T41B-1173 0800h
High pressure proton disorder in brucite
Brucite [Mg(OH)$_{2}$] is one of the simplest hydrous phases and serves as a model system for complex hydrogen bearing silicates of Earth's mantle. Brucite belongs to a class of M(OH)$_{2}$ compounds that appear to show unusual behavior at high pressure: the protons disorder while the MO sub-lattice remains crystalline. The nature and extent of proton disorder is uncertain. One can envision two types of proton disorder: 1) dynamic; the hydroxyl bond is tilted away from the c-axis and the proton hops between three symmetrically related potential wells or 2) static; each proton occupies a single well, and long-range order is frustrated by the hexagonal symmetry of the lattice. We explore the structure and physical properties of brucite over a wide range of pressure with density functional theory using the variable cell shape plane wave pseudopotential method in the local density (LDA) and generalized gradient (GGA) approximations. In this present study, we probe the energetics underlying the structure and dynamics of the proton sub-lattice by performing a series of constrained and unconstrained static calculations based on an energetically stable supercell wherein protons occupy 6i Wyckoff positions as opposed to 2d positions. We find that the equation of state and variation of lattice parameters of the superstructure with compression agree well with experiment. The displacement of the hydrogen from the three-fold axis (2d position) increases smoothly with increasing pressure. This means that even in the absence of dynamic disorder (i.e. at 0 K), the protons are frustrated and would be expected to exhibit long range disorder akin to a spin glass. In order to shed light on the dynamic nature of the proton hopping between the three energetically equivalent 6i sites, we have determined the activation barrier for such jumps. We find that the energy barrier increases with compression, possibly indicating a transition from dynamic proton disorder at lower pressures to static disorder at higher pressure. We have also investigated the possibility of proton jumps across the interlayer, by determining the potential energy well along the O\cdots\)O vector. We infer that proton jumps across the interlayer are either strongly limited, or highly cooperative since we do not find any evidence of a double well along the O\cdots\)O vector. This picture supports the view that brucite does not exhibit hydrogen bonding.
T41B-1174 0800h
Observational evidence for serpentinization of the uppermost mantle caused by bending related normal faulting offshore Middle America
Offshore Nicaragua new heat flow data show a profound deficiency in lithospheric heat loss seaward of the deep sea trench. Heat flow values decrease toward the axis. The decreasing values correspond to an increasing number of faults and fault throw of normal faults on the incoming plate as the plate approaches the trench axis. These faults are created while the ridgid lithosphere is forced and bended into the Middle America trench. We suggest that these faults support a vigorous hydrothermal circulation system in the incoming plate. In this area, new seismic refraction data suggest that both crustal and upper mantle velocities are clearly reduced with respect to the velocity structure of mature oceanic lithosphere. Reduced velocities in the crust are interpreted to be realted to crustal cracks and fissures caused by bending related faulting. Reduced upper mantle velocities are in accordance with upper mantle rocks which contain 5-10 percent serpentinites. Thus, in addition to pore and chemically bound water in the sediment and crust, altered upper mantle is an important agent to facilitate the flux of water into the deep subduction zone.
T41B-1175 0800h
Quantitative Polarized FTIR analysis of Trace OH in Populations of Disoriented Mineral Grains
Use of Fourier-transform infrared (FTIR) spectroscopy as an accurate, quantitative method to measure concentrations of hydrous species in minerals requires consideration of the anisotropic interactions of minerals with infrared light. This is normally accomplished by determining the orientation of a crystal in advance and then sectioning it perpendicular to its optic axes. Here we demonstrate a method that uses at least three randomly oriented grains, considered to be multiple samples of a homogeneous population. We explain the theory whereby (1) the orientations of the polarization vectors of measurements taken on these grains are determined by comparison to oriented standards of the same mineral, and (2) the principal-axis spectra of the sample are synthesized from the randomly oriented spectra. By comparison to complementary electron-backscattered diffraction (EBSD) data, we demonstrate that determination of orientations using the silicate overtone bands in FTIR spectra is accurate and precise. We apply the technique to determine the OH concentrations in a population of experimentally hydrated olivine grains.
T41B-1176 0800h
Electrical Resistivity of Dehydrating Serpentinite
The presence of fluids and melts in crustal rocks above subduction zones has often been attributed to the dehydration of subducting oceanic crust and hydrated mantle rocks. Dehydration processes have also been suggested to cause intermediate-depth earthquakes in the subduction zone. The most common mineral containing structurally bound water in the mantle is serpentine. By incorporating water into their crystal structure, serpentine minerals can store up to 13 wt% of water. This water is released when serpentines break down at elevated temperatures and pressures. In this study, we measured the electrical resistivity of dehydrating serpentinites and its evolution with time in laboratory experiments. Electrical properties of rocks respond highly sensitively to transiently available fluids, which are also detectable over very short time spans. The electrical resistivity measurements were performed at elevated pressure and temperature in a gas confining-medium pressure vessel at pressures of usually 600 MPa and temperatures up to 600\deg C, until a drop in resistivity was observed, indicating water was released from the rock. Our experimental setup was designed to be undrained, such that the water would be retained in the assembly. The resistance was determined with a high resolution impedance spectrometer with a spectrum of frequencies ranging from 100 mHz to 10 kHz. Several spectra were taken at each temperature step. In addition to the sample resistance, phase angles were recorded for each frequency. Generally, resistivity of the samples decreased by several orders of magnitude with increasing temperature, leveling off at about 300\deg C. The next drop in resistivity occurred at approx. 560\deg C, corresponding to the mineral phase change. Resistivities for all samples reached values of about 100 $\Omega$m, which is higher than expected for a fluid-saturated rock
T41B-1177 0800h
The Phase Diagram of Talc, 10A phase and 3.64A Phase.
The pressure-volume phase diagram for the Talc (Mg$_{3}$Si$_{4}$O$_{10}$(OH)$_{2}$), 10A phase (Mg$_{3}$Si$_{4}$O$_{10}$(OH)$_{2}$ xH$_{2}$O), 3.65A phase system has been determined using laboratory x-ray diffraction from quenched samples from a multianvil high-pressure cell and synchrotron x-ray diffraction from samples held in resistively heated diamond anvil cells. The talc data best fit a triclinic crystal structure while the 10A phase data can be fitted to a trioctahedral mica model structure. Details of the experimental techniques, phase diagram and implications for water transport into the upper mantle following subduction will be presented.
T41B-1178 0800h
Effect of Water Fugacity on Creep Strength of Anorthite
High-temperature creep of nominally water-free crustal silicates like quartz or feldspar is significantly affected by the presence of trace amounts of water. For water-saturated rocks the creep rate is expected to depend on water fugacity raised to a power of r. In this study we present first results on the fugacity exponent r and the activation volume V for creep of polycrystalline plagioclase. Fine-grained ($\sim$3 $\mu$m) anorthite aggregates were fabricated from a glass-powder by hot-isostatic pressing at 300 MPa pressure and a temperature of 1100$^{\circ}$C. To estimate the activation volume, 8 samples were subsequently dried ($\sim$ 0.003 wt% H$_{2}$O) and co-axially deformed at temperatures of 1125-1200$^{\circ}$C, pressures between 100 to 400 MPa, and axial stresses between $\sim$30 to 400 MPa, resulting in axial strain rates of about 7$\times$10$^{-7}$ - 5$\times$10$^{-4}$ s$^{-1}$. Sample deformation at constant load was dominantly accommodated by grain boundary diffusion creep for stresses below about 200 MPa. The resulting creep activation volume is about 24 cm$^{3}$mol$^{-1}$. A total of 20 samples were encapsulated in talc setting water activity aH$_{2}$O = 1. We used Fe, Ni or Cu jackets to buffer oxygen fugacity. Water content of samples after deformation was determined using Fourier-transformed infrared spectroscopy. Absorbance spectra suggest mainly molecular water located dominantly on grain boundaries. The water content is about 0.33 wt% H$_{2}$O. Wet samples were deformed at temperatures of 1000-1150$^{\circ}$C, confining pressures of 100 - 450 MPa, and stresses of $\sim$10 - 330 MPa, yielding strain rates of about 6$\times$10$^{-7}$ - 1$\times$10$^{-3}$ s$^{-1}$. In comparison to dry samples wet specimens are about 1.5 orders of magnitude weaker. In diffusion creep the fugacity exponent r is estimated to about 0.6. Samples buffered by Cu-CuO are considerably weaker than those deformed using Fe and Ni jackets. Since the strength of (dry) anorthite is independent of oxygen fugacity, our results suggest that hydrogen fugacity also affect the creep strength of wet plagioclase. The apparent hydrogen fugacity exponent is about -0.2.
