No description available.

Subduction zone volcanoes are a major natural hazard, frequently endangering lives and livelihoods. The eruptive history of many subduction zone volcanoes includes the extrusion of blocky, silicic lava that forms domes or flows, but we do not yet understand what determines the duration of dome-forming behaviour, what path magma may take to the surface, or how these systems may change over time. This thesis presents an investigation of the Santiaguito complex of lava domes and flows in Guatemala, which has been erupting continuously since its inception in 1922. The Santiaguito lavas are predominantly dacitic to andesitic, with a gradual reduction in SiO2 content from ~66 wt% in the 1920s, to ~62 wt% in 2002. This is consistent with a ~15% decrease in the extent of fractional crystallization over that time. The compositions of plagioclase phenocryst cores indicate a diminished role for magma mixing after the 1940s. I model the Santiaguito system as progressively extracting magma from an extensive, chemically-stratified storage zone. Petrological data are consistent with a storage zone extending from ~25 to ~12 km depth, and magma storage temperatures of ~940 to ~980°C. Phenocryst-hosted apatites suggest melt in this storage zone contained 401 to 1199 ppm S, 600 to 1300 ppm F, and 4100 to 6200 ppm Cl. Ascending magma may pass slowly through a conduit bottleneck, or plug, at shallow depths; groundmass texture suggests that melt rigidifies at or near the base this plug. Pre-eruptive melt volatile concentrations suggest time-averaged fluxes of 40 to 263 Mg d-1 SO2, 32 to 145 Mg d-1 HF, and 247 to 708 Mg d-1 HCl, giving ratios of 0.6 to 0.8 HF/SO2, and 2.7 to 6.2 HCl/SO2. These results are consistent with the few direct measurements of SO2 at Santiaguito, and with measured halogen emissions from other silicic dome-forming systems.

The marine silicon and carbon cycles are intrinsically linked by a unique group of primary producers; the diatoms. These siliceous-walled phytoplankton play a significant role in carbon export, making them a critical component of the global biological carbon pump with the power to affect climatic change. In this thesis, the silicon isotope composition (δ30Si) preserved in diatom opal is used together with the δ30Si of sponge opal, a powerful new proxy for deepwater silicic acid concentration, to document the Cenozoic Silicon cycle, shedding light on its role in carbon cycling and global climatic change. This study has developed a novel size-separation methodology to produce the first core top calibration of diatom δ30Si. The calibration demonstrates that diatom δ30Si exhibits a strong negative correlation with surface water silicic acid concentration, supportive of its application as a proxy for silicic acid utilisation. The refined method is used to produce a diatom δ30Si record, for the first time combined with sponge δ30Si, to gain insight into the Southern Ocean silicon cycle over one of the largest Cenozoic climatic shifts; the onset of Antarctic glaciation (~33.7 Ma). The two δ30Si records yield the first geochemical footprint to demonstrate that diatom proliferation, coincident with the onset of Antarctic circumpolar current flow, was a precursor event to the Eocene-Oligocene Transition. Diatoms are shown to have played a role in climate cooling through enhanced export and burial of organic carbon on the seafloor. The first long term reconstruction of silicic acid concentration in subsurface waters of the Southern Ocean, which spans the Late Eocene to the earliest Pliocene, provides new evidence that oceanic vertical mixing rates, coupled with the efficient removal of silicon from the surface by the diatoms, have been the most important factor in controlling the silicon chemistry of the ocean over the Cenozoic.

Isolement et caracterisation d'un materiau organique actuel d'origine algaire et resultats d'experiences de fixation "in vitro" de diverses especes metalliques cationiques et anioniques sur ce materiau. resultats d'experiences de maturation thermique conduites sous atmosphere inerte, des divers composes organo-metalliques formes par reaction entre la matiere organique et les especes metalliques cationiques

