Experimental Studies of Early Differentiation and Evolution of the Earth

为了模拟深于1000 km的岩浆海洋（MO）的核分离，我们旨在实验研究在超高压下熔融铁（MI）和硅酸盐之间的反应和元素分配。利用烧结金刚石作为劈球式多砧装置的顶砧材料，可以进行压力大于30 GPa的熔化实验。使用二次离子质谱仪的分析方法的创新使得在5 μ m直径的区域上测定小于1 ppm的元素浓度成为可能。基于这些新开发的方法，硅酸盐和硅酸铁系统的熔融关系已被检查和MI和硅酸盐液体（SL）或共存的固相之间的主要和微量元素的分配已被确定为压力的函数。主要结果如下：（1）地球原生物质的液相线相为镁橄榄石（Mw）; ...更多信息 在MO底部，深度为600-1000 km。然而，SL+Mw场的非常窄的温度间隔表明钙钛矿（Pv）在核心形成之前也将与MI共存。(2)在高于20 GPa，溶解的Si和O在MI从SL进行显着的压力。因此，深部MO的核分凝可显著提高地幔Mg/Si比值。(3)Mn，Ni和Co之间的MI和SL，Mw，和Pv到26 GPa的分配系数表明，Ni和Co的地幔丰度可以解释，如果核心材料将已与SL和Mw在MO的底部比900公里深的平衡。然而，Pv的共存需要更深的MO。(4)研究发现，在20 GPa以上的压力下，SL中的稀土元素大量溶解于MI中，轻、重稀土元素分馏明显，导致成核后原始地幔中轻稀土元素贫化。这些结果对地幔化学研究具有重要意义，因为迄今为止，原始地幔的稀土元素相对丰度一直被直观地假定为与辉长岩的稀土元素相对丰度相同。少

In order to simulate the core segregation in a magma ocean (MO) extended deeper than 1000 km depth, we aimed to examine experimentally reactions and elemental partitioning between molten iron (MI) and silicates at ultrahigh pressures. Melting experiment at pressures higher than 30 GPa can be carried out by utilizing sintered diamond as anvil material of the splitsphere multi-anvil apparatus. Innovation of analytical method using a secondary ion mass spectrometer has made it possible to determine concentration of elements less than 1ppm over an area of 5mum diameter. Based on these newly developed methods, melting relations of silicate and iron silicate systems have been examined and the partitioning of major and trace elements between MI and silicate liquid (SL) or coexisting solid phases has been determined as function of pressure. Important results obtained in the present research are summarized as follows : (1) Liquidus phase of the primary Earth's material is magnesiow ustite (Mw) … More at the bottom of MO with depths of 600-1000 km. However, a very narrow temperature interval of the SL+Mw field suggests that perovskite (Pv) would also have coexisted with MI prior to core formation. (2) At higher than 20 GPa, dissolution of Si and O in MI from SL proceeds noticeably with pressure. Thus Mg/Si ratio of the mantle could remarkably be increased by the core segregation in a deep MO.(3) Partition co efficients of Mn, Ni, and Co between MI and SL,Mw, and Pv to 26 GPa indicate that Ni and Co abundance of the mantle can be explained if the core material would have been equilibrated with SL and Mw at the bottom of MO deeper than 900 km. However, the coexistence of Pv demands a much deeper MO.(4) It has been found that appreciable amounts of rare earth elements (REE) dissolve in MI from SL at pressures higher than 20 GPa with a remarkable fractionation between light and heavy REE,resulting in depletion in light REE in the primitive mantle after the core formation. The results are of crucial importance for the mantle chemistry because the relative abundace of REE for the primitive mantle have intuitively assumed so far to be same as those of chondrites. Less