A massive star roughly 2.5 million light-years away in the Andromeda Galaxy has quietly disappeared, and the best explanation is that it collapsed directly into a black hole without producing a supernova explosion. The object, designated M31-2014-DS1, was first flagged through an infrared brightening detected in 2014, and over the following years its total light output dropped to a fraction of what it once was. This event represents one of the strongest cases yet for a “failed supernova,” a theoretical outcome where a dying star’s core gives way to gravity so completely that almost nothing escapes, not even the spectacular blast astronomers normally associate with stellar death.
Because the star sits in the nearby Andromeda Galaxy, also known as M31, astronomers have an unusually clear view of its environment compared with more distant events. Before the fade-out, archival images showed a luminous, roughly 13-solar-mass star shining steadily in visible light. After the mid-infrared flare, that steady glow dwindled until only a faint remnant remained. The emerging picture is of a star that shed a small amount of material, briefly heated its surroundings, and then collapsed so completely that almost all of its mass disappeared behind an event horizon. For researchers trying to understand how often stars die this way, M31-2014-DS1 offers a rare, well-documented case study rather than a single ambiguous snapshot.
How NEOWISE Caught a Star Fading Away
The discovery owes its existence to archival data rather than real-time observation. NASA’s NEOWISE space telescope, originally designed to hunt near-Earth asteroids, has spent years surveying the sky in infrared wavelengths. Researchers combing through that archive spotted a mid-infrared brightening in 2014 at a location in Andromeda where a luminous star had been cataloged. That infrared flare suggested the star had ejected a shell of material that heated surrounding dust, a behavior sometimes seen in stars nearing the end of their lives. What made this case unusual was what came next: instead of rebounding or detonating, the star simply got dimmer.
According to an initial analysis on preprint servers, the object’s total luminosity held roughly constant for about 1,000 days after the infrared flare, then entered a steep decline over the following 1,000 days. By around 2023, the star had faded so dramatically in visible light that ground-based and space-based optical telescopes could barely detect it. The timeline is telling. A normal supernova would have produced a weeks-long optical peak bright enough to spot across intergalactic distances, followed by a characteristic radioactive decay curve. M31-2014-DS1 did none of that. It brightened only in infrared, plateaued, and then slowly winked out, leaving behind no obvious expanding supernova remnant.
JWST and Chandra Fill in the Details
A follow-up study added critical new data from two of the most powerful observatories in operation. The James Webb Space Telescope provided mid-infrared spectroscopy and photometry of the remnant, while the Chandra X-ray Observatory placed constraints on any X-ray emission from the site. The JWST data revealed molecular absorption features including CO, CO2, H2O, and SO2 in an expanding shell of gas, consistent with material that had been weakly ejected during or just before the collapse. The expanding gas mass was estimated at roughly 0.1 solar masses, a tiny fraction of the original star, moving outward through a dust shell spanning approximately 40 to 200 astronomical units.
The bolometric luminosity of whatever remains at the site had by this point dropped to about 7 to 8 percent of the progenitor star’s output. That continued fading is hard to reconcile with a surviving star hidden behind a thick dust curtain. If the star were still intact and merely obscured, its energy would still need to go somewhere, reprocessed into infrared radiation that JWST would detect. Instead, the spectral energy distribution points toward a central object that is no longer producing significant nuclear fusion energy. The Chandra observations did not detect clear X-ray accretion signatures, which means the presumed black hole is not actively feeding on nearby material at a rate high enough to glow in X-rays. That absence is itself informative: it suggests the collapse was so efficient that very little matter was left in close orbit to spiral inward and radiate.
The N6946-BH1 Precedent
M31-2014-DS1 is not the first star suspected of vanishing this way. The earlier and better-known candidate is N6946-BH1, a star in the galaxy NGC 6946 with an estimated mass of roughly 25 solar masses. That object also brightened briefly and then faded without producing a supernova, drawing attention from Hubble, Spitzer, and the Large Binocular Telescope over multiple years of follow-up. A dedicated technical study using all three instruments reported quantitative upper limits on any optical rebrightening and continued fading in near and mid-infrared bands. The authors argued that spectral energy distribution models were inconsistent with a surviving star hidden by dust or stellar wind, keeping the failed-supernova interpretation on the table.
The comparison between the two objects is instructive but imperfect. N6946-BH1’s progenitor was nearly twice as massive as M31-2014-DS1’s estimated 13 solar masses, yet both appear to have met the same fate. If failed supernovae can occur across a wide range of stellar masses, the fraction of massive stars that die quietly could be larger than many models currently predict. That has direct consequences for how astronomers count black holes in the universe and for the expected rate of gravitational-wave events from black hole mergers. Every star that collapses without exploding is one that adds a black hole to the cosmic census without leaving behind the telltale expanding nebula that makes supernova remnants easy to find.
Why the Quiet Death Matters
Standard stellar evolution theory has long predicted that some massive stars should fail to explode. The basic physics is straightforward: if the collapsing core is massive enough, the shock wave that would normally rip the star apart can stall and fall back, producing a black hole instead of a neutron star. But observational evidence for this channel has been scarce, partly because a non-event is inherently harder to detect than an explosion. The M31-2014-DS1 case strengthens the argument precisely because it was caught in the act through years of infrared monitoring, then confirmed with spectroscopy that revealed the chemical fingerprints of cold, expanding gas rather than a hot, radioactive debris cloud. In other words, astronomers are starting to see not just that some stars vanish, but how they do so and what they leave behind.
These quiet deaths also help close the gap between theory and observation in black hole demographics. Population models of massive stars and their fates must account for the number of stellar-mass black holes implied by gravitational-wave detections and X-ray binaries. If too few stars explode as visible supernovae, the missing black holes must come from another channel. Failed supernovae like M31-2014-DS1 and N6946-BH1 offer a plausible route: stars that collapse almost directly into black holes, with only a modest burp of ejected gas to mark their passing. As more surveys accumulate long-term infrared light curves of nearby galaxies, astronomers expect to find additional candidates, gradually turning a handful of curiosities into a statistically meaningful population.
Behind the Data: Preprints, Archives, and Community Effort
Events like M31-2014-DS1 highlight not only the power of modern telescopes but also the infrastructure that lets scientists share and scrutinize results quickly. Both the initial light-curve analysis and the later JWST and Chandra study were posted as preprints on arXiv, a repository maintained by a network of institutional member organizations that collectively support its operation. By making manuscripts freely accessible before journal publication, arXiv allows researchers worldwide to test models, compare datasets, and plan follow-up observations while the evidence is still fresh.
That open ecosystem depends on more than just institutional backing. Individual scientists and the public can help sustain preprint services through voluntary donations, while detailed documentation guides authors and readers on how to submit, search, and interpret the material hosted there. In the case of failed supernova candidates, this rapid, open exchange has been crucial. Teams studying N6946-BH1 could immediately compare their findings with those emerging for M31-2014-DS1, refining criteria for what counts as a convincing disappearance and what might instead be a dust-obscured survivor. As more archival surveys and next-generation observatories come online, that combination of deep data and open sharing will be essential for catching more stars in the act of quietly becoming black holes, and for understanding what their silence tells us about the life cycle of the cosmos.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
