Sound Velocity Of (Mg, Fe)SiO₃-Perovskite At High Pressure And Its Stability And Geophysical Implications

The information of the Earth's interior, especially the understanding of the mantle and the core of the Earth, is mostly provided by the combination of the geophysical data and experiments under simultaneous high pressure and temperature conditions. In the last decades, the phase transition investigations of the candidate materials for the Earth have exceeded a great success. However, the thermodynamic stability and possible phase transitions of the (Mg, Fe)SiO₃ -perovskite, which is generally thought to be the primary constituent of Earth's lower mantle, have been controversial for several years. Especially, recent seismic imaging produced a velocity anomaly at the depth 1600-1800km in the Earth's lower mantle. We should pay more attention to the experimental investigation of the thermodynamic stability of the (Mg, Fe)SiO₃ -perovskite at simultaneous high pressure and temperature in the Earth's lower mantle conditions, which will produce great geophysical significance.The main work and achievements of this thesis can be described as followings:(1) Synthetically analyzing the shock compression data and previous results, we realized the phase transition process of the (Mg, Fe)SiO₃ enstatite detailedly under shock loading and divided it into three phase regions, which are low-pressure phase region(LPR), mixed phase region(MPR), and high-pressure phase region(HPR), corresponding to the pressure at 0-40GPa, 40-67GPa, and 68-140GPa, respectively. The D-u and P-p curve for the assemblage of (Mg_0.92, Fe_0.08O(Mw) + SiO₂(St) were calculated and compared with the experimental data, and the significantly differences suggest no possibility of chemical decomposition of perovskite to oxides during the shock compression.(2) We obtained a new relationship of shock wave velocity D and particle velocity u, which can be described linearly by D = 3.74(±0.22) + 1.49(±0.05)u (km/s), by fitting our experiment data in high-pressure phase region (at pressures between 68 to 140 GPa), representing the property ofperovskite. The Gruneisen parameter y of perovskite can be acquired in the form of y=yo(po/p)9, by fitting the high-pressure phase data, where yo =1.82(2), #=1.64(1) and />o=4.19g/cm3. Furthermore, the shock data yield a zero-pressure isoentropic bulk modulus 1^=259.6(9) GPa and its pressure derivative Kos '=4.20(5), with po=4.19g/cm3, using the third-order Birch-Murnaghan finite strain equation of state.(3) We measured the sound velocities of (Mgo.92, Feo.o8)Si03 enstatite, using the two stage gas gun. The previously results of sound velocity measurement were calculated over again, using the new D-u relationship and Gruneisen parameter y, and gained the relationship of sound velocity and pressure, as well as the curve of sound velocity versus density. The two curves showed a positive and negative jump for both compressional and shear sound velocity, at the pressure about 68GPa and 83-85GPa, respectively. However, there was no evident change on bulk sound velocity curve.(4) In order to test the applicability of Birch's low, we collected and analyzed lots of sound velocity dada of metals and materials, then detailedly described the relationship of compressional sound velocity and density, shear sound velocity and density, and bulk sound velocity and density, under isothermal, isoentropic and Hugoniot condition, respectively.(5) Base on the conclusion that compressional sound velocity and density display the same linearly relationship under three compression conditions, we compared our sound velocity results with the measurements about (Mg, Fe)SiC>3 enstatite and perovskite by static compression experiments, and previous calculated sound velocity data of perovskite by the methods of first principle and ab intio calculation. From the comparison, we proved the phase region partition further, and confirmed perovskite phase exists at the pressure 68-140GPa.(6) Using two different methods, we calculated the shock temperature of (Mg, Fe)SiC>3-Perovskite, then obtained the slope of phase transition boundary at 83-85GPa, which was very consistent with the slope estimated by the change of free energy resulted from the electronic transitions of Fe in perovskite. Thereby, we came to the conclusion that the negative jump ofsound velocity was cause by the electronic transitions of high-spin to low-spin in perovskite, which can shorten the bond of Fe-O, induce the aberration of crystal lattices, and increase the condensability, correspondingly. So (Mg, Fe)SiO3 will turn into a new perovskite phase above 85GPa. However, the positive jump of sound velocity at 68GPa was because of the influence of low-pressure phase (Enstatite). (7) Introduced the calculation of temperature coefficient of bulk sound velocity, we corrected the bulk sound velocity to PREM temperature. Then the compressional and corrected bulk sound velocities were compared with PREM sound velocity profiles, which indicate that the compressional sound velocities were parallel with PREM profile, and induced a error only by 4.89-5.08%, supporting a perovskite-predominated lower mantle model. Nevertheless, the bulk sound velocities corrected by temperature and PREM sound velocity profile intersected at the depth 1720km in the lower mantle. Thereout we can deduce that the seismic radial discontinuity at the depth 1600-1800km in the lower mantle was possibly either a phase transition interface due to electronic transitions in perovskite, or a chemical interface.