T41B-1179 0800h
Neutron powder diffraction studies of small-volume samples synthesized at high P-T
The location of the hydrogen atom and the nature of hydrogen bonding in hydrous high-pressure phases are fundamental to the understanding of their thermodynamic, rheological and physical properties. However, neutron powder diffraction data have not been collected for many high-pressure phases because the amount of sample ($\sim$10-25 mg) that can be synthesized at extreme P-T conditions (e.g., 1300 $\deg$C and 16.5 GPa for wadsleyite) is generally insufficient for most neutron experiments. Multiple-run samples cannot be used for two reasons: 1) chemical variations between individual sample runs can be significant; and 2) synthesis of larger samples ($\sim$100 mg) is highly impractical due to the very high demand for experimental time on the multi-anvil presses. To determine the feasibility of using neutron diffraction to characterize the H bonding in phases with small sample volumes, powder data were obtained for topaz-OD [Al$_{2}$SiO$_{4}$(OD)$_{2}$] and orthorhombic, Fe-free OD-wadsleyite [$\beta$- Mg$_{2}$SiO$_{4}$] with the high-flux diffractometer (D20) at the ILL. Topazes found in ultrahigh-pressure metamorphic terrains are relatively OH-rich.Wadsleyite has been proposed as the major storage site for water in the transition zone. High-pressure Raman spectra measured at room temperature for orthorhombic, Fe-free OD-wadsleyite suggest that there are at least three hydrogen sites. Data collection times for topaz-OD were 1, 2, 3 and 9 h for samples that varied in mass from 17-26 mg. The OD-wadsleyite data were collected on a 10-mg sample for a period of 24 h. Rietveld refinement was carried out using GSAS (wavelength 1.3714 $\AA$; 2$\Theta$ range 15-$154\deg$). For topaz-OD, the best fit was obtained from data collected for 2 h on a 25-mg sample [750 $\deg$C 7.5 GPa 48 h] [wRp = 0.0538, CHI**2 = 2.348 with 37 variables]. Results confirm the X-ray model with deuterium disordered over two half-occupied sites. O-D bond lengths are approximately equal [0.977(6) and 0.978(6) $\AA$]. A trifurcated hydrogen bond is observed for both D1 and D2 with bond lengths of 2.263(6), 2.257(6), 2.038(5) and 2.541(7), 2.392(6), 2.253(5), respectively. For OD-wadsleyite, the Rietveld fit was poor and the deuterium position could not be located in difference Fourier maps, suggesting that the deuterium atom was disordered over several sites. This is consistent with the IR spectra of the sample, which showed broadening of the O-D bands at 2500 and 2700 cm$^{-1}$.
T41B-1180 0800h
X-ray Diffraction and Vibrational Spectroscopy Studies on Brucite-type $\beta$-Cadmium Hydroxide to 36 GPa.
The brucite-structured hydroxides serve as a good model system to understand the compression of complex layered magnesium hydrous silicates at high pressure. We have studied the pressure behavior of $\beta$-Cd(OH)$_2$ using x-ray diffraction at APS and SSRL, and Raman and infrared-absorption spectroscopy at MIT and ALS, respectively, from 0 to 36 GPa at room temperature. The x-ray measurements show that the $c/a$ ratio decreases steeply (-0.0117/GPa) up to 7 GPa, decreases much gently (-0.0014/GPa) between 7 and 20 GPa, and drops by 1.8 % at 22 GPa. Between 22 and 27 GPa, the $c/a$ ratio does not change with pressure. A fit of pressure-volume data at 0-20 GPa yields $K_0$ = 38(2) GPa and $K_0'$=11(1). The $K_0'$ value is the highest among all the measured for brucite-type hydroxides ($K_0'$= 4-5.2). Assuming a discontinuity at 6-9 GPa, a separate P-V data fit of 0-6 GPa and 9-20 GPa yields $K_0$ = 49(2) GPa and 105(7) GPa, respectively for $K_0'=4$. The Raman-active lattice modes [$E_g$ and $A_{1g}$] of $\beta$-Cd(OH)$_2$ stiffen linearly between 0 and 17 GPa with a discontinuity at 10 GPa, and disappear at 17 GPa. The OH-stretching mode ($A_{1g}$) of $\beta$-Cd(OH)$_2$ softens linearly with pressure without discontinuity and with smallest slope (-3.62 cm$^{-1}$/GPa) among the studied brucite type hydroxides. Unlike $A_{1g}$(OH), the OH-asymmetric stretching mode ($A_{2u}$) softens at 0-8 GPa (-2.15 cm$^{-1}$/GPa) and becomes almost pressure independent above 10 GPa. Both $A{1g}$ and $A_{2u}$ OH-modes disappear at pressure above 20 GPa. Similar to other brucite-type materials, the $c$-axis of $\beta$-Cd(OH)$_2$ is much more compressible to 7 GPa and becomes comparable to that of the $a$-axis above 7 GPa. In addition, $\beta$-Cd(OH)$_2$ becomes very incompressible above 7 GPa. Together with these X-ray observations, the discontinuity at 10 GPa in the shifts of the lattice and $A_{2u}$(OH) modes indicate a displacement or rearrangement of hydrogen in the crystal structure of $\beta$-Cd(OH)$_2$ at 7-10 GPa. A phase change takes place from crystalline to amorphous states at 20 GPa.
T41B-1181 0800h
Hydrous Synthesis of Aluminum Bearing Silicate Perovskite: Implications for Hydrogen Storage in the Lower Mantle
It is suggested that the solubility of hydrogen in aluminous perovskite may be significantly enhanced because of oxygen vacancy formation associated with incorporation of Al$^{3+}$ in the structure. We have conducted perovskite synthesis experiments in the MgO-SiO$_{2}$-Al$_{2}$O$_{3}$-H$_{2}$O system in a multi-anvil press at 24 GPa and $1800\deg$ C to determine the correlation between the Al$_{2}$O$_{3}$ contents and water solubility in the perovskite structure. Starting materials were mixtures of Mg(OH)$_{2}$ and SiO$_{2}$ in a one to one mole ratio, with added Al$_{2}$O$_{3}$ contents ranging from 0-5 wt%. The mixture was loaded into a gold capsule and compressed to the stability field of silicate perovskite using an 8/3 multi-anvil cell assembly. Large perovskite crystals (up to 200 micron) were synthesized upon quenching. The product crystals were analyzed by Raman spectroscopy, and their chemical compositions were determined with electron microprobe. Hydrogen was analyzed using secondary ion mass spectrometry (SIMS) with a 6f Cameca ion probe. The experiments with low aluminum content in the starting materials produced pure large crystals of Al-bearing silicate perovskite (1.4 wt% Al$_{2}$O$_{3}$) coexisting with hydrous melt. SIMS analyses of these crystals showed the water content of 1268 ppm, indicating that the water solubility in silicate perovskites is strongly associated with the Al$_{2}$O$_{3}$ contents in the structure. The experimental results will be used to constrain the H$_{2}$O storage and to understand the role of Al$_{2}$O$_{3}$ in the Earth's mantle.
T41B-1182 0800h
Iron-nickel-water system under high pressure and high temperature
Hydrogen has been proposed as one of the light elements in the Earth's core. Information on hydrogen distribution between the mantle and core is very important because hydrogen would have profound influence on the rheological properties and phase relations of minerals and rocks and the Earth's dynamics in global scale. The behavior of water in the primordial Earth also remains unclear. Since the discovery of high solubility of hydrogen into iron at high pressure, various studies on the Fe-H and Fe-H$_{2}$O systems have been carried out. However, the Earth's core contains 5-15$%$ nickel, based on cosmochemistry arguments. In order to simulate more realistic condition for the core formation process in proto-Earth, we investigated the reaction between iron-nickel alloys, Fe$_{90}$Ni$_{10}$ and Fe$_{95}$Ni$_{5}$, and water at high pressures and high temperatures. A series of experiments were performed in the laser-heated diamond-anvil cell (LHDAC) at pressures 5 to 23 GPa and temperatures up to 1220 K with synchrotron radiation at BL13A beamline of Photon Factory (PF), High Energy Accelerator Research Organization (KEK). The reaction of iron-nickel alloys with water depends on pressure. Iron-nickel hydrides (Fe,Ni)H and iron-nickel hydroxide (Fe,Ni)(OH)$_{2}$ were formed at low pressure of 5 GPa, while iron-nickel oxide (Fe,Ni)O and iron-nickel oxyhydroxide (Fe,Ni)OOH coexisting with iron-nickel hydride (Fe,Ni)H were observed above 18 GPa. We found that the amount of hydride formed by the reaction is smaller in the iron-nickel-water system than in the iron-water system, although the hydrogen content in iron-nickel hydride is not significantly different from that in iron hydride. Our results indicate that the existence of nickel in iron tend to reduce the amount of hydrogen supplied into the Earth's core.
