated Archean oceanic crust: its thermal structure, mechanical stability and a test of the sagduction hypothesis. Geochim. Cosmochim. Acta 278, 65–77 (2020).
Aarons, S. M. et al. Titanium isotopes constrain a magmatic transition at the Hadean-Archean boundary in the Acasta Gneiss Complex. Sci. Adv. 6, eabc9959 (2020).
Zhang, Q. et al. No evidence of supracrustal recycling in Si-O isotopes of Earth’s oldest rocks 4 Ga ago. Sci. Adv. 9, eadf693 (2023).
Bauer, A. M. et al. Hafnium isotopes in zircons document the gradual onset of mobile-lid tectonics. Geochem. Perspect. Lett. 14, 1–6 (2020).
Windley, B. F., Kusky, T. & Polat, A. Onset of plate tectonics by the Eoarchean. Precambrian Res. 352, 105980 (2021).
Brenner, A. R., Fu, R. R., Kylander-Clark, A. R., Hudak, G. J. & Foley, B. J. Plate motion and a dipolar geomagnetic field at 3.25 Ga. Proc. Natl Acad. Sci. USA 119, e2084709177 (2022).
Calvert, A. J., Doublier, M. P. & Sellars, S. E. Seismic reflections from a lithospheric suture zone below the Archaean Yilgarn Craton. Nat. Commun. 12, 7245 (2021).
Campbell, I. H. & Taylor, S. R. No water, no granites - no oceans, no continents. Geophys. Res. Lett. 10, 1061–1064 (1983).
van Hunen, J. & Moyen, J. Archean subduction: fact or fiction? Annu. Rev. Earth Planet. Sci. 40, 195–219 (2012).
Wallace, P. J., Anderson, A. T. Jr & Davis, A. M. Gradients in H2O, CO2, and exsolved gas in a large-volume silicic magma system: interpreting the record preserved in melt inclusions from the Bishop Tuff. J. Geophys. Res. Solid Earth 104, 20097–20122 (1999).
Gudelius, D. et al. Zircon melt inclusions in mafic and felsic rocks of the Bushveld Complex–constraints for zircon crystallization temperatures and partition coefficients. Geochim. Cosmochim. Acta 289, 158–181 (2020).
Watts, K. E. & Mercer, C. N. Zircon-hosted melt inclusion record of silicic magmatism in the Mesoproterozoic St. Francois Mountains terrane, Missouri: origin of the Pea Ridge iron oxide-apatite-rare earth element deposit and implications for regional crustal pathways of mineralization. Geochim. Cosmochim. Acta 272, 54–77 (2020).
Laurent, O. et al. Earth’s earliest granitoids are crystal-rich magma reservoirs tapped by silicic eruptions. Nat. Geosci. 13, 163–169 (2020).
Bucholz, C. E., Stolper, E. M., Eiler, J. M. & Breaks, F. W. A comparison of oxygen fugacities of strongly peraluminous granites across the Archean–Proterozoic boundary. J. Petrol. 59, 2123–2156 (2018).
Kemp, A. I. S. et al. Hadean crustal evolution revisited: new constraints from Pb–Hf isotope systematics of the Jack Hills zircons. Earth Planet. Sci. Lett. 296, 45–56 (2010).
Blundy, J. & Wood, B. Prediction of crystal–melt partition coefficients from elastic moduli. Nature 372, 452–454 (1994).
Ferry, J. & Watson, E. New thermodynamic models and revised calibrations for the Ti-in-zircon and Zr-in-rutile thermometers. Contrib. Mineral. Petrol. 154, 429–437 (2007).
Chesner, C. A. & Luhr, J. F. A melt inclusion study of the Toba Tuffs, Sumatra, Indonesia. J. Volcanol. Geotherm. Res. 197, 259–278 (2010).
Ge, R. et al. Generation of Eoarchean continental crust from altered mafic rocks derived from a chondritic mantle: the ∼3.72 Ga Aktash gneisses, Tarim Craton (NW China). Earth Planet. Sci. Lett. 538, 116225 (2020).
