A speculation on the structure of the D″ layer: The growth of anti-crust at the core–mantle boundary through the subduction history of the Earth

The growth curve of the continental crust shows that large amounts of continental crust formed in the early part of the Earth history are missing. In order to test a hypothesis that the former crust was subducted to the deep mantle, we performed phase assemblage analysis in the systems of mid-oceanic ridge basalt (MORB), anorthosite, and tonalite–trondhjemite–granite (TTG) down to the core–mantle boundary (CMB) conditions. Results show that all these materials can be subducted to the CMB leading to the development of a compositional layering in the D″ layer. We speculate that there could be five layers of FeO-enriched melt from partial melting of MORB, MORB crust, anorthosite, TTG, and slab or mantle peridotite in ascending order. Although the polymorphic transformation of perovskite to post-perovskite in (Mg,Fe)SiO₃ may explain the seismic discontinuity at the top of the D″ layer (D″ discontinuity), the effects of solid solution on the sharpness of the transformation suggest that the compositional layering is more plausible for the origin of the D″ discontinuity. The D″ layer can be an “anti-crust” made up mostly of TTG + anorthosite derived from the former continental crust. Tectonic style of the anti-crust at the CMB is similar to that at the surface. At both places, chemically distinct layers are density stratified and are also characterized by the processes of accretion, magmatism, and metasomatism.     

In situ X-ray diffraction study of an aluminous phase in MORB under lower mantle conditions

In situ X-ray diffraction study was conducted to identify the crystal structure of the “Al-phase”, which was previously reported to form in basaltic compositions at pressures and temperatures of the uppermost part of the lower mantle. Le Bail whole-pattern fitting method was adopted to investigate the structure of the Al-phase under high pressure and temperature as well as ambient conditions. Observed patterns were satisfactorily fitted using the “hexagonal phase” with space group P6₃/m (plus minor amount of garnet) under both of these conditions. On the other hand, the calcium ferrite structure model proposed in some earlier studies based on quench experiments yielded profile-fitting results at significantly lower confidence levels, particularly at simultaneous high pressure and high temperature conditions, suggesting that this phase may not form in oceanic crust materials subducted in the uppermost lower mantle. The difference in densities of hexagonal and calcium ferrite phases, however, is only ~1% under pressures and temperatures of the uppermost part of the lower mantle conditions, which yields a negligible effect on the bulk density of the subducted oceanic crust.     

Lost primordial continents

We investigate the bulk density variations of some representative compositions for the lower mantle based on the pressure–volume–temperature equation of state of the constituent mineral phases. The density variations of pyrolite, harzburgite, mid-ocean ridge basalt (MORB), tonalite–trondhjemite–granodiorite (TTG), and anorthosite are studied at a temperature of 300 K and at lower mantle pressures. The density of MORB is greater than that of pyrolite throughout the lower mantle, while the density of harzburgite is slightly lower than that of pyrolite. The density of anorthosite is comparable to that of pyrolite in the lower mantle in general, and greater in the lowermost mantle, while the density of TTG is lower than pyrolite throughout the lower mantle. The above results have important implications for the fate of primordial continents, TTG and anorthosite crust. While subducted TTG might be stagnant in the mantle transition zone, dense subducted anorthositic crust could be expected to sink to the core–mantle boundary (CMB) and thus might be a major component of the D" layer immediately above the CMB. Thus, we propose that significant bodies of continental material could be present in the mantle in the transition zone and immediately above the CMB, in addition to the continents on the Earth's surface.     

Reduction of buried oxidized oceanic crust during subduction

The relationship among subducted oxidized oceanic crust and oxidation state of the subarc mantle, and arc magmas is one of the important aspects to evaluate convergent margin tectonics. However details of the oxidized mass transferred from buried oceanic crust to the overlying subarc mantle wedge remain obscure. Here we investigate the Songduo eclogites from south Tibet formed by the subduction of the paleo-Tethyan oceanic crust, and identify an abrupt decrease in pyrope and increase in almandine contents from the mantle to rim of garnet grains. This is coupled with a decrease in the Fe^3 + content of epidote and Fe^3 +/(Fe^2 ++ Fe^3 +) ratios from garnet core to rim domains, as well as speciation of calcite, a new mineral phase, in the rock matrix. Minor sulfates occur only as inclusions in garnet core domains, whereas sulfides are confined to the matrix as an accessory mineral phase. Aegirine augite occurs as relics or inclusions in garnet and omphacite. These features clearly suggest that oxidized components, Fe^3 + and S^6 +, were reduced as Fe^2 + and S^2 −, respectively, at the subduction zone. Thermodynamic modeling in the P–T-log₁₀f_O2 space using updated Perplex_X programs further revealed that the Songduo eclogites experienced oxygen fugacity variation of up to 8 log₁₀ units, with decreasing pressure. Petrological observations further suggest that the strong redox processes took place, after breaking of garnet, during the initial exhumation of the eclogites. CO₂ and minor sulfur are subsequently transferred from the cold oceanic subduction zone to the overlying mantle wedge, partially released by arc volcanoes to atmosphere. Our study presents a case of C and S recycling between the Earth's exterior and interior.     