T41B-1183 0800h
Influence of Protons on Fe-Mg Interdiffusion in Olivine
In this work, we present experimental quantification of the influence of water-derived protons on Fe-Mg interdiffusion in olivine. We carried out interdiffusion experiments on diffusion couples consisting of oriented single crystals of olivine of composition $\mathrm{Fo_{90}}$ and $\mathrm{Fo_{80}}$. The diffusion anneals were performed between pressures 0.1 and 6 GPa at temperatures of 1373 to 1450 K. The oxygen fugacity during the anneals was set by the Ni-NiO solid state reaction. We report diffusivity as a function of temperature, pressure, and water fugacity following the relation \[ \tilde{D}_{\mathrm{Fe-Mg}} = D_{o} \left( f_{\mathrm{H_2O}}/f_{\mathrm{H_2O}}^o\right)^r \exp{\left[-(Q + PV^*)/{\mathrm{R}T}\right]} \] where $\log{(D_o)} = (-14.8\pm2.7) \log( \mathrm{m^2 s^{-1} })$ , $r = 0.9\pm0.3$, $Q = 220\pm60$ kJ/mol, and $V^* = \left(16\pm6\right) \times 10^{-6}\mathrm{m^3 /mol}$. The water fugacity exponent of 0.9, which is approximately the same as that for hydroxyl solubility in olivine [{\it{Kohlstedt et al.}}, 1996], indicates that octahedral cation vacancies form defect associates with water-derived protons. Based on our results, cation diffusion in water-saturated olivine is $\sim$50 times faster than under water-absent conditions at a pressure of 5 GPa and a temperature of 1373 K.
T41B-1184 0800h
Effect of water on garnet-perovskite phase transformation in MORB system
Garnet lithology in Mid-Ocean Ridge basaltic (MORB) component of the subducting slab transforms to perovskite lithology at the top of the lower mantle whereas peridotite component has transformed to perovskite at shallower depths. This difference causes density crossover between MORB layer and peridotite layer in the slabs and induces the buoyancy force. Although some studies have been made to determine the phase boundary of the garnet-perovskite transformation in dry-MORB system, little is known about an influence of water. We performed in-situ X-ray diffraction experiments to examine the effect of water on the garnet-perovskite transformation in a hydrous MORB composition. High pressure and high temperature in-situ X-ray diffraction experiments were performed using a uniaxial 1500-ton press (SPEED1500) and 800-ton press (MAX III) installed in the synchrotron radiation beamline BL04B1 at SPring-8 and BL14C at the Photon Factory (PF) at the National Laboratory for High Energy Physics (KEK), respectively. The crystallization of CMA (Ca-Mg-Al) - perovskite from glass starting material was observed at lower temperatures (1273-1373K) during heating. This observation is in agreement with the previous diamond-anvil cell experiments. Then, it decomposed into orthorhombic Mg-Al-perovskite, Ca-perovskite and aluminous phase at 21.5-23.1GPa and 1473-1773K. The phase transformation of garnet lithology into perovskite-bearing assemblage in hydrous conditions is about 2GPa lower than that of the dry condition even if the discrepancy among pressure scales are taken into account. These results suggest that the conditions where the buoyancy occurs are restricted within a narrow depth region by the presence of water. Water can play an important role in dynamics of subducting slab.
T41B-1185 0800h
High pressure NMR spectroscopy of clathrate hydrates
Clathrate hydrates are molecular compounds consisting of water and non-polar gas molecules that are only stable at high pressures. Methane hydrate, the most common of these clathrates, is considered as an important constituent "mineral" of icy satellites. Its existence in the earth's hydrosphere, including seafloor sediments, has attracted much attention as an energy resource. The material undergoes phase transformations to denser clathrates at high pressure ($>$1 GPa). The origin of the stability of the high-pressure forms (i.e., filled ices) of these materials is of great interest. To address this question, we are conducting nuclear magnetic resonance (NMR) measurements at high pressures in a diamond anvil cell (Okuchi, 2004; Okuchi et al., 2004). NMR is a very sensitive method to monitor the state of hydrogen bonding. We can also obtain information on local dynamics of molecules by NMR. We will present our results and discuss the nature of intermolecular interactions in clathrate hydrates at high pressure relevant to planetary interiors.
T41B-1186 0800h
Water Partition Coefficients Between Nominally Anhydrous Minerals and Basaltic Melts: Implication on Mantle Melting
Partitioning of water between peridotite minerals and basaltic magma has a significant influence on the initiation of melting in the mantle and on the rheological structure of the lithosphere. To investigate mineral/melt and mineral/mineral partitioning of H$_{2}$O applicable to the mantle, we have conducted multiple saturation experiments consisting of hydrous basalt$\pm$ol$\pm$opx$\pm$cpx in a piston-cylinder apparatus at pressures of 1--2 GPa, temperatures of 1230--1380\deg C and bulk initial water contents of 3.3 to 6.3 wt.%. We measured H$_{2}$O in melts and minerals (ol, opx, cpx) by SIMS using methods described by [1]. Resulting liquids have 3.1-6.4 wt.% H$_{2}$O and average mineral/melt partition coefficients as follows: D$^{ol-melt}$=0.0017$\pm$0.0005 (n=9), D$^{opx-melt}$=0.019$\pm$0.004 (n=8), and D$^{cpx-melt}$=0.023$\pm$0.005 (n=2). Mineral/mineral partition coefficients are D$^{ol-opx}$=0.11$\pm$0.01 (n=4), D$^{ol-cpx}$=0.08$\pm$0.01 (n=2) and D$^{cpx-opx}$=1.4$\pm$0.3 (n=1). Observed partition coefficients are reproducible between experiments and systematic variations with pressure, temperature or concentration of H$_{2}$O are not apparent, possibly because of the relatively small range of pressures and compositions examined. The D$^{pyroxene-melt}$ increases with the Al$_{2}$O$_{3}$ content of the pyroxene due to enhanced solubility of water in Al-bearing pyroxenes. For a peridotite consisting of 58% ol, 30% opx, 10% cpx, and 2% spinel (assumed nominally anhydrous) the calculated bulk sol-liq D is 0.009$\pm$0.002 confirming that water is highly incompatible in mantle minerals. Compared to conventional trace elements, water has a behavior similar to Ce, in accordance with studies on natural basaltic glasses (e.g., [2]). If this bulk D is applicable to the deeper parts of the MORB melting regime, then following [3], we can estimate the effect of H$_{2}$O on peridotite partial melting: for mantle water concentrations of 50--200 ppm, the near-solidus melt would contain 0.6-2.3 wt.% water. Using the data of [4] for $\Delta$H$_{fusion}$, the freezing point depression is 20--60\deg C, which corresponds to initiation of melting beneath along ridge geotherms 5--20 km deeper than the anhydrous solidus, somewhat less than previous estimates ($\sim$50 km, [5]). For concentrations of 500--1000 ppm H$_{2}$O along plume geotherms, melting will begin 60--130 km deeper than the dry solidus. References [1] Koga et al. (2003) G3 4,1--20. [2] Michael (1995) EPSL 131, 301--320. [3] Hirschmann et al. (1999) J. Petrol.40, 831--851. [4] Kojitani and Akaogi (1997) EPSL 153, 209--222. [5] Hirth and Kohlstedt (1996) EPSL 144, 93--108.
