Successful management of hydrocarbon reservoirs, geothermal energy extraction sites, radioactive waste and CO2 storage sites depends on a detailed knowledge of fluid transport properties, porosity and permeability. Amongst deformation processes, fluid-rock interaction plays an important role in controlling the petrophysical properties of a rock. The presence of fluids in the rocks induce chemical and physical changes in compositions and texture, affecting porosity and permeability, hence influencing dynamic transport properties and fluid flow. Fluid-rock interaction processes have been deeply investigated in nature and in numerous experimental and numerical modelling studies. However, these studies lack a spatio-temporal characterization of the dynamic evolution of porosity and reaction microfabrics. There is no clear understanding of the spatio-temporal evolution of these properties in three dimensions, and how this evolution affects fluid percolation in the rock. Computed X-ray micro-Tomography (μCT) was applied to investigate these processes in three dimensions and observe their evolution in time (4DμCT). The combination of μCT with 2D analytical techniques (e.g. scanning electron microscope, SEM, electron microcrobe, EMPA, electron backscatter diffraction, EBSD) furthermore enables the extrapolation of the information gained from 2D analyses to the 3rd an 4th dimension (4D μCT). The thesis investigates two different categories of fluid-rock interaction processes, by using 4DμCT to monitor the evolution of mineral reactions (in the first case) and porosity (second case) in relation to strain and time. In the first case study, natural rock samples were analysed. The samples show a compositional change along a strain gradient from olivinic metagabbros to omphacite-garnet bearing eclogites in a ductile shear zone. Synchroton-based x-ray microtomography (sμCT) was applied to document the 3D evolution of garnets along the strain gradient (which represent the 4th dimension). The 3D spatial arrangement of garnet microfabrics can help determine the deformation history and the extent of fluid-rock interaction active during deformation. Results from the sμCT show that in the low strain domain, garnets form a large and well interconnected cluster that develops throughout the entire sample and garnet coronas never completely encapsulate olivine grains. In the most highly deformed eclogites, the oblate shapes of garnets reflect a deformational origin of the microfabrics. EBSD analyses reveal that garnets do not show evidence for crystal plasticity, but rather they highlight evidence for minor fracturing, neo-nucleation and overgrowth, which points to a mechanical disintegration of the garnet coronas during strain localisation. In the second case study, pressure-solution processes were investigated using NaCl as rock-analogue, to monitor the evolution of porosity and pore connectivity in four dimensions, providing a time-resolved characterization of the processes. NaCl samples were uniaxially compacted and μCT scans were taken at regular interval times to characterize the evolution of grain morphologies, pore space and macro-connectivity of the samples. Different uniaxial loads, as well as different bulk sample compositions (phyllosilicates and/or glass beads) were used to investigate their effect on the process. Greater uniaxial loads, and the presence of phyllosilicates within the deforming NaCl columns were found to enhance pressure-solution processes. The pore space becomes highly disconnected in the presence of phyllosilicates, with important implications for fluid percolation and dynamic transport properties. Mean strain rates, calculated from volumetric Digital Image Correlation (3D-DIC) analyses, were found to be higher where phyllosilicates were located. The combination of μCT with volumetric DIC and SEM imaging proved to be an efficient analytical method for investigating the dynamic behaviour of porosity and permeability during ongoing pressure-solution processes. The results showed that fluid-rock interaction critically modifies the rocks at the pore/grain scale, with important consequences on dynamic fluid transport properties. The combination of μCT with classical 2D techniques provided a better understanding on the dynamic evolution of transport properties and fluid percolation during fluid-rock interaction processes, allowing the characterization in three dimensions of reaction microfabrics and porosity.

Magister Scientiae - MSc (Earth Science) / The complexity of the Platreef stratigraphy and the generic position of the Ni-Cu-PGE mineralisation is a challenge to prospecting and mining companies in the Northern Limb of the Bushveld Complex, partially, as a result of various floor rock interactions with the reef. Therefore, this study evaluated the effects of contact metasomatic fluids on the nature and style of PGE mineralisation as the main event leading to the complexity of the Platreef stratigraphy from the contact zone near the floor rock. Fifty samples from boreholes MO009 and MO019 drilled at Moordkopje 813 LR farm for Akanani Project by Lonmin Plc were used for this study. The mineralogy and geochemistry of the Platreef samples were studied and associated with their mineralisation occurrences. Major, minor and trace element contents were analysed by XRF analysis using fused beads, and PGE contents (Pt, Pd) in 11 samples were determined by Fire Assay.