A composite geologic and seismic profile beneath the southern Rio Grande rift, New Mexico, based on xenolith mineralogy, temperature, and pressure

A complete understanding of the processes of crustal growth and recycling in the earth remains elusive, in part because data on rock composition at depth is scarce. Seismic velocities can provide additional information about lithospheric composition and structure, however, the relationship between velocity and rock type is not unique. The diverse xenolith suite from the Potrillo volcanic field in the southern Rio Grande rift, together with velocity models derived from reflection and refraction data in the area, offers an opportunity to place constraints on the composition of the crust and upper mantle from the surface to depths of  60 km. In this work, we calculate seismic velocities of crustal and mantle xenoliths using modal mineralogy, mineral compositions, pressure and temperature estimates, and elasticity data. The pressure, temperature, and velocity estimates from xenoliths are then combined with sonic logs and stratigraphy estimated from drill cores and surface geology to produce a geologic and velocity profile through the crust and upper mantle. Lower crustal xenoliths include garnet ± sillimanite granulite, two-pyroxene granulite, charnokite, and anorthosite. Metagabbro and amphibolite account for only a small fraction of the lower crustal xenoliths, suggesting that a basaltic underplate at the crust–mantle boundary is not present beneath the southern Rio Grande rift. Abundant mid-crustal felsic to mafic igneous xenoliths, however, suggest that plutonic rocks are common in the middle crust and were intraplated rather than underplated during the Cenozoic. Calculated velocities for garnet granulite are between  6.9 and 8.0 km/s, depending on garnet content. Granulites are strongly foliated and lineated and should be seismically anisotropic. These results suggest that velocities > 7.0 km/s and a layered structure, which are often attributed to underplated mafic rocks, can also be characteristic of alternating garnet-rich and garnet-poor metasedimentary rocks. Because the lower crust appears to be composed largely of metasedimentary granulite, which requires deep burial of upper crustal materials, we suggest the initial construction of the continental crust beneath the Potrillo volcanic field occurred by thickening of supracrustal material in the absence of large scale magmatic accretion. Mantle xenoliths include spinel lherzolite and harzburgite, dunite, and clinopyroxenite. Calculated P-wave velocities for peridotites range from 7.75 km/s to 7.89 km/s, with an average of 7.82 km/s. This velocity is in good agreement with refraction and reflection studies that report P_n velocities of 7.6–7.8 km/s throughout most of the Rio Grande rift. These calculations suggest that the low P_n velocities compared to average uppermost mantle are the result of relatively high temperatures and low pressures due to thin crust, as well as a fertile, Fe-rich, bulk upper mantle composition. Partial melt or metasomatic hydration of the mantle lithosphere are not needed to produce the observed P_n velocities.     

Phase transformations in descending plates and implications for mantle dynamics

The role of phase transformations in a mantle of pyrolite composition is reviewed in the light of recent experimental data. The pyroxene component of pyrolite transforms to the garnet structure at 300–350 km whilst olivine transforms to beta-Mg₂SiO₄ near 400 km. Between about 500 and 550 km, beta-Mg₂SiO₄ probably transforms to a partially inverse spinel structure whilst the CaSiO₃ component of the complex garnet solid solution exsolves and transforms to the perovskite structure. The major seismic discontinuity near 650–700 km is probably caused by disproportionation of Mg₂SiO₄ spinel into periclase plus stishovite. At a slightly greater depth, the remaining magnesian garnet transforms to the corundum or ilmenite structure. Finally, at a depth probably in the vicinity of 800–1000 km, the (Mg,Fe)SiO₃ component of the ilmenite phase transforms to a perovskite structure whilst stishovite and some of the periclase recombine to form perovskite also. The mineral assemblage so formed is about 4% denser than mixed oxides (MgO + FeO + A1₂O₃ + CaO + stishovite) isochemical with pyrolite. The above sequence of phase transformations in pyrolite provides a satisfactory general explanation of the elastic properties and density distribution in the mantle. In particular, there is no evidence requiring an increase of FeO/(FeO + MgO) ratio with depth.The depths at which major phase transformations occur in subducted lithosphere differ from those in ‘normal’ mantle. These differences are caused by two factors: (1) Temperatures within sinking plates are much lower than in surrounding mantle to depths of 700 km or more. (2) Irreversible chemical differentiation of pyrolite occurs at oceanic ridges. Lithosphere plates so formed consist of a layer of basaltic rocks underlain successively by layers of harzburgite, lherzolite, and pyrolite slightly depleted in highly incompatible elements (e.g. La, Ba, Rb, U). The phase-transformation behaviour of the first three of these layers differs from that of pyrolite.The effects of these and other factors connected with phase transformations on the dynamics of plate subsidence are discussed. It appears quite likely that plates penetrate the 650–700 km discontinuity, largely because the slope of the spinel disproportionation is probably positive, not negative as generally supposed. The former basaltic oceanic crust probably sinks deeply into the lower mantle, whilst the former harzburgite component of the plate may collect above the perovskite transition boundary. Phase transformations may thus serve as a kind of filter, leading to increased and irreversible mantle heterogeneity with time.The possible roles of phase transformations in causing deep-focus earthquakes and introducing water into the mantle in subduction zones are also briefly discussed.     

