Binding Of The Melting Curve Of Iron At High Pressure And In The Field Of Nuclear Candidate Components

The purpose of this thesis is to study the physical properties of the materials in Earth's deep interior, here mainly restricting to the Earth's core ones, at high pressure and high temperature, by means of theories of condensed matter and recently developed high pressure techniques. So this work is a research topic belonging to cross disciplines. The content of it can be divided into two parts. One is the high pressure melting curve of iron, the other is a constrained investigation for the candidate Earth's outer core composition, in which a Fe-O-S system with 90.12/7.98/1.9 wt. % is chosen as the sample materials and its Hugoniot curve, Hugoniot sound velocities and melting temperature measurements are made and analyzed. Some main and /or innovative results are summarized as follows:I Iron is the main component of Earth's core. It is also acknowledged that the Earth's core could be distinguished into a liquid outer core and a solid inner core. Therefore, the high pressure melting line of iron is a much concerned scientific problem for us since from which the temperature at the inner core boundary (ICB, at 330GPa) can be inferred as an reference data with first-order approximation (also called anchor temperature), so as to further examine the temperature profile within Earth's core. But, an unresolved scientific issue has appeared and always troubled us for nearly twenty years, i.e. the melting temperatures of iron measured by the shock compression (SC) experiment above 200 GPa are higher than those measured by diamond anvil cell (DAC) technique below 100 GPa while the former is extrapolated by Lindemann law to the corresponding lower pressure region. Although recent developed accurate X-ray diffraction and double-side laser heating techniques have been introduced in DAC measurements and superheating correction methods have been introduced in SC's data analysis, the melting temperature data discrepancies between DAC and SC experiments have been reduced but not closed nowadays. For this reason, a thermodynamic way to direct calculating the equilibrium melting temperature in SC's data analysis, instead of the superheating correction methods respectively proposed by Li Xijun (Doctoral thesis, China Academic of Engineering Physics, 2000) and Luo et al (Phys. Earth Planet. Inter, 2004,143:369), is proposed in this work. Using this new data analysis technique, an equilibrium melting temperature of 5300K at 260GPa from SC measurements can be informed, from which a surprisingly well agreement between this new SC's data and new DAC data (Shen et al., Geophys Res Lett, 1998,25: 373; Ma et al., Phys. Earth Planet. Inter., 2004, 143: 455) joined by Lindemann law has been observed. This is a main achievementmade in this paper. The rationality of this method is further supported by three kinds of the following calculations.(1) According to the exiting OK isothermal equation of state and the effective Griineisen parameter yeff and specific heat Cv, the Hugoniots relevant to solid, solid-liquid mixture and liquid regions for iron have been calculated by means of this thermodynamic analysis method, respectively. In solid and liquid regions, the calculated Hugoniots are in good agreement with experiments, which also show this calculation method is an excellent tool to distinguish solid and liquid phase from the experimental Hugoniot points. This is an additional unique merit of the method. As for in the solid-liquid mixture, a slightly scattered experimental data appear when compared with the calculation Hugoniot, because of a two-wave shock structure formation in the region.(2) Along the above-mentioned calculational Hugoniot, the Hugoniot bulk sound velocities relevant to solid and liquid phases have also been computed with same thermodynamic ways. Both are in accord with experiments. Otherwise, if we use Brown's Hugoniot parameters (J. Appl. Phys, 2000, 88: 5496) i.e. without distinguishing solid and liquid state, the computed sound velocities are 3% less than the experiments for solid iron. This is one of important conclusions revealed in this work.(3) Griineisen parameter y is one of the important thermodynamic quantities since it often appears in the equation of state and melting temperature calculations. But the available data of 7 for iron along Hugoniot are not only limited, but also diverse with each other. For this reason, we make a calculation with same thermodynamic technique for the effective Griineisen parameter along Hugoniot, denoted by yeff, based on the available data of specific heat of Cv and the Griineisen parameter y from both lattice (N) and electron (e) contributions. The results demonstrated that yeff can be in good agreement with experimental yN at lower pressure and with the calculated data deduced from experimental Hugoniot bulk sound velocity at higher pressure region, showing a rational or a correct physical basis for this yeff calculations.II In the candidate Earth's outer core compositions study, ternary alloys of Fe-O-S with 90.12/7.98/1.9 by wt.% (corresponding to Fe/FeO/FeS with 58.96/35.83/5.21 by wt.%) is chosen as the sample material, based on Alfe's suggestions (Nature, 2000 ,405:172) and the PREM model (Dziewonski, Earth Planet. Interiors, 1981,25: 297) constrained by both density and sound velocity vs pressure relations. This is the first experimental investigation for a ternary alloy candidate model of Earth's outer core composition research. (1) Hugoniot curve for this Fe-O-S system with initial density p0—6.69 (±0.06) g/cm3 hasbeen made in the range from 50 to 210 GPa. The measured Hugoniot parameters are Q—3.97 (±0.07) km/s, X=1.58 (±0.03) , which is well agreement with the calculated data deduced from the individual Hugoniot of Fe, FeO, and FeS through the volume additive modelling, Introduction to the Experimental Equation of State, Beijing: Science, Press, 1999; Lin et al., Chin J High Press Phys.Lett, 1998, 12: 40). The calculated 0 K isothermal from the measured Hugoniot through Wu's method (Chin Phys Lett, 2002, 19: 528) is also consistent well with that calculated from the individual 0 K isotherms of the ingredients through the volume additive law. Both results indicate that the volume additive model is reasonable for this kind of mixture and no detectable chemical reactions occur among Fe, FeO and FeS during shock compression.(2) Hugoniot sound velocities measurements show that the equilibrium melting temperature of 3830K at the shock-induced completely melting point of 167GPa can be inferred based on the thermodynamic energy balance calculation, and, therefore, the melting curve relevant to this kind of mixture can be calculated according to Lindemann law. To further verifying the rationality of this calculated melting curve, The Tan-Dai- Ahrens model (High Press Res, 1990; Geophys. Res. Lett, 2000) has been used to direct measure the release melting temperatures for comparison. The results are Tm=3100K at 108GPa and Tm =2880K at 90GPa, both lie on the calculated melting curve within the experimental error range. The melting temperature of this Fe-O-S is 5400K while extrapolated to ICB at 330 GPa.(3) At ICB (330 GPa), the melting temperatures of pure iron and this Fe-O-S material are 5930 K and 5400K respectively. The contribution of O and S to the melting depression is about 600K at 33OGPa. In order to compare with PREM model, the Hugoniot density and bulk sound velocity should be corrected for the temperature difference between Hugoniot state and the Earth's outer core state calculated by this thermodynamic analysis method, with an uncertainty better than that of the previously proposed " average temperature correction coefficient". The results show that the ternary alloy candidate Earth's outer core compositions of Fe-O-S system can be basically constrained with PREM model. A small concentration adjustments for O and S elements vs core depth should be made in further, due to the need for improving the coincidence degree between the PREM model and the behavior of candidate material of Fe-O-S system.