Mg-Fe partitioning between silicate spinel and magnesiowüstite at high pressure: experimental determination and calculation of phase relations in the system Mg2SiO4-Fe2SiO4

 Mg-Fe partitioning experiments between (Mg,Fe)₂SiO₄ spinel and (Mg,Fe)O magnesiowüstite were carried out at pressures of 17–21.3 GPa at temperatures of 1400 and 1600 °C, using a multi-anvil apparatus, in order to determine interaction parameters of spinel and magnesiowüstite solid solutions and also to constrain the equilibrium boundaries of the postspinel transition in the Fe-rich side in the system Mg₂SiO₄-Fe₂SiO₄. The obtained values of the interaction parameters were 3.4 ± 1.5 and 13.9 ± 1.4 kJ mol⁻¹, respectively, for spinel and magnesiowüstite solid solutions at 19 GPa and 1600 °C. The partitioning data in the system Mg₂SiO₄-Fe₂SiO₄ at 1400 and 1600 °C showed that the transition boundary between spinel and the mixture of magnesiowüstite and stishovite has a negative dP/dT slope. Using the above interaction parameters and available thermodynamic data of the Mg₂SiO₄ and Fe₂SiO₄ end members, the transition boundaries of spinel to the mixture of magnesiowüstite and stishovite were calculated. Within the uncertainties of the data used, the calculated boundaries are in good agreement with the boundaries at 1400 and 1600 °C experimentally determined in this study. The dissociation boundary of Fe₂SiO₄ spinel to wüstite and stishovite, calculated from the thermodynamic data, has a negative slope of −1.5 ± 0.6 MPa K⁻¹. Received: 18 February 1998 / Revised, accepted: 18 October 1999     

Fe–Mg partitioning between ringwoodite and magnesiowüstite and the effect of pressure, temperature and oxygen fugacity

 The partitioning of Mg and Fe between magnesiowüstite and ringwoodite solid solutions has been measured between 15 and 23 GPa and 1200–1600 ^∘C using both Fe and Re capsule materials to vary the oxidation conditions. The partitioning results show a clear dependence on the capsule material used due to the variation in Fe³⁺ concentrations as a consequence of the different oxidation environments. Using results from experiments performed in Fe capsules, where metallic Fe was also added to the starting materials, the difference in the interaction parameters for the two solid solutions (W _FeMg ^mw−W _FeMg ^ring) is calculated to be 8.5±1 kJ mol⁻¹. Similar experiments performed in Re metal capsules result in a value for W _FeMg ^mw−W _FeMg ^ring that is apparently 4 kJ higher, if all Fe is assumed to be FeO. Electron energy-loss near-edge structure (ELNES) spectroscopic analyses, however, show Fe³⁺ concentrations to be approximately three times higher in magnesiowüstite produced in Re capsules than in Fe capsules and that Fe³⁺ partitions preferentially into magnesiowüstite, with K _{D Fe3+} ^ring/mw estimated between 0.1 and 0.6. Using an existing activity composition model for magnesiowüstite, a least–squares fit to the partitioning data collected in Fe capsules results in a value for the ringwoodite interaction parameter (W _FeMg ^ring) of 3.5±1 kJ mol⁻¹. The equivalent regular interaction parameter for magnesiowüstite (W _FeMg ^mw) is 12.1±1.8 kJ mol. These determinations take into account the Fe³⁺ concentrations that occur in both phases in the presence of metallic Fe. The free energy change in J mol⁻¹ for the Fe exchange reaction can be described, over the range of experimental conditions, by 912 + 4.15 (T−298)+18.9P with T in K, P in kbar. The estimated volume change for this reaction is smaller than that predicted using current compilations of equation of state data and is much closer to the volume change at ambient conditions. These results are therefore a useful test of high pressure and temperature equation of state data. Using thermodynamic data consistent with this study the reaction of ringwoodite to form magnesiowüstite and stishovite is calculated from the data collected using Fe capsules. Comparison of these results with previous studies shows that the presence of Fe³⁺ in phases produced in multianvil experiments using Re capsules can have a marked effect on apparent phase relations and determined thermodynamic properties. Received: 13 September 2000 / Accepted: 25 March 2001     

