Science Articles

What's at the centre of the earth?

Thu, 7th Jan 2016

Jon Cartwright


New experiments that crush material between two diamonds to simulate the extraordinarily high temperatures and pressures found in the earth's interior are providing answers to the age-old questions of what our planet is made of, and where its ingredients came from.

'We are measuring things that were not measured before,' says geophysicist Prof. James Badro of the Paris Institute of Earth Physics, France.

Prof. Badro has been working with colleagues on the DECORE project, backed by the EU's European Research Council (ERC), to understand the composition of the earth's core. The core is thought to be mostly iron, yet analyses of seismic waves suggest that it is not dense enough to be iron on its own.

The best way to understand the precise composition is to replicate the conditions of the core in the lab. That means temperatures of thousands of degrees and pressures of about 100 gigapascals - over a million times the pressure of our atmosphere.

To mimic such conditions, Prof. Bardo and colleagues used a diamond anvil cell - a device that crushes a sample between two diamond tips - which they heated with a laser. They began with two samples in the cell: one made of silicates, like the earth's mantle, and one made of iron.


Once the samples were crushed and heated to core-like conditions, the researchers used a combination of two techniques - ion-beam microscopy and electron microscopy - to deduce the composition of the resultant alloy.

They found that some of the silicate's elements - vanadium, chromium, nickel and cobalt - seeped into the iron in large quantities, whereas others - silicon and oxygen - only went in a bit. The density of the resultant alloy was just right to explain the type of seismic waves received from the earth's actual core, suggesting that the composition of the alloy was the same as the core.

Data interpretation

In some sense, diamond anvil cells offer an easy route to replicating extreme conditions. But the data is not always easy to interpret.

Professor Dan Frost of the University of Bayreuth in Germany says that, in particular, the exact pressure that has been applied to the sample during the experiment is hard to pinpoint. This is important because it is the parameter that corresponds to depth beneath the earth's surface.

Although basic physics says that pressure is the force applied per unit area, the quantity is hard to calculate because some of the force supplied by the diamond tips always 'leaks' back into the retaining gasket.

As part of another ERC-funded project called DEEP, Prof. Frost and others have pioneered a method to record reliable pressures inside diamond anvil cells. They measured two different parameters - the sample volume and its compressibility - which can be entered into an equation to solve for pressure.

Making these measurements required apparatus that could simultaneously perform two analytical techniques. 'We had to put it together from bits - it was a custom job,' said Prof. Frost, who recently won a EUR 2.5 million Leibniz Prize from the German Research Foundation for his work.

As a result, Prof. Frost and colleagues managed to draw up a table of how the volume of a sample relates to temperature and pressure. In the future, then, all researchers need to do to find the accurate pressure is to measure the sample volume.


The work has borne fruit already. Having done studies of minerals at various pressures, corresponding to various depths, the researchers believe that the earth's lower mantle should have a similar chemical composition to the upper mantle.

That result throws the origin of the earth into a different light. Some scientists had thought that the earth was at least partly made from meteorites which most likely originate from the asteroid belt, and since these did not have the exact composition of the upper mantle, the assumption was that the missing constituents had sunk into the lower mantle.

If the two regions of mantle have the same composition, says Prof. Frost, that theory goes out the window - it could be that the earth was made from matter elsewhere in the early solar system.

'That's what we're really fascinated about, the formation of the earth,' he adds. 'Tracking down where all those building blocks come from.'


Subscribe Free

Related Content


Make a comment

Planet earth is millions and millions of years old. Why does it take so long for earth to cool down? The inner temperature is 5,500 C the outer 15 C. I would think there should be a great amount of heat from the inner earth to the surface. But it seems the heat is well isolated. Why? Hans, Sat, 9th Jan 2016

Three times as much water as in the earth's oceans was discovered very recently 400 miles deep in the earth's crust .. that seems to upset conventional ideas about earth's masses and makeup. Alohascope, Sat, 16th Jan 2016

There are several reasons:

One is that the Earth is a large body and so there's already a lot of embodied energy to lose, so cooling is relatively slow on even geological timescales.

