The effects of composition, pressure and oxygen fugacity on partition coefficients between amphibole and hydrous basaltic melt were studied at 1.5 to 2.5 GPa and 1000 to 1130°C. Partition coefficients (D i = concentration of element i in amphibole/concentration of i in melt) of large-ion-lithophile elements (LILE: Rb, Sr, Ba), high-field-strength elements (HFSE: Y, Zr, Nb, Ta, Hf), and rare-earth elements (REE: La to Lu) were determined between amphiboles and coexisting quenched melts created by partial crystallization of seven different starting compositions in a piston-cylinder high-pressure apparatus. Trace elements were analyzed by laser-ablation, microprobe inductively coupled plasma-mass spectrometer (LAM-ICP-MS). The effects of premium, temperature and oxygen fugacity on the partition coefficients are minor, but statistically measurable. Amphibole composition affects partitioning of these trace elements by a maximum factor of 3.5 in the range of pressures and temperatures studied with an oxygen fugacity range of 2 orders of magnitude above and below nickel-nickel oxide buffer. Experiments specifically investigating the role of Ti demonstrate that a positive correlation exists between amphibole VITi 4+ content and DBa, D Sr, DTa, D Zr, DLa, DCe, DPr, and DNd. Increasing pressure from 1.5 GPa, to 2.2 or 2.5 GPa (depending upon composition) increases DLILE, but decreases DHFSE and DREE. Raising the oxygen fugacity at 1.5 or 2.5 GPa by 3 orders of magnitude increases DRb, DBa, DLa, and D Nd, whereas DTi, D Hf, and DZr decrease; however, the maximum difference between partition coefficients measured at low and high oxygen fugacities is only a factor of 1.7. All of the effects of composition, pressure, and oxygen fugacity reflect the control of crystal chemistry on the partitioning of trace elements between amphibole and basaltic melt. No effects of melt composition were discerned in this study. The measured partition coefficients were used to investigate tr

A review of immiscibility data in 62 silicate, borate and germanate binaries permits identification of four groups of cations displaying different immiscibility behaviour. The first group consists of network-modifier cations which have ionic radii ≳ 87.2 pm and coordination numbers equal to, or higher than, 5 (e.g. Ca2+, La3+, U4+). The second group involves cations with ionic radii ≲ 87.2 pm. They have at least two coordination numbers: the first one is always 4 and the second is ≥5; for this reason they are called amphoterics (e.g. Li1+, Mg2+, Al3+). The third group contains cations with variable crystal field stabilization energies (e.g. Fe2+, Ni2+, Cr3+), and the fourth group includes cations with a lone pair of electrons (e.g. Pb 2+, Bi3+, Te4+). / Immiscibility data suggest that the origin of phase separation is associated with coulombic repulsions between poorly screened cations. The larger the ionic potential of a cation, the greater the repulsions with its neighbours, and the larger the size of its immiscibility field. However, amphoterics do not obey this rule because network-formers like SiO2 exert a structural control upon immiscibility which creates a selective solution mechanism that affects cations with radii ≲ 87.2 pm. Such small cations appear to be capable of fitting in pentagonal-like cages where they adopt a 4-fold coordination. In tetrahedral coordination, the bonds have a greater covalent character and the oxygens are more polarized towards the amphoterics which efficiently shield their positive charges, reducing coulombic repulsions and thus immiscibility. / Experiments performed in the system CaO-MgO-SiO2 at 1.0 GPa show that pressure has little effect on miscibility gaps associated to network-modifiers such as Ca2+ and "weak" amphoterics like Mg 2+. However, it is shown that amphoterics with substantial fractions of cations in 4-fold coordination and cations with variable crystal field stabilization energies capable of high spin to low spin transitions are expected to enlarge their immiscibility fields at high pressure. Magmas rich in these cations are potential candidates to develop phase separation at depth. / Experiments were also performed in the systems CaO-SiO2 and MgO-SiO2 at 1.0 GPa and results were combined with all the currently available phase equilibria and thermodynamic data at 1 bar in the same systems to critically optimize the thermodynamic properties of the liquid phase at low and high pressures. Assessments were made with the modified quasi-chemical model of Blander & Pelton using a computer program that simultaneously optimized all reliable data to give a small set of excess Gibbs free energy parameters. Pressure was found to have little effect on the topology of the CaO-SiO2 system but a pronounced one on the MgO-SiO2 binary. These contrasting behaviours from two homovalent and isochemical cations are linked to the polymorphic transition of the magnesium metasilicate. The amphoteric nature of the Mg2+ cation in the liquid phase makes MgO-SiO2 melts compressible but this effect appears to be small.