Silicon (Si) isotopes have been extensively studied in low temperature environments but the science of Si isotopes in igneous material has been comparatively ignored. This is because the degree of isotopic fractionation at high temperatures is relatively small, making the accurate measurement of these variations extremely challenging. Using state-of-the-art analytical techniques and instrumentation, which deliver high levels of precision, the objective of this research is to rectify this omission. Specifically, this study aims to investigate whether there are systematic Si isotope variations within, and provide robust Si isotopic compositions for, the major silicate reservoirs on Earth. To this end, a broad range of mantle and crustal lithologies, sourced globally and from various tectonic regimes, have been analysed using high resolution MC-ICP-MS. Analyses indicate that the Si isotopic compositions of mantle-derived mafic and ultramafic material are extremely homogeneous. These data are used to calculate a Bulk Silicate Earth (BSE) average of δ³⁰Si = -0.29 ± 0.08 ‰ (2 s.d.). The degree of Si isotopic fractionation as a result of magmatic differentiation has also been assessed, and found to be small but resolvable between basalt and rhyolitic end-members. Finally, this research shows that, although igneous rocks and sediments derived from the continental crust can be relatively heterogeneous with respect to Si isotopes, bulk averages calculated for the upper, middle and lower continental crust are all very similar to that of BSE. Providing robust estimates for these reservoirs has greatly improved our knowledge of the behaviour of Si isotopes in silicate lithologies and provides a framework for further Si isotopic investigations of such material.

The Palaeoproterozoic Birimian terrane of the West African Man Shield is a gold province of global importance. Gold mineralization at the Damang deposit, however, is unique amongst known gold deposits in Ghana. It comprises a stratigraphically controlled auriferous quartz-pebble conglomerate, which is overprinted by later orogenic gold contained in a sub-horizontal fault-fracture quartz vein array. Furthermore, this gold mineralization is hosted in sediments of the Tarkwaian System, rather than volcanics of the underlying Birimian Supergroup. This thesis integrates petrological, geochemical, thermodynamic and geochronological data to investigate the Damang deposit in the context of the regional tectonic and metallogenic framework. Staurolite-grade, amphibolite facies peak metamorphism at Damang occurred at around 595ºC and 5.5 kbar at 2005±26 Ma. This was immediately followed by a short period (~ 5 Ma) of rapid exhumation, at a rate of approximately 2.6 mm/yr. This decompression led to the formation of the fault-fracture mesh, which was enhanced by the silicification of host rocks during an earlier ‘Pink Haematite’ hydrothermal event that was associated with the intrusion of voluminous dolerite bodies. Orogenic gold mineralization occurred at the end of this exhumation period at conditions of 400–450ºC and 1–2 kbar. These conditions are in broad agreement with other Ghanaian deposits. A gold-bearing, CO₂-rich (X(CO₂) ~ 0.7), aqueous fluid generated through metamorphic devolatilization reactions close to peak metamorphism and during the subsequent exhumation fed the fault-fracture mesh. This fluid imparted a potassic–carbonation–sulphidation alteration overprint similar to that seen at orogenic gold deposits globally. This alteration signature is characterized and identified with portable infrared spectroscopic techniques. Exhumation rates following gold mineralization were less than 0.01 mm/yr, while ⁴⁰Ar/³⁹Ar dating of biotites give ages ranging between 1978.8±6.2 Ma and 1898±11 Ma, indicating extremely prolonged (~ 80 Ma) cooling through the interval 300–250ºC. Ultimately, this thesis demonstrates that the Damang deposit is part of the same regional metallogenic event that formed the other Ghanaian deposits. This implies that the Tarkwaian System elsewhere in Ghana may be prospective for orogenic gold mineralization.

