V23A-1221
Experimental Constraints on Al Solubility in Model Granitic Fluids at 700 °C and 1 GPa
High P and T fluids associated with metamorphism and magmatism play an important role in crustal evolution and mass transfer. To understand aqueous fluids associated with model granitic crust at high P and T, we investigated the solubilities of K-feldspar (Ksp), muscovite (ms) and corundum (co) in H2O at 700°C and 10 kbar. Starting materials of natural microcline, single corundum crystals, reagent KSi3O6.5 (KS3) were equilibrated with H2O in a piston-cylinder apparatus. Fluid compositions were determined by weight loss and phase-equilibrium bracketing methods. With increasing KS3 from 0.021 to 0.45 molal, Al increased from 0.012 to 0.170 molal and coexisting mineral assemblage changed from co, through ms, to Ksp. Our results locate two invariant points: I1, co + ms + fluid (0.082 molal Al, 1.32 molal KS3) and I2, Ksp + ms + fluid (0.16 molal Al, 2.69 molal KS3). Total solubility of co in the presence of dissolved KS3 is up to fifty times higher than in pure H2O at the same P and T. Experiments at quartz saturation at the same P and T provide the pH buffering assemblages Ksp + ms + qtz (I1) and ky + ms + qtz (I2) and give nearly identical results, which implies that the presence or absence of quartz does not significantly effect Al solubility for the investigated mineral assemblages. Al enhancement due to a possible Al-K species was investigated by measuring co solubility in KOH solutions at 700°C and 10 kbar. Measured logAl solubility increased from -2.55 to 0.43 molal with increasing logKOH from -2.89 to 0.59 molal. Total Al solubility in the system Al-K-O-H assuming an Al-K species was calculated using thermodynamic properties of Pokrovskii and Helgeson (Chem. Geol. 137, 221, 1997) for Al species extrapolated to 10 kbar using the Guentelberg equation for activity coefficients. Calculated solubilities agree closely with the experimental results. Addition of silica species (SiO2, aq, Si2O4, and HAlSiO4) allowed prediction of Al solubilities of -1.87 molal logAl (I1) and -1.82 molal logAl (I2), which are substantially lower than measured. The underestimate of solubility using only ionic, monomeric and dimeric species points to the presence of more polymerized aqueous species in the fluid phase. Our experimental results show that Al solubility is much higher than previously appreciated in aqueous fluids equilibrated with granitic bulk compositions, and that substantial Al transfer may occur at deep-crustal metamorphic conditions.
V23A-1222
Experimental Constraints on the Transformation of Granulite-Facies Metapelites to Eclogites and the Role of Brines in Subduction-Zone Processes
A common feature of HP and UHP terranes is the subduction of lower crustal rocks to great depths. Investigations in the Bergen Area in Norway have shown that this process is triggered by fluids present during eclogite-facies metamorphism (e.g. Austrheim, 1987, EPSL, 81, 221-232). Fluid inclusions in eclogite-facies minerals range from dilute solutions to chloride-rich brines and previous studies have shown that highly saline fluids may occur in subduction zones where continental material is involved. A key example of these processes is exposed at Val Savenca of the Sesia-Lanzo Zone, Italy, where Eo-Alpine eclogite-facies metamorphism and fluid flow led to partial transformation of Hercynian amphibolite-eclogite facies metapelites (garnet + biotite + sillimanite + K-feldspar + plagioclase + quartz) to zoisite + jadeite + kyanite + phengite + quartz with associated jadeite veins. Application of the K-feldspar-jadeite-quartz barometer to the plagioclase domains yields P-T conditions of 1.7-2.1 GPa at 600°C and low a(H2O) of 0.3-0.6 (Tropper et al., 1999, JMG, 17, 195-209). In order to understand the role of brines in the transformation exposed at Val Savenca, we carried out piston- cylinder experiments with a fresh, natural granulite from the Moldanubic Unit in upper Austria with the assemblage garnet + biotite + K-feldspar + plagioclase + quartz. The experiments were conducted in the presence of H2O-NaCl fluids at 600°C and 2 GPa for 2-4 days. The fluids had the compositions X(H2O) = 1.0, 0.90, 0.80 and 0.70, and fluid/rock ratio varied from 1:1 to 1:10. Oxygen fugacity was buffered at NNO and HM in most experiments. Preliminary data from unbuffered experiments show increasing reaction progress with increasing salinity in the fluid. At X(H2O) = 1.0, no visible reaction occurred, whereas at X(H2O) = 0.7, the protolith assemblage is almost completely replaced by large (up to 100 mm long), euhedral crystals of jadeite + phengite + quartz. Biotite is the only relict phase. The experiments also show a strong indication of Na-Ca exchange: the anorthite component of plagioclase goes readily into solution, producing zoisite needles only upon quench. Future experiments will be conducted on fluids in the system H2O-NaCl-CaCl2 to quantify this behavior. These experiments show that brines are highly effective in promoting reaction progress in subduction zone processes.
V23A-1223
Thermodynamics of SiO2-H2O at the upper critical end point
The hydrous melting curve of quartz terminates at a critical end point (CEP) near 10 kbar and 1080°C (Kennedy et al, AJS 260, 501, 1962), where the compositions of coexisting aqueous fluid and hydrous melt become identical near mol fractions (SiO2: XS; H2O: XH) of 0.5. To explore this behavior we investigated the thermodynamic mixing properties of SiO2-H2O near the CEP. We measured quartz solubility at 10 kbar and 700°-1050°C by weight loss of single crystals encased with H2O in welded Pt capsules in a piston-cylinder apparatus. The SiO2 mol fractions are, at each T(°C): (700, 0.0123); (800, 0.0217); (850, 0.0295); (900, 0.0398); (950, 0.0575); (1000, 0.0824) (1035, 0.138); (1050, 0.199); (1060, 0.320). The last datum is from Kennedy et al (1962; 9.5 kbar). These data along with solubility measurements on enstatite-forsterite (Newton and Manning, GCA 66, 4165, 2002) were used to retrieve activity coefficients (γS) of SiO2 (reference state Si(OH)4): ln γS = -0.0554(ln XS)2 - 1.064(ln XS) - 4.057, which is valid in the range of the measurements. The activity coefficient of H2O (γH) in quartz-saturated solutions determined from the Gibbs- Duhem equation varies from 1.003 at 700°C to 1.378 at 1060°C, with a projected value of 1.653 at 1080°C. The strong T dependence of XS above 1000°C and the γH near the theoretical regular solution value of e1/2 = 1.649 at a critical point verify that a CEP must exist near 10 kbar, 1080°C and XS = 0.5. Furthermore, the activity of the monomer Si(OH)4 and that of the dimer SiO(OH)6 are so low at the CEP that other more highly polymerized species must dominate. We propose a change of reference state at high T and P in the system SiO2-H2O from Si(OH)4 to SiO(OH)2 as XS approaches 0.5. For compositions near 1:1, the mixture is virtually a regular solution. The reference state may consist of charge-balanced rings and/or long chains of stoichiometry near H2SiO3 or a random mixture of charged or uncharged groups of near 1:1 composition.