Contaminant release and transport in platinum tailings are poorly studied in literature. This study serves to characterise these processes. The tailings facility is located in Steelpoort, Mpumalanga, South Africa on Critical Zone rocks of the Rustenburg Layered Suite in The Bushveld Igneous Complex. Tailings material samples were collected by hand auger- and direct push probe drilling at specific locations to represent the different materials present in the tailings facility. Water samples were collected from monitoring boreholes as well as the Steelpoort River. The samples were analysed using XRD, XRF, ABA, NAG, Reflected Light Microscopy, Acid Leaching Tests and ICP scans. Using the collected data, a geochemical model was constructed for the interpretation of mineral phase dissolution and to trace the mineral phases releasing contaminants. Analysis-, test- and modelling results showed that alteration mineral phases formed within ten years in the tailings material and that the existing alteration phases viz. talc and chlorite, as well as sulphides, are the major contributors of contaminants. Elevated pH values as well as major cation and anion concentrations were found in the fluid discharging from the tailings as well as the in groundwater, with little to no heavy metals, which were traced directly to the chromite phase. This suggests that platinum tailings do not contribute to heavy metal contamination or acid rock drainage but may increase aquifer salinity and alkalinity. The flow through the tailings, underlying vadose zone and fractured rock aquifer was characterised using permeameter- and pumping tests. From the data collected, an unsaturated flow model was developed to characterise the flow through the tailings. From the model, discharge from the tailings was calculated to take place at 0.7m per decade into the underlying vadose zone with fracture flow in the aquifer ranging from 0.46-0.026m/d, as calculated from pumping test results. Contaminant migration into the Steelpoort River is possibly inhibited by the Dwarsriver Fault, based on the chemical data and hydraulic conductivities calculated. Therefore, groundwater is considered to be the major receptor in the system and groundwater users may be negatively impacted by increasing groundwater salinity and major ion concentrations. / Dissertation (MSc)--University of Pretoria, 2013. / Geology / unrestricted

Metamorphic hydrous, CO2-bearing fluids play a critical role in the global carbon cycle. However, how big this influence is on the global carbon cycle and therefore on global climatic processes, is unknown. The actual amount of CO2 which is released into the atmosphere due to metamorphic processes is still debated. For this purpose, fluid-driven reactions in metamorphic rocks must be studied by tracking fluid-rock interactions along pathways of ancient fluids. In the study presented in this thesis, we study fluid-rock interaction in the southeastern part of the Greek island Syros in the Cycladic Archipelago (Aegean). On Syros fluid-rock interaction is recorded by the preservation of blueschist facies assemblages at greenschist facies conditions along a normal shear zone. Blueschist preservation is caused by a combination of metasomatic addition of SiO2 and Na2O and elevated XCO2 which is maintained by high fluxes of a CO2-bearing, hydrous fluid along the shear zone. This research aims to provide a better understanding of the role of mountain building in the carbon cycle. Flux estimates for climate-forcing fluid components (e.g. carbon) require that their concentration in the fluid, fluid volumes and velocities are known. This will be the focus of future work. Further, whole rock chemistry and the availability of specific minerals will be studied to achieve knowledge about which kind of parameters influence and enhance the propagation of fluids through rocks.

