DI53B-01 INVITED
Oxygen, a possible light element in the Earth's inner core?
The composition of the Earth's core continues to be controversial. It is well known that the density of hcp- (Fe,Ni) solid solutions are too high as compared to seismic observations of the core. Thus, one or more light elements are needed to resolve this discrepancy. However, the nature of the light elements remains unknown but likely candidates include silicon, sulfur, and oxygen. Several recent experimental and theoretical studies have focused on sulfur and silicon. In this study we consider hcp-(Fe1-x,Nix), hcp-(Fe1-x,Six), and hcp-(Fe1-x,Ox) type solid solutions and several stochiometric intermetallic phases. The preliminary results of our static first-principle simulations show, in agreement with experimental observations, that Ni dissolves readily in hcp-Fe at temperatures higher than ~2000-2500 K at inner core pressures . These moderate temperatures are well within the thermal energy budget of the Earth's inner core and indicate a complete solid solution at inner core conditions (~5000 – 7000 K). Considering solid solutions between hcp-Fe and eps-(Fe,Si) we find that stochiometric compounds of composition Fe3Si and Fe5Si3 are less stable than a competing mechanical mixture of hcp-Fe and eps-(Fe,Si). This suggests the absence of intermetallic phases between hcp-Fe and eps-(Fe,Si) and the stability of hcp-Fe derived (Fe1-x,Six) solid solutions at least at high temperatures. The bulk moduli along the hcp-Fe, eps-(Fe,Si) joint vary linearly with silicon content. This finding supports previous experimental and theoretical studies that interpolate elasticity data linearly between hcp-Fe and eps-(Fe,Si) at high pressure and temperature to infer inner core compositions. In contrast, hcp- (Fe0.875O0.125) solid solutions become stable only at inner core densities. For lower densities we find that at least the C44 elastic constant is negative. The elastic properties of hcp-(Fe1- x,Ox) are significantly different from the hcp-(Fe1-x,Six) solid solutions. These differences may be sufficient to use currently available seismic observations to evaluate the significance of hcp-(Fe1- x,Ox) solid solutions and the structural modification we find . Thus, if oxygen is accommodated in the new structural form it may provide an explanation for the seismically observed layering of the Earth's inner core and oxygen should be reconsidered as a candidate for the light element in the Earth's core.
DI53B-02 INVITED
A Glass Transition in the Lowermost Outer Core?
New theories for the viscosity of metallic melts at core pressures and temperatures, together with observations of translational modes of oscillation of the Earth's solid inner core, have suggested that the viscosity of Earth's liquid outer core may approach 1011 Pa-sec near the inner core boundary. If the viscosity of the lowermost 450 km of the outer core (F region) were in this range, it may be in a glassy state, characterized by a frequency dependent shear modulus and increased viscoselastic attenuation. The amplitudes of compressional seismic waves reflected from the inner core boundary (PKiKP) and diffracted around its upper surface (PKP-C diff) can be used to explore the range of allowable non-zero shear modulus and viscoelastic attenuation consistent with seismic observations. The absolute amplitudes of partially reflected PKiKP in the 30-80° range are consistent with a shear modulus at the bottom of the F region as high as one-third of that in the crystallized inner core. The frequency dependent decay of PKP-Cdiff phase at distances greater than 155 ° is consistent with attenuation in a metallic slurry of viscosity on the order 1011 Pa-sec, with the effect of increased viscoelastic attenuation trading off with the effect of increased shear modulus. Confirmation of lateral variations of seismic velocities and attenuation in the F region would also support high viscosity in F. In modeling PKiKP absolute amplitudes, care must be taken in modeling the incident source spectrum, and tradeoffs exist in the shear modulus and density discontinuities at the inner core boundary. In modeling distance decay of PKP-Cdiff, the competing effects of inner core topography need to be assessed. With these caveats, seismic body waves cannot reject the hypothesis of a glassy region in the lowermost outer core. If a glassy F region exists, its effect on the observed flattening of the depth gradient in compressional wave velocity must be included in estimates of the depth dependence of light alloying elements in the F region and their relation to the solidification process of the inner core.
