Archaean Cratons Research Articles

Archaean multi-stage magmatic underplating drove formation of continental nuclei in the North China Craton

The geodynamic processes that formed Earth’s earliest continents are intensely debated. Particularly, the transformation from ancient crustal nuclei into mature Archaean cratons is unclear, primarily owing to the paucity of well-preserved Eoarchaean–Palaeoarchaean ‘protocrust’. Here, we report a newly identified Palaeoarchaean continental fragment—the Baishanhu nucleus—in northeastern North China Craton. U–Pb geochronology shows that this nucleus preserves five major magmatic events during 3.6–2.5 Ga. Geochemistry and zircon Lu–Hf isotopes reveal ancient 4.2–3.8 Ga mantle extraction ages, as well as later intraplate crustal reworking. Crustal architecture and zircon Hf–O isotopes indicate that proto-North China first formed in a stagnant/squishy lid geodynamic regime characterised by plume-related magmatic underplating. Such cratonic growth and maturation were prerequisites for the emergence of plate tectonics. Finally, these data suggest that North China was part of the Sclavia supercraton and that the Archaean onset of subduction occurred asynchronously worldwide.

The late Archaean granite paradox: A case study from the Zimbabwe Craton

Late- to post-tectonic high-K granites are found in many Archaean cratons and are thought to be the product of a major, crustal-scale melting event in the underlying TTG crust leading to the stabilisation of the craton. However, despite the TTG melting model being an obvious explanation for the origin of late-Archaean high-K post-tectonic granites, experimental studies show that TTGs are insufficiently fertile to produce large volumes of potassic granites. This is the late Archaean granite paradox. Here we argue that the paradox can be resolved if the TTG protolith is more potassic than might be expected from a straightforward partial melt of an Archaean basalt. We propose that a likely fertile protolith for late Archaean granites is TTG crust which has incorporated a partial melt of older felsic crust during its emplacement. This hypothesis is validated with a case study from the Neoarchaean rocks of the Zimbabwe Craton.This paradox reflects a more fundamental problem when considering the origin of Archaean TTGs, for the ‘enriched’ basaltic protolith invoked in many models of TTG genesis is not abundant in Archaean terrains, nor should it be if the basaltic protolith is a melt of primitive or depleted mantle. This means that fertile, K-rich, TTGs are not simply the product of the melting of a basaltic protolith, but involve an additional process. Three models of TTG petrogenesis are discussed which might lead to K-enrichment in the melt – the hydrothermal potassic enrichment of the basaltic protolith and the influences of fractional crystallisation and/or crustal contamination on the TTG magmas. We conclude that to produce a sufficiently fertile, K-rich TTG source in the Zimbabwe Craton the contribution of a melt phase from older TTG crust is most consistent with the major and trace element and isotopic geochemistry.

Tectonic processes and the evolution of the continental crust

The Earth is the only known planet where plate tectonics operates. We review the features of Archean and early Proterozoic geology that constrain tectonic environments and inform discussions of the onset of plate tectonics. There is the question of scale and how the results of individual case studies are put into a wider global context. Global models may be difficult to test and we seek to integrate evidence for plate tectonics being active with ancient records of subduction. We explore evidence for when the continental crust became rigid enough to facilitate plate tectonics based on the occurrence of widespread dyke swarms and large sedimentary basins, relatively widespread granulite facies metamorphism and evidence for crustal thickening. We argue that it remains difficult to constrain tectonic settings from contemporaneous metamorphic events without spatial controls. Archean cratons stabilized at different times in different areas from 3.1 to 2.5 Ga, juvenile continental crust changed from mafic to more intermediate compositions, there was a reduction in crustal growth at c. 3 Ga and increasing evidence for the lateral movement of crustal fragments. These, together with the other changes at the end of the Archean, are taken to reflect the onset of plate tectonics as the dominant global regime.