京都大学 / 0048 / 新制・課程博士 / 博士(理学) / 甲第19511号 / 理博第4171号 / 新制||理||1599(附属図書館) / 32547 / 京都大学大学院理学研究科地球惑星科学専攻 / (主査)准教授 河上 哲生, 教授 平島 崇男, 教授 山 明 / 学位規則第4条第1項該当 / Doctor of Science / Kyoto University / DFAM

Fluids, mainly H2O and CO2, are released from H- and C-bearing phases during prograde metamorphism. Because of the buoyancy of these fluids, they rise within the crust towards the surface of the Earth. Metamorphic fluids take advantage of permeable horizons, shear zones, fold hinges, fractures, and are channelled into high-flux zones. Fluid fluxes for channelized fluid flow may exceed background pervasive fluxes by several orders of magnitude. Metamorphic fluids react with the surrounding rock during fluid flow, and altered zones are commonly observed adjacent to high-flux conduits. Fluid-altered rock is texturally, mineralogically, chemically, and isotopically different from rock unaffected by fluid flow. In this thesis, fluid-rock interaction is studied at two localities in the Scottish Highlands: Glen Esk and the Isle of Islay. Glen Esk is one of the type localities used by George Barrow (1853-1932) to propose the concept of metamorphic zones and metamorphic index minerals as an approximate determination of metamorphic grade. In several of the metamorphic zones in Glen Esk, index mineral distribution is highly dependent on proximity to veins. The occurrence of index minerals is therefore not only controlled by pressure and temperature, but also by the availability of metamorphic fluids. Evidence of a retrograde fluid flow event from the North Esk Fault is observed in Glen Esk, for which a time-averaged fluid flux of 0.0003 – 0.0126 m3∙m-2∙yr-1 is calculated. The duration of the fluid event is estimated to between 16 and 334 kyr. On the Isle of Islay, kyanite is observed in rocks of chlorite or lower-biotite metamorphic grade, i.e. much lower temperatures than usually associated with kyanite formation. The favoured explanation for this is retrograde infiltration of extremely high-CO2 fluids, at least locally XCO2 &gt; 0.7, at ~340°C, which altered these rocks and stabilised kyanite in a carbonate-bearing assemblage. Oxygen and carbon stable isotope profiles across the Islay Anticline reveals highly channelized fluid flow along the axial region of this fold, with fluid:rock ratios at least four times higher than in rock farther away from the fold. Although carbon and oxygen isotope ratios of metacarbonate rocks were altered along the Islay Anticline, negative anomalies observed below and above the Port Askaig Tillite Formation cannot solely be attributed to metamorphic fluid flow, which implies that these rocks to varying degree retain their primary paleoclimatological isotopic signatures. / Stora volymer H2O och CO2 frigörs som fluider under prograd metamorfos. Metamorfa fluider har lägre densitet än det omgivande berget, varför de stiger genom jordskorpan mot jordytan. Metamorfa fluider kanaliseras i permabla lager, skjuvzoner, veckaxlar, sprickor och andra högflödeszoner. Kanaliserade fluidflöden kan vara flera storleksordningar högre än bakgrundsvärdet för fluidflöde inom en bergart. Metamorfa fluider reagerar under transport med det omgivande berget och bildar fluidomvandlade zoner i anslutning till högflödeskanaler. Fluidomvandlat berg uppvisar texturella, mineralogiska, kemiska och isotopsammansättningsmässiga skillnader i jämförelse med berg som inte utsatts för fluidomvandling. I denna avhandling behandlas reaktioner mellan fluid och berg som studerats i två lokaler i de skotska högländerna: Glen Esk och Islay. Glen Esk är en av de typlokaler som George Barrow (1853-1932) använde för att lägga fram konceptet om metamorfa zoner och metamorfa indexmineral som används för att ungefärligt uppskatta metamorf grad. I flera av de metamorfa zonerna är förekomsten av indexmineral i hög grad beroende av närhet till kvartsådror, vilket visar att bildandet av indexmineral inte bara styrs av tryck och temperatur, utan också av åtkomst till metamorfa fluider. I Glen Esk finns också spår av ett fluidflöde från North Esk-förkastningen, under retrograda metamorfa förhållanden, för vilket mededfluidflödet över tid uppgår till 0.0003 – 0.0126 m3∙m-2∙år-1. Denna fluidflödeshändelse beräknas ha pågått mellan 16 000 och 334 000 år. På ön Islay i de sydvästra högländerna återfinns bergarter, som trots sin låga metamorfa grad i