Subcratonic Lithosphere Research Articles

Geochemistry of forty‐one eclogitic and pyroxenitic mantle xenoliths from the Central Slave Craton, Canada (Ekati Diamond Mine)

AbstractThis article describes a novel dataset on non‐diamondiferous eclogite and garnet pyroxenite xenoliths from four kimberlite pipes of the Ekati Diamond Mine (Central Slave Craton, Canada). Xenoliths brought to the surface by kimberlite eruptions are direct sources of information on the composition and evolution of the Earth's mantle. Eclogite and garnet pyroxenite xenoliths, specifically, are testimony of subduction into, and metasomatism of, the mantle beneath cratons. Furthermore, these rocks are major hosts for diamond and thus an important part of the deep carbon cycle. The sample suite consists of 41 small xenoliths (2–5 cm) recovered from drill cores. The dataset includes major and trace element concentrations for garnet, clinopyroxene and ilmenite, as well as stable oxygen isotope compositions of garnets. Strontium and neodymium isotopic compositions are reported for garnet and clinopyroxene for four samples which were large enough to allow for analysis. Overall, this dataset significantly expands and complements existing datasets on diamondiferous and non‐diamondiferous xenoliths from the Slave Craton in Canada, furthering our understanding of the composition of the Slave subcratonic lithosphere. The dataset includes several samples with rare mineral assemblages, including an olivine‐bearing eclogite as well as ilmenite and apatite‐bearing garnet‐pyroxenites, and thus provides data shedding light on rarely reported compositional nuances in xenolith suites found in kimberlites.

Zirconium/hafnium fractionation and rare earth element systematics in sub-cratonic garnet pyroxenites, Norway

Ten previously studied garnet pyroxenites and one eclogite from orogenic peridotites in the Western Gneiss Region of Norway were analysed for whole-rock major and trace and mineral trace elements to characterise the evolutionary stages of the East Greenland sub-cratonic mantle. A continuous range of whole-rock MgO contents (17–30 wt.%) correlates inversely with hbox {Na}_2hbox {O}, hbox {Al}_2hbox {O}_3, and CaO contents. In contrast, rare earth element (REE) compositions allow the samples to be devided into two types. Type I samples have relatively flat primitive mantle (PM) normalised REE patterns (0.4-6.8 times PM values) with minor fractionation of middle–heavy REEs. Type II samples are simlar to type I samples but have higher light REE contents (6.1–57 times PM values) that reflect metasomatic enrichment. Whole-rock Zr contents (1–27,upmu hbox {g},hbox {g}^{-1}) and Zr/Hf ratios (17–39) are positively correlated. Type I samples have low Zr/Hf ratios that mostly differ from type II samples, indicating that metasomatism modified the initially sub-chondritic Zr/Hf ratios. Garnet Zr–Zr/Hf–Y–Ti systematics suggest that metasomatic enrichment transformed the type I samples into type II samples, consistent with processes observed in other cratonic areas (e.g., the northern East European Platform and the Kaapvaal Craton). The formation of type I pyroxenite in dunite from a melt at sim 100,text {km} depth implies the sub-chondritic Zr/Hf ratios were inherited from that melt. Such melts are thought to form by melting of garnet-bearing, melt-depleted mantle, which is consistent with models for recycling of Archaean palaeo-oceanic crust throughout the mantle prior to the formation of the sub-cratonic lithosphere beneath East Greenland.

Ascent rate of the Udachnaya-East kimberlite melts from olivine diffusion chronometry

Kimberlite melts are low volume, volatile-rich melts formed in the asthenosphere at depth >150-250 km. Due to their low viscosity and the exsolution of CO2 triggered by their interaction with the sub-cratonic mantle, these melts are highly buoyant and rise rapidly through the lithosphere, ultimately reaching the surface and erupting at sonic to super-sonic speed. Studies based on fluid/melt dynamics, fracturing and disaggregation of the transported mantle-derived cargo, as well as on H diffusion in olivine and Ar diffusion in magmatic phlogopite showed that the ascent rate of kimberlite melts can vary between 0.1 and 30 m/s. However, these rates are extrapolated from processes starting at different depths, and thus may not be representative for the whole ascent path of kimberlite melts through the sub-cratonic lithosphere. To shed light on this issue, we applied diffusion chronometry to the deepest feature ascribable to the liquid line of descent of the melts, i.e. early liquidus olivine formed around mantle-derived xenocrystic cores from the Udachnaya-East pipe (Siberian craton). Our results showed that the time elapsed between the formation of the Internal Zone II (early olivine on the liquidus) or rims (typical magmatic olivine) around xenocrystic cores and final cooling is between 5 and 74 days. Considering travelled distances of 80-110 km in accordance with thermobarometric calculations, this implies overall average ascent rates of kimberlite melts of 0.02 to 0.23 m/s, which are comparable to those of typical alkali basalts, albeit starting at much greater depth. This indicates that kimberlite melts start with moderate ascent rates and only later increase their speed, likely due to massive CO2 exsolution triggered by decarbonation reactions during melt/fluid-rock interactions in the lithospheric mantle. The overall timescales of 5 to 74 days obtained in our study overlap with the lowest values reported in the literature from chemical zoning of mantle-derived minerals (25 days to >100 years), confirming that the time elapsed between kimberlite-related mantle metasomatism and the ascent of kimberlite melts was short. Importantly, our data show that Fe-Mg interdiffusion in olivine can be used as a trustable chronometer in kimberlite systems, opening new perspectives on the study of ascent dynamics in kimberlites worldwide.