Second, the Earth is its own nuclear reactor. The core is enriched for a number of radioactive isotopes like potassium-40 and thorium; the steady decay of these elements leads to radiogenic heating, supplying additional heat energy to the Earth's core. chris, Sat, 16th Jan 2016

I have to say that I have never been satisfied with the iron core model.
Gravitationally it has never made sense. If we follow the thinking that the Earth when it was first formed was basically a molten ball of materials that could all mix and differentiate, then the heaviest elements ending up at the centre make no sense at all.
The force applied by the effect of Gravity, unlike the pressure, reduces as you travel towards the centre. At the very centre of the Earth although the pressure is at a maximum, there should be zero Gravity. Gravity is after all calculated by the amount of mass within a radius. At radius 0 there can be no gravitational effect.

Now if we go back to the molten Earth beginning and assume a constant density (well mixed hot soup), the maximum gravitational force would be felt at the surface, reducing as you travel towards the centre. With the maximum pressure in the centre reducing as you travel towards the surface.
This scenario does not in any way say that the heaviest elements should end up in the centre. In fact they should end up somewhere between the upper core and lower mantle area. This puts the heavier elements where the unstable ones produce the most work in heating the planet. The pressures involved bellow this level coupled with the reducing gravity should make the core a very compacted representation of the Upper Mantle.
Not pure iron as we have always been told.
Just my two cents worth...
arthur.manousakis, Sat, 16th Jan 2016

One wild card variable is water. I did research growing gem quality crystals, many years back, using various methods, such as hydrothermal. Hydrothermal water, is high pressure and high temperature water above it critical point. This phase of water is very corrosion to almost all minerals.

Water in a confined space, such as experiment equipment, will dissolve its way in the direction of increasing temperature; even if the direction of higher heat is downward, since higher temperature will increase mineral solubility in water and represent the direction of increasing entropy; 2nd law. In experimental equipment the mineral rich hot water flows upward due to convection and deposits the excess mineral on the top surface. This renews the water. The result is a moving pocket of water eating downward, and depositing mineral at the cooler top side. The net affect is water in sealed pockets, can theoretically dissolve its way toward to the mantle, even though water is lighter. This is driven by chemical potentials not density.

Studies have shown that there is an ocean of water the size of the Arctic ocean, in the mantle below SE Asia. Once water reaches the mantle the temperature and pressure of the water will exceed the parameters of hydrothermal water. The water will change phase to become what is called superionic water. This is a nasty phase of water, that would explode like TNT if there is a pressure drop. Shifts in the crust, to create a pressure drop, good way to blast holes in the crust, or further detach the crust/mantle so the plates can slide. 

As the water diffuses even deeper toward the core; solubility, this water changes phase again, into ionic water, at about the conditions observed at the inner mantle and outer core. It is possible the layers of the earth reflect the phases of water. If we go even deeper, at the pressure and temperatures of the iron core, water become metallic.

I can see amalgams forming between the iron of the core and dissolved metallic water. This will resulting the iron will become oxidized to oxide, resulting in electron flow. The oceans are slightly alkaline with a slight negative charge; extra electrons. This could be mediated by water phases right to the oceans.

puppypower, Sun, 17th Jan 2016


I think this is right, but perhaps for a different reason...

Natural systems tend to approach the state of minimum energy. If you had an arrangement inside the Earth where:
the densest material (eg Thorium or Uranium metal) was half-way between the center and the surface of the Earth
and less dense material (eg iron and nickel) was at the center...
Then you could reduce the total energy by swapping a lump of lead for a same-sized lump of nickel-iron in the center of the Earth.
This suggests that over time, the densest material will tend to migrate towards the center (all other things being equal).

However, there are some other factors at play, such as the solubility of Uranium and Thorium in other minerals. We don't have rock samples from deep inside the Earth, but is thought that these radioactive elements are more soluble in silicate-type minerals (in the crust and mantle) than in the nickel-iron alloy thought to exist in Earth's core.

So we are no longer comparing the density of a lump of Uranium with the density of a lump of iron, but we are comparing the density of a Uranium-containing silicate mineral with the density of iron. Iron is more dense, and so tends to sink to the center, over time.

Perhaps the best way to estimate the distribution of radioactive minerals in the Earth is to look at the concentrations of radioactive elements in meteorites shattered from differentiated solar system bodies, comparing concentrations of radioactive elements in nickel/iron meteorites vs rocky meteorites. evan_au, Sun, 17th Jan 2016

Another thing to be taken into account here is the Earth's relatively large moon. Taking into account the size of the Moon and the leading theory that it was also a lot closer in the past with a correspondingly faster spinning Earth, puts the centre of gravity not that far under the surface in the beginning and even today the Gravitational centre of the system is like a mixmaster rotating through the layer some 1,700 Kms under the surface. That is also not conducive for the heaviest elements sinking to the core.
arthur.manousakis, Mon, 18th Jan 2016

I agree that ocean tides on Earth's surface would have been enormous in the distant past, because if you halve the Earth-Moon distance, tidal forces increase by a factor of 8.


