If you have an ACS member number, please enter it here so we can link this account to your membership. (optional)

ACS values your privacy. By submitting your information, you are gaining access to C&EN and subscribing to our weekly newsletter. We use the information you provide to make your reading experience better, and we will never sell your data to third party members.



Chemists journey to the center of the Earth

The chemistry of the core drives our planet’s magnetic field and holds clues about Earth’s history. Geochemists are going to extremes to understand it

by Katherine Bourzac
October 2, 2022 | A version of this story appeared in Volume 100, Issue 35
Brightly colored illustration showing Earth cut in half to reveal distinct interior layers.

Credit: Johan Swanepoel/Science Source | The nature of Earth's interior has long fascinated scientists.


In brief

The chemistry of Earth’s core tells the story of how the planet evolved and what sustains its magnetic field, without which Earth would be lifeless. But chemists don’t agree on the core’s elemental recipe. Geochemists journey to the center of the Earth in the lab, conducting experiments at extreme temperatures and pressures to unlock the planet’s chemistry. This work will help us understand our planet and guide astronomers looking for chemical clues of life elsewhere in the universe.

Bombastic German geochemist Otto Lidenbrock, his reluctant nephew, and a laconic guide climbed into a dormant volcano in Iceland and traveled deep into the Earth. On their way into the planet’s core, they nearly died of dehydration before encountering underground rivers, oceans, and dinosaurs. Finally, they were ejected from a volcano in Sicily.

That’s the story of Jules Verne’s 1864 science fiction classic, Journey to the Center of the Earth. The actual center of the Earth doesn’t feature oceans or ichthyosauruses, but it does hold deep chemical mysteries.

Most notably, geochemists still aren’t sure what Earth’s core is made of. It definitely contains iron, but a range of geophysical observations suggest that other elements are lurking down there. Scientists also don’t know how iron melts and solidifies or how conductive it is at the extreme pressures and temperatures at the center of our planet.

These open questions leave a lot to learn about the core, which makes up about a third of Earth’s mass. “It’s huge, but we don’t know what its composition is. And it has huge implications,” says Kei Hirose, a geoscientist at the University of Tokyo.

Answering these questions about Earth’s core will help us undertand how the planet formed and what sustains its magnetic field. We owe a lot to that magnetic field. Without it, Earth would be bombarded with ionizing radiation from the sun that would have made it impossible for the atmosphere and possibly the oceans to form. Also, without the shielding provided by the magnetic field, life probably could not have begun on Earth.

Earth's interior
While Earth was forming over 4.5 billion years ago, its surface was made up of a hot magma ocean. As objects in the early solar system hit the young planet, they melted, and their iron sank into the planet's center, forming the planet's core. Today's Earth consists of a solid-iron inner core, a liquid-iron outer core, a silicate mantle, and the planet's crust, which contains the continents and oceans.
Illustrations showing the interior structure of the Earth over 4.5 billion years ago and in the present day. In the early Earth illustration, elements include: an impactor about to hit the young planet, its upper layer made up of a magma ocean with regions of iron precipitating from silicate, an underlying mantle where the precipitating iron pieces have become larger, and an iron core. In the illustration of Earth’s current day interior structure, the planet’s crust is shown on the outside, with a mantle underneath that, and finally an outer and inner core.
Credit: Nature/Shutterstock/C&EN

Lidenbrock had a 16th-century alchemical script to guide him to Earth’s interior. Geochemists looking to solve the mysteries of the core make their journeys in the lab with the help of high-pressure chambers and lasers. Using these setups, researchers can replicate the core’s temperatures and pressures. Today’s inner core is as hot as the surface of the sun, and its pressure is a few million times what we experience on the surface. Much of the chemistry that scientists have learned from experiments at standard temperature and pressure does not apply to these core-like conditions. But by exploring this extreme chemistry, geochemists may find answers to some big questions about our planet.

Deepening mystery

The current picture of Earth’s interior formed by the middle of the 20th century.

