Ultrabasic magmas and high-degree melting of the mantle

As the degree of melting of mantle peridotite increases, the liquids that are formed become more basic and less viscous, and the spacing between residual crystals increases. The settling velocities of residual crystals in partial melts consequently will increase by several orders of magnitude, from 9.4 × 10^–4 cm/s to 4.3 × 10^–1 cm/s for a 1 cm olivine grain, as the proportion of liquid increases from 15 to 60%.To produce an ultrabasic komatiitic magma from a source with commonly assumed mantle composition requires 50 to 80% melting. Before this degree of melting can be reached, a highly fluid picritic magma produced by 30 to 50% melting will segregate from the source. Ultrabasic magmas probably form by a sequential melting process and are derived from a residuum composed of refractory minerals and trapped liquid left by previous episodes of partial melting and magma extraction. Trace element concentrations in ultrabasic komatiite lavas are consistent with this theory.     

Subsolidus and melting phase relations of basaltic composition in the uppermostlower mantle

The phase relations and the element partitioning in a mid-oceanic ridge basalt composition were determined for both above-solidus and subsolidus conditions at 22 to 27.5 GPa by means of a multianvil apparatus. The mineral assemblage at the solidus changes remarkably with pressure; majorite and stishovite at 22 GPa, joined by Ca-perovskite at 23 GPa, further joined by CaAl₄Si₂O₁₁-rich CAS phase at 25.5 GPa, and Mg-perovskite, stishovite, Ca-perovskite, CF phase (approximately on the join NaAlSiO₄-MgAl₂O₄), and NAL phase ([Na,K,Ca]₁[Mg,Fe²⁺]₂[Al,Fe³⁺,Si]_5.5-6.0O₁₂) above 27 GPa. The liquidus phase is Ca-perovskite, and stishovite, a CAS phase, a NAL phase, Mg-perovskite, and a CF phase appear with decreasing temperature at 27.5 GPa. Partial melt at 27 to 27.5 GPa is significantly depleted in SiO₂ and CaO and enriched in FeO and MgO compared with those formed at lower pressures, reflecting the narrow stability of (Fe,Mg)-rich phases (majorite or Mg-perovskite) above solidus temperature. The basaltic composition has a lower melting temperature than the peridotitic composition at high pressures except at 13 to 18 GPa (Yasuda et al., 1994) and therefore can preferentially melt in the Earth’s interior. Recycled basaltic crusts were possibly included in hot Archean plumes, and they might have melted in the uppermost lower mantle. In this case, Ca-perovskite plays a dominant role in the trace element partitioning between melt and solid. This contrasts remarkably with the case of partial melting of a peridotitic composition in which magnesiowüstite is the liquidus phase at this depth.     

Electrical conductivity of hydrous basaltic melts: implications for partial melting in the upper mantle

The Earth’s uppermost asthenosphere is generally associated with low seismic wave velocity and high electrical conductivity. The electrical conductivity anomalies observed from magnetotelluric studies have been attributed to the hydration of mantle minerals, traces of carbonatite melt, or silicate melts. We report the electrical conductivity of both H₂O-bearing (0–6 wt% H₂O) and CO₂-bearing (0.5 wt% CO₂) basaltic melts at 2 GPa and 1,473–1,923 K measured using impedance spectroscopy in a piston-cylinder apparatus. CO₂ hardly affects conductivity at such a concentration level. The effect of water on the conductivity of basaltic melt is markedly larger than inferred from previous measurements on silicate melts of different composition. The conductivity of basaltic melts with more than 6 wt% of water approaches the values for carbonatites. Our data are reproduced within a factor of 1.1 by the equation log σ = 2.172 − (860.82 − 204.46 w ^0.5)/(T − 1146.8), where σ is the electrical conductivity in S/m, T is the temperature in K, and w is the H₂O content in wt%. We show that in a mantle with 125 ppm water and for a bulk water partition coefficient of 0.006 between minerals and melt, 2 vol% of melt will account for the observed electrical conductivity in the seismic low-velocity zone. However, for plausible higher water contents, stronger water partitioning into the melt or melt segregation in tube-like structures, even less than 1 vol% of hydrous melt, may be sufficient to produce the observed conductivity. We also show that ~1 vol% of hydrous melts are likely to be stable in the low-velocity zone, if the uncertainties in mantle water contents, in water partition coefficients, and in the effect of water on the melting point of peridotite are properly considered.     