T41B-1187 0800h
Micro-analysis of absolute water concentration in minerals
Accurate water concentrations in hydrous and nominally anhydrous minerals are necessary for describing both the size of reservoirs in the global water cycle and the transfer processes between those reservoirs. We present a new method for the determination of absolute hydrogen abundance in solid materials. Hydrogen extraction and manometry is a standard technique for absolute water content measurements, but requires large masses of sample (typically 10's of mg to grams). Major contributions have been made by measuring relative concentration on very small samples using Fourier transform infrared (FTIR) spectroscopy and secondary ionization mass spectrometry (SIMS) but these methods must be calibrated by absolute methods. Our method allows absolute determination of hydrogen content while reducing sample sizes by two to three orders of magnitude, requiring less than 100 \mug of materials with 1 wt% H$_{2}$ and less than 10 mg for those with 1000 ppm H$_{2}$. This micro-analytical technique will allow the measurement of samples that were previously impractical or impossible to measure by current techniques. Our method uses continuous flow mass spectrometry to measure mols of hydrogen. Mineral samples are dehydrated by combustion in a flow of helium carrier gas. Water is cryogenically collected, converted by uranium reduction to hydrogen gas, and analysed by isotope ratio mass spectrometry (IRMS). The measured signal intensity of H$_{2}$$^{+}$2 ion current is calibrated by repeated analyses of the mineral zoisite, which contains a stoichiometric concentration of 1.98 wt % water. Signal response is linear over the calibrated range of 62 to 809 nmols, with scatter about the line of $\pm$6% (n = 8). In the course of this analysis the hydrogen isotopic ratio is also determined. For our Tanzanian zoisite samples the average \deltaD = -37.5 $\pm$1.2 per mil (n = 8). This system is flexible enough to work with a variety of minerals and other geologic materials. For each material a proper heating protocol must be developed. We use Raman and FTIR spectroscopy to establish that samples are not dehydrated during pre-treatment to remove surface water and that samples are completely dehydrated after combustion. These methods have confirmed that we successfully dehydrate samples of zoisite, basaltic glass, tourmaline, and garnets in 30-90 minutes. Duplicate analyses of a tourmaline sample (expected H$_{2}$O content of $(<)$\sim$4.0 wt. %) give a measured value of 3.76 $\pm$0.04 %. Repeated analyses of ALV 526-1, a mid-ocean ridge glass with a known water content of 0.21 wt % water, yield a value of 0.21 $\pm$0.06 wt % (n = 10) total water contents ranging from 43-181 nmols. The high relative error for ALV 526-1 is a result of uncertainties introduced by blank corrections or by surface water contributions. These are issues that must be addressed with continued work. The small sample sizes required by this analysis open up possibilities for studies of many phases of interest in mantle geochemistry. The water contents of mantle pyroxenes and garnets are within a range where 5-50 mg of sample are adequate for analysis. These measurements will help describe the mantle water budget, one of the least understood components of the global hydrologic cycle.
T41B-1188 0800h
High-pressure elasticity of Fo90 hydrous ringwoodite
If the Earth accreted with chondritic amounts of hydrogen, three quarters of the planet's water has either been lost to space, or is now tied up in minerals of the interior. The nominally anhydrous polymorphs of (Mg,Fe)$_{2}$SiO$_{4}$ (olivine, wadsleyite, and ringwoodite) in the upper mantle constitute one of the largest potential reservoirs of water (as hydroxyl) in the planet. New constraints on the hydrogen content of Earth's potentially hydrous transition zone are emerging from seismology and electrical conductivity profiles, but these geophysical observations require quantitative laboratory measurements on the effects of hydration on e.g. velocities and conductivity. We have measured the sound velocities and elastic constants of Fo90 ringwoodite containing about 1 wt% H$_{2}$O to 10 GPa using gigahertz ultrasonic interferometry in the diamond anvil cell. The aggregate bulk and shear moduli of hydrous ringwoodite are 176(7) and 103(5) GPa, respectively, with pressure derivatives of 4.8(6) and 1.8(3). The elastic constants of hydrous ringwoodite represented by {\it C}$_{11}$, {\it C}$_{44}$, and {\it C}$_{12}$ are 298(13), 112(6), and 115(6) GPa, respectively, with pressure derivatives of 8.2(6), 1.4(1), 3.2(4). We calculate that compressional and shear-wave velocities are reduced by about 40 m/s for every 0.1 wt% H$_{2}$O added to ringwoodite. The Vp/Vs ratio of ringwoodite increases by roughly 2% upon hydration to 1 wt% H$_{2}$O, and this difference does not appear to change significantly with pressure. At ambient pressure, we observe a reduction of P-wave velocity equivalent to an increase in temperature of $\sim$600$\deg$C and on S-wave velocity of $\sim$1000$\deg$C. Hydration is therefore expected to have a much larger effect on velocity than does temperature or variations in Fe/Mg within reasonable ranges of these parameters in the transition zone, and may be recognized by elevated Vp/Vs ratios.
T41B-1189 0800h
Investigation into the Effect of H$_{2}$O on Olivine-Ringwoodite Transformation Kinetics
The rapid transformation of metastable olivine subducted into the transition zone has been proposed as a mechanism leading to deep focus earthquakes, as well as general rheological weakening of the subducting slab. In evaluation of this hypothesis, several experimental studies have derived olivine-ringwoodite transformation rates for dry systems. However, double-seismic zones in subducting slabs suggest significant hydration and serpentinization. This likely presence of H$_{2}$O could be expected to significantly enhance the kinetics of the olivine-ringwoodite transformation, reducing the depth to which metastable olivine would persist in the transition zone. We have initiated a series of experiments to systematically investigate the effect of dissolved water on the transformation kinetics of olivine to its high-pressure polymorphs, ringwoodite and wadsleyite. Single crystals of San Carlos olivine were milled and sieved to produce spheres with a diameter of 450-500 microns. These crystals were then hot-pressed in a matrix of olivine and enstatite powder, and deuterated by welding liquid D$_{2}$O inside the sample capsule during piston-cylinder experiments. The use of D$_{2}$O as a proxy for H2$_{2}$O was made to ease SIMS analysis because of the high background for H$_{2}$O. All deuteration experiments were conducted with a nickel/nickel oxide oxygen fugacity buffer. Run conditions ranged from 2-3.2 GPa and 1000-1050 $^{o}$C. The single-crystal San Carlos olivine spheres in the run products were found to contain 150-400 ppm D$_{2}$O by weight. The deuterated olivine spheres and hot-pressed matrix were cut into 1.5 mm cubes and are being used as starting material in experiments conducted in the wadsleyite and ringwoodite stability fields. These experiments are being performed in a multi-anvil apparatus for durations ranging from 30-240 minutes. Run products will be prepared as thin sections and the thickness of the reaction rims will be measured and related to run duration to infer transformation rates as a function of temperature, pressure, and water content. TEM analysis will be used to determine the effect of D$_{2}$O on the transformation mechanisms.
T41B-1190 0800h
Upper Mantle P and S Velocity Structure Beneath Eastern Mexico Derived Through Waveform Inversion
The majority of detailed models of seismic velocity in the transition zone have been derived from observations of triplicated P and S waves that have turning depths within the transition zone. The waveforms of these waves have been modeled largely through trial and error and thus resolution is difficult to assess. Furthermore, it is very rare that P and S studies are done with similar source-receiver geometries making P to S ratio studies uncertain. We present P and S models for the upper mantle beneath eastern Mexico derived through waveform inversion of triplicated P and S waves produced from the same earthquake. The models were derived from seismic waveform data produced by an event located at the border of Mexico and Guatemala recorded by the La Ristra passive seismic array. The La Ristra array consisted of 54 broadband seismometers arranged linearly from west Texas to southeastern Utah spanning about 950 kms. The data span from 18.5 to 26.5 degrees in distance. The orientation of the array lies approximately along a great circle from the Guatemala source making the data set ideal for investigating the upper mantle. We shifted the data in time to account for shallow crustal and upper mantle variations beneath the array using the results of a local tomography and receiver function study. The waveforms were inverted for mantle velocity from 40 to 800 km depth using a conjugate gradient algorithm. In the inversion we tried a suite of starting models with different depths of the 410 and 660 discontinuities and different gradients. The best fitting models have discontinuities of 8% for P and 8% for S at 410 km depth and 3% for P and 6% for S at 670 km depth. A common feature of the models is a low velocity zone above the 410 km discontinuity that is much more prominent in the S model than the P model. This feature may be due to partial melt induced by water release from the transition zone as proposed by Bercovici and Karato. The overall jump in velocity at 410 km is also larger than in previously published models with a lower gradient below. A second feature is an unusually high gradient beginning at about 550 km depth extending to the 660 km discontinuity. This may be a thermal anomaly due to a flat lying slab or perhaps represents a phase change in the transition zone. The data do not require a discontinuity near 520 km depth.
