High-pressure polymorphs of olivine and the 660-km seismic discontinuity

Abstract

It had long been accepted that the 400-km seismic discontinuity in the Earth's mantle results from the phase transition of (Mg,Fe)2-SiO4-olivine to its high-pressure polymorph beta-spinel (wadsleyite), and that the 660-km discontinuity results from the breakdown of the higher-pressure polymorph gamma-spinel (ringwoodite) to MgSiO3-perovskite and (Mg,Fe)O-magnesiowüstite. An in situ multi-anvil-press X-ray study indicated, however, that the phase boundary of the latter transition occurs at pressures 2 GPa lower than had been found in earlier studies using multi-anvil recovery experiments and laser-heated diamond-anvil cells. Such a lower-pressure phase boundary would be irreconcilable with the accuracy of seismic measurements of the 660-km discontinuity, and would thus require a mineral composition of the mantle that is significantly different from what is currently thought. Here, however, we present measurements made with a laser-heated diamond-anvil cell which indicate that gamma-Mg2SiO4 is stable up to pressure and temperature conditions equivalent to 660-km depth in the Earth's mantle (24 GPa and 1,900 K) and then breaks down into MgSiO3-perovskite and MgO (periclase). We paid special attention to pressure accuracy and thermal pressure in our experiments, and to ensuring that our experiments were performed under nearly hydrostatic, inert pressure conditions using a variety of heating methods. We infer that these factors are responsible for the different results obtained in our experiments compared to the in situ multi-anvil-press study.

The Effect of Water on Mantle Phase Transitions

“…Ignorance of the pressure effect on thermocouple Electromotive force (EMF) in multianvil experiments underestimates temperatures (e.g., Ohtani 1979), and it further reduces discrepancy. The results obtained by Chudinovskikh and Boehler (2001) for Mg 2 SiO 4 using the DAC are consistent with the data of Fei et al (2004a). The data determined by Shim et al (2001) using the DAC indicate that the post-spinel transformation boundary (Figs.…”

Ultrasonic measurements of single-crystal gold under hydrostatic pressures up to 8 GPa in a Kawai-type multi-anvil apparatus

“…Anderson et al's EOS model of gold ( 0 T K ′ = 5.5) [7] was used as the pressure scale in the first in situ X-ray diffraction study of the post-spinel phase transformation of Mg 2 SiO 4 [26] . The determined phase boundary, however, is about 2 GPa lower than the 660-km discontinuity, while the post-spinel boundaries determined by other pressure scales such as platinum and SrB 4 O 7 :Sm 2+ agree well with the 660-km discontinuity [27,28] . A simultaneous high-pressure and high-temperature experiment on Au and MgO indicated that Anderson et al's EOS model underestimates pressures by 1.4 ± 0.3 GPa relative to the Matsui's MgO scale at 20-24 GPa and 1873 K [29] .…”

Uniqueness of Steiner minimal trees on boundaries in general position

“…The boundary determined by in situ diffraction study using the Anderson scale [14] is situated at 20.5 GPa and 2100 K, which is ∼2 GPa lower than the earlier estimates. Recent studies [15,16] show that the discrepancy is due to the error in the Anderson scale. Figure 3 shows the pressure dependence of the thermal expansion coefficients around 1000 K. There are discrepancies between the results from the EXAFS experiments and Anderson's EOS, and they become larger with the rise of pressure.…”