MR23D-01 INVITED
Partitioning of iron between lower mantle minerals
The Mg-Fe partitioning between lower mantle minerals of perovskite and ferropericlase has been long searched but the results exhibit large inconsistency. The partitioning of iron into recently discovered post-perovskite phase has been controversial as well. This could be in part due to multiple valence and spin states of iron in the lower mantle conditions as well as heterogeneous distribution of iron in laser-heated DAC samples. Here we examined the Mg-Fe partitioning between perovskite/post-perovskite and ferropericlase based on sub-micron scale chemical analyses of recovered samples. The high-pressure phases were synthesized from (Mg0.9Fe0.1)2SiO4 gel at pressures between 30 and 126 GPa and temperatures around 2000 K in a laser-heated DAC. The thin section of recovered samples were obatined by Ar-ion milling using PIPS and Ion Slicer. A typical grain size was about 100-nm. The chemical analyses were made with TEM, FE-EPMA and FE-SEM. The Fe3+/total Fe ratios were also determined from the ELNES spectroscopy attached with TEM. For samples thinned by the Ion Slicer, large areas were observed under the TEM. The analyses demonstrated a remarkable chemical gradient in iron content according to a temperature gradient during laser-heating. Such a spatial variation in iron content was found only in ferropericlase, whereas perovskite and post-perovskite had limited chemical variations. This results in apparently large variations in the Mg-Fe partitioning. The partitioning was therefore calculated from the Fe/Mg ratios of perovskite/post-perovskite and original bulk composition. The Fe3+/total Fe ratios were obtained to be 0.1 to 0.2 for all perovskite, post-perovskite and ferropericlase, indicating the minor presence of ferric iron in our samples. Our results show that KD(Fp/Pv) =(XFe2+/XMg)Fp/(XFe2+/XMg)Pv is approximately constant at 4 between 30 and 105 GPa. This value is lower than those obtained at about 1500 K by Kobayashi et al. [2005], likely due to the higher temperature in this study. The KD(Fp/PPv) value increased to around 10 for post-perovskite, suggesting a strong enrichment of iron in ferropericlase in the lowermost mantle.
MR23D-02
Iron partitioning and the self-oxidation of the lower mantle
Magnesium silicate perovskite (Mg,Fe)SiO3 (Mg-pv) and ferropericlase (Mg,Fe)O (fp) are the dominant phases in the lower-mantle. Their physical and chemical properties determine the dynamics of the deep Earth. It is thus of prime importance to constrain element partitioning at high pressure for improving the geochemical models of the Earth. We investigated iron partitioning between Mg-pv and fp synthetised under lower-mantle conditions (up to 115 GPa and 2200 K) in a laser heated diamond anvil cell (LH-DAC). Recovered samples were thinned to electron transparency by focussed ion beam (FIB) and characterized by analytical transmission electron microscopy (ATEM) and nanometer scale ion probe (nanoSIMS). Iron concentrations in both phases were obtained from EDX measurements and nanoSIMS. Our results are the first to show that recently reported transitions in the lower-mantle (Badro et al, 2003; Murakami et al., 2004) directly affect the evolution of Fe-Mg partitioning between both phases. Mg-pv is increasingly iron-depleted above 70-80 GPa possibly due to the high spin-low spin transition of iron in fp. Conversely, the perovskite to post-perovskite transition is accompanied by a strong iron enrichment of the silicate phase. Iron concentrations determined by ATEM and nanoSIMS are in excellent agreement. Nanoparticles of metallic iron were observed in the Mg-pv bearing runs (figure), suggesting the disproportionation of ferrous iron and the self-oxidation of the mantle, but were not observed when the post- perovskite (ppv) phase was present. Implications on the oxidation state of the Earth and core segregation will be discussed. References J. Badro, G. Fiquet, F. Guyot, J.-P. Rueff, V.V. Struzhkin, G. Vanko, and G. Monaco, Science 300 789-791 (2003) M. Murakami, K. Hirose, N. Kawamura, N. Sata, and Y. Ohishi, Science 304 855-858 (2004)
MR23D-03
Spin crossover and short-range order in lower mantle (Mg,Fe)O
It is now well established that spin crossover occurs in (Mg,Fe)O at high pressure, but conflicting results have been reported of the pressure range over which the transition occurs. Since the spin state can influence the physical properties of (Mg,Fe)O and hence the properties and dynamics of the lower mantle, we made a detailed study of (Mg,Fe)O containing 5, 13 and 20 mole % FeO using Mössbauer spectroscopy in the diamond anvil cell. In the pressure range up to 20 GPa we observed a non-reversible change in quadrupole splitting in both hydrostatic and non-hydrostatic experiments that can be interpreted from an analysis of the quadrupole splitting distribution to indicate a rapid increase in short-range order with a tendency to Fe clusterisation. Further experiments varying the P,T path during synthesis suggest that (Mg,Fe)O likely exhibits significant short-range order under lower mantle conditions. We observed the onset of spin crossover near 50 GPa in the Mössbauer spectra of all samples, but with a transition width that is highly composition dependent. Analysis of literature data together with the current results suggest that spin crossover is a thermal equilibrium process without phase transition. The composition and temperature dependence of spin crossover in (Mg,Fe)O can be described relatively well using such a model, taking into account the local structure of the solid solution. The model further suggests that Fe ions in different local surroundings can have different critical pressures of spin crossover. The total fraction of low-spin Fe2+ depends therefore not only on pressure, temperature and composition, but also on short-range order. Consequences for the physical properties and dynamics of the lower mantle will be discussed.