The trapped fluid phase in upper mantle xenoliths from Victoria,Australia: implications for mantle metasomatism

Mantle-derived xenoliths of spinel lherzolite, spinel pyroxenite, garnet pyroxenite and wehrlite from Bullenmerri and Gnotuk maars, southwestern Victoria, Australia contain up to 3 vol.% of fluids trapped at high pressures. The fluid-filled cavities range in size from fluid inclusions (1–100 m) up to vugs 11/2 cm across, lined with euhedral high-pressure phases. The larger cavities form an integral part of the mosaic microstructure. Microthermometry and Raman laser microprobe analysis show that the fluids are dominantly CO₂. Small isolated inclusions may have densities 1.19 g/cm³, but most inclusions show microstructural evidence of partial decrepitation during eruption, and these have lower fluid densities. Mass-spectrometric analysis of gases released by crushing or heating shows the presence of He, N₂, Ar, H₂S, CO_s and SO₂ in small quantities; these may explain the small freezing-point depressions observed in some inclusions. Petrographic, SEM and microprobe studies show that the trapped fluids have reacted with the cavity walls (in clinopyroxene grains) to produce secondary amphiboles and carbonates. The trapped CO₂ thus represents only a small residual proportion of an original volatile phase, which has undergone at least two stages of modification — first by equilibration with spinel lherzolite to form amphibole (±mica±apatite), then by limited reaction with the walls of the fluid inclusions. The inferred original fluid was a CO₂-H₂O mixture, with significant contents of (at least) Cl and sulfur species. Generation of this fluid phase in the garnet-peridotite stability field, followed by its migration to the spinel peridotite stability field, would provide an efficient mechanism for metasomatic enrichment of the upper mantle in LIL elements. This migration could involve either a volatile flux or transport in small volumes of silicate melt that crystallize in the spinel peridotite field. These observations suggest that some portions of the subcontinental upper mantle contain large reservoirs of free fluid CO₂, which may be liberated during episodes of rifting or magmatism, to induce granulite-facies metamorphism of the lower crust.     

Permo–Triassic intraplate magmatism and rifting in Eurasia: implications for mantle plumes and mantle dynamics

At the transition from the Permian to the Triassic, Eurasia was the site of voluminous flood-basalt extrusion and rifting. Major flood-basalt provinces occur in the Tunguska, Taymyr, Kuznetsk, Verkhoyansk–Vilyuy and Pechora areas, as well as in the South Chinese Emeishen area. Contemporaneous rift systems developed in the West Siberian, South Kara Sea and Pyasina–Khatanga areas, on the Scythian platform and in the West European and Arctic–North Atlantic domain. At the Permo–Triassic transition, major extensional stresses affected apparently Eurasia, and possibly also Pangea, as evidenced by the development of new rift systems. Contemporaneous flood-basalt activity, inducing a global environmental crisis, is interpreted as related to the impingement of major mantle plumes on the base of the Eurasian lithosphere. Moreover, the Permo–Triassic transition coincided with a period of regional uplift and erosion and a low-stand in sea level. Permo–Triassic rifting and mantle plume activity occurred together with a major reorganization of plate boundaries and plate kinematics that marked the transition from the assembly of Pangea to its break-up. This plate reorganization was possibly associated with a reorganization of the global mantle convection system. On the base of the geological record, we recognize short-lived and long-lived plumes with a duration of magmatic activity of some 10–20 million years and 100–150 million years, respectively. The Permo–Triassic Siberian and Emeishan flood-basalt provinces are good examples of “short-lived” plumes, which contrast with such “long lived” plumes as those of Iceland and Hawaii. The global record indicates that mantle plume activity occurred episodically. Purely empirical considerations indicate that times of major mantle plume activity are associated with periods of global mantle convection reorganization during which thermally driven mantle convection is not fully able to facilitate the necessary heat transfer from the core of the Earth to its surface. In this respect, we distinguish between two geodynamically different scenarios for major plume activity. The major Permo–Triassic plume event followed the assembly Pangea and the detachment of deep-seated subduction slabs from the lithosphere. The Early–Middle Cretaceous major plume event, as well as the terminal–Cretaceous–Paleocene plume event, followed a sharp acceleration of global sea-floor spreading rates and the insertion of new subduction zone slabs deep into the mantle. We conclude that global plate kinematics, driven by mantle convection, have a bearing on the development of major mantle plumes and, to a degree, also on the pattern of related flood-basalt magmatism.     

