Isotope composition and volume of Earth’s early oceans

Abstract

Oxygen and hydrogen isotope compositions of Earth’s seawater are controlled by volatile fluxes among mantle, lithospheric (oceanic and continental crust), and atmospheric reservoirs. Throughout geologic time the oxygen mass budget was likely conserved within these Earth system reservoirs, but hydrogen’s was not, as it can escape to space. Isotopic properties of serpentine from the approximately 3.8 Ga Isua Supracrustal Belt in West Greenland are used to characterize hydrogen and oxygen isotope compositions of ancient seawater. Archaean oceans were depleted in deuterium [expressed as δD relative to Vienna standard mean ocean water (VSMOW)] by at most 25 ± 5‰, but oxygen isotope ratios were comparable to modern oceans. Mass balance of the global hydrogen budget constrains the contribution of continental growth and planetary hydrogen loss to the secular evolution of hydrogen isotope ratios in Earth’s oceans. Our calculations predict that the oceans of early Earth were up to 26% more voluminous, and atmospheric CH₄ and CO₂ concentrations determined from limits on hydrogen escape to space are consistent with clement conditions on Archaean Earth.

The geologic record provides scant definitive evidence of early Archaean seawater chemistry. Oxygen and hydrogen isotope compositions of oceanic sedimentary rocks provide a commonly used proxy record, but interpretations remain inconclusive (1–4). Average δ¹⁸O values of Archaean marine sediments (cherts and carbonates) are tens of per mil lower than present-day values, and systematically increase over geologic time (2, 3). This trend has been attributed either to lower in the Archaean (1, 3), or to ocean temperatures up to approximately 70 °C (5–8). Both interpretations have been questioned based on the susceptibility of chemical sediments to diagenetic alteration over geologic time (2, 4), and for the latter, a lack of geologic evidence for the extreme greenhouse gas concentrations required to sustain such high temperatures under a less luminous young Sun (9).

Minerals formed by metasomatic reactions between seawater and oceanic lithosphere provide an alternative proxy for early ocean chemistry (10–14). Here we report hydrogen and oxygen isotope compositions of 114 silicate mineral separates from the ca. 3.8 Ga Isua Supracrustal Belt (ISB) of West Greenland, including metamorphosed basalts, gabbros, and ultramafic rocks that represent fragments of Eoarchaean oceanic crust (see Fig. S1, Table S2). Minerals formed during heterogeneous late-stage CO₂- or meteoric-dominated metamorphic overprinting are distinguished by characteristic trends in their coupled δD and δ¹⁸O values (Fig. S2). In contrast, serpentine minerals formed by reaction of seawater with ultramafic rocks (peridotites) preserved in a low-strain enclave of the ISB (15) define a different isotopic trend. Here we use these serpentines to constrain D/H and ¹⁸O/¹⁶O ratios of the Eoarchaean ocean.

Serpentine Isotope Geochemistry

Serpentine-group minerals are hydrous magnesian silicates commonly formed by hydrothermal infiltration of seawater into ultramafic oceanic crust, hydration of the mantle wedge by slab-derived fluids, or fluid-rock interaction in continentally emplaced ultramafic lithologies (16–18). Serpentinized oceanic peridotites are dominantly composed of low-temperature (< 300 °C) polymorphs lizardite and chrysotile, and to a lesser extent antigorite, which forms at temperatures > 250 °C (17). In modern oceanic serpentinites, antigorites have δD between -46 and -30‰ (Figs. 1 and 2A, filled squares), values that indicate equilibrium with modern-day seawater at temperatures between 150 and 300 °C (19). In contrast, chrysotiles and lizardites have δD predominantly < 50‰ (Figs. 1 and 2A, open squares). They likely formed by recrystallization in the presence of low-δD fluids originating from dehydration reactions in hydrothermally altered oceanic crust (20, 21). Due to its more refractory nature (12, 22, 23), antigorite retains its primary δD value, even when lizardite and chrysotile reequilibrate (21) (Fig. 1).

Fig. 1.