However, I think that the impact on Earth's interior today might be a little overstated.

The Moon's mass is about 1/80th of the Earth's mass.

The tidal forces decrease as you move from the surface of the Earth towards the center (see URL above)

a 0.3 m tidal displacement for rock at the surface of the Earth represents a 0.1m tidal displacement at 1700km from the center of the Earth.

For some serious tidal stirring, have a look at Io. The Io-Jupiter distance is a bit larger than the Earth-Moon distance. But Jupiter's mass is about 30,000 times greater than Io, resulting in tides in the solid surface thought to reach 100m.  evan_au, Mon, 18th Jan 2016

Evan, I obviously placed too much emphasis on tidal effects with the mixmaster mention. It took attention away from the point I was trying to convey.
With the centre of Gravity that any differentiation is going to follow not at any time being synonymous with the centre of the Earth, I do not see any logic in the claim that the core is made of the heavier elements. I certainly can't see any way to form an iron core. The mechanism to transport it there has never existed.  There may well be some iron there just like there is some up here, but I believe the core to be a huge mixture of matter including water and silicates under extreme pressure. An alloy that we probably have a difficult time understanding.
Anyway the mixmaster is on very slow setting.. :-} arthur.manousakis, Mon, 18th Jan 2016

The density difference argument, which has the heavier elements sink to the core of the earth, assumes all the materials are inert and will not chemically interact with each other. If we assume chemical interaction, this changes the equation. For example, if we added rock salt to a glass of water, the rock salt will sink, due to its higher density. But in the longer term, the water, although much lighter, will diffuse downward, into the salt, and will begin to solubilize the salt, beginning on the outer core of the rock salt. The salt ions will then diffuse upward into the water, driven by entropy, to form a uniform solution. The oceans have a little bit of everything.

The main element found on the earth is oxygen, with elemental oxygen very reactive with nearly all the elements. If you had a plasma of atoms so there are no chemicals, just atoms, and then cool this, the oxygen ends up with extra electrons. Oxygen will strip electrons off metal stops, for example, to form metal oxides. Once you have hydrothermal water, oxides that may not be soluble in the surface water, become very soluble. Silicates are silicon metal plus oxygen; silicon oxide. This is quartz with is stable in ocean water, but is very soluble in hydrothermal water.

If we had water diffusing and solubilizing its way toward the iron core of the earth, driven by 2nd law solubility, the high temperature and pressure of the earth's interior would induce water phases. Each phase is aggressive toward minerals in its own way. This would make the earth interior form layers influenced by water phases. 

The metallic iron core, is electron rich, with respect to oxygen, with chemical potential favoring the formation iron oxide (hydroxide). The electrons released will follow the chemical potential to the surface, causing the oceans to become slightly negative.

The water on the surface is evaporated by the sun. This evaporation and the formation of clouds results in positive charge forming in the atmosphere. This is an external wild card, driven by the sun, that helps react and solubilize the core, with the release of heat. The sun sets a potential in the atmospheric water, than needs the negative charge of electrons to cancel the induced positive charge. Lightning has electrons flowing from the earth (ground) to the cloud.

Here is an interesting relevant observation. Seismic waves travel faster north-south than east-west, about four-seconds faster pole-to-pole than through the equator. Although the earth is wider at the equator than it is pole to pole, the four-second difference exceeds the distance difference, if we assume uniform materials. The higher net speed north-south means the core of the earth has to be denser north-south compare to east-west.

This makes sense since the equator is heated by the sun, to form positive charge in the clouds. The result in more electrons will flow from the equator to the atmosphere, than from the poles. This imbalance will causes core solid crystals to become different  north-south. East and west will have more oxygen; iron oxide blend. 

Mars is the red planet, due to iron oxide in its surface. Marks lacks a magnetic field. This suggests its water penetrated to the interior, thereby removing the positive charge in its atmosphere, that helps to align the field. The field vector cancels. puppypower, Mon, 18th Jan 2016

See the whole discussion | Make a comment

Not working please enable javascript
Powered by UKfast
Genetics Society