MR23D-04
Spin Transition and Equation of State of Ferropericlase at High Temperature: Implications for Density Model of the Lower Mantle
Iron-bearing lower mantle mineral, ferropericlase, undergoes an electronic transition of high-spin to low-spin in Fe2+ at high pressure. It has been documented that the spin transition causes a volume decrease and it occurs in the mantle pressure range at room temperature. In order to determine if the transition occurs in the Earth's lower mantle and understand its effect on mantle density profile, we have performed a series of compression experiments on (Mg0.80Fe0.20)O (fp20) and (Mg0.61Fe0.39)O (fp39) at high temperatures up to 3000 K. The high-PT experiments were performed at synchrotron facilities, using the externally-heated and laser-heated diamond-anvil cell techniques. Three sets of diffraction data of fp39 up to 80 GPa and 1000 K (external heating), 60 GPa and 3000 K (laser-heating), 100 GPa and 2400 K (laser-heating) were used to establish the P-V-T equation of state and determine the effect of temperature on the spin transition. Diffraction data of fp20 were also obtained up to 110 GPa and 2000 K to evaluate the compositional effect. The experimental data indicate that the spin transition pressure increases with increasing the iron content and temperature. For any reasonable mantle composition models, low-spin state ferropericlase is expected in the Earth's lower mantle. Our comprehensive P-V-T data of ferropericlase allow us to precisely evaluate the effect of the spin transition on the density profile of the lower mantle.
MR23D-05
New Constraints on the Pyrolitic Model: In Situ X-ray Diffraction Measurements on KLB-1 Peridotite Under the Lower Mantle Conditions
Composition of the Earth\'{ }s lower mantle remains an essential question in Earth sciences. Modeling of lower mantle composition is generally obtained by comparing mineral physics data and seismic observations, using measured properties of the lower mantle end-member phases such as perovskite and ferropericlase. In this study, in situ X-ray diffraction patterns were collected from 30 GPa to 110 GPa and from 1400 K to 2500 K using KLB-1 peridotite as starting material. The experiments were performed in laser-heated diamond-anvil cell at the Advanced Photon Source and at the European Synchrotron Radiation Facility. The sample was loaded in neon pressure medium. We observe the three lower mantle phases at all pressures, i.e., magnesium-iron silicate perovskite, ferropericlase and calcium silicate perovskite, and report molar volume changes as a function of pressure and temperature. We notice the high spin to low spin transition of iron in the ferropericlase at room temperature and also at high temperature. The equations of state (EOS) of the two perovskites and ferropericlase are determined using Au as internal pressure standard. We also examine the Fe/Mg partitioning between Mg- perovskite and ferropericlase with analytical TEM carried out on recovered samples, and the phase proportion from the Rietveld refinement of X-ray diffraction patterns. Using the volume data and chemical composition information, we calculate the densities of peridotite along geotherm of the Earth. The calculations explore the effect of Fe/Mg partitioning variation and the accompanied change in the phase proportion on the densities. Our calculated density profile allows us to make direct comparison with the seismic observations and evaluate current composition mantle models.
MR23D-06 INVITED
Constraints on Mantle Composition from Ultrasonic Velocity Measurements at High P and T
Elasticity measurements at high pressure and high temperatures provide crucial data for testing and constraining possible mineralogical composition of the Earth's deep interior. Advanced ultrasonic techniques in conjunction with state-of-the-art synchrotron facilities can now allow us to conduct simultaneous measurements of sound velocities and density to P-T conditions up to 25-28 GPa and 1600K. Using these techniques, new results have been obtained for many important mantle phases, including olivine, wadsleyite, calcium silicate perovskite, and magnesium silicate perovskite through a thermodynamically self-consistent finite strain data analysis approach, proving an updated mineral physics elasticity database for the study of physical properties of mantle phases as well as aggregate models with different mineralogical compositions. We have calculated phase equilibrium and volume fractions for pyrolite and piclogite models by minimizing the free energy of the mineralogical assemblages at mantle depths, the compressional and shear wave velocities and density profiles are then compared with global and regional seismic data from the upper mantle to the bottom of the transition zone. Details of the comparison to distinguish pyrolite and piclogite at the different regions of the upper mantle and transitions zone as well as their comparison with the velocity gradient and discontinuities in the transition zone, will be discussed.