Hydrogen and oxygen isotope values of Isua serpentines. ISB serpentines = purple circles (Table 1). Shown relative to: modern ocean serpentinites [gray region; filled squares—antigorite, open squares—lizardite and chrysotile, hatched squares—undifferentiated serpentine; from Hess Deep [East Pacific Ridge] (52), the Marianas Trench (23, 53), the Puerto Rican Trench, the Mid-Atlantic Ridge and the Blanco Fracture Zone (16)], the meteoric water line (MWL) and modern seawater composition (VSMOW), and the isotope composition of serpentine that would be in equilibrium with (1) modern-day seawater (black line), (2) Archaean seawater from this study (δD = -25 ± 5‰; δ¹⁸O = 2.3‰; purple region), and (3) Archaean seawater as proposed by Hren et al. (3) (δD = -60‰; δ¹⁸O = -10‰; gray line). Based on the temperature-dependent fractionation of Saccocia et al. (19). Errors for Isua serpentines are 1σ (SD) or analytical error, whichever is larger (see Table 1, Table S1 and Analytical Methods; N = 2–4 for δD, N = 1–2 for δ¹⁸O).

Fig. 2.

Isotopic features of modern ocean serpentine. A. Hydrogen isotope composition of serpentine from various discrete locations (16, 21, 23, 52, 53). B. Histogram of δ¹⁸O values of modern ocean serpentinites (blue) (16, 21, 23, 52–56). Also shown is the distribution of δ¹⁸O of ISB serpentinites (purple).

Oxygen isotopes in oceanic serpentines range from +0.8 to +12.4‰, with the majority between +3 and +7‰ (Fig. 2B). Isotopic variation is due to the temperature dependence of serpentine-water δ¹⁸O fractionation (19) and the exchange of oxygen isotopes between the fluids and basaltic ocean crust (24). This can be seen in the δ¹⁸O of fluids venting at mid-ocean ridge hydrothermal systems, which are typically enriched in ¹⁸O relative to seawater by 2 to 3‰ (24–26).

Serpentinites in ophiolites range in age from Tertiary to the Eoarchaean, and show a wider range in δD than their present-day oceanic counterparts (Fig. 3), often due to postemplacement isotope exchange with crustal fluids, including meteoric waters. The highest δD values, particularly those of antigorite, are commonly considered to have preserved their original isotopic compositions established during serpentinization by seawater in oceanic environments (10, 12, 27), and have been used to make inferences about the extent to which δD_SEA WATER changes over geologic time (11).

Fig. 3.

δD of serpentine in modern ocean crust and in ophiolites as a function of time. Antigorite = solid squares, lizardite and chrysotile = open squares, whole rock data = crossed squares. Oceanic serpentines are from the same references as Fig. 1. Older serpentines are from ophiolite or greenstone belt sequences (10–12, 16, 27, 57–59, this study). Serpentine-forming fluids (δD = -25 ± 5‰) at 3.8 Ga are the minimum estimate of Archaean seawater, based on equilibrium fractionation of unaltered ISB antigorite serpentine. Blue line represents the approximate increase in seawater δD with time (calculated from average of values in Table 2) resulting from gradual, steady-state continental growth, and hydrogen escape before the rise of oxygen. The different slopes before and after approximately 2.4 Ga represent the elimination of appreciable hydrogen escape after the rise of oxygen (Fig. S7). We note that seawater evolution as represented by this line is approximately parallel to maximum δD of serpentine over geologic time (purple line). Phan = Phanerozoic.

Serpentines in the Isua Supracrustal Belt

The Isua Supracrustal Belt represents the best preserved oceanic crust of Eoarchaean age (15, 28, 29). It has undergone regional greenschist- to amphibolite-facies metamorphism (SI Text), but both geochemical (15) and structural preservation (14) of protolith material occurs within localized low-strain exposures. For example, original bedding is visible in low-δ¹³C carbon-rich marine sediments (30), and pillow lavas and sheeted dike structures associated with metagabbroic and serpentinized ultramafic rocks indicate that the ISB contains preserved fragments of oceanic crust (14, 29).

Hydrogen and oxygen isotopes of the analyzed minerals distinguish two metasomatic events that have selectively altered ISB lithologies (Fig. S2). First, CO₂-rich fluids likely associated with ca. 3.81 - 3.74 Ga granitoid intrusions (14, 15, 28) intensely metasomatized ultramafic lithologies and adjacent amphibolites and felsic schists along the western and southern portions of the ISB. A second, later-stage meteoric-overprinting in the ISB is restricted to highly deformed regions and along contacts with late-stage intrusions (14). Isotopically, the first metasomatic event is distinguished by a large increase in δ¹⁸O (up to +18‰) with an insignificant change in δD, and the second resulted in a progressive decrease in δD (as low as -129‰) with very little change in δ¹⁸O. The isotopic trends associated with both events apparently originate at a common protolith isotopic composition (Fig. S2).