First there’s the crust, with its continents and oceans. The continental crust’s thickness averages about 35 km; oceanic crust is about 6 km deep. No human has penetrated deeper than this layer. The deepest hole ever dug, the Kola Superdeep Borehole in northern Russia near the Norwegian border, reaches a mere 12 km. If Earth were an apple, the borehole would not even fully pierce through its skin.

The core is “huge, but we don’t know what its composition is. And it has huge implications.”
Kei Hirose, geoscientist, University of Tokyo

Below the crust lies a hot mantle of solid silicates that extends to a depth of about 2,891 km below the surface. Deeper still is Earth’s iron core. A layer of liquid iron, similar in viscosity to water, extends to a depth of 5,150 km. The planet’s deepest depths hold a giant frozen ball of iron that is over 7.5 billion km3 in volume. The very center of the planet is about 6,378 km from its surface.

Scientists have deduced this basic structure from a few data points. For example, geoscientists have studied how two kinds of seismic waves travel from earthquake epicenters through the planet’s interior. P-waves are fast-moving pressure waves and can travel through both solids and liquids. S waves move slowly and only through solids. By watching where these different seismic waves are detected on Earth’s surface, scientists deduced that the planet has a solid mantle, a liquid outer core, and a solid inner core.

The second important measurement is the planet’s density, which is derived by dividing its mass by its volume. The planet’s volume is relatively straightforward to measure because we know the Earth’s circumference. Without a cosmic scale, however, the mass is trickier to calculate. Scientists have used measurements of the planet’s gravity and Newton’s gravitational laws to get the mass. These values suggest that Earth’s density is on average 5.25 gm/cm3. Scientists have concluded that Earth must have something fairly heavy inside it to lead to such a high density. An iron core fits that description.

Earth's magnetic dynamo
A stirring motion (yellow coils) in the electrically conductive liquid iron of the planet’s outer core acts as a dynamo that generates Earth’s magnetic field. The magnetic field's direction is indicated with black arrows.
Illustration showing the planet’s magnetic dynamo. Illustration shows a red inner core surrounded by an orange outer core. Yellow coils in the outer core indicate stirring motion in the liquid layer. Black lines running from the top of the image to the bottom show the direction of the planet’s magnetic field.
Credit: Azcolvin429/Wikimedia Commons/C&EN

An iron core also provides an explanation for Earth’s magnetic field. Bruce Buffett, a planetary scientist at the University of California, Berkeley, compares Earth’s outer core to a pot of water on the stove that’s been brought to a boil and then taken off the flame. As the water on top cools, it sinks to the bottom, stirring the liquid by thermal convection. Earth was very hot when it first formed around 4.5 billion years ago, and as the material deep inside the planet continues to cool, the liquid outer core stirs. This stirring causes the liquid iron to move, and because iron is electrically conductive, this motion generates a magnetic field.

“This is the dynamo process,” Buffett says. “We need fluid motion to continually regenerate the magnetic field.” Scientists know that the planet’s magnetic field has existed for at least 3.5 billion years, according to observations of exceedingly rare and ancient single zircon crystals found in rock deposits. Magnetite inclusions found in these crystals have had their magnetic orientations set for billions of years—and some researchers believe they show evidence of a strong magnetic field on Earth dating back 4 billion years (Proc. Natl. Acad. Sci. U.S.A. 2020, DOI: 10.1073/pnas.1916553117).

But this model of Earth isn’t complete. In the past 70 years, geochemists have found two major problems with the story of a pure iron core and a magnetic dynamo driven by thermal convection.

Francis Birch raised the first major problem. Birch worked on the Manhattan Project and at Harvard University and was one of the founders of modern geophysics. In the 1950s, he calculated that the outer core is about 10% less dense than it would be if it were made of pure iron (J. Geophys. Res. 1952, DOI: 10.1029/JZ057i002p00227). The core must include some lighter elements. Contemporary scientists have revised this estimate from 10% to about 8%, the University of Tokyo’s Hirose says, but Birch’s challenge still stands.

A 2012 finding challenged scientists’ certainty about what powers the dynamo. A team led by Dario Alfè at University College London used density functional theory to estimate iron’s thermal and electrical conductivity at extreme pressure and temperature. The researchers’ models suggest that iron is two to three times as thermally and electrically conductive under core-like conditions as had been assumed (Nature 2012, DOI: 10.1038/nature11031).

