Tech & Science · September 12, 2026
NASA finds 84 hidden X-ray sources that may solve two cosmic mysteries
A Chandra archive search has exposed a population of unusually cool, extremely luminous objects in six galaxies. Their light could point toward long-sought Type Ia supernova progenitors — and toward a missing source of the ultraviolet energy that shapes galactic gas.
Some of the most energetic objects in nearby galaxies may have been hiding in one of astronomy’s least convenient places: the narrow frontier between extreme ultraviolet light and very soft X-rays. A study published September 9 in Nature Astronomy, using observations from NASA’s Chandra X-ray Observatory, reports a distinct population that is bright at the bottom edge of Chandra’s energy range and then almost disappears at higher X-ray energies. The researchers call them hypersoft X-ray sources, or HSSs.
The result is notable because the team did not uncover a single unusual star or one exotic outlier. It found 84 sources across six galaxies. Two of those galaxies are spirals — Andromeda, also known as M31, and the Pinwheel Galaxy, M101 — while four are ellipticals. The sources turn up in both younger, star-forming environments and older stellar populations. That breadth is one reason the authors do not argue that every HSS must be the same kind of system. Instead, the new category may be a spectral signature shared by several kinds of compact binary systems operating in an observational blind spot.
The central idea is simple even if the astrophysics is not: these objects radiate most strongly at energies that are difficult to see. Typical X-ray binaries are easiest to recognize above about 0.3 kilo-electron volts. The newly reported sources are detected primarily or exclusively below that threshold. Spectral modeling indicates that much of their total output may emerge at still lower energies in the extreme ultraviolet, a band that interstellar gas readily absorbs. In effect, the sources can be intrinsically brilliant while looking strangely faint to telescopes tuned to more familiar X-ray bands.
What “hypersoft” means here
The name describes the spectrum, not a gentle or low-power object. The published sample is unusually weighted toward photons below 0.3 keV. The brightest members can still reach roughly 1038 ergs per second in the narrow X-ray band, while their total energy output may be greater once the unseen extreme-ultraviolet component is included.
The discovery
A population emerged when astronomers looked at the bottom of the X-ray band
Lead author Mustafa Muhibullah of the University of Alabama, working with Jimmy A. Irwin and Rosanne Di Stefano of the Center for Astrophysics | Harvard & Smithsonian, searched public Chandra observations for point-like sources that behaved differently from familiar X-ray binaries. The decisive signature was not merely that a source looked soft. It had to be dramatically stronger in Chandra’s lowest-energy images and weak or absent when the same field was examined at higher X-ray energies.
The blind spot
The universe is bright in a band that is hard to observe
The electromagnetic spectrum does not contain convenient borders chosen for astronomers. Extreme ultraviolet, or EUV, blends into the softest X-rays. Unfortunately, neutral hydrogen and helium are very effective at absorbing energetic ultraviolet photons. Even modest columns of gas between a source and Earth can erase much of the signal. That is why a system whose energy output peaks in the EUV can be astrophysically powerful yet undercounted in surveys.
Chandra is not an EUV telescope, but its sensitivity extends low enough in X-ray energy to catch the high-energy tail of such a source. The new work turns that edge response into a search strategy. If an object is conspicuous between roughly 0.15 and 0.3 keV and then collapses in brightness between 0.3 and 1 keV, the spectrum is telling astronomers that the unseen peak may lie below the conventional X-ray window.
The Nature Astronomy paper reports that the most luminous examples reach around 1038 ergs per second in the narrow low-energy X-ray band. That level is not the signature of a minor stellar flare. It points toward accretion: matter falling onto a compact object and converting gravitational energy into heat and radiation. The question is which compact object, under what physical conditions, and for how long.
A dense stellar remnant pulling hydrogen-rich material from a companion could burn or radiate at very soft energies. Some such systems are candidates in the long-running search for Type Ia supernova progenitors.
Accretion onto an ultradense remnant can generate luminous emission with complex spectra. The new class may include systems whose geometry or temperature pushes most detectable photons unusually low.
An accreting stellar-mass black hole can also produce very soft thermal states. The authors leave room for a mixed population rather than assigning all 84 sources to one engine.
Possible engines
One spectral class may hide several kinds of binary systems
A compact object by itself does not make an HSS. The leading scenarios involve a binary partner feeding it. Gas pulled from a companion star forms a stream or disk, heats as it falls inward and releases radiation before reaching the surface of a white dwarf or neutron star, or disappearing through a black hole’s event horizon. Small changes in temperature, accretion rate, composition and geometry can shift where most of that energy emerges in the spectrum.
Mystery one
Could some of them be Type Ia supernovae before the explosion?
Type Ia supernovae are among the most important explosions in modern cosmology. Their calibrated brightness allows astronomers to estimate distances across the universe, and observations of these events helped establish that cosmic expansion is accelerating. Yet the stellar systems that produce them remain an active field of research. Astronomers know a white dwarf is involved, but multiple evolutionary channels may lead to the thermonuclear runaway.
The wording has to stay careful. The study does not identify 84 imminent supernovae, nor does it show that HSSs are a single precursor stage. The authors propose that accreting white dwarfs and post-nova systems are among the physical classes that can produce the observed spectra. The scientific value lies in converting a theoretical possibility into a searchable population. Future work can ask which HSSs have the right temperatures, luminosities, variability and companion-star properties to fit specific supernova models.
A new window does not solve the Type Ia progenitor problem by itself. It gives astronomers a new population in which the missing systems might be hiding.