Mantle-Derived Cargo vs Liquid Line of Descent: Reconstructing the P–T–fO2–X Path of the Udachnaya–East Kimberlite Melts during Ascent in the Siberian Sub-Cratonic Lithosphere

Abstract Disentangling the physico-chemical evolution of the melts that give rise to kimberlites during their genesis, ascent through the sub-cratonic lithosphere and emplacement in the crust is challenging. This is because the extensive entrainment and assimilation of, and reaction with, mantle-derived material makes kimberlites mixtures of xenocrystic, magmatic components and alteration minerals, rarely preserving evidence of their original melt composition. Here, a detailed textural and compositional study of coherent and volcaniclastic kimberlite units from the Udachnaya–East pipe (Siberian craton) was performed to reconstruct the pressure–temperature–oxygen fugacity-compositional (P–T–fO2–X) path of kimberlite melts during their ascent through the sub-cratonic lithosphere. Routine and high-precision electron microprobe analyses of olivine, phlogopite and Fe–Ti oxides enabled to discriminate the mantle-derived cargo from the magmatic components, and thus make inferences on the liquid line of descent (LLD) of proto-kimberlite to kimberlite melts. Most of the olivine cores in Udachnaya–East kimberlite are xenocrystic and record conditions of last equilibration in the Siberian sub-cratonic lithospheric mantle at T–P ranging from 812–1227°C at 3.1–5.4 GPa, to 871–1170°C at 4.6–7.0 GPa, depending on the chosen model geotherm (35 mW/m2 vs 40 mW/m2). Based on their Ni, Cr, Mn, Al and P vs Mg/Fe systematics, olivine core populations were associated to the sheared, granular garnet-bearing or spinel-bearing cratonic peridotites. The occurrence of olivine Internal Zones (I) having the same composition as Fe-rich cores, as well as rare mantle-derived xenocrystic cores of phlogopite, bear witness of mantle metasomatism preceding kimberlite ascent. The assimilation of mantle material by initially H2O- and P-rich proto-kimberlitic melts is recorded by the cotectic formation of magmatic olivine Internal Zones (II) and phlogopite Internal Zones around resorbed xenocrystic cores. The LLD then evolved by inducing cotectic precipitation of olivine, phlogopite rims and Cr-spinel at P of 1.5–3.0 GPa, T of 1120–1250°C and fO2 from −2.8 to −1.6 ΔFMQ. Ilmenite–magnetite pairs in the groundmass record the later conditions of crystallization that occurred at P of ~1.0 GPa, T from 1133°C down to 1000°C and fO2 of +0.3 to +0.9 ΔFMQ. These results confirm the progressive oxidation of the melt that finally led to the formation of calcite, apatite and Mg-rich olivine and phlogopite rinds. The F-enrichment of phlogopite rinds reflects a late decrease of the H2O/CO2 activity of the melt that evolved toward alkali-carbonated composition. Our results showed that the crystal cargo of kimberlites can be used to track all processes acting in between the melt-rock reactions in the mantle and the emplacement in the crust.

An exotic Cretaceous kimberlite linked to metasomatized lithospheric mantle beneath the southwestern margin of the São Francisco Craton, Brazil