The contribution of Rob Westaway to the study of fluvial archives

Robert Westaway was a structural and hard-rock geologist who turned his attention to the study of Late Cenozoic fluvial archives, believing that the preservation of staircases of river terraces, particularly representing the Middle and Late Pleistocene, could only be explained in terms of crustal activity in response to surface processes, the latter affected by climatic change. His entry into this research area coincided with the realisation that such terrace sequences required surface or crustal uplift to have taken place over the time interval represented. Workers were unable to explain the observed amounts of uplift in terms of erosional isostasy, a process that could potentially have explained such one-way crustal movement. Westaway envisaged a mechanism by which mobile lower crust migrated to beneath uplifting areas, maintaining and reinforcing their uplift. The mechanism requires complex mathematics to explain it, as well as lending itself to mathematical modelling of the process, based on varying crustal properties and changes in the rates of surface processes in response to climatic fluctuation. Essentially the lower-crustal effect can be envisaged as a positive-feedback enhancement of erosional isostasy.It became apparent that Westaway's theories could explain geomorphological and sedimentary fluvial archives that were otherwise difficult to elucidate. Mantle-based erosional isostasy could not explain terrace staircases, for example. Many of these occur in regions that are tectonically inactive, and so cannot be attributed to neotectonic activity. A game-changer in terms of persuading the wider community came from the recognition of crustally ultrastable regions in which progressive long-timescale uplift has not occurred: Archaean cratons. Westaway's envisaged lower-crustal flow would not be expected in such regions, which have cold, brittle and immobile crust to its full depth. Ancient fluvial deposits are found close to modern valley-floor levels in such areas. Regions of younger Precambrian crust (Proterozoic) showing intermediate situations were subsequently identified. Other dilemmas could be resolved, such as the ‘back-tilting’ of the early-Middle Pleistocene Bytham River in the English Midlands, caused by its drainage crossing crustal blocks with different properties that have accordingly experienced differential uplift. Although glacio-isostasy, mostly seen in the effects of rebound since the Last Glacial Maximum (LGM), is largely accommodated in the mantle, and thus is reversed as a response to glacial loading and unloading, in areas of suitable crustal type there is evidently a small lower-crustal component that is less readily reversible.Westaway's important contribution has yet to be fully integrated into received wisdom in geomorphological and Quaternary circles, although much of it is now widely accepted and more will be explored and published in due course.

Heat flow and thermal structure in the Koyna seismic zone, western India: Implications for recurrent reservoir triggered seismicity

Scientific core drilling to a depth of 1522 m in the Koyna seismic zone, western India, provides a rare opportunity to characterize the geothermal regime and study its implications on seismogenesis. The borehole KBH-1 penetrated the 932.5 m thick Deccan Trap and continued for 589.5 m in the granite-gneiss basement. The equilibrium temperature profile was obtained from measurements made 10 months after drilling. Thermal conductivity was measured on core samples of Deccan basalt and basement granitoids. Radiogenic heat production of the basement granitoids was estimated from U, Th and K analyses. Salient results are as follows. (i) The temperature gradients in Deccan basalt and granite-gneiss basement, 26.5 mKm−1 and 15.7 mKm−1, respectively, combined with thermal conductivity measurements for the two rock types, yield a mean heat flow of 43 mW m−2. (ii) The low heat flow in Koyna is similar to values characteristic of Archaean cratons. (iii) Mean Th concentration of 6.8 ppm, U 0.9 ppm, and K 1.8% yielded a mean radiogenic heat production of 0.9 μWm−3 for the basement granitoids. (iv) New geological and geothermal constraints lead to improved estimates of temperature in the crust. Although geothermal considerations support a ∼ 25 km deep seismic-aseismic transition, the seismic activity is limited to the uppermost 10 km. (v) The infrequent occurrence of earthquakes below ∼10 km could be attributed to the low permeability of the basement granitoids, which reduces the influence of pore fluids in triggering failures along faults at deeper levels.

Intensive chemical weathering in the Arctic during the Miocene Climatic Optimum

The Arctic today has shallow, chemically immature, and frost-dominated weathering regimes. At Sokli, Finland (68°N), a 70 m deep palaeo-weathering profile is developed in a Devonian carbonatite pipe that represents fundamentally different past weathering environments. Formation of the apatite-francolite P-ore likely began under Palaeogene warm, humid climates. Later, cryptomelane (K-Mn oxide) crusts developed within the ore that have yielded peak 40Ar/39Ar ages of 16.20 ± 0.13 Ma (2σ)., The crusts formed at the redox front during the Miocene Climatic Optimum (∼16.9–14.7 Ma) under mean annual temperatures ∼12–14 °C warmer than today. The presence of the cryptomelane crust at shallow depth (15 m) indicates very low erosion rates since formation, consistent with its position on a tectonically stable Archaean craton and in the cold-based ice-divide zone of successive Fennoscandian ice sheets. The Miocene Climatic Optimum triggered a pulse of intensive weathering in mid- and low latitudes; the Sokli cryptomelane ages demonstrate that intensive chemical weathering extended into the Arctic.

