U41A-01 INVITED
Chemical interaction of the core and mantle
New geochemical techniques, available only since the last decade, have opened the study of the mantle geochemistry of siderophile elements, i.e., those elements that are preferentially present in the metallic core (Fe, Ni, W, Re, Os, Pt, etc.). Evidence that the core and mantle chemically interact was first recognized by the presence of radiogenic Os isotope ratios in Hawaiian lavas. Such Os isotope signals are generated by the growth of the solid inner core due to preferential partitioning of the parental Re and Pt isotopes in the outer core. Successful models of Os isotope evolution for the outer core require an early differentiation of the inner core and strong solid metal-liquid metal partitioning of Re, Os and Pt. An early growth (pre-3.5 Ga) of the inner core is supported by Os isotope analysis of 2.7 Ga Kostomuksha komatiites. Such models represent an extreme endmember among models of the timing of inner core differentiation. The partitioning of Os and Pt are testable, with some difficulty, by ingenious new experiments using a Fe-Ru solid solution to stabilize the hcp-Fe phase at pressures accessable by the multi-anvil cell. Measurements of the platinum group element abundances coupled with Os isotope ratios in the Kostomuksha, and other komatiites, indicate that the mechanism of core-mantle interaction takes place by isotopic and chemical equilibration between the outer core and lower mantle. No evidence is found for the large- scale addition of outer core material to the base of the mantle. A chemical exchange near equilibrium is possibly consistent with Fe/Mn data for Hawaii, as well. The mantle is composed of Fe-Mg silicates and variation of the ratio of Fe/Mg is an important potential source of density variation observed by seismic tomography. In the lower mantle, the mineralogy is Mg-silicate perovskite or post-perovskite phase and ferropericlase, an FeO-MgO solid solution. The core is composed of Fe-Ni-X alloy, where X is a light element. Potentially, oxygen enters the core as FeO. If FeO is a component of the core, the chemical exchange of FeO should be controlled by equilibrium between ferropericlase and metallic liquid. Evidence for this is apparent in the systematically higher Fe/Mn ratios observed in many mantle plume-derived basalt lavas. Mantle geochemistry is finally coming to grips with the single most important question that geophysicists may ask of it: major element variation in mantle plumes that reach the surface to form volcanic islands.
U41A-02
Compositional and Thermal Effects on Transition Zone Structure
Our present picture of Earth's chemical and dynamical evolution points to an uncertain knowledge of the composition of the mantle, which is important for us to understand the thermal and dynamical evolution of Earth. Dynamical models suggest a mantle made of a mechanical mixture of basalt and harzburgite, especially perhaps a basaltic gradient and a concentration of basalt in the lower mantle. We use a self-consistent thermodynamic calculation of phase equilibria and physical properties to explore the Vs structure of the mechanical mixture in the transition zone based on the mantle composition of Workman and Hart [2004]. We performed calculations for a mechanical mixture model with varying fractions of basalt (0% to 40%), along adiabats with potential temperature ranging from 1400K to 1800K. In general, seismic structure is insensitive to basalt fraction in the transition zone, except the location and thickness of the 410 and 660 km discontinuities, which are linearly dependent on the basalt fraction. Another important characteristic is that the 660 km discontinuity shows a double-step discontinuity, in which the second step is caused by the sudden transition of garnet- perovskite+Ilmenite. Compared with the influence of basalt fraction, the influence of the geotherm is very significant for the shear-wave velocity, including the location and thickness of 410 and 660 km discontinuities, which are linearly related to temperature, in the transition zone. The non-global 520 km discontinuity structure could be explained by the variation of temperature. In fact it is not caused by the global transformation of wadsleyite to ringwoodite. Instead, the existence of the 520 km discontinuity corresponds to low temperature, which causes large amount of garnet to transform to cpv in the transition zone of wadsleyite to ringwoodite, and the reverse is true at high temperature. When comparing our results with PREM, we find a velocity deficit that increases in magnitude from 400 to 740 km depth, which indicates a ~100 K colder subadiabatic geotherm. Besides the deficit in shear wave velocity, the gradient from 410 to 660 km would also indicate a basaltic gradient in the mantle.