Avant de commencer le bilan de mes travaux et les principaux résultats obtenus j'aimerais faire une introduction, un peu à la marge des préoccupations scientifiques ou presque, mais qui permet de positionner cette HdR. « Rappeler l'histoire c'est gérer efficacement le présent et mieux s'armer pour réussir le futur » : C'est avec cette phrase que j'ai envie de situer ce mémoire. Tracer le cheminement de mes connaissances aura un double écho. Le premier, important, c'est de mieux placer mon expérience de recherche dans le contexte de cette vaste discipline qui est la géologie de la MO. Le deuxième impact, peut être subsidiaire, c'est de confronter mes acquis dans ce domaine et peut être rejoindre mes collègues habilités à diriger des recherches, étape considérée comme importante aux yeux de la plupart des collègues étape qui reste plus administrative que scientifique pour d'autres. Pour ma part j'ai tranché, il faut concilier ces deux mondes : je le fais. Il est vrai que l'exercice ne doit pas être uniquement considéré comme une obligation académique mais également comme une nécessité scientifique qui permet de présenter l'état des lieux des recherches dans mon domaine. Généralement cet exercice est profitable pour le futur habilité, mais doit également l'être par un nécessaire recul scientifique en présentant une vision la plus globale possible au service de notre communauté scientifique. Pour cela, je vais essayer de faire passer comme même un message, ne serait ce que pour les idées que je défends les leçons tirées de notre manière de faire la SCIENCE et les orientations que je souhaite voir émerger pour le futur de notre discipline puis à une échelle plus modeste les projets que je désire développer. Le baromètre d'un chercheur scientifique comme chacun le sait se mesure essentiellement par le nombre de publications. Qu'en est-t-il d'un l'habilité à diriger des recherches. Pour moi, en plus de ces considérations comptables pour tout chercheur, un habilité à diriger des recherches doit l'être par ses compétences à la fois techniques et scientifiques, c'est quelqu'un qui s'investit énormément dans la formation par la recherche des masters et de thésard, c'est un entraineur, un meneur de troupes. Ce n'est pas un acteur de la recherche, c'est plutôt l'auteur et le metteur en scène. En posant cette définition je place la barre très haute et rends difficile ma tache dans le sujet qui nous concerne notamment dans l'élaboration de ce mémoire.

Mass dependent isotope fractionation of stable isotopes between meteorites and planetary materials has been used to assess processes that occurred during formation of Earth and its core. However, thus far little is known about the mass dependent isotope fractionation of Mo in the solar system, and at high temperatures in the Earth, in particular during mantle processes. Molybdenum is a refractory and moderately siderophile element. The processes that might have fractionated Mo in the early solar system include condensation and evaporation of dust grains, metal-silicate segregation, core crystallization, silicate and sulphide melting and aqueous alteration. In order to investigate the processes fractionating Mo isotopes, it is first necessary to assess how much fractionation takes place during mantle melting, estimate the isotopic composition of the bulk silicate Earth, and then make comparisons with primitive and differentiated meteorites. I present double spike Mo isotope data for forty-two mafic and seven ultramafic samples from diverse locations, and nineteen extra-terrestrial samples. The delta^98/95Mo values of all the terrestrial samples (normalized to NIST SRM 3134) exhibit a significant range from +0.53±0.21 to -0.56±0.09‰. The compositions of mid-ocean ridge basalts (MORBs) (+0.03±0.07‰, 2s.d.) and ultramafic rocks (+0.38±0.15‰, 2 s.d.) are relatively uniform and well resolved, providing evidence of fractionation associated with partial melting. In contrast intraplate and ocean island basalts (OIBs) display significant variability within a single locality from MORB-like to strongly negative (-0.56‰). The most extreme values measured are for nephelinites from the Cameroon Line and Trinidade, which also have anomalously high Ce/Pb and low Mo/Ce relative to normal oceanic basalts. The observed relationships between delta^98/95Mo and Ce/Pb, U/Pb and Mo/Ce provide evidence that sulphide plays a critical role in retaining Mo in the mantle and fractionating its isotopic composition in basaltic magmas. If residual sulphides are responsible the Mo isotopic composition, Mo budget of the bulk silicate Earth will be misrepresented by values estimated from basalts. On this basis a revised best estimate of the Mo content in the bulk silicate Earth (BSE) ranging between 251 to 268 ppb is derived, approximately 6 times higher than previously assumed, and similar to the levels of depletion in refractory siderophile elements such as W, Ni and Co. This significantly ameliorates the argument for Mo removal via late stage sulphide extraction to the core. The Mo isotopic composition of the BSE (0.35‰) is distinct from the delta^98/95Mo values found in primitive and iron meteorites. Although Mo isotopic fractionation varies between different phases within a single iron meteorite, and occurs during fractional crystallization in asteroidal cores, most iron meteorites have ddelta^98/95MoSRM3134 (-0.14 to -0.06‰) that are similar to ordinary and CI carbonaceous chondrite (-0.12 to -0.09‰). This range of delta^98/95Moo is not only significantly lighter than the BSE, but also enstatite chondrites, which have delta^98/95Mo values of 0.04 to 0.13‰. Several possible explanations are proposed. (A) Core-mantle differentiation fractionates Mo isotopes. The recently proposed Mo effect of sulphide liquid removal is likely to be minor because this should have generated a light Mo isotope composition for the BSE. However, isotopic fractionation associated with metal-silicate partitioning may be responsible for the heavy Mo in the BSE. (B) A distinct isotopic composition for the late material that contributed Mo to the BSE. Enstatite chondrites (or other putative groups of chondrites with a heavy Mo isotope composition) and sulphur-rich components form the cores of impacting bodies are the most likely candidates that could deliver heavy Mo to Earth. (C) The Mo isotopic composition of the Solar System is heterogeneous in a mass dependent fashion such that heavier Mo isotopes are enriched in the section of the disk from which Earth accreted. There are some difficulties behind each of these models and further work is needed to determine which is correct.