Evidence for a very thick Kaapvaal craton root: Implications for equilibrium fossil geotherms in the early continental lithosphere

The presence of diamonds of Archaean age within the cratonic lithosphere is surprising because their formation implies a cool early mantle lithosphere despite the early Earth's hotter state. There remains a widely (albeit not exclusively) held concept that the prevailing geothermal gradient within Archaean cratons was not much different from today. This ‘Archaean paradox’ is supported by pressure and/or temperature (PT) estimates of potentially ancient diamond inclusion (DI) falling close to modern cratonic geotherms. Solutions to the paradox include more effective Archaean heat loss mechanisms (e.g. heat pipes), deeper lithospheric mantle roots or the early lithosphere forming in a disequilibrium state.Here we demonstrate the effects of a thermally evolving Earth on equilibrium geotherms within the Kaapvaal craton of South Africa. We modelled geotherms from the present-day conditions backward in time to 3.0 Ga at thermal equilibrium in the lithosphere by adjusting three dominant variables: (i) time-corrected crustal heat production; (ii) a secularly cooling mantle potential temperature (Tp) in the underlying convecting mantle; and, for illustrative purposes, (iii) variable depth of the lithosphere-asthenosphere boundary (LAB) or variable heat flow at the LAB. Computed model geotherms were compared to DI PT data from the Kaapvaal craton, as well as PT conditions recorded in mantle xenoliths of variable age. It is demonstrated that equilibrium geotherms in the lithosphere cannot be reconciled with PT conditions recorded in Kaapvaal craton DI nor with those recorded in Proterozoic xenoliths if either the lithosphere has remained of constant thickness or held a constant basal heat flux since 3.0 Ga. Rather, provided that PT conditions recorded by DI from the Kaapvaal craton reflect equilibration along fossil ancient equilibrium geotherms, it is required that the early lithosphere was substantially thicker than what is preserved today. The inferred former thickness of Kaapvaal craton could have reached up to ∼350 km if the prevailing mantle Tp at the time of diamond encapsulation was high and they are of Archaean age. If, however, the Archaean asthenosphere Tp was more modest or if the diamonds formed later than the Archaean the inferred LAB depths are shallower (∼250-275 km), but still substantially deeper than at the present day.Although only the Kaapvaal craton has the necessary DI PT constraints for this exercise, the fundamental nature of radioactive heat production rate and mantle cooling mean that our findings could imply that Earth's continental lithosphere in general attained its maximal vertical extent early in the Archaean and that cratons have been variably eroded and weakened since that time.

Fluid processes in the early Earth and the growth of continents

Water is an essential ingredient in transforming primitive mantle-derived (mafic) rocks into buoyant (felsic) continental crust, thereby driving the irreversible differentiation of Earth's lithosphere. The occurrence in Archaean cratons of sodic granites of the tonalite–trondhjemite–granodiorite (TTG) series, high-MgO variolitic basalts, high-Mg diorites (sanukitoids) and diamonds with harzburgitic inclusion assemblages, all require the presence of hydrous fluids in Earth's deep crust and upper (lithospheric) mantle since at least the Paleoarchaean (3.6–3.2 billion years ago). However, despite its importance, where and how water was stored in Archaean crust, and how some water was transported into the upper mantle, are poorly understood. Here, we investigate Archaean crustal fluid budgets through calculated phase equilibria for three protolith compositions — a low-MgO mafic (basaltic) composition, a high-MgO (picritic) composition and an ultrahigh-MgO ultramafic (komatiitic) composition — that are representative of mafic to ultramafic magmatic rocks in Archaean greenstone belts. We show that the mode and stability of hydrous minerals, in particular chlorite, is positively correlated with protolith MgO content, such that high-MgO basalts can store up to twice the amount of crystal-bound H2O than low-MgO basalts. Importantly, ultrahigh-MgO rocks such as komatiite can store four times as much H2O, most of which is retained until temperatures exceeding 700°C. Warmer geotherms in the early Archaean favoured dehydration of hydrated high-MgO and ultramafic rocks in the deep crust, leading to hydration and/or fluid-fluxed melting of overlying basaltic rocks to produce ‘high-pressure’ TTG magmas. Burial of Archaean mafic–ultramafic crust along cooler geotherms resulted in dehydration of ultramafic material within the lithospheric mantle, providing the source of enriched Archaean basalt that was parental to large volumes of ancient TTG-dominated continental crust.