Anhydrous partial melting of an iron-rich mantle I: subsolidus phase assemblages and partial melting phase relations at 10 to 30 kbar

Anhydrous partial melting experiments, at 10 to 30 kbar from solidus to near liquidus temperature, have been performed on an iron-rich martian mantle composition, DW. The DW subsolidus assemblage from 5 kbar to at least 24 kbar is a spinel lherzolite. At 25 kbar garnet is stable at the solidus along with spinel. The clinopyroxene stable on the DW solidus at and above 10 kbar is a pigeonitic clinopyroxene. Pigeonitic clinopyroxene is the first phase to melt out of the spinel lherzolite assemblage at less than 20°C above the solidus. Spinel melts out of the assemblage about 50°C above the solidus followed by a 150° to 200°C temperature interval where melts are in equilibrium with orthopyroxene and olivine. The temperature interval over which pigeonitic clinopyroxene melts out of an iron-rich spinel lherzolite assemblage is smaller than the temperature interval over which augite melts out of an iron-poor spinel lherzolite assemblage. The dominant solidus assemblage in the source regions of the Tharsis plateau, and for a large percentage of the martian mantle, is a spinel lherzolite.     

Phase relations in the phlogopite-apatite system at 20 kbar; implications for the role of fluorine in mantle melting

Phase relations have been determined at 20 kb in the simple, Fe-, Ti-free systems hydroxyphlogopite-hydroxyapatite and hydroxyfluorphlogopite-hydroxyfluorapatite in order to determine distribution of fluorine between phlogopite, apatite and melt under mantle conditions. No excess H₂O was present in the hydroxyphlogopite-hydroxyapatite system and the F/(OH) ratio was unity in the F-bearing system. Both systems are pseudobinary and contain forsterite at phlogopite-rich compositions. In the F-absent system, the minimum melting occurs at 1225°C and Phl₈₅Ap₁₅, whereas in the F-bearing system this temperature is 1260°C and Phl₆₆Ap₃₄. Phlogopite in the F-absent system has lower Al than in the F-bearing system with both showing Si+Mg=^[IV]Al+^[VI]Al as the principal substitution. Increase in CaO in forsterite increases with increasing apatite in the bulk composition and is more pronounced in the F-absent system. Distribution of fluorine between phlogopite and liquid and apatite and liquid shows that D _F ^(Phl/glass) ranges from 2–1.25 depending on temperature and bulk composition, whereas the D _F ^(Apat/glass) is about unity. These results suggest that fluorine will tend to remain in the solid phases rather than the melt during partial melting in the mantle. Hence the enrichment of fluorine in ultrapotassic magmas and its role in their evolution are constrained.     