V23A-1224
Speciation of Aqueous Silica Using Raman Spectroscopy and First Principles Calculations
This study presents Raman spectra of high-pH silica solutions taken at ambient conditions with varying silica concentrations. Dissolved silica plays an important role in lithospheric fluid chemistry. Over the range of crustal temperatures and pressures, silica concentrations in quartz-saturated aqueous fluids vary sufficiently to allow for significant mass transport of silica via fluid-rock interaction. Polymerization of aqueous silica plays an important role in elevating dissolved SiO2 concentrations, and could afford silicate-melt-like or crystal-like sites into which otherwise insoluble elements such as titanium could substitute, leading to enhanced mobility for those elements. It would therefore be useful to understand what the independent effects of concentration, composition (pH and incorporation of other elements), pressure, and temperature are on silica polymerization. Raman spectra of silica solutions have previously been obtained [Ref. 1, 2], but those studies did not vary aqueous silica concentration at a fixed P and T. As a foundation for future studies of polymerization of aqueous silica at high P and T, we collected Raman spectra of high-pH silica solutions at ambient conditions with varying silica concentrations. Total silica concentration was varied while keeping pH, P, and T constant. First principles calculations of explicitly solvated silica monomers and dimers are used to interpret the experimental spectra. The spectra show that as total silica concentration increases, the ratio of silica in dimers to silica in monomers increases as well, shown by the ratio of the dimer peak height at 600 cm-1 to the monomer peak height at 780 cm-1. This has been thermodynamically predicted [Ref. 3], and has been indirectly observed in high P-T solubility measurements [Ref. 4], but ours is the first experiment to directly observe that this is the case while keeping temperature and pressure constant. These results are a promising first step towards hydrothermal diamond anvil cell experiments aimed at exploring the effects of temperature, pressure, and concentration on silica polymerization. 1. Zotov and Keppler (2002). Chem. Geol., 184: 71-82. 2. Dutta and Shieh (1985). Appl. Spectrosc., 39: 343-346. 3. Gerya, et al. (2005). Eur. J. Mineral., 17: 269-283. 4. Newton and Manning (2003). Contrib. Min. Pet., 146: 135-143.
V23A-1225
Experimental Investigation of the Dissolution and Alteration of Olivine at Low Temperatures and Pressures
We have determined the surface characteristics of single crystals of olivine that have been altered under a variety of experimental conditions relevant to processes occurring on the Martian surface and in shallow crustal environments (i.e. mid-ocean ridge systems and hydrothermal environments). Most previous experiments have focused on dissolution rates of mineral powders at near-surface conditions of temperature (25 °C), pressure (1 atm), and pH (2--7) (see Brantley, 2003, for summary). The intent of our study is to explore the surface physical, chemical and mineralogical features attendant with reactions of single crystals of olivine. Surface analytical methods such as SEM, XRD, TEM, SIMS, IR and XPS are being used to interrogate the reacted mineral surfaces. Our experiments were performed over a range of T=25--150 °C, t=1--30 d, and pH=2--7. The starting material for these experiments consisted of single grains of San Carlos olivine (Fo90) that were cut to a size of ~4 mm in diameter and ~1 mm in thickness. The larger, flat surfaces of these olivine ‘slices' were polished to ensure a uniform surface layer. Initial starting solutions included doubly-distilled water (pH=7), HCl (pH=2) and H2SO4 (pH=2). Two olivine crystals per experiment were loaded into a pyrex tube, and the accompanying solution was added with a fluid to rock ratio of either 5:1 or 15:1. Solutions were frozen and tubes were evacuated prior to sealing, ensuring that any reaction or dissolution occurred at an internal pressure corresponding to the temperature of interest along the liquid-vapor curve for the solution. Preliminary results indicate that all olivine samples, regardless of experiment duration, temperature, or pH conditions, have some measurable amount of weight loss (0.1--0.3%). SEM analysis shows the development of dissolution pits in most of the experimental samples, and the presence of secondary phases on the surface of olivine samples reacted with acidic starting solutions (pH=2). BSE and SE images indicate that these secondary phases (as yet not quantitatively determined) are compositionally distinct from starting olivine. A characteristic of these experiments is that the reaction rapidly increases the pH of the solution, even in experiments with a fluid to rock ratio of 15:1. At a fluid to rock ratio of 5:1, the short duration experiments (t=1d, T=150 °C) exhibited a change in pH from 2 to 6.17 for HCl (aq) starting solutions, and a similar change was also observed for H2SO4 (aq) starting solutions. These results have an important bearing on the interpretation of olivine reactivity in acidic solutions relevant to numerous planetary crustal settings.
V23A-1226
Mass Transfer as a Function of Metamorphic Fluid Flux
Fluid-rock interaction is an integral part of metamorphism in the crust, and causes significant mass transfer most clearly demonstrated by mineral veins, ore deposits, and alteration fronts that are ubiquitous throughout metamorphic belts. Quantification of major and trace element fluxes during metamorphism and fluid-rock interaction is critical for complete understanding of global element cycles. The SW Highlands of Scotland are an ideal natural laboratory for investigation of major and trace element fluxes during metamorphism. Syn- metamorphic time integrated fluid fluxes within the 6km wide Ardrishaig anticline have been quantified and vary from <10m2/m3 on the limbs of the anticline to >300 m2/m3 in the axial zone (Skelton et al., 1995, J Petrol 36, 563). This study investigates the major and trace element fluxes that occur as a function of the variation in fluid flux within the Ardrishaig anticline. Both depletion and enrichment in chemical elements occurs and the scale of mass transfer increases with increasing fluid flux. Mobile elements include Si, C, K, Na, Sr, As, and Sb, and element mobility varies strongly with lithology. This well-constrained, localized study correlates well with larger regional scale studies (Pitcairn et al., 2006, Econ Geol 101, 1525) where many elements were shown to be locally mobile but very few were shown to have undergone large scale transport.
V23A-1227
Crystal Size Distribution of Periclase in Contact Metamorphic Marbles as Record of Fluid Infiltration
Crystal size distributions (CSD) of periclase in contact metamorphic marbles are combined with geochemical and petrologic information to deduce the controls that acted on the periclase forming reaction. Data are presented for two profiles in a dolomite xenolith in mafic intrusive rocks at the Cima Uzza, southern Adamello massif (Italy). Stable isotope data and the presence of a sharp periclase reaction front on hand specimen scale shows that the formation of periclase is the consequence of high temperature fluid infiltration. Stable isotope data show depletion for 13C and 18O in a narrow region (~40cm) near the igneous contact, whereas the periclase forming reaction front extends up to 4m into the host rock. The carbon and the oxygen front are located at the same place, which would require an X(CO2) of 0.5, if the isotope fronts are interpreted using a standard infiltration model, even if modelled as the side of a front. A similar amount of reaction progress, calculated from measured volume of periclase (corrected for retrograde brucite formation), was found over the entire profiles. Surprisingly, dolomite is still present as prograde leftovers in most samples demonstrating that reaction did not go to completion. The median grain size of periclase crystals remains constant over both profiles. Nevertheless, CSD\'{ } s flatten systematically, reflecting a larger proportion of bigger grains with increasing distance from the contact. We interpret variations in grain sizes to be the result of changing reaction affinities along an infiltration front flattened (dispersed) by diffusion/dispersion and kinetics. A numerical model is presented, based on the textural analyses and geochemistry data from the field, describing the dynamic nucleation and crystallization of periclase in this infiltration driven system.