Kielman, R., Whitehouse, M., Nemchin, A. & Kemp, A. A tonalitic analogue to ancient detrital zircon. Chem. Geol. 499, 43–57 (2018).
Mole, D. R. et al. The formation of Neoarchean continental crust in the south-east Superior Craton by two distinct geodynamic processes. Precambrian Res. 356, 106104 (2021).
Zong, K. Q. et al. The generation and evolution of Archean continental crust in the Dunhuang block, northeastern Tarim Craton, northwestern China. Precambrian Res. 235, 251–263 (2013).
Xiang, H. & Connolly, J. A. D. GeoPS: an interactive visual computing tool for thermodynamic modelling of phase equilibria. J. Metamorph. Geol. 40, 243–255 (2022).
Holland, T. J. B. & Powell, R. An improved and extended internally consistent thermodynamic dataset for phases of petrological interest, involving a new equation of state for solids. J. Metamorph. Geol. 29, 333–383 (2011).
Holland, T. J. B., Green, E. C. R. & Powell, R. Melting of peridotites through to granites: a simple thermodynamic model in the system KNCFMASHTOCr. J. Petrol. 59, 881–900 (2018).
Green, E. et al. Activity–composition relations for the calculation of partial melting equilibria in metabasic rocks. J. Metamorph. Geol. 34, 845–869 (2016).
Holland, T. J. B., Green, E. C. R. & Powell, R. A thermodynamic model for feldspars in KAlSi3O8–NaAlSi3O8–CaAl2Si2O8 for mineral equilibrium calculations. J. Metamorph. Geol. 40, 587–600 (2022).
White, R. W., Powell, R. & Johnson, T. E. The effect of Mn on mineral stability in metapelites revisited: new a–x relations for manganese-bearing minerals. J. Metamorph. Geol. 32, 809–828 (2014).
Kress, V. C. & Carmichael, I. S. The compressibility of silicate liquids containing Fe2O3 and the effect of composition, temperature, oxygen fugacity and pressure on their redox states. Contrib. Mineral. Petrol. 108, 82–92 (1991).
Bédard, J. H. A catalytic delamination-driven model for coupled genesis of Archaean crust and sub-continental lithospheric mantle. Geochim. Cosmochim. Acta 70, 1188–1214 (2006).
Watson, E. B. & Harrison, T. M. Zircon saturation revisited: temperature and composition effects in a variety of crustal magma types. Earth Planet. Sci. Lett. 64, 295–304 (1983).
Furnes, H., Dilek, Y. & De Wit, M. Precambrian greenstone sequences represent different ophiolite types. Gondwana Res. 27, 649–685 (2015).
Bea, F. & Montero, P. Behavior of accessory phases and redistribution of Zr, REE, Y, Th, and U during metamorphism and partial melting of metapelites in the lower crust: an example from the Kinzigite Formation of Ivrea-Verbano, NW Italy. Geochim. Cosmochim. Acta 63, 1133–1153 (1999).
This work was supported by grants from the National Natural Science Foundation of China (41872191, 41922017 and 42025202).
The authors declare no competing interests.
Peer review information
Nature thanks Elizabeth Bell and Marco Fiorentini for their contribution to the peer review of this work. Peer reviewer reports are available.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Extended data figures and tables
Modified after ref. 66.
