Experimental study of silica diffusion during metasomatic reactions in the presence of water at 550°C and 1000 bars

Metasomatic reactions between quartz and incompatible oxides or hydroxides were experimentally studied at 550°C and 1000 bars water pressure. Two porous pellets of the initial reagents pressed one against the other were used. Reaction rims in the millimeter range develop at the initial boundary in the oxide pellet. All the experiments show that an important transfer of silica occurs by diffusion in the stationary intergranular solution.The chemical transfer of silica through the intergranular fluid is quantitatively determined by studying the kinetics of growth of the forsterite rim in the system quartz-brucite. The kinetics limiting stage being silica transfer, the experiments allow the determination of the diffusion coefficient of silica through the solution. At 550°C and 1000 bars, a value of 2.4 × 10^?1cm² s¹ is found.This high value shows the importance of chemical diffusion in the intergranular fluid of rocks during metamorphic processes.     

Alteration geochemistry of the Nopal I uranium deposit (Sierra Peña Blanca,Mexico), a natural analogue for a radioactive waste repository in volcanic tuffs

Natural analogues provide an approach to characterize and test the long‐term modelling of a repository performance. This article presents geochemical information about the alteration conditions of the Nopal I uranium deposit, Mexico, an analogue for the proposed Yucca Mountain radioactive waste repository. Mineralization and hydrothermal alteration of volcanic tuffs are contemporaneous, according to petrographic observations. Trace element geochemistry (U, Th, REE) provides evidence for local mobilization of uranium under oxidizing conditions and further precipitation under reducing conditions. O‐ and H‐isotope geochemistry of kaolinite, smectite, opal and calcite suggests that argillic alteration proceeded at shallow depth with meteoric water at 25–75 °C, a low‐temperature context, unusual for volcanic‐hosted uranium deposits. This temperature range is compatible with some post‐closure evolution models of the proposed Yucca Mountain repository.     

Contrasting isotopic signatures between anthropogenic and geogenic Zn and evidence for post-depositional fractionation processes in smelter-impacted soils from Northern France

Zinc isotopes have been studied along two smelter-impacted soil profiles sampled near one of the largest Pb and Zn processing plants in Europe located in northern France, about 50 km south of Lille. Mean δ⁶⁶Zn values along these two soil profiles range from +0.22 ± 0.17‰ (2σ) to +0.34 ± 0.17‰ (2σ) at the lowest horizons and from +0.38 ± 0.45‰ (2σ) to +0.76 ± 0.14‰ (2σ) near the surface. The δ⁶⁶Zn values in the lowest horizons of the soils are interpreted as being representative of the local geochemical background (mean value +0.31 ± 0.38‰), whereas heavier δ⁶⁶Zn values near the surface of the two soils are related to anthropogenic Zn. This anthropogenic Zn occurs in the form of franklinite (ZnFe₂O₄)-bearing slag grains originating from processing wastes at the smelter site and exhibiting δ⁶⁶Zn values of +0.81 ± 0.20‰ (2σ). The presence of franklinite is indicated by EXAFS analysis of the topsoil samples from both soil profiles as well as by micro-XANES analysis of the surface horizon of a third smelter-impacted soil from a distant site. These results indicate that naturally occurring Zn and smelter-derived Zn exhibit significantly different δ⁶⁶Zn values, which suggests that zinc isotopes can be used to distinguish between geogenic and anthropogenic sources of Zn in smelter-impacted soils. In addition to a possible influence of additional past sources of light Zn (likely Zn-sulfides and Zn-sulfates directly emitted by the smelter), the light δ⁶⁶Zn values in the surface horizons compared to smelter-derived slag materials are interpreted as resulting mainly from fractionation processes associated with biotic and/or abiotic pedological processes (Zn-bearing mineral precipitation, Zn complexation by organic matter, and plant uptake of Zn). This conclusion emphasizes the need for additional Zn isotopic studies before being able to use Zn isotopes to trace sources and pathways of this element in surface environments.     

Study of two alteration systems as natural analogues for radionuclide release and migration

U-deposit hosted in hydrothermally altered tuffs in Mexico, together with weathering profiles from Cameroon were studied as natural analogues of radionuclide release and migration. Using petrological and spectroscopic methods (infrared and electron paramagnetic resonance), we have distinguished successive secondary mineral parageneses and the behaviour of radionuclides.