Fission track and fault kinematics analyses for new insight into the Late Cenozoic tectonic regime changes in West-Central Sulawesi (Indonesia)

A 3-D density model for the Cretan and Libyan Seas and Crete was developed by gravity modelling constrained by five 2-D seismic lines. Velocity values of these cross-sections were used to obtain the initial densities using the Nafe–Drake and Birch empirical functions for the sediments, the crust and the upper mantle. The crust outside the Cretan Arc is 18 to 24 km thick, including 10 to 14 km thick sediments. The crust below central Crete at its thickest section, has values between 32 and 34 km, consisting of continental crust of the Aegean microplate, which is thickened by the subducted oceanic plate below the Cretan Arc. The oceanic lithosphere is decoupled from the continental along a NW–SE striking front between eastern Crete and the Island of Kythera south of Peloponnese. It plunges steeply below the southern Aegean Sea and is probably associated with the present volcanic activity of the southern Aegean Sea in agreement with published seismological observations of intermediate seismicity. Low density and velocity upper mantle below the Cretan Sea with ρ  3.25 × 10³ kg/m³ and V_p velocity of compressional waves around 7.7 km/s, which are also in agreement with observed high heat flow density values, point out at the mobilization of the upper mantle material here. Outside the Hellenic Arc the upper mantle density and velocity are ρ ≥ 3.32 × 10³ kg/m³ and V_p = 8.0 km/s, respectively. The crust below the Cretan Sea is thin continental of 15 to 20 km thickness, including 3 to 4 km of sediments. Thick accumulations of sediments, located to the SSW and SSE of Crete, are separated by a block of continental crust extended for more than 100 km south of Central Crete. These deep sedimentary basins are located on the oceanic crust backstopped by the continental crust of the Aegean microplate. The stretched continental margin of Africa, north of Cyrenaica, and the abruptly terminated continental Aegean microplate south of Crete are separated by oceanic lithosphere of only 60 to 80 km width at their closest proximity. To the east and west, the areas are floored by oceanic lithosphere, which rapidly widens towards the Herodotus Abyssal plain and the deep Ionian Basin of the central Mediterranean Sea. Crustal shortening between the continental margins of the Aegean microplate and Cyrenaica of North Africa influence the deformation of the sediments of the Mediterranean Ridge that has been divided in an internal and external zone. The continental margin of Cyrenaica extends for more than 80 km to the north of the African coast in form of a huge ramp, while that of the Aegean microplate is abruptly truncated by very steep fractures towards the Mediterranean Ridge. Changes in the deformation style of the sediments express differences of the tectonic processes that control them. That is, subduction to the northeast and crustal subsidence to the south of Crete. Strike-slip movement between Crete and Libya is required by seismological observations.     

Origin of high-velocity anomalies beneath the Siberian craton: A fingerprint of multistage magma underplating since the Neoarchean

Despite the violent eruption of the Siberian Traps at ~ 250 Ma, the Siberian craton has an extremely low heat flow (18–25 mW/m²) and a very thick lithosphere (300–350 km), which makes it an ideal place to study the influence of mantle plumes on the long-term stability of cratons. Compared with seismic velocities of rocks, the lower crust of the Siberian craton is composed mainly of mafic granulites and could be rather heterogeneous in composition. The very high V_p (> 7.2 km/s) in the lowermost crust can be fit by a mixture of garnet granulites, two-pyroxene granulites, and garnet gabbro due to magma underplating. The high-velocity anomaly in the upper mantle (V_p = 8.3-8.6 km/s) can be interpreted by a mixture of eclogites and garnet peridotites. Combined with the study of lower crustal and mantle xenoliths, we recognized multistage magma underplating at the crust-mantle boundary beneath the Siberian craton, including the Neoarchean growth and Paleoproterozoic assembly of the Siberian craton beneath the Markha terrane, the Proterozoic collision along the Sayan-Taimyr suture zone, and the Triassic Siberian Trap event beneath the central Tunguska basin. The Moho becomes a metamorphism boundary of mafic rocks between granulite facies and eclogite facies rather than a chemical boundary that separates the mafic lower crust from the ultramafic upper mantle. Therefore, multistage magma underplating since the Neoarchean will result in a seismic Moho shallower than the petrologic Moho. Such magmatism-induced compositional change and dehydration will increase viscosity of the lithospheric mantle, and finally trigger lithospheric thickening after mantle plume activity. Hence, mantle plumes are not the key factor for craton destruction.     