T41B-1191 0800h
High pressure neutron powder diffraction study of OD-chondrodite
Chondrodite [Mg$_{5}$Si$_{2}$O$_{8}$(F,OH)$_{2}$], a member of the humite group, has been the subject of a number of recent studies because of its high P-T stability, association with mantle-derived rocks and structural similarity to olivine. In particular, OD-chondrodite is a potential repository for water in the upper mantle and serves as a model for the OH defect structure in olivine. High-pressure neutron powder diffraction studies of the intermediate F-bearing solid solution have shown that hydrogen bonding in this phase strengthens slightly with pressure in response to a rotation of the OH vector. In this study, neutron powder data have been collected to 5.87 GPa for the OD-chondrodite in order to investigate the affect of the OD $<$=$>$ F substitution on the hydrogen-bond geometry and the mechanism of compression. In contrast to F-bearing chondrodite, there are two D atoms disordered over four possible positions in the OD end-member. The two O-D bond lengths remain essentially constant with increasing pressure. Both H-bond lengths associated with the D1 site show significant increases with pressure, in contrast to F-bearing chondrodite. At the D2 site, one H-bond length increases whereas the second bond length decreases. The largest angular change is associated with O-D1$^{...}$O5, which increases from $\sim$155 to 165$\deg$. The increase in H-bond lengths appears to be consistent with the positive pressure shifts ($\sim$30 cm$^{-1}$) observed for the OH bands in infrared spectra of OD-chondrodite. Strain calculations show that the minimum axis of compression is perpendicular to the closest packed layers in both OD and F-bearing chondrodite. The maximum direction of compression, which is approximately parallel to the b crystallographic axis in F-bearing chondrodite, is rotated $\sim$24$\deg$ clockwise in the bc plane in OD-chondrodite.
T41B-1192 0800h
High Pressure IR And XRD Studies For Hydrous Minerals
The knowledge of the high pressure behavior of hydrous minerals is important to an understanding of the source of water, and its role in deep focus earthquakes in the Earth's mantle. The structural information of hydrogen could be attained from the spectroscopic studies, which are sensitive to the change relative to hydrogen bonds. We performed in situ high pressure IR spectra measurements up to 25 GPa for Al(OH)3, In(OH)3, Zn(OH)2, Ni(OH)2 at U2A beamline, NSLS. Corresponding high pressure XRD experiments were carried out at X17B3 beamline, NSLS, and ID-B beamline of HPCAT, APS for their compressibility and phase stability. These high pressure researches could extend our understanding of the hydrogen role during compressions and pressure induced phase transition in the well concerned hydrous minerals.
T41B-1193 0800h
How well do we understand the dissolution of water in minerals?
Dissolution of H$_{2}$O in nominally anhydrous minerals is a multi-step process, which is still insufficiently understood in spite of its far-reaching consequences for understanding the deep Earth's water cycle. To study this process on a fundamental level we use the structurally simplest model mineral, MgO, which crystallizes in a densely packed, face-centered cubic structure and is predominantly ionic. When H$_{2}$O dissolves in MgO, two H$^{+}$ substitute for one Mg$^{2+}$ introducing OH$^{-}$ anions and Mg$^{2+}$ vacancies. The energetically most favorable, and hence, most likely defect is that of OH$^{-}$ pairs next to Mg$^{2+}$ vacancies. However, these OH$^{-}$ pairs are not stable. Upon cooling below ~600°C they rearrange electronically: they split off an H$_{2}$ molecule and their two oxygen anions change their valency from 2- to 1- undergoing spin pairing to form a peroxy anion, O$_{2}$$^{2-}$. We present experimental evidence (i) for the appearance of the H-H stretching band in the infrared spectrum of large MgO single crystals grown from an H$_{2}$O-laden melt, due to lattice-bound H$_{2}$ molecules, and (ii) for the copious H$_{2}$ evolution from nanocrystalline OH$^{-}$-doped, ultrahigh purity MgO. We further show that the peroxy anions control the dielectric polarization of the MgO, i.e. effective dielectric constant, and the electrical conductivity. Recognizing that the dissolution of water does not stop at the stage of forming OH$^{-}$, that OH$^{-}$ can "disappear" from the IR spectrum through their conversion to H$_{2}$, and that oxygen can adopt another valency than the usual 2- is likely to profoundly affect the way how we look at and should study the Earth's deep water cycle.
T41B-1194 0800h
Water in Olivine - not a Simple Question
How much "water" is present in olivine is of great interest for the upper mantle. Thermodynamics mandates that small amounts of H$_{2}$O become structurally incorporated, when olivine crystallizes in an H$_{2}$O-laden environment. It is widely believed that the formation of hydroxyl (OH$^{-}$ or Si-OH) is the end point of the H$_{2}$O dissolution mechanism. Therefore it is also believed that quantitative solute H$_{2}$O contents can be obtained by measuring the O-H stretching bands in olivine crystals by infrared (IR) spectroscopy. However, we can take olivine crystals, which are devoid of OH$^{-}$ according to the IR criteria, and fracture them in front of a mass spectrometer. Instantly upon fracture, H$_{2}$ molecules evolve from the freshly formed surfaces, followed by H$_{2}$O and then again by H$_{2}$. This observation indicates that an apparently "dry" olivine can contain dissolved H$_{2}$O, though not in form of IR-recognizable OH$^{-}$ or Si-OH. When we measure the effective dielectric constant \epsilon$_{eff}$ at the limit of 0 Hz, we observe that \epsilon$_{eff}$ changes reversibly in two steps, around 200°C and 400°C. At about 400°C the olivine turns into a p-type semiconductor. Our observations point (i) at the conversion of hydroxyl, probably Si-OH pairs, into H$_{2}$ plus peroxy links, Si-OO-Si, (ii) at a two-step dissociation of the peroxy links. We conclude that, to determine the amount of solute "water" in olivine, it is not enough to measure its hydroxyl content by IR spectroscopy. Other methods have to be used or developed to provide information about the number of H$_{2}$ molecules and peroxy links in the olivine matrix.
T41B-1195 0800h
Solubility of hydrogen and aluminum in CaSiO$_3$-perovskite and the partitioning of water among silicates in the lower mantle
The incorporation of water into nominally anhydrous silicates is associated with mechanical weakening and lowered melting points such that small amounts of water in mantle minerals can serve to alter planetary dynamics. The effects of minor phases have been largely neglected, yet CaSiO$_{3}$-perovskite, which may contain a large vacancy population, could serve as a repository for significant quantities of trace volatile elements. Here we present first principles, pseudopotential calculations of fully relaxed structures on CaSiO$_{3}$-perovskite with aluminum and hydrogen substitutions. We assume the hydrogen enters the structure via the coupled substitution of Si$^{4+}$ for Al$^{3+}$ + H$^{+}$. The most stable hydrogen location has an OH bond length of 1.02 $\AA$ and is virtually independent of pressure. This substitution is associated with a 1.5% softening of the bulk modulus for a structure containing 3.1 mol% H$_{2}$O and Al$_{2}$O$_{3}$, a lesser effect than recently reported results for Si$^{4+}$ for Al$^{3+}$ + H$^{+}$ softening in MgSiO$_{3}$-perovskite [Panero and Stixrude, EOS trans. AGU 85(17), 2004] or SiO$_{2}$-stishovite [Panero and Stixrude, EPSL 2004]. We calculate that the enthalpy of solution of Ca$^{2+}$+Si$^{4+}$=2Al$^{3+}$ is about half that of the Si$^{4+}$ for Al$^{3+}$ + H$^{+}$ reaction, with both solution enthalpies generally increasing with pressure between 0 and 100 GPa. This is in contrast to MgSiO$_{3}$ in which the solution enthalpies for equivalent reactions are nearly equal and generally constant with pressure. Assuming ideal entropy of solution, at 2000 K we find Ca-perovskite saturated in aluminum and water to contain 100 ppm wt H$_{2}$O and 1.7 wt % Al$_{2}$O$_{3}$ at the top of the lower mantle. Due to the differing behavior of the solution enthalpy with pressure, the partitioning of aluminum and hydrogen between MgSiO$_{3}$ and CaSiO$_{3}$ perovskite increasingly favors MgSiO$_{3}$ under water- and aluminum- undersaturated conditions. Aluminum partitions preferentially into MgSiO$_{3}$ with K$_{Al}^{Mgpv/Capv}$ $\sim$ 10 at 25 GPa and 2000 K, consistent with Irifune (1994) and Murakami et al., (2002), increasing to $\sim$ 50 at 100 GPa. The effect of pressure on water partitioning is more dramatic with K$_{H}^{Mgpv/Capv}$ increasing from $\sim$15 at 25 GPa to $\sim$ 120 at 100 GPa.