a, The effect of zircon crystallization in evolved, residual melts after fractional crystallization, illustrated using data from the Palaeoarchaean (3.46-Ga) Barberton volcanic rocks27,47. The orange curve shows the relationship based on median whole-rock and zircon compositions according to the zircon Ce4+/Ce3+ oxybarometer11, whereas the dark purple curve is calculated using the median composition of quartz-hosted melt inclusions that represent residual melts27,47. For a given ΔFMQ value (median −0.13 for the Barberton volcanic zircons), this results in approximately 1.5 wt% lower H2O content (compare the orange and dark purple arrows). However, as most zircons crystallized earlier than quartz, according to thermodynamic–geochemical modelling27, the slightly evolved melts (for example, the light purple curve) only have limited effect on the estimate of magmatic H2O content (within 1 wt%). b, The effect of metamorphic disturbance of whole-rock composition, illustrated using data from the Eoarchaean (3.72-Ga) Aktash gneisses, Tarim Craton31,53. Similarly, the orange curve shows the relationship between magmatic H2O content and ΔFMQ for the median zircon (this study) and whole-rock compositions (Supplementary Table 3), whereas the other coloured curves show the effects of various metamorphic modifications of the whole-rock composition, assuming other elements being constant. The vertical dashed lines with arrows show the corresponding changes in estimated magmatic H2O content for a given ΔFMQ value (median 0.14 for the Aktash gneisses). This simple modelling shows that loss of U and Th during high-grade metamorphism or anatexis may result in overestimates of H2O content, whereas loss of LREE has an opposite effect. Preferential U loss is indicated by the high Th/U ratios (7.1–38.2) of several Archaean TTG suites (for example, the 3.72-Ga Aktash complex31,53), probably because of high-grade metamorphism and anatexis67. By contrast, the Th and LREE contents of Archaean TTGs seem to be conservative, probably because of retention of these elements in monazite67. Moreover, the opposite effect of Th-U and LREE loss means that, even if these elements were mobilized (for example, by partial dissolution of monazite in metamorphic fluids or anatectic melts), their effects tend to compensate each other. Accordingly, we only corrected for the effect of U loss for TTG suites with high Th/U ratios (>7), by assuming a Th/U ratio of 5, the median value of the global sodic TTGs15, and the results are considered as minimum estimates of equilibrium magmatic H2O content during zircon crystallization.
a, The 3.72-Ga Aktash gneisses and the 3.2–3.0-Ga Heluositan gneisses of the Tarim Craton. b, The 4.02–2.95-Ga Acasta gneisses. c, The Toba Tuff (about 0.07 million years ago) and the Pea Ridge rhyolite (about 1,450 million years ago). The negative correlations between ΔFMQ and zircon Hf contents in the Archaean TTGs (a,b) suggest that, with crystallization of the magma, as indicated by increasing Hf in the residual melt and therefore crystallizing zircon, the magmatic fO2 decreases, owing to crystallization of amphibole and ilmenite that have slightly higher Fe3+/ΣFe ratios than the primary TTG melts with relatively low Fe3+/ΣFe ratios. This implies that the median ΔFMQ values, and therefore the calculated magmatic H2O contents, are probably minimum estimates of the primary TTG melts. The negative correlations between zircon Eu/Eu* and Hf contents suggest that zircon co-crystallized with plagioclase.
a and b correspond to the T–H2O diagrams calculated at 0.2 and 0.5 GPa, respectively, assuming initial ΔFMQ = 0 at 800 °C, whereas c and d are the corresponding isopleth diagrams. In c and d, paths A′ and A″ correspond to zircon crystallization in initially H2O-undersaturated magmas with lower H2O content and predict a negative correlation between fO2 and H2O, which is not observed. By contrast, paths B′ and B″ correspond to zircon crystallization in H2O-saturated magmas at 0.2 and 0.5 GPa, respectively, and can explain the observed magmatic H2O content of about 6 wt% and about 10 wt% in most Archaean granitoids.
