Determination of the three phase region of the post-spinel transition in (Mg,Fe)2SiO4: explanation of the extreme sharpness of the 660-km discontinuity and implication for chemical structure and dynamics of the deep mantle

地震观测表明，660公里的不连续面厚度不到2公里，对应的气压为&lt；0.1 Gpa。这种锐度与410公里的不连续形成鲜明对比，后者的厚度估计为7公里。这种660公里的不连续通常可以用(Mg，Fe)2SiO4中的尖晶石后转变来解释，在这种转变中，(Mg，Fe)2SiO4环木分解成(mg，Fe)SiO_3钙钛矿和方镁石。由于这三种矿物都有镁-铁固溶体，所以尖晶石后转变应该有一个有限的间隔，因此我们需要一个特殊的解释来解释660公里不连续的极端尖锐。为此，必须确定尖晶石后转变的转变间隔。如果转换的压力区间为&gt；0.1 Gpa，我们将不得不重新考虑深部地幔的结构和动力学。首先，可能是上地幔和下地幔的化学成分不同，地幔对流至少部分是分层的。另一种解释是，缓慢的成核动力学阻止了尖晶石后转变的启动，一旦成核发生，转变就会非常迅速地进行。如果证明尖晶石后过渡的时间间隔非常薄，则可以通过660公里不连续面厚度的全球变化来评估地幔垂直流动的存在。以往的实验研究没有成功地确定尖晶石后转变的压力区间，这是因为它们在压力测定方面缺乏足够的精度，并且受到动力学迟滞的影响。在以往的研究中，压力测定的精度不优于0.3 Gpa，这太大了，无法确定过渡间隔，可以小于0.1 Gpa。相比之下，申请人已经建立了测定样品压力的实验技术，其精度为0.04 Gpa，这对本项目来说应该是足够的。他还建立了一种实验技术，通过使用助熔剂来获得矿物的平衡组成。结合这两种技术，我们将在2000K的压力-组成空间中确定环木石+钙钛矿+方镁石的整个三相区，并根据三相区的几何形状，考虑预期的与镁铁矿层的交换，估算出真实地幔中尖晶石后转变的压力区间。

Seismic observations indicate that the 660-km discontinuity is less than 2 km thick, corresponding to <0.1 GPa in pressure. Such sharpness is in contrast with that of the 410-km discontinuity, whose thickness is estimated to be 7 km. The 660-km discontinuity is usually explained by the postspinel transition in (Mg,Fe)2SiO4, in which (Mg,Fe)2SiO4 ringwoodite dissociates into (Mg,Fe)SiO3 perovskite plus periclase (Mg,Fe)O. Because these three minerals have Mg-Fe solid solutions, the postspinel transition should have a finite interval, and therefore we need a special explanation to account for the extreme sharpness of the 660-km discontinuity. For this reason, the transition interval of the postspinel transition must be determined. If the pressure interval of the transition is >0.1 GPa, we will have to reconsider the structure and dynamics of the deep mantle. Firstly, it is possible that the chemical compositions are different between the upper and lower mantles, and that mantle convection is at least partially layered. An alternative explanation is that the sluggish kinetics of nucleation prevents initiation of the postspinel transition, and once nucleation occurs, the transition proceeds very rapidly. If it is proved that the interval of the postspinel transition is extremely thin, the presence of vertical mantle flow could be assessed by global variation of the thickness of the 660-km discontinuity. Previous experimental studies have not successfully determined the pressure interval of the postspinel transition because they have lacked sufficient precision in pressure determination and suffered from the effects of sluggish kinetics. The precision of pressure determination in previous studies was no better than 0.3 GPa, which was too large to determine the transition interval, which could be less than 0.1 GPa. In contrast, the applicant has already established experimental techniques to determine sample pressure with a precision of 0.04 GPa, which should be sufficient for the present project. He has also established an experimental technique to obtain the equilibrium compositions of minerals by using a flux. Combining these two techniques, the whole three phase region of ringwoodite+perovskite+periclase will be determined in pressure-composition space at a constant temperature of 2000 K. The pressure interval of the postspinel transition in the real mantle will be estimated based on the geometry of the three phase region by considering the expected Mg-Fe exchange with majorite.