Trace-element fractionation in Hadean mantle generated by melt segregation from a magma ocean

Calculations of the energetics of terrestrial accretion indicate that the Earth was extensively molten in its early history. Examination of early Archaean rocks from West Greenland (3.6-3.8 Gyr old) using short-lived 146Sm-142Nd chronometry indicates that an episode of mantle differentiation took place close to the end of accretion (4.46 +/- 0.11 Gyr ago). This has produced a chemically depleted mantle with an Sm/Nd ratio higher than the chondritic value. In contrast, application of 176Lu-176Hf systematics to 3.6-3.8-Gyr-old zircons from West Greenland indicates derivation from a mantle source with a chondritic Lu/Hf ratio. Although an early Sm/Nd fractionation could be explained by basaltic crust formation, magma ocean crystallization or formation of continental crust, the absence of coeval Lu/Hf fractionation is in sharp contrast with the well-known covariant behaviour of Sm/Nd and Lu/Hf ratios in crustal formation processes. Here we show using mineral-melt partitioning data for high-pressure mantle minerals that the observed Nd and Hf signatures could have been produced by segregation of melt from a crystallizing magma ocean at upper-mantle pressures early in Earth's history. This residual melt would have risen buoyantly and ultimately formed the earliest terrestrial protocrust.     

Late formation and prolonged differentiation of the Moon inferred from W isotopes in lunar metals

The Moon is thought to have formed from debris ejected by a giant impact with the early 'proto'-Earth and, as a result of the high energies involved, the Moon would have melted to form a magma ocean. The timescales for formation and solidification of the Moon can be quantified by using 182Hf-182W and 146Sm-142Nd chronometry, but these methods have yielded contradicting results. In earlier studies, 182W anomalies in lunar rocks were attributed to decay of 182Hf within the lunar mantle and were used to infer that the Moon solidified within the first approximately 60 million years of the Solar System. However, the dominant 182W component in most lunar rocks reflects cosmogenic production mainly by neutron capture of 181Ta during cosmic-ray exposure of the lunar surface, compromising a reliable interpretation in terms of 182Hf-182W chronometry. Here we present tungsten isotope data for lunar metals that do not contain any measurable Ta-derived 182W. All metals have identical 182W/184W ratios, indicating that the lunar magma ocean did not crystallize within the first approximately 60 Myr of the Solar System, which is no longer inconsistent with Sm-Nd chronometry. Our new data reveal that the lunar and terrestrial mantles have identical 182W/184W. This, in conjunction with 147Sm-143Nd ages for the oldest lunar rocks, constrains the age of the Moon and Earth to Myr after formation of the Solar System. The identical 182W/184W ratios of the lunar and terrestrial mantles require either that the Moon is derived mainly from terrestrial material or that tungsten isotopes in the Moon and Earth's mantle equilibrated in the aftermath of the giant impact, as has been proposed to account for identical oxygen isotope compositions of the Earth and Moon.     

Oxygen-isotope evidence from ancient zircons for liquid water at the Earth's surface 4,300 Myr ago

Granitoid gneisses and supracrustal rocks that are 3,800-4,000 Myr old are the oldest recognized exposures of continental crust. To obtain insight into conditions at the Earth's surface more than 4 Gyr ago requires the analysis of yet older rocks or their mineral remnants. Such an opportunity is presented by detrital zircons more than 4 Gyr old found within 3-Gyr-old quartzitic rocks in the Murchison District of Western Australia. Here we report in situ U-Pb and oxygen isotope results for such zircons that place constraints on the age and composition of their sources and may therefore provide information about the nature of the Earth's early surface. We find that 3,910-4,280 Myr old zircons have oxygen isotope (delta18O) values ranging from 5.4+/-0.6% to 15.0+/-0.4%. On the basis of these results, we postulate that the approximately 4,300-Myr-old zircons formed from magmas containing a significant component of re-worked continental crust that formed in the presence of water near the Earth's surface. These data are therefore consistent with the presence of a hydrosphere interacting with the crust by 4,300 Myr ago.     