Serpentinites are preserved in regions of the ISB not affected by the CO₂-metasomatic event (Fig. S1) (28). Antigorite, lizardite, and chrysotile occur with varying modes, and have a cumulative range in δD of -99 to -53‰ and δ¹⁸O of +0.1 to +5.6‰ (Table 1; Fig. 1, purple circles). Serpentines with the highest δD (samples from low-strain outcrops) exhibit little variation in δ¹⁸O, but samples with δD ≲ -80‰ (highly deformed samples or those composed predominantly of chrysotile/lizardite) have δ¹⁸O that decrease with decreasing δD. This hockey-stick trend in values of δD and δ¹⁸O of hydrothermal minerals is characteristic of progressive meteoric-hydrothermal alteration (31), where decreasing δD and δ¹⁸O values of crustal rocks reflects increasing water-rock (W/R) ratios. Hydrogen isotopes in the rock are more significantly affected by secondary water-rock reaction than oxygen isotopes because of the greater oxygen content in rock-forming minerals. Thus at low W/R mass ratios, δD_ROCK will reflect isotope exchange with meteoric fluids but will not, whereas at elevated W/R ratios, both the and δD_ROCK will change.

Table 1.

Isotope analyses of Isua serpentinites

Sample	Mineral*	δD	1σ†	δ18O	1σ
03-10B	A	−53 ±	0.5	4.6 ±	0.28
03-10C	A+L	−53	0.7	4.1	-‡
15-05	A	−56	1.9	4.9	0.07
03-10A	L	−56	1.3	5.3	1.13
01-10	L	−61	1.8	4.1	-
28-05	U	−69	2.2	5.1	0.42
940094	A+L	−69	2.2	3.5	-
07-08A	U	−73	1.1	4.8	0.35
940095	U	−74	0.2	4.5	0.35
09-08	A+L	−78	1.3	5.2	0.42
20-07A	L	−84	4.0	4.1	-
16-05	A+L	−85	2.1	5.6	-
08-08	L+C	−91	1.5	2.9	-
810000	A	−95	1.9	4.4	0.78
07-07	U	−98	1.2	0.1	0.42
06-07AA	A+C	−99	0.2	2.7	0.42

Relative to VSMOW.

^*Primary phase: A = Antigorite, L = Lizardite, C = Chrysotile, U = all serpentine polymorphs.

^†1σ = 1SD for N = 2–4 samples. Does not include analytical error.

^‡1σ not applicable; N = 1. Further details of analyses in Table S1.

Hypothetical water-rock reaction paths between the most deuterium-rich serpentine (δD = -53‰, δ¹⁸O = +4.8 to +5.6‰) and meteoric water (δD = -90‰, δ¹⁸O = -12.5‰) are shown in Fig. 4. δD_{METEORIC-WATER} was estimated based on approximately 40‰ fractionation (an approximate ΔD_MINERAL-H2O for Fe-Mg minerals at 100–300 °C; SI Text) between meteoric water and ISB mineral samples from regions associated with late-stage intrusions or mylonitization. The lowest δD value of all minerals analyzed in the ISB is -129‰ (Table S2). was calculated using the equation for the meteoric water line (32). With the exception of two outliers (16-05, 07-07, Table 1), the hydrogen and oxygen isotope properties of ISB serpentines are well approximated by a W/R reaction path where temperatures are approximately 100–200 °C (Fig. 4). For samples with δD > -85‰, W/R ratios were likely ≲0.1 by weight, regardless of the temperature of isotope exchange. In view of these observations, we interpret serpentines with δD > -85‰ as preserving oxygen isotope values close to their original composition.

Fig. 4.

ISB serpentine data shown relative to hypothetical reaction paths between serpentines and meteoric water. Starting composition of rock is δD = -53‰ and δ¹⁸O = +4.1 to +5.3‰ (highest δD and δ¹⁸O values of ISB serpentines) and meteoric water is δD = -90‰ and δ¹⁸O = -12.5‰. Reaction paths shown for varying temperature conditions, with water-rock ratios (W/R) denoted. Errors for Isua serpentines are 1σ (SD) or analytical error, whichever is larger (see Table 1, Table S1 and Analytical Methods; N = 2–4 for δD, N = 1–2 for δ¹⁸O).