The chemical evolution of peridotites and felsic rocks included in or adjacent to them, and fluids during the hydrothermal alteration of these rocks is the subject of this study. Experiments using a flow-through hydrothermal apparatus, and studies of natural samples from the JM Asbestos chrysotile deposit in the Asbestos ophiolite, Quebec, were the means by which the various topics of this thesis were investigated. Open-system fluid-rock interaction was studied by flowing deionized water or a 0.1 m CaCl2 solution through olivine (Fo91) and lherzolite at a temperature of 300°C and a pressure of 300 bars. During these experiments, the original minerals (olivine, orthopyroxene and clinopyroxene) were altered to an assemblage, which consisted of serpentine-group minerals (conical, polygonal, and cylindrical fibers, and plates), brucite and magnetite. Reaction of lherzolite with deionized water and thermodynamic modelling indicate that the highest pH and calcium concentration are attained when pyroxenes are present in the unaltered rock, and that these parameters are at maximum in the presence of the buffering assemblage serpentine + brucite + diopside. Alteration of olivine by 0.1 m CaCl2 aqueous solution does not lead to the formation of secondary calcium phases. These results, supported by thermodynamic calculations, suggest that calcium-rich fluids can circulate through forsteritic olivine-rich lithologies without inducing calcium metasomatism. / In the JM Asbestos mine, these Si-Al lithologies are largely felsic dykes, which, based on geological relationships, are interpreted to have been emplaced after serpentinization of the host peridotites began, but before obduction of the Asbestos ophiolite onto the Laurentian continental margin. Slates underlying and in thrust contact with serpentinite are also locally rodingitized. Mass balance calculations suggest that the calcium responsible for the rodingitization of the slates may have been derived locally, whereas the calcium responsible for the rodingitization of the felsic dykes was most probably derived from a source remote from the dykes. The strong dilution in rodingites and nephrites of the low-solubility minor and traces elements P, Cr and Zr suggests that very large amounts of fluids were involved in calc-silicate metasomatism of the felsic dykes and adjacent serpentinites. / The absence of CO2 and presence of hydrocarbons (alkanes from methane to pentane) in the fluid inclusions of both episodes is an indication that the fluids responsible for calc-silicate alteration were derived from the serpentinites. (Abstract shortened by UMI.)

The influence of parent-solution chemistry on the precipitation of various Pb-carbonate phases (cerussite, PbCO3(s); hydrocerussite, Pb3(CO3)2(OH)2(s); and phosgenite, Pb2(CO3)Cl2(s)) is studied at 25ºC, 1 bar total pressure in the systems Pb(NO3)2-CO 2-H2O and PbCl2-CO2-H2O. Analysis of both fresh and aged (8 and 16 days), fine-grained Pb-carbonate precipitates resulting from the addition of 0.035 M Pb-bearing (PbCl 2 or Pb(NO3)2) into 0.035 M C-bearing (NaHCO 3 and Na2CO3) reagents (or vice versa ) by powder X-ray diffraction and variable-pressure scanning electron microscopy reveals that: (1) a potential relationship exists between crystal morphology and aqueous speciation (i.e. in the presence of NO3- and Cl-); and (2) solid-phase interconversions between cerussite ↔ hydrocerussite and cerussite ↔ phosgenite readily occurs until equilibrium conditions are achieved in a period of time < 8 days. Further, the precipitation conditions corresponding to various morphologies and habits of cerussite (prismatic, bladed, platy-hexagonal) and phosgenite (tabular, intergrown) is documented.