Tungsten isotope evidence from approximately 3.8-Gyr metamorphosed sediments for early meteorite bombardment of the Earth

The 'Late Heavy Bombardment' was a phase in the impact history of the Moon that occurred 3.8 4.0 Gyr ago, when the lunar basins with known dates were formed. But no record of this event has yet been reported from the few surviving rocks of this age on the Earth. Here we report tungsten isotope anomalies, based on the (182)Hf (182)W system (half-life of 9 Myr), in metamorphosed sedimentary rocks from the 3.7 3.8-Gyr-old Isua greenstone belt of West Greenland and closely related rocks from northern Labrador, Canada. As it is difficult to conceive of a mechanism by which tungsten isotope heterogeneities could have been preserved in the Earth's dynamic crust mantle environment from a time when short-lived (182)Hf was still present, we conclude that the metamorphosed sediments contain a component derived from meteorites.     

Hf-Nd isotope evidence for a transient dynamic regime in the early terrestrial mantle

Modern basalts have seemingly lost all 'memory' of the primitive Earth's mantle except for an ambiguous isotopic signal observed in some rare gases. Although the Earth is expected to have reached a thermal steady state within several hundred million years of accretion, it is not known how and when the initial chemical fractionations left over from planetary accretion (and perhaps a stage involving a magma ocean) were overshadowed by fractionations imposed by modern-style geodynamics. Because of the lack of samples older than 4 Gyr, this early dynamic regime of the Earth is poorly understood. Here we compare published Hf-Nd isotope data on supracrustals from Isua, Greenland, with similar data on lunar rocks and the SNC (martian) meteorites, and show that, about 3.8 Gyr ago, the geochemical signature of the Archaean mantle was partly inherited from the initial differentiation of the Earth. The observed features seem to indicate that the planet at that time was still losing a substantial amount of primordial heat. The survival of remnants from an early layering in the modern deep mantle may account for some unexplained seismological, thermal and geochemical characteristics of the Earth as observed today.     

Super-chondritic Sm/Nd ratios in Mars, the Earth and the Moon

Small isotopic differences in the atomic abundance of neodymium-142 (142Nd) in silicate rocks represent the time-averaged effect of decay of formerly live samarium-146 (146Sm) and provide constraints on the timescales and mechanisms by which planetary mantles first differentiated. This chronology, however, assumes that the composition of the total planet is identical to that of primitive undifferentiated meteorites called chondrites. The difference in the 142Nd/144Nd ratio between chondrites and terrestrial samples may therefore indicate very early isolation (<30 Myr from the formation of the Solar System) of the upper mantle or a slightly non-chondritic bulk Earth composition. Here we present high-precision 142Nd data for 16 martian meteorites and show that Mars also has a non-chondritic composition. Meteorites belonging to the shergottite subgroup define a planetary isochron yielding an age of differentiation of 40 +/- 18 Myr for the martian mantle. This isochron does not pass through the chondritic reference value (100 x epsilon(142)Nd = -21 +/- 3; 147Sm/144Nd = 0.1966). The Earth, Moon and Mars all seem to have accreted in a portion of the inner Solar System with approximately 5 per cent higher Sm/Nd ratios than material accreted in the asteroid belt. Such chemical heterogeneities may have arisen from sorting of nebular solids or from impact erosion of crustal reservoirs in planetary precursors. The 143Nd composition of the primitive mantle so defined by 142Nd is strikingly similar to the putative endmember component 'FOZO' characterized by high 3He/4He ratios.     

Geomagnetic field strength 3.2 billion years ago recorded by single silicate crystals

The strength of the Earth's early geomagnetic field is of importance for understanding the evolution of the Earth's deep interior, surface environment and atmosphere. Palaeomagnetic and palaeointensity data from rocks formed near the boundary of the Proterozoic and Archaean eons, some 2.5 Gyr ago, show many hallmarks of the more recent geomagnetic field. Reversals are recorded, palaeosecular variation data indicate a dipole-dominated morphology and available palaeointensity values are similar to those from younger rocks. The picture before 2.8 Gyr ago is much less clear. Rocks of the Archaean Kaapvaal craton (South Africa) are among the best-preserved, but even they have experienced low-grade metamorphism. The variable acquisition of later magnetizations by these rocks is therefore expected, precluding use of conventional palaeointensity methods. Silicate crystals from igneous rocks, however, can contain minute magnetic inclusions capable of preserving Archaean-age magnetizations. Here we use a CO2 laser heating approach and direct-current SQUID magnetometer measurements to obtain palaeodirections and intensities from single silicate crystals that host magnetite inclusions. We find 3.2-Gyr-old field strengths that are within 50 per cent of the present-day value, indicating that a viable magnetosphere sheltered the early Earth's atmosphere from solar wind erosion.     