DI53B-03
The dynamical influence of the inner core on the geomagnetic field
The Earth's magnetic field is generated in the liquid outer core by convective flows, powered by the release of buoyant material at the inner core boundary. Although uncertain, the age of the inner core is estimated by energy budget arguments to be substantially less than the known age of the geomagnetic field (Labrosse et al, 2001). In the absence of an inner core, thermal convection can power the dynamo although this requires a high initial temperature and remains rather controversial. Nevertheless, without any definitive theory of Earth's evolution however, we must accept the possibility that the geodynamo can robustly operate both with and without an inner core, and it is of significant interest to characterise any differences in the associated mechanisms of magnetic field growth. On a dynamical level, the existence of an inner core has a fundamental impact on the form of the convection splitting the outer core into two regions, separated by the tangent cylinder, an imaginary cylinder coaxial with the rotation axis and tangent to the inner core. Indeed, there is now mounting evidence, both observational (e.g. Hulot et al., 2002) and computational (e.g. Sreenivasan and Jones, 2006) of different dynamics in these two regions indicated for instance, by retrograde vortices close to the poles. As far as the magnetic field is concerned however, the influence of the tangent cylinder is not so obvious. Observations cannot probe beneath the core-mantle boundary and as such, it is not possible to predict what differences in the field structure exist, if any, between these two regions. Geodynamo models would be an excellent tool to investigate this further but, out of computational necessary, operate in parameter regimes remote from Earth's core. Consequently it is unclear what resemblence their internal structure bears to the real Earth, despite geophysically plausible fields on the core-mantle boundary. The correct dynamical regime in the Earth's core is described by the so-called magnetostrophic balance between Coriolis, pressure, buoyancy and the magnetic Lorentz forces. Solutions of the associated equations have remained out of reach because of a certain necessary condition on the magnetic field, the so-called Taylor constraint (Taylor, 1963), requiring zero net magnetic torque on any fluid cylinder coaxial with the rotation axis. However, in an exciting recent development, we showed that magnetic fields that satisfy the Taylor constraint in a full sphere can be constructed, paving the way for more realistic models of Earth's dynamo (Livermore et al., submitted). In this work, we build on these ideas and explore the construction of simple magnetic fields that satisfy Taylor's constraint in a spherical shell. Here, the influence of the inner core has a significant impact on the structure of the constraint, splitting up the set of cylinders into two distinct classes, those inside and outside the tangent cylinder. In comparison to the full sphere, the effective number of constraints on the magnetic field is greatly increased by this more realistic geometry, and we attempt to identify what fundamental changes in the permissible morphology of the magnetic fields arise.
DI53B-04 INVITED
Deformation of Directionally Solidifying Alloys, and the Earth's Inner Core
Most explanations for the elastic anisotropy of the Earth's inner core (IC) rely on a lattice preferred orientation (LPO) of hexagonal close-packed (hcp) Fe crystals, though it is not known with certainty that this is the stable phase of Fe throughout the IC. The explanations for the LPO fall broadly into two classes, solidification and deformation texturing. However, it seems increasingly likely that no one explanation may suffice to understand the complex IC structure that seismologists are beginning to reveal. In this study we examine experimentally the high temperature deformation of an hcp zinc-rich tin alloy that has been directionally solidified. The directionally solidified castings have the columnar, dendritic structure that has been proposed for the IC. We then heat a slice of a casting to a high homologous temperature, at which the small fraction of interdendritic tin melts. While held at this temperature, we give the slice a differential twist to produce a constant strain rate, while measuring the torque to infer the stress. We examine each slice before and after deformation for changes in crystalline orientation, microstructure (morphology and grain size), and chemical variations. The goals of this study are to understand the high temperature deformation mechanism of directionally solidifying alloys, the relevant lengthscale for deformation (grain size or dendritic spacing), the role of dynamic recovery and recrystallization, and the resulting changes in microstructure and LPO. We will present preliminary results as we examine a range of homologous temperatures, strain rates, and total strains. The hope is that this study will help to interpret IC elastic and attenuation anisotropies, and to give insight on the relevant lengthscale and viscosity of the IC, both of which relate to the deformation mechanism.