The role of subduction erosion in the generation of Andean and other convergent plate boundary arc magmas,the continental crust and mantle

Subduction erosion, which occurs at all convergent plate boundaries associated with magmatic arcs formed on crystalline forearc basement, is an important process for chemical recycling, responsible globally for the transport of ~1.7 Armstrong Units (1 AU = 1 km³/yr) of continental crust back into the mantle. Along the central Andean convergent plate margin, where there is very little terrigenous sediment being supplied to the trench as a result of the arid conditions, the occurrence of mantle-derived olivine basalts with distinctive crustal isotopic characteristics (⁸⁷Sr/⁸⁶Sr ≥ 0.7050; ε_Nd ≤ −2; ε_Hf ≤ +2) correlates spatially and/or temporally with regions and/or episodes of high rates of subduction erosion, and a strong case can be made for the formation of these basalts to be due to incorporation into the subarc mantle wedge of tectonically eroded and subducted forearc continental crust. In other convergent plate boundary magmatic arcs, such as the South Sandwich and Aleutian Islands intra-oceanic arcs and the Central American and Trans-Mexican continental margin volcanic arcs, similar correlations have been demonstrated between regions and/or episodes of relatively rapid subduction erosion and the genesis of mafic arc magmas containing enhanced proportions of tectonically eroded and subducted crustal components that are chemically distinct from pelagic and/or terrigenous trench sediments. It has also been suggested that larger amounts of melts derived from tectonically eroded and subducted continental crust, rising as diapirs of buoyant low density subduction mélanges, react with mantle peridotite to form pyroxenite metasomatites that than melt to form andesites. The process of subduction erosion and mantle source region contamination with crustal components, which is supported by both isotopic and U-Pb zircon age data implying a fast and efficient connectivity between subduction inputs and magmatic outputs, is a powerful alternative to intra-crustal assimilation in the generation of andesites, and it negates the need for large amounts of mafic cumulates to form within and then be delaminated from the lower crust, as required by the basalt-input model of continental crustal growth. However, overall, some significant amount of subducted crust and sediment is neither underplated below the forearc wedge nor incorporated into convergent plate boundary arc magmas, but instead transported deeper into the mantle where it plays a role in the formation of isotopically enriched mantle reservoirs. To ignore or underestimate the significance of the recycling of tectonically eroded and subducted continental crust in the genesis of convergent plate boundary arc magmas, including andesites, and for the evolution of both the continental crust and mantle, is to be on the wrong side of history in the understanding of these topics.     

Role of the subducted slab,mantle wedge and continental crust in the generation of adakites from the Andean Austral Volcanic Zone

 All six Holocene volcanic centers of the Andean Austral Volcanic Zone (AVZ; 49–54°S) have erupted exclusively adakitic andesites and dacites characterized by low Yb and Y concentrations and high Sr/Y ratios, suggesting a source with residual garnet, amphibole and pyroxene, but little or no olivine and plagioclase. Melting of mafic lower crust may be the source for adakites in some arcs, but such a source is inconsistent with the high Mg# of AVZ adakites. Also, the AVZ occurs in a region of relatively thin crust (<35 km) within which plagioclase rather than garnet is stable. The source for AVZ adakites is more likely to be subducted oceanic basalt, recrystallized to garnet-amphibolite or eclogite. Geothermal models indicate that partial melting of the subducted oceanic crust is probable below the Austral Andes due to the slow subduction rate (2 cm/year) and the young age (<24 Ma) of the subducted oceanic lithosphere. Geochemical models for AVZ adakites are also consistent with a large material contribution from subducted oceanic crust (35–90% slab-derived mass), including sediment (up to 4% sediment-derived mass, representing approximately 15% of all sediment subducted). Variable isotopic and trace-element ratios observed for AVZ adakites, which span the range reported for adakites world-wide, require multistage models involving melting of different proportions of subducted basalt and sediment, as well as an important material contribution from both the overlying mantle wedge (10–50% mass contribution) and continental crust (0–30% mass contribution). Andesites from Cook Island volcano, located in the southernmost AVZ (54°S) where subduction is more oblique, have MORB-like Sr, Nd, Pb and O isotopic composition and trace-element ratios. These can be modeled by small degrees (2–4%) of partial melting of eclogitic MORB, yielding a tonalitic parent (intermediate SiO₂, CaO/Na₂O>1), followed by limited interaction of this melt with the overlying mantle (≥90% MORB melt, ≤10% mantle), but only very little (≤1%) or no participation of either subducted sediment or crust. In contrast, models for the magmatic evolution of Burney (52°S), Reclus (51°S) and northernmost AVZ (49–50°S) andesites and dacites require melting of a mixture of MORB and subducted sediment, followed by interaction of this melt not only with the overlying mantle, but the crust as well. Crustal assimilation and fractional crystallization (AFC) processes and the mass contribution from the crust become more significant northwards in the AVZ as the angle of convergence becomes more orthogonal. Received: 1 March 1995 / Accepted: 13 September 1995     