Hafnium isotope evidence for a transition in the dynamics of continental growth 3.2 Gyr ago

Earth's lithosphere probably experienced an evolution towards the modern plate tectonic regime, owing to secular changes in mantle temperature. Radiogenic isotope variations are interpreted as evidence for the declining rates of continental crustal growth over time, with some estimates suggesting that over 70% of the present continental crustal reservoir was extracted by the end of the Archaean eon. Patterns of crustal growth and reworking in rocks younger than three billion years (Gyr) are thought to reflect the assembly and break-up of supercontinents by Wilson cycle processes and mark an important change in lithosphere dynamics. In southern West Greenland numerous studies have, however, argued for subduction settings and crust growth by arc accretion back to 3.8 Gyr ago, suggesting that modern-day tectonic regimes operated during the formation of the earliest crustal rock record. Here we report in situ uranium-lead, hafnium and oxygen isotope data from zircons of basement rocks in southern West Greenland across the critical time period during which modern-like tectonic regimes could have initiated. Our data show pronounced differences in the hafnium isotope-time patterns across this interval, requiring changes in the characteristics of the magmatic protolith. The observations suggest that 3.9-3.5-Gyr-old rocks differentiated from a >3.9-Gyr-old source reservoir with a chondritic to slightly depleted hafnium isotope composition. In contrast, rocks formed after 3.2 Gyr ago register the first additions of juvenile depleted material (that is, new mantle-derived crust) since 3.9 Gyr ago, and are characterized by striking shifts in hafnium isotope ratios similar to those shown by Phanerozoic subduction-related orogens. These data suggest a transitional period 3.5-3.2 Gyr ago from an ancient (3.9-3.5 Gyr old) crustal evolutionary regime unlike that of modern plate tectonics to a geodynamic setting after 3.2 Gyr ago that involved juvenile crust generation by plate tectonic processes.     

A crystallizing dense magma ocean at the base of the Earth's mantle

The distribution of geochemical species in the Earth's interior is largely controlled by fractional melting and crystallization processes that are intimately linked to the thermal state and evolution of the mantle. The existence of patches of dense partial melt at the base of the Earth's mantle, together with estimates of melting temperatures for deep mantle phases and the amount of cooling of the underlying core required to maintain a geodynamo throughout much of the Earth's history, suggest that more extensive deep melting occurred in the past. Here we show that a stable layer of dense melt formed at the base of the mantle early in the Earth's history would have undergone slow fractional crystallization, and would be an ideal candidate for an unsampled geochemical reservoir hosting a variety of incompatible species (most notably the missing budget of heat-producing elements) for an initial basal magma ocean thickness of about 1,000 km. Differences in 142Nd/144Nd ratios between chondrites and terrestrial rocks can be explained by fractional crystallization with a decay timescale of the order of 1 Gyr. These combined constraints yield thermal evolution models in which radiogenic heat production and latent heat exchange prevent early cooling of the core and possibly delay the onset of the geodynamo to 3.4-4 Gyr ago.     

Record of mid-Archaean subduction from metamorphism in the Barberton terrain, South Africa

Although plate tectonics is the central geological process of the modern Earth, its form and existence during the Archaean era (4.0-2.5 Gyr ago) are disputed. The existence of subduction during this time is particularly controversial because characteristic subduction-related mineral assemblages, typically documenting apparent geothermal gradients of 15 degrees C km(-1) or less, have not yet been recorded from in situ Archaean rocks (the lowest recorded apparent geothermal gradients are greater than 25 degrees C km(-1)). Despite this absence from the rock record, low Archaean geothermal gradients are suggested by eclogitic nodules in kimberlites and circumstantial evidence for subduction processes, including possible accretion-related structures, has been reported in Archaean terrains. The lack of spatially and temporally well-constrained high-pressure, low-temperature metamorphism continues, however, to cast doubt on the relevance of subduction-driven tectonics during the first 1.5 Gyr of the Earth's history. Here we report garnet-albite-bearing mineral assemblages that record pressures of 1.2-1.5 GPa at temperatures of 600-650 degrees C from supracrustal amphibolites from the mid-Archaean Barberton granitoid-greenstone terrain. These conditions point to apparent geothermal gradients of 12-15 degrees C-similar to those found in recent subduction zones-that coincided with the main phase of terrane accretion in the structurally overlying Barberton greenstone belt. These high-pressure, low-temperature conditions represent metamorphic evidence for cold and strong lithosphere, as well as subduction-driven tectonic processes, during the evolution of the early Earth.     