We present major and trace element compositions of mineral concentrates comprising garnet xenocrysts, ilmenite, phlogopite, spinel, zircon, and uncommon minerals (titanite, calzirtite, anatase, baddeleyite and pyrochlore) of a newly discovered Late Cretaceous kimberlite (U-Pb zircon age 90.0 ± 1.3 Ma; 2σ) named Osvaldo França 1, located in the Alto Paranaíba Igneous Province (APIP), southeastern Brazil. Pyrope grains are lherzolitic (Lherz-1, Lherz-2 and Lherz-3), harzburgitic (Harz-3) and wehrlitic (Wehr-2). The pyrope xenocrysts cover a wide mantle column in the subcratonic lithosphere (66–143 km; 20–43 kbar) at relatively low temperatures (811–875 °C). The shallowest part of this mantle is represented by Lherz-1 pyropes (20–32 kbar), which have low-Cr (Cr2O3 = 1.74–6.89 wt.%) and fractionated middle to heavy rare earth elements (MREE-HREE) pattern. The deepest samples are represented by Lherz-2, Lherz-3, Harz-3, and Wehr-2 pyropes (36–43 kbar). They contain high-Cr contents (Cr2O3 = 7.36–11.19 wt.%) and are characterized by sinusoidal (Lherz-2 and Wehr-2) and spoon-like (Lherz-3 and Harz-3) REE patterns. According to their REE and trace elements, pyrope xenocrysts have enriched nature (e.g., Ce and Yb vs. Cr2O3), indicating that the cratonic lithosphere has been affected by a silicate melt with subalkaline/tholeiite composition due to their low Zr, Ti and Y concentrations. Besides minerals with typical kimberlitic signatures, such as ilmenite and zircon, the exotic compositions of phlogopite and ulvöspinel suggest an enriched component in the magma source. The formation of rare mineral phases with strong enrichment of light-REE (LREE) and high field strength elements (HFSE) is attributed to the late-stage kimberlitic melt. We propose a tectonic model where a thermal anomaly, represented by the low-velocity seismic anomaly observed in P-wave seismic tomography images, supplied heat to activate the alkaline magmatism from a metasomatized cratonic mantle source during the late-stages of Gondwana fragmentation and consequent South Atlantic Ocean opening. The metasomatism recorded by mineral phases is a product of long-lived recycling of subducted oceanic plates since the Neoproterozoic (Brasiliano Orogeny) or even older collisional events, contributing to the exotic character of the Osvaldo França 1 kimberlite, as well as to the cratonic lithospheric mantle.

Subcratonic and tectonic evolution of pyroxenite and eclogite with lamellar inclusions in garnet, Western Gneiss Region, Norway

Abstract Oriented lamellar inclusions of pyroxene and rutile in mantle garnet often serve as evidence for majoritic and titaniferous precursor garnets, respectively. We investigated ten new such microstructure-bearing samples from six orogenic peridotite bodies in SW Norway, which originated in the E Greenland mantle lithosphere, petrologically and thermobarometrically. All pyroxenite (nine) and eclogite (one) samples have large (mainly porphyroclastic) garnet containing silicate and oxide inclusions with shape-preferred orientation relationship. These inclusions vary—dependent on their size—systematically in shape (acicular to subprismatic), width (∼50 μm to submicron size), spacing (several 100 to ∼10 μm) and phase (pyroxene to Ti-oxide ± pyroxene). Smaller inclusions can fill the space between larger inclusions, which support the idea of consecutive generations. The larger, early formed lamellae occur least frequent and are most poorly preserved. A younger generation of other inclusions decorates healed cracks cutting across cores but not rims of garnet. These inclusions comprise oxides, silicates, carbonates (aragonite, calcite, magnesite) and fluid components (N2, CO2, H2O). The older, homogeneously distributed inclusions comply texturally and stoichiometrically with an origin by exsolution from excess Si- and Ti-bearing garnet. Their microstructural systematic variation demonstrates a similar early evolution of pyroxenite and eclogite. The younger inclusions in planar structures are ascribed to a metasomatic environment that affected the subcratonic lithosphere. The microstructure-bearing garnets equilibrated at ∼3.7 GPa (840 °C) and ∼3.0 GPa (710 °C), at a cratonic geotherm related to 37–38 mW m−2 surface heat flow. Some associated porphyroclastic grains of Mg-rich pyroxene have exsolution lamellae of Ca-rich pyroxene and vice versa that indicate a preceding cooling event. Projected isobaric cooling paths intersect isopleths for excess Si in garnet at ∼1550 °C, if an internally consistent thermodynamic data set in the system Na2O–CaO–MgO–Al2O3–SiO2 (NCMAS) is applied (or ∼1600 °C if using CMAS). This temperature may confine the crystallisation of the unexsolved garnets at 100–120 km depths of the E Greenland subcratonic lithosphere. Tectonism is indicated in coastal and hinterland samples by porphyroclastic orthopyroxene with Al2O3 concentrations showing W-shaped profiles. Cores of associated large (>200 μm) recrystallised grains have low Al2O3 contents (0.18–0.23 wt.%). Both characteristics typify relatively short intracrystalline Al diffusion lengths and a prograde metamorphism into the diamond stability field. We assign this event to subduction during the Scandian orogeny. Porphyroclastic orthopyroxene in other samples shows U-shaped Al2O3 concentration profiles paired with long Al diffusion lengths (several 100 μm) that exceed the radius of recrystallised grains. Their cores contain high Al2O3 contents (0.65–1.16 wt.%), consistent with a diffusional overprint that obliterated prograde and peak metamorphic records. Unlike Al2O3, the CaO content in porphyroclastic orthopyroxene cores is uniform suggesting that early exhumation was subparallel to Ca isopleths in pressure–temperature space. The depth of sample origin implies that rock bodies of Scandian ultra-high pressure metamorphism occur in nearly the entire area betweenNordfjord and Storfjord and from the coast towards ∼100 km in the hinterland, i.e. in a region much larger than anticipated from crustal eclogite.