Activity-composition relations of the magnesiowüstite solid solution series in equilibrium with metallic iron in the temperature range 1050–1400 K

The equilibrium $${\text{(1}} - y{\text{)Fe}}_{(s)} + \tfrac{{\text{1}}}{{\text{2}}}{\text{O}}_{{\text{2(g)}}} \rightleftarrows {\text{Fe}}_{{\text{1}} - y} {\text{O}}_{{\text{(}}s,{\text{ in MW)}}} $$ was studied by measuring oxygen potentials for a range of different magnesiowüstite compositions relative to those of the iron-wüstite system in an oxygen concentration cell involving yttria stabilized zirconia as the solid electrolyte. The temperature range covered was 1050 to 1400 K. Separate measurements of the mole fraction of trivalent iron in magnesiowüstite (x(Fe³⁺)) were made and the composition dependence ofx(Fe³⁺) was taken into account in calculations of the activity-composition relations of FeO, Fe_2/3O and MgO.     

Experimental determination of the effects of pressure and temperature on the stoichiometry and phase relations of wüstite

When quenched metastable wüstite (Fe_.924O and Fe_.947O) is held at 300°C at pressures up to 200 kbar in a diamond anvil cell, a mixture of magnetite, metallic iron and wüstite is found. We interpret this to indicate that magnetite plus metallic iron constitute the stable phase assemblage at pressures and temperatures below this boundary is stoichiometric FeO ($\text{a}{}_0\text{= 4.332 ± 0.001}\text{A}\text{?}$) at pressures below 110 kbar at 300°C. However, just below the boundary in the pressure range 110 kbar to 200 kbar at 300°C, the residuál wüstite is non-stoichiometric ($\text{a}{}_0\text{< 4.332}\text{A}\text{?}$). Data collected at pressures and temperatures above the boundary indicate that non-stoichiometric wüstite (Fe_xO) plus metallic iron constitute the stable phase assemblage and that the value of x in Fe_xO increases as pressure is increased isothermally to 100 kbar and then decreases as pressure is increased above 100 kbar.     

Magmatic evolution of the material of the Earth’s lower mantle: Stishovite paradox and origin of superdeep diamonds (Experiments at 24–26 GPa)

The ultrabasic–basic magmatic evolution of the lower mantle material includes important physicochemical phenomena, such as the stishovite paradox and the genesis of superdeep diamonds. Stishovite SiO₂ and periclase–wüstite solid solutions, (MgO · FeO)_ss, associate paradoxically in primary inclusions of superdeep lower mantle diamonds. Under the conditions of the Earth’s crust and upper mantle, such oxide assemblages are chemically impossible (forbidden), because the oxides MgO and FeO and SiO₂ react to produce intermediate silicate compounds, enstatite and ferrosilite. Experimental and physicochemical investigations of melting phase relations in the MgO–FeO–SiO₂–CaSiO₃ system at 24 GPa revealed a peritectic mechanism of the stishovite paradox, (Mg, Fe)SiO₃ (bridgmanite) + L = SiO₂ + (Mg, Fe)O during the ultrabasic–basic magmatic evolution of the primitive oxide–silicate lower mantle material. Experiments at 26 GPa with oxide–silicate–carbonate–carbon melts, parental for diamonds and primary inclusions in them, demonstrated the equilibrium formation of superdeep diamonds in association with ultrabasic, (Mg, Fe)SiO₃ (bridgmanite) + (MgO · FeO)_ss (ferropericlase), and basic minerals, (FeO · MgO)_ss (magnesiowüstite) + SiO₂ (stishovite). This leads to the conclusion that a peritectic mechanism, similar to that responsible for the stishovite paradox in the pristine lower mantle material, operates also in the parental media of superdeep diamonds. Thus, this mechanism promotes both the ultrabasic–basic evolution of primitive oxide–silicate magmas in the lower mantle and oxide–silicate–carbonate melts parental for superdeep diamonds and their paradoxical primary inclusions.     