V23A-1228
The Olivine Enigma: Why Olivine Does Not Affect Seismic Velocities in Oceanic Gabbros
Seismic velocities in oceanic gabbros are expected to increase with increasing olivine content because of the very high elastic moduli of Fo90, but this relationship has not been confirmed by laboratory studies of gabbros recovered by scientific ocean drilling. An inversion of modal analyses and measured elastic moduli based on a Voigt-Ruess-Hill (VRH) model shows that the effective elastic moduli of olivine in these rocks are anomalously low relative to olivine with a composition of Fo90. This anomaly might be explained by the low Mg numbers of the recovered oceanic gabbros, by a high density of open cracks in the olivine, by fine-scale alteration of olivine, or by a combination of these factors. The magnesium contents olivine grains in the gabbros have highly variable, but lower Mg numbers (median Fo73) and hence somewhat lower moduli than those of olivines that are typical of upper mantle rocks (~Fo90). SEM analyses indicate that cracks in the olivine grains are filled with talc/serpentine. This observation suggests that the cracks are filled in situ and hence not in themselves a cause of low moduli. However, cracks containing talc and/or serpentine make up 5-10% of the olivine grain volume; with additional alteration along the margins of some grains, total alteration ranges from 5 to 30%. More detailed electron microprobe analyses of selected gabbro samples from ODP Hole 923A indicate that olivine grains in these rocks are typically 10 to 20% altered to talc and serpentine, plus small amounts of tremolite and magnetite. Numerical (VRH) modeling indicates that the low Mg numbers and about 10% alteration of olivine to talc and/or serpentine in these rocks will account for the observed low effective olivine moduli and for the fact that there is no observed correlation between the olivine content and measured seismic velocity is oceanic gabbro samples. Thus, seismic velocities are not indicative of olivine content in a gabbroic lower crust owing to fine-scale in situ alteration of olivine to talc and/or serpentine.
V23A-1229
Transition Metal Systematics of Opx-Enriched Harzburgites From the Cascades Arc With Implications for the Origin of Cratonic Peridotites
A number of peridotite xenoliths collected from the Simcoe volcanic field region of the Cascades arc exhibit notable enrichment of modal orthopyroxene. The process driving this enrichment is most likely metasomatism of the mantle wedge by Si-rich fluids derived ultimately from the underlying slab. By investigating the resultant elemental systematics associated with subduction zone metasomatism of this type, we hope to shed light on the origin of other opx-rich peridotites, such as those seen in many cratonic xenolith suites. The xenoliths found in the Simcoe volcanic field provide a rare opportunity to examine the composition of sub arc mantle, as it is unusual to find mantle xenoliths in volcanic arc lavas. The samples were analyzed using laser ablation ICPMS and their bulk compositions were reconstructed from point-counted mineral modes. Two-pyroxene mineral thermometry of the samples yield temperatures of approximately 1000 degrees C, corresponding to a depth of origin at uppermost mantle pressures if typical arc geotherms are assumed. Most of the peridotites are harzburgites or olivine-orthopyroxenites (Mg#s 0.88-0.9; opx mode 0.15-0.9), with small amounts of clinopyroxene (<0.02). Clinopyroxenes are significantly enriched in the light rare earths, consistent with a metasomatic origin for these opx-rich harzburgites. Of note is the counterintuitive systematics of Zn. Whole-rock Zn decreases with opx, but Zn in olivine also decreases with opx mode while Zn in opx increases with opx mode, hence the decrease in whole- rock Zn is not simply due to mechanical segregation of harzburgite into opx- and ol-rich zones. In summary, the REE signatures suggest the subducting slab as the most likely candidate for the source of the fluids that caused the opx enrichment. The opx-enrichment itself and the unusual trends in Zn suggest a reaction between a silicic fluid and normal harzburgite. Moreover, the concomitant decrease in olivine and whole-rock Zn with opx mode suggests significant leaching of Zn from the peridotite during this process. Because the bulk partitioning of Zn in anhydrous peridotite melting is unity, low Zn contents are anomalous. The best explanation for these low values is that Zn partition coefficients decrease in hydrous environments. Many opx-enriched Archean cratonic peridotite xenoliths have anomalously low Zn contents, supporting the suggestion that such peridotites formed in arc environments.
V23A-1230
The Paddy's Flat Gold District (Murchison, Western Australia), insight on a rheologically and structurally control lode gold deposit
In the North-West of the world class gold producer Yilgarn craton lies the Murchison Goldfield (Western Australia). Several rich deposits mined since the turn of the 19th century suggest that this late-Archaean terrane is highly prospective for gold exploration. Located in the northern part of the Murchison goldfield, the Paddy's Flat District recorded a production of ~2 Moz. The Paddy's Flat Gold District area is located on an N-S trending, steeply east dipping shear on the north western limb of an upright isoclinal fold known as the Polelle syncline. It is hosted in a strained and metamorphosed volcanic and minor sedimentary sequence. Mineralisation within the Paddy's Flat District is not limited to a particular rock type and it is associated with a wide range of lithologies including: (i) intermediate to felsic extrusive volcanics, (ii) intrusive porphyrytic microgranite, (iii) ultramafic volcanic assemblage and (iv) iron rich banded chert. Instead, the mineralisation is highly structurally controled by the Paddy's Flat shear zone which provided fluid-pathways for the mineralising fluid. Regardless of the lithological composition, the mineral assemblage associated with the gold mineralisation consists of quartz, carbonate (ankerite, siderite), sulphides (pyrite, arsenopyrite), fuchsite and sericite assemblage suggesting the introduction of an oxidizing fluid rich in Au, S, As and K ± SiO2. At Paddys Flat the mineralisation is spatially associated with rheologicaly competent units or with domains located at the boundary between rheologically distinct lithologies within the main shear. The fluid pathways provided by the shear zone and the rheological contrast are suggested to be key parameters in the formation of economic lode gold deposit in the Paddy's Flat district.
V23A-1231
Fluid-Rock Interaction in the Miocene Tejeda Caldera, Gran Canaria, Canary Islands
Rhyolite-trachyte tuffs deposited within the Miocene Tejeda caldera (Mogán Group 14-13.3Ma) show evidence of severe fluid-rock interaction. The altered tuffs are restricted to a peripheral zone directly inside the caldera margin, and occur at four distinct horizons within the mid-upper Mogán ignimbrite succession. Upper-Mogán tuffs display pervasive intermediate argillic alteration (smectite+illite+zeolites+adularia) and silicification (microcrystalline quartz+amorphous silica), indicative of low-temperature (≤250°C) near-neutral pH conditions. Quartz+kaolinite+muscovite+chlorite+calcite alteration of mid-Mogán tuffs and breccias may reflect boiling of higher-temperature, acidic hydrothermal fluids at depth. Si+Na+K+Pb+Sr+Rb were highly mobile during fluid-rock interaction, whereas Ti+Zr+Nb were dominantly refractory. Altered intra-caldera tuffs (n=65) have higher δ18O values than equivalent unaltered extra-caldera ignimbrites, reflecting an overall low-temperature near-surface environment in which meteoric water (δD ca.-15‰, δ18O ca.-3‰) was the dominant fluid source. A decrease in δ18O from upper- to mid- Mogán altered tuffs is consistent with an increase in fluid temperature with depth. Unaltered ignimbrites have δD values of -110 to -168‰ (n=6) and ≤0.2wt% H2O, indicative of Rayleigh-type H2O-exsolution. Altered tuffs have δD values of -52 to -117‰ (n=75) and up to 4wt% H2O, reflecting interaction with steam (δD\ll-15‰) or an evolved low-δD fluid. Apparently unaltered ignimbrites between altered horizons (n=13), and shield basalts directly outside the caldera margin (n=6), have elevated δD and H2O values relative to equivalent unaltered rocks, indicative of minor alteration. Supported by numerical modelling, our Gran Canaria data reflect an intrusion-related, structurally controlled epithermal system, in which fluids and/or vapours migrated through intra-caldera tuffs via channelised, porous flow. This study may help to unravel the complex processes of fluid-rock interaction characteristic of both active and fossil caldera-hosted epithermal systems that are presently inaccessible or poorly exposed.