Composition of Reduced Mantle Fluids: Evidence from Modeling Experiments and Fluid Inclusions in Natural Diamond

Abstract —Experimental modeling in the C–O–H, C–O–H–N, and peridotite–C–O–H–N systems, combined with analyses of fluid inclusions in natural diamonds, is used to reconstruct the compositions of fluids that can be stable in the reduced mantle. Hydrocarbons (HCs) in the upper mantle can form either by reactions of carbonates with iron/wüstite and water or by direct hydrogenation of carbon phases (graphite, diamond, and amorphous carbon) interacting with reduced fluids. Carbon required for the formation of HCs can come from diamond, graphite, or carbonates. Mainly light alkanes are stable at the mantle pressures and temperatures in the C–O–H and C–O–H–N systems as well as in the peridotite–fluid system under ultrareduced to moderately reduced redox conditions at the oxygen fugacity from –2 to +2.5 lg units relative to the IW (Fe–FeO) buffer. Some oxygenated HCs can be stable in fluids equilibrated with carbonate-bearing peridotite. Ammonia and, to a lesser degree, methanimine (CH3N) are predominant nitrogen species in reduced fluids in the conditions of the subcratonic lithosphere or the Fe0-bearing mantle. The presence of HCs as common constituents of reduced mantle fluids is supported by data on inclusions from natural diamonds hosted by kimberlites of the Yakutian province and from placer diamonds of the northeastern Siberian craton and the Urals. Fluid inclusions have minor amounts of H2O, methane, and other light alkanes but relatively high concentrations of oxygenated hydrocarbons, while the H/(H + O) ratio varies from 0.74 to 0.93. Hydrocarbon-bearing fluids in some eclogitic diamonds have high CO2 concentrations. Also, the fluid inclusions have significant percentages of N2 and N-containing species, Cl-containing HCs, and S-containing compounds. Both the experimental results and the analyses of fluid inclusions in natural diamonds indicate that HCs are stable in the upper mantle conditions. The set of hydrocarbons, mainly light alkanes, might have formed in the mantle from inorganic substances. Further research should focus on the causes of the difference between experimental and natural fluids in the contents of methane, light alkanes, oxygenated hydrocarbons, and water and on the stability of N-, S-, and Cl-containing fluid components.

Phase Relations in the Harzburgite–Hydrous Carbonate Melt at 5.5–7.5 GPa and 1200–1350°С

Phase relations are studied experimentally in the harzburgite–hydrous carbonate melt system, the bulk composition of which represents primary kimberlite. Experiments were carried out at 5.5 and 7.5 GPa, 1200–1350°С, and $${{X}_{{{\text{C}}{{{\text{O}}}_{2}}}}}$$ = 0.39–0.57, and lasted 60 hours. It is established that olivine–orthopyroxene–garnet–magnesite–melt assemblage is stable within the entire range of the studied parameters. With increase of temperature and $${{X}_{{{\text{C}}{{{\text{O}}}_{2}}}}}$$ in the system, Ca# in the melt and the olivine fraction in the peridotite matrix significantly decrease. The composition of silicate phases in run products is close to those of high-temperature mantle peridotite. Analysis of obtained data suggest that magnesite at the base of subcontinental lithosphere could be derived by metasomatic alteration of peridotite by asthenospheric hydrous carbonate melts. The process is possible in the temperature range typical of heat flux of 40–45 mW/m2, which corresponds to the conditions of formation of the deepest peridotite xenoliths. Crystallization of magnesite during interaction with peridotite matrix can be considered as experimentally substantiated mechanism of CO2 accumulation in subcratonic lithosphere.