The Paleo-Mesoarchaean Gondpipri Mafic-Ultramafic Intrusions, Western Bastar Archaean Craton, Central India: Insights from Bulk-Rock Geochemistry and Sm-Nd and S Isotope Studies on the Formation of Ni-Cu-PGE Mineralization

Abstract Magmatic Ni-Cu-platinum group element (PGE)-Te mineralization in the Gondpipri mafic-ultramafic layered intrusion of ca. 3323 ± 74 Ma age, western Bastar craton, central India, is one of the most prospective exploration targets for magmatic sulfides in India. The Gondpipri layered intrusion is divided into two distinct groups of rocks based on their mineralization potential, which include (1) mineralized layered gabbro and pyroxenite and (2) a barren olivine gabbro intrusion. The host rocks show Cu + Ni concentrations up to 5,000 ppm with a Cu/Ni ratio <1 and all PGE values between 0.1 and 1.1 ppm. Mineralization occurs in two modes: type I mineralization occurring as blebs, specks, and dissemination and type II mineralization occurring as stringers and minor veins. The geochemical data suggest that the parental magma of the host rock was generated at depths between spinel and garnet peridotite mantle source regions and subsequently modified by assimilation fractional crystallization (AFC) of the continental crust. High large ion lithophile elements, Th/Yb ratios of the studied rocks, and Sm-Nd isotope studies are consistent with a depleted mantle source. The geochemical proxies such as Th versus Ba/Th and (Ta/La)PM versus (Hf/Sm)PM and higher Sr/Nd (2.21–82.58) ratios indicate involvement of fluid-related subduction metasomatism and enrichment processes in an island-arc tectonic setting. Mineral assemblages and textural relationships between platinum group minerals (PGMs) and base metal sulfides suggest that sulfide-silicate liquid immiscibility was brought about by the precipitation of magnetite/Cr magnetite resulting in sulfide saturation in the melt by decreasing S solubility. Sulfur isotope compositions (δ34S: 1.61–3.30‰) and Sm-Nd geochemistry suggest that the sulfur was added in the tholeiitic magma by magmatic process. Crustal contamination played a significant role in sulfide saturation and in bringing about PGE and Te, As, Bi, Sb, Se (TABS) mineralization. PGM-NiTeBi developed at relatively low temperatures, whereas moncheite (PtPd)Te2 and merenskyite (PdTe) were formed at 650°C. The identification of Ni-Cu-PGM-Te in the margin of the western Bastar craton boosts deeper subsurface exploration.

Evidence for a Single Large Igneous Province at 2.11 Ga across Supercraton Superia

Abstract The Superia supercraton palaeogeographic reconstruction (c. 2.65–2.00 Ga) is predominantly based on the shared large igneous province (LIP) record of the Superior, Hearne, Wyoming, and Karelia-Kola Archaean cratons. Palaeogeographic reconstruction relies on U–Pb geochronology (i.e. magmatic barcodes), palaeolatitudes from palaeomagnetic studies, and geometry of mafic dyke swarms as part of LIPs, as well as similarities in cover stratigraphy and/or basement geology. If contemporaneous mafic units from these cratons are indeed fragments of the same LIP, then integration of their chemistry can provide insight into the overall LIP plumbing system. A geochemical evaluation of whole-rock major and trace elements, and Nd isotopes, is presented for c. 2.11 Ga mafic units from the Marathon dykes (Superior), Griffin gabbro sills and Kazan dykes (Hearne), Bear Mountain dykes (Wyoming), and Tohmajärvi-Pirtguba dykes, Misi gabbro sills, and Oravaara and Hirsimaa volcanic rocks (Karelia-Kola). These units include the Nieminen dyke, Western Karelia, for which we present a new U–Pb CA-ID-TIMS date of 2114.4 ± 3.0 Ma. Four geochemical groups are identified: Groups 1 (mid-ocean ridge basalt-like; in Karelia) and 2 (within plate basalt; in Hearne, Wyoming, and Karelia) are distinct end-members of upper mantle and plume melts, respectively; Group 3 (in Superior, Wyoming, and Karelia) is a hybrid of Groups 1 and 2; and Group 4 (in all cratons) formed by assimilation and fractional crystallisation of Groups 1, 2, and 3. We present a model of the LIP plumbing system and define three magmatic stages by combining our geochemical interpretations with existing geochronology. The early stage (2135–2130 Ma) is limited to Group 1 dykes. The middle stage (2130–2113 Ma) includes Groups 1, 3, and 4. Onset of the late stage (2113–2101 Ma) is marked by the introduction of Group 2 but Groups 1 and 4 are also abundant. Finally, with our integrated tectonic and geochemical model, we discuss the mineralisation potential and prospectivity of the entire c. 2.11 Ga magmatic system.

