The clay is the clue. Larissa and Galatea are two small, dark moons tucked close to Neptune, where surface temperatures hover near 50 kelvin and liquid water cannot persist. Yet the James Webb Space Telescope has detected a spectral signature of magnesium-rich phyllosilicates on both moons and in Neptune’s rings. These are minerals produced when rock is altered by liquid water for a prolonged period.
The present moons are far too small to have supplied the required heat. The leading explanation is that the minerals formed inside much larger primordial satellites, before Neptune captured Triton and the newcomer destabilised the original moon system. Collisions broke those worlds apart, exposed material once buried in their interiors and left some of the rubble to gather into the inner moons and rings seen today.
That is a compelling reconstruction, not a direct replay of the catastrophe. The team also leaves room for another large icy body to have wandered too close to Neptune and been torn apart. What Webb has measured securely is the mineral fingerprint. The parent bodies and the precise sequence of destruction must be inferred from chemistry and orbital dynamics.
Webb pulled four faint spectra out of Neptune’s glare
The researchers used the integral-field unit on Webb’s Near-Infrared Spectrograph, NIRSpec, to observe three inner moons: Proteus, Larissa and Galatea. They also combined light from the Adams, Arago and Le Verrier rings. These are punishing targets. The moons are faint, the rings are fainter and all sit beside a planet bright enough to spill scattered light across the instrument.
A custom reduction separated the targets from that background. The resulting spectra, described in the peer-reviewed Science Advances paper and its open manuscript, cover roughly 1.7 to 4.5 micrometres. Instead of the familiar surface chemistry expected for icy outer Solar System bodies, all three moons and the rings displayed an unusually deep, broad absorption around three micrometres associated with hydroxyl, or OH, bonds.
The surprise deepened because the spectra show no clear water-ice bands at 1.5, 1.65, 2.0 or 4.5 micrometres, and no obvious 3.1-micrometre Fresnel peak. Webb has found hydrated material, but not an ordinary veneer of exposed water ice.
Two moons and the rings carry a clay checkmark
Larissa, Galatea and the ring spectrum contain an additional sharp absorption centred at 2.72 micrometres. Its checkmark-like shape closely matches magnesium-rich, serpentine-like phyllosilicates measured in heavily altered CM carbonaceous chondrite meteorites. The resemblance is also strong to the clay-rich spectrum of the dwarf planet Ceres.
Phyllosilicates form through aqueous alteration. Primary silicate rock reacts with liquid water and is converted into hydrated minerals. The band shape points towards extensive alteration lasting at least roughly one to ten million years at moderate temperatures below about 300 to 400 kelvin. The team notes that no comparable phyllosilicate feature had previously been detected on an outer Solar System body beyond Jupiter.
Proteus provides an important control. It has the same deep three-micrometre hydrated-material band but lacks, or has a much weaker contribution from, the sharp clay feature. The headline therefore refers to two moons, Larissa and Galatea, plus the rings, rather than all three satellites observed.
Tiny frozen moons could not cook these minerals
Larissa and Galatea are only about 200 kilometres across. Bodies that small could not retain enough formation heat or radioactive energy to melt large amounts of internal ice for millions of years. An impact might produce brief local hydrothermal activity, but the study finds that impacts on moons this size would not warm material from about 50 kelvin to water’s melting point across the volume needed to explain the spectra.
The clay must therefore predate the current moons. In the favoured picture, it formed deep inside one or more large, differentiated icy satellites. Radioactive decay and accretional heat melted part of their internal water ice. Liquid circulated through rocky interiors and drove the same broad kind of alteration recorded in some meteorites and on Ceres.
The later destruction also had to be selective. Heating the clay above roughly 700 kelvin would dehydrate it, so the collisions that exposed and dispersed the material could not have thermally erased most of the mineral record. In my reading, that constraint turns the spectrum into more than a label: it begins to limit the violence of events that happened billions of years ago.
Triton arrived travelling the wrong way
Triton contains more than 99 per cent of the mass in Neptune’s satellite system and orbits in the direction opposite Neptune’s rotation. Its retrograde, inclined path and Pluto-like character indicate that it formed in the Kuiper Belt and was captured rather than growing in a regular disk around Neptune.
Capture would initially have placed Triton on a large, elongated orbit cutting through the region occupied by any native moons. Gravitational kicks raised their orbital eccentricities, driving close encounters, collisions and ejections while interactions with the debris helped circularise Triton’s path. This general scenario has been studied for decades and is summarised in SpaceDaily’s account of how Neptune’s inner satellites were repeatedly broken and rebuilt.
The new clay detection supplies evidence that orbital models alone could not. If the current rings and small moons contain altered rock from the interiors of large bodies, then a larger satellite generation really did exist and was catastrophically opened. The researchers estimate that perhaps only about one per cent of the destroyed material remained to build the compact system now circling inside Triton.
Nereid may be the survivor while Proteus records sorting
A separate Webb study strengthens the family history. Nereid, a medium-sized moon on a distant eccentric orbit, has a spectrum more like the regular moons of Uranus than captured Kuiper Belt objects. Simulations show that Triton’s arrival could have kicked a native moon into a Nereid-like orbit. The Caltech-led result therefore identifies Nereid as a possible intact survivor of Neptune’s original satellite system.
Proteus, meanwhile, shows that the reaccreted rubble was not perfectly mixed. Its weak clay signature may mean that it assembled from a different radial zone of the debris disk. Alternatively, later heating could have dehydrated its phyllosilicates. The same broad unidentified hydrated mineral appears on Proteus, Larissa, Galatea and the rings, suggesting a shared reservoir even though the detailed mineral proportions differ.
The discovery team favours the primordial-moon explanation because it requires no extra major event. Neptune should have formed regular moons, Triton was captured, and the capture was already expected to destroy much of that system. Clay-rich rubble fits naturally into the gaps.
A second catastrophe has not been ruled out
Natural does not mean proven. A differentiated Kuiper Belt object, perhaps comparable in scale to Pluto, could have passed within Neptune’s Roche limit and been tidally shredded after Triton settled down. If that interloper had experienced internal melting and aqueous alteration, its fragments could also seed clay-rich moons and rings.
Distinguishing the histories will require better dynamical simulations, more laboratory spectra obtained under outer Solar System conditions and observations of the other small Neptunian moons. The identity of the broad three-micrometre absorber is especially valuable. It may represent relatively primitive hydrated rock or material from hotter, deeper layers where phyllosilicates began to change.
Webb’s first Neptune portrait made the rings and moons visually crisp. NIRSpec has now made them geological samples. Larissa and Galatea may be only rubble piles today, but the clay on their surfaces carries the thermal history of worlds large enough to melt water inside, and the dynamical scar of whatever tore those worlds apart.