V23A-1232
The Aeolian Volcanic Arc: New Insights From Subduction Zone Thermal Models and Mineral Solubility Scaling Relationships
The Calabrian subduction zone, situated southeast of the Italian ‘boot' in the Ionian Sea, is the latest manifestation of African-Eurasian plate interaction. This plate interaction has been remarkably dynamic since the Mesozoic, hosting episodes of mountain belt and volcanic arc formation including, for example, the Alpine, Carpathian and Apennine orogenic belts and Hellanic and, most recently, Aeolian volcanic arcs. Subduction of cold oceanic lithosphere beneath Europe initiated around 80 Ma, and the last 30 Ma have been characterized by alternating episodes of rapid back-arc rifting and back-arc spreading (up to 6-8 cm/yr) mediated by dip-parallel and/or trench-parallel tears in the descending slab resulting from differential trench rollback (Wortel and Spakman 2000). Backarc extension effectively moved the plate boundary from the European continental margin in the north to the African continental margin in the south, creating the modern Western Mediterranean basins. The Tyrrhenian oceanic basin was opened during the latest episode of trench rollback, from 5-2 Ma, followed by initiation of the subduction-related Aeolian volcanism by 1.3 Ma (Beccaluva et al. 1982) and complete cessation of extension of the overriding plate around 0.8-0.5 Ma (Goes et al. 2004). The seven subaerial volcanoes of the Aeolian volcanic arc sit atop thin (16-30 km) continental crust, and collectively tap a heterogeneous mantle source. Slab geometry in the depth range of 150 to 500 km has been refined using the hypocenter relocation procedure of Engdahl et al 1998 for teleseismic events beneath the Tyrrhenian Sea, in conjunction with recent tomographic results. The thermal state of the Calabrian subduction zone at depths relevant to dehydration and magma genesis has been investigated using a 2-dimensional time-dependent thermal model of the descending slab and convecting mantle wedge based on seismic, geologic and geodetic observational data. Modeling methodology follows van Keken et al. 2002 and includes power-law olivine rheology and shear-heating scaled to fit forearc heat flow. The computed thermal model is used with a thermodynamically-based mineral solubility scaling relationship between fluid solute content and the relative permittivity of the fluid to infer the solute content of a slab-derived aqueous fluid existing in equilibrium with slab and mantle wedge mineralogy. The relative abundances of different solutes in slab-derived fluids are critical in determining the trace element characteristics of subduction zone magmas. Among some of the Aeolian arc volcanic centers, the recent (<30 ky) evolution in magma composition from calc-alkaline to high-potassium has been variably attributed to melting of distinct source regions, or the increasing abundance of aqueous fluids in a source region associated with a young/incipient arc. Our approach of thermal-field and solute content modeling can contribute to the resolution of this and other petrogenesis questions for Aeolian arc lavas.
V23A-1233
New Models of Crustal Fluid Flow Incorporating Magmatism and Porosity Evolution During Orogenesis
We present a 2d numerical model simulating the flow of fluid in an orogenic overthrust setting during magmatic intrusion and metamorphic devolatilization. The model is intended to test the existing hypotheses on the direction of the regional fluid flow in the deep crust. Advective (single-pass, pervasive or focused), convective (multi-pass), and sub-horizontal (up-temperature) fluid flow regimes have been proposed on the basis of various geological observations and models. Our numerical simulation reveals complicated local fluid flow patterns, with fluid moving upward and downward, as well as in the direction of decreasing temperature. Most of the fluid flux is concentrated along the fault zone and around magmatic intrusions, which take the form of repetitive basaltic sills. The sills have variable permeability for fluid depending on their temperature, and may serve as sources/sinks for volatiles and as a barrier for fluid flow, thus introducing anisotropy to our model. In addition, we present a "next- generation" numerical model for fluid flow in a deformable two-phase (fluid, solid) media with visco-elastic rheology. The necessity of the new model for crustal fluid flow is indicated by the difficulty that traditional one- phase formulations have in reproducing the near-lithostatic fluid pressure in the deep crust with realistic rock permeability and in the absence of extra fluid sources. As fluid motion is driven by the gradients in fluid pressure, the use of two-phase models with solid-fluid interaction and evolving porosity may be critical for numerical simulations of deep crustal fluid flow. We test the traditional one-phase formulation by comparing the fluid patterns inferred from the one-phase and the two-phase models. Furthermore, we use the two-phase formulation to reconstruct the profile of the fluid pressure and associated matrix permeability within the model crustal section, and to estimate the characteristic time scales of the fluctuations in fluid pressure due to hydrofracturing and fracture healing.
V23A-1234
Major and Trace element Geochemistry of Hydrothermal Minerals From the Sudbury Structure: Implications for Fluid Architecture
The Sudbury structure mainly comprises a layered intrusive body (the 2.5 km thick Sudbury Igneous Complex; SIC), and the overlying 1.4 km thick Onaping (OF) and Vermilion (VF) Formations. This sequence and the underlying basement have been affected by extensive hydrothermal fluid circulation, postulated to involve deep formational brines, sea water, and/or orthomagmatic fluids derived from the SIC. Late-stage magmatic fluids are indicated by miarolitic cavities (Kfs-Qtz-Ep-Chl-Hbl), which are common throughout the upper part of the SIC but also occur in the overlying Onaping Intrusion (OI) and towards the footwall of the SIC. Mineralogically similar veins (Ep-Qtz-Kfs-Chl) also occur in the SIC and OI, and possibly represent the migration of these orthomagmatic fluids away from their source. Hydrothermal epidote shows similar major element chemistry (e.g. Fe/Fe+Al) irrespective of epidote type or stratigraphic position. In general, epidote from the SIC displays average Fe/Fe+Al ratios from of 0.16 to 0.32. Similar values (0.19 to 0.32) are reported for epidote from the OF. Epidote from cavities are characterized by relatively flat HREE patterns, variable degrees of LREE fractionation, and moderate to strong positive Eu anomalies (Type I). However, some show a relatively flat LREE pattern, slightly enriched HREE and moderate positive Eu anomalies (Type II). A third group (Type III), show steep REE patterns with moderate to strong fractionation between HREE and LREE, and moderate positive Eu anomalies. Vein epidote is characterized by Type I patterns only. These data suggest that fluids of comparable temperature and composition precipitated epidote and related minerals in both miarolitic cavities and veins through the SIC and OF. Calcite occurs as a replacement mineral in the OF, in amygdules (Qtz-Chl-Cal-Ep-Sulph) in syn-depositional aphanitic dikes in the OF, and as bands, discordant aggregates and replacement phases (blocky calcite) in the VF, which lies above the OF. Amygdule calcite displays either relatively flat REE patterns, with minor LREE fractionation and no Eu anomalies (not observed in Ep), or the Type III patterns seen in epidote. Replacement carbonate in the OF show the Type I and II patterns characteristic of epidote. Banded and unmineralized blocky calcite displays flat, yet irregular, REE patterns, with minor negative Ce anomalies, and no Eu anomaly. Discordant and blocky calcite associated with minor sulphide mineralization displays moderate LREE fractionation and flat to slightly enriched HREE patterns, with a strong positive Y anomaly and negative Ce and Eu anomalies, suggesting seawater involvement. The REE geochemistry of VF carbonates is thus distinctly different from that of epidote and carbonate deeper in the system. Similar REE patterns observed in hydrothermal epidote and carbonates suggest that fluids of similar composition and temperature circulated through both the SIC and OF. However, the distinct character of some sulphide-related VF carbonates suggests the involvement of different fluids, or deposition under different conditions. The distinct positive Eu anomaly in epidote and carbonate is consistent with precipitation from relatively hot, reduced fluids. Variations in REE chemistry within single samples suggests that the REE systematics are complex and that fluids evolved with time. Finally, a potential fluid connection between the footwall and hanging wall of the SIC, as has been previously suggested, is potentially demonstrated by the ubiquitous Eu anomaly found across the entire Sudbury hydrothermal system.