Fate of fluids at the base of subcratonic lithosphere: Experimental constraints at 5.5–7.8 GPa and 1150–1350 deg C

Phase relations in the lherzolite-C-O-H-N system are studied experimentally at 5.5–7.8 GPa, 1150–1350 deg C, and oxygen fugacity (fO2) from 2.5 log units below to 3.5 log units above the iron-wüstite (IW) equilibrium, in 10- to 150-h runs. The two-capsule technique is applied to maintain hydrogen fugacity (fH2) at the IW (Fe-FeO), MMO (Mo-MoO2), and HM (Fe2O3-Fe3O4) equilibria. The mineral assemblage stable in the Fe0- and graphite-saturated lherzolite-C-O-H-N system, at 6.3–7.8 GPa and 1200–1350 deg C, consists of olivine, orthopyroxene, clinopyroxene, and garnet. The metal phase occurs either as iron carbide (Fe3C) or iron nitride (Fe3N) at low and high concentrations of nitrogen, respectively. Carbide and nitride phases contain progressively more Ni (5–6 to 25 wt.%) as fluids become more hydrous. Fluids equilibrated with lherzolite consist of CH4 and C2H6, minor amounts of other alkanes, H2O, NH3, and methaneimine (CH3N). Fluids with high nitrogen contents are mainly composed of NH3, N2, light alkanes, and water. As fO2 increases, Fe3C and Fe3N oxidize and silicate phases (olivine, orthopyroxene, and garnet) contain more FeO, while fluids become richer in H2O and more depleted in hydrocarbons (HCs). Fluids synthesized at lithospheric P = 5.5–6.3 GPa, T = 1150–1200 deg C and fO2 near the carbon-saturated water maximum (CW) contain up to 85 rel.% H2O and at least 14 rel.% of total HCs. The presence of HCs stable to oxidation suppresses water activity in fluids while the solidus of the lherzolite-C-O-H-N system at fO2 near CW becomes at least 150 deg C higher than that of the carbon-free lherzolite-H2O system. Further oxidation of HCs and C0 leads to partial carbonation of olivine and orthopyroxene and their replacement by magnesite and clinopyroxene. The total content of HCs (mainly C2-C13 alkanes and oxygenated HCs) in fluids from magnesite-bearing lherzolite at 5.5 GPa and 1200 deg C is never below 9 rel.% even in 150-h runs, while CO2 does not exceed 3–5 rel.%. Melt inclusions composed of Mg and Ca carbonates in olivine record the first batches of carbonate melt at these P-T and redox conditions. The solidus of the lherzolite-C-O-H-N system at 5.5–7.8 GPa with fO2 between metal and carbonate saturation is above the typical upper mantle temperatures at a heat flux of 40 mW/m2. Thus, the experiments demonstrate that HC-H2O fluids can ascend from metal-saturated asthenosphere to more oxidized lithospheric mantle in a typical thermal regime and supply volatiles for carbonatite metasomatism and generation of carbonate-silicate melts.

Asthenospheric kimberlites: Volatile contents and bulk compositions at 7 GPa

During ascent, kimberlites react with the lithospheric mantle, entrain and assimilate xenolithic material, loose volatiles and suffer from syn- and post-magmatic alteration. Consequently, kimberlite rocks deviate heavily from their primary melt. Experiments at 7 GPa, 1300–1480 °C, 10–30 wt% CO2 and 0.46 wt% H2O on a proposed primitive composition from the Jericho kimberlite show that saturation with a lherzolitic mineral assemblage occurs only at 1300–1350 °C for a carbonatitic melt with <8 wt% SiO2 and >35 wt% CO2. At asthenospheric temperatures of >1400 °C, where the Jericho melt stays kimberlitic, this composition saturates only in low-Ca pyroxene, garnet and partly olivine.We hence forced the primitive Jericho kimberlite into multiple saturation with a lherzolitic assemblage by adding a compound peridotite. Saturation in olivine, low- and high-Ca pyroxene and garnet was obtained at 1400–1650 °C (7 GPa), melts are kimberlitic with 18–29 wt% SiO2 + Al2O3, 22.1–24.6 wt% MgO, 15–27 wt% CO2 and 0.4–7.1 wt% H2O; with a trade-off of H2O vs. CO2 and temperature. Melts in equilibrium with high-Ca pyroxene with typical mantle compositions have ≥2.5 wt% Na2O, much higher than the commonly proposed 0.1–0.2 wt%.The experiments allow for a model of kimberlite origin in the convective upper mantle, which only requires mantle upwelling that causes melting at the depth where elemental carbon (in metal, diamond or carbide) converts to CO2 (at ∼250 km). If primary melts leading to kimberlites contain a few wt% H2O, then adiabatic temperatures of 1400–1500 °C would yield asthenospheric mantle melts that are kimberlitic (>18 wt% SiO2 + Al2O3) but not carbonatitic (<10 wt% SiO2 + Al2O3) in composition, carbonatites only forming 100–200 °C below the adiabat. These kimberlites represent small melt fractions concentrating CO2 and H2O and then acquire part of their chemical signature by assimilation/fractionation during ascent in the subcratonic lithosphere.