Mystery two
They may help explain who keeps galactic gas ionized
The second mystery is less famous outside astrophysics but just as consequential for understanding how galaxies work. Gas between stars can be ionized — stripped of electrons — when it absorbs sufficiently energetic radiation. Hot, massive stars provide a major supply of ionizing photons in star-forming galaxies. But in some environments, especially where the stellar population is older, the observed state of the gas is difficult to explain with young stars alone.
Soft X-rays, location and variability
Chandra can identify the low-energy tail, establish whether a source is point-like, compare it across epochs and place it within the stellar environment of its host galaxy.
The hidden EUV power
Most of the modeled luminosity may emerge where interstellar absorption is strongest, so researchers must infer the unseen component from the accessible X-ray spectrum and physical models.
The M101 clue
The Pinwheel Galaxy shows why a population study matters
M101 is a nearby face-on spiral galaxy popular with professional and amateur astronomers because its structure is easy to see. For the HSS study, it offers something more useful than beauty: a large disk containing diverse stellar neighborhoods and a deep Chandra observing history. NASA’s Chandra team notes that the observations used for the public M101 composite span 25 pointings from 2000 to 2005 and total about 274 hours of exposure.
Why archive science matters
The discovery was hiding in public data, not waiting for a new telescope
There is a tendency to equate astronomical progress with a new launch. New instruments are crucial, but this study illustrates another route: ask a new question of old data. Chandra’s archive contains observations optimized for many original purposes — black holes, supernova remnants, hot gas, star-forming regions, galaxy clusters and more. Once those observations are calibrated and preserved, later researchers can combine them into a survey no single proposal could have assembled.
What the paper does not claim
A “new class” is a starting point, not a final identity card
The most important caution is categorical. “Hypersoft X-ray source” describes an observed spectrum. It does not yet tell astronomers with certainty what object sits at the center of every system. The paper explicitly allows several physical classes, including accreting white dwarfs, post-nova systems and black-hole binaries. That diversity could be a feature rather than a flaw: the same radiative state may emerge from different compact objects under certain conditions.
The second caution concerns completeness. An object that is easiest to detect at very low X-ray energies is also unusually sensitive to foreground absorption and instrument response. The 84-source catalog is therefore not a simple count of all HSSs in the surveyed galaxies. It is a detected sample shaped by exposure time, line-of-sight gas, distance and the performance of the telescope at the low-energy edge.
Do individual sources remain persistently hypersoft, cycle through states, disappear after nova eruptions or flare in ways that separate white-dwarf systems from black-hole binaries?
Deep optical and infrared observations can search for companion stars, star clusters or nova remnants at the same positions, narrowing the physical interpretation.
Longer X-ray exposures and better low-energy spectral modeling can constrain temperature, absorbing material and total luminosity below the directly observed band.
Comparing star-forming disks, old bulges and elliptical galaxies can reveal whether one evolutionary channel dominates in particular stellar populations.
What comes next
The fastest progress may come from combining wavelengths
Because the defining energy peak is partly hidden, no single telescope is likely to provide the full answer. Chandra can isolate the low-energy X-ray tail and precisely locate a source. Hubble-class optical imaging can resolve crowded stellar fields and identify possible counterparts. Infrared observations can probe cool companions and dust. Ultraviolet data, where available, can test how steeply the spectrum rises toward the absorbed EUV band. Radio observations may reveal jets or other signs of accretion in black-hole systems.
The larger meaning
A discovery at the edge of an instrument can redraw the center of a problem
The appealing paradox of hypersoft X-ray sources is that they are both luminous and concealed. Their physics may be extreme enough to matter for supernova explosions and galactic ionization, yet the universe places much of their light behind an absorbing curtain. The Chandra work does not remove that curtain. It finds a way to recognize the glow leaking around its edge.
For now, the most defensible conclusion is also the most exciting: an observational blind spot was large enough to hide dozens of powerful sources in galaxies that astronomers have studied for decades. The data were already there. A different way of sorting the photons made the population visible.
Reader questions
Hypersoft X-ray sources, explained
Are these 84 objects a completely new kind of star?
Not necessarily. They are a newly defined observational class. The underlying systems may include more than one known type of compact binary, such as accreting white dwarfs, neutron stars or black holes operating in an unusually soft spectral state.
Why can something be powerful but hard to detect?
Because most of its radiation may emerge in the extreme ultraviolet and very soft X-rays. Interstellar hydrogen and helium absorb those photons efficiently, so a luminous source can look weak or disappear at the energies most surveys use.
Did NASA prove these are Type Ia supernova progenitors?
Why does galactic gas ionization matter?
Ionization affects the temperature, chemistry and emission of gas between stars. It also influences how astronomers infer star-formation conditions and energy sources inside galaxies.
Was a new telescope needed for this discovery?
No. The team mined public archival observations from NASA’s Chandra X-ray Observatory. The result shows how long-lived mission archives can produce discoveries long after the original observations were taken.
Primary sources and further reading
- NASA Science — “NASA’s Chandra Unveils Mysterious X-Ray Objects,” September 9, 2026
- Nature Astronomy — Muhibullah, Irwin and Di Stefano, “Hypersoft X-ray sources as a low-energy class of luminous cosmic emitter,” published September 9, 2026
- Chandra X-ray Center — M101 release materials, observation details and visual descriptions
- Chandra Blog — lead-author overview of the extreme-ultraviolet observational blind spot
Comments
Post a Comment