The stability of sapphirine + clinopyroxene: implications for phase relations in the CaO-MgO-Al2O3-SiO2 system under deep-crustal and upper mantle conditions

Reaction textures and chemographic relations in sapphirine-bearing basic granulites at Finero, Italy, suggest that sapphirine and aluminous diopside were formed in mutual equilibrium from an inferred early spinel-bearing assemblage. Finero appears to be the only known locality where this association has been found in situ, although it is also known from kimberlite and breccia pipe nodules elsewhere. The reactions deduced to have occurred in these rocks suggest the existence of stable invariant points involving the phases sapphirine-spinel-orthopyroxene-clinopyroxene-garnet-anorthite and sapphirine-two pyroxenes-garnetanorthite-kyanite (or sillimanite) in the CMAS end-member system. P-T estimates for the relevant rocks, and the available experimental data, suggest that these points lie at around 800°–900° C, 9–11 kbar. A semi-quantitative petrogenetic grid, incorporating these invariant points with previously determined univariant reactions, is proposed. It is inferred that sapphirine+diopside are stable relative to spinel-bearing assemblages below 900° C. The relatively low temperature explains why sapphirine has not to date been reported from experimental work on the CMAS system. It also suggests that sapphirine may be an important aluminous phase in Mg-rich metagabbros under conditions corresponding to the base of the continental crust, despite the observed rarity of such rocks at the surface.     

Phase relations of a natural MARID composition and implications for MARID genesis,lithospheric melting and mantle metasomatism

Melting experiments were performed on a natural mica-amphibole-rutile-ilmenite-clinopyroxene (MARID) sample from the Kaapvaal mantle lithosphere (AJE137) at 20 to 35 kbar and 800 to 1450°C. A solidus was determined at 1260°C and 30 kbar above which phlogopite, clinopyroxene and olivine were stable with an alkali-rich silicate melt. Olivine is the only crystallizing phase just below the liquidus of the AJE137 bulk composition and K-richterite was only stable in the subsolidus region ( 1100°C at 30 kbar). These results are consistent with previous studies in more simple systems. In experiments with 10 wt% added water the solidus was depressed by ca. 300°C and K-richterite was stabilized above this solidus. MARIDs represent a potential lowtemperature component in the lithospheric mantle beneath the Kaapvaal Craton of southern Africa. The addition of > 10 wt% water (with less than a 120°C rise of temperature above the geotherm) to this mantle region would create conditions for the melting of this component. This may then be incorporated in any continental flood basalt parent magma that traverse this lithospheric mantle. The derivation of MARIDs from a silicate melt of their bulk composition, even if water saturated, is considered unlikely as such small degree melts could not sustain the elevated liquidus temperatures required (> 1200°C at 30 kbar) in a cold (< 800°C at 30 kbar) mantle lithosphere. MARID xenoliths may be produced by the interaction of an alkali-rich fluid with a peridotite or as the residue to a group II kimberlitic parent magma that has undergone fractionation of olivine and the exsolution of a carbonatite component.     

Oxygen fugacity and geochemical variations in the martian basalts: implications for martian basalt petrogenesis and the oxidation state of the upper mantle of Mars

The oxygen fugacity of the Dar al Gani 476 martian basalt is determined to be quartz-fayalite-magnetite (QFM) −2.3 ± 0.4 through analysis of olivine, low-Ca pyroxene, and Cr-spinel and is in good agreement with revised results from Fe-Ti oxides that yield QFM −2.5 ± 0.7. This estimate falls within the range of oxygen fugacity for the other martian basalts, QFM −3 to QFM −1. Oxygen fugacity in martian basalts correlates with ⁸⁷Sr/⁸⁶Sr, ¹⁴³Nd/¹⁴⁴Nd, and La/Yb ratios, indicating that the mantle source of the basalts is reduced and that assimilation of crust-like material controls the oxygen fugacity. This allows constraints to be placed on the oxidation state of the martian mantle and on the nature of assimilated crustal material. The assimilated material may be the product of early and extensive hydrothermal alteration of the martian crust, or it may be amphibole- or phlogopite-bearing basaltic rock within the crust. In either case, water may play a significant role in the oxidation of basaltic magmas on Mars, although it may be secondary to assimilation of ferric iron-rich material.     