A possible nitrogen crisis for Archaean life due to reduced nitrogen fixation by lightning.

Nitrogen is an essential element for life and is often the limiting nutrient for terrestrial ecosystems. As most nitrogen is locked in the kinetically stable form, N2, in the Earth's atmosphere, processes that can fix N2 into biologically available forms-such as nitrate and ammonia-control the supply of nitrogen for organisms. On the early Earth, nitrogen is thought to have been fixed abiotically, as nitric oxide formed during lightning discharge. The advent of biological nitrogen fixation suggests that at some point the demand for fixed nitrogen exceeded the supply from abiotic sources, but the timing and causes of the onset of biological nitrogen fixation remain unclear. Here we report an experimental simulation of nitrogen fixation by lightning over a range of Hadean (4.5-3.8 Gyr ago) and Archaean (3.8-2.5 Gyr ago) atmospheric compositions, from predominantly carbon dioxide to predominantly dinitrogen (but always without oxygen). We infer that, as atmospheric CO2 decreased over the Archaean period, the production of nitric oxide from lightning discharge decreased by two orders of magnitude until about 2.2 Gyr. After this time, the rise in oxygen (or methane) concentrations probably initiated other abiotic sources of nitrogen. Although the temporary reduction in nitric oxide production may have lasted for only 100 Myr or less, this was potentially long enough to cause an ecological crisis that triggered the development of biological nitrogen fixation.     

Chronological evidence that the Moon is either young or did not have a global magma ocean

Chemical evolution of planetary bodies, ranging from asteroids to the large rocky planets, is thought to begin with differentiation through solidification of magma oceans many hundreds of kilometres in depth. The Earth's Moon is the archetypical example of this type of differentiation. Evidence for a lunar magma ocean is derived largely from the widespread distribution, compositional and mineralogical characteristics, and ancient ages inferred for the ferroan anorthosite (FAN) suite of lunar crustal rocks. The FANs are considered to be primary lunar flotation-cumulate crust that crystallized in the latter stages of magma ocean solidification. According to this theory, FANs represent the oldest lunar crustal rock type. Attempts to date this rock suite have yielded ambiguous results, however, because individual isochron measurements are typically incompatible with the geochemical make-up of the samples, and have not been confirmed by additional isotopic systems. By making improvements to the standard isotopic techniques, we report here the age of crystallization of FAN 60025 using the (207)Pb-(206)Pb, (147)Sm-(143)Nd and (146)Sm-(142)Nd isotopic systems to be 4,360?±?3 million years. This extraordinarily young age requires that either the Moon solidified significantly later than most previous estimates or the long-held assumption that FANs are flotation cumulates of a primordial magma ocean is incorrect. If the latter is correct, then much of the lunar crust may have been produced by non-magma-ocean processes, such as serial magmatism.     

Early history of Earth's crust-mantle system inferred from hafnium isotopes in chondrites

The 176Lu to 176Hf decay series has been widely used to understand the nature of Earth's early crust-mantle system. The interpretation, however, of Lu-Hf isotope data requires accurate knowledge of the radioactive decay constant of 176Lu (lambda176Lu), as well as bulk-Earth reference parameters. A recent calibration of the lambda176Lu value calls for the presence of highly unradiogenic hafnium in terrestrial zircons with ages greater than 3.9 Gyr, implying widespread continental crust extraction from an isotopically enriched mantle source more than 4.3 Gyr ago, but does not provide evidence for a complementary depleted mantle reservoir. Here we report Lu-Hf isotope measurements of different Solar System objects including chondrites and basaltic eucrites. The chondrites define a Lu-Hf isochron with an initial 176Hf/177Hf ratio of 0.279628 +/- 0.000047, corresponding to lambda176Lu = 1.983 +/- 0.033 x 10-11 yr-1 using an age of 4.56 Gyr for the chondrite-forming event. This lambda176Lu value indicates that Earth's oldest minerals were derived from melts of a mantle source with a time-integrated history of depletion rather than enrichment. The depletion event must have occurred no later than 320 Myr after planetary accretion, consistent with timing inferred from extinct radionuclides.     

Aerobic respiration in the Archaean?

