Diffusion in the DAC: Probing the physical state of the Earth's inner core

Lead Research Organisation: University of Bristol
Department Name: Earth Sciences


The inner core is the deepest and most inaccessible layer within the Earth. It is a sphere of solid iron, alloyed with some nickel and one or more 'light' elements (such as silicon, sulfur and carbon) and is 2500 km in diameter and grows by 1 mm every year. The conditions within it are unimaginably extreme, with pressures up to 3.6 million times the pressure at the surface of the Earth, and temperatures near 6000 C, similar to the surface of the sun. The inner core is central to the Earth system. As it crystallises it produces latent heat that helps drive convection in the liquid iron outer core above it, which drives the Earth's heat engine. Because some light elements such as oxygen prefer liquid iron to the solid, the growth of the inner core also changes the chemistry of the outer core, which may in turn effect the chemistry of the whole Earth. Many numerical models of the Earth's geodynamo - the mechanism whereby the convecting outer core produces the Earth's magnetic field - require the presence of the inner core for several reasons, including the extra heat produced by its growth. The geomagnetic field shields us from the harmful effects of the solar wind. Thus the inner core is important to the way in which life has developed on our planet and for maintaining the clement conditions on its surface today.

Yet there is much we do not understand about its structure and evolution. Much of what we do know comes from the study of seismic waves that pass through the inner core. These studies tell us that it is a complex place, with seismic waves travelling faster from pole to pole than they do through its equator. There is also a so-called 'hemispheric dichotomy' where seismic waves in the surface of the inner core travel faster in the eastern hemisphere than the western hemisphere. There is also evidence that there is an 'innermost inner core' around 1000 km in diameter with a different seismic signature to the rest of the inner core. Some studies have even suggested that the whole inner core may be rotating faster than the rest of the Earth.

Understanding and interpreting these surprising discoveries requires knowledge of the physical properties of the iron alloy from which the inner core is made. One of the most important of these properties is viscosity, for which no direct measurements have yet been made. Just such a measurement is the aim of this research.

Viscosity can be determined by measuring how fast iron atoms diffuse through crystals of iron. This will be done in two ways. Firstly, using the laser-heated diamond anvil cell at the School of Earth Sciences, University of Bristol. This equipment consists of two opposing gem-quality diamonds with flat tips, between which discs of iron, less than the diameter of a human hair (around 100 microns) and around 5 microns thick, are compressed to enormous pressures up to 200 million bar. The discs will be coated with a layer of iron, enriched in one of its isotopes, to act as a tracer. While at high-pressure, the sample is heated to temperatures up to 4000 C using infrared lasers and causing the tracer atoms to diffuse through the iron. Using a technique known as secondary ion mass spectrometry (SIMS) at the NERC ion microprobe facility, University of Edinburgh, we can strip away the iron, a few atomic layers at a time, measuring the tracer concentration at each stage. How far the tracer atoms managed to diffuse during a certain heating time tells us the diffusion rate, from which the viscosity can be determined.

Even with this technology, it will be difficult to reach the extreme conditions of the inner core so we will use a second method, ab initio computer simulation at the Department of Earth Sciences, UCL. In this method, a box of iron atoms is simulated within a computer using quantum mechanical methods. Hypothetical tracer atoms can be followed as they diffuse through the box, again allowing us to calculate the rate of diffusion and the viscosity.

Planned Impact

The proposed research has potential benefits for those outside the immediate scientific field of the geosciences. The techniques and equipment supported by research such as that proposed here are often of utility to other scientific groups, especially materials scientists. Over the past year I have been working with members of the Interface Analysis Centre at the University of Bristol, using the diamond anvil cell and Raman spectroscopy to understand the behaviour, under applied stress, of the thermal barrier coatings used on turbine blades. This work has been very successful and papers are being prepared for submission; the results have a potential for direct economic benefit to industry. This synergy between high-pressure experimental petrology and the materials sciences is still to be exploited to its fullest; I intend to continue to develop these links during the lifetime of the proposed project (see Pathways to Impact document for details).

In addition to fostering multidisciplinary research, the proposed project also has the potential to enthuse the general public, especially school age pupils. The nature of this research, with its 'extreme conditions' and high-tech experimental and computational approach to exploring a seemingly remote and unimaginable place can inspire awe in those with curious minds. The potential for interest in the results generated from this project from the popular scientific press and the wider media is significant and is something I intend to pursue alongside the usual pathways of publication in the peer-reviewed scientific literature. More specifically, this kind of research is ideally suited for use as a method of encouraging pupils in school to take an active interest in STEM (Science, Technology, Engineering and Mathematics) subjects and ultimately to study them. This is especially true of GCSE and A-level students who are actively deciding on the next step in their future educational path. I intend to visit local schools and educate students about not just the nature of the Earth and the findings of the proposed project but also to try and actively encourage students to study science, especially the geosciences at A-level and university (see the Pathways to Impact plan). I am currently involved in a project at UCL, entitled 'box-office blunders' which aims to educate pupils at local schools about the nature of the Earth through a 'myth-busting' exercise. The idea is to show disaster movies that have a geophysical aspect to the plot (e.g.: "The Core" and the recent "2012") and have the students draw up a list of scientific flaws. Afterwards, a workshop session will be held to explain how the Earth really works, how geoscientists go about studying it, and what academic research is really like.