T41B-1196 0800h
Kinetics of Antigorite Dehydration by in Situ X-ray Diffraction
The dehydration of serpentine minerals during the burial of oceanic slabs plays a significant role in the origin of magmatism and sismicity at subduction zones. Although recent high-pressure experiments studies show that antigorite is the only stable serpentine mineral at mantle conditions, the mechanisms and kinetics of its breakdown are still misunderstood. The dehydration of antigorite under pressure (1-5.5 GPa) and temperature (until 800°C) has been studied in situ by XRD, using a Paris-Edinburgh press installed at ESRF. Three kinds of starting materials were used: (a) a powder of natural antigorite, (b) a mix of antigorite powder (95 wt%) and crystal seeds of forsterite + enstatite (5 wt%), (c) samples a or b saturated in water, to study the influence of the water-content on the P-T conditions for dehydration. The X-rays being absorbed by metallic capsules (Au, Pt), the samples were loaded in boron nitride (BN) containers. BN capsules do not fully prevent fluids to escape from the sample, so that the water activity during the experiments was lower than one. The samples were first pressurised, and then slowly heated (10°C/min). The collection of diffraction spectra every 1-2 minutes during the temperature increase enables to follow, in real-time, the mineralogical reactions of antigorite breakdown. The sequence of XRD spectra reveals that the high-temperature stability limit of antigorite is reduced by 50-100°C under water-unsaturated conditions. Moreover, the decomposition of antigorite to the high-temperature products, forsterite + enstatite, proceeds via an intermediate assemblage of forsterite + "talc-like" phase, observed within a temperature interval of 130 +/- 20°C. The analysis of the transformation-time data using the Avrami model suggests that the breakdown of antigorite and the "talc-like" phase is kinetically controlled by surface growth processes at the edges of grains. The overall transformation rates are 10 to 100 times faster than those observed in water-saturated experiments, and show that the water activity is a strong driving force for dehydration. In the light of these kinetic results, we examine the possibility of dehydration induced seismicity in subducting oceanic slabs.
T41B-1197 0800h
Water under extreme conditions of the mantle
The form in which water exists in the Earth's mantle and crust is currently not very well understood. We used Raman spectroscopy to study fluid water at approximately 1000 K and 2 to 60 GPa in a laser heated diamond anvil cell. First principles molecular dynamics (MD) simulations have also been employed to simulate water at 2000 K up to 120 GPa. The experimental Raman intensity of the O-H stretch mode was observed to decrease with pressure, and beyond 50 GPa this mode was no longer visible. Consistent with this experimental observation, the MD simulations show that water under these conditions consists of very short lived (<10 fs) H$_{2}$O, H$_{3}$O$^{+}$ and O$^{2-}$ species. These results indicate that water at extreme conditions is highly chemically reactive and conductive, which is important for understanding of elementary chemical and physical processes in the Earth.
T41B-1198 0800h
Molecular Modeling of the 10-\AA\ Phase at Subduction Zone Conditions
Molecular dynamics (MD) modeling of the 10-\AA\ phase, Mg$_3$Si$_4$O$_{10}$(OH)$_2$$\dot$$x$H$_2$O, with $x$ = 2/3, 1.0 and 2.0 shows complex structural changes with pressure, temperature and water content and provides new insight into the structures and stabilization of these phases under subduction zone conditions. The structure(s) of this phase and its role as a reservoir of water in the mantle have been controversial, and these calculations provide specific predictions that can be tested by in situ diffraction studies. At ambient conditions, the computed structures of talc ($x$=0) and the 10-\AA\ phases with $x$=2/3 and 1.0 are stable over the 350 ps period of the MD simulations. Under these conditions, the $x$=2/3 and 1.0 10-\AA\ phases show phlogopite-like layer stacking in good agreement with previously published structures based on powder X-ray diffraction data for samples quenched from high pressure and temperature experiments. The calculations show that the 10-\AA\ phase with $x$=2.0 is unstable at ambient conditions. The computed structures at $P$=5.5 GPa and $T$=750K, well within the known stability field of the 10-\AA\ phase, change significantly with water content, reflecting changing H-bonding configurations. For $x$=2/3, the layer stacking is talc-like, and for $x$=1.0 it is phlogopite-like. The calculations show that transformation between these two stackings occurs readily, and that the talc-like stacking for $x$=2/3 composition is unlikely to be quenchable to ambient conditions. For $x$=2.0, the layer stacking at $P$=5.5 GPa and $T$=750K is different than any previously proposed structure for a 10-\AA\ phase. In this structure, the neighboring basal oxygens of adjacent TOT layers are displaced by $b/3$ (about 3 \AA) resulting in the Si atoms of one siloxane sheet being located above the center of the six-member ring across the interlayer. The water molecules are located 1.2 \AA\ above the center of all six-member rings and accept H-bonds from the OH groups located below the rings. The $b/3$-displaced structure does not readily transform to either the talc-like or phlogopite-like structure, because neither of these stackings can accommodate two water molecules per formula unit. There is likely to be a compositional discontinuity and phase transition between the $b/3$-displaced phase and the phase with phlogopite-like stacking.
T41B-1199 0800h
Structure and Decompression Melting of a Novel 2-D High Pressure Ice Phase Formed in Nano-Confinement Between Two Talc Surfaces
Molecular dynamics (MD) simulation of water confined in nano-spaces between layers of talc (composition Mg$_3$Si$_4$O$_{10}$(OH)$_2$ + 2H$_2$O) at 300K and pressures of approximately 0.45 GPa show the presence of a novel 2-D ice structure for the interlayer water, and the structural evolution with decreasing pressure provides insight into the mechanisms of its decompression melting. Talc is hydrophobic at ambient pressure and temperature, but at high pressure and temperature, weak hydrogen bonding between the talc and water plays an important role in stabilizing the structure. The results suggest that experimentally accessible elevated pressures may cause a wide range of previously unknown water structures in nano-confinement. At pressures significantly greater than 0.45 GPa, the water molecules are isolated and can not form an H-bonds with each other. At about 300K and 0.45 GPA, however, the interlayer spacing is large enough that a continuous H-bond network forms among the water molecules, basal oxygens, and OH-groups. In the talc 2-D ice, each water molecule is coordinated by six O$_{\rm b}$ atoms of one basal siloxane sheet and three water molecules. The water molecules are arranged in a buckled hexagonal array in the $a$-$b$ directions with two sub-layers along [001]. Each H$_2$O molecule has four H-bonds, accepting one from a talc OH-group and one from another water molecule, and donating one to an O$_{\rm b}$ and one to another water molecule. In plan view, the molecules are arranged in six-member rings reflecting the substrate talc structure. Decompression melting occurs by vacancy and interstitial formation due to migration of water molecules from the six-member rings to the centers of other six-member rings and eventual formation of liquid-like regions and large vacant cavities in the interlayer space. The results provide specific, experimentally testable predictions about the interlayer phase transition and suggest that pressure alone may cause significant effects on the structure of nano-confined water.