The crustal tongue melting model and the origin of massive anorthosites

Recent detailed field studies in several anorthosite complexes have shown that anorthosites are frequently associated with weakness zones in the crust which may have favoured their emplacement at mid-crust levels. Recent experimental data have shown that the parent magma compositions of various anorthosite massifs lie on thermal highs in the relevant phase diagrams at 10–13 kbar, indicating that these magmas cannot be derived by fractionation of peridotitic mantle melts but by melting of gabbronoritic sources in the lower crust at 40–50 km depths. In the Sveconorwegian Province terne boundaries have been traced in deep seismic profiles to Moho offsets or to tongues of lower crustal material underthrust to depths higher than 40 km. In Southern Norway, we suggest that a lithospheric-scale weakness zone (the Feda transition zone?) has channelled the Rogaland anorthosites through linear delamination, asthenospheric uprise and melting of a mafic lower crustal tongue.     

Exhumation of oceanic eclogites: thermodynamic constraints on pressure,temperature, bulk composition and density

The exhumation mechanism of high‐pressure (HP) and ultrahigh‐pressure (UHP) eclogites formed by the subduction of oceanic crust (hereafter referred to as oceanic eclogites) is one of the primary uncertainties associated with the subduction factory. The phase relations and densities of eclogites with MORB compositions are modelled using thermodynamic calculations over a P–T range of 1–4 GPa and 400–800 °C, respectively, in the NCKFMASHTO (Na₂O–CaO–K₂O–FeO–MgO–Al₂O₃–SiO₂–H₂O–TiO₂–Fe₂O₃) system. Our modelling suggests that the mineral assemblages, mineral proportions and density of oceanic crust subducted along a cold P–T path are quite different from those of crust subducted along a warm P–T path, and that the density of oceanic eclogites is largely controlled by the stability of low‐density hydrous minerals, such as lawsonite, chlorite, glaucophane and talc. Along a cold subduction P–T path with a geotherm of ~6 °C km^?1, lawsonite is always present at 1.1 to >4.0 GPa, and chlorite, glaucophane and talc can be stable at pressures of up to 2.3, 2.6 and 3.6 GPa respectively. Along such a P–T path, the density of subducted oceanic crust is always lower than that of the surrounding mantle at depths shallower than 110–120 km (< 3.3–3.6 GPa). However, along a warm subduction P–T path with a geotherm of ~10 °C km^?1, the P–T path is outside the stability field of lawsonite, and the hydrous minerals of chlorite, epidote and amphibole break down completely into dry dense minerals at relatively lower pressures of 1.5, 1.85 and 1.9 GPa respectively. Along such a warm subduction P–T path, the subducted oceanic crust becomes denser than the surrounding mantle at depths >60 km (>1.8 GPa). Oceanic eclogites with high H₂O content, oxygen fugacity, bulk‐rock X_Mg [ = MgO/(MgO + FeO)], X_Al [ = Al₂O₃/(Al₂O₃ + MgO + FeO)] and low X_Ca [ = CaO/(CaO + MgO + FeO + Na₂O)] are likely suitable for exhumation, which is consistent with the bulk‐rock compositions of the natural oceanic eclogites on the Earth's surface. On the basis of natural observations and our calculations, it is suggested that beyond depths around 110–120 km oceanic eclogites are not light enough and/or there are no blueschists to compensate the negative buoyancy of the oceanic crust, therefore explaining the lack of oceanic eclogites returned from ultradeep mantle (>120 km) to the Earth's surface. The exhumed light–cold–hydrous oceanic eclogites may have decoupled from the top part of the sinking slab at shallow depths in the forearc region and are exhumed inside the serpentinized subduction channel, whereas the dense–hot–dry eclogites may be retained in the sinking slab and recycled into deeper mantle.     

Ultrahigh-pressure eclogite transformed from mafic granulite in the Dabie orogen, east-central China