klorit- eller biotitzonen innehåller mineralet kyanit, dvs. temperaturer långt under vad som vanligen associeras med kyanitbildning. Detta förklaras med infiltration av fluider med extremt hög CO2-halt, åtminstone lokalt så högt som XCO2 &gt; 0.7, vid ca. 340°C. Fluidomvandling av dessa bergarter stabiliserade kyanit tillsammans med karbonatmineral. Syre- och kolisotopprofiler över Islayantiklinen påvisar hög kanalisering av fluider längs dess veckaxeln. Förhållandet mellan fluid och berg var mer än fyra gånger så högt i närheten av veckaxeln jämfört lokaler längre ifrån densamma. Påverkan av metakarbonatbergarters isotopförhållanden har skett längs Islayantiklinen, men fluidpåverkan kan inte ensamt förklara de isotopanomalier som observerats under och ovan Port Askaig-tilliten, varför dessa bergarter kan ha bibehållit sin primära paleoklimatologiska isotopsignatur. / <p>At the time of the doctoral defense, the following papers were unpublished and had a status as follows: Paper 2: Submitted. Paper 3: Manuscript.</p>

La monazite est présente comme minéral accessoire dans la plupart des environnements géologiques. Elle est souvent riche en U et Th, n’incorpore pas (ou peu) de Pb initial et son système isotopique (U-Th-Pb) est résistant à la diffusion dans la majorité des conditions crustales, ce qui en fait un chronomètre très attractif. De plus, elle constitue la principale source de Th et une des principales sources de terres rares (REE) contenus dans la croûte terrestre. En présence de fluide, la monazite peut recristalliser par un processus de dissolution-précipitation couplée, avec une composition chimique et/ou isotopique différente de la monazite initiale. Ces recristallisations impliquent alors une redistribution des éléments contenus par la monazite (REE, Th, U, Pb) et la compréhension de la mobilité de ces éléments stratégiques est cruciale pour leurs aspects économiques (ressources en lanthanides et actinides) et environnementaux (contexte de stockage des déchets).Cette thèse vise à étudier le comportement de la monazite lors des interactions fluide-monazite et son potentiel comme traceur chronomètre et traceur géochimique des minéralisations hydrothermales. Pour cela une approche pluridisciplinaire a été adoptée, combinant minéralogie, pétrologie expérimentale, géochronologie, et tectonique. Les travaux présentés s’organisent en deux parties: l’une concernant la datation de monazites hydrothermales de fentes alpines, et l’autre des expériences d’altérations hydrothermales en laboratoire.Une quarantaine de cristaux de monazite et une dizaine de cristaux de xénotime ont été collectés dans des fentes alpines (veines hydrothermales se formant durant l’exhumation) des domaines externes (Argentera, Belledonne et Mont-Blanc) et internes (zone Briançonnaise). La datation U-Th-Pb in-situ par LA-ICP-MS a permis de mieux contraindre l’âge et la durée des circulations hydrothermales pendant les épisodes de déformation tardifs liés à l’exhumation des Alpes occidentales. Les analyses d’inclusions fluides dans la monazite couplées à des âges traces de fission sur zircons ont apporté de nouvelles contraintes sur le gradient géothermique induit par les circulations fluides dans les veines hydrothermales. L’analyse systématique des produits expérimentaux (monazite et fluide) de 18 expériences hydrothermales a permis de confirmer la mobilité des éléments comme l’uranium ou les terres rares lourdes lors des réactions hydrothermales. L’étude à l’échelle nanométrique des domaines de monazite recristallisée a mis en évidence un nouveau mécanisme de remplacement caractérisé par la propagation du front de réaction à l’aide de nano-pores et nano-fractures. Ce mécanisme conduit à un remplacement anisotrope et à un mélange de nano-domaines de monazite primaire et recristallisée. Ces observations ont des implications majeures pour le stockage des déchets radioactifs ou en géochronologie pour expliquer les perturbations des âges monazites ayant réagi avec un fluide dans les milieux hydrothermaux ou métamorphiques. / Monazite is commonly found in most of geological environments. Monazite can be rich in uranium and thorium, does not incorporate lead, and its isotopic (U-Th-Pb) system is very robust to diffusion in most of crustal conditions, which makes it a very attractive chronometer. In addition, it represents the main source of thorium