DI52A-03
MELTING OF BRIDGMANITE TO 135 GPA: TOWARD A COHERENT MODEL FOR THE MELTING BEHAVIOR IN THE LOWER MANTLE

Friday, 18 December 2015: 10:50
303 (Moscone South)
Denis Andrault, University Blaise Pascal Clermont-Ferrand II, Clermont-Ferrand, France, Giacomo Pesce, Laboratoire Magmas et Volcans, Clermont-Ferrand Cedex, France and Nohamed Mezouar, ESRF European Synchrotron Radiation Facility, Grenoble, France
Abstract:
Our knowledge on the melting behavior in the deep mantle remains based on a limited number of experimental and theoretical works. Today, thanks to (i) availability of very brilliant X-ray synchrotron sources and (ii) improved control of the P-T conditions in the laser-heated diamond anvil cell (LH-DAC), the experimental results should reach some agreement about the melting diagrams. However, it is not the case and major controversies remain. For example, liquidi of peridotitic (1) and chondritic-type (2) mantles are reported with a temperature difference of ~1000 K at a pressure of ~90 GPa (corresponding to ~2000 km depth), which cannot be explained by the relatively small compositional difference between these two materials.

To bring new insights about the melting properties of the deep mantle, our strategy has been to study the melting curve of the end-member liquidus phase, the (Mg,Fe)(Al,Si)O3 bridgmanite (Bg), before applying a basic thermodynamical approach to the mineralogical system made of Bg, CaSiO3-perovskite and (Mg,Fe)O ferropericlase. Our approach cannot be as formal as currently done for melting in the shallow mantle, due to lack of major thermodynamical parameters. Still, our analysis yields original information, such as the degree of partial melting as a function of P, T and fraction of Bg in the geological material. The Mg/Si ratio in melts can also be addressed.

Concerning the controversy between LH-DAC experiments, it can be solved taking into account migration in the temperature gradient of the pseudo-eutectic melt, when the sample starts to melt. This effect is expected to occur more extensively in absence of an insulating material between the sample and the diamond anvils. It yields an overestimation of the liquidus temperature for a given chemical composition, due to loss of the most fusing elements.

References:

1. Fiquet et al. (2010) Melting of Peridotite to 140 Gigapascals. Science 329, 1516-1518.

2. Andrault et al. (2011) Melting curve of the deep mantle applied to properties of early magma ocean and actual core-mantle boundary. Earth Planet. Sci. Lett. 304, 251-259.