V23A-1235
Distribution of boron and lithium in serpentinites using laser ablation ICP-MS
Boron and lithium are fluid-mobile elements that can be used as tracers for seawater and slab-derived fluids in subduction zones. Hydrothermally altered oceanic lithospheric mantle (in the form of serpentinite) may be the major sink for seawater-derived B and Li. In this study, we investigate how these elements are distributed among the primary and secondary minerals in variably serpentinized peridotites, helping us understand how B and Li may be released in slab-derived fluids during subduction. Serpentinite sections from the Feather River Ophiolite in California were mounted in epoxy and then hand-polished. They were then analyzed for major elements and certain trace elements, including Li and B, using a ThermoFinnigan Element II laser ablation ICP-MS. These measurements complement bulk rock measurements done by ICP-AES and presented in another paper. All samples were shown previously to have substantial bulk-rock enrichment of B and Li, indicating an external contribution from fluids, perhaps seawater. We examined two lithologies on the micron scale: partially serpentinized harzburgites and serpentinite veins. Relatively fresh olivine has 5-10 ppm B, which is much higher than that expected for the mantle, but still lower than the bulk rock B. B contents increase up to 50 ppm with serpentinization of olivine. Some serpentinized pyroxenes also show B contents up to 50 ppm, while others are not enriched, having about bulk rock B. The behavior of Li is more complicated. Fresh olivine was found to be enriched in Li, containing from 0.5 up to 2.5 ppm. Serpentinized olivines lost Li and mostly contained less than 0.5 ppm, generally close to or slightly less than bulk rock. Serpentinized pyroxenes are highly enriched in Li, containing up to 10 ppm. Collectively, our data confirm the notion that serpentinites are sinks for B and Li. Our data suggest, however, that much of the bulk rock enrichment in B and Li is stored in the unaltered olivines as well as in the serpentine. These observations suggest that significant amounts of B and Li (particularly Li) may be retained in the subducting slab, even after serpentine breakdown. Thus, some B and Li will be liberated into the mantle wedge and transferred to arc magmas, but a non-negligible fraction will not be liberated and hence will be subducted deep into the interior of the Earth.
V23A-1236
Long-term and short-term erosion rates in river catchments of the Rhenish Massif and the Black Forest, Germany
We constrained long-term erosion rates from the concentration of cosmogenic 10Be in stream sediments in order to quantify the Late Quaternary denudation history of mountain ranges in central Europe. Four different catchments in Germany, ranging in size from 8 to 379 km2 were investigated. Two of them, the Aabach and Möhne catchments drain predominantly low-grade Paleozoic metasediments. The other two, the Gutach and Acher catchments in the Black Forest are situated in Late Paleozoic granites. Erosion rates derived from the 10Be concentrations range from 29 to 86 mm/ka in the Rhenish Massif and from 26 to 91 mm/ka in the Black Forest. These spatially-averaged erosion rates integrate over the past 7 to 23 ka. Central to our investigation are questions concerning the relative importance of lithology and catchment relief on long-term erosion rates. Short-term erosion rates for all catchments were quantified by combining the amounts of suspended and dissolved loads in water samples with water discharge data and basin area. By analyzing the stable isotope signatures δ18O of river water and δ13C of dissolved anorganic carbon and by taking into account the precipitation and evaporation we corrected the dissolved load for organic, atmospheric and anthropogenic inputs. The preliminary short-term erosion rates vary between 9 and 33 mm/ka and are only about one third of the erosion rates derived from 10Be. The short-term erosion rates are complemented by erosion rates derived from the volume of sediment stored behind reservoirs of known age. These erosion rates range from 2 to 13 mm/ka and are lower than the erosion rates derived from river loads, as they do not take into account the dissolved load. Furthermore, we focused on dependence of lithology and land use on short-term erosion rates.
V23A-1237
Geochemical and Isotopic Estimates of Eolian Dust in Soils of the San Juan Mountains, USA.
Eolian dust deposition in the San Juan Mountain Range in southern Colorado has increased 5-7 fold in the past two centuries. This dust deposition contributes an exogenous supply of biologically relevant elements such as Ca, K, Mg, and P to these alpine ecosystems in the form of fine textured mineral particulates. The deposition of eolian dust may be an underestimated factor of soil formation and soil chemistry in these alpine settings. The importance of eolian dust relative to the weathering of local bedrock likely varies across bedrock types. The San Juan Range is geologically diverse with distinct regions of Meso-proterozic crystalline granites in the Weminuche Wilderness, Mesozoic sedimentary layers near Molas Pass in the San Juan National Forest, and Tertiary volcanic geology found on Red Mountain Pass in the Uncompahgre National Forest. Principle component analysis of element chemistry shows that bedrock and soils from these sites cluster by geology. In addition, these groups are chemically distinct from eolian dust collected from snow in the San Juan Range. Several elements seem to drive the difference of dust from soils and bedrock including Ca, Sr, Cu and Cd. The purpose of this research was to estimate the relative contribution of eolian dust to alpine soil element pools in the San Juan Mountains across a range of local geologic parent material. A calculation of element mass- balance shows that Cu and Cd are enriched in the surface soils of both volcanic and sedimentary soils relative to concentrations in local bedrock. However, Ca is enriched only in volcanic soils. These observations support the notion that eolian dust contributes to soil formation and that the relative contribution of dust across the landscape varies with geology. In addition to element mass-balance estimates we utilize Sr and Nd isotope measurements of soil, bedrock, and dust to further constrain the importance of eolian dust to these alpine soils.