DI53B-05
Inner Core Boundary Properties From Amplitudes and Anti-Correlation in Observations of Body Waves Reflected at the Boundaries of the Core
The inner core boundary of the Earth is characterised by a discontinuous change in elastic properties between the outer and inner core. The size and nature of these discontinuities provides direct constraints on the age of the inner core and the energy needed to sustain Earth's magnetic field. In ray theory, a measure of the density ratio at the inner core boundary is given by the amplitude ratio of P waves reflected by the outer core boundary (PcP) and inner core boundary (PKiKP). We present observations of high-quality arrivals of both PcP and PKiKP waves originated from earthquakes and nuclear explosions. A new method that considers microseismic and event-generated noise is introduced for reliable measurements of absolute and relative amplitudes and their uncertainties. A number of numerical experiments were conducted to model the amplitude ratio (PKiKP/PcP), including a search for the mechanism for reproducing an observed anti- correlation of the amplitude of these wave types. Bounds were placed on the inner core discontinuity properties using the new amplitude measurements and their uncertainties, and a number of mechanisms were eliminated as an explanation for the anti-correlation of PKiKP and PcP amplitudes. We favour a mechanism of heterogeneity in earthquake and explosion radiation patterns that can selectively increase or decrease energy in the direction of the PcP or PKiKP wave because rate of change of PKiKP with increasing vertical take-off angle is opposite to that of PcP.
DI53B-06
Constraints from normal modes and exotic body waves on the radial structure of the Earth's inner core
The thermal and compositional structure of the inner core is key to understanding the inner workings of our planet. Seismology is the only technique that can measure its elastic parameters and density, providing constraints for mineral physics and dynamical modeling of the Earth's core. Unfortunately, even the core's radial structure remains largely unknown due to different and uneven sampling of body waves and whole Earth oscillations. For example, widely ranging values of inner core shear attenuation and impedance contrast at the inner core boundary are invoked to explain observations of inner core shear waves such as PKJKP. Here, we aim to better constrain the radial structure by combining observations of exotic body waves and whole Earth oscillations. Normal modes tightly constrain the inner core compressional velocity at 11.15 km/sec and the shear wave velocity at 3.55 km/sec, in agreement with PKJKP observations. The normal modes also provide information on the shear attenuation structure at long periods, which will be compared with PKJKP observations at shorter periods. The detailed structure at the inner core boundary will be studied using other exotic inner core waves. PKIIKP is strongly sensitive to the jump in compressional velocity and PKJKP provides information on the shear wave velocity contrast. Using constraints from normal mode data, we search large earthquakes for observations of these waves. The (non)-identification of exotic inner core waves provides contraints on the core's radial structure.
DI53B-07
Windows into the solid-state viscosity and seismic anisotropy of Earth's inner core inferred from experiments on micro-fabricated, controlled-geometry samples
The diffusion coefficients of core materials provide important constraints on the solid-state viscosity of the inner core, which governs the dominant mechanism for observed seismic anisotropy. Of considerable interest is the grain size and defect structure present in Earth's solid inner core, and what effect that may have on rates of diffusion. To quantify Fe/Ni diffusion under high pressure and temperature conditions, multi-layered, controlled-geometry samples were prepared for use in the laser-heated diamond anvil cell. These samples are composed of a Fe64Ni36 alloy foil substrate with 300 nm of Ni sputter-deposited atop the alloy foil to introduce an initial discontinuity in the samples. Fe and Ni were used because they are the two elements that comprise ~90% of the inner core. Experiments were performed at pressures of 20-50 GPa and laser heated to homologous temperatures of 0.8-1.1 Tm. We present results on TEM (transmission electron microscopy) analysis of prepared foils both before and after heating in order to gain a better understanding of the defect structure in our samples, and whether there is sufficient time for the defects to anneal themselves under the time constraints of laser heating. Results of EDS (energy dispersive x-ray spectroscopy) analysis through a cross-section of the hotspot provide compositional profiles across the initial compositional discontinuity to determine diffusion coefficients of core materials. The combined result of diffusivities and defect densities then allow for an examination of creep mechanisms in core materials under conditions of high pressure and temperature, resulting in insight into the solid-state viscosity of the inner core and the deformation mechanism responsible for observed seismic anisotropy.
