No description available.

The oceanic lithosphere forms as a residue of partial melting of the mantle beneath the mid-ocean ridge axis. Subduction of this residual layer has a profound impact on the Earth’s thermal and geochemical cycles as the recycling of this layer facilitates heat loss from the Earth’s interior and induces geochemical heterogeneities in the mantle. The goal of this study is to understand the thermal and geochemical evolution of the oceanic lithospheric mantle from a petrological perspective. An empirical geobarometer is calibrated for ocean island xenoliths in order to understand the thermal structure of the oceanic lithospheric mantle. The results of 0.1 MPa experiments from this study and high-pressure experiments from previous studies are used in the calibration. The uncertainties on pressures derived using the above geobarometer are high and hence could not be tested against thermal models for the oceanic lithosphere. The geochemical evolution of the oceanic lithospheric mantle involves post-melting geochemical modifications such as metasomatism. The geochemical evolution of the uppermost oceanic lithospheric mantle is studied using harzburgites from Hess Deep ODP Site 895, which are depleted in moderately incompatible elements relative to the global suite of abyssal peridotites. A comparison between Yb-abundances in Hess Deep harzburgites iii iv and those of a model depleted MORB mantle (DMM) residue reveals that the harzburgites have undergone up to 25% melting, assuming 0.5% melt porosity. Higher light and middle rare earth elements in the Hess Deep harzburgites than the model DMM melting residue are interpreted as the result of plagioclase crystallisation from melts being extracted by diffuse porous flow through the upper mantle. The effect of plagioclase crystallisation does not affect the chemistry of residual mineral phases as evidenced from the depleted light rare earth element abundances in clinopyroxene relative to the bulk rock. Ocean island xenoliths are studied to understand when and where metasomatism occurs in the deeper portion of the oceanic lithosphere. The median values of measured and reconstructed bulk concentration of Al2O3 in most ocean island xenoliths is lower than in abyssal peridotites, which generally would be interpreted as indicating a higher extent of melting in the former. However, a comparison between Yb- abundances in ocean island xenoliths and abyssal peridotites with a model DMM melting residue suggests that the extents of melting in the suites of rocks are broadly similar. Although fewer in number than ocean island xenoliths, abyssal peridotites from several locations have low concentrations of moderately incompatible elements. Metasomatism is observed in both, ocean island xenoliths and abyssal peridotites in the form of higher bulk rock Ce and Nd concentration than the model DMM melting residue but the extent of metasomatism is higher in ocean island xenoliths. There is no correlation between the concentrations of bulk rock Ce, Nd, Sm and Eu of ocean island xenoliths and age of the oceanic lithosphere from which the xenoliths originate. It is interpreted that metasomatism in the lower oceanic lithospheric mantle occurs near the ridge axis above the wings of the melting column. / Graduate / 0996 / 0372

Au cours de la dernière phase d’accrétion, les planètes terrestres ont connu des impacts géants violents et très énergétiques. A la suite du chauffage causé par les impacts, la Terre primitive était partiellement ou totalement fondue, et un océan magmatique a été formé dans la couche externe de la Terre. Le refroidissement successif de l’océan magmatique a causé la cristallisation fractionnée du manteau primitif. Cependant, il reste beaucoup d’incertitudes à propos de l’accrétion de la Terre primitive, comme la profondeur et la durée de vie d’un (ou plusieurs) océan(s) magmatique(s), l’effet de la recristallisation du manteau sur la ségrégation chimique entre les différents réservoirs de la Terre et ainsi de suite. La connaissance des propriétés de fusion du manteau profond est important aussi pour examiner la possibilité d’une fusion partielle actuellement. L’objectif était d’aborder quelques problèmes concernant le manteau inférieur terrestre : Quelle est la séquence de fusion entre les phases dominantes dans le manteau inférieur ? Est-ce qu’on peut expliquer la zone à ultra-basse vélocité (ULVZ) avec la fusion partielle d’un manteau pyrolytique (ou chondritique) ? Quel est le partage du fer entre les phases silicatées liquides et solides dans le manteau profond ? Est-ce qu’on peut donner des informations nouvelles sur les propriétés d’un océan magmatique profond à partir des courbes de fusion du manteau primitif ? Dans cette étude les courbes de fusion et les relations de fusion ont été analysées en utilisant la cellule à enclume de diamant chauffé au laser (LH-DAC) pour des pressions entre 25 et 135 GPa et des températures jusqu’à plus que 4000 K, i.e. pour des conditions de P-T qui correspondent au manteau inférieur terrestre entier. Les compositions utilisées ont été le raccord entre MgO et MgSiO3 et une composition de type chondritique pour le manteau terrestre. J’ai utilisé deux techniques in-situ de radiation-synchrotron pour déduire les propriétés de fusion à hautes pressions ; la diffractométrie au rayons-X et la fluorescence au rayons-X. Les nouveaux résultats obtenus dans cette étude sont : (...) / During the final stage of accretion, terrestrial planets experienced violent and highly energetic giant impacts. As a consequence of impact heating, the early Earth was partially or wholly molten, forming a magma ocean in the outer layer of Earth. Subsequent cooling of the magma ocean has led to fractional crystallization of the primitive mantle. Many unknowns remain about accretion of the early Earth, such as extension depth and life time of the magma ocean(s), role of mantle recrystallization on the chemical segregation between the different Earth reservoirs, and so on. The knowledge of melting properties of the deep mantle is also important to investigate the possibility of partial melting at the present time. The aim of this study was to tackle a few major questions concerning the Earth lower mantle : What is the melting sequence between the main lower mantle phases ? Can we explain the ultra-low-velocity zones (ULVZ) by partial melting of pyrolitic (or chondritic) mantle ? How does iron partition between liquid and solid silicate phases in the deep mantle ? Can we provide new information on the properties of the deep magma ocean based on the melting curve of the primitive mantle ? Melting curves and melting relations have been investigated using the laser-heated diamond anvil cell (LH-DAC) for pressure between 25 and 135 GPa and temperature up more than 4000 K, i.e. at P-T conditions corresponding to the entire Earth’s lower mantle. Compositions investigated were the join between MgO and MgSiO3 and a model chondritic-composition for the Earth mantle. Two different in situ synchrotron radiation techniques have been used to infer melting properties at high pressures ; X-ray diffraction and X-ray fluorescence spectroscopy. The new results obtained in this study include : (...)

