Giant Stars Starved of Metals Seem to Run Out of Puff

A massive star sheds its skin with the help of borrowed metal: ions of carbon, oxygen and iron soak up its fierce ultraviolet light, and the push they get flings gas off the surface at speeds of thousands of kilometers per second. With less metal on board, that push ought to weaken. The winds of massive stars poor in metals matter far beyond the stars themselves, because the models astronomers use to read the light of the James Webb Space Telescope’s early galaxies take much of their stellar physics from metal-rich stars, and a Hubble Space Telescope survey of 29 massive stars in six nearby dwarf galaxies hints that below about 10 percent of the Sun’s metallicity those winds drop off faster than the trend at higher metallicities would predict. Hints, mind you, and the team says as much.

Webb, launched in 2021, has spent its first years turning up young galaxies that refuse to behave. It “opened up a whole bunch of new questions about the evolution of these early galaxies,” says Grace Telford, an assistant professor of physics and astronomy at the University of Utah who led the work, and her verdict on the galaxies themselves is blunter still: they’re weird.

Stars above about 20 times the Sun’s mass spend their short main-sequence lives as O-type stars, so hot that much of their output lands in the extreme ultraviolet, where it can strip electrons from hydrogen (the same process that, early in cosmic history, reionized the gas between galaxies). Astronomers lump every element heavier than hydrogen and helium together as metals, and the first galaxies had precious little of them. The Milky Way, replete with the stuff, makes a poor stand-in. So the survey, the Treasury of Extremely Metal-Poor O Stars (TEMPOS), went looking in the next best places, six dwarf galaxies within about 5 million light-years, each with metal abundances below 20 percent of the Sun’s.

Faint Stars, Expensive Hubble Hours

The catch is distance. Individual stars that far off are so faint that each one eats into Hubble’s schedule, and Telford does the sums plainly: “It’s a sample of 29 stars, which doesn’t sound like a lot, but when each one costs up to 35 hours of Hubble time to observe, it gets really expensive.”

The survey added fresh far-ultraviolet spectra of 12 stars to archival observations of 17 more, then reduced all 29 the same way so they could be compared like for like. Before it, only 11 O stars in galaxies this metal-poor had wind speeds from detailed modeling, too few to span the range of temperatures and life stages. The new set covers most of the O-star types these small galaxies are known to hold, with gaps where no suitable star has turned up (early-type supergiants in the poorest galaxies, for one).

Where the Winds Begin to Falter

The measurement at the heart of it is the terminal wind speed, roughly the top velocity gas reaches as it streams away, read off how far a carbon absorption line smears toward the blue. Across the sample it falls as metallicity falls, much as radiation-driven wind theory expects, but in the three poorest galaxies it seems to fall harder: “There’s sort of a smooth trend and then suddenly for lowest-metallicity stars, the wind speed really drops off,” Telford says, and she admits she was “so excited to find that fun surprise in the data.”

Some of the contrasts are stark. Two giants of the same spectral class, one in a comparatively metal-rich dwarf galaxy and one in Sextans A, show wind absorption that differs in reach by more than 600 kilometers per second, and the Sextans A star lacks the telltale emission bump altogether. Several stars in the poorest galaxies show no visible wind at all, with apparent speeds under about 400 kilometers per second, while the fastest winds in the survey top 2,000 kilometers per second, quick enough to cross the continental United States in a little over two seconds.

Set alongside 43 O stars in the Small Magellanic Cloud and 47 in the Large, measured the same way from earlier Hubble spectra, the survey’s merely metal-poor stars slot onto a fairly smooth downward slope; the extremely metal-poor ones sit below it. The iron tells a parallel story, since at a given spectral type the stars in the less impoverished galaxies absorb noticeably more ultraviolet light in dense thickets of iron lines, a hint that the sample spans a real spread in iron, the element thought to do much of the work of launching these winds.

Tentative is the authors’ own word for the steep fall, and not a courtesy. The poorest group holds just 15 stars, their metallicities are borrowed from older supergiants in the same galaxies rather than measured in the O stars themselves, and the empirical wind speeds run about 15 percent high against a model-based method.

Nor is speed the same thing as mass lost, which is what decides how much of a star is left by the end. Pinning that down needs full atmosphere models fitted to the ultraviolet spectra and to visible-light spectra from the Keck Observatory in Hawaii, work the team has under way.

If the drop holds up, the consequences could run a long way outward. Stars that shed less of themselves would keep more mass for later, which could change how they evolve, how they die and how much ionizing light they pour into their surroundings, and those are exactly the ingredients baked into the population models used to interpret Webb’s galaxies. There’s a stranger thread in the data too, as eight of the 29 stars move oddly for their place in their galaxy’s rotation, making them candidate runaways, possibly kicked loose by a companion’s supernova or a close encounter in a crowded cluster. Nobody yet knows how common such partnerships are at these metallicities.

The processed spectra are going public through the Mikulski Archive for Space Telescopes, so anyone can check the fall-off for themselves. With a handful of stars, Telford notes, “you don’t see these trends,” and with 29 you start to; whether the winds of the universe’s first massive stars were feebler still is the question the next round of modeling gets to chew on.

FAQ