Anomalously high concentrations of arsenic and phosphate are found in the sediments of the Saguenay Fjord in comparison to those of the Gulf and St. Lawrence Estuary. The source of arsenic is unclear. It could be supplied to the Fjord by the marine waters which enter from the St. Lawrence Estuary or may have accumulated as a result of historically high levels of effluent input. / The waters of the Saguenay Fjord are also characterized by anomalously high particulate aluminum concentrations (i.e., [Al]tot/[Al] diss > 1), introduced as a result of the aluminum refining activity along the shores of its tributaries. Gibbsite (Al(OH)3), a by-product of this activity, is known to strongly adsorb phosphate and arsenate from low ionic strength aqueous solutions, and may act as a vector for these two elements to sediments of the Saguenay Fjord. / The small amount of arsenate and phosphate adsorbed onto gibbsite from seawater indicates that their adsorption to particulate aluminum in the water column cannot account for the elevated levels of arsenic and phosphate in the sediments of the Saguenay Fjord. (Abstract shortened by UMI.)

An experimental investigation of metal-organic co-adsorption was conducted using cadmium and citrate in the presence of corundum. The experiments were performed in the pH range 3.0 to 10.0 in a 0.01M NaCl matrix with citrate:Cd ratios of 0.00 and 10.00. The presence of citrate enhances cadmium adsorption between pH 3.5 and 7.2, and reduces it above pH 7.2. The results from the experiments provide constraints on the thermodynamic properties of the surface complexes. The adsorption of Cd is quantified by the competition between aqueous CdCit$ sp{-1}$ and the surface species $>$AlCitCd$ sp0$, $>$AlOCd$ sp+$, and $>$AlOCdCl$ sp0$. The experimentally determined equilibrium constants provide a means to predict Cd mobilities in groundwater systems by describing the relative stabilities of surface and aqueous metal-organic complexes as a function of metal-to-ligand and solute-to-sorbent ratios.

<p> The global iron and sulfur cycles are linked to some of the most ancient metabolisms on our planet, and, therefore, possibly other planetary bodies. They are also linked to some of our most pressing environmental problems on Earth. Understanding their interactions and monitoring their occurrence is, therefore, an important aspect of exploring planetary bodies and sustainable resource management. This dissertation reports investigations into the identification and analysis of iron and sulfur bearing phases in three extreme environments: Acid-gas/CO₂ co-sequestration, acid mine drainage (AMD), and Mars. </p><p> H₂S and SO₂ (acid gases) are often co-contaminants in CO₂ streams. Co-sequestration of these gases lowers the cost of sequestration, but these sulfur-bearing gases can increase the reactivity of the injection site, particularly if iron-bearing minerals are present. Analyzing iron and sulfur reactivity in these systems is, therefore, crucial to further development of carbon sequestration technology. Experiments examining the simulated co-injection of CO₂, H₂S and/or SO₂ are reported for hematite-bearing and hematite-free sandstones. The iron mineralogy of the sandstones, which is monitored using M&ouml;ssbauer spectroscopy, is key to the reactivity of the sandstones. </p><p> Jarosite is an iron sulfate found primarily in areas of AMD, but it has also been identified on Mars. On Earth, using remote sensing to quantify jarosite concentrations could be a time and cost saving measure for AMD detection and cleanup. On Mars, quantifying jarosite may help constrain surface processes responsible for its formation. Quantitative mineral abundance determination using remote sensing requires knowledge of the optical constants of minerals present on the surface. The optical constants of jarosite are determined and the methodology for optical constant determination (Hapke formulation), including an open source computer code, is reported.</p>