Transmission electron microscopic study of the fine-grained vein matrix in the Suizhou L6 meteorite

The mineralogy of shock vein matrix in the Suizhou meteorite has been investigated by optical and transmission electron microscopy. It was revealed that the vein matrix is composed of majorite-pyrope garnet, magnesiowüstite, and ringwoodite, with FeNi–FeS intergrowths. The observation and character of ring-like selected electron diffraction (SAED) patterns indicate that the idiomorphic garnet crystals in the vein matrix have different orientations. The polycrystalline nature of magnesiowüstite is also confirmed by a ring-like SAED pattern. Both garnet and magnesiowüstite crystals showed sharp diffraction spots, signifying the good crystallinity of these two minerals. The SAED pattern of cryptocrystalline ringwoodite shows only diffuse concentric diffraction rings. FeNi metal and troilite (FeS), which were molten during the shock event, occur in the matrix as fine eutectic FeNi–FeS intergrowths filling the interstices between garnet and magnesiowüstite grains. Based on the phase diagram of the Allende chondrite and the results of this TEM study, it is inferred that majorite-pyrope garnet first crystallized from the Suizhou chondritic melt at 22–26 GPa, followed by crystallization of magnesiowüstite at 20–24 GPa, and then ringwoodite at 18–20 GPa. The eutectic intergrowths of FeNi-metal and troilite are proposed to have crystallized during meteorite cooling and solidified at the last stage of vein formation.     

P-V-T equation of state of magnesiowüstite (Mg0.6Fe0.4)O

The lattice parameter of magnesiowüstite (Mg_0.6Fe_0.4)O has been measured up to a pressure of 30 GPa and a temperature of 800 K, using an external heated diamond anvil cell and diffraction using X-rays from a synchrotron source. The experiments were conducted under quasi-hydrostatic condition, using neon as a pressure transmitting medium. The experimental P-V-T data were fitted to a thermal-pressure model with the isothermal bulk modulus at room temperature K _T0 = 157 GPa, (?K _TO /?P) _T =4, (?K _T /?T) _P =-2.7(3) × 10^-2 GPa/K, (?K _T /?T) _v =-0.2(2) × 10^-2 GPa/K and the Anderson-Grüneisen parameter δ _T =4.3(5) above the Debye temperature. The data were also fitted to the Mie-Grüneisen thermal equation of state. The least-squares fit yields the Debye temperature θ _DO = 500(20) K, the Grüneisen parameter γ ₀=1.50(5), and the volume dependence q=1.1(5). Both thermal-pressure models give consistent P-V-T relations for magnesiowüstite to 140 GPa and 4000 K. The P-V-T relations for magnesiowüstite were also calculate by using a modified high-temperature Birch-Murnaghan equation of state with a δ _t of 4.3. The results are consistent with those calculated by using the thermal-pressure model and the Mie-Grüneisen relation to 140 GPa and 3000 K.     

Effects of defect and pressure on the thermal expansivity of Fe x O

Pressure–volume–temperature measurements have been carried out using synchrotron X-ray diffraction for wüstite at static pressures of 1.9, 2.6, and 5.4 GPa. Our results revealed that the composition change of wüstite and, hence, rearrangements of defect structures are primarily caused by the magnetite (Fe₃O₄) exsolution at temperatures of 523–723 K. Based on the isobaric volume–temperature data collected during cooling, the contribution of compositional variations to the unit-cell volumes of wüstite in the ranges of 300–673 K and 723–1073 K is negligibly small, within the experimental uncertainties. These observations suggest that the measured volume changes in the range of 300–673 K and 723–1,073 K can be attributed to the metal–oxygen bond expansion. Owing to the magnetite exsolution, thermal expansion data are obtained in each experiment at 1.9, 2.6, and 5.4 GPa for wüstite of two different compositions, Fe_0.987O and Fe_0.942O. At all three pressures, Fe_0.942O shows a thermal expansion that is about 30% larger than Fe_0.987O. Such findings represent the first experimental evidence of a substantial effect of nonstoichiometry on thermal expansivity, and based on previous thermodynamic calculations of the defect formation and interaction, this effect is likely associated with the distinct defects arrangements in iron-rich and more iron-deficient wüstite. This study also presents thermal equations of state for wüstite of two different compositions. Such volume-related properties at high temperatures are experimentally difficult to obtain in wüstite but important for thermodynamic studies in the binary Fe–O system.     