Although ultrahigh‐pressure (UHP) metamorphic rocks are present in many collisional orogenic belts, almost all exposed UHP metamorphic rocks are subducted upper or felsic lower continental crust with minor mafic boudins. Eclogites formed by subduction of mafic lower continental crust have not been identified yet. Here an eclogite occurrence that formed during subduction of the mafic lower continental crust in the Dabie orogen, east‐central China is reported. At least four generations of metamorphic mineral assemblages can be discerned: (i) hypersthene + plagioclase ± garnet; (ii) omphacite + garnet + rutile + quartz; (iii) symplectite stage of garnet + diopside + hypersthene + ilmenite + plagioclase; (iv) amphibole + plagioclase + magnetite, which correspond to four metamorphic stages: (a) an early granulite facies, (b) eclogite facies, (c) retrograde metamorphism of high‐pressure granulite facies and (d) retrograde metamorphism of amphibolite facies. Mineral inclusion assemblages and cathodoluminescence images show that zircon is characterized by distinctive domains of core and a thin overgrowth rim. The zircon core domains are classified into two types: the first is igneous with clear oscillatory zonation ± apatite and quartz inclusions; and the second is metamorphic containing a granulite facies mineral assemblage of garnet, hypersthene and plagioclase (andesine). The zircon rims contain garnet, omphacite and rutile inclusions, indicating a metamorphic overgrowth at eclogite facies. The almost identical ages of the two types of core domains (magmatic = 791 ± 9 Ma and granulite facies metamorphic zircon = 794 ± 10 Ma), and the Triassic age (212 ± 10 Ma) of eclogitic facies metamorphic overgrowth zircon rim are interpreted as indicating that the protolith of the eclogite is mafic granulite that originated from underplating of mantle‐derived magma onto the base of continental crust during the Neoproterozoic (c. 800 Ma) and then subducted during the Triassic, experiencing UHP eclogite facies metamorphism at mantle depths. The new finding has two‐fold significance: (i) voluminous mafic lower continental crust can increase the average density of subducted continental lithosphere, thus promoting its deep subduction; (ii) because of the current absence of mafic lower continental crust in the Dabie orogen, delamination or recycling of subducted mafic lower continental crust can be inferred as the geochemical cause for the mantle heterogeneity and the unusually evolved crustal composition.     

The fate of subducted oceanic crust: a mineral segregation model

Plate subduction and mantle plumes are two of the most important material transport processes of the silicate Earth. Currently, a debate exists over whether the subducted oceanic crust is recycled back to the Earth's surface through mantle plumes, and can explain their derivation and major characteristics. It is also puzzling as to why plume heads have huge melting capacities and differ dramatically from plume tails both in size and chemical composition. We present data showing that both ocean island basalt and mid-ocean ridge basalt have identical supra-primitive mantle mean Nb/U values of ～46.7, significantly larger than that of the primitive mantle value. From a mass balance calculation based on Nb/U?we have determined that nearly the whole mantle has evolved by plate subduction-induced crustal recycling during formation of the continental crust. This mixing back of subducted oceanic crust, however, is not straightforward, because it generally would be denser than the surrounding mantle, both in solid and liquid states. A mineral segregation model is proposed here to reconcile different lines of observation. First of all, subducted oceanic crustal sections are denser than the surrounding mantle, such that they can stay in the lower mantle, for billions of years as implied by isotopic data. Parts of subducted oceanic crust may eventually lose a large proportion of their heavy minerals, magnesian-silicate-perovskite and calcium-silicate-perovskite, through density segregation in ultra-low-velocity zones as well as in very-low-velocity provinces at the core-mantle boundary due to low viscosity. The remaining minerals would thus become lighter than the surrounding mantle, and could rise, trapping mantle materials, and forming mantle plumes. Mineral segregation progressively increases the SiO₂ content of the ascending oceanic crust, which enhances flux melting, and results in giant Si-enriched plume heads followed by dramatically abridged plume tails. Therefore, ancient mineral-segregated subducted oceanic crust is likely to be a major trigger and driving force for the formation of mantle plumes.     

The beginnings of hydrous mantle wedge melting

This study presents new phase equilibrium data on primitive mantle peridotite (0.33 wt% Na₂O, 0.03 wt% K₂O) in the presence of excess H₂O (14.5 wt% H₂O) from 740 to 1,200°C at 3.2–6 GPa. Based on textural and chemical evidence, we find that the H₂O-saturated peridotite solidus remains isothermal between 800 and 820°C at 3–6 GPa. We identify both quenched solute from the H₂O-rich fluid phase and quenched silicate melt in supersolidus experiments. Chlorite is stable on and above the H₂O-saturated solidus from 2 to 3.6 GPa, and chlorite peridotite melting experiments (containing ~6 wt% chlorite) show that melting occurs at the chlorite-out boundary over this pressure range, which is within 20°C of the H₂O-saturated melting curve. Chlorite can therefore provide sufficient H₂O upon breakdown to trigger dehydration melting in the mantle wedge or perpetuate ongoing H₂O-saturated melting. Constraints from recent geodynamic models of hot subduction zones like Cascadia suggest that significantly more H₂O is fluxed from the subducting slab near 100 km depth than can be bound in a layer of chloritized peridotite ~ 1 km thick at the base of the mantle wedge. Therefore, the dehydration of serpentinized mantle in the subducted lithosphere supplies free H₂O to trigger melting at the H₂O-saturated solidus in the lowermost mantle wedge. Alternatively, in cool subduction zones like the Northern Marianas, a layer of chloritized peridotite up to 1.5 km thick could contain all the H₂O fluxed from the slab every million years near 100 km depth, which suggests that the dominant form of melting below arcs in cool subduction zones is chlorite dehydration melting. Slab P–T paths from recent geodynamic models also allow for melts of subducted sediment, oceanic crust, and/or sediment diapirs to interact with hydrous mantle melts within the mantle wedge at intermediate to hot subduction zones.     