The highest δD_SERPENTINE value recorded in the ISB provides a minimum estimate of their primary hydrogen isotope composition. However, it is possible that the serpentines with δD ≈ -53‰ do not represent the initial rock composition as we model in Fig. 4, but have also been altered by meteoric water interaction. To test this, reaction paths between the same meteoric fluid and a hypothetical starting rock with δD = -30‰ (uppermost δD of modern ocean antigorites) were also calculated. These paths are shown in Fig. S4A, and are less consistent with the Isua serpentine data than those in Fig. 4, but cannot be unequivocally ruled out. We conclude that samples with the highest δD (-56 to -53‰) and δ¹⁸O between +4.1 and +4.9‰ (Fig. 4), have undergone minimal late-stage metasomatism (W/R ≲ 0.01), and retain an isotope signature of seawater serpentinization of Eoarchaean oceanic crust.

Critical to our interpretation are the paragenetic, petrofabric, and geochemical features of the most deuterium-rich ISB serpentine samples. These samples exhibit primary pseudomorph textures after peridotitic olivine (Fig. S5), and contain modally abundant antigorite that has a tendency to seal pathways for secondary fluids and inhibit hydrogen isotope exchange (SI Text) (12, 22, 23). We note that the four samples with δD≥-56‰ are all from a region in the ISB that has been structurally and geochemically characterized as a well-preserved fragment of oceanic crust (Figs. S1, S3) (14, 29). Pons et al. (33) compared the source region for these serpentinized ultramafics to serpentinite mud volcanoes of the Mariana forearc on the basis of zinc isotopes. The presence of magnetite in the ISB serpentinites also indicates formation by low-temperature (50–300 °C) serpentinization on the ocean floor, as opposed to formation during high-grade serpentinization of the mantle wedge during crustal accretion (18, 34) (SI Text).

Over the range of temperatures typical for antigorite formation in oceanic settings (250–300 °C, SI Text), samples 15-05, 03-10A, 03-10B, and 03-10C (Table 1) would form in equilibrium with fluids having a δD between -28 and -21‰ (-25 ± 5‰, incorporating analytical error) and δ¹⁸O between +0.8 and 3.8‰ (Fig. 1, blue cross) (19). We consider these values to represent a minimum (the most negative possible) estimate for δD of Eoarchaean seawater. Additionally, we note our modeled oxygen isotope ratios for early Archaean oceans (Fig. 1) are not significantly different from those of modern-day seawater (). As in modern hydrothermal alteration of ocean crust, ISB serpentine-forming fluids may be up to approximately 2.8‰ higher than starting seawater δ¹⁸O, because of isotopic exchange with oceanic crust (25). From Fig. 2B, it is clear that the δ¹⁸O distribution in ISB serpentinites (purple values) is consistent with that of modern ocean serpentinites (blue values), particularly for samples where δD > -85‰.

Models for a Global Hydrogen Budget

Geologic processes that control the secular increase in δD_SEA WATER from -25 ± 5‰ in the Eoarchaean to 0‰ in modern oceans can be evaluated through mass balance constraints on fluxes between global reservoirs and hydrogen loss to space (35). Global reservoirs include water contained in the modern ocean, trapped in hydrous minerals of the continental crust and sediment, in surficial and glacial waters, in the biosphere, in the hydrated ocean crust and in the mantle (Table S3). Relative to the other hydrogen reservoirs, the volume of water trapped in the biosphere is negligibly small. We assume that the extent of hydration of ocean crust by reaction with seawater has likely remained approximately constant over geologic time. In addition, we tentatively consider a net zero flux of water into and out of the mantle, as modern ingassing fluxes in subduction zones and outgassing fluxes at mid-ocean ridges are comparable (see Table S4, SI Text).