In the U-deposit, the mineral parageneses show that uranium migration is mainly controlled by the redox potential and silica activity of the altering solutions. The high silica content of the solutions is caused by the intense alteration of volcanic rocks. Two types of secondary clay mineral parageneses are evidenced: a kaolinization, intense where uranium is accumulated in the welded tuffs, and a smectitization mainly developed in the underlying weakly welded tuffs.

Several types of kaolinite have been defined according to their genesis (fillings in fissures and feldspar pseudomorphs), their location relative to a breccia pipe where uranium has accumulated (core and rim of the pipe; surrounding rhyolitic tuffs), and particle morphology, structural order and substitutional Fecontent. It is shown that the variations of the concentration of paramagnetic defect centres, always more than ten times as important than those measured in weathering kaolinites, are only correlated to the location of the kaolinites. The highest values correspond to the breccia pipe kaolinites, e.g. kaolinites located in the uranium accumulation zones. Moreover, one or two main defects centres are detected depending on the intimate association of kaolinites with uranium-bearing minerals. Besides, in weathering kaolinites from U-depleted laterites, defect centre concentrations are correlated to the total Fe203 content in bulk samples. This means that the defect centre acts as a memory of the travel of uranium when this element was sorbed onto iron gels in the first stage of weathering.

It is concluded that paramagnetic defect centres in kaolinites might allow an efficient fingerprint of successive irradiations in the natural analogues under study and could be an useful tool to control radionuclides migration through kaolinite-containing clayey materials such as those used for waste repository.

A better understanding of radiation efficiency as well as accurate dose-ratekaolinite-containing clayey materials such as those used for waste reposit estimation are needed for a quantitative tracing of the migration ofA better understanding of radiation efficiency as well as accurate dose-ratekaolin radionuclide elements. With this aim, a simulation has been undertaken withestimation are needed for a quantitative tracing of the migration of various radiations sources. We have determined for each irradiation the parameters of the paramagnetic centres created in order to understand the way they are forming. The knowledge of the parameters governing the formation and the stability of the radiation centres in kaolinites allow to use this mineral as a natural dosimeter.     

Remobilization of arsenic from buried wastes at an industrial site: mineralogical and geochemical control

An industrial area contaminated by As was studied to determine the source of this element and its speciation in As-bearing solids and in run-off waters. Mineral precipitates and water samples were collected and analyzed to assess processes controlling As mobility at this site. The integrated study of a contaminated industrial area allowed identification of the source of the As and of the nature of secondary As-bearing phases. The results obtained both on solid and water samples were used to model As behavior during waste leaching on carbonate rocks. At the upper end of a topographic transect across the site, run-off waters (pH=7.9) interact with surficial waste piles (containing arsenolite, arsenopyrite and pyrite), becoming acidic (pH=2.2) and concentrated in dissolved arsenate species (As⁵⁺) (ΣAs ranging from 0.961 to 3.149·10⁻³ mol/l). Those acidic waters interact with the limestone substratum, providing dissolved Ca which reacts with As to precipitate 1:1 Ca arsenates (weilite CaHAsO₄, haidingerite CaHAsO₄.H₂O and pharmacolite CaHAsO₄.2H₂O) and, in minor amounts, Ca–Mg arsenates (picropharmacolite (Ca,Mg)₃(AsO₄)₂ 6H₂O). The 1:1 Ca arsenates identified are known to precipitate at low pH (3–6) and seem to be stable in media with high dissolved CO₂, in comparison with other types of Ca arsenates. However, due to their high solubilities, they are not strictly relevant candidates to immobilize As in contaminated surficial environments. Although reported solubilities decrease to values close to the French and US drinking standards in Ca-rich solutions, a thorough examination of the precipitation/dissolution kinetics of Ca arsenates should be undertaken to assess their long-term stability and their efficiency in rapidly immobilizing As in contaminated surficial environments.     