Possible contribution of lower mantle magnesiowüstite metallization into geomagnetic data

Phase change of dielectric magnesiowüstite in the lower mantle may leave signatures in geomagnetic records of the globally distributed array of observatories. We investigate theoretically which may be the contribution of magnesiowüstite metallization to geomagnetic data and how the variations of magnetic susceptibility associated with this phase change may influence the Earth's field. The modeling is performed using spherical harmonic analysis (SHA) of mantle electromagnetic (EM) responses in observatory geomagnetic data at periods of decades, 11 years, 1 year, and 27 days. The existence of a lower mantle conductor is checked against monthly means of real observatory records from 1920 through 2009 obtained by preliminary processing.     

Texture development and elastic stresses in magnesiowűstite at high pressure

Cubic magnesiowűstite has been deformed in a diamond anvil cell at room temperature. We present results for (Mg_0.4Fe_0.6)O, (Mg_0.25Fe_0.75)O, and (Mg_0.1Fe_0.9)O up to 37, 16, and 18 GPa, respectively. The diffraction images, obtained with the radial diffraction technique, are analyzed using both single peak intensities and a Rietveld method. For all samples, we observe a [100] fiber texture but the texture strength decreases with increasing iron content. This texture pattern is consistent with {110}〈1-10〉 slip. The images were also analyzed for stress, elastic strains, and elastic anisotropy. In general, the stress measured in magnesiowűstite samples is lower than previously measured on MgO. The elastic anisotropy deduced from the X-ray measurements shows a broad agreement with models based on measurements with other techniques.     

Melting of wüstite and iron up to pressures of 600 kbar

The effect of pressure on melting temperature of wüstite and iron has been measured with laser-heated diamond anvil cell. The temperature was determined by measuring the thermal radiation emitted by the sample as a function of wavelength in the range from 600 nm to 900 nm to which Planck's radiation function was fitted; the pressure was measured by ruby-fluorescence technique. The melting curve of wüstite in this study when extrapolated to low pressures agrees with Lindsley's (1966) data. Our data are similar to the recent data of Boehler (1992) and close to that of Ringwood and Hibberson (1990) at pressure of 160 kbar, but the melting temperature does not rise as rapidly with increasing pressure as reported by Knittle and Jeanloz (1991). If tungsten emissivity is used in the temperature calculation, the melting curve of iron matches those of Boehler et al. (1990). Use of emissivity of iron in the temperature calculation results in somewhat higher temperatures than those reported by Boehler et al. (1990).     

Dependence on pressure of conduction by hopping of small polarons in minerals of the Earth's lower mantle

Electrical conductivity of the lower mantle-like assemblage (Mg,Fe)SiO₃ perovskite-(Mg,Fe)O magnesiowüstite is usually analyzed using the quasi-chemical Arrhenian approach of diffusion. The conductivity of this assemblage has often been attributed to hopping of small polarons, because of the low value of the activation energy and the small negative activation volume. However, the solid-state physics approach can provide more arguments, for or against conduction by polarons. We have tried to bridge the gap between the two approaches and identify the physical quantities entering the phenomenological activation parameters. In particular, we have investigated the pressure dependence of the activation energy, and the physical meaning of the activation volume. Hopping is controlled by the binding energy of the polaron and by the value of the exchange integral, which increases with pressure causing the observed decrease of the activation energy. From the physical theory and the results of experiments at pressures up to 40 GPa and temperatures up to 400 ^∘C, we have estimated the values of parameters characteristic of polarons: radius, mobility, time between jumps and adiabaticity. These values are compatible with conduction by small adiabatic polarons. The consequences for extrapolations to lower mantle conditions of the presence of a temperature dependent preexponential term in the expression for conductivity have been examined. It was found that the extrapolations are not significantly different from those using the Arrhenius equation. Received: 5 November 1998 / Revised, accepted: 4 May 1999     

