Mercury, the smallest and innermost of the terrestrial planets, has long puzzled planetary scientists. Its surface is dark, its core is anomalously large and light, and its chemistry is unlike anything else in the inner solar system. Now, as the joint European and Japanese BepiColombo mission prepares for the final leg of its cruise toward the planet, a series of laboratory studies from the University of Liège and KU Leuven offers a coherent explanation for two of Mercury’s most enigmatic features at once: the graphite-rich layer that once capped the young planet, and the composition of the vast metallic core beneath it.
The research, led by geologists Bernard Charlier of the University of Liège and Olivier Namur of KU Leuven, rests on a specialised discipline known as experimental petrology. Because no spacecraft has ever returned samples from Mercury and no meteorite has ever been convincingly linked to the planet, scientists cannot simply analyse rock in the way they do for the Moon or Mars. Instead, they recreate the planet’s interior in miniature. Using high-pressure, high-temperature apparatus, the teams subjected synthetic mixtures approximating Mercury’s interior to temperatures between roughly 1,250 and 2,170 degrees Celsius, at pressures equivalent to those found deep within a rocky planet, and watched how carbon behaved as molten metal separated from molten silicate.
The logic of the approach follows from how all the terrestrial planets formed. Mercury, Venus, Earth and Mars each began as accumulating material from the disc of gas and dust surrounding the young Sun. The heat released during accretion, together with contributions from radioactive decay and giant impacts, was sufficient to melt the growing planets, producing global oceans of magma. This molten stage was decisive: as a magma ocean crystallised, it sorted elements between a dense metallic core sinking to the centre and a silicate mantle left behind above. The mantle later partially melted to generate the magmas that built the crust, and the gases exsolved during this early evolution may have formed the planet’s first atmosphere. Whatever the magma ocean stage decided, the planet inherited for the next four and a half billion years.
The key variable governing carbon’s fate turned out to be oxygen. Geochemists quantify this with a parameter called oxygen fugacity, an effective measure of how oxidising or reducing the environment is. Under relatively oxidising conditions, the experiments showed, carbon is strongly siderophile, meaning it prefers dissolving in metallic iron and follows the metal down into the core. Earth, which formed under comparatively oxidising conditions, exemplifies this behaviour. Mercury, however, is the most reduced of the terrestrial planets, its rocks remarkably poor in oxidised iron. Under those highly reducing conditions, the experiments revealed that carbon becomes far less attracted to metal and instead stays dissolved in the silicate magma.
That single shift in chemical affinity has dramatic consequences. As the reduced magma ocean cooled and crystallised, the carbon it hosted could no longer remain dissolved and precipitated as graphite, the soft, low-density crystalline form of pure carbon. Because graphite is too light to sink through the molten silicate, it floated. Accumulating at the top of the magma ocean, it formed a primitive crust of graphite, a crust made not of silicate rock like every other planetary crust in the inner solar system, but of carbon. The team’s model indicates this primordial graphite crust was on the order of 40 to 120 metres thick, a figure that matches the carbon-rich surface layer of roughly 1 to 3 per cent by mass inferred from observations by NASA’s MESSENGER spacecraft, which orbited Mercury from 2011.
The primordial graphite crust did not survive intact. Meteorite bombardment and the vigorous volcanism that built Mercury’s younger volcanic plains disrupted and redistributed the carbon-rich layer, mixing it into the surface deposits we see today. This history helps explain why MESSENGER detected abundant carbon darkening Mercury’s terrain, a property that had previously been difficult to reconcile with any simple crustal model. The new experiments provide the missing physical mechanism: a flotation crust of graphite, generated by the planet’s own extreme chemistry rather than by any exotic delivery of carbon from outside.
The same experiments resolved a second, long-standing puzzle about Mercury’s interior. Geodetic data from MESSENGER show that the planet’s core is extraordinarily large, accounting for roughly 70 per cent of Mercury’s total mass, and yet it is less dense than pure iron would be. Something light must be dissolved in the metal. Carbon had been proposed as a candidate light element, and the idea was appealing: if carbon stayed in the core, it would simultaneously explain the density deficit and the surface graphite. The new measurements close that door. Under precisely the reducing conditions required to float a graphite crust, the core remains strikingly poor in carbon, containing less than 5,000 micrograms per gram, or less than 0.5 per cent. That concentration is far too low to account for the observed density shortfall.
The culprit, the researchers conclude, is primarily silicon, with sulphur playing a secondary role. These elements dissolve in metallic iron under reducing conditions and are light enough to depress the core’s density to the value MESSENGER measured. They carry an additional consequence with far-reaching implications: both silicon and sulphur substantially lower the melting point of iron alloy. That depression helps explain how Mercury’s core has remained at least partially liquid for 4.5 billion years, despite the planet’s small size and the steady loss of its primordial heat. A liquid, convecting layer in the core is the dynamo that sustains Mercury’s magnetic field, one of the surprising discoveries of the Mariner 10 and MESSENGER eras. The same light elements that lighten the core, in other words, may also be what keeps the planet’s generator running.
By tying Mercury’s extreme chemical reduction to both its graphite crust and its core composition, the work provides a unified framework that extends well beyond one planet. It speaks to the nature of the original, more massive proto-Mercury that some models propose existed before a giant impact stripped away much of its mantle. It informs the study of hypothetical super-Mercuries, dense rocky exoplanets detected around other stars whose interiors may follow similar rules. It bears on carbon-rich exoplanets generally, where reduced conditions might likewise produce graphite flotation crusts rather than carbon sequestered in cores. And it even reaches back to the early Earth, which accreted largely from reduced materials, meaning the partitioning behaviour measured in these experiments constrains how our own planet distributed its carbon between core, mantle, atmosphere and the biosphere that eventually depended on it.
The timing of the work is no accident. BepiColombo, a joint mission of the European Space Agency and the Japan Aerospace Exploration Agency, is entering the final phase of its long journey and will place two spacecraft in orbit around Mercury, carrying instruments far more sensitive than those of MESSENGER. As Charlier notes, future observations could confirm and quantify graphite as a major phase on Mercury’s surface, turning a laboratory inference into a measured planetary fact. If the orbiters detect the spectral signature and thickness of the carbon-rich layer the experiments predict, scientists will have a rare example of a planet whose earliest crust, core chemistry and magnetic history can all be traced back to a single parameter: how little oxygen was present when its magma ocean cooled, more than four billion years ago.
Subject of Research: Carbon distribution during Mercury's magma ocean evolution and the formation of its graphite crust and core composition
Article Title: How Mercury formed its graphite crust and core
Article References: How Mercury formed its graphite crust and core. (n.d.). Original publication
Image Credits: AI Generated
DOI: Not provided
Keywords: Mercury, graphite crust, magma ocean, experimental petrology, oxygen fugacity, planetary cores, MESSENGER, BepiColombo, silicon, sulphur, magnetic field, reduced planets
Cite Scienmag News
Grant Pearson. (October 10, 2026). Mercury’s Graphite Crust Traced to a Carbon Starved Core. Scienmag. https://scienmag.com/mercurys-graphite-crust-traced-to-a-carbon-starved-core/
Grant Pearson. "Mercury’s Graphite Crust Traced to a Carbon Starved Core." Scienmag, 10 October 2026, https://scienmag.com/mercurys-graphite-crust-traced-to-a-carbon-starved-core/. Accessed 10 October 2026.
Grant Pearson. "Mercury’s Graphite Crust Traced to a Carbon Starved Core." Scienmag. October 10, 2026. https://scienmag.com/mercurys-graphite-crust-traced-to-a-carbon-starved-core/

