Chandra’s archive held 84 X-ray sources nobody had classified

Eighty-four point-like X-ray sources, scattered across six nearby galaxies, were hiding in plain sight in NASA’s Chandra X-ray Observatory archive. They give off so little high-energy radiation that ordinary X-ray surveys passed over them, yet they pour out ultraviolet light. A team led by Mustafa Muhibullah at the University of Alabama has now grouped them into a single, newly defined class and named it hypersoft X-ray sources.

The work appeared in Nature Astronomy on September 9, 2026, with Jimmy Irwin of the University of Alabama and Rosanne Di Stefano of the Center for Astrophysics, Harvard and Smithsonian, as co-authors.

Eighty-four sources in six galaxies

NASA’s Chandra release and the Chandra X-ray Center’s version both report the same tally: 84 hypersoft sources in six galaxies, two spiral and four elliptical. The spirals are the Andromeda Galaxy, catalogued as M31, and the Pinwheel Galaxy, M101. The ellipticals named in the paper’s own coverage include NGC 3379 and NGC 4472; the two remaining elliptical galaxies are not named in the releases, and no source breaks the 84 down galaxy by galaxy.

All of the sources came from data Chandra had already collected. Nobody pointed the telescope at a new target. Instead the team went back through archived observations and looked at the lowest-energy end of the instrument’s range, where a hypersoft source shows up in the faintest-energy images and then vanishes in harder X-ray bands. Di Stefano said that combing the Chandra archive eliminated what used to be a blind spot for telescopes, a remark carried in the Newswise copy of the release.

Hypersoft means below 0.3 keV

The Nature Astronomy paper defines hypersoft sources as luminous, point-like X-ray objects that sit outside the centre of their host galaxy and are detected mainly below 0.3 kiloelectronvolts. Their X-ray output runs near 10³⁸ ergs per second, but spectral models imply that the total energy they radiate is far larger, concentrated in the extreme ultraviolet where Chandra cannot see.

A Chandra blog entry describes them as radiating hundreds of thousands to millions of times the energy of the Sun while still being cooler than a typical X-ray binary. Muhibullah said in the release that the team had never met a group of objects that behave like this one, and that working out what they are was the obvious next job, a task made harder because a source that radiates almost entirely in the extreme ultraviolet leaves only a faint X-ray signature for Chandra to measure. The Chandra blog dates its post to September 8, a day before the journal paper’s publication date.

No new observing time was needed.

The scale of the archive search is worth stating plainly. Chandra has been operating since 1999, and each of the six host galaxies has been observed repeatedly over those years. The sources were present in that data the whole time; what changed was the question asked of it. Muhibullah has worked on the project as a doctoral student since fall 2021, according to the Chandra blog, so the identification of a coherent class took roughly five years of analysis before the Nature Astronomy paper appeared.

Candidates: white dwarfs, neutron stars and black holes

The authors do not claim a single identity. The release says the objects probably involve a black hole, neutron star or white dwarf pulling material from a companion star, with the heated gas producing both low-energy X-rays and strong ultraviolet light. The paper singles out accreting white dwarfs and post-nova systems as likely members, with accreting black holes possible for the most luminous.

The presence of these sources in both star-forming regions and old stellar populations suggests more than one kind of system. Muhibullah’s group points to a possible link with Type Ia supernova progenitors, a longstanding puzzle. Irwin remarked that finding a way to spot such explosions before they go off would be very important. A hypersoft source could be a white dwarf in a long phase of steady nuclear burning on its way to exploding, which the paper treats as a hypothesis, not a result.

A second open problem is the gas between stars. The release notes that the intense ultraviolet from hypersoft sources may help explain what ionises that gas in galaxies, though no measurement of that effect is reported. Sci.News summarises the discovery as evidence for large, previously undetected populations of binary systems with intense ultraviolet output.

Chandra sees only the thin X-ray tail of what these objects emit, a limit the paper’s spectral models make explicit. That gap is the main unresolved point: the 84 sources have a measured X-ray brightness and a modelled total output, but the ultraviolet energy that dominates their budget has been inferred from spectra, not observed directly, and none of the releases says which instrument would be used to close that gap.

This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.

More from Morning Overview