T41B-1200 0800h
High-temperature single-crystal neutron diffraction study of natural chondrodite
Chondrodite [Mg$_{5}$Si$_{2}$O$_{8}$(F,OH)$_{2}$] is a member of the humite group, which is a family of hydrous magnesium silicates with varying Mg$_{2}$SiO$_{4}$ : Mg(F,OH)$_{2}$ ratio. All humite minerals are structurally related to forsterite (Mg$_{2}$SiO$_{4}$), which is regarded as the most important mineral of the Earth's upper mantle. Based on their P-T stability, humites are candidate materials for transporting fluid phases and, in particular, OH into the deep Earth. Chondrodite is of particular interest because it exhibits an unusually high thermal stability. Unlike other dense, hydrous Mg-silicates (DHMS), OH-chondrodite remains stable at ambient pressure up to 873 K. In hydrous minerals, the thermal stability must be related, at least in part, to the behavior of the O-H$^{...}$O bond at high temperatures since dehydration is often the first breakdown reaction of such minerals. Raman studies of both OH- and the F-bearing chondrodite show a slight negative temperature dependence of the O-H vibrational frequency. This is unusual when compared with other DHMS phases, since it implies a decrease in the force-constant for the covalent O-H bond, which may be linked to a shortening of the hydrogen bond. The temperature dependence of the H-atom environment in F-bearing chondrodite (Mg$_{4.64}$Fe$_{0.28}$Mn$_{0.014}$Ti$_{0.023}$(Si$_{1.01}$O$_{4}$)$_{2}$F$_{1.16}$OH$_{0.84}$) was investigated from 10 K to 900 K using single crystal neutron diffraction data collected in this study (500, 700 and 900 K), and previously published low-temperature data collected for the same crystal (10 K, 100 K and 300 K). The objectives of these experiments were: 1) to determine how the H-atom environment affects thermal stability; and 2) to investigate the correlation between Raman and neutron data. Atomic positions at each temperature were corrected for the partial substitution of O by F, and the apparent shortening of the O-H bond at high temperatures due to thermal motion of H and O/F. The refined temperature dependent O/F positions showed a physically reasonable evolution of the anisotropic displacement parameter with temperature. The Burgi-Capelli method (Burgi and Capelli 2000) was used to calculate the full mean-square displacement matrix and correct the O-H bond for the effect of thermal motion without any assumptions as to the correlation of O and H displacements. With this method, the observed temperature dependent displacement parameters for O and H, respectively, are used to refine the vibrational eigenvectors and their frequencies for the O-H unit. The refined stretching and bending frequencies ($\sim$2900 cm$^{-1}$ / $\sim$760 cm$^{-1}$) compare favorably with observed values ($\sim$3000 to 3500 cm$^{-1}$ / $\sim$500 to 800 cm$^{-1}$), lending credibility to the method. Results suggest that the O-H$^{...}$F geometry is controlled by the thermal expansion of the silicate framework whereas the H position is governed by the bond-valence requirement that maintains the H atom at a position of ideal bond valence sum. Burgi, H.B. and Capelli, S.C. (2000): Dynamics of molecules in crystals from multi-temperature anisotropic displacement parameters. I. Theory. Acta Crystallographica, A56, 403 - 412.
T41B-1201 0800h
Mantle H$_{2}$O-Activity Estimated From Amphibole Equilibria
Determining values of H$_{2}$O activity (aH$_{2}$O) for mantle rocks will yield a better understanding of those mantle processes that are controlled, in part, by the availability of H$_{2}$O (e.g., melting and deformation). Two different types of amphibole equilibria can be used to estimate H$_{2}$O activity in the mantle. The first method relies on the equilibrium between iron oxy-component and hydroxy-component, as described in the dehydrogenation/ oxidation reaction Fe$^{2+}$ + OH$^{-}$ = Fe$^{3+}$ + O$^{2-}$ + 1/2 H$_{2}$, for which the equilibrium constant (Ke) can be expressed in terms of thermodynamic mole fractions ([ ] = H-vacancy on the O(3) anion position) as Ke = fH$_{2}$ (28.94) ((XFe$^{3+}$)$^{2}$(X[ ])$^{2}$)/((XFe$^{2+}$)$^{2}$(XOH)$^{2}$). The variation in Ke was quantified experimentally by annealing three different amphiboles (two mantle-derived kaersutitic amphiboles and a crustal pargasite) over the range 700-1000°C, 1-10 kbar, and fH$_{2}$ from that of the HM to GM solid buffer assemblages. The equation below relates log Ke to T, P, and amphibole composition: log Ke = 4.23 - 4380/T(K) + [1.37((Ti+Al$_{total}$ apfu) - 2.49)] + [(88/T(K)) (P-1 (kbar))] If the T, P, and amphibole composition are known, values of log Ke calculated from the equation predict the equilibrium log fH$_{2}$ to within 0.20 to 0.40 log units. If log fO$_{2}$ at the time of equilibration can be independently estimated, the H$_{2}$O activity can also be estimated. An alternate approach for estimating H$_{2}$O activity from amphibole-bearing mantle rocks is to use a variety of H$_{2}$O-buffering equilibria among end-member components in olivine, two-pyroxenes, amphibole and other phases (e.g., 2tr + 2fo = 5en + 4di + 2H$_{2}$O). A self-consistent thermodynamic database (THERMOCALC, Holland and Powell, 1990) can be used to determine the aH$_{2}$O of such univariant H$_{2}$O-buffering equilibria as a function of P and T. A mantle amphibole assemblage from Dish Hill, CA (sample DH101-E, McGuire et al., 1991) was used to calculate aH$_{2}$O using the two different methods. The mean value of aH$_{2}$O determined from seven different dehydration reactions is 0.04, with a 1-standard-deviation range from 0.006 to 0.06. That range of water activity is in good agreement with the value of aH$_{2}$O = 0.013 obtained using the dehydrogenation/oxidation equilibrium, along with an estimate of log fO$_{2}$. The use of xenolith amphiboles to infer values of aH$_{2}$O in the mantle requires that the H-content of the amphibole does not change during ascent or eruption. Changes in H-content have significantly different effects on the dehydration and dehydrogenation equilibria. Thus, comparison of the aH$_{2}$O estimates from these two different methods may permit quantification of H-loss.
T41B-1202 0800h
Limited Hydration of the Juan de Fuca, Gorda and Explorer Plates and its effect on the Intraslab Seismicity
Using data collected offshore Nicaragua and Costa Rica, Ranero et al. (2003) showed that seismic reflection imaging may be a useful tool for determining the depth and extent of hydration of the oceanic lithosphere approaching a subduction zone. The water stored in subducting plates is released landward of the trenches through dehydration and is believed to strongly affect a number of processes of importance to natural hazard studies. The released water promotes partial melting responsible for arc magmatism and leads to dehydration embrittlement that is often considered to be the most plausible earthquake mechanism for intraslab events at intermediate depths of 50-300 km. Free water can also affect the physical properties of rocks at the megathrust by reducing the temperature at which transition from brittle to ductile deformation occurs, thus possibly having significant impact on the location of the downdip limit of seismogenic zones. To study hydration processes at the Cascadia convergent margin, we compiled a database of all regional seismic reflection surveys (streamers 3 km or longer) done across the Juan de Fuca, Gorda and Explorer plates. Unlike at the Middle America trench, oceanic lihthosphere subducted at the Cascadia margin is young, warm, and descends at a relatively shallow angle. Bending-related normal growth faulting is sparser and subtler at the Cascadia margin but, surprisingly, it begins much further seaward of the trench than offshore Nicaragua and Costa Rica and the fault density is mostly uniform. Fault planes dip steeply and are challenging to image. Where fully imaged, these faults offset the sediments, cut across the crust to Moho and penetrate into the upper mantle. However, the depth of fault penetration into the mantle appears to be far less than at the Middle America trench, suggesting limited depth of hydration for the subducting plates at the Cascadia margin. The difference in the depth and extent of hydration at the Cascadia and Middle America margins seems to correlate well with the difference in the density and magnitude of intraslab seismicity observed at both margins. This limited depth of hydration of the subducting lithosphere at the Cascadia margin may restrict the maximum magnitude of the Cascadia intraslab earthquakes to about Mw=7.
T41B-1203 0800h
Estimating Upper Mantle Hydration from {\it In Situ} Electrical Conductivity
The electrical conductivity of 35-40 Ma Pacific plate has been measured {\it in situ}; one robust result is the presence of bulk anisotropy in the lithospheric upper mantle. We interpret this anisotropy to be a result of hydrothermal circulation into the upper mantle along spreading-ridge-parallel normal faults: the associated zones of serpentinized peridotite provide the pathways of enhanced electrical conductivity required by the data. Our modeling bounds the range of possible anisotropic ratios, which are then used to estimate the amount of water required to serpentinize the requisite amounts of peridotite. These data sets, however, do not indicate anisotropy in the bulk conductivity of the crust, nor in the asthenospheric mantle. This second point is significant, as recent measurements of sub-continental asthenospheric conductivity have been interpreted to indicate anisotropy aligned with present plate motion, with the diffusion of hydrogen through olivine advanced as an explanation.
