Far from the warmth needed to reshape rock with liquid water, Neptune’s small inner moons should be geologically quiet and deeply frozen. Instead, new observations have uncovered minerals that require millions of years of interaction with liquid water, hinting that these modest moons are not what they appear to be at all. Rather than preserving their own history, they may be the surviving fragments of much larger worlds that were destroyed long ago, leaving behind rare pieces of their once-hidden interiors.
The small moons orbiting close to Neptune have always presented a puzzle. Unlike the giant icy satellites elsewhere in the Solar System, these bodies are irregularly shaped, relatively small, and thought to have formed after an earlier generation of satellites disappeared. Their present-day appearance offers few clues about that violent history.
Now, observations made with the James Webb Space Telescope (JWST) have uncovered an unexpected chemical fingerprint on these moons and Neptune’s rings—one that appears to preserve evidence from deep inside ancient icy worlds that no longer exist.
The study focuses on Proteus, Larissa, Galatea, and Neptune’s rings, using infrared spectroscopy to investigate the minerals on their surfaces. Instead of finding simple frozen material expected in the outer Solar System, the researchers detected abundant hydrated minerals, including phyllosilicates that are remarkably similar to minerals found in highly altered carbonaceous meteorites.
Those minerals tell a story that begins long before the current moons were ever assembled.
Frozen moons unexpectedly displayed minerals made by liquid water
The key discovery emerged from infrared spectra collected by JWST’s Near Infrared Spectrograph (NIRSpec).
Every observed moon and the rings exhibited a strong absorption feature near 3 micrometers, indicating the presence of hydroxyl (OH)-bearing hydrated material. Two of the moons—Larissa and Galatea—also displayed an especially distinctive absorption centered near 2.72 micrometers.
That particular spectral feature immediately attracted attention because it closely matches Mg-rich serpentine phyllosilicates, minerals formed through prolonged interaction between rock and liquid water.
These are not minerals that form quickly.
According to the study, producing this degree of aqueous alteration generally requires more than 1 to 10 million years of exposure to liquid water at relatively modest temperatures of roughly 300 to 400 kelvin or below.
Those conditions simply do not exist on the frozen surfaces of Neptune’s present inner moons.
Instead of explaining the moons themselves, the minerals appear to preserve evidence from an earlier chapter in Neptune’s history.
The mineral fingerprints closely resemble some of the Solar System’s most altered rocks
To determine exactly what they were seeing, the researchers compared the JWST spectra with laboratory measurements of meteorites and with observations of well-studied bodies elsewhere in the Solar System.
The match proved surprisingly specific.
The spectral signature closely resembles CM2 carbonaceous chondrites that have experienced extensive aqueous alteration. As alteration progresses within these meteorites, their infrared absorption shifts toward the same sharp 2.7-micrometer “checkmark” shape observed on Larissa and Galatea.
The comparison also extended to larger planetary bodies.
Among known objects, the Neptunian moons most closely resemble Ceres, whose surface is rich in Mg-bearing phyllosilicates produced through widespread aqueous alteration.
That similarity carries important implications.
Ceres underwent extensive internal alteration driven by liquid water. If Neptune’s small moons display comparable mineralogy, then the material now visible on their surfaces likely experienced similar processes before these moons ever existed in their present form.
The moons themselves are simply too small to have produced these minerals
Finding hydrated minerals raised an obvious question.
Could Larissa and Galatea have generated them internally?
The study concludes that this explanation is highly unlikely.
Larissa measures roughly 194 kilometers across, while Galatea is about 176 kilometers in diameter.
Objects this small cannot realistically generate enough internal heat to melt substantial amounts of water ice for the prolonged periods required to produce the observed phyllosilicates.
The researchers point to thermal models of Miranda, one of Uranus’s moons, which is approximately 470 kilometers across—about two and a half times larger than Larissa and Galatea. Even Miranda is thought to have struggled to generate enough internal heat to melt water ice without an additional source such as tidal heating.
If Miranda barely approaches those conditions, the much smaller Neptunian moons are even less capable of sustaining them.
This effectively rules out the idea that the minerals formed inside the present-day moons.
Impacts also fail to explain the observations
Another possible explanation involves collisions.
Large impacts can briefly generate heat, and laboratory experiments have shown that shock processes are capable of producing localized hydrated minerals under some conditions.
However, the evidence does not fit this scenario either.
The temperatures generated by impacts on bodies the size of Larissa and Galatea would generally be too low to melt water ice starting from Neptune’s frigid surface temperatures of around 50 kelvin.
Even more importantly, hydrothermal systems created by impacts last far too briefly.
The phyllosilicates detected by JWST require alteration lasting millions of years, whereas impact-generated hydrothermal activity would persist for much shorter timescales.
The researchers also note that an impact energetic enough to maintain those conditions would probably destroy bodies as small as Larissa and Galatea altogether.
Taken together, these constraints make impact-driven formation an unlikely explanation.
The minerals instead appear to come from worlds that no longer exist
With both local formation mechanisms ruled out, the researchers arrive at a different possibility.
The current moons may have reaccreted from fragments of much larger differentiated icy bodies whose interiors had already undergone extensive aqueous alteration.
In other words, the surfaces visible today may actually be pieces of the deep interiors of ancient worlds.
The study argues that the most plausible source is a primordial system of regular satellites that once orbited Neptune before being catastrophically destroyed.
Those original satellites would have been much larger than today’s inner moons.
Their interiors could have remained warm enough for liquid water to circulate through rock over millions of years, producing the Mg-rich phyllosilicates now detected by JWST.
When those larger moons later broke apart, fragments of their interiors became incorporated into the smaller bodies that eventually reassembled.
Triton’s arrival may have triggered the destruction
How did those original satellites disappear?