Anomalously high concentrations of arsenic and phosphate are found in the sediments of the Saguenay Fjord relative to those of the Gulf and St. Lawrence Estuary. Whereas the source of phosphate is likely anthropogenic, arsenic appears to be scavenged from the bottom marine waters by settling detrital and authigenic particles. The surface waters of the Saguenay Fjord show a particulate aluminum anomaly ([Al]tot/[Al]diss > 1) that decreases downstream or with increasing salinity. The mineralogy of the aluminum particulate matter may be akin to gibbsite given the refining activities in the region. The adsorption of phosphate and arsenic to various mineral oxides is well established but the scavenging capacity of gibbsite in this estuarine environment is not known. / In order to simulate the behavior of the adsorbent/adsorbate relationship during estuarine mixing, the adsorption capacity of gibbsite for arsenate and phosphate was measured in pure water; 0.67 M NaCl; 10 mM CaCl2; 10 mM CaCl2 + 0.64 M NaCl and in seawater in the absence and presence of the fluoride ion.

The oxygen isotope fractionation between HCO3-/CO 32- and H2O was determined experimentally in order to elucidate the precipitation mechanisms of orthorhombic carbonate minerals in solutions. The oxygen isotope composition of BaCO3 (witherite), precipitated quantitatively from solutions of various pH and the relative contribution of carbonic acid species to DIC (total dissolved inorganic carbon) at each pH in these solutions were used to generate the following fractionations at 25&deg;C: 1000lnaHCO -3-H2O =30.53+/-0.08 1000lna&parl0;CO 2-3-H2 O&parr0;=23.17+/-0.08 / Based upon the above oxygen isotope fractionation and the oxygen isotope systematics of aragonite and witherite obtained from various precipitation experiments, it was discovered that CO32- rather than the more abundant HCO3- ions are preferentially incorporated into the growing carbonate crystal. Furthermore, it was proposed that the faster deprotonation kinetics of isotopically light HCO 3- ions and the preferential incorporation of light CO3 2- isotopologues could account for the kinetic isotope effects observed in these carbonate minerals. / Aragonite was inorganically precipitated from Na-Ca-Mg-Cl-HCO3 solutions at 0, 5, 10, 25, and 40&deg;C to determine the temperature dependence of the equilibrium oxygen isotope fractionation between this mineral and the parental solution. To establish a reliable relationship between temperature and oxygen isotope fractionation as well as discount potential experimental artifacts, three different precipitation protocols were employed but statistically identical results were obtained under the conditions investigated. In addition, the extent of kinetic isotope effects induced by variations of the precipitation rate, pH, and Mg2+ concentration of the parent solutions, were thoroughly tested to establish that equilibrium oxygen isotope fractionation occurs. / In order to obtain accurate oxygen isotope compositions of the aragonite precipitates by the conventional phosphoric acid dissolution method, a statistically reliable acid fractionation factor for aragonite (and calcite) was also determined from pure natural aragonite (and calcite) specimens. From a large number (i.e., 29 and 60 individual analyses for aragonite and calcite, respectively) of replicate total oxygen isotope analyses and the isotopic composition of the acid-liberated CO2 between 25 to 75&deg;C, the following two new equations are proposed: 1000lnaCO2&parl0;ACID &parr0;-Calcite=3.59&parl0; 103/T&parr0;-1.79 1000lnaCO2&parl0;ACID &parr0;-Aragonite=3.39&parl0; 103/T&parr0;-0.83 / By combining the new acid fractionation factor of 1.01063 for aragonite at 25&deg;C with the oxygen isotope data from the synthesis experiments determined in this study, a new relationship is reported for the temperature dependence of the aragonite-water oxygen isotope fractionation over the temperate range of 0-40&deg;C: 1000lnaaragonite-water=17. 88+/-0.13&parl0;103/T&parr0; -31.14+/-0.46 / This new experimental calibration of the aragonite-water fractionation yields a positive aragonite-calcite oxygen isotope fractionation over the temperature range investigated, in agreement with theoretical calculations. This result also provides a baseline to quantify vital or kinetic effects that are frequently observed in natural abiogenic or biogenic aragonite minerals.