Stability of methane in reduced C–O–H fluid at 6.3 GPa and 1300–1400°C

The composition of a reduced C–O–H fluid was studied by the method of chromatography–mass spectrometry under the conditions of 6.3 GPa, 1300–1400°C, and fO2 typical of the base of the subcratonic lithosphere. Fluids containing water (4.4–96.3 rel. %), methane (37.6–0.06 rel. %), and variable concentrations of ethane, propane, and butane were obtained in experiments. With increasing fO2, the proportion of the CH4/C2H6 peak areas on chromatograms first increases and then decreases, whereas the CH4/C3H8 and CH4/C4H10 ratios continually decrease. The new data show that ethane and heavier HCs may be more stable to oxidation, than previously thought. Therefore, when reduced fluids pass the “redox-front,” carbon is not completely released from the fluid and may be involved in diamond formation.

Fluid-induced transition from banded kyanite- to bimineralic eclogite and implications for the evolution of cratons

Heterogeneous, modally banded kyanite-bearing and bimineralic eclogites from the lithospheric mantle, collected at the Roberts Victor Diamond mine (South Africa), show a reaction texture in which kyanite is consumed. Geothermobarometric calculations using measured mineral compositions in Perple_X allowed the construction of a P-T path showing a steep, cool prograde metamorphic gradient of 2°C/km to reach peak conditions of 5.8GPa and 890°C for the kyanite eclogite. The kyanite-out reaction formed bimineralic eclogite and is probably an integral part of the mineralogical evolution of most archetypal bimineralic eclogites at Roberts Victor and potentially elsewhere. The kyanite-out reaction occured at close to peak pressure (5.3GPa) and was associated with a rise in temperature to 1380°C. Mass balance calculations show that upon breakdown, the kyanite component is fully accommodated in garnet and omphacite via a reaction system with low water fugacity that required restricted fluid influx from metasomatic sources. The δ18O values of garnets are consistently higher than normal mantle values. Each sample has its characteristic trend of δ18O variance between garnets in the kyanite-bearing sections and those in the bimineralic parts covering a range between 5.1‰ and 6.8‰. No systematic change in O-isotope signature exists across the sample population. Differences in garnet trace element signatures between differing lithologies in the eclogites are significant. Grossular-rich garnets coexisting with kyanite have strong positive Eu-anomalies and low Gd/Yb ratios, while more pyrope-rich garnets in the bimineralic sections have lost their positive Eu-anomaly, have higher Gd/Yb ratios and generally higher heavy rare earth element contents. Garnets in the original kyanite-bearing portions thus reflect the provenance of the rocks as metamorphosed gabbros/troctolites. The kyanite-out reaction was most likely triggered by a heating event in the subcratonic lithosphere. As kyanite contains around 100ppm of H2O it is suggested that the kyanite-out reaction, once initiated by heating and restricted metasomatic influx, was promoted by the release of water contained in the kyanite. The steep (high-P low-T) prograde P-T path defining rapid compression at low heating rates is atypical for subduction transport of eclogites into the lithospheric mantle. Such a trajectory is best explained in a model where strong lateral compression forces eclogites downward to higher pressures, supporting models of cratonic lithosphere formation by lateral collision and compression.

Composition of a carbonatitic melt in equilibrium with lherzolite at 5.5–6.3 GPa and 1350°C

Generation of ultra-alkaline melts by the interaction of lherzolite with cardonatites of various genesis was simulated at the P–T parameters typical of the base of the subcratonic lithosphere. Experiments with a duration of 150 h were performed at 5.5 and 6.3 GPa and 1350°C. The concentrations of CaO and MgO in melts are buffered by the phases of peridotite, and the concentrations of alkalis and FeO depend on the composition of the starting carbonatite. Melts are characterized by a low (<7 wt %) concentration of SiO2 and Ca# from 0.40 to 0.47. It is demonstrated that only high-Mg groups of carbonatitic inclusions in fibrous diamonds have a composition close to that of carbonatitic melts in equilibrium with lherzolite. Most likely, the formation of kimberlite-like melts relatively enriched in SiO2 requires an additional source of heat from mantle plumes and probably H2O fluid.

Composition of primary kimberlite magma: constraints from melting and diamond dissolution experiments