La couverture pédologique de Guyane Française est un compartiment majeur de stockage et de transformation du mercure (Hg) naturel et anthropique. Le fonds géochimique en Hg des sols guyanais est naturellement élevé (0.11-0.50 µg g-1). Dans les sols hydromorphes du bas fond de la crique Combat, les concentrations très élevées en Hg (0.82 ± 0.35 µg g-1) sont héritées de l'orpaillage ancien. L'étude des flux de mercure dans le bassin versant de la crique Combat a montré que les sols jouent un rôle de « puits » pour le Hg inorganique et de source de mono-méthylmercure (MMHg). Nos résultats démontrent que l'orpaillage est à l'origine des conditions géochimiques qui favorisent la méthylation du Hg. Seules la modification des pratiques minières et la réhabilitation des sites après exploitation peuvent limiter la bioaccumulation du MMHg le long des chaînes alimentaires.

The global marine distributions of Cd and phosphate are closely correlated, which has led to Cd being considered as a marine micronutrient. Recent developments in Cd stable isotope mass spectrometry have revealed that Cd uptake by phytoplankton causes isotopic fractionation in the open ocean and in culture. The explanation for this nutrient-like behaviour is unknown as there is only one identified biochemical function for Cd, an unusual Cd/Zn carbonic anhydrase (CdCA1). This thesis investigates why Cd appears to act as an algal nutrient by performing subcellular analyses of microorganisms genetically-modified to express the CdCA1 gene. It was found that CdCA1 was not a significant contributor to whole-cell Cd isotope compositions. Instead, a large proportion of the internalized Cd is sequestered into cell membranes with a similar direction and magnitude of Cd isotopic fractionation as seen in surface seawater. This observation is explained if Cd is mistakenly imported with other divalent metals and subsequently managed by binding within the cell to avoid toxicity. This result implies that surface seawater Cd isotope compositions, if captured by an appropriate archive, may be invaluable for reconstructions of past marine productivity. The role of environmental factors in modulating the inorganic partitioning of Cd isotopes into calcite was investigated through a series of laboratory analogue experiments. In seawater, the light isotopes of Cd are always preferred in calcite. The magnitude of fractionation showed no response to temperature, ambient [Mg], or precipitation rate. To further identify suitable palaeaoceanographic archives, the Cd isotopic composition of a suite of modern deep-sea corals were investigated. It was found that the Cd/Ca and Cd isotope composition of coralline calcium carbonate followed the predicted trend for closed-system Rayleigh fractionation in the calcifying space. The lack of isotopic offsets between some corals and seawater will simplify the application of Cd isotopes in deep-sea corals -- and potentially other marine calcifying organisms that vacuolize seawater prior to calcium carbonate precipitation -- to palaeoceanography.