Twelve dust grains caught in Earth’s stratosphere hold clays that, by the team’s calculation, formed with water carrying up to about 15% more oxygen-18 than ocean water, far heavier than the water that altered familiar carbon-rich meteorites.
Peer reviewed · Laboratory study of 12 dust grains · Water composition calculated, not measured directly
Tiny dust grains that drifted into Earth’s upper atmosphere carry chemical traces of ancient water rich in heavy oxygen, according to a study published in Science Advances on 11 November 2026. Lindsay Keller of NASA’s Johnson Space Center and colleagues at NASA and UCLA argue the grains came from the Kuiper belt beyond Neptune, so the water points to a long-predicted reservoir in the outer solar system. The catch: the water itself is gone. Its make-up was worked out from minerals, and the Kuiper belt origin is inferred from radiation damage.
What you need to know
- Finding: Clay-rich grains point to water with an oxygen-18 value of about +155‰ (per mil) relative to ocean water, an upper-limit estimate from the team’s model.
- Method: Electron microscopes at NASA Johnson mapped each grain’s minerals, and an ion microprobe at UCLA measured its three oxygen isotopes; a mass-balance calculation then recovered the water.
- Source: Keller, Snead, Liu and McKeegan, “A primordial heavy water reservoir in the outer solar system”, Science Advances, 11 November 2026 (open access).
- Why it matters: Ice rich in oxygen-17 and oxygen-18 has long been expected in the cold outer solar system, but meteorites almost never preserve it.
- Limitation: Only 12 grains, each 10 to 30 micrometres across, and both their Kuiper belt origin and the water’s composition are inferences, not direct measurements.
Evidence at a glance
| Item | Assessment |
|---|---|
| Evidence status | Peer reviewed. Laboratory analysis of 12 interplanetary dust particles; the water’s isotopic composition is calculated from a mass-balance model. |
| Primary source | Lindsay P. Keller, Christopher J. Snead, Ming-Chang Liu and Kevin D. McKeegan, “A primordial heavy water reservoir in the outer solar system”, Science Advances 12(46), eaeg1602. Received 4 February 2026, accepted 9 October 2026, online 11 November 2026. CC BY-NC 4.0. |
| Evidence strength | Moderate for the measured isotopes, which carry stated 2-sigma errors and were checked against standards. Limited for the conclusion about a large outer reservoir, because it rests on 12 grains, an estimated starting-rock composition and an origin inferred from earlier track-density work. |
| Principal limitation | Nobody measured the water. The +155‰ figure is back-calculated, depends on the assumed composition of the starting rock, and the authors describe it as an upper limit. |
| Coverage at a glance | 3 distinct credible publishers found as of 17 November 2026, 08:50 UTC (ScienceAlert, Phys.org and The Brighter Side of News), from SciQuest web searches for the paper title and topic. The EurekAlert release itself, a republished copy of it and rewrite or aggregator sites were excluded. Coverage is not evidence that a finding is correct. |
Sources: the Science Advances paper (full text via Europe PMC, PMC13753991), the AAAS release on EurekAlert and SciQuest’s coverage search at the time stated.
Is this a big deal?
It’s a solid step on an old question, not a settled answer. Planetary scientists have long expected ice in the cold outer solar system to be rich in heavy oxygen, yet meteorites almost never preserve it. The clearest traces were a rare mineral in one meteorite and Rosetta’s measurements of water at comet 67P. These grains add a second line of evidence, and one tied, through their radiation history, to a region beyond the giant planets. But 12 particles are a small sample, and the key number comes from a model.
What scientists found
All 12 grains are “hydrated”, meaning they contain minerals that formed when liquid water reacted with rock. Most of their mass is clay, chiefly saponite with some serpentine, threaded with tiny iron-nickel sulphide grains, magnesium-iron carbonates and abundant carbon. On average they hold about four times more carbon than CI chondrites, the meteorites usually taken as the benchmark for the Sun’s own composition.
Their oxygen is unusual. On average the grains have a δ18O of +11.6‰ and a δ17O of +8.5‰. Those symbols measure how far the ratio of a heavy isotope to ordinary oxygen-16 sits above ocean water, in parts per thousand. The pattern plots along a slope-1 line on the scientists’ standard isotope chart, unlike the shallower trends seen in water-altered meteorites.
From those numbers the team worked back to the water. Their answer: it had δ18O of about +155 ± 17‰ and δ17O of about +143 ± 32‰. That means roughly 15% more oxygen-18 than ocean water. The authors call it “a major heavy water reservoir beyond the orbits of the gas giant planets.”