Greater thermal conductivity for iron challenges the notion that the planet’s magnetic field originates solely from thermal convection of conductive iron in the outer core. At those higher levels of conductivity, the core would have cooled much faster than scientists previously thought, meaning either the planet started out much hotter than scientists can explain or some other force is stirring the core to sustain the dynamo. But what other force? Hirose calls this “the new core paradox.”

Core recipes

A more detailed understanding of the core’s elemental composition is key to solving these mysteries. Geochemists suspect that a large amount of nickel is in the core. But besides iron, only a few other elements are likely present at levels that matter. “Some people prefer oxygen. These people like silicon. He likes sulfur,” says Hirose, who likes hydrogen. “We have so many different stories, so many papers. They are so inconsistent with each other.”

The unknown light elements should be abundant in the solar system and capable of forming alloys with iron, Hirose says. Magnesium is abundant, but “it wants to be MgO” instead of hanging out with iron, Hirose says. Off the list. “Nitrogen wants to be N2,” so it’s probably not a big player in the planet’s interior, he says. According to Hirose, there are five candidates for light elements in Earth’s core: hydrogen, carbon, silicon, sulfur, and oxygen.

Closeup of two diamonds arranged tip to tip.
Credit: S.-H. Dan Shim/Arizona State University
A peek inside a diamond anvil cell shows the tips of two of these super hard gems. Scientists can place a sample between the tips and then bring them together to exert extremely high pressures.

This narrowing of possible elements mirrors how geochemists look at the periodic table. They tend to divide elements into categories linked to Earth’s history and the part of the planet in which the elements often appear. Earth, like other bodies in today’s solar system, started out as a hot ball that expanded as it was bombarded with planetoids and other objects. When these objects hit the young Earth, their material landed in a deep, hot magma ocean. Denser iron—and iron-loving, or siderophile, elements like nickel and other metals—sank to the core. Lithophilic elements such as silicon that like to form rocks mostly ended up in what became the mantle. Other more buoyant elements, like nitrogen, ended up in the oceans and atmosphere.

So how do geochemists evaluate possible core recipes in the lab? They create little simulated cores. Researchers mix likely ingredients, bring them to extreme pressures and temperatures, and identify how much of a given light element is incorporated into the resulting simulated iron core and whether that core’s properties match those of the actual one.

The workhorse of high-pressure science is a simple mechanical device called a diamond anvil cell. “I can carry it in my hand,” says Bethany Chidester, a research fellow at Los Alamos National Laboratory. The diamond anvil cell consists of two of the durable gems mounted to metal plates, tip to tip. Scientists place a sample to study between these tips and apply pressure by turning a set of screws with a handheld screwdriver. Because pressure is equal to force divided by area, a relatively small force applied over a minuscule area, like the tip of a diamond, can yield pressures as great as those at the center of the Earth: about 360 GPa and higher.

Once geochemists put a sample under these pressures, they heat the sample with lasers or resistive heaters. Then they can gather data, often using X-ray spectrometers at synchrotrons to view the material’s crystal structure. Scientists can also recover a sample to perform isotopic analysis, study it under a microscope, or measure the velocity of sound waves in the material to compare data with the core’s known properties derived from seismological data.

Hirose has been using these techniques to test alloys of iron with his favorite potential light element, hydrogen. He favors the FeH alloy because the way it carries sound waves matches seismological data. Under conditions at Earth’s surface, hydrogen isn’t a big fan of iron, but at core-like conditions, it’s “a very iron-loving element,” Hirose says. Hirose’s group is studying the melting behavior of FeH.

Anat Shahar, a planetary scientist at the Carnegie Institution for Science, takes a different approach to studying possible core recipes. She examines how isotopes of iron and other elements behave at high pressure to understand how elements separated during the formation of the planet.

Microscopy image of a small metallic ball surrounded by a shiny, rocky-looking outer layer.
Credit: Anat Shahar
After being placed under pressures and temperatures similar to those in the interior of Earth, an experimental sample has separated into an Earth-like structure, about 5 mm in diameter, with a metal core and a silicate shell.

