About

It is an exhilarating time in Earth science, where we can use computations based only on fundamental physics to accurately predict properties of materials important for understanding the Earth that cannot yet be measured, or use these methods to understand the behaviour of Earth materials. Until recently, our accuracy was limited depending on the complexity of the material, but now we can accurately predict properties of all Earth materials. For the same reasons, it is an exciting time in materials science, as new materials can be designed computationally, using suitably accurate methods.

In order to understand the behaviour of the Earth, we model it, but to model it we require knowledge of the inherent properties of its materials.

The Mystery of Iron Oxide - The journey that led to ToMCaT

The Earth has long been thought of as a stony, insulating mantle surrounding a metallic core however recent results suggest it may not be so simple. There are probably significant metallic minerals in the Earth, and maybe even a metallic layer at the base of the mantle. How FeO metallisation behaves in solid solution with MgO in ferropericlase is crucial to understanding the Earth.

FeO is the simplest iron bearing mineral, and an endmember of magnesiowustite or ferropericlase (Mg,Fe)O, believed to be the second most common mineral in the deep Earth. Whereas MgO is a simple ionic crystal with a large band gap, for which simple electronic structure methods work well, FeO is electronically complex, with an open-shelled ferrous Fe²⁺ ion. Nevertheless, there is almost perfect solid solution from pure MgO to FeO. Complicating the experimental study of FeO is that it has abundant iron vacancies, so that even in equilibrium with metallic iron at zero pressure it is more like Fe0.92O, containing Fe³⁺ clusters. At high pressures, or with the introduction of MgO in solid solution, it becomes stoichiometric.

Picture right: Space-filling model of part of the crystal structure of Manganese(II) oxide, MnO. Structural data from the CrystalMaker 8.1 structure library, originally from Zhang J Physics and Chemistry of Minerals 26 (1999) 644-648.

Using standard methods, FeO is predicted to be a metal, but it is really a good insulator at ambient conditions. At high pressures and temperatures, shock experiments showed that FeO became metallic but it was believed to be due to a change in structure to the NiAs structure. At high pressures and lower temperatures, FeO has a large rhombohedral distortion - like pulling on the corner of a cube - and was believed to remain insulating.

Recently, using a new approach termed DFT-DMFT, some remarkable predictions were made, the most surprising being

1. there is no change in crystal structure associated with this insulator to metal transition,
2. it does not occur in any of the known ways that other materials metallise (thus a “new kind of metal) and
3. temperature is an important parameter, as well as pressure.

It turns out that this discovery has major implications for the Earth, many in far flung fields ranging from modelling the propagation of Earth’s magnetic field, to core mantle coupling leading to variations in the length of the day and consideration of the behaviour of exoplanets and super-Earth’s.

In addition to the geophysical and geochemical importance of the metallisation in FeO, it is of great interest in physics and materials science. The type of metallisation discovered is different than any known transition. The typical model for metallisation of a Mott insulator is growth of normal metallic bands in between the upper and lower Hubbard bands via band widening due to increase in pressure, and would not be very temperature sensitive. A second expected method would be broadening of electronic states due to temperature to fill the band gap. The third expected method would be a structural phase transition to a crystal structure that is metallic. None of these methods have been observed in the new studies, rather, there is a collapse of the Mott state due to quantum and thermal fluctuations between high- and low-spin states. This is a new type of process not previously considered, and may be of importance in a whole new class of metallic materials, perhaps even a new class of superconductors.

Project Outline

ToMCaT (Theory of Mantle, Core and Technological Materials) is an ERC funded research project at University of Munich (LMU) starting in 2013 and funded for 5 years. The research group will predict properties of Earth and
technological materials using fundamental physics, compute properties of minerals and melts to better
understand them, and estimate properties when data is unavailable. This project will constrain properties of major problematic minerals, melts, and aqueous fluids crucial to modelling of the Earth, including equations of state, phase transitions, electrical and thermal conductivity, chemical diffusivity, viscosity and rheology. This project will concentrate on iron and other transition metal bearing Earth materials, for which conventional electronic structure methods are inadequate. Whereas properties of closed-shell atomic systems and simple metals can now be computed to accuracy limited only by available computing time, open-shelled systems continue to be problematic.

European Union’s Seventh Framework Programme

As part of the European Union’s Seventh Framework Programme, the ToMCaT project was granted a € 2.7 million award for the 5 year project. The project is hosted by University of Munich (LMU).