The stability of Mg(2)SiO(4), a major constituent in the Earth's mantle, has been investigated experimentally by in situ observation with synchrotron radiation. A cubic-type high-pressure apparatus equipped with sintered diamond anvils has been used over pressures of 11 to 15 gigapascals and temperatures of 800 degrees to 1600 degrees C. The phase stability of alpha-Mg(2)SiO(4) and beta-Mg(2)SiO(4) was determined by taking account of the kinetic behavior of transition. The phase boundary between alpha-Mg(2)SiO(4) and beta-Mg(2)SiO(4) is approximated by the linear expression P = (9.3 +/- 0.1) + (0.0036 +/- 0.0002)T where P is pressure in gigapascals and T is temperature in degrees Celsius.

Abstract. The phase boundary between wadsleyite and ringwoodite in Mg2SiO 4 composition was determined by in situ observation using synchrotron X-ray and multi anvil apparatus in KEK, Tsukuba, Japan. An energy dispersive method was employed using the Ge solid state detector and the white X-ray beam from the synchrotron radiation source. The pressure was determined by the equation of state of NaCl. The stability field was identified by the change in intensities of diffraction lines of each phases. As a result, the phase boundary is expressed as a linear equation P=I0.32(28)+0.00691(9)xT, where P is pressure in gigapascals and T is temperature in degrees Celsius. IntroductionOlivine is the major constituent in the Earth's upper mantle, and the significant seismic discontinuities, which locate 400 km and 670 km deep, are considered to be caused by the phase transition of olivine to wadsleyite (modified spinel structure) and the decomposition of ringwoodite (spinel structure) to magnesiowustite and There are a number of quench experiments so far on the phase boundary between wadsleyite and ringwoodite [e.g., Kawada, 1977; $uito, 1977;Katsura and Ito, 1989], however, the in situ X-ray determination have not been conducted yet.In the quench experiments, pressure was estimated from the calibration curve, which is based on the fixed point at room and/or high temperatures, and the considerable uncertainty remains in pressure (e.g., +I.5GPa above 14 GPa

A double‐stage multianvil system was developed to conduct in situ X ray diffraction study at high pressures. The system generates pressures and temperature above 25 GPa and 1200°C and has been used in conjunction with intense synchrotron radiation to study the phase transitions of MgSiO3 and the thermal expansion of MgSiO3 perovskite. At pressures between 20 and 30 GPa, we observed a series of phase transitions from spinel+stishovite through ilmenite to perovskite. Spinel+stishovite transforms into ilmenite at 20.0±1.0 GPa, and ilmenite transforms into perovskite at 24.0±0.5 GPa at 800°C. The phase boundary between ilmenite and perovskite is estimated to be P (GPa) = (24.0±0.5) ‐ (0.0025±0.0025)(T‐800)(°C) by combining the present results and previous constraints on the slope of the phase boundary from the thermodynamic properties. Measured unit cell parameters of MgSiOs perovskite indicate that the average volumetric thermal expansion coefficient of MgSiOs perovskite over 25°–1200°C is 2.0±0.4×10−5 K−1 at 25 GPa. These results provide the phase stability relation of MgSiO3 and the thermal equation of state of MgSiO3 perovskite at pressures corresponding to the boundary between the upper and lower mantle. However, uncertainties involved in extrapolating to the actual mantle temperatures do not allow evaluation of the chemical structure near the 670 km discontinuity in the mantle.

A high pressure and high temperature in‐situ X‐ray diffraction experiment of MgSiO3 perovskite was conducted using a sintered diamond multianvil apparatus (SDMA8) combined with the synchrotron radiation. The unit cell parameters of the perovskite were measured up to 1095°C at about 20.5 GPa. The thermal expansion coefficient is 0.8(±0.8)×10−5 K−1 at 20.5 GPa and room temperature.

Numerical calculations have been done to reveal temperature distributions and cooling speeds of laser-heated samples in a diamond anvil cell. The distributions were calculated for variable experimental parameters including the diameter of the laser beam, anvil gap, and sample. The results show that the radial temperature distribution in all samples is Gaussian. The axial temperature gradient is ∼10 K/μm in samples heated by a broad laser beam of 100 μm diameter, and is ∼102 K/μm when a narrow laser beam of 10 μm diameter is used. The broad beam can generate a less extreme temperature gradient in both radial and axial directions as compared with the narrow beam, whereas the temperature gradient strongly depends on the anvil gap, although this is minimized when a narrow beam is used. When the narrow beam is used to heat samples, the surface temperature of the anvil culet can be kept below 400 K; thus, the narrow beam is suitable for heating samples under high pressure while keeping anvil temperatures low. Cooling speeds of the laser-heated samples were found to be from 100 to 102 K/μs. The discrepancy between actual sample temperature and measured temperature was also evaluated. To measure the peak temperature of a sample heated by a laser with a Gaussian power distribution within 5% accuracy, radiation from a circular area of less than (1/2)σ of the Gaussian must be observed.