Deep structure of the Central Alps in the light of recent seismic data

The deep seismic reflection traverses across the Central Alps (NFP 20, ECORS-CROP) contain a new set of data on the lower crust which has been interpreted in different ways. One currently fashionable model depicts the European lower crust (ELC) as gently dipping below the Adriatic crust. However, this model requires that an observed sharp termination of the ELC under the internal border of the External Massifs is due to the non-transmission of organized seismic energy through the complex upper crust. This explanation is questioned as other reflections in this and similarly complex areas are recorded, and as the same sharp termination of the ELC under the internal border of the External Massifs is observed on all seismic lines for a length of 300 km. A tectonic — metamorphic cause appears to more satisfactorily explain the obeservations, and therefore an alternative model combining surface and deep geophysical data is proposed. It consists of three mutually largely decoupled tectonic levels. (1) The shallow obducted part or lid, bounded at its base by the combined Late Miocene Jura and Lombardic basal thrusts. Estimates of shortening based on balanced sections are at least about 100 km. (2) The intermediate level between the brittle-ductile transition and the top of the subducted mantle. It contains a stack of lower crust imbrications (with a minor admixture of upper mantle) accommodated by (inducted into) the ductile middle crust. Estimates of shortening based on area balancing are again of the order of slightly more than 100 km. (3) The subducted upper mantle, for which there are no reflection data.In the Central Alps the Late Miocene phase was dextrally transpressive, producing flower structures at the shallow level (External Massifs); the stacks of lower crust imbrications at the intermediate level may be the equivalent of the External Massifs at that level. Inverted flower structures of the subducted mantle are possible, but no detailed data are available.     

Experimental constraints on phase relations in subducted continental crust

. Synthesis piston cylinder experiments were carried out in the range 2.0-4.5 GPa and 680-1,050 °C to investigate phase relations in subducted continental crust. A model composition (KCMASH) has been used because all major ultrahigh-pressure (UHP) minerals of the whole range of rock types typical for continental crust can be reproduced within this system. The combination of experimental results with phase petrologic constraints permits construction of a UHP petrogenetic grid. The phase relations demonstrate that the most important UHP paragenesis consists of coesite, kyanite, phengite, clinopyroxene, and garnet in subducted continental crust. Below 700 °C talc is stable instead of garnet. As most of these minerals are also stable at much lower pressure and temperature conditions it is thus not easy to recognize UHP metamorphism in subducted crust. A general feature, however, is the absence of feldspars at H₂O-saturated conditions. Plagioclase is never stable at UHP conditions, but K-feldspar can occur in H₂O-undersaturated rocks. Mineral compositions in the experiments are fully buffered by coexisting phases. The Si content of phengite and biotite increase with increasing pressure. At 4.0 GPa, 780 °C, biotite contains 3.28 Si per formula unit, which is most probably caused by solid solution of biotite with talc. Above 800 °C, the CaAl₂SiO₆ component in clinopyroxene buffered with kyanite, coesite and a Mg-phase increases with increasing temperature, providing a tool to distinguish between 'cold' and 'hot' eclogites. Up to 10% Ca-eskolaite (Ca_0.5[]_0.5AlSi₂O₆) in clinopyroxene has been found at the highest temperature and pressure investigated (>900 °C, 4.5 GPa). Garnet buffered with coesite, kyanite and clinopyroxene displays an increase of grossular component with increasing pressure for a given temperature. Although the investigated system represents a simplification with respect to natural rocks, it helps to constrain general features of subducted continental crust. The observed phase relations and phase compositions demonstrate that at pressures >3.0 GPa and temperatures >800 °C continental crust can retain significant amounts of H₂O (>1 wt%), whereas K-free mafic or ultramafic rocks are dry at these conditions. UHP parageneses are only preserved if the whole exhumation path is situated within the stability field of phengite, i.e. if there is cooling during exhumation or if the whole exhumation occurred at T <700 °C. In contrast, break down of phengite and concomitant partial melting in terranes that show isothermal decompression may lead to a complete recrystallization of the subducted crust during exhumation. The density of UHP rocks can be estimated on the basis of the established phase relations. Pelitic rocks are likely to have a density close to mantle rocks (3.3 g/cm³) because of significant amounts of dense garnet and kyanite whereas granitic rocks are less dense (3.0 g/cm³). Hence, subducted average continental crust is most probably buoyant with respect to mantle rocks and tends to get exhumed as soon as it is detached from the down-going slab. Electronic supplementary material to this paper can be obtained by using the Springer LINK server located at http://dx.doi.org/10.1007/s00410-001-0336-3.     