Craton Formation in Early Earth Mantle Convection Regimes

AbstractHow the geological record of cratons reconciles with the tectonic environments in which they formed has remained debated. We use 2D Cartesian geometry numerical models of mantle convection varying temperatures from present day to Archaean‐inferred values, to address the formation of cratons, accounting for melt depletion‐dependent rheological stiffening. For mantle temperatures comparable to present day, melting is negligible and the convective regime depends on the strength of the thermal lithosphere. For mantle potential temperatures higher than present day, high depletion degree and large depleted mantle volumes are formed at low lithospheric strength and high surface mobility, whereas these are negligible beneath a poorly mobile lithosphere. When compared to the models, the record of tectonics and large volumes of high‐degree depleted mantle in Archaean cratons is best explained by a lithosphere initially prone to yielding and mobility. At high mobility, large depletion favors the progressive differentiation of the thermochemical lithosphere, which stiffens and thickens with increasing mantle temperatures. The ensuing reduced heat flow atop a hotter mantle is in agreement with the inferred Archaean thermal evolution, and may rule out the viability of a stagnant lid for the early Earth. Large‐scale depletion stiffening resists plate margin formation and this wanes as heat production decreases, thus may hold the key for the establishment of plate tectonics during secular cooling.

Seismic reflections from a lithospheric suture zone below the Archaean Yilgarn Craton

Seismic reflectors in the uppermost mantle, which can indicate past plate tectonic subduction, are exceedingly rare below Archaean cratons, and restricted to the Neoarchaean. Here we present reprocessed seismic reflection profiles from the northwest Archaean Yilgarn Craton and the Palaeoproterozoic Capricorn Orogen of western Australia that reveal the existence of a ~4 km thick south-dipping band of seismic reflectors that extends from the base of the Archaean crust to at least 60 km depth. We interpret these reflectors, which lie south of a ~50 km deep crustal root, as a relict suture zone within the lithosphere. We suggest that the mantle reflectors were created either by subduction of an oceanic plate along the northern edge of the Yilgarn Craton, which started in the Mesoarchaean and produced the rocks in northern Yilgarn greenstone belts that formed in a supra-subduction zone setting, or, alternatively, by underthrusting of continental crust deep into the lithosphere during the Palaeoproterozoic.

Geochemical reconnaissance of the Guéra and Ouaddaï Massifs in Chad: evolution of Proterozoic crust in the Central Sahara Shield

Abstract In 1964, W.Q. Kennedy suggested that the crust of Saharan Africa is different from the rest of Africa. To date, the geologic evolution of this region remains obscure because the age and composition of crystalline basement are unknown across large sectors of the Sahara. Most of Africa comprises Archaean cratons surrounded by Palaeo- to Mesoproterozoic orogenic belts, which together constitute Africa’s three major shields (the Southern, Central and West African Shields), finally assembled along belts of Pan-African rocks. By contrast, central Saharan Africa (5.3x106 km2), an area just over half the size of Europe, is considered either as a Neoproterozoic region constructed of relatively juvenile crust (0.5 to 1.0 Ga), or as an older (North African) shield that was reactivated and re-stabilized during that time, a period commonly referred to as “Pan African”. Here, using U-Pb zircon age determinations and Nd isotopic data, we show that remote areas in Chad, part of the undated Darfur Plateau stretching across ¾ million km2 of the central Sahara, comprise an extensive Neoproterozoic crystalline basement of pre-tectonic gabbro-tonalite-granodiorite and predominantly post-tectonic alkali feldspar granites and syenites that intruded between ca. 550 to 1050 Ma. This basement is flanked along its western margin by a Neoproterozoic continental calc-alkaline magmatic arc coupled to a cryptic suture zone that can be traced for ~2400 km from Tibesti through western Darfur into Cameroon. We refer to this as the Central Saharan Belt. This, in a Gondwana framework, is part of a greater arc structure, which we here term the Great Central Gondwana Arc (GCGA). Inherited zircons and Nd isotopic ratios indicate the Neoproterozoic magmas in the central Sahara were predominantly derived from Mesoproterozoic continental lithosphere. Regional deformation between 613 to 623 Ma marks the onset of late alkaline granite magmatism that was widespread across a much larger area of North Africa until about 550 Ma. During this magmatism, the region was exhumed and eroded, leaving a regional peneplain on which early Palaeozoic (Lower-Middle Cambrian) siliciclastic sediments were subsequently deposited, as part of a thick and widespread cover that stretched across much of North Africa and the Arabian Peninsula. Detrital zircons in these cover sequences provide evidence that a substantial volume of detritus was derived from the central Sahara region, because these sequences include ‘Kibaran-age’ zircons (ca. 1000 Ma) for which a source terrain has hitherto been lacking. We propose that, in preference to calling the central Sahara a “ghost” or “meta” craton, it should be called the Central Sahara Shield.