V23A-1238
Solute Export and Carbon Dioxide Trends of the Altamaha River Basin
The Altamaha River Basin (ARB), a major drainage of the Atlantic seaboard, was monitored near Jesup, Georgia, on a biweekly frequency during April 2006 through June 2007. Alkalinity, pH, temperature, and Total Dissolved Solids (TDS) were measured in the field. Geographic Information Systems and historical precipitation and stream flow were used to calculate mean annual: (1) precipitation flux into the basin; (2) runoff from the basin; (3) solute export; and (4) partial pressure of carbon dioxide (pCO2). The mean annual precipitation flux to the ARB is 46.4 km3. Runoff near the mouth of the Altamaha River is 12.1 km3 or about 26% of the precipitation input, implying that roughly three quarters of all the rain that falls on the basin is lost to evapotranspiration. The ARB has discharge-weighted average TDS concentration of 53.3 mg/L and annually exports 646,445 metric tons of dissolved solutes to the Atlantic Ocean. The pCO2 values range between close to atmospheric equilibrium in May to 106 times above atmospheric equilibrium in December. For the rising limb of the hydrograph, pCO2 trend mimics stream flow with pCO2 peak occurring about a month ahead of discharge. The low values probably indicate CO2 drawdown by aquatic photosynthesis and higher values indicate discharge of wastewater/shallow groundwater charged with bacterially respired carbon dioxide. Thus, the ARB was a net source of carbon dioxide to the atmosphere during the water year 2006-2007.
V23A-1239
Variation of U-Th disequilibria in the different size fractions of river sediments: Evidence from the Himalayan river system.
U-series disequilibria in river sediments have the potential to bring main constrains on the time scale of erosion processes at the scale of watershed. However such a determination requires to characterize and to understand better the variation of 238U-234U-230Th disequilibria in sediment following their nature and/or their size grain. In order to answer this question, we analysed U-Th disequilibria in sediments collected at different depths of the water column of the the Ganges river and one of its main tributary: the Gandak river. We use the natural mineralogical sorting done by the river on the carried sediments to evaluate the sediments U-Th variations as a function of their grain size. These data highlight the significant variations of U-Th disequilibria with the grain size of sediments, implying a difference in the source and transfer time of sediments with their granulometry. For the Ganges river system, it can be proposed from 238U-234U-230Th data that bedload and river-bank sediments originate from Himalayan range and are marked by long transfer time into the plain. By contrast 238U-234U-230Th disequilibria in suspended sediment are explained by a scenario involving an old U gain followed by a recent U-Th fractionation. The U-gain is related to U vegetation recycling in the plain, which is all the more intense than the sediments are far from the high range. Overall, these results indicate that a large proportion of MES in the Ganges river system come from the erosion of soils horizon in the plain, and that their transit time on the plain is short. The detailed study of U-series disequilibria in the different size fraction of river sediments seems therefore important for a correct determination of the transport laws of sediments in alluvial plain.
V23A-1240
Use of B Isotopes to Investigate the Influence of the Vegetation on the Rate of Soil Weathering
The rate at which soil minerals are weathered and the influence of the vegetal cover on the intensity and the nature of the reactions that release solutes from parent rocks to natural waters are still poorly understood. A way to investigate interactions between soil weathering and plant development is to focus on geochemical tools that are distinctively affected by these two processes and to monitor their behavior along yearly time scales, relevant to the soil/plant system. According to this approach, boron isotopes are particularly interesting because they undergo great isotopic fractionation during water/rock interactions and they are micro-nutrients essential for the life development. To monitor seasonal variations of the B fluxes in the soil/plant system, we benefit from the equipped experimental Strengbach catchment (Vosges, France) where vegetation samples, rainfalls, throughfalls, soil solutions at various depths (5, 10, 30, 60 cm) and spring waters are collected every 6 weeks from 2003 up to now. The observed B fluxes at the watershed scale reveal that it is controlled by the vegetation cycling because the amount of B involved in the vegetation cycling if about 5 times greater that its discharge at the outlet of the basin. Analyses of the B isotopic compositions of vegetation samples and throughfalls reveal unexpected isotopic fractionation that makes the vegetation-related B fluxes quit distinguishable from those controlled by water/rock interactions. The fact that the B geochemical cycle is regulated by the vegetation cycle together with an easy-to- follow isotopic signature makes then possible to determine the relative part of the vegetal activity from the soil weathering rate. We developed a 1D model of the B bio-geochemical cycle at the soil/plant scale that takes into account the transient features of the vegetation cycling and the water transport in soil. The results of the model reveal distinct reaction zones in soils according to its mineralogy and the ratio between the intensities of the vegetal activity and mineral weathering. This approach makes then possible to calculate the rate at which B is released from soil minerals and offer the opportunity to go further in the characterization of the controls and feedbacks that regulate interactions between soil and plant. http://ohge.u-strasbg.fr/indexuk.html
V23A-1241
Tropical volcanic islands: best first cycle erosion laboratories.
Lesser Antilles islands are located in a tropical climate with high temperatures (24 to 28 ° C), high precipitation (can reach 12 m/yr), very dense vegetation, sharp relief with a NS gradient of age. Antilles rivers have a torrential hydrologic regime with extreme erosion conditions. Chemical weathering rates are among highest world values 140-200 t/km2/yr on surface and 2 to 5 time higher when subsurface water circulations are taken account (Rad et al., 2007). Timescales and physical rates of erosion are calculated from U-Th isotopic compositions. Our results, with 3 different methods (all based on mass budgets between the river bedrock and its erosion products) show a good agreement for physical erosion rates with a maximum value of 2500 t/km2/yr. We show that solid loads of these volcanic rivers are dominated by sand, which represent more than 80% of the soil profile. Erosion rates are directly correlated to the age of basins. Indeed, among all parameters (climat, runoff, slopes, vegetation…) the age basins is a key control parameter. The younger the basin is the higher the weathering rate is. This correlation asserts that younger volcanic rocks are more easily weathered than old ones: young fresh material is easily mobilized by erosion, while for older rocks with thick soil covers, chemical and physical erosion rates are much lower, they reach a threshold with low chemical and physical rates once soils are constituted. It seems that rivers draining young lava flows are also the ones, which are not at steady state with the lowest mechanical denudation rates. Erosion processes in Lesser Antilles are typical of a first cycle erosion with simple relationship between rates, timescale of erosion and basins ages. It appears that first stage of erosion are characterized by high chemical denudation rates and low physical denudation rates, the erosion products are then close to the bedrock. It is then followed in a second stage by constant chemical weathering rates with higher mechanical denudation rates. \small Setareh Denise Rad, Claude Jean Allegre and Pascale Louvat, 2007, Hidden erosion on volcanic islands, Earth and Planetary Science Letters. In Press.
V23A-1242
Nutrient Sourcing of Ten Plant Species in the Southwest U.S. using Strontium Isotopes: Effects of Rooting Depth, Bedrock Type, and Landscape Age
For decades, researchers have been examining chronosequences in Hawaii to quantify mineral weathering rates and tropical plant nutrient pools. Within El Malpais National Park, New Mexico, well-dated basalt flows allow for comparison of the Hawaiian data to a semi-arid ecosystem. We measured 87Sr/86Sr ratios in cellulose and bedrock to gauge tree, shrub, & grass (Pinus ponderosa, Pinus edulis, Juniperus monosperma, Juniperus scopulorum, Populus tremuloides, Chrysothamus nauseosus, Fallugia paradoxa, Rhus trilobata, Bouteloua gracilis, and Xanthoparmelia lineola (Berry) Hale) dependence on atmospheric dust as a nutrient source. Sampling sites varied by bedrock type (limestone, sandstone, granite, cinder and basalt) and by age (Quaternary to Precambrian) providing a wide and discrete range of 87Sr/86Sr ratios. Thus, we can pinpoint the roles landscape age (3 ka to greater than 200 ka) and bedrock recalcitrance play in mineral weathering versus eolian dust influence. This study suggests that dust dominates the nutrient cycle on younger landscapes (3 ka), shows a mixture of mineral weathering-dust inputs by 9 ka, and is rock-dominated by 120 ka. Rates of soil nutrient depletion vary in older, non-basalt landscapes (>250 ka), depending on the type the parent bedrock. For example, landscapes on Precambrian gneiss and Paleozoic limestone still show significant mineral contributions while the quartz-rich, carbonate-cemented Zuni Sandstone is almost completely eolian-dominated. Cellulose 87Sr/86Sr variation by plant species at a single site allows us to monitor plant rooting depths and interspecies competition for vital nutrients. Within semiarid ecosystems, nutrient concentrations exhibit both vertical and lateral heterogeneity. The reasons for this variation include vertical and lateral heterogeneity in soil moisture and foliar trapping of nutrient-rich dust followed by incorporation of the throughfall into the underlying soil. This study shows that throughfall does play a significant role for certain species (e.g. J. monosperma) but not for others. A species' ability to trap dust and its overall rooting depths both influence its nutrient intake.
