Researches from the Max Planck Institute for Chemistry in Mainz/Germany studied the stability and solidus of materials relevant to the lower mantle at very high pressures and temperatures. These laboratory experiments help explain seismic measurements from the deep Earth (Science, vol. 280, 26 June 1998 and Science, vol 282, 10 Juli 1998).
Seismological observations reveal several anomalies in the lowermost part of the Earth´s mantle: A discontinuous increase in velocity some 200 km above the core-mantle boundary and an ultra-low velocity zone in the very near vicinity of the core. While the first observation has been interpreted as being due to a possible breakdown of the major lower-mantle constituent (Mg,Fe)SiO3-perovskite to its component (Mg,Fe)O and a dense form of SiO2, the second observation has been interpreted to be due to partial melting.
The breakdown of perovskite not only requires that the resulting mineral assemblage is denser than that of perovskite, it would also result in significant changes in the chemical and rheological properties of the deep mantle. Partial melting would require a melting behavior that is in drastic contrast to previous melting experiments on the major lower mantle components and our knowledge of the eutectic behavior observed at low pressure.
R. Boehler, G. Sergiou, and A. Zerr, scientists from the High Pressure Group at the Max Planck Institute for Chemistry in Mainz/Germany, have examined the stability of perovskite by heating stoichiometric mixtures of (Mg,Fe)O and SiO2 in a diamond anvil cell with a high power CO2-laser to about 3000 K in a hydrostatic medium at pressures up to 100 Gpa (106 atm). This technique avoids both large pressure and temperature gradients which is an essential requirement for a reliable assessment of the stability of solids, especially under extreme conditions. In all experiments (Mg,Fe)SiO3-perovskite was formed which unambiguously shows that this phase is stable under lower mantle conditions. Moreover, both MgSiO3 crystal and glasses subjected to similar pressure/temperature conditions also formed perovskite as a single phase, as evident from Raman spectra. No SiO2 was detected. Furthermore, the fluorescence spectra of chromium(III)-doped perovskite were recorded at analogous conditions, which in the case of decomposition would produce a strong fluorescence line of MgO:Cr3+. No MgO was detected. All these results are a strong indication that decomposition of perovskite in the lower mantle is highly unlikely and is not the cause for unusual seismic patterns (Science, vol. 280, 26 June 1998).
In another study (Science, vol. 282, 10 Juli 1998), the same authors investigated the solidus of the mantle relevant material to about 60 GPa (equivalent to the 1500 km depth inside the Earth); in earlier examinations the solidus was constrained to about 25 GPa being equivalent to about 700 km depth. The small samples (70 m m x 70 m m x 15 m m) embedded in the argon pressure medium were heated using CO2-laser radiation in a diamond anvil cell. Changes in the texture on the recovered sample surface were used to verify partial melting of the sample material using a Scanning Electron Microscope (SEM) and an Atomic Force Microscope (AFM). This new approach of the melting temperature determination at high pressures in a diamond anvil cell is usable for any one- or multicomponent material, especially if no other signs of melting (for example change in the absorption of the heating laser radiation) can be used. The results show that the melting temperature of the multicomponent material studied is significantly lower than the melting temperatures of its end-members phases. Even though the solidus in the deep mantle lies over 1500 K above the average present-day geoterm, its value at the core-mantle boundary (CMB) approaches the outer core temperature. Thus, partial melting of mantle material close to the outer core may explain observed seismic anomalies in this region.
It should be generally stressed that the solidus of multicomponent mantle material is the key parameter in modeling the early Earth. It is possible that after the Earth´s formation a significant part of the Earth was at least partially molten. Even the presence of a small amount of melt results in a significant weakening of mantle material. Moreover, the temperatures at which mantle material solidify marks a change in the character of convection features during the Earth’s evolution.
Published: 13-07-98
Contact: Reinhard Boehler
Max-Planck Institute for Chemistry
Mainz/Germany
Phone: 49-6131-305252
Fax: 49-6131-305330
Journal
Science