10 25 50
Briggs R (2017) High-pressure melting behavior of tin up to 105 GPa in Physical Review B
Dobson D (2016) The phase diagram of NiSi under the conditions of small planetary interiors in Physics of the Earth and Planetary Interiors
Lord O (2014) The NiSi melting curve to 70GPa in Physics of the Earth and Planetary Interiors
Lord O (2014) The melting curve of Ni to 1 Mbar in Earth and Planetary Science Letters
Lord OT (2015) The equation of state of the Pmmn phase of NiSi. in Journal of applied crystallography
Description Since the beginning of this award, three major results have reached the publication stage:

1. The melting curve of Ni to 100 GPa.

In this study, the laser-heated diamond anvil cell (LH-DAC) was used to subject Ni metal to pressures of over 1 million times that of the atmosphere and reach temperatures of over 4000 degrees Kelvin (around 3700 Celsius) while determining whether the sample was in the liquid or solid state. This was achieved by performing X-ray diffraction at the European Synchrotron Research Facility in Grenoble, France., which can differentiate between the solid and liquid state. The melting curve that we determined from these measurements agrees well with existing studies in which melting was determined in quantum mechanics based computer simulations and 'shock' experiments in which samples are compressed and heated during hyper-velocity impacts. However, it strongly disagrees with earlier measurements in the LH-DAC which employed different methods of determining the temperature at which the sample had melted. This result is important, because similar discrepancies in the literature exist for a number of transition metals. The new data suggest this disagreement is an artifact of the earlier measurements, thus invalidating numerous physical models proposed to explain the discrepancy.

2. The melting curve of NiSi to 70 GPa

In this study, similar techniques were used to determine the melting curve of nickel silicide as well as to determine its crystal structure, which changes with both pressure and temperature. NiSi is one 'end-member' within the chemical system that can be used to describe the composition of Earth's core (and indeed the cores of other terrestrial planets). This study provides valuable data to those in the geoscience community who try to model the cores of terrestrial planets. Combining data on end-members like NiSi can allow models to be developed in which the density and sound velocity of any composition within the system can be determined. These values can then be compared with those determined from seismology - the only direct probe we have for the deep parts of the Earth, including the core - potentially allowing the composition of the core to be constrained. This in turn has implications for the structure and evolution of our planet.

3. Melting in the systems Enstatite-Magnesite and Magnesite-Calcite from 15 - 80 GPa

This result was selected as a 'notable paper' at American Mineralogist, the journal in which it was published. The paper was summarized in the following way on the journal's website: Tomson et al. find that low-degree partial melts of subducted carbonates may be widespread in the convective lower mantle. They hypothesize that during subduction, partial melting of carbonates is initiated near the upper/lower mantle boundary, and that this may explain so-called superdeep diamonds, which are thought to derive from this depth interval. Their results also indicate that the slopes of melting curves in this system (dT/dP) approach zero, or perhaps even become negative, at higher pressures, so that carbonate-rich partial melts may then be equally or more dense than bulk equilibrium solids. Such a characteristic would allow carbonate to be trapped in the lower mantle in the liquid state, providing the possibility of a carbon-rich reservoir in the deep mantle.

In addition, I have completed the ab initio simulations proposed for iron in the fcc structure and am now embarking on similar calculations for hcp iron. These results are currently being analysed and prepared for publication.
Exploitation Route The results listed above should be of interest to the high-pressure physics and geophysics communities; they should stimulate further work because the results innevitably lead to further questions. In the case of the melting of transition metals, this is likely to lead to some lively discussion of the efficacy of various techniques used in high pressure science, and should improve confidence in future results, both experimental and computational.
Sectors Other
Description It is too early in the life-cycle of this award for the findings to have been used by others; the publications listed have only just appeared (or are just about to appear) in print.
First Year Of Impact 2014
Sector Other
Description Studies of mantle melting with the Centre of Earth Evolution and Dynamics, University of Oslo, Norway. 
Organisation University of Oslo
Country Norway, Kingdom of 
Sector Academic/University 
PI Contribution A PhD student, Marzena Baron, has spent a significant portion of the last three years in the high-pressure labs at the School of Earth Sciences in Bristol using our microfabrication and laser-heated diamond anvil cell facilities to perform novel measurements of the melting phase relations of simplified mantle lithologies.
Collaborator Contribution CEED have provided the PhD student (Marzena Baron) as well as expertise, diamond anvils (the major laboratory consumable) and have organised the analytical side of the project (Focussed Ion Beam sample recovery and Transmission Electron Microscopy analysis at IPGP, Paris and the Bayerisches Geoinstitut, Bayreuth, Germany).
Impact This collaboration has yielded a novel dataset concerning mantle melting processes that is in the final stages of preparation and pending submission early in 2016. It has also led to the development of a novel technique, namely micro-fabrication fully encapsulated samples for the laser-heated diamond anvil cell, which will be of great utility to petrologists and those studying materials at high pressure in general.
Start Year 2013
Description Transition metal melting with the Institute of Mineralogy, physical materials and Cosmochemistry, Paris 
Organisation National Center for Scientific Research (Centre National de la Recherche Scientifique CNRS)
Department Institute of Mineralogy, physical materials and Cosmochemistry IMPMC, Paris
Country France, French Republic 
Sector Public 
PI Contribution I have been performing experiments to try to determine the high-pressure melting lines of various transition metals, which are highly controversial, due to dramatic differences in the predictions made by ab initio computer simulations and earlier experiments.
Collaborator Contribution My collaborators (Guillaume Morard and others) have provided equipment time; specifically, access to Focussed Ion Beam technology which has yielded new data, as well as expertise
Impact This collaboration is only in the early stages; it has not yielded any formal outcomes so far.
Start Year 2014