Crystal Structure and Equation of State of Mantle and Core Materials to 100 GPa and 2500 K

Thomas Duffy (duffy@princeton.edu)¹, Sang-Heon Shim (sangshim@princeton.edu)¹, Abby Kavner (abby@princeton.edu)¹, Sergio Speziale (speziale@princeton.edu)¹ & Guoyin Shen (shen@cars.uchicago.edu)²

¹ Department of Geosciences, Princeton University, Princeton, NJ 08648, USA

² GSECARS, The University of Chicago, Chicago, IL 60637, USA

We report new measurements of the crystal structure and equation of state of mantle and core materials directly under the pressure and temperature conditions of the deep Earth. Experiments were performed using double-sided laser heating in the diamond anvil cell together with synchrotron diffraction at the GSECARS sector of the Advanced Photon Source. X-ray diffraction was carried out using both energy dispersive and angle dispersive geometries. Previous studies of MgSiO₃perovskite have reached conflicting conclusions regarding the presence of a phase transition at very high pressures. In our experiments at 85 GPa and at 100 GPa, the dominant feature we observe up to 2500 K is the diffraction pattern of MgSiO₃ demonstrating the stability of this material under deep mantle conditions. For CaSiO₃ perovskite, we also observe no phase transformation to 96 GPa and 2400 K. The stress state of the sample before, during, and after heating was characterized using lattice strain theory. This analysis shows that previous studies of CaSiO₃ perovskite at 300 K may have been contaminated by deviatoric stresses. From our data, a complete P-V-T equation of state was determined. The bulk modulus at lower mantle conditions is 10% lower than found in previous studies. We have also studied the sulfides FeS and PbS to 35 GPa and 2000 K. For FeS, these experiments directly cover the P-T range of the core of Mars. Two separate polymorphs of FeS are found to be stable at conditions of the Martian core. The measured densities of FeS were used to evaluate models of the Martian interior and place new constraints on the core thickness.

High Pressure Partitioning of Pt in An-Di Eutectic Melt: Implications for the Accretion History of the Earth

Werner Ertel (wertel@lpl.arizona.edu)¹, Michael J. Drake (drake@lpl.arizona.edu)¹ & Paul. J. Sylvester (pauls@sparky2.esd.mun.ca)²

¹ Lunar and Planetary Laboratory, The University of Arizona, 1629 East University Boulevard, Tucson, Az 85721, USA

² Department of Earth Science, Memorial University of Newfoundland, 300 Prince Philip Drive, St. John's Newfoundland, Canada

Highly siderophile elements (HSE) like Pt can monitor core formation processes based on their siderophile - Fe-loving - geochemical behavior. Tremendous effort in the last 10 years contributed to a better understanding of their partitioning behavior at low pressures (1 bar) in basaltic melt systems. However, little is known about the effect of pressure on the partitioning behavior of these elements, an issue of interest in respect to Earth accretion scenarios (deep magma ocean versus late veneer). We are, therefore, investigating the partitioning behavior of Pt at pressures up to 90 kbar. The starting melt composition is the same as in lower pressure experiments (1 bar, 2 and 4 kbar), the eutectic composition of An-Di. Charges were contained in Pt capsules coated with Al₂O₃ cement to prevent graphite penetration. Experiments up to 20 kbar were performed in a Quickpress piston cylinder apparatus, while experiments up to 90 kbar were performed in a Walker-type multi anvil apparatus. For each set of piston cylinder runs (10 and 20 kbar) time series experiments were performed at constant temperature to determine the minimum run duration for the attainment of equilibrium. A minimum of 6 hours for piston cylinder experiments was determined to be sufficient for attainment of equilibrium. A second set of experiments investigated the temperature dependence of Pt solubility at both 10 and 20 kbars. Platinum solubility increases slightly with increasing temperature while an increase in pressure from 10 to 20 kbar decreases the Pt solubility by about an order of magnitude. Further piston cylinder and multi anvil experiments are currently under way to extend our knowledge to pressure conditions at the base of a deep magma ocean, believed to be present during the late stages of accretion and core formation.

Re Solubility in Haplobasaltic Melts at Variable Oxygen Fugacity: Experimental Determination Using Mechanically Assisted Equilibration and LA-ICP-MS

Werner Ertel (wertel@lpl.arizona.edu)¹, Don B. Dingwell (don.dingwell@uni-bayreuth.de)², Paul J. Sylvester (pauls@sparky2.esd.mun)³ & Hugh St. C. O'Neill (hugh.oneill@anu.edu.au)⁴

¹ Lunar and Planetary Laboratory, The University of Arizona, 1629 East University Boulevard, Tucson AZ 85721, USA

² Bayerisches Geoinstitut, Universitaet Bayreuth, 95440 Bayreuth, Germany

³ Department of Earth Science, Memorial University of Newfoundland, 300 Prince Philip Drive, St. John's, Nfld, Canada

⁴ Research School of Earth Sciences, The Australian National University, Canberra ACT 0200, Australia

The geochemical behavior of Re and other highly siderophile elements (HSE) makes them key monitors for core formation processes. Their apparent overabundances in the upper mantle (chondrite relative) have been the spur for the introduction of new accretion models to explain these findings. Despite the great need for accurate partitioning data for Re, experiments aimed at unravelling their 1 atm partitioning behavior have been fraught with difficulties, both experimental and analytical. Here we have combined new experimental and analytical techniques to clarify the picture.

We investigated the solubility behavior of Re at 1 atm in a haplobasaltic silicate melt using the mechanically assisted equilibration technique of Dingwell et al. (1994) over an oxygen fugacity range of -12 < logfO₂ < -7 at a constant temperature of 1400°C. Quenched samples were checked for major element composition by electron microprobe. Re concentrations in these samples were in a first attempt determined by instrumental neutron activation analyses (INAA) resulting solubilities from 40000 ppb down to about 1000 ppb. The severe scatter in these results suggested huge nugget formation problems, and we choose to duplicate the analyses applying LA-ICP-MS techniques in time resolved measuring mode. Thus, a newly developed analytical technique - laser-ablation inductively-coupled plasma mass spectrometry (LA-ICP-MS) was attempted to evaluate the possible presence of micronuggets in the melts.

Solubilities determined by LA-ICP-MS in the identical samples used for INAA before ranged from 10000 ppb down to 10 ppb, and are to our knowledge the lowest Re solubilities ever reported. Re is dissolved as oxidized species, changing its oxidation state from 2+ at the lowest fO₂s over intermediate oxidation states to 6+ at the highest fO₂s investigated.

The present results emphasize the importance of low oxidation states and extremely high metal-silicate partition coefficients of Re at the reducing conditions relevant to geochemical partitioning in the deep Earth.