Oxygen fugacity dependence of Ni,Co, Mn,Cr, V,and Si partitioning between liquid metal and magnesiowüstite at 9–18 GPa and 2200°C

The oxidation states of Ni, Co, Mn, Cr, V and Si in magnesiowüstite have been determined in metal-oxide distribution experiments using a multi anvil apparatus at 9 and 18 GPa and 2200°C as a function of oxygen fugacity. Despite limitations to control oxygen fugacity by applying conventional buffering methods in high pressure experiments, a wide range of redox-conditions (3 log bar units) has been imposed to the metal-oxide partitioning experiments by varying the Si/O ratio of the starting material. The oxygen fugacity was calculated according to the Fe-FeO equilibrium between the run products. The ability to impose different oxygen fugacities by varying the starting material is confirmed by the large variation of element partitioning coefficients obtained at constant pressure and temperature. The calculated valences at both pressures investigated are divalent for Co, Mn, V and 4+ for Si. The results for Cr (∼2.5+) and Ni (∼1.5+) indicate non-ideal mixing of Ni and Cr in at least one of the product phases. Because the application of 1 bar activity coefficients for Ni and Cr in metal alloys does not change these valences, non-ideal mixing in magnesiowüstite or significantly larger non-ideal mixing properties of Ni and Cr in metal alloys at high pressure are likely to be responsible for the apparent valences. Omitting such non-ideal mixing properties when extrapolating high-pressure element partitioning data may be significant. The elements Cr, V and Mn become siderophile (D_M^met/ox > 1) at 9–18 GPa and 2200°C at oxygen fugacities below IW-2.7 to IW-3.7. Considering, in addition, the influence of temperature, the depletion of Cr, Mn and V in the Earth’s mantle may be due, at least partly, to siderophile behavior at high pressure and temperature.     

Fe–Mg Interdiffusion in (Mg,Fe)O

We have performed a series of interdiffusion experiments on magnesiowüstite samples at room pressure, temperatures from 1,320° to 1,400°C, and oxygen fugacities from 10^?1.0 Pa to 10^?4.3 Pa, using mixed CO/CO₂ or H₂/CO₂ gases. The interdiffusion couples were composed of a single-crystal of MgO lightly pressed against a single-crystal of (Mg_1-x Fe _x )_1-δO with 0.07<x<0.27. The interdiffusion coefficient was calculated using the Boltzmann–Matano analysis as a function of iron content, oxygen fugacity, temperature, and water fugacity. For the entire range of conditions tested and for compositions with 0.01<x<0.27, the interdiffusion coefficient varies as $$\tilde D\, =\,2.9\times10^{ - 6}\,f_{{\text{O}}_2 }^{0.19}\,x^{0.73}\,{\text{e}}^{ - (209,000\, -\,96,000\,x)/RT}\,\,{\text{m}}^{\text{2}} {\text{s}}^{ -1} $$ These dependencies on oxygen fugacity and composition are reasonably consistent with interdiffusion mediated by unassociated cation vacancies. For the limited range of water activity that could be investigated using mixed gases at room pressure, no effect of water on interdiffusion could be observed. The dependence of the interdiffusion coefficient on iron content decreased with increasing iron concentration at constant oxygen fugacity and temperature. There is a close agreement between our activation energy for interdiffusion extrapolated to zero iron content (x=0) and that of previous researchers who used electrical conductivity experiments to determine vacancy diffusivities in lightly doped MgO.     

