Mercury’s Surface May Contain Just 37% Silicon Dioxide

Mercury’s surface may be substantially poorer in silicon dioxide than previously believed. A new study estimates its silicate surface at roughly 37% SiO₂ by mass, compared with earlier estimates of about 49–60%.140 If confirmed, the result would point to volcanic rocks produced by unusually hot, extensive melting deep inside the planet. But the measurement also has an important caveat: Mercury formed under extremely oxygen-poor conditions, so some silicon may exist in metallic or carbide forms rather than as silicon dioxide.3342

How researchers estimated Mercury’s silica content

No spacecraft has landed on Mercury to analyze rocks directly, and no Mercury samples have been returned to Earth. The researchers therefore used an indirect infrared method based on the Christiansen Feature, a characteristic emissivity maximum whose position changes with the silicon-dioxide content of silicate materials.130

The team first measured laboratory glass beads made with compositions relevant to planetary surfaces. Those measurements established an empirical relationship between the Christiansen Feature’s wavelength and the amount of SiO₂ in the glass. They then applied the calibration to high-resolution infrared observations of the Moon, where researchers could compare the resulting map with Apollo, Luna, and Chang’e samples. That lunar comparison provided a way to test the method before using it on telescopic observations of Mercury.1

The resulting Mercury estimate is not a direct chemical assay. It is best understood as the approximate silicate SiO₂ equivalent inferred from the planet’s infrared signal.

What 37% SiO₂ could mean for Mercury

A surface this silica-poor would support a more extreme picture of Mercury’s volcanic history. The planet’s smooth plains cover large areas and are widely understood to have formed through volcanism.3637 Producing such low-silica lavas could require unusually high degrees of partial melting in a mantle that was hotter and melted more deeply than conventional models suggest.140

The finding would therefore connect Mercury’s present-day surface to an especially intense early period of planetary differentiation and volcanism. It does not, however, determine a single mantle temperature, melting depth, or geological pathway. Those conclusions depend on how representative the laboratory silicate glasses are of Mercury’s chemically unusual surface.

The alternative: silicon may not be in SiO₂

Mercury is one of the most chemically reduced terrestrial planets known. Measurements have found high sulfur and relatively low iron and oxygen, consistent with formation under conditions in which oxygen was scarce.334244

In that environment, not all of Mercury’s silicon necessarily has to be bonded to oxygen. Some could occur in reduced forms such as metallic silicon or silicon carbide. Those phases could alter the infrared signal in a way that resembles a silicate surface with unusually little SiO₂.40

That possibility changes how the headline number should be interpreted. The study does not prove that only 37% of all silicon-bearing material on Mercury is silicon dioxide. Instead, it provides a low estimate for the surface’s silicate silica content under the assumptions used in the calibration.

How BepiColombo can test the result

The ESA–JAXA BepiColombo mission is approaching the phase that could resolve this uncertainty. Its Mercury Transfer Module is scheduled to separate on 3 September 2026, while the spacecraft’s orbit-insertion sequence is scheduled to begin on 21 November 2026.1724 Full science operations are planned for April 2027.26

BepiColombo’s MERTIS instrument is designed to observe Mercury in thermal infrared wavelengths, including the spectral region containing the Christiansen Feature.30 Unlike earlier disk-integrated observations, MERTIS should be able to connect infrared measurements with individual regions and geological units across the planet.

Those observations can answer several questions:

• Is the low inferred SiO₂ abundance broadly representative of Mercury’s surface?
• Does it vary between volcanic plains, cratered terrain, and other geological units?
• Do regional patterns favor unusually deep mantle melting?
• When combined with elemental and mineralogical measurements from other instruments, do they indicate that some silicon is stored in reduced metallic or carbide phases?

Until those measurements are available, the strongest conclusion is cautious but significant: Mercury’s volcanic surface may be far less silica-rich than assumed, revealing either an exceptionally hot and extensively melted mantle, an even more unusual reduced chemistry, or a combination of both.