U41A-03
Grain size variations in the Earth's mantle and preservation of primordial mantle heterogeneities
Mantle convection causes changes in the grain sizes of mantle minerals due to dynamic recrystallization, phase transitions, grain growth and Ostwald ripening. These processes are highly non-uniform and may produce large spatio-temporal variations in the grain size and the viscosity. Various estimates suggest that the grain size is likely to vary from 0.01 to 1 cm, which is consistent with observations of xenoliths and recent seismic constraints. The primordial crystal sizes established upon crystallization of the early magma ocean can reach several centimeters in diameter. No subsolidus or supersolidus process seems to be capable of producing a locally averaged grain size larger than 10 cm. The spinel-perovskite phase transformation is the most significant one for grain size reduction. Other phase transformations cause a transient grain size reduction over length scales of about several kilometers within the phase boundary, although nanometer-size grains can be produced over large regions in the cold subducting slabs. Numerical simulations of mantle convection with grain size evolution shows that while the spinel-perovskite phase transformation keeps the grain size in the most of the lower mantle small, some material can preserve its original grain size. The large viscosity of these regions prevents them from mixing and preserves their primordial composition. The amount of the preserved heterogeneities depends on various factors including the vigor of convection, the compositional buoyancy of the heterogeneities and the kinetics of grain growth.
U41A-04
Chemical dynamics in the core: Contributions from high-pressure phase equilibria of Fe-X systems
We discuss advances in mineral physics of Fe-X binary systems, and their implications for dynamics of the core and core-mantle interactions. Recent progress, using both experimental and theoretical methods, has improved our understanding of phase diagrams and equations of state of Fe-rich binary systems. Synchrotron x-ray diffraction methods, in particular, have allowed a more detailed picture of several Fe-X phase diagrams at high pressures and temperatures, and also improved the high-P,T equations of state of relevant metallic phases. The liquidus slope, dT/dC, is an important parameter driving the rate of chemical buoyancy release during crystallization of the core. Experimental constraints on this parameter are improving: dT/dC for the Fe-O system is estimated to be about 4 times lower than dT/dC in the Fe-S system at 60 GPa. An oxygen rich core would consequently generate more compositional buoyancy, per degree of cooling, during inner core crystallization than a sulfur rich core would. The composition of a eutectic, relative to the outer core composition, is an important constraint on allowable phase diagrams to describe the core. The Fe-S eutectic decreases rapidly with increasing pressure from 31.6 wt% at 1 bar to 15 wt% at 20 GPa, but changes very little at higher pressures to at least 60 GPa. The Fe-O eutectic is only 2 wt% O at 16 GPa, which is not consistent with the requirements of crystallization of metal from the outer core; however, recent work demonstrates that the eutectic increases in O content with pressure, reaching 10 wt% O at 90 GPa. Ternary systems are less well investigated, but are becoming increasingly important as more detailed geochemical estimates of core composition are proposed.
U41A-05
Plate mantle coupling and the state of stress of the lithosphere
The internal deformation of plates is controlled by the state of stress of the lithosphere. The sources of stress include variations in density and thickness of the lithosphere as well as basal and edge tractions that arise from plate driving forces. The coupling between mantle stresses and the lithosphere is determined by the rheology of the plates themselves and that of the mantle underneath. We present a study of the state of stress of the lithosphere that examines the importance of lithospheric structure, including continental root, and mantle shear tractions that arise from mantle density and viscosity heterogeneity. We compute global mantle flow in the presence of strong lateral viscosity variations using the finite element code CitComS and the lithospheric stresses with the finite element package ABAQUS. We examine the effects of variations in lithospheric structure, looking at different end-members of compensation (Pratt and Airy) as well as a new global model of lithospheric density and thickness, whose properties are determined via a self-consistent thermodynamic method. We find that both mantle and lithospheric heterogeneity are important in determining the state of stress in continents and oceans and the mantle signal is particularly strong in SE Asia and other areas of long-lived subduction. The primary effect of lateral variations in mantle viscosity is to strongly couple continents and deeper mantle, enhancing deformation near continental roots.
U41A-06 INVITED
The coupled modeling of mantle convection and continental tectonics.
The continental crust is a small fraction of the volume of the mantle and yet it has a disproportionate influence on the dynamics of the Earth. This is due to its relatively high buoyancy which ensures that it largely remains within the cool thermal boundary layer of the Earth's inner convection cycle. The continental crust has both a thermal and mechanical influence on the underlying convection. We will briefly summarize a number of interesting observations which relate to the interplay between the negative buoyancy of the crust, its tendency to insulate the mantle, and the heterogeneity in strength of the continental lithosphere. We next examine how the positive buoyancy of the continental crust and the negative buoyancy of subducting slabs interact - can subducted buoyant material interrupt subduction, and can subduction disrupt a strong continental block.