Optical signatures of low spin Fe3+: a new probe for the spin state of bridgmanite and post-perovskite

Sergey S. Lobanov,Han Hsu,Jung-Fu Lin,Takashi Yoshino,Alexander F. Goncharov

DOI: https://doi.org/10.48550/arXiv.1610.01048

2016-10-04

Abstract:Iron spin transition directly affects properties of lower mantle minerals and can thus alter geophysical and geochemical characteristics of the deep Earth. While the spin transition in ferropericlase has been vigorously established at P ~ 60 GPa and 300 K, experimental evidence for spin transitions in other rock-forming minerals, such as bridgmanite and post-perovskite, remains controversial. Multiple valence, spin, and coordination states of iron in bridgmanite and post-perovskite are difficult to resolve with conventional spin-probing techniques. Optical spectroscopy, on the other hand, is sensitive to high/low spin ferrous/ferric iron at different sites; thus, it can be a powerful probe for spin transitions. Here we establish the optical signature of low spin Fe3+O6, a plausible low spin unit in bridgmanite and post-perovskite, by optical absorption experiments in diamond anvil cells. We show that the optical absorption of Fe3+O6 in NAL (new aluminous phase) is very sensitive to the iron spin state and represents a model behavior of bridgmanite and post-perovskite in the deep lower mantle across a spin transition. Specifically, an absorption band centered at ~ 19000 cm-1 is characteristic of the 2T2g to 2T1g (2A2g) transition in low spin Fe3+ in NAL at 40 GPa. This new spectroscopic information constrains the crystal field splitting energy of low spin Fe3+ to ~ 22200 cm-1 which we also independently confirm by our first-principles calculations. Together with available information on the electronic structure of Fe3+O6-compounds, we constrain the spin-pairing energy of Fe3+ in an octahedral field to ~ 20000-23000 cm-1. This implies that octahedrally-coordinated Fe3+ in bridgmanite is low spin at P > ~ 40 GPa.

Geophysics,Materials Science