Thermal conductivity of granitoids of varying composition up to 300 °C and implications for crustal thermal models

SUMMARY A thermal conductivity profile through the upper crustal column is an essential ingredient in any thermal modelling. Granitoid is one of the major constituents of the upper crust in the Archaean cratons. Although granitoids have a wide range in composition, yet data on their thermal conductivity at elevated temperatures are very limited. At present, a single value is commonly used to characterize the decrease in thermal conductivity with temperature for the upper crust. We are reporting thermal conductivity measured at 25 °C, 50 °C and thereafter at 50 °C intervals up to 300 °C on 34 granitoid samples of four compositionally different types. The samples are alkali granite, biotite granite, granodiorite and metasomatised granodiorite from one of the Archaean cratons of the Indian shield, known as Bundelkhand Craton. Before studying the samples at elevated temperatures, these have been studied for their physical, petrological and geochemical characteristics. By 300 °C, the thermal conductivity decreases on the average by 28–31 per cent for metasomatised granodiorite, alkali granite and biotite granite, and in stark contrast by 16 per cent for granodiorite. Expressing the thermal conductivity variation with temperature as λT = λRT (1 + bT)–1, two distinct temperature coefficient (b) values have been found, 1.1 × 10–3 to 2.2 × 10–3 K–1 for alkali feldspar granite to monzogranite and 0.4 × 10–3 to 1.2 × 10–3 K–1 for granodiorite to tonalite to quartz diorite. One of the implications of this outcome is illustrated by applying these two distinct temperature coefficients for the upper crust for a 1-D generic model with a surface heat flow and appropriate radiogenic heat production of the crustal column in arriving at crustal temperature–depth profiles. The temperature differences at the base of a 40-km crust vary as much as 90 °C. Further, the temperature coefficient can be expressed as b = 0.71 × λRT– 0.63 for the alkali feldspar granite to monzogranite, whereas b = 0.83 × λRT– 1.26 for the granodiorite to tonalite to quartz diorite, which will be useful in determining the temperature coefficient of various types of granitoid from thermal conductivity at room temperature (λRT).

Depth-dependent peridotite-melt interaction and the origin of variable silica in the cratonic mantle

Peridotites from the thick roots of Archaean cratons are known for their compositional diversity, whose origin remains debated. We report thermodynamic modelling results for reactions between peridotite and ascending mantle melts. Reaction between highly magnesian melt (komatiite) and peridotite leads to orthopyroxene crystallisation, yielding silica-rich harzburgite. By contrast, shallow basalt-peridotite reaction leads to olivine enrichment, producing magnesium-rich dunites that cannot be generated by simple melting. Komatiite is spatially and temporally associated with basalt within Archaean terranes indicating that modest-degree melting co-existed with advanced melting. We envisage a relatively cool mantle that experienced episodic hot upwellings, the two settings could have coexisted if roots of nascent cratons became locally strongly extended. Alternatively, deep refractory silica-rich residues could have been detached from shallower dunitic lithosphere prior to cratonic amalgamation. Regardless, the distinct Archaean melting-reaction environments collectively produced skewed and multi-modal olivine distributions in the cratonic lithosphere and bimodal mafic-ultramafic volcanism at surface.

Thermochemical lithosphere differentiation and the origin of cratonic mantle.

