Implications for a Magma Ocean from in Situ High Pressure-Temperature Melting Experiments on the Allende Meteorite and Peridotite

Crystallization of chondritic and peridotitic material in a magma ocean can be used to simulate accretion and segregation of a bulk planet. Although chondritic and peridotitic phase equilibria below 20 GPa have been studied in detail, experiments above 20 GPa, and particularly >23 GPa, are lacking. Because this pressure range represents the transition into the Earth's lower mantle, as well as the potential bottom of a shallow magma ocean, experiments conducted in this pressure range are critical to understanding early crystallization of the deepest planetary mantles. The objective of this study is to determine liquidus phases in situ at high pressures and temperatures for a number of planetary mantle analog materials.

Experiments were conducted at the 10 MN (1,000 ton) LVP in 13-ID-D, GSECARS, APS. A 3mm TEL beamline modified assembly was used in all experiments - a Re furnace with lanthanum chromite insulating sleeve, alumina or MgO end caps and graphite capsule. Starting materials are fine powder, packed into the capsule. X-ray windows are a slit in the Re furnace and alumina or graphite plugs in the lanthanum chromite. MgO was used as a diffraction pressure standard. Runs were pressurized to 400, 600, 650 and 700 tons, and diffraction data were collected at temperature intervals between 1700 °C and 2400 °C.

Preliminary results for peridotite suggest the liquidus at 22 GPa occurs at 2350 °C, with ferropericlase as the liquidus phase. For Allende, Mg,Fe perovskite is the liquidus phase above 21 GPa and 2050 °C. A shallow perovskite liquidus slope, steep ferropericlase slope is consistent with deeper, hotter magma ocean. The ferropericlase slope may be very steep, and become the liquidus phase above 25 GPa. In a magma ocean crystallizing from the bottom up, this scenario may not crystallize enough perovskite to drive an oxidation pump of the mantle.