Experiments are applied to constrain the composition of primary kimberlitic magmas which were in equilibrium with lithospheric peridotite and could resorb the entrained diamond to form typical dissolution features. The experiments are run on samples of a model carbonatite and a melt of the Udachnaya kimberlite at 6.3 GPa and 1400 °C, and at unbuffered or Re–ReO2-buffered oxygen fugacity (1–2 log units above Ni–NiO). Near-liquidus dry Fe3+-free carbonatitic melt (derived from carbonated harzburgite) is saturated with the Ol–Grt–Opx–Mgs assemblage and is almost inert to diamond. Carbonatitic melts that bear 4.6–6.8 wt% Fe2O3 or 1.5 wt% H2O are in equilibrium only with Mgs ± Ol near the liquidus. Dissolution of diamond by these melts produces surface textures uncommon (corrosion sculptures) or common (negative-oriented trigons, shield-shaped laminae and elongate hillocks) to kimberlitic diamonds. The near-liquidus melt of the Udachnaya kimberlite (Yakutia) with 10–12 wt% H2O is saturated with the Ol–Grt–Cpx assemblage and may result from melting of carbonated garnet-bearing wehrlite. Hydrous kimberlitic melt likewise resorbs diamonds forming typical negative-oriented trigons, shield-shaped laminae and elongate hillocks on their surfaces. Therefore, the melts that could originate in the thermal conditions of subcratonic lithosphere, entrain diamond and dissolve it to produce dissolution features on crystal surfaces, were compositionally close to kimberlite (16–19 wt% SiO2) and rich in H2O. Dry Fe3+-bearing carbonatites with fO2 controlled by the ferric/ferrous equilibrium slightly above the Ni–NiO buffer cannot be diamond carriers.

Comparison of mantle lithosphere beneath early Triassic kimberlite fields in Siberian craton reconstructed from deep-seated xenocrysts

Mantle xenocrysts from early Triassic kimberlite pipes from Kharamai, Ary-Mastakh and Kuranakh fields in the Anabar shield of Siberia revealing similar compositional trends were studied to estimate the superplume influence on the subcratonic lithosphere mantle (SCLM). Pressure-temperature (PT) reconstructions using monomineral thermobarometry for 5 phases show division of the SCLM beneath the Kharamai field into 6 units: pyroxenitic Fe-rich (1–2 GPa) and Mg-rich (2–3 GPa) layers; middle with two levels of Gar-Sp pyroxenites at ∼3 and 4–5 GPa; Gar-dunite–harzburgites ∼4.5–6.5 GPa subjected to Ilm-Px vein metasomatism; and a Mg-rich dunite lower part. In the Anabar shield (Ary-Mastakh, Dyuken and Kuranakh fields) mantle lithosphere is composed of three large units divided into two parts: upper part with amphiboles and phlogopite; two levels of pyroxenites and eclogites at 3 and 4 GPa, and a lower part composed of refertilized dunites. Diagrams showing P-Fe#Gar clusters for garnets and omphacites illustrate the differences between SCLM of these localities. Differences of Triassic SCLM from Devonian SCLM are in simple layering; abundance of Na-Cr-amphiboles and metasomatism in the upper SCLM part, thick pyroxenite-eclogite layer and lower part depletion, heated from SCLM base to 5.0 GPa.Kharamai mantle clinopyroxenes represent three geochemical types: (1) harzburgitic with inclined linear REE, HFSE troughs and elevated Th, U; (2) lherzolitic or pyroxenitic with round TRE patterns and decreasing incompatible elements; (3) eclogitic with Eu troughs, Pb peak and high LILE content. Calculated parental melts for garnets with humped REE patterns suggest dissolution of former Cpx and depression means Cpx and garnets extraction. Clinopyroxenes from Ary-Mastakh fields show less inclined REE patterns with HMREE troughs and an increase of incompatible elements. Clinopyroxenes from Kuranakh field show flatter spoon-like REE patterns and peaks in Ba, U, Pb and Sr, similar to those in ophiolitic harzburgites. The PT diagrams for the mantle sections show high temperature gradients in the uppermost SCLM accompanied by an increase of P-Fe#Ol upward and slightly reduced thickness of the mantle keel of the Siberian craton, resulting from the influence of the Permian–Triassic superplume, but with no signs of delamination.

The composition of garnet as an indicator of the conditions of peridotite-carbonatite interaction in the subcratonic lithosphere (experimental data)

The composition of garnet as an indicator of the conditions of peridotite-carbonatite interaction in the subcratonic lithosphere (experimental data)

Conditions of phlogopite formation upon interaction of carbonate melts with peridotite of the subcratonic lithosphere

Conditions of phlogopite formation upon interaction of carbonate melts with peridotite of the subcratonic lithosphere