Experimental melting of crustal xenoliths from Kilbourne Hole,New Mexico and implications for the contamination and genesis of magmas

Experiments (P=6.9 kb; T=900–1000°C) on four crustal xenoliths from Kilbourne Hole demonstrate the varying melting behavior of relatively dry crustal lithologies in the region. Granodioritic gneisses (samples KH-8 and KH-11) yield little melt (<5–25%) by 925°C, but undergo extensive (30–50%) melting between 950 and 1000°C. A dioritic charnockite (KH-9) begins to melt, with the consumption of all modal K-feldspar, by 900°C. It is as fertile a melt source as the granodiorites at lower temperatures, but is outstripped in melt production by the granodiorite gneisses at high temperature, yielding only 26% melt by 1000°C. A pelitic granulite (KH-12) proved to be refractory (confirming earlier predictions based on geochemistry) and did not yield significant melt even at 1000°C. All melts have the composition of metaluminous to slightly peraluminous granites and are unlikely to be individually recognizable as magma contaminants on the basis of major element chemistry. However, the relative stability of K-feldspar during partial melting will produce recognizable signatures in Ba, Eu, K/Ba, and Ba/Rb. Melts of KH-11, which retains substantial K-feldspar throughout the melting interval, are generally low in Ba (<500–800 ppm), have high K/Ba and low Ba/Rb (est.) (62–124 and 1–3, respectively). Melts of KH-9, in which all K-feldspar disappears with the onset of melting, are Ba-rich [2000–2600 ppm, K/Ba=16–22; Ba/Rb (est.) =25–47]. Melts of KH-8 have variable Ba contents; <500 ppm Ba at low temperature but >900 ppm Ba in high-temperature melts coexisting with a K-feldspar-free restite. Although REE were not measured in either feldspar or melt, the high Kspar/melt Kds for Eu suggests that the melts coexisting with K-feldspar will have strong negative Eu anomalies. Isotopic and trace element models for magma contamination need to take into account the melting behavior of isotopic reservoirs. For example, the most radiogenic (and incompatible element-rich) sample examined here (the pelitic granulite,⁸⁷Sr/⁸⁶Sr=0.757) is refractory, while samples with far less radiogenic Sr (⁸⁷Sr/⁸⁶Sr=0.708-0.732) produced substantial melt. This suggests that, in this area, the isotopic signature of contamination may be more subtle than expected. The experimental results can be used to model the petrogenesis of Oligocene volcanic rocks exposed 150 km to the NW of Kilbourne Hole, in the Black Range in the Mogollon-Datil volcanic field. The experimental results suggest that a crustal melting origin for the Kneeling Nun and Caballo Blanco Tuffs is unlikely, even though such an interpretation is permitted by Sr isotopes. Curstal contamination of a mantle-derived magma best explains the chemical and isotopic characteristics of these tuffs. Both experimental and geochemical data suggest that the rhyolites of Moccasin John Canyon and Diamond Creek could represent direct melts of granodiorite basement similar, but not identical, to the Kilbourne Hole granodiorites, perhaps slightly modified by crystal fractionation. The absence of volcanic rocks having⁸⁷Sr/⁸⁶Sr>0.74 in the region is consistent with the refractory character of the pelitic granulite.     

Possible implications of modal mineralogy for melting in mantle lherzolites

The melting reaction at the solidus of mantle peridotite is commonly peritectic in nature, with liquid and one or more solid phases produced upon melting. In some situations, one of the phases participating on the reactant side of the reaction is present in low abundance. This article explores the possible effects of the low abundance of a reactant phase on the melting behavior of mantle peridotite.For example, spinel lherzolite begins to melt via the peritectic reaction, clinopyroxene + orthopyroxene + spinel = olivine + liquid in the ∼1- to 2-GPa pressure range. In natural spinel lherzolites, spinel is a modally minor mineral and may be infrequently in contact with both clinopyroxene and orthopyroxene. If these mutual contacts are insufficient to generate an interconnected melt, then significant melting may not occur until a combination of minerals that are modally abundant and in contact begin to melt. This scenario could have implications for the physical process of melting and for the timing of formation of an interconnected melt network and separation of the melt from the residue.To begin to investigate this possibility, the spatial relationships between the constituent minerals in two fertile spinel lherzolites were determined by elemental mapping with the electron microprobe. Olivine, orthopyroxene, and clinopyroxene are of similar size, whereas the spinel was smaller and interstitial. Spinel and clinopyroxene are frequently in contact, but mutual contacts of spinel, clinopyroxene, and orthopyroxene are rare. Because of the changes in modal mineralogy anticipated for these lherzolites with increasing temperature, these mutual contacts will be even less common at the solidus. Therefore, an interconnected, potentially extractable, melt may not occur by the solidus spinel + orthopyroxene + clinopyroxene melting reaction.     

