Alumina solubility in periclase determined to lower mantle conditions and implications for ferropericlase inclusions in diamonds

Titelangaben

Man, Lianjie ; Fei, Hongzhan ; Kim, Eun Jeong ; Néri, Adrien ; Xie, Longjian ; Frost, Daniel J.:
Alumina solubility in periclase determined to lower mantle conditions and implications for ferropericlase inclusions in diamonds.
In: Geochimica et Cosmochimica Acta. Bd. 375 (2024) . - S. 36-49.
ISSN 0016-7037
DOI: https://doi.org/10.1016/j.gca.2024.05.002

Volltext

Link zum Volltext (externe URL):

Abstract

A series of high-pressure multi-anvil experiments were conducted in the MgO-Al2O3 system between 15 and 50 GPa and at temperatures up to 2623 K, to determine the solubility of Al2O3 in periclase in the stability fields of coexisting spinel, corundum, the Mg2Al2O5 modified ludwigite phase, the MgAl2O4 calcium ferrite phase, and the MgAl2O4 calcium titanate phase. The Al2O3 solubility in periclase is strongly temperature-dependent over the conditions investigated. Conversely, periclase Al2O3 solubility exhibits a negative relationship with pressure. Experiments with up to 40 mol. FeO in ferropericlase show Al2O3 solubilities that are near identical to those of periclase, within experimental uncertainties. A simple thermodynamic model incorporating significant literature on ambient pressure measurements is able to describe periclase Al2O3 solubility up to 40 GPa. This model is used to investigate the Al2O3 contents of ferropericlase inclusions observed in natural diamonds, which vary up to 0.35 mol. . Pressure–temperature curves along which particular inclusions formed can be produced if equilibrium can be assumed with Al-bearing minerals found in the same diamond. Alternatively, the maximum possible Al2O3 content in ferropericlase can be determined for a certain pressure–temperature profile and inclusions with Al2O3 contents that exceed this curve can be excluded from those conditions. To obtain such solubility curves, calculations are performed for periclase coexisting with the Al-rich phases spinel, garnet, bridgmanite and the MgAl2O4 calcium ferrite phase. The calculations indicate that periclase, and by inference ferropericlase, Al2O3-contents cannot be greater than 0.5 mol. under present day adiabatic mantle temperatures and go through a minimum at mantle transition zone conditions. This excludes a number of Al-rich ferropericlase inclusions found in natural diamonds from being formed in the transition zone, unless temperatures were super-adiabatic. This subset of inclusions likely formed either at the base of the upper mantle or the top of the lower mantle, but must have formed at near adiabatic temperatures. The majority of ferropericlase inclusions have Al2O3 contents that would be consistent with formation in the transition zone at near slab temperatures, but could still have been formed at higher temperatures if Al2O3 activities were low.