MR23D-07
Stability and Equation of State of post-perovskite phase in the system MgSiO3-Al2O3 to 2 Mbar
Al2O3 is an important oxide component in the Earthfs interior, and the lower mantle is thought to contain about 5 mol% Al2O3. It is also known that dissolution of the Al2O3 content can affect phase transition pressure and elastic properties of mantle phases. We have studied the effects of Al2O3 content on stability and equation of state of post-perovskite (pPv) phase in MgSiO3 using laser-heated diamond cells. Powdered glasses with the bulk composition of X mol% MgSiO3 and (1-X) mol% Al2O3 (X = 100, 95, 90, and 85) were used as starting materials. They were mixed with Pt powder that served as both a pressure standard and a laser absorber. NaCl was used as a pressure medium. In situ x-ray diffraction experiments at high pressures and temperatures were conducted at GSECARS in Advanced Photon Source using angle-dispersive diffraction and double-sided laser heating. The pPv transition boundary was determined to be 130(2) GPa at 1800 K for X=85. It is likely that the pressure interval of the two-phase stability field (binary roop) of pPv + perovskite is probably a few GPa for X=85, which is inconsistent with both the experimental results for X=75 by Tateno et al. (2005) and theoretical predictions by Akber-Knutson et al. (2005). Pressure (P) -volume (V) data for the pPv phase were obtained for all the single- phase pPv with respective bulk compositions up to 2 Mbar. By fitting the P-V data of the pPv phase with X=95 at 123 - 180 GPa to the 2nd order Birch-Murnaghan equation of state with the reference state set to 120 GPa, bulk modulus at 120 GPa was determined to be 602(21) GPa. Contrary to the case for the perovskite phase in the MgSiO3-Al2O3 system, systematic trend between P-V-X for the pPv phase was found to be insignificant, and the pPv phase in this system has the smallest volume as well as pure MgSiO3 among the silicate pPv phases in various compositions investigated so far. By comparing the P-V data for the pPv phase of X=95 composition with those for perovskite phase with the same chemical composition obtained by Walter et al. (2004), it is estimated that the volume reduction associated with the pPv transition with X=95 bulk composition is 3%.
MR23D-08
Oxygen Fugacity Buffers at Conditions of the Deep Earth
Oxygen fugacity, a proxy for the chemical potential of oxygen, not only drives redox reactions, element partitioning and structural phase transitions, but also controls some transport and rheological properties, especially in minerals like silicates and oxides in which oxygen vacancies can play a large role. Therefore, the importance of oxygen fugacity in the deep Earth can hardly be overstated. In this study we have constructed oxygen fugacity buffers for the metal-oxide systems Fe-FeO, Ni-NiO, and Re-ReO2 at high pressures and temperatures, extending to the conditions of the lower mantle. Pressure-volume-temperature relations for the Fe-FeO, Ni-NiO, and Re-ReO2 metal-oxide pairs were measured by synchrotron X-ray diffraction in both a multi-anvil press and a laser heated diamond anvil cell. Simultaneous measurement of both the metal and its oxide provided a measure of the volume difference between the two phases that is more precise than comparisons between independently determined equations of state. This allows more precise evaluation of the thermodynamics of the metal-oxide system, including oxygen fugacity buffer curves at high pressure. Nonstoichiometric effects in wustite were eliminated by high-PT equilibration with Fe, allowing the equation of state of stoichiometric FeO to be measured. We show that the differences (in log fO2 units) between the IW and NNO buffers, and also between the IW and RRO buffers, decrease significantly with increasing pressure. These results can augment our understanding of metal-silicate partitioning at high pressures, for example, and form a basis for further investigations into the redox state of the deep Earth.