“Heavy water” can mislead. In everyday use it means water made with deuterium, the heavy form of hydrogen, used in some nuclear reactors. Here it means ordinary H₂O whose oxygen atoms are more often the heavier oxygen-17 (9 neutrons) or oxygen-18 (10 neutrons) instead of oxygen-16 (8 neutrons). The study measured oxygen, not hydrogen. (For the hydrogen kind, see our report on deuterium in the interstellar comet 3I/ATLAS.)
How they found it
The grains come from NASA’s cosmic dust collection. Since 1981, NASA has flown high-altitude aircraft, the ER-2 and WB-57, with collector plates coated in foam or sticky silicone oil that open only in the stratosphere, above most Earthly pollution. Keller’s team picked 12 smooth, compact “C-type” particles at random from the catalogues kept by the curation facility at Johnson Space Center.
Each grain was sliced into sections only 50 to 70 nanometres thick and examined with a transmission electron microscope, which identified the minerals and also counted radiation tracks. Those tracks are trails of damage left in crystals by energetic particles from solar flares, and they build up the longer a grain drifts through space. Densities of roughly 10 billion to 100 billion tracks per square centimetre are about 100 times higher than expected for dust from the asteroid belt or from Jupiter-family comets. In earlier work, Keller and Flynn argued that only a long trip in from the Kuiper belt explains so much damage.
The rest of each grain went to UCLA, where a Cameca ims-1290 ion microprobe fired a beam of caesium ions at it and counted the oxygen-16, -17 and -18 ions knocked loose. The researchers corrected for a molecular signal (oxygen-16 bonded to hydrogen) that falls almost on top of oxygen-17, and checked their calibration against a slice of the Orgueil meteorite.
Think of it like mixing paint. If you know the colour you started with, the colour you ended with, and that one part in eight came from a second tin, you can work out what colour that tin was. That is the mass-balance step. The analogy breaks down in one place: the “starting colour” here, the dust’s original glassy silicate, was not measured in these grains. The team estimated it from earlier measurements of dry dust particles, and a different starting value would shift the answer.
Show the maths: from clay back to water
The team’s main reaction turns glassy silicate and iron metal into saponite clay. Of the oxygen atoms in the products, about 1 in 8 comes from the water and 7 in 8 from the rock. So the grain’s value is a weighted average:
grain ≈ (7/8) × rock + (1/8) × water, which rearranges to water ≈ 8 × grain − 7 × rock.
For oxygen-18: 8 × (+11.6) − 7 × (−8.8) = 92.8 + 61.6 = +154.4‰. The paper gives +155 ± 17‰.
For oxygen-17: 8 × (+8.5) − 7 × (−10.6) = 68.0 + 74.2 = +142.2‰. The paper gives +143 ± 32‰.
A δ18O of +155‰ means the oxygen-18 to oxygen-16 ratio is 1.155 times that of ocean water, about 15% higher. The authors left out loosely bound water, which is lost in space or in the instrument’s vacuum, so they call these values an upper limit. SciQuest recalculated these figures from the paper’s Tables 1 and 2.
Why it matters to you
This won’t change anything in your daily life. It is a wonder question: what the raw ingredients of the solar system were and where they sat. It also has a curious personal edge. Estimates cited in the paper suggest up to 25% of the interplanetary dust reaching the inner solar system comes from Kuiper belt objects, so some of the fine dust settling through Earth’s atmosphere may have begun its journey beyond Neptune.
What the evidence does not show
It does not show water from the Kuiper belt directly. No liquid or ice survives in the grains; the composition is reconstructed. It also relies on two assumptions the authors state: that the dry, glassy dust is the right starting material, and that the reactions they model are the main ones.
The Kuiper belt origin is an interpretation too. It depends on how fast solar-flare tracks accumulate, a rate Keller and Flynn recalibrated in earlier work. One grain, L2071E35, had no crystals able to record tracks at all. Per-grain results also vary widely: the calculated water ranges from +111 to +182‰ in oxygen-18 across the 12 grains.
Finally, the paper does not identify the parent bodies or say how they got warm enough for liquid water.
How this fits what we already knew
The Sun is richer in oxygen-16 than Earth and most meteorites, as NASA’s Genesis mission showed from captured solar wind. A leading explanation is that ultraviolet light broke apart carbon monoxide in the young solar system’s gas in a way that left water ice enriched in oxygen-17 and oxygen-18. Finding that heavy ice has been hard.
| Sample | What it is | Water Δ17O (‰) | Type of value |
|---|---|---|---|
| CI and CM chondrites | Water-altered carbon-rich meteorites | about +3 to +7 | Calculated from minerals |
| Two-reservoir models of CM and CI meteorites | Earlier estimate of primordial water | about +7 to +9 | Modelled |
| 12 Kuiper belt dust grains (this study) | Water that altered the grains | +62 ± 34 (per grain +26 to +150) | Calculated, upper limit |
| Cosmic symplectite in the Acfer 094 meteorite | Rare iron-oxide and sulphide intergrowth | about +80 | Measured in the mineral |
Δ17O measures how far a sample sits from the line that ordinary, mass-dependent Earth processes follow (Δ17O = δ17O − 0.52 × δ18O); ocean water is 0. All values as given or cited in Keller et al. (2026). Uncertainties are 2-sigma.