T41B-1204 0800h
Geophysical and Petrological Consequences of the `Slab Water Cycle'
Water released at the bottom of the transition zone within a plunging lithospheric slab and percolating upward through that slab accounts for several phenomena observed within and around subduction zones. The pressure of that water as it rises permits brittle faulting, and therefore deep-focus earthquakes, anywhere from the bottom of the transition zone upward. Water percolating upward above the transition zone also explains the lower of the paired seismic zones observed at intermediate depths, either through its own pressure, or by dehydration of lithospheric mantle serpentinized by it. Eventually, water that does not rise through the hot mantle wedge may be rising in serpentine seamounts. Incompatible elements, and perhaps carbon, should follow the water in that cycle and therefore become progressively concentrated in the subducting slab. When the slab becomes deflected and thrust horizontally along the base of the transition zone, its eventual foundering releases water, which must then rise through normal mantle rather than the slab. What is not dissolved in the transition zone may account for some kimberlites and other `wet spot rocks' far away from the subduction zone. Similarly, a detached slab has to release its water within upwelling mantle and may thus contribute to late arc magmatism that is more abundant and geochemically different than before detachment.
T41B-1205 0800h
Dependence of Water Solubility in Mantle Olivine on the Silica Activity
Trace amount of hydrogen stored in nominally anhydrous minerals in the Earth's mantle plays an important role in the dynamics of solid state processes. It may change mineral physical properties such as strength, diffusion and creep rates, electrical conductivity, and transformation kinetics. There have been tremendous efforts to study the incorporation of hydrogen in mantle minerals. However, answers to problems such as incorporation mechanism and solubility of hydrogen in olivine, the most abundant mineral in the upper mantle, remain controversial and somewhat illusive. We present here the results from experimental studies of dependence of hydrogen solubility in olivine on the silica activities, which in turn help our understanding of the incorporation mechanism of hydrogen. Off-line high-pressure experiments were performed on SAM85 press equipped with a DIA module at beamline X17B2 of NSLS at 5-10 GPa and 800 - 1200C. Durations range from one to four hours. Standard cell assembly for DIA was used, each consisting pressure medium of amorphous boron mixed with epoxy, graphite heater, and BN sample capsule. San Carlos olivine was used as starting material. Three layers of sample separated by Pt foil were packed into the capsule, with a fine powder of olivine, a mixture of olivine and orthopyroxene, and a mixture of olivine and periclase in each layer. Several larger grains of olivine (few hundred microns) were embedded in each powder sample. Volatiles were provided by the decomposition of epoxy at high temperature. After experiments samples were cut and double-polished for FTIR analyses, which were conducted at beamline U2A of NSLS. Both powder and single crystal FTIR data revealed the following trend for water solubility in olivine in three samples: olivine + MgO $>>$ olivine $>$ olivine + opx. Water content in olivine in low silica environment (i.e., in sample olivine + MgO) is al least an order of magnitude higher than that equilibrated with high silica activity (i.e., in sample olivine + opx). However, no significant difference among IR patterns in terms of hydroxyl stretching bands for different samples was found, in contrary to findings in some earlier studies. Our results suggest the silica activity may be the dominant factor for water solubility in mantle olivine, and that hydrogen could prefer to enter Si vacancies in olivine.
T41B-1206 0800h
Experimental Study of the PVTX Properties of Water-Methane
Hydrothermal fluids containing small amounts of methane are common in many geological environments, including sedimentary basins, submarine hydrothermal systems and low grade metamorphic rocks. To better understand the behavior of these fluids, the phase equilibrium properties of water containing small amounts of methane ($<$ 4 mol%) were determined using the synthetic fluid inclusion technique. Methane in the experiments was generated by the reaction of aluminum carbide with water to produce methane and aluminum hydroxide. Reaction products (methane and aluminum hydroxide) were verified by Raman spectroscopy and X-ray powder diffraction analyses, respectively. Concentrations of methane in the inclusions were calculated based on the reaction stoichiometry. A series of experiments demonstrated that using fresh aluminum carbide (from newly opened container) was critical to obtaining accurate compositions. This is because aluminum carbide decomposes slowly when exposed to humid air. Error analyses based on mass balance showed that with appropriate sample handling, the errors associated with the experimental technique were $<$ 5% (relative) for methane concentrations $<$ 4 mol%. Synthetic fluid inclusions containing $<$ 4 mol% methane were trapped at 500-$700\deg$C, 1-3 kilobars. For a given formation temperature and pressure, the homogenization temperature decreases with increasing methane concentration. This behavior reflects both the changing slope of the fluid isochore as well as the increasing pressure on the solvus with increasing methane. Phase diagrams summarizing the PTX properties of the water-methane system will be presented.
T41B-1207 0800h
Equation of State of Aluminous H-bearing Stishovite to 60 GPa
The equation of state of water-bearing aluminous stishovite has become an important issue in relation to the both the properties of subducting slabs and also the transport and retention of water in the deep mantle. There were numerous studies on elasticity of stishovite, however most of them were performed on pure SiO$_{2}$ compositions, which is not representative of natural stishovites formed in MORB layer of subducting slabs. We performed elasticity measurements of SiO$_{2}$ stishovites with various Al$^{3+}$ and H$^{+}$ contents with Brillouin scattering and x-ray diffraction techniques at ambient and high pressures. We have determined the P-V equation of state of Al-rich H-bearing SiO$_{2}$ stishovite by x-ray powder diffraction at pressures up to 60 GPa using synchrotron radiation. The sample contained 1.8 %wt Al$_{2}$O$_{3}$ and about 450 ppm or more of H$^{+}$. This composition corresponds to stishovite that would coexist with aluminous iron-bearing Mg-silicate perovskite in a subducted slab. Our results indicate that Al$^{3+}$ and H$^{+}$ have a less pronounced effect on the elastic properties of stishovite than it was reported earlier. Our results suggest that Al and H are retained in stishovite under extreme P-T conditions and that stishovite is an agent for transporting water to the deepest lower mantle. Brillouin scattering measurements of acoustic velocities on stishovites with various chemical compositions will also be discussed.
T41B-1208 0800h
Effects of hydrogen impurities on the lattice thermal diffusivity of quartz and quartzites up to 1000\deg C
The phonon contribution to thermal diffusivity (D$_lat$) in quartz single-crystals was measured using a laser flash apparatus between room temperature and 1000\deg C. Our measurements revealed differences in D$_lat$ between samples which most likely result from hydroxyl impurities. Variations in D$_lat$ between wet and dry samples persist above the $\alpha$ to $\beta$ transition. For all samples, D$_lat$ follows a 1/T trend. Unlike previous measurements, thermal diffusivity of $\beta$-quartz remains constant or decreases slightly with increasing temperature, showing that the technique removes radiative transfer effects. The difference between samples is of greatest interest. The two orientations of a quartz single-crystal containing 171 H/10$^6$ Si as OH defects have D$_lat$ vales at room temperature that are 22% and 41% lower than equivalent orientations for a dry sample. The effect of defects on D$_lat$ is amplified when the OH-dipole is parallel to the electromagnetic vector of the diffusing heat. Oriented milky quartz samples with 1260 to 2100 H/$^6$ Si have D$_lat$ 12-15% lower than dry quartz values, suggesting that fluid inclusions slow thermal diffusion to a smaller degree. Room temperature D$_lat$ of 5 different quartzite samples vary up to 24% from the highest quartzite value of 4.04 mm$^2$/s. For all samples, D$_lat$ is lower than that expected for randomly oriented, dry quartz. In quartzites, differences in D$_lat$ between samples decrease with temperature, such that little variation is seen for the $\beta$ phase. This behavior is expected if porosity, which ranges from 5% to 19%, hinders thermal transport in quartzite. Grain size does not appear to affect D$_lat$. If water plays an important role in heat transport, we should see it, since quartzite contains much more water than the single crystals. Because of the small number of samples examined, the dual speciation as silanol complexes on grain surfaces of some quartzites (up to 420 H/$^6$ Si) and as molecular water in all samples (ranging from 328 to 5103 H/$^6$ Si), combined with porosity variations, our understanding of water's effect in quarzites is impeded. Additional experiments are planned to probe the sensitivity of D$_lat$ to H impurities in these samples. Small amounts of water in quartz are known to affect physical properties such as electrical conductivity and strength. Studies of D$_lat$ presented here suggest that water, even in minute amounts, alters the thermal properties of quartz as well. Finding a similar effect in other anhydrous minerals would add to our understanding of mantle and lithosphere dynamics.