The Station Creek Igneous Complex (SCIC) is one of the largest Middle-Late Triassic plutonic bodies in the northern New England Orogen of Eastern Australia. The igneous complex comprises of five plutons - the Woonga Granodiorite (237 Ma), Woolooga Granodiorite (234 Ma), Rush Creek Granodiorites (231 Ma) and Gibraltar Quartz Monzodiorite and Mount Mucki Diorite (227 Ma respectively), emplaced as high-level or epizonal bodies within the Devonian-Carboniferous subduction complex that resulted from a westward subduction along the east Australian margin. Composition of the SCIC ranges from monzogabbro to monzogranite, and includes diorite, monzodiorite, quartz monzodiorite and granodiorite. The SCIC has the typical I-type granitoid mineralogy, geochemistry and isotopic compositions. Its geochemistry is characteristics of continental arc magma, and has a depleted-upper mantle signature with up to 14 wt% supracrustal components (87Sr/86Srinitial = 0.70312 to 0.70391; Nd = +1.35 to +4.9; high CaO, Sr, MgO; and low Ni, Cr, Ba, Rb, Zr, Nb, Ga and Y). The SCIC (SiO2 47%-76%) has similar Nd and Sr isotopic values to island-arc and continentalised island-arc basalts, which suggests major involvement of upper mantle sourced melts in its petrogenesis. SCIC comprises of two geochemical groups - the Woolooga-Rush Greek Granodiorite group (W-RC) and the Mount Mucki Diorite-Gibraltar Quartz Monzodiorite group (MMD-GQM). The W-RC Group is high-potassium, calc-alkalic and metaluminous, whereas the MMD-GQM Group is medium to high potassium, transitional calc-alkalic to tholeiitic and metaluminous. The two geochemical groups of the SCIC magmas are generated from at least two distinct sources - an isotopically evolved Neoproterozoic mantle-derived source with greater supracrustal component (10-14 wt%), and an isotopically primitive mafic source with upper mantle affinity. Petrogenetic modeling using both major and trace elements established that the variations within respective geochemical group resulted from fractional crystallisation of clinopyroxene, amphibole and plagioclase from mafic magma, and late fractionation of alkalic and albitic plagioclase in the more evolved magma. Volcanic rocks associated with SCIC are the North Arm Volcanics (232 Ma), and the Neara Volcanics (241-242 Ma) of the Toogoolawah Group. The major and trace element geochemistry of the North Arm Volcanics is similar to the SCIC, suggesting possible co-magmatic relationship between the SCIC and the volcanic rock. The age of the North Arm Volcanics matches the age of the fractionated Rush Creek Granodiorite, and xenoliths of the pluton are found within epiclastic flows of the volcanic unit. The Neara Volcanics (87Sr/86Sr= 0.70152-0.70330, 143Nd/144Nd = 0.51253-0.51259) differs isotopically from the SCIC, indicating a source region within the HIMU mantle reservoir (commonly associated with contaminated upper mantle by altered oceanic crust). The Neara Volcanics is not co-magmatic to the SCIC and is derived from partial melting upper-mantle with additional components from the subducting oceanic plate. The high levels emplacement of an isotopically primitive mantle-derived magma of the SCIC suggest periods of extension during the waning stage of convergence associated with the Hunter Bowen Orogeny in the northern New England Orogen. The geochemical change between 237 to 227 Ma from a depleted-mantle source with diminishing crustal components, to depleted-mantle fractionate, reflects a fundamental change in the source region that can be related to the tectonic styles. The decreasing amount of supracrustal component suggests either thinning of the subduction complex due to crustal attenuation, leading to the late Triassic extension that enables mantle melts to reach subcrustal levels.