In total compositional data of

In total 18 compositional data of Rw have been collected from the high-P experiments with some bulk compositions broadly appropriate for the upper thrombopoietin receptor agonist such as the depleted harzburgite with 82 wt% olivine removal (Irifune and Ringwood, 1987), pyrolite (Irifune, 1994), Tinaquillo lherzolite (Wood, 2000), MPY-90 pyrolite (Wood, 2000), K-doped KLB-1 (Wang and Takahashi, 2000), and KLB-1 (Hirose, 2002) at variant P-T conditions, ranging from 19 to 24 GPa and 1573–2373 K (Table 6). The stable phase assemblages observed in these high-P experiments were Rw ± Mj ± Ca-Pv ± Wd ± St (stishovite) ± Mg-Pv (MgSiO3-rich perovskite) ± Mw (magnesiowüstite) ± Ak (MgSiO3-rich akimotoite) ± K phase II ((K,Mg) (Al,Si)O4 phase) ± Melt. From these high-P experiments, we did not observe any correlation between the XFe of the Rw and the XFe of the bulk composition. In order to see the P effect on the composition of the Rw, we removed the data with either too high experimental T, or too low experimental T, or without Mj phase, and obtained a data set of 6 experiments (Runs E293 and E291 from Irifune (1994), Run 3312 from Wang and Takahashi (2000), and Runs C190, C189 and C179 from Hirose (2002)). A nominally negative correlation between the XFe of the Rw and the XFe of the bulk composition emerged, which made no sense and was thus interpreted as an artifice. Since the experimental T range was from 1773 to 1873 K only, the real cause to this correlation was pressure, as shown in Fig. 6. The correlation between the Rw composition and P illustrated in Fig. 6 can be described with the following equation:where P is in GPa (17 < P < 24 GPa). It should be emphasized that this correlation seems neither strongly dependent to the bulk compositions nor severely affected by the temperature in the lower part of the MTZ, which varies by ∼50 K only along the normal mantle geotherm from ∼17 to 24 GPa (generally ∼1773 K at 17 GPa increasing to ∼1823 K at 24 GPa; Ito and Katsura, 1989; Jackson, 1998; Jackson and Rigden, 1998). Accordingly, the XFe of the Rw decreases from ∼0.13 to 0.09 as P increases from ∼17 to 24 GPa. At P higher than 24 GPa, Rw breaks down to the phase assemblage Mg-Pv + Mw (Liu, 1976b; Fei et al., 2004). Several lines of data support the composition variation of the Rw in the lower part of the MTZ, as displayed in Fig. 6. High-P experiments in the simple system Mg2SiO4–Fe2SiO4 demonstrated that the compositions of the Wd-Rw pair moved to the Mg2SiO4 end as P increased (Ito and Katsura, 1989; Katsura and Ito, 1989; Fei et al., 1991; Frost, 2003b). With the introduction of Mj to the Ol-Wd-Rw system, a negative correlation between the XFe of the Wd/Mj and P was experimentally demonstrated by Irifune and Isshiki (1998); a similar negative correlation between the XFe of the Rw/Mj and P has not been experimentally demonstrated, but highly possibly remains valid, considering the similar crystal chemical behavior of Rw and Wd in coexistence with Mj (Frost, 2003a, 2003b). Furthermore, Ol inclusions in diamonds from the mantle (Hutchison et al., 2001; Sobolev et al., 2008), with some probably being a retrograde phase of Rw from the lower part of the MTZ (Hutchison et al., 2001), had various XFe values, from ∼0.06 to 0.14, implying a possible correlation between the XFe and P. To precisely establish such a correlation for the real Earth environment, more quantitative composition and inclusion-formation P data, as documented by Nestola et al. (2011b), should be reported for the natural samples first.
Conclusion With complete summarizing and critical analyzing of the relevant experimental data in the literature, we have arrived at the following conclusions:
Acknowledgments We thank Dr Q. Liu for providing some old reference papers and valuable discussions. We thank Dr F. Nestola, Dr M. Walter and an anonymous scientist for commenting on an early version of this manuscript. This investigation was financially supported by the National Natural Science Foundation of China (41273072 and 41090371), the Key Laboratory of Earth\'s Deep Interior, Chinese Academy of Sciences (DQSB201101), and the State Key Laboratory of Ore Deposit Geochemistry, Institute of Geochemistry, Chinese Academy of Sciences (201114).