a and b correspond to the T–H2O diagrams of the median sodic TTGs at about 0.2 and 0.5 GPa, respectively (Extended Data Fig. 6), whereas c and d correspond to the T–H2O and P–H2O diagrams of the median-enriched Archaean basalts, respectively (Extended Data Figs. 8 and 5d). The top abscissas show the initial H2O/Ce ratios of the system according to the model H2O contents and Ce concentrations of the median sodic TTGs (40.4 ppm)15 and enriched Archaean basalts (14.7 ppm)30. The colour scale shows the logarithm Ce concentrations in melts (in ppm), which are largely controlled by melt proportions (approximately 1/F) because Ce is highly incompatible in major phases (LREE-rich accessory phases, for example, apatite and allanite, are not modelled). The solid white and dashed black lines are the isopleths of melt H2O contents (wt%) and H2O/Ce ratios, respectively. a and b show that the melt H2O/Ce ratios did not change substantially during the crystallization of TTG melts (paths A′, A″, B′ and B″) and correspond well with the initial H2O/Ce ratios, unless fluid saturation is reached when melt H2O content is buffered by fluid and the melt H2O/Ce ratio becomes lower than the initial value. c and d show that the H2O/Ce ratios of TTG melts produced by isobaric heating (paths J and K) and isothermal decompression (paths D′, E′ and F′) approach the initial H2O/Ce ratios of the magma sources, except at low-temperature (<900 °C) conditions in which low-degree melting (or extreme fractionation) produces extremely H2O-rich melts with high H2O/Ce ratios even from relatively H2O-poor sources. However, this process is unlikely to generate large-scale TTG melts because the melt proportions are too low (<10 wt%). Instead, the high H2O/Ce ratios of the groups 2 and 3 Archaean TTGs can be explained by partial melting (or fractionation) of mafic sources with 1.5–2.0 and 3–4 wt% H2O at ≥900 °C conditions, respectively.
This diagram is calculated at 1.1 GPa, at which the solidus temperature is the lowest (about 637 °C). The thick red line shows the solidus, whereas the dashed red line corresponds to fluid saturation. The dotted black line shows the H2O content used in the modelling in Fig. 3b,d, which slightly oversaturates the solidus at the lowest temperature. The dashed white lines show the melt isopleths in wt%.
Data for Archaean granitoids (orange diamonds) and Phanerozoic and Archaean PCDs (purple circles and yellow diamonds, respectively) are from this study (Supplementary Table 3), whereas data for Phanerozoic IAB (blue circles) are from ref. 32. Error bars are 2 s.e.m. a, The positive correlation between H2O/Ce and Ba/La suggests that the magma sources of Archaean granitoids were variably influenced by influx of fluids, as in Phanerozoic arc magmas. The large variations in Ba/La ratios of Archaean granitoids probably result from metamorphic disturbance of whole-rock Ba, whereas the lower Ba/La and H2O/Ce ratios of Archaean rocks compared with Phanerozoic arc magmas probably suggest that Archaean subduction was less efficient in H2O transport than Phanerozoic subduction. The dashed black line and shaded grey area are the linear regression and 95% confidence level for all data, respectively. The inset shows the distribution and correlation for the three groups of Archaean granitoids identified in the main text. b, The positive correlation between H2O/Ce and Sr/Y suggests that the H2O-rich sources of Archaean TTGs occurred in the deep lower crust or the mantle. Symbols are the same as in a.
Extended Data Fig. 10 The effects of varying αTiO2 on Ti-in-zircon temperature, magmatic fO2 and H2O content.
This is illustrated using zircon and melt inclusion data for the Ig2E unit of the Bishop Tuff. The vertical dashed line and grey bar show the αTiO2 and standard error of the Bishop Tuff, according to the magnetite–ilmenite data12. It is apparent that varying αTiO2 values may introduce large (>50 °C) uncertainties in Ti-in-zircon temperature but do not notably affect the estimates of fO2 (ΔFMQ) or H2O content.
This file contains Supplementary Tables 1–4 and references of the data sources.
This file is a spreadsheet that is used to calculate magmatic water content using the equilibrium zircon and melt/whole-rock compositions, as well as the ΔFMQ values.
About this article
Cite this article
Ge, RF., Wilde, S.A., Zhu, WB. et al. Earth’s early continental crust formed from wet and oxidizing arc magmas. Nature 623, 334–339 (2023). https://doi.org/10.1038/s41586-023-06552-0