V23A-1243
Volcanic CO2 as a major agent of weathering in volcanic regions
According to high erosion rates, weathering of volcanic areas is one of the main processes controlling the atmospheric CO2 levels (e.g. Louvat, 1997, Dessert et al., 2003). So far, the origin of CO2 was assumed to be mainly atmospheric. The origin and consumption rates of CO2 can be estimated using concentration of Dissolved Inorganic Carbon (DIC ~ bicarbonates) in the rivers and carbon stable isotopes. The processes governing the CO2 consumption by chemical weathering were studied in four volcanic areas, with climates from tropical to sub-polar, in different geodynamic contexts. Lesser Antilles, Reunion, Iceland and French Massif Central are ideal sites for the study of weathering due to the gradients of rainfall (up to 14m/yr), rock ages (tertiary to subactual) and volcanic activities, inducing variable weathering rates (50- 400t/km\2/yr). δ13CDIC and major elements chemistry were studied in streams, springs and soil solutions of these four areas. The δ13CDIC and major elements concentrations are highly variable, and allow us to identify the origin of DIC as a mixing between biogenic CO2 (average value of δ13C ~ - 29.1±2.0‰ in tropical areas, -26‰ in temperate to sub-polar climate) and volcanic CO2 (δ13CDIC = -5 to 3‰). We found that volcanic CO2 is a major source of carbon, the highest contribution being for zones with high infiltration of water (up to 100\percent of DIC for Piton de la Fournaise in Reunion). As inferred by Rad et al. (2007) from major elements, this isotopic study of the DIC cycling in the river demonstrates the importance of the volcanic fluids on chemical weathering processes.
V23A-1244
Geochemical Evolution of Loess-derived soils Predicted Using a GCM and a Reactive Transport Model
A North-South transect along the Mississippi River valley provides an opportune environmental gradient across which to investigate chemical weathering. Soil profiles along this transect are interpreted to have developed from a uniform parent, the Peoria Loess, with pedogenesis commencing between 13 - 10 14C ka BP. At the pedon scale, we examine mineral evolution in these soils based on XRF elemental and mineralogical analysis. To interpret concentration changes as a function of depth and their relation to climate variations along this transect, we simulate climate at specific times during the Holocene using the GENESIS global climate model (GCM). Model outputs of predicted temperatures and moisture fluxes from the GCM are used to drive the WITCH model to calculate chemical weathering as a function of time. The WITCH model describes mineral dissolution/precipitation based on laboratory kinetic rate laws. Additionally, this model calculates soil CO2 levels at varying depths. We compare observed to calculated soil profiles to investigate questions related to how temperature and precipitation drive weathering over the last 13 ka. This is the first example of the use of a GCM to drive a geochemical code to predict soil evolution.
V23A-1245
Effect of Free Energy and Dislocation Density on Calcite Dissolution Kinetics
Mineral dissolution plays a central role in regulating various geochemical processes at both global and local scales and consequently has been a vital research subject in geosciences. To date, the accelerated buildup of atmospheric CO2 is spurring a growing interest in understanding carbonate dissolution. Significant progress has been made on this topic through intensive studies in the past decades. Yet, the dissolution behavior of carbonate minerals remains controversial when it comes to the relationship between kinetics and free energy as well as the significance of dislocation density in the overall dissolution processes. Here, we investigate the dissolution of calcite at near and far from equilibrium conditions to examine the controlling factors for the reaction kinetics. Mixed-flow experiments were conducted at various saturation states under room temperature. Fragmental and powdered calcite samples with different defect densities were used to explore the effect of dislocation density on dissolution kinetics, and geometric surface area of the mineral grains was used to normalize the dissolution rate. The experiments were conducted in both closed and open (to air) settings. Experimental observations show a highly nonlinear dependence of dissolution rate on the Gibbs free energy. Furthermore, dissolution rates do not seem to be affected by dislocation density near and far from equilibrium. Finally, dissolution rates measured at same saturation conditions are similar regardless the experimental settings, closed or open. These results suggest that: (i) The classic TST model may not be sufficient to depict the relation between dissolution rate and Gibbs free energy for calcite in all saturation conditions. The sigmoidal trend in the R- ΔG relationship indicates that, though the TST rate equation yields a fairly precise description of calcite dissolution kinetics when ΔG < -12 KJ/mol, it clearly overestimates the dissolution rate when the system approaches far-from equilibrium. (ii) The effect of high-energy surface sites associated with crystal imperfections may be overwhelmed by that of the growing steps generated by dissolution. (iii) The partial pressure of CO2 in ambient environment bears little importance to calcite dissolution once the saturation conditions of the solutions are fixed.
V23A-1246 [WITHDRAWN]
Surface Complexation Modeling of Radionuclide Sorption in the Saturated Zone of Yucca Mountain Rocks
The U.S. DOE is preparing to submit a license application to the Nuclear Regulatory Commission (NRC) to create a geologic repository at the Yucca Mountain, Nevada, for the disposal of spent nuclear fuel and high-level radioactive waste. In the event of a radionuclide release, the ground water beneath the Yucca Mountain is the primary medium through which most radionuclides might move from the geologic repository to the accessible environment. Sorption of radionuclides onto rock surfaces in the saturated zone of Yucca Mountain (SZ) is one of the most important processes retarding their release to the accessible environment. For this reason, a considerable experimental effort has been devoted over the last two decades to the measurements of sorption distribution coefficients (Kd) for various radionuclides in rock samples from the vicinity of the repository site at the Yucca Mountain. Despite the quantity and quality of the data, they are strictly valid only under the experimental conditions at which they were measured, whereas the Kd distributions used as inputs in performance assessment calculations need to represent the range of geochemical conditions and rock types expected to occur along the transport pathways. Hence geochemical modeling was used to calculate and predict chemical speciation of elements of interest in solid and solution under a variety of different conditions. The computer code PHREEQC v2.3 and the thermodynamic database PHREEQCDATA025.DAT were used for this geochemical modeling. The modeling provides a basis for extrapolating the experimentally derived Kd's, and provides improved understanding of the underlying sorption mechanisms, thus justifying and defending the Kd's selected for further radionuclide transport modeling development. This presentation focuses on the elements Am, U, Np and Pu which sorb in the SZ primarily via surface complexation reactions. We discuss quantitatively the influence of groundwater compositions, rock surface area, binding constants, and thermodynamic data on the sorption of the above mentioned radionuclides onto the minerals commonly present in the Yucca Mountain geologic setting. In addition, calculated sorption parameters are compared with the results of experimental data. The site specific probability distributions for Kd's selected as inputs to the numerical transport model of site-scale saturated zone beneath the Yucca Mountain are also presented.