Three model scenarios that consider fluxes among the remaining hydrogen reservoirs are presented in Fig. 5 as a function of total ocean mass and δD_SEA WATER, and are compared with values inferred from ISB serpentinites. In scenarios A and B, Earth behaves as a closed system for hydrogen. In A, all the hydrogen currently sequestered in the cryosphere, biosphere, surface- and ground-waters, sediments, and continental crust is allocated to the ocean reservoir, characterizing seawater before continental growth sequestered deuterium-poor fluids from the oceans into hydrous minerals (SI Text). In scenario B, 100% of modern continental crust is present, but it assumes nonglacial conditions. In these two scenarios, the ocean would have been 4 to 20% larger than modern oceans, and lower in deuterium by 9–18‰. Model scenario C is an estimate of the effect of hydrogen escape to space via methane photolysis in the preoxygenic Archaean atmosphere (36), given ΔD_CH4-H2O = -150 to -300‰ (37), and diffusion-limited hydrogen escape, where no D/H fractionation occurs (36, 38) (SI Text). The amount of hydrogen escape required to oxidize Earth’s crust indicates that preescape oceans would have been approximately 6% larger, and have a δD 10 to 20‰ lower than Vienna standard mean ocean water (VSMOW).

Fig. 5.

Hydrogen isotope composition and mass of ocean in the early Archaean. The terrestrial hydrogen budget [A, B; e.g., Lécuyer et al. (35)] and Catling et al. (36) data (C) represents the required δD and mass of the ocean to satisfy the parameters of their models. The terrestrial hydrogen budget model represents a primordial ocean in which either (A) continental crust is not yet present (δD = -18‰), or (B) where 100% of modern continental mass exists (δD = -9‰). Model (C) estimates that oxidization of all modern continents, oceans and the atmosphere due to H-escape caused 6% mass loss from the ocean and enriched the ocean in deuterium by approximately 10 to 20‰ (36). Point (D) shows the mass of the ocean required to satisfy the Archaean δD_SEA WATER values predicted by chert analyses from Hren et al. (3), given hydrogen loss via escape to space from photolysis of methane (range in ocean mass results from uncertainty in methane-water fractionation). These data are shown relative to the minimum δD values of the Archaean ocean determined from serpentine analyses (this study, Table 2), and the ocean mass as a function of the volume of continents present at 3.8 Ga (0 to 70%).

Here we consider the effects of both hydrogen sequestration in terrestrial reservoirs and hydrogen escape to space to account for the secular increase of δD_SEA WATER from -25 ± 5‰ at ca. 3.8 Ga (see SI Text for model parameters, calculations, and uncertainties). The calculated amount of hydrogen that must be lost via escape to accommodate the estimated maximum and minimum δD_{EOARCHAEAN OCEAN} values based on ISB serpentinites (-30 and -20‰) are compiled in Table 2, assuming different amounts of continental crust at approximately 3.8 Ga. If significant continental mass existed at approximately 3.8 Ga, the isotopic shift of seawater since that time would be dominantly controlled by hydrogen escape. If no continental crust existed at approximately 3.8 Ga, less hydrogen would have to escape to space to result in the observed change in δD, as some low-δD fluids would be incorporated into gradually growing continental crust. For our modeled range of δD_SEA WATER, ocean volumes would have been 9 to 26% greater than present-day, requiring a mass loss of H to space of between approximately 2 × 10²¹ mol and 18 × 10²¹ mol (Table 2; Fig. 5). Such a difference in mass may have a significant impact on current hypotheses of early Earth processes, including shallow mid-ocean ridge-crests and associated models for δ¹⁸O buffering of the Archaean ocean (1), continental freeboard in early Earth, and rates of heat and volatile transport via ocean spreading ridges (39).

Table 2.

Ocean volume at approximately 3.8 Ga based on δD_SERPENTINE = -54.5 ± 1.5‰

Archaean δDSEA WATER	% Continents Formed	H lost via escape (mol H)*	O2 Equivalent (mol O2)	Ocean size:
				(mol H2O)	(kg H2O)	(% modern)
−20	0	2.40 × 1021	0.60 × 1021	9.29 × 1022	1.67 × 1021	119
	15	3.43 × 1021	0.86 × 1021	9.16 × 1022	1.65 × 1021	118
	70	7.18 × 1021	1.80 × 1021	8.71 × 1022	1.57 × 1021	112
	100	9.23 × 1021	2.31 × 1021	8.46 × 1022	1.52 × 1021	109
−30	0	11.93 × 1021	2.98 × 1021	9.76 × 1022	1.76 × 1021	126
	15	12.82 × 1021	3.21 × 1021	9.63 × 1022	1.73 × 1021	124
	70	16.11 × 1021	4.03 × 1021	9.16 × 1022	1.65 × 1021	118
	100	17.91 × 1021	4.48 × 1021	8.90 × 1022	1.60 × 1021	114

^*Elemental hydrogen.