http://dx.doi.org/10.1016/j.gsf.2016.07.005

The Hadean history of Earth is shrouded in mystery and it is considered that the planet was born dry with no water or atmosphere. The Earth-Moon system had many features in common during the birth stage. Solidification of the dry magma ocean at 4.53 Ga generated primordial continents with komatiite. We speculate that the upper crust was composed of fractionated gabbros and the middle felsic crust by anorthosite at ca. 21 km depth boundary, underlain by meta-anorthosite (grossular + kyanite + quartz) down to 50–60 km in depth. The thickness of the mafic KREEP basalt in the lower crust, separating it from the underlying upper mantle is not well-constrained and might have been up to ca. 100–200 km depending on the degree of fractionation and gravitational stability versus surrounding mantle density. The primordial continents must have been composed of the final residue of dry magma ocean and enriched in several critical elements including Ca, Mg, Fe, Mn, P, K, and Cl which were exposed on the surface of the dry Earth. Around 190 million years after the solidification of the magma ocean, “ABEL bombardment” delivered volatiles including H₂O, CO₂, N₂ as well as silicate components through the addition of icy asteroids. This event continued for 200 Myr with subordinate bombardments until 3.9 Ga, preparing the Earth for the prebiotic chemical evolution and as the cradle of first life. Due to vigorous convection arising from high mantle potential temperatures, the primordial continents disintegrated and were dragged down to the deep mantle, marking the onset of Hadean plate tectonics.     

Fate of carbonates within oceanic plates subducted to the lower mantle, and a possible mechanism of diamond formation

We report on high-pressure and high-temperature experiments involving carbonates and silicates at 30–80 GPa and 1,600–3,200 K, corresponding to depths within the Earth of approximately 800–2,200 km. The experiments are intended to represent the decomposition process of carbonates contained within oceanic plates subducted into the lower mantle. In basaltic composition, CaCO₃ (calcite and aragonite), the major carbonate phase in marine sediments, is altered into MgCO₃ (magnesite) via reactions with Mg-bearing silicates under conditions that are 200–300°C colder than the mantle geotherm. With increasing temperature and pressure, the magnesite decomposes into an assemblage of CO₂ + perovskite via reactions with SiO₂. Magnesite is not the only host phase for subducted carbon—solid CO₂ also carries carbon in the lower mantle. Furthermore, CO₂ itself breaks down to diamond and oxygen under geotherm conditions over 70 GPa, which might imply a possible mechanism for diamond formation in the lower mantle.     

Chlorite-rich ultramafic reaction zones in Colorado Plateau xenoliths: recorders of sub-Moho hydration

 Three chlorite-rich and one garnet-pyroxenite xenolith from the diatreme at Moses Rock, Utah, document storage and transport of water and consequent metasomatism in the mantle within the stability field of garnet peridotite, probably at depths of at least 75 km. Three mineral assemblages are present in zones in one chlorite-rich xenolith: in that xenolith, an assemblage of chlorite+enstatite+diopside+ ilmenite+titanian chondrodite is separated by diop- side+“talc” from an assemblage of chlorite+diopside+ilmenite+pyrite. Euhedral grains of enstatite (0.02% Al₂O₃, 0.05% CaO)+diopside record low temperatures, and high Mn/Fe in these pyroxenes was caused by growth in chlorite-dominated rock. Derivation from garnet lherzolite is established by relict pyrope (Py₇₁Gr₁₁Alm₁₈). The “talc” has Fe/Mg unusually high relative to that of associated chlorite, and electron probe analyses of the “talc” sum low, consistent with excess water; the unusual composition may be due solely to alteration and consequent submicroscopic intergrowths of other phases, but the “talc” could be an analogue of the high-pressure synthetic 10-Å phase. Garnet pyroxenite has a retrograde assemblage of chlorite-garnet-omphacite. The chlorite-rich rocks formed at contacts between garnet peridotite and other mantle rock in response to fluid flow. Pressures ≥2.2 GPa are consistent with stability of enstatite + aqueous fluid and of diopside + talc, with the occurrence of titanian chondrodite, and with the stability of garnet lherzolite. A chlorite separate has δ¹⁸O=6.9, consistent with mantle hydration. The small-scale reaction zones could have formed in a geologically brief time, plausibly just before eruption at about 25 Ma, and the responsible fluids probably also catalyzed recrystallization of associated eclogites. The hydration may have been restricted to shear zones that traversed the lower crust and the mantle to at least 75 km depth. The chlorite-rich rocks may be from the deepest part of the mantle that was sampled by the diatreme eruption. Chlorite-garnet pairs in garnet pyroxenites and pyrope megacrysts yield temperatures in the range 410–510^° C. Low temperatures in the mantle of the Colorado Plateau are consistent with an unusually low mantle heat flux and with cooling of lithosphere by an underlying subducted slab. Received: 14 April 1994/Accepted: 23 December 1994