[S-IT20] 地球深部科学

[SIT20-24] Valence/spin states of iron in peridotite glass to megabar pressure implications for dense iron-rich silicate melt at the bottom of the mantle

★Invited Papers

*増野 いづみ¹、芳野 極¹、三井 隆也²、藤原 孝将²、井上 紗綾子³、境 毅³ (1.岡山大学惑星物質研究所、2.量子科学技術研究開発機構、3.愛媛大学地球深部ダイナミクス研究センター)

キーワード：ペリドタイトガラス、下部マントル、電気伝導度、鉄のスピン状態

In this study, we have conducted high-pressure electrical conductivity and energy-domain synchrotron Mössbauer spectroscopic measurements of peridotite glass up to ~150 GPa at room temperature. We observed that the shape of the Mössbauer spectra changes above 60 GPa, corresponding to the emergence of the new components of Fe²⁺. The center shift (CS)-quadrupole splitting (QS) values of the new component are in the middle between high spin (HS) Fe²⁺ and intermediate spin (IS) Fe²⁺ from 60 GPa to 100 GPa, and then become almost the same as the values of IS Fe2+ above 100 GPa. This means that the spin state of Fe²⁺ gradually changes from HS to IS with pressure. We did not observe the discontinuous change in the electrical conductivity profile as a function of pressure due to the gradual transition of HS-IS. However, we observed the change in the trend of the electrical conductivity profile around ~83 GPa, which is likely induced by the increase of the relative abundance of the new component. The CS value of Fe³⁺ also decreases above 100 GPa, suggesting the HS to low-spin (LS) transition. Generally, the ionic radii of IS and LS should be smaller than HS. We suggest that the spin state changes to IS Fe²⁺ and LS Fe³⁺ may be associated with the structural transition in Si-O coordination number in the glass because the structure of the glass will become more compact.
The analogy between melt and glass is still under debate, however, the decreases of iron partition coefficient with increasing pressure previously observed in Nomura et al. (2011) and Tateno et al. (2014) may be due to the spin state change of Fe2+ in the melts. In addition to the effects of the changes of the coordination number and the iron partition coefficient between melt and crystal, the spin transitions possibly occur in the silicate melts at the pressure shallower than the ULVZs and CMB would make the melts denser. Therefore, silicate melts would be gravitationally stable at the base of the mantle and the silicate melts could be the cause of the ULVZs.