We consider the values determined from our hydrogen mass balance model to be reasonable first order approximations based upon average masses and isotope compositions of major hydrogen reservoirs on Earth (SI Text), which should be supported by additional evidence from the geologic record. Namely, changes in the isotopic composition of seawater with time will be apparent in other well-preserved rock suites. In Fig. 3, an extrapolation of δD_SEA WATER from present day to approximately 3.8 Ga based on the average values from Table 2 is shown as a blue line (SI Text). If Archaean δD_SEA WATER were lower; e.g., approximately -60‰ as is proposed by an alternative study based on the 3.4 Ga Buck Reef Cherts in South Africa (3), the slope of this line would be much steeper; if δD_SEA WATER had remained constant with time, the line would be horizontal. The average isotope composition of serpentine that would be in equilibrium with our modeled δD_SEA WATER is denoted by the parallel purple line, which is closely correlated to the maximum measured (and thus the best-preserved) δD values of serpentinite over geologic time.

Although an ambitious study, the Buck Reef Cherts analyzed by Hren et al. (3) (also shown in Fig. 5 as scenario D) are a less well-suited proxy for seawater δD than serpentine, resulting in the discrepancy between our two datasets. δD of chert is measured from small concentrations of water (typically less than 1.4 wt% compared to up to 14 wt% in serpentine) that are susceptible to contamination with organic matter, clay minerals, or secondary fluids (40), and the temperature dependence of H/D fractionation between chert and water is not straightforward (3, 41).

Another important consideration is the relationship between the δD and mass of Archaean oceans and the extent to which methane photolysis controlled oxidation of Earth’s surface (36). δD_SEA WATER of -60‰ (3) would result in early oceans that were 25 to 67% larger than their present size (Fig. 5), thus releasing 4 to 10 times as much oxygen by H escape from photolysis of biogenically produced methane than in Catling et al. (36)’s calculation for oxidation of the crust, summarized by the reaction:

[1]

An additional oxygen sink, such as oxidation of the mantle via subduction, would be necessary to accommodate the oxidation associated with such an extreme hydrogen flux to space over Earth history. In contrast, the amount of oxygen released due to hydrogen escape from our models in Table 2 is 2.5 ± 1.9 × 10²¹ mol O₂, a range consistent with the estimated 2.0 to 2.9 × 10²¹ mol O₂-equivalent necessary to oxidize the Earth’s crust before the rise of oxygen at ca. 2.4 Ga (36).

Finally, our model indicates that hydrogen escape is limited to at most 18 × 10²¹ mol H, releasing approximately 4.5 × 10²¹ mol of free O₂ to the Earth system (Table 2). If this much oxygen was released over approximately 1.4 billion years (3.8 to 2.4 Ga), then atmospheric CH₄ at 3.8 Ga must have been at most 480 ppmv (64 to 480 ppmv for the range of O₂ released in Table 2; SI Text) (36). Haqq-Misra et al., (42) has shown that at CH₄∶CO₂ ratios above 1.0 or 0.2 (depending on atmospheric model), a climatically cooling organic haze would form, lowering surface temperatures. For an atmospheric mixing ratio of CH₄ in the Archaean of approximately 64 to 480 ppmv, the pCO₂ needed to maintain a haze-free atmosphere would be between 10^-4.2 and 10^-2.6 bar, or at a minimum approximately 0.2 to 6 times present atmospheric levels. These values are in the pCH₄ and pCO₂ ranges suggested by Precambrian paleosols, methanogenic metabolic constraints, and magnetite/siderite stability in early Archaean banded iron formations (9, 43, 44) (Fig. S7).

δ¹⁸O of Eoarchaean Seawater

The of Archaean oceans we calculate from ISB serpentinites (+0.8 to 3.8‰; Fig. 1) is significantly higher than estimates based on carbonate and chert analyses that assume a temperate Archaean climate, which are as low as -13.3‰ (2, 3). Our values are, however, consistent with observations made from other Archaean volcanic rocks (13), including those from nearby pillow basalts in the ISB (14), from biogenic phosphates preserved in the 3.5 to 3.2 Ga Barberton Greenstone Belt (4), and from the ophiolite record of the past approximately 3.5 Ga (10).

