A team of researchers has produced the strongest experimental evidence yet that Earth’s iron core holds enormous quantities of hydrogen, potentially equivalent to 9 to 45 oceans’ worth of water. The findings, published in Nature Communications, are based on high-pressure laboratory experiments that simulate conditions deep inside the planet. If confirmed by independent methods, the results could reshape scientific models of how Earth formed, where its water came from, and what drives its magnetic field.
Squeezing Metal at 5,100 Kelvin
The central challenge in studying Earth’s core is that no drill or probe can reach it. The outer core begins roughly 1,800 miles below the surface, where pressures exceed millions of atmospheres and temperatures rival the surface of the sun. To replicate those extremes in a lab, the research team used laser-heated diamond anvils to compress tiny samples of metal and silicate, then measured how hydrogen distributed itself between the two materials. They paired this with atom probe tomography, or APT, a technique that maps individual atoms in three dimensions, to directly count hydrogen atoms trapped inside the metallic phase after each run.
The experiments reached pressures of roughly 30 to 60 gigapascals and temperatures around 5,100 Kelvin. Under those conditions, hydrogen showed a strong preference for the metal side of the sample, behaving as what geochemists call a siderophile element. The team estimated that Earth’s core contains between 0.07 and 0.36 weight percent hydrogen. Translated into a quantity readers can picture, that range corresponds to roughly 9 to 45 oceans’ worth of hydrogen locked inside the deepest part of the planet. Because the experiments directly visualized hydrogen at the atomic scale instead of inferring it from bulk properties alone, they significantly tighten previous constraints on how much of this light element the core can realistically hold.
Why Hydrogen Sinks Toward Iron
The idea that hydrogen might concentrate in Earth’s core is not new, but earlier work relied on smaller datasets or computational models rather than direct atom-scale measurements. A 1997 experiment published in Science first demonstrated that hydrogen dissolves readily into molten iron at high pressure. That study argued that if the early Earth’s magma ocean contained significant water, much of that H2O could have reacted with sinking iron droplets to form iron hydride, dragging hydrogen downward during core formation. Follow-up work using multi-anvil presses and synchrotron X-ray beams, summarized in the original Science report, showed that this process could operate efficiently over the entire depth of a global magma ocean, making hydrogen sequestration into the core a planet-scale phenomenon rather than a localized curiosity.
A separate set of experiments published in an earlier Nature Communications paper measured the metal-silicate partition coefficient of hydrogen at 30 to 60 GPa and 3,100 to 4,600 K, finding that hydrogen is strongly siderophile with a partition coefficient near 30. Those results implied that 0.3 to 0.6 weight percent hydrogen could have been incorporated into the core under plausible accretion scenarios. Although the new atom probe measurements slightly revise that range downward, they are broadly consistent with the earlier partitioning experiments. Together, these studies indicate that during the first tens of millions of years of Earth’s history, when metal and silicate were vigorously mixing and separating, hydrogen had a clear chemical pathway into the growing core.
Simulations Back Up the Lab Work
Computational work has reinforced the experimental picture by exploring conditions that are difficult to reach in the laboratory. Using ab initio molecular dynamics and thermodynamic integration, one group modeled how water divides between iron and silicate melts across an even broader pressure range of roughly 20 to 135 GPa and 2,800 to 5,000 K. Their simulations confirmed that water partitions strongly into liquid iron under core-formation conditions, especially at higher pressures more representative of the deep magma ocean. These calculations also suggested that once hydrogen enters the metallic phase, it remains stable there over geologic timescales, rather than diffusing back into the mantle.
Additional modeling of volatile behavior during planet formation, accessible through the Nature sign-in portal, explores how different starting compositions of planetary building blocks affect the final hydrogen inventory of the core. These studies vary the relative contributions of dry inner-solar-system material and wetter, more distant planetesimals. Across a wide range of plausible mixtures, they still find that the core acts as the dominant sink for hydrogen as long as metal and silicate equilibrate at high pressures. The emerging consensus is that, unless early Earth was improbably dry, a substantial fraction of its primordial hydrogen ended up locked away in the metallic interior.
Carbon Complicates the Picture
A confident estimate of core hydrogen depends on knowing what other light elements are present, and that is where the story gets harder to pin down. A recent preprint circulated through the arXiv server highlights that hydrogen and carbon do not partition independently. When both elements are present in a high-pressure melt, their interaction changes the effective partition coefficients for each. In practical terms, the amount of carbon already dissolved in the core could either raise or lower the amount of hydrogen the metal can absorb, shifting the inferred hydrogen budget by a meaningful margin. This coupling means that resolving the core’s hydrogen content will require parallel progress on constraining its carbon inventory.
The authors of the new Nature Communications study acknowledge several other sources of uncertainty. Residual hydrogen trapped in experimental samples can inflate measured concentrations, a systematic bias that is difficult to eliminate entirely. The silicon content of the core also remains debated; different assumptions about how much silicon is present change the calculated hydrogen capacity because silicon competes for space in the iron lattice. And the models assume a specific amount of hydrogen was available during accretion, an assumption that depends on how wet or dry the building blocks of Earth actually were. None of these caveats invalidate the findings, but they widen the error bars enough that the difference between 9 oceans and 45 oceans is real, not just a rounding issue, and future work will need to narrow that range.
What a Hydrogen-Rich Core Means for Earth
If the upper end of the estimate holds, most of the hydrogen that arrived with Earth’s raw materials ended up not on the surface as oceans but buried thousands of miles below. That reframes a long-standing puzzle. Scientists have struggled to explain why Earth retains as much surface water as it does, given the violence of giant impacts during late accretion that could have stripped volatile elements into space. A core that acts as a massive hydrogen reservoir suggests the planet’s total volatile budget was far larger than surface inventories alone imply. In this view, the oceans we see today represent only the thin, mobile portion of a much deeper hydrogen system that extends from the crust down into the metallic core.
The presence of hydrogen also matters for understanding Earth’s magnetic field. The geodynamo depends on convection currents in the liquid outer core, and the density and viscosity of that liquid are sensitive to its composition. Adding a light element like hydrogen lowers the density of iron alloy, which in turn affects how heat and buoyancy drive motion in the outer core. A hydrogen-rich composition could help explain why the core remains partially liquid and convecting after more than four billion years, sustaining a magnetic shield that protects the atmosphere from the solar wind. As geophysicist Li Li of Brookhaven National Laboratory has emphasized, geological processes visible at the surface, from earthquakes to volcanic eruptions, are expressions of what happens deep inside the planet, and revising the core’s composition inevitably feeds back into how scientists interpret surface observations.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