and a major source of rare earth elements (REE), in the crust. During fluid-monazite interaction, monazite can recrystallize by a coupled dissolution-precipitation process, with a chemical/isotopic composition different from the initial monazite. These recrystallizations involve a redistribution of the elements contained in monazite (REE, Th, U, Pb) and understanding of the mobility of these strategic elements is crucial for their economic (lanthanides resources and actinides) and environmental (storage of radioactive waste) aspects.The aim of this thesis is to investigate the behaviour of monazite during fluid-monazite interactions and its potential as chronometer and geochemical tracer of fluid mineralization, via a multi-disciplinary approach including mineralogy, experimental petrology, geochronology and tectonic. The work presented here is organized in two parts: one on Alpine hydrothermal monazite dating and the other on the results of hydrothermal alteration experiments in laboratory.About forty monazite and ten xenotime crystals were collected in Alpine clefts (hydrothermal veins formed during exhumation) of the external (Argentera, Belledonne, Mont-Blanc) and the internal (Briançonnais Zone) domains. The LA-ICP-MS in-situ U-Th-Pb dating allowed to better constrain the age and duration of hydrothermal circulation during the late deformation stages related to the exhumation of the western Alps. The fluid inclusion analysis of monazite crystals coupled with zircon fission-track dating have brought new constrains on the geothermal gradient induced by fluid circulations in the hydrothermal veins. Systematic analysis of experimental products (monazite and fluid) of 18 experiments confirmed the mobility of elements such as uranium or heavy REE during hydrothermal reactions. The nanoscale study of monazite recrystallized domains showed a new replacement mechanism characterized by the propagation of the reaction front through nano-pores and nano-fractures. This mechanism leads to anisotropic replacement and a mixture of nano-domains of primary and recrystallized monazite. These observations have major implications for the storage of radioactive waste or in geochronology to explain the disturbances of monazite ages that have reacted with fluid in hydrothermal or metamorphic environments.

Plagioclase is the most abundant mineral in the lower crust and it is thus important to constrain the behaviour of plagioclase during deformation. Anorthosites, which are plagioclase-rich rocks, are common in Archaean cratons but their origin and rheological importance is still debated. The aims of this thesis are to 1) describe a newly discovered Archaean anorthosite complex (Naajat Kuuat, SW Greenland), investigate its origin and a possible genetic relationship between the anorthosite and associated mafic-ultramafic rocks and 2) to study the rheology and deformation mechanisms in plagioclase-rich rocks. The main focus of this thesis is on the deformation studies. (1) Geochemical whole-rock analyses from the Naajat Kuuat complex are indicative for an origin near a subduction zone setting. A genetic link by crystal fractionation between the anorthosite and associated mafic-ultramafic units is inferred. (2) Deformation behaviour of plagioclase is assessed from analyses of three anorthosite units deformed during different conditions. Samples were analysed using the electron backscatter diffraction technique (EBSD) in combination with optical and chemical analyses. All three case studies show significant strain localisation related to grain size reduction. A wet anorthosite deformed at dry conditions (T ~675-700°C) was dynamically recrystallised. Continuous bands of recrystallised grains developed a texture yet display microstructures and grain relationships indicative for grain size sensitive creep, suggesting that the rheology followed a Newtonian flow law. In the other two studies, samples with initially dry and wet composition, respectively, have experienced deformation during fluid present conditions at T ~550-620°C. These two samples show that fluids effectively caused reactions, replacements and aided strain localisation during deformation at mid crustal conditions. / At the time of the doctoral defence the following papers were unpublished and had a status as follows: Paper 1: Manuscript; Paper 2 Manuscript; Paper 3 Manuscript.