Some implications of xenolith glasses for the mantle sources of alkaline mafic magmas

The assumption that mafic alkaline magmas are derived from mantle sources with a lherzolite mineralogy has become entrenched in the petrologic literature. Although it is commonly assumed that highly alkaline magmas require metasomatised mantle sources, there is little understanding of the spatial relation of such sources with respect to those of associated more Si-rich transitional magmas. Glasses developed in mantle xenoliths represent natural experiments which may provide some insight on this problem. Highly silica undersaturated glasses developed in the amphibole-garnet clinopyroxenite portion of a composite xenolith from Nunivak Island, Alaska, become quartz normative where they penetrate adjacent spinel lherzolite. A comparison of glass compositions in mantle pyroxenite and lherzolite xenoliths reveals that glasses developed in amphibole pyroxenite xenoliths are in general more silica undersaturated than those in lherzolite xenoliths. This suggests that some highly silica undersaturated magmas such as nephelinites may in fact be derived by the preferential melting of amphibole or amphibole-garnet pyroxenite veins and that the spectrum from nephelinite to transitional alkaline basalt that characterizes many individual alkaline volcanic suites is produced by mixing with melt derived from the host lherzolite as the degree of partial melting increases.     

Contrasted melting relations in a pyrolite upper mantle under mid-oceanic ridge,stable crust and island arc environments

The pyrolite model composition provides a satisfactory source composition for mantle-derived magmas insofar as major elements and “compatible” trace elements are concerned but there is evidence for mantle inhomogeneity in the abundances of “incompatible” minor and trace elements (e.g., K, Ti, P, Rb, Sr, light rare earth elements etc.). The composition of a magma, assuming a constant source composition, varies according to the pressure, temperature and water pressure or water content of the source region. The latter two variables essentially determine the degree of partial melting of the source region and in considering the chemical composition of the melt and nature of the residual phases, this parameter is of prime importance.For high degrees (> 20% approx.) of partial melting of a pyrolite source region, magmas are of tholeiitic character but are of increasingly undersaturated and alkaline type for lower degrees of partial melting and high pressures. For any chosen degree of melting and fixed water content of the source region, magmas are more olivine-rich at higher pressures. For any chosen pressure and chosen degree of partial melting, magmas are less olivine-rich at high water contents (and thus lower temperatures). Quartz tholeiite magmas may be derived by ~ 30% melting of pyrolite under water-saturated conditions at pressures up to between 17 kbar and 20 kbar. These generalizations may be applied to understand the characteristic magmatism of mid-oceanic ridges, island chains, oceanic islands and orogenic regions.     

Single-crystal elastic properties of chondrodite: implications for water in the upper mantle