“The core composition is really important because it affects the evolution of the planet so much,” she says. Her studies have helped establish that pressure plays a previously unappreciated role in driving isotopic fractionation during planetary evolution.

Shahar makes 5 mm wide miniature Earths. She mixes silicate and iron powders with light-element candidates inside a piston, applies pressure, and then checks which elements and isotopes of those elements fractionate into a core-like center and which go into a mantle-like outer layer in the lab-made planetoid.


When she adds a lot of hydrogen, heavier iron isotopes favor the core-like layer, and lighter ones favor the mantle, matching what is observed on Earth. The same is true when she adds silicon. Because of these data, Shahar’s preferred light-element spice mixture is Si, H, and O.

If light elements are hanging out in the core, that would do more than solve the density problem that Birch first identified. They could also explain the persistence of the dynamo. As the iron in the core continually freezes, “some of these impurities—silicon, oxygen, or hydrogen—are probably excluded,” UC Berkeley’s Buffett says. This process would resemble what happens when sea ice freezes and expels salt into the water below. Likewise, when Earth’s core freezes, it could be ejecting these light elements. The movement of these elements could provide a chemical convection that helps stirs the outer core and maintains the dynamo.

Deep chemistry

The chemistry of Earth’s core could be more complex still. Scientists like to assume that there is no exchange of material between mantle and core in present-day Earth. The mantle’s solid silicate is immiscible with liquid iron, so the two probably don’t chemically react with each other today, Hirose says. But that doesn’t mean it is or has always been a hard boundary with no exchange—particularly if the lower mantle was liquid in the past.

Buffett is open to the possibility of reactions between mantle and core across the planet’s time span. Precipitation of MgO from the core into the mantle could have helped power core dynamo motion in the planet’s early days, he says. Even today, Buffett says, “there could be a reaction at the boundary, like rust on the surface of a piece of metal.” Such a reaction might be self-limiting, building up a rind at the bottom of the mantle that blocks off further interaction.

He also notes that the mantle’s oxygen-rich rock could transport that element into the core along grain boundaries. “Just because the diffusion is low, that does not mean it’s zero,” Buffett says. “The geochemists don’t like this story,” he says. But it should be testable. Scientists are looking for signals of this exchange in isotopes in lava from the planet’s deepest volcanoes.

Other elements—even heavy ones—could play an important role in the core and the dynamo, Chidester says. She discovered this by accident in graduate school. “I had this whole plan for my PhD,” she says. At the University of Chicago, she set out to prove that there could not be uranium in the core.

But she found that uranium, which had been considered lithophilic and not at all an iron lover, joined with iron in her high-pressure experiments. She then worked with computational modelers to perform what she calls a “heroic feat of math.” They compiled data from hundreds of experiments on various potential core compositions at a range of temperatures and pressures and then used a computer model to grow a planet in 100 stages from those data. At each stage, they ran calculations and identified where all the elements were.

In this computational planet, the core could hold a large amount of uranium. The presence of uranium suggested a substantial source of core-stirring energy to sustain the dynamo: heat produced by radioactive decay (Geochem., Geophys., Geosyst. 2022, DOI: 10.1029/2021GC009986).

So the scientific jury is still out on the composition of Earth’s core. “Each element has very strong grounds for why it should be in the core,” Hirose says. But if you sum the amount of each element that different scientists have proposed the core holds—and he has—“it’s too much in total.” Scientists still have work to do.

Under pressure

Coming to a consensus on core chemistry might be easier if scientists better understood the basic properties of iron under high pressures and temperatures. It was not until this year, for instance, that researchers published the first experimental measurements of the melting point of iron at Earth-core pressures and higher.

Few facilities can reach the necessary pressures in a controlled fashion while making measurements. It took scientists at Lawrence Livermore National Laboratory (LLNL) nearly 20 years to develop the techniques for this recent iron study, which required the use of one of the world’s most precise, powerful laser systems, the National Ignition Facility. The resulting experiment allowed researchers to study iron’s properties at pressures found in Earth’s core and at higher ones possibly found at the center of large exoplanets called super-Earths.