This thesis aims to provide a better understanding of the role of mountain building in the carbon cycle. The amount of CO2 released into the atmosphere due to metamorphic processes is largely unknown. To constrain the quantity of CO2 released, fluid-driven reactions in metamorphic rocks can be studied by tracking fluid-rock interactions along ancient fluid flow pathways. The thesis is divided into two parts: 1) modeling of fluid flow rates and durations within shear zones and fractures during greenschist- and blueschist-facies metamorphism and 2) the assessment of possible mechanisms of fluid infiltration into rocks during greenschist- to epidote-amphibolite-facies metamorphism and controlling chemical and mineralogical factors of reaction front propagation. On the island Syros, Greece, fluid-rock interaction was examined along a shear zone and within brittle fractures to calculate fluid flux rates, flow velocities and durations. Petrological, geochemical and thermodynamic evidence show that the flux of CO2-bearing fluids along the shear zone was 100-2000 times larger than the fluid flux in the surrounding rocks. The time-averaged fluid flow velocity and flow duration along brittle fractures was calculated by using a governing equation for one-dimensional transport (advection and diffusion) and field-based parameterization. This study shows that fluid flow along fractures on Syros was rapid and short lived. Mechanisms and controlling factors of fluid infiltration were studied in greenschist- to epidote-amphibolite-facies metabasalts in SW Scotland. Fluid infiltration into metabasaltic sills was unassisted by deformation and occurred along grain boundaries of hydrous minerals (e.g. amphibole) while other minerals (e.g. quartz) prevent fluid infiltration. Petrological, mineralogical and chemical studies of the sills show that the availability of reactant minerals and mechanical factors, e.g. volume change in epidote, are primary controls of reaction front propagation. / <p>At the time of the doctoral defense, the following papers were unpublished and had a status as follows: Paper 2: Manuscript. Paper 3: Manuscript. Paper 4: Manuscript.</p><p> </p>

Les processus affectant le coin mantellique situé au-dessus d’une zone de subduction (déformation et interaction avec les fluides/liquides magmatiques libérés par la plaque inférieure) ont des implications importantes sur la dynamique de la subduction et le budget géochimique global de la Terre.Afin de mieux contraindre ces processus, ma thèse a porté sur l’étude de l’unité rubanée de l’ophiolite du Semail. Cette unité de 200-500m d’épaisseur s’est (dé)formée, juste au-dessus de l’interface interplaque, pendant l’initiation de la subduction ou du chevauchement intra-océanique (qui a mené, à terme, à l’obduction de l’ophiolite). Elle est en effet située au-dessus de la semelle métamorphique HT (amphibolites à granulites ; 750-850°C et 0.9-1.1GPa) interprétée comme des écailles de la plaque inférieure métamorphisées et sous-plaquées à la plaque supérieure (actuelle ophiolite) lors des premiers stades de la convergence.Après une caractérisation structurale de terrain de cette unité rubanée et la collecte de plus de200 échantillons tout le long de l’ophiolite, j’ai mené une analyse intégrée (Microscopie optique, MEB, microsonde, EBSD, (LA-)ICPMS) sur une sélection d’entre-eux, afin de caractériser l’évolution P-T, pétrologique, géochimique et structurale des péridotites de l’unité rubanée pendant cet épisode de déformation.Les résultats montrent que cette déformation a mené à la formation de zones de cisaillement (proto)mylonitiques (~850-750°C) puis ultramylonitiques (~750-650°C) et que cette déformation BT s’est faite en décompression (d’~3kbar, i.e. 10km). Les résultats pétrologiques indiquent que des fluides silicatés hydratés ont percolé à travers (et intéragi avec) ces péridotites pendant cette déformation. Ces processus d’interaction ont mené à (1) la précipitation de minéraux métasomatiques (Ol+Opx+Cpx+Spl+Amp±Sulf), et (2) l’enrichissement des phases en éléments mobiles dans les fluides (surtout B, Li et Cs :concentrations de 1 à 40 fois celles du manteau primitif).L’analyse des isotopes du bore (δ11B des péridotites métasomatisées jusqu’à +25‰) démontre que ces fluides ont une signature de subduction et qu’ils sont vraisemblablement issus de la déshydratation de la semelle HT lors de sa formation à 750-850°C.En combinant ces résultats avec une analyse microstructurale, j’ai ensuite étudié les mécanismes et les rétroactions entre la circulation de ces fluides, la déformation des péridotites et la localisation de cette déformation. A l’échelle macroscopique, on observe une focalisation des fluides dans les zones de cisaillement actives, associée à une localisation progressive de la déformation.Nous avons aussi exploré les conséquences rhéologiques de l’hydratation de ce manteau sur le régime, couplé