DI53B-08
Phase relations of Fe-Si-Ni alloys at core conditions: Implications for the Earth inner core
The Earth core consists of a liquid outer core and a solid inner core, which are believed to be made predominantly of iron (Fe). Among all crystallographic structures proposed, a consensus has more or less emerged with the hexagonal closed packed structure -hcp- for iron. The question of the structure of this alloy at core conditions, in particular in vicinity of the melting line is however still largely debated. Among others, a possible thermal and chemical stabilization of body-centered cubic iron in the Earth's core has indeed been proposed with the theoretical calculations of Vocadlo et al. [Nature, 424, 536, 2003]. Recent X-ray experiments have shown the existence of such a bcc structure above 220 GPa at high-temperature for iron- nickel alloys [Dubrovinsky et al., Science, 316, 1880, 2007]. It is also known from density systematics that the Earth's core is made of iron alloyed with light elements [see Poirier, Phys. Earth Planet. Int., 85, 319, 1994]. We recently proposed a compositional model for the Earth's inner core from a systematic study of the effect of light elements on sound velocities at high pressure. Our preferred core model is an inner core which contains 2.3 wt % silicon and traces of oxygen [see Badro et al., Earth Planet. Sci. Lett., 254, 233, 2007 for more details]. Recent studies, however, suggest that small amount of silicon or nickel can substantially affect the phase relations and thermodynamic properties of iron alloys. We present results from an X-ray diffraction carried out at ESRF at high-pressure and high-temperature, using a state-of-the-art double sided laser heating system. We address the question of the structure of this alloy at core conditions. Two different alloys have been synthesized for this experiment, with Fe : 92.4, Si : 3.7, Ni 3.9 and Fe: 88.4, Si: 7.3, Ni: 4.3 in wt %, so as to satisfy the core preferred compositional model described in Badro et al. [2007]. The samples were loaded in a diamond anvil cell with neon as pressure transmitting medium transmitting medium, and subsequently analyzed by diffraction collected on a CCD detector during laser-heating at pressure. Experiments were carried out between 20 and 200 GPa, and 1500-5000 K. Our results show an increase of the pressure transition from bcc to hcp with increasing silicon content, with much more precise pressure transitions than previously published. X-ray diffraction pattern contain fcc or hcp at high-temperature and high-pressure conditions. If an expansion of the fcc stability field is observed with increasing silicon and/or nickel content, our observations show a wide stability of hcp-iron alloys up to 200 GPa and high-temperature. These results are discussed in the light of recent experimental and theoretical investigations.
DI53B-09 INVITED
Geodynamic Constraints on the Inner Core Super-Rotation
Seismic observations suggest that the inner core is rotating at a slightly faster rate than the mantle. This may represent a steady super-rotation, or an oscillating inner core which has been rotating on average faster than the mantle for the past few decades. Both scenarios are dynamically feasible from the perspective of core dynamics. However, in both of these scenarios, gravitational coupling between the heterogeneous density structures of the mantle and inner core should resist a differential rotation of the inner core, and this places limits on the latter's rate of rotation. In this work, I use simple models of angular momentum balance between the inner core, fluid core and mantle to derive constraints on the rates of both the steady and time- dependent inner core rotation. I show that a steadily rotating inner core is limited by the torque on the mantle from surface forces at the core-mantle boundary. As for the rate of rotation of an oscillating inner core, it is constrained by the changes in mantle rotation induced by gravitational coupling, which must not exceed the observed changes in length of day. The maximum rate of inner core rotation, for either scenarios, can be derived from a few parameters such as the strength of the gravitational coupling, the inner core viscosity and the electrical conductivity of the lower mantle. These parameters are not well known, but their values may be estimated from a range of different observations. Based on these, it is found that the maximum rotation amplitude of an inner core oscillation at a period of 60 yr is ~0.03 deg/yr, while the maximum amplitude of a steadily rotating inner core is ~0.3 deg/yr. The general conclusion that is reached is that if the inner core rotation rate is as large as reported in some studies, approximately 0.2 deg/yr, then it is more likely to represent a steady rotation than a slow oscillation. For smaller rotation rates, either scenarios are possible options.