I don’t know if we’ll come to a consensus. But I really hope so because there are so many other things we should study.
Anat Shahar, planetary scientist, Carnegie Institution for Science

Richard Kraus, a shock physicist at LLNL, says the experiment was designed to determine whether super-Earths could have magnetic fields generated by Earth-like dynamos. The carefully crafted experiment emulates the conditions that a lump of iron would experience while falling to the core of a planet.

A set of 16 lasers fires at an iron sample placed in a carefully designed holder called a target. An ablator on top of the iron is the rocket fuel for these experiments, says Jon Eggert, chief scientist at LLNL’s high-energy-density science group. Made of beryllium, this ablator absorbs the light from the high-energy laser beams and sends an intense shock wave through the iron. During the experiments, the researchers first subjected an iron sample to a high-pressure, high-temperature state. Then they increased the pressure to simulate what might happen as iron descends to a planetary core. Throughout this process, they used X-ray diffraction to check whether the iron was solid or liquid.

Kraus and Eggert’s team had only a few chances to get the experiment right. The National Ignition Facility does about 400 experiments, or “shots,” per year. It takes 6 months to make the target. “We take a shot, analyze it, and have many meetings and discussions,” Eggert says. “A huge amount of planning goes into these shots.” Of the 12 shots used for the iron experiment, which was carried out over 3 years, 7 worked well enough to yield publishable data (Science 2022, DOI: 10.1126/science.abm1472).

A metal stick holding a strip of material made of white and metallic materials and a microscope image showing a gray metallic circle in the middle of a rectangle of metal.
Credit: Richard Kraus/LLNL
This experimental target was used to study iron's melting behavior at extreme pressures at the National Ignition Facility. The trapezoid in the center of the white bar at left is the roughly 3 mm target, shown close-up at right.

Before this work, the melting curve of iron had been measured up to only 290 GPa. The LLNL study extends the data to Earth-core pressures, 360 GPa, and beyond, up to 1,000 GPa.

Iron’s melting properties at these extreme pressures led the researchers to conclude that super-Earths could have a similar core structure to our own, with a solid inner core and a molten outer one sustaining a dynamo. But, like with Earth’s core, scientists need to know more about how other elements in the super-Earth cores could change that picture. The LLNL team wants to do more high-pressure melting-point experiments on iron and its alloys to explore this question.

And melting curves aren’t the only physical properties of iron that researchers are studying. Iron’s thermal conductivity at high pressures is still up for debate. A few years after the theoretical prediction of higher-than-expected thermal conductivities for iron threw off accepted explanations for the magnetic dynamo, researchers experimentally measured lower-than-predicted conductivities at core pressures (Nature 2016, DOI: 10.1038/nature18009). But conductivity is very difficult to measure directly in high-pressure experiments, Buffett says, and the matter is not yet settled.

In general, the field is desperate for more data, Eggert says. He laments that today’s ultra-high-energy lasers and the best diagnostic instruments aren’t always at the same facilities. “We can get to extreme conditions, but we can’t diagnose them as well as we’d like,” he says. However, the power of X-ray laser systems is increasing, and better combinations of high energy and precise diagnostics are on the horizon.

Theoretical modeling can play a critical role in guiding experimental directions and design, but predicted material properties must be verified in the lab. Scientists can’t just rely on extrapolations based on experimental results at standard temperature and pressure, because strange things happen in planetary cores. “This is outside our realm of intuition,” Kraus says.

It’s been 70 years since Birch threw into question the composition of Earth’s core. Hirose hopes that with more data, agreement on the composition will emerge in the next 10 years.

“I don’t know if we’ll come to a consensus,” the Carnegie Institution for Science’s Shahar says. “But I really hope so because there are so many other things we should study.” Geochemists could, for example, use what they know about the Earth to help astronomers look for chemical signs that might be characteristic of exoplanets with strong magnetic fields. These planets are some of the best places to look for signs of life elsewhere in the universe. Geochemists are eager to tackle such projects more fully, but first they need to get on more solid ground when it comes to Earth’s core chemistry.


This article has been sent to the following recipient:

Chemistry matters. Join us to get the news you need.