MR23D-09
Iron-carbon system at Earth's lower mantle conditions
The abundance of the carbon near the surface of the Earth is significantly lower than that expected from a cosmochemical grounds. As was shown in recent studies (Shcheka, 2006) the solubility of carbon in the major mantle forming silicates is very low and other minor carbon-rich phases should dominate the carbon budget within the bulk Earth's mantle. One of the them, magnesite (MgCO3) doesn't undergo dissociation at lower mantle conditions (~115 GPa and ~2200 K) and can be one of the major hosts for carbon throughout most parts of the Earth's lower mantle (Isshiki, 2003). It is well known that presence even small amount of carbon can significantly change the path of multi elemental chemical reactions and thereby physical properties of the end products. Hence, the affect of carbon on the phase transformation, chemical reaction, partitioning, melting temperatures etc of the core-mantle forming elements can significantly change our understanding of Earth's lower mantle and the influence of carbon on evolution of Earth. In this work we have studied in-situ the chemical stability and physical properties of the iron-carbon system under appropriate Earth's lower mantle core conditions. The double-sided laser heating technique combined with high resolution angle-dispersive micro x-ray diffraction system at GSECARS (Sector 13, APS, Argonne) was used to characterize samples in the diamond anvil cell (DAC). Iron carbide was synthesized in-situ in DAC from a mixture of Fe and C powders at pressures above 6 GPa and temperature ~1500K. We did not found dissociation of iron carbide in the entire pressure-temperatures range studied: up to ~170 GPa and ~4000K. High pressure high temperature stability of iron carbide phases in the D'' layer, is essential for interpreting the observed seismic anomaly in that region and understanding of the early Earth differentiation. Implications of these results to the composition of the Earth's interior will be discussed.
MR23D-10 INVITED
Indications of crustal carbon in lower-mantle Juina diamond as revealed by FIB/TEM and NanoSIMS investigations.
Indications of crustal carbon in lower-mantle Juina diamond as revealed by FIB/TEM and NanoSIMS investigations. R. Wirth, D.L. Pinti, Y. Sano, N. Takahata, F. Kaminsky Superdeep diamonds from Juina (Mato Grosso, Brazil) contain inclusions such as Fe-periclase, Ca-Si walstromite, majoritic garnet (Kaminsky et al., 2001) and most recently a hydrous aluminium silicate phase "Phase Egg" was found (Wirth et al. 2007). Phase Egg in diamond might suggest that crustal material has been subducted to a depth of the lower Transition Zone. The present paper demonstrates how the combination of Focused Ion Beam (FIB) specimen preparation combined with TEM and NanoSIMS can provide a comprehensive data set from a TEM foil with the dimensions of 15 x 10 x 0.200 micrometer. TEM investigation of the foil revealed the presence of three phases: diamond host carbonate single crystal ilmenite nanocrystal amorphous quench phase. The carbonate is composed of Mg 45.4 at. %, Ca 49.8 at. %, Fe 4.3 at. %, Mn 0.2 at. % and Sr 0.2 at. %. The amorphous quench phase contains Si Zr K. The carbon and nitrogen isotopic composition of the diamond was analyzed from the same TEM foil against internal standard (carbonado diamond GM02 from Brazil) with a Camaca NanoSIMS 50. A 13 spot analyses for the carbon isotopic composition and the N content were measured following two precise transects across the sample. The δ13-C of the diamond varies between -13.9 ± 1.9 ‰ and -25.1± 1.8 ‰. The δ13-C values plotted against the C/N ratios suggest the occurrence of two carbon sources. The first source shows lower C/N ratio and enriched δ13-C (- 14 ‰). The second source of carbon shows higher C/N ratio and depleted δ13-C values (-25 ‰). The occurrence of predominantly depleted δ13-C values (-14 to -25 ‰) suggests that this diamond formed from or contains partially carbon from biological origin. A recent paper supports this idea that reports carbonate inclusions in diamonds from the same locality (Juina, Brazil) thus indicating that the Earth's global CO2-cycle has an ultra-deep extension (Brenker et al., 2007). References Brenker, F.E., Vollmer, C., Vincze, L., Vekemans, B., Szymanski, A., Janssens, K., Szaloki, I, Nasdala, L., Joswig. W., Kaminsky, F. (2007) Carbonates from the lower part of transition zone or even the lower mantle. EPSL, doi: 10.1016/j.epsl.2007.02.038. Kaminsky, F.V., Zakharchenko, O.D., Davies, R., Griffin, W.L., Shiryaev, A.A. (2001) Superdeep diamonds from the Juina area, Mato Grosso State, Brazil, Contributions to Mineralogy and Petrology 140(5), 734-753. Wirth, R., Vollmer, C., Brenker, F., Matsyuk, S., Kaminsky, F. (2007) Inclusions of nanocrystalline hydrous aluminium silicate "Phase Egg" in superdeep diamonds form Juina (Mato Grosso State, Brazil). EPSL, 259, 384- 399.