The researchers discuss two broad possibilities.
One proposes that Neptune’s original satellite system was destroyed during the capture of Triton, Neptune’s largest moon.
The other suggests that a large differentiated Kuiper Belt Object wandered too close to Neptune and was torn apart by tidal forces within the planet’s Roche limit.
The study favors the first explanation.
One reason comes from separate JWST observations of Nereid, Neptune’s third-largest moon.
Those observations indicate that Nereid’s composition is inconsistent with known Kuiper Belt Objects and instead supports the idea that it may be the lone surviving member of Neptune’s original regular satellite system.
Combined with dynamical simulations showing that Triton’s inward migration could have displaced one original satellite onto Nereid’s present orbit, this provides observational support for the existence—and later destruction—of an earlier satellite system.
Under this interpretation, today’s inner moons and rings formed afterward from debris left behind by that catastrophic event.
The authors note that this scenario also requires fewer separate dynamical events than the alternative involving the tidal disruption of a passing Kuiper Belt Object.
One mystery became even harder to explain
Although the new observations strongly support a reaccreted origin, they also expose an unexpected problem.
The researchers found no evidence for water ice on any of the observed inner moons or rings.
That absence is surprising.
If the precursor bodies melted enough ice internally to produce phyllosilicates, they should also have contained abundant water-ice-rich outer layers.
Similarly, if the moons formed from the disruption of a differentiated Kuiper Belt Object, significant amounts of water ice would also be expected to survive.
Instead, the observations point toward material that is remarkably poor in volatile ice.
The authors suggest several possibilities.
Perhaps much of the water ice was lost during later collisional evolution. Alternatively, the reaccreted material may have originated preferentially from rocky interior regions rather than icy outer layers.
At present, however, the study cannot determine which explanation is correct.
Proteus tells a different part of the story
The largest inner moon, Proteus, behaves differently from Larissa and Galatea.
It shares the broad 3-micrometer hydroxyl absorption found throughout the system, indicating abundant hydrated material.
Yet it lacks strong evidence for the Mg-rich phyllosilicates seen on the smaller moons.
This difference could reflect several possibilities discussed by the authors.
Proteus may have reaccreted from material that never experienced extensive aqueous alteration.
Alternatively, it may once have contained phyllosilicates that were later heated enough to dehydrate them, weakening the characteristic spectral signature.
The researchers note that impacts capable of ejecting material from Proteus—possibly including the event thought to have formed the nearby small moon Hippocamp—might have altered its surface mineralogy.
Another possibility is that Proteus formed from a different portion of the ancient debris disk.
The available observations cannot yet distinguish among these explanations.
An unidentified hydrated mineral remains one of the study’s biggest puzzles
Every observed moon and the rings display an exceptionally deep 3-micrometer OH absorption band.
Yet despite comparisons with multiple laboratory spectral libraries, the researchers could not identify the mineral responsible.
Its spectral characteristics differ from known water ice and from previously cataloged hydrated minerals.
Several possibilities remain.
The material could represent relatively unaltered rocky components incorporated into Neptune’s original satellites.
Alternatively, it may have formed deeper inside those ancient bodies under temperatures and pressures that altered phyllosilicates into a different hydrated phase.
Resolving this mystery will require new laboratory measurements of candidate minerals under outer Solar System conditions.
Until then, the identity of the dominant hydrated material remains unknown.
A possible hint of ammonia adds another layer
The spectra also contain a subtle feature near 3.07 micrometers.
The researchers suggest that it could indicate the presence of ammonia-bearing material, although they treat this interpretation cautiously.
On Ceres, a similar feature has been attributed to ammoniated phyllosilicates.
However, the Neptunian observations differ in an important way.
The possible ammonia-related feature is strongest on Proteus, which is actually the least phyllosilicate-rich moon.
Because of this mismatch, the study argues that ammonia is more likely associated with the unidentified hydrated material than with the phyllosilicates themselves.
If correct, this could indicate that ammonia survived within precursor material that never became hot enough to release it completely.
The minerals place new limits on Neptune’s violent past
The phyllosilicates themselves also constrain the physical history of the Neptunian system.
Once formed, these minerals cannot survive temperatures much above approximately 700 kelvin without dehydrating.
That means the catastrophic disruption of Neptune’s original satellites—and all subsequent collisions that eventually produced today’s moons—must have avoided heating most of the material beyond that limit.
This provides new observational constraints for future models of Triton’s capture, satellite destruction, debris-disk evolution, and moon formation.
The researchers emphasize that many details remain unresolved.
Better understanding the unidentified hydrated material, the apparent absence of water ice, and the processes governing debris reaccretion will be necessary before Neptune’s early history can be reconstructed with greater confidence.
Small moons may have become rare windows into hidden planetary interiors
The study concludes that the hydrated minerals detected on Neptune’s inner moons almost certainly formed deep inside much larger bodies where liquid water persisted for millions of years.
Whether those bodies were primordial Neptunian satellites or a large differentiated Kuiper Belt Object remains an open question, although the available evidence most naturally supports the destruction of Neptune’s original satellite system during Triton’s capture.
If that interpretation is correct, Larissa, Galatea, and their neighboring moons are far more than small frozen satellites.
They are surviving fragments of ancient worlds whose interiors would otherwise remain forever inaccessible.
In the researchers’ view, these moons may represent the only place in the Solar System where scientists can directly examine material from the deep interior of an icy satellite—or possibly even a dwarf planet—that has been naturally excavated and preserved after catastrophic destruction, offering an unusually direct glimpse into processes that shaped the outer Solar System billions of years ago.
Publication details
M. Ryleigh Davis et al, Neptune’s inner moons and rings are exposed icy body interiors, Science Advances (2026). DOI: 10.1126/sciadv.aeb1437