Chondrodite, a member of the humite group of minerals, forms by hydration of olivine and is stable over a range of temperatures and pressures that includes a portion of the uppermost mantle. We have measured the single crystal elastic properties of a natural chondrodite specimen at ambient conditions using Brillouin spectroscopy. The isotropic aggregate bulk (K) and shear (μ) moduli calculated from the single-crystal elastic moduli, C_ij, are: K_S=118.4(16) GPa and μ=75.6(7) GPa. A comparison of the structures and elasticity of olivine and chondrodite indicate that the replacement of O with (OH,F) in M²⁺O₆ octahedra has a small effect on the elasticity of humite-group minerals. The slightly diminished elastic moduli of humite-group minerals (as compared to olivine) are likely caused by a smaller ratio of strong structural elements (SiO₄ tetrahedra) to weaker octahedra, and perhaps a more flexible geometry of edge-sharing MO₄(O,OH,F)₂ octahedra. In contrast to the humite-olivine group minerals, the incorporation of water into garnets and spineloids leads to a more substantial decrease in the elastic properties of these minerals. This contrasting behavior is due to formation of O₄H₄ tetrahedra and vacant hydroxyl-bearing octahedra in the garnets and spineloids, respectively. Therefore, the mechanism of incorporation of H/OH into mineral phases, not only degree of hydration, should be taken into account when estimating the effect of water on the elastic properties of minerals. The bulk elastic wave velocities of chondrodite and olivine are very similar. If humite-like incorporation of OH is predominant in the upper mantle, then the reaction of OH with olivine will have a minor or possibly no detectable effect on seismic velocities. Thus, it may be difficult to distinguish chondrodite-bearing rocks from “anhydrous” mantle on the basis of seismically determined velocities for the Earth. Received: 25 February 1998 / Revised, accepted: 18 August 1998     

MnNCKFMASH phase relations in cordierite-orthopyroxene migmatitic gneisses, southern india: implications for low-pressure crustal melting under granulite-facies

Cordierite-orthopyroxene migmatitic gneisses exposed in Achankovil unit of the Kerala Khondalite Belt, southern India show evidences of melting, melt extraction and in-situ crystallization of melt under granulite-facies conditions. The sequential mineral assemblages garnet + biotite + orthopyroxene + plagioclase + quartz (± melt) in the mesosomes and garnet + biotite + orthopyroxene + cordierite + plagioclase + K-feldspar + quartz + melt in the melanosomes makes the Achankovil cordierite-orthopyroxene migmatitic gneisses a good example of anatectic rocks, where substantial melt fractions remained in-situ during decompression and cooling. Therefore, the rocks provide an opportunity to investigate deep crustal processes and record of rheological (thermal and mechanical) reequilibration prevailed during the final stages of orogeny. The significance of cordierite formation and its possible relationship with melt formation are investigated applying theoretical calculations in the MnO-Na₂O-CaO-K₂O-FeO-MgO-Al₂O₃-SiO₂-H₂O (MnNCKFMASH) system. Results of numerical modelling of the mineral assemblages in pressure-temperature-composition (P-T-X) pseudosections using Perple_X infer that the sequence of reactions involving formation of cordierite-orthopyroxene-melt assemblage is consistent with an isothermal decompression (with a pressure drop of >1.5 kbars) at high temperatures (>800 °C), forming leucosomes. Biotite dehydration melting reactions, occurring above 4.5 kbars constrain prograde arm of the P-T trajectory and is interpreted as a product of crustal thickening, which was followed by rapid decompression. The final stage of exhumation is characterized by rehydration of cordierites in the melanosome by melt-solid interactions at exceptionally low-pressure (??3.2 kbars) conditions. The high-temperature isothermal decompression inferred from the mineral reactions and P-T-X pseudosections constitute a clockwise P-T path for the exhumation of the lower crust. This clockwise P-T path is consistent with the common tectonic model accepted for the genesis of granulite-facies migmatites during crustal thickening and later unroofing, accompanied with arc-continent collision. Our conclusions indicate low-P metamorphism and anatexis can be traced to convergent setting, where melt buoyancy considerably decreases density of the lithosphere and modifies rheology leading to rapid exhumation of the lower crust. Therefore, the crustal evolution in the Kerala Khondalite Belt is correlated with two stage processes: (i) thickening of the crust in relation to a continental-arc setting, followed by (ii) exhumation along a high-temperature stable geotherm with sufficient pressure release associated with syn- to post-convergence transpression and transtension.     