Cratons record the early history of continental lithosphere formation, yet how they became the most enduring part of the lithosphere on Earth remains unknown1. Here we propose a mechanism for the formation of large volumes of melt-depleted cratonic lithospheric mantle (CLM) and its evolution to stable cratons. Numerical models show large decompression melting of a hot, early Earth mantle beneath a stretching lithosphere, where melt extraction leaves large volumes of depleted mantle at depth. The dehydrated, stiffer mantle resists further deformation, forcing strain migration and cooling, thereby assimilating depleted mantle into the lithosphere. The negative feedback between strain localization and stiffening sustains long-term diffused extension and emplacement of large amounts of depleted CLM. The formation of CLM at low pressure and its deeper re-equilibration reproduces the evolution of Archaean lithosphere constrained by depth-temperature conditions1,2, whereas large degrees of depletion3,4 and melt volumes5 in Archaean cratons are best matched by models with lower lithospheric strength. Under these conditions, which are otherwise viable for plate tectonics6,7, thermochemical differentiation effectively prevents yielding and formation of margins: rifting and lithosphere subduction are short lived and embedded in the cooling CLM as relict structures, reproducing the recycling and reworking environments that are found in Archaean cratons8,9. Although they undergo major melting and extensive recycling during an early stage lasting approximately 500million years, the modelled lithospheres progressively differentiate and stabilize, and then recycling and reworking become episodic. Early major melting and recycling events explain the production and loss of primordial Hadean lithosphere and crust10, whereas later stabilization and episodic reworking provides a context for the creation of continental cratons in the Archaean era4,8.

The evolution of the continental crust and the onset of plate tectonics.

The Earth is the only known planet where plate tectonics is active, and different studies have concluded that plate tectonics commenced at times from the early Hadean to 700 Ma. Many arguments rely on proxies established on recent examples, such as paired metamorphic belts and magma geochemistry, and it can be difficult to establish the significance of such proxies in a hotter, older Earth. There is the question of scale, and how the results of different case studies are put in a wider global context. We explore approaches that indicate when plate tectonics became the dominant global regime, in part by evaluating when the effects of plate tectonics were established globally, rather than the first sign of its existence regionally. The geological record reflects when the continental crust became rigid enough to facilitate plate tectonics, through the onset of dyke swarms and large sedimentary basins, from relatively high-pressure metamorphism and evidence for crustal thickening. Paired metamorphic belts are a feature of destructive plate margins over the last 700 Myr, but it is difficult to establish whether metamorphic events are associated spatially as well as temporally in older terrains. From 3.8-2.7 Ga, suites of high Th/Nb (subduction-related on the modern Earth) and low Th/Nb (non-subduction-related) magmas were generated at similar times in different locations, and there is a striking link between the geochemistry and the regional tectonic style. Archaean cratons stabilised at different times in different areas from 3.1-2.5 Ga, and the composition of juvenile continental crust changed from mafic to more intermediate compositions. Xenon isotope data indicate that there was little recycling of volatiles before 3 Ga. Evidence for the juxtaposition of continental fragments back to ~2.8 Ga, each with disparate histories highlights that fragments of crust were moving around laterally on the Earth. The reduction in crustal growth at ~ 3 Ga is attributed to an increase in the rates at which differentiated continental crust was destroyed, and that coupled with the other changes at the end of the Archaean are taken to reflect the onset of plate tectonics as the dominant global regime.

Spinel, olivine, and pyroxene chemistry of the Eoarchaean Manfred Complex (Yilgarn Craton, Western Australia), with implications for the tectonic setting of Archaean layered mafic intrusions and the stabilisation of continental nuclei

Anorthosite complexes are distinctive components of Archaean cratons and their petrogenesis has implications for ancient mantle evolution, early Earth geodynamics, and growth of continental crust. However, the origin and tectonic setting of Archaean anorthosites and their complementary mafic and ultramafic cumulates is uncertain due to metamorphism and deformation. Here, we report spinel, olivine, and pyroxene chemical compositions from rare, well preserved ultramafic cumulates of the Eoarchaean Manfred Complex in the Yilgarn Craton of Western Australia, which contains the oldest anorthosites on Earth. Relict magmatic spinel in peridotite and pyroxenite enclaves define a distinctive high Al, low Ti-Fe3+ layered mafic intrusion trend that establishes a petrogenetic link to the ca. 3730 Ma anorthosites. Olivine chemistry (Fo85), Ni concentrations, FeO/MgO ratio trends for coexisting olivine and liquid, and spinel Mg# identify the parental magma as picritic to tholeiitic. Unique spinel Cr#, Fe#, and Ti contents are attributed to derivation of the parental magmas from a spinel peridotite source, in contrast to younger Archaean anorthosite complexes, reflecting an evolution of Archaean mantle. Magmatic spinel Al3+, Fe3+, and Cr3+ systematics are consistent with the emplacement and differentiation of the ca. 3730 Ma Manfred Complex parental magma within a thickened oceanic plateau. This implies a similar, intraplate tectonic environment for formation of the surrounding, coeval tonalite-trondhjemite-granodiorite gneisses in the Narryer Terrane. We therefore infer that early Archaean oceanic plateaus may be important tectonic settings for producing and preserving ancient continental crust, as well as the variety of supracrustal rocks commonly observed in gneiss complexes of Archaean cratons.

