Iron-bearing Oxides, Silicate Glasses and Carbonates at Lower Mantle Pressures

Abstract

Iron is the fourth most abundant element in Earth’s mantle and it affects geophysically observable properties, such as the density, sound velocities, viscosity and transport properties of mantle phases. Additionally, the concentration of iron in minerals and melts of the lower mantle dictates their structure and stability. Thus, understanding the effect of iron is important to interpret seismically observable complexities and understand their effect on processes in the deep Earth. In this thesis, I use experimental and theoretical methods to improve our understanding of the high-pressure behavior of iron-bearing periclase (“ferropericlase”), silicate glasses and carbonates. Ferropericlase is thought to represent a significant fraction of Earth’s lower mantle, and may explain the slow compressional and shear sound velocities of ultra-low velocity zones at the core-mantle boundary. To understand the effect of iron concentration on the (Mg,Fe)O solid solution, the equation of state and hyperfine parameters of (Mg,Fe)O with 48 mol% FeO were measured using X-ray powder diffraction and time-domain Mossbauer spectroscopy, respectively, and the spin crossover behavior was compared to that of (Mg,Fe)O with 10 to 60 mol% FeO. I find that iron-rich ferropericlase at core-mantle boundary pressures likely contains a significant fraction of high-spin iron, contributing a positive buoyancy to promote topographic relief of ultra-low velocity zones in the lowermost mantle. Some ultra-low velocity zones, particularly those at the base of the central parts of large low shear velocity provinces, may be best explained by the presence of iron-bearing silicate melts. The behavior of iron in silicate melts is poorly understood, but may be approximated by iron-bearing silicate glasses. I measured the hyperfine parameters of iron-bearing rhyolitic glass up to ~120 GPa and basaltic glasses up to ~90 GPa using time-domain Mossbauer spectroscopy. Iron within these glasses experiences changes in coordination environment with increasing pressure without undergoing a high-spin to low-spin transition. Thus, ferrous iron in chemically–complex silicate melts likely exists in a high-spin state throughout most of Earth’s mantle. Decomposition of carbonates may be responsible for creating silicate melts within the lower mantle by lowering the melting temperature of surrounding rock. Identifying and characterizing the stability of carbonate phases is therefore a necessary step towards understanding the transport and storage of carbon in Earth’s interior. Dolomite is one of the major mineral forms in which carbon is subducted into the Earth’s mantle. Although iron-free dolomite is expected to break down upon compression into single-cation carbonates, high-pressure polymorphs of iron-bearing dolomite may resist decomposition. Using a genetic algorithm that predicts crystal structures (USPEX), I have found a monoclinic phase with space group C2/c that has a lower energy than all previously reported dolomite structures at pressures above 15 GPa, and the substitution of iron for magnesium stabilizes monoclinic dolomite with respect to decomposition at certain pressures of the lower mantle. In this thesis, I demonstrate that iron undergoes a spin transition in (Mg,Fe)O, (Mg,Fe)CO3 and Ca(Mg,Fe)(CO3)2, while iron in basaltic and silicate glasses likely does not experience a spin transition up to lowermost mantle pressures. Additionally, I find that the amount of iron in (Mg,Fe)O and Ca(Mg,Fe)(CO3)2 dictates the dynamic and thermodynamic stabilities of those phases within Earth’s lower mantle.