Ferropericlase—a lower mantle phase in the upper mantle

Experiments on compositions along the join MgO–NaA³⁺Si₂O₆ (A=Al, Cr, Fe³⁺) show that sodium can be incorporated into ferropericlase at upper mantle pressures in amounts commonly found in natural diamond inclusions. These results, combined with the observed mineral parageneses of several diamond inclusion suites, establish firmly that ferropericlase exists in the upper mantle in regions with low silica activity. Such regions may be carbonated dunite or stalled and degassed carbonatitic melts. Ferropericlase as an inclusion in diamond on its own is not indicative of a lower mantle origin or of a deep mantle plume. Coexisting phases have to be taken into consideration to decide on the depth of origin. The composition of olivine will indicate an origin from the upper mantle or border of the transition zone to the lower mantle and whether it coexisted with ferropericlase in the upper mantle or as ringwoodite. The narrow and flat three phase loop at the border transition zone—lower mantle together with hybrid peridotite plus eclogite/sediments provides an explanation for the varying and Fe-rich nature of the diamond inclusion suite from Sao Luiz, Brazil.     

Storage of F and Cl in the upper mantle: geochemical implications

Electron microprobe analyses yielded mean values of $\text{F}\text{0.43}\text{and}\text{Cl}\text{0.08}\text{wt.}\text{\%}$ for primary-textured phlogopites in coarse, depleted garnet-lherzolite xenoliths from kimberlites. Most secondary-textured phlogopites have too low Cl (0.01–0.08 wt.%) to be metamorphic precursors of primary-textured phlogopites. MARID-suite phlogopites and many megacrysts in kimberlites have low Cl ($\text{～ 0.02}\text{wt.}\text{\%}$), and some but not necessarily all secondary micas may result from infiltration of kimberlite into peridotite xenoliths. A good correlation between P and F in some oceanic basalts and gabbros might suggest that these elements are derived mainly from F-rich apatite inthe mantle, and that whitlockite is not present in the source region. Mantle-derived mica and amphibole have such low Cl that it is necessary to attribute Cl in oceanic basalts and gabbros either to substantial Cl in the source apatite, or to Cl from invading solutions, or both: three apatites from the mantle contain 0.8–1.0 wt.% Cl, and others contain lower amounts. The halogen contents of kimberlitic magmas can be explained by incorporation of Cl-bearing mica and F-rich apatite during melting of peridotites, but compositional constraints are weak.     

Trace elements in ocean ridge basalt glasses: implications for fractionations during mantle evolution and petrogenesis

Seven well-documented and fresh glassy selvages from ocean floor basalt pillows were analyzed by radiochemical neutron activation analysis for Ag, Au, Bi, Br, Cd, Cs, Ge, In, Ir, Ni, Os, Pd, Rb, Re, Sb, Se, Te, Tl, U and Zn. The samples came from active spreading centers in the Indian and Atlantic Ocean. Glasses from DSDP Leg 24, site 238 (Indian Ocean) have a somewhat peculiar trace element pattern, but this is thought to reflect secondary processes operating at shallow depth, not an anomalous source region in the mantle. Our data rather indicate that heterogeneities in the mantle are confined to the highly incompatible lithophile elements.Chemical fractionations during petrogenesis of tholeiitic basalts are discussed in the light of literature data for primitive peridotitic upper mantle nodules. (Ir, Os), Au, Pd, Ni and Re are strongly fractionated from each other in igneous processes; the unfractionated chondritic mantle pattern thus imposes firm constraints on mantle evolution models. The potentially chalcophile elements Ag, Cd, In and Zn do not behave differently from lithophile elements of the same valency and comparable ionic radius. Residual sulfides are not abundant enough to efficiently control the partitioning of these elements during basalt petrogenesis. However, the poor coherence of Tl to Rb and U in ocean floor basalts could point to retention of Tl by residual sulfides during depletion of the MORB source regions. Sb is strongly depleted in the source regions of ocean ridge basalts; most likely, it was present as a highly incompatible Sb⁵⁺ cation. The limited Rb/Cs fractionation in oceanic tholeiites, as opposed to continental tholeiites and acidic rocks, appears to reflect the low abundance of volatile constituents and hydrous silicates in normal ocean ridge basalts.