Phase relationships in the system K2CO3-CaCO3 at 6 GPa and 900-1450 C

Phase relations in the system K 2 CO 3 -CaCO 3 have been studied in the compositional range, X(K 2 CO 3 ), from 100 to 10 mol%, at 6.0 GPa and 900-1450 °C. At 900-950 °C, the system has three intermediate compounds: K 6 Ca 2 (CO 3 ) 5 , K 2 Ca(CO 3 ) 2 , and K 2 Ca 3 (CO 3 ) 4 . The K 2 Ca(CO 3 ) 2 compound decomposes to the K 6 Ca 2 (CO 3 ) 5 + K 2 Ca 3 (CO 3 ) 4 assembly above 950 °C. The K6Ca2(CO 3 ) 5 and K 2 Ca 3 (CO 3 ) 4 compounds melt congruently slightly above 1200 and 1300 °C, respectively. The eutectics were established at 64 and 44 mol% near 1200 °C and at 23 mol% near 1300 °C. K 2 CO 3 remains as a liquidus phase at 1300 °C and 75 mol% and melts at 1425 ± 20 °C. Aragonite remains as a liquidus phase at 1300 °C and 20 mol% and at 1400 °C and 10 mol%. CaCO 3 solubility in K 2 CO 3 and K 2 CO 3 solubility in aragonite are below the detection limit (&lt;0.5 mol%). Infiltration of subduction-derived K-rich Ca-Mg-Fe-carbonatite into the Fe 0 -saturated mantle causes the extraction of (Mg,Fe)CO 3 components from the melt, which shifts its composition toward K-Ca-carbonatite. According to our data this melt can be stable at the P-T conditions of subcratonic lithosphere with geothermal gradient of 40 mW/m 2 corresponding to temperature of 1200 °C at 6 GPa.

Temperature-dependent transient creep and dynamics of cratonic lithosphere

Large-scale mantle convection forms the upper boundary layer (lithosphere) where the vertical temperature drop is about 1300 K. Theoretical rheology and laboratory experiments with rock samples show that transient creep occurs while creep strains are sufficiently small. The transient creep is described by the temperature-dependent Andrade rheological model. Since plate tectonics allows only small deformations in lithospheric plates, creep of the lithosphere plates is transient whereas steady-state creep, described by non-Newtonian power-law rheological model, takes place in the underlying mantle. The solution of stability problem shows that the lithosphere is stable but small-scale convective oscillations are attenuated very weakly in regions of thickened lithosphere beneath continental cratons (subcratonic roots) where the thickness of the lithosphere is about 200 km. These oscillations create small-scale convective cells (the horizontal dimensions of the cells are of the order of the subcratonic lithosphere thickness). Direction of motion within the cells periodically changes (the period of convective oscillations is of the order of 3 × 108 yr). In this study, the oscillations of cratonic lithosphere caused by initial relief perturbation are considered. This relief perturbation is assumed to be created by overthrusting in orogenic belts surrounding cratons. The perturbation of the Earth's surface relief leads to a fast isothermal process of isostatic recovery. In the presence of vertical temperature gradient, vertical displacements, associated with the recovery process in the lithosphere interior, instantly produce the initial temperature perturbations exciting thermoconvective oscillations in the cratonic lithosphere. These small-amplitude convective oscillations cause oscillatory crustal movements which form sedimentary basins on cratons.

Transient creep and convective instability of the lithosphere

SUMMARY Laboratory experiments with rock samples show that transient creep, at which strain grows with time and strain rate decrease at constant stress, occurs while creep strains are sufficiently small. The transient creep at high temperatures is described by the Andrade rheological model. Since plate tectonics allows only small deformations in lithospheric plates, creep of the lithosphere plates is transient whereas steady-state creep, described by non-Newtonian power-law rheological model, takes place in the underlying mantle. At the transient creep, the effective viscosity, found in the study of postglacial flows, differs significantly from the effective viscosity, which characterizes convective flow, since timescales of these flows are very different. Besides, the transient creep changes the elastic crust thickness estimated within the power-law rheology of the lithosphere. Two problems of convective stability for the lithosphere with the Andrade rheology are solved. The solution of the first problem shows that the state, in which large-scale convective flow in the mantle occurs under lithospheric plates, is unstable and must bifurcate into another more stable state at which the lithospheric plates become mobile and plunge into the mantle at subduction zones. If the lithosphere had the power-law fluid rheology, the effective viscosity of the stagnant lithospheric plates would be extremely high and the state, in which large-scale convection occurs under the stagnant plates, would be stable that contradicts plate tectonics. The mantle convection forms mobile lithosphericplatesiftheeffectiveviscosityoftheplateisnottoomuchhigherthantheeffective viscosity of the underlying mantle. The Andrade rheology lowers the plate effective viscosity corresponding to the power-law fluid rheology and, thus, leads to instability of the state in which the plates are stagnant. The solution of the second stability problem shows that the state, in which the lithospheric plate moves as a whole with constant velocity, is stable but smallamplitude oscillations are imposed on this motion in regions of thickened lithosphere beneath continental cratons (subcratonic roots) where the thickness of the lithosphere is about 200 km. These oscillations create small-scale convective cells (the horizontal dimensions of the cells are of the order of the subcratonic lithosphere thickness). Direction of motion within the cells periodically changes (the period of oscillations is of the order of 10 8 yr). The small-amplitude convective oscillations cause small strains and do not destroy the thickening of the lithosphere beneath cratons. Thus, the transient creep of the lithosphere explains not only mobility of the lithospheric plates but longevity of subcratonic roots as well.