The progressive increase in δ¹⁸O of chemical sediments over geologic time to their modern average value of approximately 0‰ [PDB] we attribute to postdepositional exchange with hydrothermal or pore water fluids on the seafloor in accord with conclusions drawn from the Barberton phosphate data (4). In modern oceans, sediments are on average approximately 500–1,000 m thick (45), and the near-surface geothermal gradient is about 25 °C/km. If the Earth’s heat flow in the early Archaean was approximately three times greater than at present (46, 47), then a thick package of seafloor sediments on the Archaean ocean floor could isotopically exchange with seawater-derived pore waters at temperatures approaching 70 °C, as on average, the temperature gradient of the Earth’s crust will scale linearly with heat flow. A decline in Earth’s heat flow with geologic time could explain the progressive increase in δ¹⁸O of cherts and carbonates, although we note that such a decrease in heat flow over geologic time has been disputed (48, 49).

Conclusions

Petrogenetic and geochemical evidence preserved in ISB serpentinites constrains a minimum δD_SEA WATER of early Archaean oceans to -25 ± 5‰ relative to present-day values (VSMOW). In the context of the metasomatic history of the ISB, the structural setting of the samples considered, the secular trend in hydrogen isotope compositions of serpentinites over geologic time, and the corroboration with independent models of surface oxidation and atmospheric methane levels in the Eoarchaean, these data appear to genuinely reflect seawater-oceanic crust interaction and are the best constraint thus far for δD of early Archaean seawater. Mass balance considerations demonstrate that low-deuterium waters were primarily sequestered in continental crust during its progressive growth since the Hadean, and later also into groundwater and glacial reservoirs. At most, hydrogen lost to space is limited to 18 × 10²¹ mol H, releasing ≲4.5 × 10²¹ mol of free O₂ to the Earth system, and constraining atmospheric methane concentrations at 3.8 Ga to ≲480 ppmv. This supports the argument that the combined greenhouse effect of atmospheric CO₂ and CH₄ cannot independently reconcile the faint early sun paradox. Additional forcing, such as a lower Earth-albedo (9), is necessary to maintain temperate conditions in the early Archaean.

Oxygen isotope compositions of ISB serpentines suggest that they formed in the presence of seawater with a δ¹⁸O similar to modern oceans, consistent with oxygen isotope studies of Archaean biogenic phosphates (4), volcanic rocks from other Archaean greenstone Belts (13, 43), and mafic pillow lavas and sheeted dikes in the ISB (14). We therefore suggest that the low δ¹⁸O of Archaean chemical sediments is a result of postdepositional exchange with shallow ground- or pore waters.

Analytical Methods

Mineral separates were hand-crushed or powdered using a microdrill for hydrogen and XRD analyses; rock samples were crushed and sieved to a size of approximately 0.5 mm, then handpicked using a stereo microscope for oxygen analysis. Mineral purity was visually assessed, and samples with better than 95% purity were used for analysis. Hydrogen and oxygen isotope analyses of mineral separates of serpentine, talc, tremolite, anthophyllite, hornblende, biotite, and muscovite from ultramafic lenses and adjacent amphibolite and felsic rocks within all major lithologies in the ISB were performed at the Stanford University Stable Isotope Biogeochemistry Lab, following the methods of Sharp (50) and Sharp et al. (51). For hydrogen isotope analysis, 1–5 mg of powdered mineral separate were dried at vacuum for > 24 h, then dropped into a Finnigan high temperature conversion elemental analyzer (TC-EA) using an autosampler flushed with helium. H₂ gas produced from sample combustion in a 1,450 °C carbon reduction furnace was introduced into a Finnigan DeltaPlusXL mass spectrometer in a helium gas stream. For oxygen isotope analysis, approximately 1 mg of mineral separates were dried at vacuum for > 1 h in a nickel sample holder. Samples were heated using a CO₂-infrared laser in a vacuum chamber containing the oxidizing agent BrF₅. The oxygen gas was directly analyzed using a dual-inlet Finnigan MAT 252 mass spectrometer. Isotope compositions of the samples were corrected relative to National Bureau of Standards samples as well as laboratory standards, and are correct within ± 3.4‰ for hydrogen and ± 0.2‰ for oxygen of accepted values. Results of these analyses are presented in the standard delta notation as parts per thousand (‰), relative to the VSMOW standard. X-ray diffraction was performed on serpentine samples to distinguish antigorite, lizardite, and chrysotile polymorphs, at the Geballe Laboratory at Stanford University.