A primitive dwarf galaxy with almost none of the heavy elements found in the Milky Way is somehow producing iron and silicon-carbide dust — materials astronomers thought early galaxies would struggle to make

The dwarf galaxy Sextans A is poor in the elements astronomers collectively call metals. Yet observations from the James Webb Space Telescope indicate that two of its ageing stars are making dust from silicon carbide and, probably, almost pure metallic iron.

That pairing matters because silicon and iron are not ingredients these stars can manufacture in abundance for themselves. They inherited only a small supply from earlier generations of stars. Models have therefore tended to predict that dust containing those elements should become scarce as metallicity falls.

The result does not show an ancient galaxy caught in the early universe. Sextans A is nearby, about four million light-years away. Its value is as a local chemical analogue: astronomers can resolve individual stars in an environment containing only about 1 to 7 per cent of the Sun’s metal abundance.

Six old stars under Webb’s infrared eye

The study, led by Martha Boyer of the Space Telescope Science Institute and published in The Astrophysical Journal, used the low-resolution spectrometer on Webb’s Mid-Infrared Instrument, or MIRI. The team examined six asymptotic giant branch stars in Sextans A: five carbon-rich stars and one oxygen-rich M-type star.

Asymptotic giant branch, or AGB, stars are low- and intermediate-mass stars in a late stage of evolution. They have exhausted hydrogen in their cores and later swell, pulsate and shed their outer layers through stellar winds. As the expelled gas moves away and cools, atoms and molecules can condense into microscopic solids. Those grains enter interstellar space, where some survive to become material for later stars, asteroids and planets.

The oxygen-rich target, catalogued as star 90034, probably began with about four to five times the Sun’s mass. The carbon star central to the other result is catalogued as 90428. Webb did not detect the same dust around both. It found evidence for two different routes through the same shortage of heavy elements.

Silicon carbide left a recognisable fingerprint

Star 90428 showed emission near a wavelength of 11.3 micrometres. That is a characteristic infrared feature of silicon carbide, or SiC. Its spectrum also contained strong absorption by acetylene, consistent with a carbon-rich atmosphere. The authors describe the SiC identification as confirmed.

Carbon stars have one advantage in a metal-poor environment. During the AGB phase, repeated mixing events can dredge freshly synthesised carbon from the stellar interior to the surface. Once carbon outnumbers oxygen there, the excess can condense into amorphous carbon grains. The star is partly supplying its own raw material.

Silicon is different. An AGB star does not replenish it in the same way. The silicon available for SiC grains largely reflects the material from which the star formed. A 2024 independent review of dust sources in the early universe notes that SiC and iron-grain production should become less efficient as metallicity falls.

Sextans A is now the lowest-metallicity galaxy known to host a star with detected SiC dust. That record belongs to one object in a sample of six. It does not establish that silicon carbide is abundant throughout the galaxy, and the study notes that the few metal-poor carbon stars with comparable SiC could be unusual rather than typical.

The iron result is a model-based inference

Star 90034 presented a different problem. Its infrared emission showed a strong excess, indicating warm circumstellar dust, but it lacked the clear silicate emission normally expected near 10 micrometres. The spectrum was comparatively smooth.

The researchers compared the observed spectral energy distribution with models containing different grain mixtures. A model dominated by metallic iron reproduced it best. Fits using amorphous carbon or large silicate grains performed poorly. The preferred model allowed less than one per cent silicate dust, although such a small contribution could not be excluded.

It is therefore reasonable to say the star is likely producing iron dust. It would be too strong to say Webb directly identified iron in the same manner as the SiC feature. Metallic iron absorbs light efficiently but produces a largely featureless infrared continuum, so the conclusion rests on model comparison and on alternative compositions failing to match the data as well.

There is another linguistic trap here. The star is not forging fresh iron nuclei. It is condensing inherited iron atoms into solid grains in its wind. The surprise is that this process appears efficient despite the star beginning with such a small inventory.

Why the usual silicate route may fail

In a metal-rich, oxygen-bearing AGB star, magnesium, silicon and oxygen can combine into silicates such as olivine. Sextans A began with very little magnesium and silicon. Star 90034 may also be undergoing hot bottom burning, in which the base of its convective envelope becomes hot enough for nuclear processing. That can reduce the surface abundances of oxygen and magnesium required for familiar silicate grains.

Under those conditions, silicate formation may stall while metallic iron still condenses. Iron is not a substitute manufactured after the other ingredients disappear. It may simply be the solid that this particular wind’s chemistry permits.

The estimated output is potentially substantial. If the measured production rate continued through the last 20,000 to 30,000 years of the star’s AGB evolution, the paper calculates that it would make roughly 0.9 to 3.7 times the iron dust mass predicted by existing models, depending on the assumed stellar mass. The range extends from close agreement to a clear excess. It should not be collapsed into a claim that every model has failed.

A nearby analogue, not a photograph of the first galaxies

NASA describes Sextans A as containing roughly 3 to 7 per cent of the Sun’s metal content, while the paper discusses a broader range down to about one per cent for its stellar population. The phrase “almost none” is defensible as a comparison with the Milky Way, but it does not mean the galaxy is metal-free.

Nor is low metallicity the same thing as great age. Sextans A is a nearby dwarf irregular galaxy with ongoing star formation. Astronomers study it because its chemistry resembles conditions more common in young galaxies, while its proximity lets Webb separate stars that would blur together at high redshift.

That makes it a useful test case, not a perfect reconstruction. Early galaxies experienced different radiation fields, gas densities, merger histories and rates of star formation. A dust pathway seen in Sextans A still has to be shown to operate widely before it can be inserted into models of the distant universe with confidence.

Why faster ageing stars matter

Dust in early galaxies presents a timing problem. Webb has detected substantial dust in systems seen when the universe was young, but many low-mass carbon-producing AGB stars take hundreds of millions of years to evolve. Supernovae and the growth of grains in interstellar clouds are therefore usually given much of the early work.

Higher-mass AGB stars evolve faster. The Boyer paper argues that stars near the upper end of the AGB mass range could begin returning dust within roughly 30 to 50 million years of forming. If metallic iron production like that inferred around star 90034 is common, these stars could contribute earlier and with a different grain mixture than many dust-evolution models currently assume.

That “if” carries most of the scientific weight. The study observed six stars. The confirmed SiC feature belongs to one, and the likely iron-dominated wind belongs to another. The authors explicitly call for observations of a larger population of massive, metal-poor AGB stars before drawing conclusions about whole-galaxy dust budgets.

Sextans A has not settled the origin of dust in the early universe. It has exposed a plausible route that standard assumptions may undercount. Even where familiar rock-forming ingredients are scarce, ageing stars can apparently turn a thin inheritance of heavy elements into solids, and the composition of those grains may matter as much as their total mass.