Aluminium X-ray absorption Near Edge Structure in model compounds and Earth’s surface minerals

Aluminium K-edge X-ray absorption near edge spectra (XANES) of a suite of silicate and oxides minerals consist of electronic excitations occurring in the edge region, and multiple scattering resonances at higher energies. The main XANES feature for four-fold Al is at around 2 eV lower energy than the main XANES feature for six-fold Al. This provides a useful probe for coordination numbers in clay minerals, gels, glasses or material with unknown Al-coordination number. Six-fold aluminium yields a large variety of XANES features which can be correlated with octahedral point symmetry, number of aluminium sites and distribution of Al-O distances. These three parameters may act together, and the quantitative interpretation of XANES spectra is difficult. For a low point symmetry (1), variations are mainly related to the number of Al sites and distribution of Al-O distances: pyrophyllite, one Al site, is clearly distinguished from kaolinite and gibbsite presenting two Al sites. For a given number of Al-site (1), variations are controlled by changes in point symmetry, the number of XANES features being increased as point symmetry decreases. For a given point symmetry (1) and a given number of Al site (1), variations are related to second nearest neighbours (gibbsite versus kaolinite). The amplitude of the XANES feature at about 1566 eV is a useful probe for the assessment of Al^IV/Al^total ratios in 2/1 phyllosilicates. Al-K XANES has been performed on synthetic Al-bearing goethites which cannot be studied by ²⁷Al NMR. At low Al content, Al-K XANES is very different from that of α-AlOOH but at the highest level, XANES spectrum tends to that of diaspore. Al-K XAS is thus a promising tool for the structural study of poorly ordered materials such as clay minerals and natural alumino-silicate gels together with Al-subsituted Fe-oxyhydroxides. Received: 18 December 1996 / Revised, accepted: 2 June 1997     

Dike distribution in the Oman-United Arab Emirates ophiolite

Mafic dikes and dunite veins are observed in the mantle section of the Oman – United Arab Emirates (O-UAE) ophiolites, as well as diabase dikes and hydrothermal veins in the crust section. They have been systematically measured during the mapping of this ophiolite and are represented by their trajectories in the folded map 3 in the back of this volume, and by local stereoplots included in this study. Mafic dikes in the mantle section correspond to basaltic melt being injected at decreasing temperatures from above or at peridotite solidus, down to below 450°C. Hydrothermal veins associated with dioritic dikes issued from hydrous melting of host gabbros are observed down to the base of the crust, bearing evidence for sea water penetration into basal gabbros at or above 900°C, that is very close to the ridge axis. Dike orientations record the stress field at the time of their injection. In most places, all types of dikes are dominantly parallel to the general trend of the nearest sheeted dike complex; thus the stress field has not visibly changed from melt injection in the asthenosphere below the ridge of origin to injection in a lithosphere up to a few Myr old, at distances beyond 100 km from the axis. Local preferred orientations, when they are considered in the frame of the paleo-ridge system of O-UAE, result in a coherent model throughout the belt: the sheeted dike complex dips moderately away from the presumed ridge axis and the mantle dikes, toward this axis. These opposite directions are explained by the presumed effect of subsidence toward the axis for the sheeted dikes and by the central feeding from an asthenospheric uprise for the mantle dikes.     

A new method to estimate the oxidation state of basaltic series from microprobe analyses

The oxygen fugacity and therefore the iron redox state of a melt is known to have a strong influence on the liquid line of descent of magmas and thus on the composition of the coexisting melts and crystals. We present a new method to estimate this critical parameter from electron probe microanalyses of two of the most common minerals of basaltic series, plagioclase and clinopyroxene. This method is not based on stoichiometric calculations, but on the different partitioning behaviour of Fe³⁺ and Fe²⁺ between both minerals and a melt phase: plagioclase can incorporate more Fe³⁺ than Fe²⁺, while clinopyroxene can incorporate more Fe²⁺ than Fe³⁺. For example, the effect of oxidizing a partly molten basaltic system (Fe³⁺ is stabilized with respect to Fe²⁺) results in an increase of FeO_total in plagioclase, but a decrease in the associated clinopyroxene. We propose an equation, based on published partition coefficients, that allows estimating the redox state of a melt from these considerations. An application to a set of experimental and natural data attests the validity of the proposed model. The associated error can be calculated and is on average < 1 log unit of the prevailing oxygen fugacity.In order to reduce the number of different variables influencing the Fe²⁺/Fe³⁺ mineral/melt equilibrium, our model is restricted to basaltic series with SiO₂ < 60% that have crystallized at intermediate to low pressure (< 0.5 GPa) and under relatively oxidizing conditions (?FMQ > 0; where FMQ is the fayalite–magnetite–quartz oxygen buffer equilibrium), but it may be parameterized for other conditions. A spreadsheet is provided to assist the use of equations, and to perform the error propagation analysis.