ou découplé, de l’interface. Les lois rhéologiques indiquent que l’affaiblissement des (proto)mylonites de l’unité rubanée, par hydratation, peut expliquer le couplage de l’interface de subduction à 850-750°C et, ainsi, l’accrétion de la semelle HT. Nous interprétons les zones de cisaillement ultramylonitiques ultérieures (~750-650°C) comme étant liées au stade d’exhumation simultanée de l’unité rubanée et de la semelle HT d’environ 10km au-dessus de l’interface, jusqu’à leur position actuelle sous l’ophiolite.Les résultats de cette étude suggèrent donc que l’interface semelle HT/unité rubanée représente une interface de subduction fossilisée et le manteau (proto)mylonitique sus-jacent, un coin mantellique qui s’est (dé)formé et a intéragi avec des fluides de subduction pendant l’initiation de la subduction. L’unité rubanée de l’ophiolite du Semail représente donc un des rares objets géologiques permettant d’étudier les processus à l’oeuvre dans un coin mantellique, et d’en traquer, avant sa fossilisation, les transformations mécaniques et chimiques sur ~1 million d’années. / The processes affecting the mantle wedge atop a subduction zone (deformation and interaction with fluids/melts released by the downgoing plate) play a major role on subduction zones dynamics and the global geochemical budget of the Earth.To better constrain these processes, my Ph.D. research project has focused on studying the basal banded unit of the Semail ophiolite. This 200-500m thick peridotitic basal unit was (de)formed, directly above the interplate interface, during the intra-oceanic subduction (or underthrusting) initiation (that ultimately led to the ophiolite obduction). The banded unit indeed overlies the HT metamorphic sole (amphibolites to granulites ; 750-850°C and 0.9-1.1GPa) interpreted as slices of the downgoing plate underplated to the upper plate (the ophiolite) during early subduction (or subduction "infancy").After a field-based structural characterization of this banded unit and more than 200 samples collected all along the strike of the ophiolite, I carried out an integrated analysis (Optical microscopy, SEM, microprobe, EBSD, (LA-)ICPMS) on selected samples, in order to constrain the P-T, petrological, geochemical and structural evolution of the banded unit peridotites during this deformation event.Our results show that this deformation led to the formation of (proto)mylonitic (at ~850-750°C) then ultramylonitic (at ~750-650°C) shear zones and that this deformation was associated with peridotites decompression (of ~3kbar, i.e. 10km). Petrological results suggest that hydrated silicate fluids have percolated through (and interacted with) these peridotites during their deformation. These interaction processes triggered (1) the precipitation of metasomatic minerals (Ol+Opx+Cpx+Spl+Amp±Sulf), and (2) the enrichment of phases in fluid mobile elements (parti- cularly B, Li and Cs;concentrations from 1 to 40 times higher than those of the primitive mantle).The analysis of boron isotopes (δ11B of metasomatized peridotites up to +25‰) demonstrated that these fluids had a "subduction signature" and that they presumably derived from HT sole dehydration while forming at 850-750°C.By combining these results with microstructural analyses, I then studied the mechanisms and feedbacks between the circulation of these fluids, peridotites ductile deformation and strain localization. At the macroscopic scale, we observe a focusing of fluids in actively deforming peridotites associated to progressive strain localization during peridotites cooling.We also investigated the rheological consequences of banded unit peridotites hydration on the regime (coupled or decoupled) of the interface. Rheological laws indicate that the hydration-related weakening of banded unit (proto)mylonites is able to explain the coupling of the subduction inter- face at 850-750°C and, thereby, HT sole slicing and accretion. We interpret the later development of the ultramylonitic shear zones (at ~750-650°C) as being associated to the subsequent exhumation stage, i.e. the coeval exhumation of the banded unit and the HT metamorphic sole over around10km along the interface, up to their present-day position under the ophiolite.The results of this work suggest that the HT sole/banded unit contact represents a fossilized subduction interface and the overlying (proto)mylonitic mantle, a frozen-in mantle wedge that was (de)formed and interacted with subduction fluids during subduction infancy. The Semail ophiolite banded unit therefore provides a rare glimpse of processes affecting a mantle wedge, and enables tracking its mechanical and geochemical transformations over 1My (prior to its fossilization).The processes highlighted in this Ph.D. research project thus bring new constraints on the (petrological-geochemical-rheological) consequences of mantle wedge peridotites interaction with subduction fluids.