Experimental modeling of ferropericlase formation under upper mantle conditions
Final Report Abstract
In this project, we investigated the possibility of crystallising ferropericlase, (Fe,Mg)O, at pressure-temperature conditions corresponding to Earth’s upper mantle. Two different types of experiments were performed: i) one to determine the composition of silico-carbonate melt that is in equilibrium with ferropericlase and olivine as a function of pressure and temperature, and ii) the other to simulate ferropericlase crystallization during reduction of silicocarbonate melt, along with the concomitant formation of graphite or diamond. The equilibrium experiments were conducted at 5-12 GPa and 1500-1700°C and the reduction experiments were performed at 7.5 and 10 GPa and 1400-1600°C. In the equilibrium experiments, melts saturated in ferropericlase and olivine have SiO2 contents of 2-12 wt %, which increases with increasing temperature and MgO+FeO (Bulatov et al. submitted). On the other hand, SiO2 content varies inversely with CaO and calculated CO2 concentrations. The mineral-melt Fe-Mg partitioning, defined as KDmin-melt = (Fe/Mg)min x (Mg/Fe)melt is much stronger for ferropericlase than for olivine (KDol-melt = 0.4-0.6, while KDper-melt shifts from ~1.5 to ~2.5 as Mg# of the ferropericlase is lowered from 0.9 to 0.7). Inter-mineral Mg-Fe partitioning is also variable, with KDper-ol changing from 2.5 to 5.3 with increasing overall Fe content. Although the compositions of our silico-carbonate melts are similar to those obtained from experimental studies of melting carbonated harzburgite or lherzolite, which are saturated in olivine and low-Ca pyroxene, there are some differences that tend to make them mutually exclusive. Although the melt compositions seem to converge at high pressures (~12 GPa), we conclude that melts saturated in ferropericlase cannot be produced by crystallization of melts derived by melting carbonated peridotite under upper mantle conditions. Experiments involving the partial reduction of silico-carbonate melt reveal the co-crystallisation of ferropericlase and diamond by spontaneous nucleation. This can be understood by the reaction: MgCO3 (melt) = MgO (ferropericlase) + C (dia/gr) + O2. The ferropericlase in contact with the melt has a Mg# = 0.6-0.7. The configuration of the experiments is such that the reducing metallic Fe is physically separated from the silico-carbonate melt by a layer of olivine. This means that our results are not due to direct interaction between melt and metal and can be generalised to reducing conditions, even without metal saturation. The only requirement for ferropericlase crystallisation is that low-Ca pyroxene cannot be part of the mineral assemblage, implying that some preconditioning of the peridotite is required, potentially by previous interaction with percolating melts or fluids. The resulting ferropericlase crystallizing in equilibrium with olivine will be richer in Fe compared with lower mantle ferropericlase in equilibrium with bridgmanite, or indeed that in equilibrium with wadsleyite or ringwoodite in the transition zone. Thus, the considerable variation in Mg# values observed for some suites of ferropericlase inclusions in diamond could in part be attributable to ferropericlase formation in the upper mantle.
Publications
-
(2019) Ferropericlase crystallization under upper mantle conditions: implications for inclusions in diamond. European Geophysical Union (EGU) annual meeting, Vienna, EGU2019-9216
Woodland AB, Bulatov VK, Girnis AV, Brey GP, Höfer HE