About short and long range ordering in wüstites, Fe1-xO

This short contribution consists of several comments on a paper recently published by T.R. Welberry and A.G. Christy in Physics and Chemistry of Minerals (volume 24, Number 1, January 1997, pages 24–38). Due to the fact that some of the previous works were not taken into account by these authors, it seems necessary to present briefly a short review of the main results concerning the short and long range ordering of point defects in wüstites Fe_1-xO. Using direct imaging found in the literature, it was clearly shown in papers of our group, from 1981 to 1991, that one of the most probable and numerous clusters in wüstite under equilibrium is the (10/4) one, which is of blende ZnS type. The P' and P'' quenched phases were also described from this cluster.     

Comparative study of the kinetics of the thermal decomposition of synthetic and natural siderite samples

The mechanism of the thermal decomposition of two siderites (a pure synthetic and a natural Mg-containing sample) has been determined from comparison of the results obtained from linear heating rate (TG) and constant rate thermal analysis (CRTA) experiments in high vacuum. The thermal decomposition of the synthetic siderite takes place approximately 200 K below the decomposition temperature of the natural sample. The mechanism and the product of the thermal decomposition are different for the siderite samples. In fact, an A₂ kinetic model describes the thermal decomposition of the synthetic siderite, whereas the thermal decomposition of the natural sample obeys an F₁ kinetic law. Decomposition products of the synthetic siderite are iron and magnetite, those of the natural siderite are wüstite and minor magnetite. Received: 22 July 1999 / Accepted: 12 February 2000     

Internal reduction of (Mg,Ni) O: Effect of a NiO-Concentration gradient

Single crystals of (Mg, Ni) O with a Ni-concentration gradient from the outer surface to the interior have been reduced so that metallic nickel particles precipitate in the wüstite matrix. The internal reduction morphology is very different from the one observed in reduced homogeneous (Mg, Ni) O single crystals: pores develop and the precipitates grow along preferential crystallographic directions. The addition of iron ions to the inhomogeneous mixed oxide crystals suppresses the porosity. The observations are described and discussed in terms of point defects and transport kinetics.     

Clinopyroxene-perovskite phase transition of FeGeO3 under high pressure and room temperature

In order to confirm the possible existence of FeGeO₃ perovskite, we have performed in situ X-ray diffraction measurements of FeGeO₃ clinopyroxene at pressures up to 40 GPa at room temperature. The transition of FeGeO₃ clinopyroxene into orthorhombic perovskite is observed at about 33GPa. The cell parameters of FeGeO₃ perovskite are a=4.93(2) Å, b=5.06(6) Å, c=6.66(3) Å and V=166(3) ų at 40 GPa. On release of pressure, the perovskite phase transformed into lithium niobate structure. The previously reported decomposition process of clino-pyroxene into Fe₂GeO₄ (spinel)+GeO₂ (rutile) or FeO (wüstite) +GeO₂ (rutile) was not observed. This shows that the transition of pyroxene to perovskite is kinetically accessible compared to the decomposition processes under low-temperature pressurization.     

Detecting a magnesiowüstite phase transition in the lower mantle by inversion of geomagnetic data

Global geomagnetic data are inverted for detecting a high-conductivity layer at depths of 1500–2000 km to test the hypothesis of a magnesiowüstite phase transition in the lower mantle. We present the results of processing of both synthetic and global data—average monthly values of the geomagnetic field from 1920 to 2009. The inverted global data are consistent with the possible existence of a high-conductivity layer at great depths in the lower mantle.     

High pressure equilibrium of nickel and cobalt between metal and mantle minerals

In order to test the effect of very high pressures on the siderophile behaviour of two elements, Ni and Co, we have carried out diamond anvil cell experiments on Ni- and Co-bearing systems, up to 70 GPa. Observation of recovered samples by analytical transmission electron microscopy shows that Ni and Co remain siderophile at least up to 70 GPa, but that their siderophile character decreases with pressure, as already observed in previous studies at lower pressures. Our results also suggest that the abundances of Ni and Co observed in the Earth’s upper mantle cannot be explained by very high pressure equilibrium between silicate perovskite, magnesiowüstite, and metal.