Iron oxyhydroxides are ubiquitous in natural systems and have been recognized as strong adsorbents for Group V elements, such as arsenic and phosphorus. Consequently, the mobility and bioavailability of phosphate and arsenate in the marine environment are mostly regulated by the fate of iron oxyhydroxides. However, to this day, the majority of adsorption studies of phosphate and arsenate were performed in low ionic strength solutions. In this study, phosphate and arsenate complexation on goethite in a 0.7 M NaCl solution and seawater was characterized over a pH range of 3.0 to 10.0 in batch adsorption experiments. Phosphate complexation is described using electric double layer models which consider the presence of three monodentate surface complexes, each characterized by an intrinsic formation constant. Arsenate shows a similar adsorption pattern on goethite but a higher affinity than phosphate. A model including three surface complexes describes the arsenate adsorption at relatively high initial concentrations (i.e., 23 and 34 mumol/l) but overestimates the adsorption at a lower concentration (i.e., 8.8 mumol/l). The equilibrium model derived by combining the formation constants obtained in the single oxyanion subsystems predicts the shape of the competitive adsorption data but fails to reproduce it quantitatively. In competitive experiments, phosphate adsorption is underpredicted whereas arsenate adsorption is overpredicted. / Magnesium, calcium and sulphate complexation on goethite was studied in a 0.7 M NaCl solution at 25°C and a pH range of 3.0 to 10.0 and their surface complexation constants were derived. Phosphate adsorption in a 0.7 M NaCl solution increases slightly in the presence of Mg and Ca and decreases at low pH upon the addition of SO4. In contrast, arsenate adsorption is not affected by the presence of Mg or Ca. Phosphate adsorption in seawater is reduced at low pH and enhanced at neutral pH relative to 0.7 M NaCl whereas arsenate adsorption is identical in both solutions except for a decrease at low pH in seawater. The equilibrium model derived by combining the single adsorbate subsystems predicts phosphate and arsenate adsorption in the PO 4-SO4-goethite and AsO4-Ca-goethite systems well, but fails to accurately reproduce the adsorption data in the PO4-Ca-goethite, PO4-Mg-goethite, PO4-seawater-goethite, AsO4-Mg-goethite and AsO4-seawater-goethite systems. The inclusion of ternary surface complexes >FeOMgHPO4-, >FeOMgH2PO 40, >FeOCaHPO4-, >FeOCaH 2PO40, >FeOMgHAsO4- and >FeOMgH2AsO40 in the latter systems improves model fits significantly. Phosphate adsorption in the competitive experiments in seawater is identical to that in the 0.7 M NaCl solution whereas arsenate adsorption in seawater is greater at pH > 7. In competitive experiments in seawater, phosphate adsorption is underpredicted at pH < 6.5 whereas arsenate adsorption is reproduced well using the Constant Capacitance Model (CCM) with the inclusion of the ternary surface complexes. / Phosphate and arsenate adsorption by ferrihydrite, in both 0.7 M NaCl and artificial seawater solutions, was studied at particle concentrations of 0.24 g/l and 0.025 g/l between pH 3.0 and 10.0. Both phosphate and arsenate surface complexation can be described using a model comprising three surface intrinsic constants which decrease with decreasing particle concentration. Ferrihydrite is a stronger adsorbent of PO4 and AsO4 on a per mass basis whereas goethite is a stronger sorbent in terms of chemical affinity. The CCM is able to predict PO4 and AsO4 adsorption on a mixture of goethite and ferrihydrite by combining the complexation constants derived from the single adsorbent subsystems. There is no evidence that adsorbed PO4 and AsO4 is released to solution when ferrihydrite is converted to hematite, an iron oxide with a much lower affinity for these oxyanions. The coprecipitation of both PO4 and AsO4 with hematite during the conversion may be responsible for these observations. / Finally, the surface complexation constants reported in this thesis may be incorporated into other models and contribute to the development of an experimental and thermodynamic database for marine geochemical systems.