DI13A-01
Optical Absorption Spectra of Hydrous Wadsleyite to 32 GPa
Optical absorption spectra of high-pressure minerals can be used as indirect tools to calculate radiative conductivity of the Earth's interior [e.g., 1]. Recent high-pressure studies show that e.g. ringwoodite, γ-(Mg,Fe)2SiO4, does not become opaque in the near infrared and visible region, as previously assumed, but remains transparent to 21.5 GPa [2]. Therefore, it has been concluded that radiative heat transfer does not necessarily become blocked at high pressures of the mantle and ferromagnesian minerals actually could contribute to the heat flow in the Earth's interior [2]. In this study we use gem-quality single-crystals of hydrous Fe-bearing wadsleyite, β-(Mg,Fe)2SiO4, that were synthesized at 18 GPa and 1400 °C in a multianvil apparatus. Crystals were analyzed by Mössbauer and Raman spectroscopy, electron microprobe analysis and single-crystal X-ray diffraction. For absorption measurements a double-polished 50 μm sized single-crystal of wadsleyite was loaded in a diamond-anvil cell with neon as pressure medium. Optical absorption spectra were recorded at ambient conditions as well as up to 32 GPa from 400 to 50000 cm-1. At ambient pressure the absorption spectrum reveals two broad bands at - 10000 cm-1 and -15000 cm-1, and an absorption edge in the visible-ultraviolet range. With increasing pressure the absorption spectrum changes, both bands continuously shift to higher frequencies as has been observed for ringwoodite [2], but is contrary to earlier presumptions for wadsleyite [3]. Here, we will discuss band assignment along with the influence of iron, compare our results to previous absorption studies of mantle materials [2], and analyze possible implications for radiative conductivity of the transition zone. References: [1] Goncharov et al. (2008), McGraw Yearbook Sci. Tech., 242-245. [2] Keppler & Smyth (2005), Am. Mineral., 90 1209-1212. [3] Ross (1997), Phys. Chem. Earth, 22 113-118.
DI13A-02
Melting of Fe-Si Alloys up to 14 GPa
Melting temperature (Tm) experiments of Fe-17wt%Si and Fe-9wt%Si were carried out up to 14 GPa in a Walker module press. Compositions of the quenched and recovered samples were checked by electron microprobe. The determination of melting was based on analysis of the quenched sample texture. The partial melt texture was determined by comparing the melted and unmelted textures. The melting boundary of Fe- 17wt%Si increases with pressure and is consistent with previous melting data at higher pressure and with studies that have shown a single stable solid structure up to 124 GPa. The range of partial melting increases with pressure and when extrapolated to 21 GPa, is much greater than the range of partial melting indicated by the experiments of Kuwayama and Hirose (2004). The Tm of Fe-9wt%Si increases with pressure below 7 GPa but with a lower pressure-dependence than Fe-17wt%Si. After intersecting the Fe-17wt% Si liquidus at approximately 7.5 GPa, the Tm of Fe-9wt%Si increases rapidly with pressure, suggesting the possibility of a triple point. This is consistent with the extension of the bcc-fcc solid phase boundary determined over the narrow range of temperature (800-10000C) by Zhang and Guyot (1999). The range of partial melting shrinks with pressure for Fe-9wt% Si below 7 GPa. The effect of Si on the depression of Tm of Fe will be discussed.
DI13A-03
Stability and Compressibility of RbAlSi3O8 at High Pressure Conditions
Potassium and Rubidium are minor and trace elements of the Earth's mantle, both of which have long-lived radioactive isotopes. The decay of 87Rb to 87Sr in particular serves as a geochemical tracer for long-lived mantle processes. While Rb in the crust is stored as a feldspar (RbAlSi3O8), the most abundant mineral of the Earth's crust, the mineral host of Rb in the mantle is uncertain. Here we report the phase stability and compressibility of RbAlSi3O8, rubicline, from laser-heated diamond anvil cell (LHDAC) experiments with synchrotron-based x-ray diffraction. The high-pressure phase diagram of rubicline is similar to that of microcline, transforming to the hollandite structure at ∼15 GPa, then to the hollandite II structure at ∼25 GPa, suggesting that the host mineral for K and Rb in the mantle are identical. Preliminary results yield the bulk modulus of Rubidium hollandite-I to be 210 (± 10) GPa, assuming a K'=4.
DI13A-04
Geoelectric Structural Directions And Dimensionality Of The Damara Belt and Surrounding Cratons
Magnetotelluric (MT) soundings were made along three transects, as part of the broader Southern African Magnetotelluric Experiment (SAMTEX), crossing the Kalahari Craton, the Neo-Proterozoic Ghanzi- Chobe/Damara belts (collectively termed the DMB) and the southern Congo/Angola craton in an effort to define the geo-electric structure across the region. The McNeice and Jones (2001) distortion decomposition technique was applied to the MT data and indicate significant depth and along-profile variations in geo-electric strike and dimensionality on all the three transects crossing these three tectonic units (i.e. Kalahari craton, Congo/Angola craton and the DMB). The geo-electric strikes are generally parallel to the North-East trending tectonic fabric as inferred from the magnetic data, but the significant strike variations with depth are expressions of heterogeneity in the lithospheric structure. The Kalahari craton, south of the DMB, exhibits a fairly one dimensional (1D) structure with preferred strike directions between 40-50 degrees for the entire period (i.e. depth) range, indicating little crust-mantle decoupling. The DMB appears to be strongly 2D at lower crustal and upper mantle depths (10-100s) with no consistent/preferred strike direction and significant phase differences between the conductive and resistive directions. North of the DMB and into the southern Congo/Angola craton there are significant variations in geo-electric strike direction and dimensionality at most sites for crust and upper mantle lithosphere. Although strike directions of between 50-60 degrees are observed on most sites for periods greater than 100s, the large phase differences indicate that the regional structure is 3D. These variations in strike directions and dimensionality suggest that sectional 2D inversion modelling of each profile is necessary and models with period dependent direction must be analyzed.
DI13A-05
A dendritic solidification experiment under large gravity - implications for the Earth's inner core solidification regime.
The Earth's inner core solidification regime is usually thought to be dendritic, which should results in the formation of a mushy layer at the inner core boundary, possibly extending deep in the inner core. The release of latent heat and solute associated with crystallization provides an important boyancy source to drive thermo- chemical convection in the core. In the laboratory, two modes of convection associated with the crystallization of mushy layers have been observed. One is a boundary layer mode originating from the destabilisation of the chemical boundary layer present at the mush-liquid interface; the second is the so-called 'mushy layer mode' which involves the whole mushy layer. In the mushy layer mode, convection usually takes the form of narrow plumes rising through crystal free conduits called chimneys. One particularity of inner core crystallization is its extremely small solidification rate compared to typical outer core convective timescales. We have designed and build an experiment devoted to the study of crystallization under a large gravity field, using a centrifuge, of an aqueous solution of ammonium chloride, which is a good analogue to metallic alloys. The large gravity field allows to reach Rayleigh numbers much larger than in typical solidification experiments. Under large gravity fields, we observe the disappearance of chimney convection and show that the large gravity field promotes the boundary layer convection mode at the expent of the mushy layer mode. As the gravitationnal forcing is increased, convective heat and solute transport are significantly enhanced, which results in larger solid fraction directly below the mush-liquid interface. The increase in solid fraction results in a dramatic decrease of the permeability in the mushy layer, which eventually becomes subcritical in respect to the mushy layer mode. Because of the very slow solidification rate of the inner core, convective transport of heat and solute from the ICB is expected to be very effective and would induce a large solid fraction directly below the ICB.