The Earth's atmosphere during the Archaean era (3,800-2,500 Myr ago) is generally thought to have been anoxic, with the partial pressure of atmospheric oxygen about 10(-12) times the present value. In the absence of aerobic consumption of oxygen produced by photosynthesis in the ocean, the major sink for this oxygen would have been oxidation of dissolved Fe(II). Atmospheric oxygen would also be removed by the oxidation of biogenic methane. But even very low estimates of global primary productivity, obtained from the amounts of organic carbon preserved in Archaean rocks, seem to require the sedimentation of an unrealistically large amount of iron and the oxidation of too much methane if global anoxia was to be maintained. I therefore suggest that aerobic respiration must have developed early in the Archaean to prevent a build-up of atmospheric oxygen before the Proterozoic. An atmosphere that contained a low (0.2-0.4%) but stable proportion of oxygen is required.     

Boron and oxygen isotope evidence for recycling of subducted components over the past 2.5 Gyr

Evidence for the deep recycling of surficial materials through the Earth's mantle and their antiquity has long been sought to understand the role of subducting plates and plumes in mantle convection. Radiogenic isotope evidence for such recycling remains equivocal because the age and location of parent-daughter fractionation are not known. Conversely, while stable isotopes can provide irrefutable evidence for low-temperature fractionation, their range in most unaltered oceanic basalts is limited and the age of any variation is unconstrained. Here we show that delta(18)O ratios in basalts from the Azores are often lower than in pristine mantle. This, combined with increased Nb/B ratios and a large range in delta(11)B ratios, provides compelling evidence for the recycling of materials that had undergone fractionation near the Earth's surface. Moreover, delta(11)B is negatively correlated with (187)Os/(188)Os ratios, which extend to subchondritic values, constraining the age of the high Nb/B, (11)B-enriched endmember to be more than 2.5 billion years (Gyr) old. We infer this component to be melt- and fluid-depleted lithospheric mantle from a subducted oceanic plate, whereas other Azores basalts contain a contribution from approximately 3-Gyr-old melt-enriched basalt. We conclude that both components are most probably derived from an Archaean oceanic plate that was subducted, arguably into the deep mantle, where it was stored until thermal buoyancy caused it to rise beneath the Azores islands approximately 3 Gyr later.     

Isotopic evidence for Mesoarchaean anoxia and changing atmospheric sulphur chemistry

The evolution of the Earth's atmosphere is marked by a transition from an early atmosphere with very low oxygen content to one with an oxygen content within a few per cent of the present atmospheric level. Placing time constraints on this transition is of interest because it identifies the time when oxidative weathering became efficient, when ocean chemistry was transformed by delivery of oxygen and sulphate, and when a large part of Earth's ecology changed from anaerobic to aerobic. The observation of non-mass-dependent sulphur isotope ratios in sedimentary rocks more than approximately 2.45 billion years (2.45 Gyr) old and the disappearance of this signal in younger sediments is taken as one of the strongest lines of evidence for the transition from an anoxic to an oxic atmosphere around 2.45 Gyr ago. Detailed examination of the sulphur isotope record before 2.45 Gyr ago also reveals early and late periods of large amplitude non-mass-dependent signals bracketing an intervening period when the signal was attenuated. Until recently, this record has been too sparse to allow interpretation, but collection of new data has prompted some workers to argue that the Mesoarchaean interval (3.2-2.8 Gyr ago) lacks a non-mass-dependent signal, and records the effects of earlier and possibly permanent oxygenation of the Earth's atmosphere. Here we focus on the Mesoarchaean interval, and demonstrate preservation of a non-mass-dependent signal that differs from that of preceding and following periods in the Archaean. Our findings point to the persistence of an anoxic early atmosphere, and identify variability within the isotope record that suggests changes in pre-2.45-Gyr-ago atmospheric pathways for non-mass-dependent chemistry and in the ultraviolet transparency of an evolving early atmosphere.     

Coupled 142Nd-143Nd evidence for a protracted magma ocean in Mars

Resolving early silicate differentiation timescales is crucial for understanding the chemical evolution and thermal histories of terrestrial planets. Planetary-scale magma oceans are thought to have formed during early stages of differentiation, but the longevity of such magma oceans is poorly constrained. In Mars, the absence of vigorous convection and plate tectonics has limited the scale of compositional mixing within its interior, thus preserving the early stages of planetary differentiation. The SNC (Shergotty-Nakhla-Chassigny) meteorites from Mars retain 'memory' of these events. Here we apply the short-lived 146Sm-142Nd and the long-lived 147Sm-143Nd chronometers to a suite of shergottites to unravel the history of early silicate differentiation in Mars. Our data are best explained by progressive crystallization of a magma ocean with a duration of approximately 100 million years after core formation. This prolonged solidification requires the existence of a primitive thick atmosphere on Mars that reduces the cooling rate of the interior.     