Secular cooling and crystallization of partially molten Archaean continental crust over 1 Ga

The protracted tectonic and magmatic record of cratons over the Archaean Eon has been classically interpreted in terms of long-lived shallow-dipping subduction or repeated mantle plumes. In this paper, we use the 1D conductive heat equation to model the evolution of the geotherm of a generic felsic-dominated Archaean cratonic nuclei solely considering the secular decay of radioactive isotopes (238U, 235U, 232Th, and 40K), responsible for heat production in the crust. Using a range of plausible parameters for crustal thickness, lithospheric thickness, and surface heat flux, this modelling shows that Archaean crust was characterized by an initially high geothermal gradient at 3.5 Ga, with a Moho temperature close to 900 °C, and that it might have remained partially molten for about one billion years. The existence of a partially molten crust for an extended period of time offers an alternative option to shallow-dipping subduction or repeated mantle plumes for the understanding of the peculiar tectonic evolution of Archaean cratons marked by (i) protracted high-temperature metamorphism and magmatism associated with crustal differentiation, and (ii) widespread deformation characterized by structural domes attributed to the development of crustal-scale gravitational instabilities.

Configuration of Late Archaean Chilimanzi and Razi Suites of Granites, South-Central Zimbabwe Craton, From Gravity Modelling: Geotectonic Implications

The subsurface geometry of five representative late Archaean ‘Chilimanzi and Razi’ suite plutons in the Zimbabwe craton (ZC) has been investigated by gravity modelling constrained in part by surface geology, density measurements and seismic information, to determine their 3D configuration and infer tectonic context of emplacement. The generally K-rich, massive, homogeneous monzogranites are characterised by large Bouguer gravity lows (up to − 30 mGal amplitude) whose gradients outline their spatial extent well. The southernmost plutons and their anomalies have general trends paralleling the North Marginal Zone (NMZ) of the Limpopo orogenic belt (LB). Predictive gravity models indicate that the density contrast of the Chivi batholith (CB) adjacent to the ‘volcanic arc-like’ Belingwe greenstone belt extends to a depth of about 13 km. The nearby Razi pluton (RP) which intrudes the ZC-LB boundary appears to have been emplaced at shallower depths/levels. The gravity model suggests a thickness of about 5–6 km, and a moderate to shallow dip to the southeast under the NMZ, compatible with syn-kinematic intrusion during overthrust of the LB over the ZC. The smallest Nalatale granite (Ng) is on average 2.5 km thick under the Fort Rixon greenstone belt but includes a root up to 4.5 km thick under the anomaly peak, and a steep contact with the tonalite/gneiss to the east. These granites follow the general power-law for pluton dimension and are similar in this respect to the classical wedge-shaped plutons, extending largely in one direction, with large aspect ratios (Length (L)/Thickness (T) > 7). However, the overall shape of the RP is typical of a diapir (Width (W) 5; L/T > 15). Overall, the CB and the ZG are interpreted as massive, deep-rooted batholithic intrusions (L/T ≅ 10), contrary to some geological interpretations of these late, post-kinematic intrusions as sheet-like bodies emplaced at relatively shallow levels in the crust. On the other hand, the ChB appears to be a tabular intrusion, probably fed by dykes; it exhibits a lateral extent much greater than the vertical one, outlining a sheeted geometry (W/T > 7; L/T > 18). The geophysical evidence, together with geological and fabric data, support and/or confirm the two main granite configurations: sheets and batholith; and thus also confirm the two main modes of emplacement: dyke and diapirism or ballooning plutonism. This is consistent with other known batholiths on the ZC but considered unusual for plutons of the same age and spatially close when compared to other Archaean cratons.