V23A-1247
Selenate and Selenite Reduction by Nanometer-Scale Zerovalent Iron Particles
Selenium oxyanions can be present in agricultural drainage waters, coal mining effluent, and as fission products in radioactive wastes. The objective of this work was to evaluate the effectiveness of both nanometer scale zerovalent iron (nano-Fe) and 100 mesh Fe filings for reduction and immobilization of aqueous selenate Se(VI) and selenite Se(IV). The uptake of Se(VI) and Se(IV) using batch equilibrium, kinetics, and X-ray absorption spectroscopic (XAS) techniques was investigated. In addition, a thorough investigation of the solid phase corrosion products by X-ray diffraction was conducted. The crystalline corrosion product was similar to magnetite, though some distinct differences in the XRD results were noted between Se(IV)- and Se(VI)-treated samples. Application of quantitative X-ray absorption near edge spectroscopy (XANES) revealed that both Se(VI) and Se(IV) were reduced to a mixture of elemental Se(0) plus iron(II) selenide (Se(-II)). The Se local atomic structure in Se(VI)- and Se(IV)-treated nano-Fe was determined using extended x-ray absorption fine structure spectroscopy (EXAFS) and a Se-Se interatomic distance of 2.44 angstroms was revealed. This work suggests that nano-Fe is an efficient material for removing dissolved Se(VI) and Se(IV) from waste waters by formation of an insoluble, reduced FeSe product.
V23A-1248
Transport of Nanoparticles in Heterogeneous Systems: Methodology and Applications
Engineered nanoparticles are making their way into natural environment as a result of a growing nanochemical industry. However, processes that govern the deposition and transport of nanoparticles are yet to be understood and often complicated by technical difficulties in tracing these nanoparticles once they enter into and/or mixed with heterogeneous minerals in the natural environment. In this presentation, we discuss the transport of nanoparticles in heterogeneous systems such as in a mixed colloidal system of silica, ferrihydrite and natural organic matter in the light of new methodologies that may be used for such complicated systems. Transport of heterogeneously charged nanoparticles was investigated in column studies using quartz or iron-oxide-coated quartz as collector surfaces. Monodisperse SiO2 (~40 and ~80 nm) and ferrihydrite (~100 nm) particles and a natural humic acid was used. Results indicate that, even under favorable conditions, ferrihydrite nanoparticles show a conservative transport through oppositely charged quartz media when they coexist with the humic acid or with oppositely charged silica nanoparticles. The ratio of oppositely charged nanoparticles is critical in determining their mobility. Similarly, the transport and detachment of iron oxide nanoparticles from iron oxide- coated quartz were observed when humics were present in the feed solution. Our results imply that transport or co-transport of oppositely charged nanoparticles can occur simultaneously under conditions that are relevant to natural geochemical environment. Further studies are needed to understand detailed mechanisms and processes that govern the deposition and transport of engineered nanoparticles under realistic environmental conditions and in the presence of heterogeneous sediment collectors. New tools such as the used of various labeled nanoparticles (such as fluorescent-labeled nanoparticles) and neutron scattering techniques could be useful in studies of such complicated systems.
V23A-1249
Investigation of Li in Clay Interlayers at Different Temperatures by NMR
We used 6Li and 7Li MAS NMR to investigate the environmental changes of Li in the interlayer of clay minerals at room and high (250oC) temperatures. We also checked the usefulness 6Li NMR in studying Li in interlayer. Hectorite, Wyoming-montmorillonite, beidellite, and lepidollite were used for our study. 6Li NMR spectra show narrower peaks than those of 7Li NMR, but S/N ratio is low and there are no noticeable chemical shift changes, which makes it difficult to apply 6Li NMR to get information on Li environment in clay interlayers. 7Li NMR spectra, however, show changes in the peak width and quadrupole pattern, providing information on the Li environment in the interlayer, even though the change in chemical shift is not observed. In montmorillonite, two different environments of Li are observed, one having narrow peak with typical quadrupole pattern, and another having broad peak without that pattern. At high temperature, the broad peak becomes relatively, narrow which was also obserbed in the 7Li NMR spectra obtained from beidellite, but not hectorite. This changes are attributed to the coordination changes in the water molecules around Li which is tightly bonded on the basal oxygen of Si tetrahedra as inner-sphere complexes. The narrow peak in montmorillnoite can be assigned to the Li bonded as outer-sphere complexes.
V23A-1250
Generation of Oxidants From the Reaction of Nanoparticulate Zero-Valent Iron and Oxygen for the use in Contaminant Remediation
The reaction of zero-valent iron (ZVI) with oxygen can lead to the formation of oxidants, which may be used to transform recalcitrant contaminants including non-polar organics and certain metals. Nanoparticulate iron might provide a practical mechanism of remediating oxygen-containing groundwater and contaminated soil. To gain insight into the reaction mechanism and to quantify the yield of oxidants, experiments were performed with model organic compounds in the presence of nanoparticulate zero-valent iron and oxygen. At pH values below 5, ZVI nanoparticles were oxidized within 30 minutes with a stoichiometry of approximately two Fe0 oxidized per O2 consumed. Using the oxidation of methanol and ethanol to formaldehyde and acetaldehyde, respectively, we found that less than 2% of the consumed oxygen was converted to reactive oxidants under acidic conditions. The yield of aldehydes increased with pH up to pH 7, with maximum oxidant yields of around 5% relative to the mass of ZVI added. The increase of aldehyde yield with pH was attributable to changes in the processes responsible for oxidant production. At pH values below 5, the corrosion of ZVI by oxygen produces hydrogen peroxide, which subsequently reacts with ferrous iron [Fe(II)] via the Fenton reaction. At higher pH values, the aldehydes are produced when Fe(II), the initial product of ZVI oxidation, reacts with oxygen. The decrease in oxidant yield at pH values above 7 may be attributable to precipitation of Fe(II). The oxidation of benzoic acid and 2-propanol to para-hydroxybenzoic acid and acetone, respectively, followed a very different trend compared to the primary alcohols. In both cases, the highest product yields (approximately 2% with respect to ZVI added) were observed at pH 3. Yields decreased with increasing pH, with no oxidized product detected at neutral pH. These results suggest that two different oxidants may be produced by the system: hydroxyl radical (OH-·) at acidic pH and a more selective oxidant such as the ferryl ion [Fe(IV)] at neutral pH. This provides insight into the type of compounds that may be oxidized using the zero-valent iron and oxygen system. The addition of certain compounds such as oxalate and polyoxometalate (POM) may improve contaminant remediation efficiencies by enhancing oxidant yields. The introduction of 1 mM oxalate improved the formaldehyde yield by approximately 20% at neutral pH. Oxalate accelerates the Fenton reaction and limits the passivation of the ZVI surface by increasing iron solubility. The presence of excess POM greatly enhanced the yield of formaldehyde, with maximum yields of 60 and 35% with respect to ZVI added at pH 2 and 7, respectively. The mechanism of POM enhancement is a function of solution pH. At acidic pH, POM acts an electron shuttle by directly transferring electrons from ZVI to oxygen to increase the hydrogen peroxide production. At neutral pH, POM may act by forming soluble iron-complexes and preventing the build-up of an iron oxide layer on the ZVI surface.