Atmospheric carbon dioxide concentrations before 2.2 billion years ago

The composition of the Earth's early atmosphere is a subject of continuing debate. In particular, it has been suggested that elevated concentrations of atmospheric carbon dioxide would have been necessary to maintain normal surface temperatures in the face of lower solar luminosity in early Earth history. Fossil weathering profiles, known as palaeosols, have provided semi-quantitative constraints on atmospheric oxygen partial pressure (pO2) before 2.2 Gyr ago. Here we use the same well studied palaeosols to constrain atmospheric pCO2 between 2.75 and 2.2 Gyr ago. The observation that iron lost from the tops of these profiles was reprecipitated lower down as iron silicate minerals, rather than as iron carbonate, indicates that atmospheric pCO2 must have been less than 10(-1.4) atm--about 100 times today's level of 360 p.p.m., and at least five times lower than that required in one-dimensional climate models to compensate for lower solar luminosity at 2.75 Gyr. Our results suggest that either the Earth's early climate was much more sensitive to increases in pCO2 than has been thought, or that one or more greenhouse gases other than CO2 contributed significantly to the atmosphere's radiative balance during the late Archaean and early Proterozoic eons.     

High 3He/4He ratios in picritic basalts from Baffin Island and the role of a mixed reservoir in mantle plumes

The high 3He/4He ratio of volcanic rocks thought to be derived from mantle plumes is taken as evidence for the existence of a mantle reservoir that has remained largely undegassed since the Earth's accretion. The helium isotope composition of this reservoir places constraints on the origin of volatiles within the Earth and on the evolution and structure of the Earth's mantle. Here we show that olivine phenocrysts in picritic basalts presumably derived from the proto-Iceland plume at Baffin Island, Canada, have the highest magmatic 3He/4He ratios yet recorded. A strong correlation between 3He/4He and 87Sr/86Sr, 143Nd/144Nd and trace element ratios demonstrate that the 3He-rich end-member is present in basalts that are derived from large-volume melts of depleted upper-mantle rocks. This reservoir is consistent with the recharging of depleted upper-mantle rocks by small volumes of primordial volatile-rich lower-mantle material at a thermal boundary layer between convectively isolated reservoirs. The highest 3He/4He basalts from Hawaii and Iceland plot on the observed mixing trend. This indicates that a 3He-recharged depleted mantle (HRDM) reservoir may be the principal source of high 3He/4He in mantle plumes, and may explain why the helium concentration of the 'plume' component in ocean island basalts is lower than that predicted for a two-layer, steady-state model of mantle structure.     

Statistical geochemistry reveals disruption in secular lithospheric evolution about 2.5 Gyr ago

The Earth has cooled over the past 4.5 billion years (Gyr) as a result of surface heat loss and declining radiogenic heat production. Igneous geochemistry has been used to understand how changing heat flux influenced Archaean geodynamics, but records of systematic geochemical evolution are complicated by heterogeneity of the rock record and uncertainties regarding selection and preservation bias. Here we apply statistical sampling techniques to a geochemical database of about 70,000 samples from the continental igneous rock record to produce a comprehensive record of secular geochemical evolution throughout Earth history. Consistent with secular mantle cooling, compatible and incompatible elements in basalts record gradually decreasing mantle melt fraction through time. Superimposed on this gradual evolution is a pervasive geochemical discontinuity occurring about 2.5?Gyr ago, involving substantial decreases in mantle melt fraction in basalts, and in indicators of deep crustal melting and fractionation, such as Na/K, Eu/Eu* (europium anomaly) and La/Yb ratios in felsic rocks. Along with an increase in preserved crustal thickness across the Archaean/Proterozoic boundary, these data are consistent with a model in which high-degree Archaean mantle melting produced a thick, mafic lower crust and consequent deep crustal delamination and melting--leading to abundant tonalite-trondhjemite-granodiorite magmatism and a thin preserved Archaean crust. The coincidence of the observed changes in geochemistry and crustal thickness with stepwise atmospheric oxidation at the end of the Archaean eon provides a significant temporal link between deep Earth geochemical processes and the rise of atmospheric oxygen on the Earth.