HR: 14:10h
AN: MR33B-03 INVITED     [Abstracts]
TI: Solving the Controversy about the Role of Al-Defects on the Equation of State of the Main Lower Mantle Perovskite Phase
AU: * Andrault, D
EM: andrault@impmc.jussieu.fr
AF: IMPMC, 140 rue de Lourmel, Paris, 75015 France
AU: Bolfan-Casanova, N
EM: n.bolfan@opgc.univ-bpclermont.fr
AF: Magmas et Volcans, 5 rue Kessler, Clermont-Ferrand, 63038 France
AU: Bouhifd, M A
EM: ali.bouhifd@earth.ox.ac.uk
AF: Earth Sciences Dpt, parks Road, Oxford, OX1 3PR United Kingdom
AU: Guignot, N
EM: guignot@esrf.fr
AF: ESRF, 6 rue Jules Horowitz, Grenoble, 38000 France
AU: Kawamoto, T
EM: kawamoto@bep.vgs.kyoto-u.ac.jp
AF: Geothermal Sciences, Kyoto University, Beppu, 874-0903 Japan
AB: We performed a new compression study of Al-bearing (Mg,Fe)SiO3 perovskite phases (Al-Pv) synthesized at variable pressure and temperature conditions and using different composition of starting materials by varying the excess of MgO or SiO2. For the different runs we refined bulk moduli (K0) from 235 to 270 GPa, in agreement with the range of values reported in the literature. Such a scattering relies on the complex crystal chemistry of this phase. In particular, the Al-substitution can occur via two competitive mechanisms producing oxygen vacancies (Sub-1) or coupled Al-substitution on both perovskite sites (Sub-2). Previous results of ab-initio calculations suggested that only Sub-1 should have a significant effect on K0, by increasing the compressibility due to the presence of oxygen vacancies (1). Our different compression curves can all be explained by considering that the competitive Sub-1 and Sub-2 mechanisms are favored at higher temperatures and higher pressures, respectively. These trends are compatible with previous reports (1-3). On the other hand, the effect of MgO and SiO2 activities appears to be minor for the pressures and temperatures investigated in this study. Note that the synthesis conditions are dominant in the result. This can be explained as the diffusion processes are certainly slowered due to the Al-Pv grain growth, and therefore the phase crystal chemistry is unlikely to evolve along the compression curve. Consequently, the large range of K0 reported from the different groups is understandable, because each used specific experimental techniques, and therefore produced Al-Pv samples with specific defect populations. Concerning the Fe-bearing Al-Pv relevant to the lower mantle, we stress that there is no strong controversy about its K0 at present. The previous reports (4, 5) and our new data set for a MORB-type Pv show almost no change or a slight increase of the K0-values. This effect is compatible with a large majority of Sub-2 expected when both Al and Fe3+ are present in the Al-Pv phase (6). Sub-2 should also be strongly favored at the very high pressures involved in the deep lower mantle. Therefore, even if some questions may remain about a possible lower K0 value just below the 670 km discontinuity, because of significant amount of Sub-1 at moderate pressures, the extrapolation to higher depths of these low K0 values is certainly irrelevant to the deep Earth. (1) Brodholt, Nature 407, 620-622 (2000). (2) Navrotsky et al., JGR, 108, 2330 (2003). (3) Akber-Knutson and Bukowinski, EPSL, 222, 317-330 (2004). (4) Ono et al., PEPI, 145, 9-17 (2004). (5) Andrault et al., EPSL, 193, 501-508 (2001). (6) Lauterbach et al., Contrib. Mineral. Petrol. 138, 17-26 (2000).
DE: 3900 MINERAL PHYSICS
DE: 3904 Defects
DE: 3919 Equations of state
DE: 3924 High-pressure behavior
DE: 3954 X-ray, neutron, and electron spectroscopy and diffraction
SC: Mineral and Rock Physics [MR]
MN: Fall Meeting 2005