Stellar Vanishing Act in Andromeda Confirms a Black Hole Born Without an Explosion

Black holes do not always announce themselves with violence. Theory has long allowed for a quieter path: a massive star whose core collapses without generating the explosive shock wave of a supernova, leaving behind a black hole and a slowly fading dust cloud. These events would be nearly impossible to detect, because they produce not a brilliant flash but a gradual disappearance - a star there one decade and gone the next.

The challenge of detecting this kind of event led astronomer Kishalay De and colleagues to search archival infrared observations for the distinctive signature theorists predicted: a brief brightening followed by a sustained, deep fade. What they found, published February 12 in Science, is now described as the clearest evidence yet for direct stellar collapse into a black hole.

The star that disappeared

M31-2014-DS1 was a hydrogen-depleted supergiant in the Andromeda Galaxy, the Milky Way's nearest large neighbor at about 2.5 million light-years away. In infrared data from NASA's NEOWISE mission, the star brightened starting around 2014 and sustained that increased brightness for roughly two years. Then, beginning in 2016, its luminosity collapsed - not in the explosive outward sense, but inward. By 2022, the star was effectively invisible in optical and near-infrared light. Follow-up observations with Hubble and ground-based facilities found only a faint remnant in mid-infrared wavelengths, heavily obscured by dust.

De and colleagues interpret the sequence as physically coherent with the failed supernova model. When the star's core collapsed, convective motions in its outer layers prevented a clean infall. Instead of falling directly into the nascent black hole, the outer envelope settled into an orbiting disc. As that material slowly accreted, it heated the surrounding dust shed by the star during its final years - producing the infrared brightening. As accretion slowed, the infrared signal faded, but very gradually, because only about 1 percent of the stellar envelope is estimated to fall in over the first decades.

What distinguishes this from a supernova

The observational distinction is stark. Supernovae outshine their host galaxies for weeks in visible light and are detected in enormous numbers. M31-2014-DS1 produced no optical transient detectable as a supernova. Its infrared brightening was modest by astrophysical standards, and its disappearance was slow. "Imagine if the star Betelgeuse suddenly disappeared," De said. "Everybody would lose their minds. The same kind of thing was happening with this star in the Andromeda Galaxy." The difference is that nobody noticed for years - the signal was buried in archival data waiting to be analyzed.

The De team found the event only after conducting what they described as the largest study of variable infrared sources ever attempted, systematically tracking every star in the Milky Way and nearby galaxies in archived NEOWISE data. M31-2014-DS1 fit theoretical predictions precisely: the brightening profile, duration, and subsequent fade matched what models of direct collapse and convective outer-envelope ejection would produce.

A second case and a pattern emerging

The analysis of M31-2014-DS1 also enabled reinterpretation of an earlier event. NGC 6946-BH1, observed around 2010 in a galaxy roughly 25 million light-years away, had been tentatively classified as a failed supernova but remained debated because it was 100 times fainter and the data were sparse. The physical framework now established for M31-2014-DS1 fits NGC 6946-BH1's behavior as well. Both appear to belong to the same category of objects.

Two events do not define a population. But they do confirm that the mechanism exists, is observable with current instruments, and was probably occurring throughout the universe undetected until systematic infrared surveys made it accessible. Future infrared surveys with higher sensitivity and cadence are likely to find more.