The table brings out the paper’s main argument. The water that altered familiar meteorites looks only mildly heavy, which the authors read as primordial water that had mixed with oxygen-16-rich gas and dust closer in. The dust grains point to a much heavier store that overlaps with the rare symplectite, now tied to a region beyond the giant planets. The grains also lack the calcium-rich carbonates and the magnetite typical of CI meteorites, and calcium carbonates turn up in samples returned from asteroid Bennu too. The authors suggest the grains’ watery phase may have been short-lived or limited in water.
The authors propose that the outer reservoir stayed largely cut off from the inner solar system, most likely by Saturn alone or together with the ice giants Uranus and Neptune, citing simulations of a pressure bump outside Saturn’s orbit that would have slowed inward-drifting dust. That part is a proposal resting on other teams’ models.
Independent perspective
SciQuest found no independent comment on this paper at the time of writing and did not obtain one. When Keller and Flynn first presented the Kuiper belt origin for track-rich dust in 2019, Science News reported that planetary scientist Carey Lisse of the Johns Hopkins Applied Physics Laboratory welcomed the idea but said many more samples were needed. With 12 grains, that point still stands.
How other headlines framed it
ScienceAlert went for scale, describing primordial water hidden at the edge of the solar system, and its opening treated the Kuiper belt origin as settled before adding hedges later. Phys.org named the isotope, calling it a heavy-oxygen reservoir, and covered the Saturn barrier idea. The Brighter Side of News placed the reservoir beyond Neptune and was clearest in separating heavy oxygen from deuterium. None of the three led with the fact that the water’s composition is calculated rather than measured.
What happens next
The obvious test is more grains, and NASA’s curation team says it continues to collect cosmic dust with high-altitude aircraft and plans to add high-altitude balloons. Measuring the starting silicate inside the same grains, rather than borrowing it from other particles, would tighten the water estimate. The paper does not name a mission or a date for follow-up work.
In one sentence
Clay in 12 grains of dust that most likely came from beyond Neptune suggests the outer solar system once held water unusually rich in oxygen-17 and oxygen-18, kept apart from the inner planets, though the water’s make-up is a calculated upper limit from a small sample.
Quick check: did it click?
- Is this “heavy water” the deuterium kind? No. It means water with more of the heavy oxygen isotopes, oxygen-17 and oxygen-18.
- How do researchers link the dust to the Kuiper belt? From the very high density of solar-flare tracks in its crystals, which implies a long journey through space.
- Why is the water value much heavier than the grains themselves? Only about one oxygen atom in eight in the clay came from water, so the water had to be strongly enriched to shift the grains even slightly.
Read next
- 3I/ATLAS methane holds more deuterium than any measured in the Solar System: the hydrogen kind of “heavy”, measured in an interstellar comet.
- Argon-40 in Uranus and Neptune may trace their origins more than their rock: another isotope clue to how the outer solar system formed.
- Earth’s core carbon may be isotopically heavy, Hawaiian lava droplets suggest: isotopes linking Earth to outer solar system meteorites.
Sources and disclosure
- Keller, L. P., Snead, C. J., Liu, M.-C. and McKeegan, K. D. A primordial heavy water reservoir in the outer solar system. Science Advances 12, eaeg1602 (2026). Full text: PMC13753991.
- AAAS, “Dust from the Kuiper Belt suggests the outer solar system contains water with heavy oxygen”, EurekAlert, 11 November 2026.
- Keller, L. P. and Flynn, G. J. Evidence for a significant Kuiper belt dust contribution to the zodiacal cloud. Nature Astronomy 6, 731–735 (2022).
- NASA Johnson Space Center Astromaterials Curation, Aircraft Collected Particles collection description.
- L. Grossman, “Kuiper Belt dust may be in our atmosphere (and NASA labs) right now”, Science News, 27 March 2019.
Disclosure: This article was drafted with AI assistance and checked against the sources above. The mass-balance check and the 15% conversion are SciQuest calculations from the paper’s tables; the graphic is a SciQuest original based on the paper’s data. The study authors declare no competing interests. Spotted an error? See our corrections policy.
Discover more from SciQuest
Subscribe to get the latest posts sent to your email.

Leave a comment