Astronomers may just have watched a star collapse directly into a black hole — minus the supernova. The failed supernova could help us understand how stars die.

Hubble observations of failed supernova
This pair of Hubble Space Telescope photos shows the massive star before and after it vanished, presumably by imploding to form a black hole. The left image shows the 25-solar-mass star as it appeared in 2007. But then it seemed to vanish, as seen in the right panel image from 2015. A faint infrared glow still emanates from where the star used to be. This radiation might come from debris falling onto a black hole.
NASA / ESA / C. Kochanek (OSU)

Just two weeks ago, Sky & Telescope reported that amateur astronomers had discovered a massive star exploding in the so-called Fireworks Galaxy (NGC 6946), known for its abundance of supernovae.

Yet years before that a quieter, even more extraordinary event had taken place in the same galaxy. Another massive star had brightened briefly in a faint mimicry of a supernova before it disappeared — presumably, into a black hole.

Scott Adams (Caltech and Ohio State University), Christopher Kochanek, Jill Gerke (both at Ohio State University), and colleagues caught the rare event using the Large Binocular Telescope in Arizona. They were using the pair of 8.4-meter telescopes to monitor a million aging stars in nearby galaxies, waiting for the stars to go pop.

Finally, after four years of watching and waiting, the star known as N6946-BH1 did exactly that — but not in the way you might expect.

Failed Supernova: How to Not Explode a Star

An ordinary massive star, one more than eight times as massive as the Sun, ends its life with a bang. Its heavy, inert core gives way to gravity — becoming either a neutron star or black hole — and the shock of that sudden collapse sends the star’s outer layers flying.

Yet theorists actually have a hard time making a simulated star explode. The problem is, the rebound shock tends to stall, halted by the flow of material onto the crushed core. Only recently have computer simulations come close to detonating a star, but even those incredibly detailed simulations were only 10 milliseconds long. No one has yet modeled a supernova shock wave breaking completely out of a star — those kinds of simulations are at least a couple of years away.

From the simulations we have now, though, it appears that some stars, those whose centers are too compact, will never explode. “The ‘explodability’ of a star seems to be dependent on the density of the layers just outside of the iron core,” Adams explains. When too-dense stars run out of elements to fuse to stave off gravity’s pull, they’ll just collapse.

The First “Failed Supernova” Candidate

Supernovae are rare, and disappearing stars may be even rarer. So Kochanek and colleagues started keeping an eye on a million aging stars in 27 nearby galaxies with the hopes of catching just one on the brink of collapse.

And by 2016, they had found seven. Six of these were standard core-collapse supernovae. But one was something else. The star, about 25 times more massive than the Sun and 22 million light-years away in NGC 6946, had simply winked out of existence.

Now, Adams and colleagues are reporting archival and follow-up observations of this disappearing act in the Monthly Notices of the Royal Astronomical Society (full text here). New analysis of archived Large Binocular Telescope observations show that the star’s light stayed roughly constant over a period of ten years before it changed suddenly In 2009, brightening to become 1 million times more luminous than our sun for several months. While bright, this outburst was nowhere near supernova levels. Within several months, the star’s visible light disappeared, and new observations from the Hubble Space Telescope confirm it’s gone.

But Spitzer Space Telescope images reveal a faint infrared glow — there’s still something there. The authors argue that the glow could come from stellar debris falling back onto the newly formed black hole, heating up and radiating before disappearing into the maw.

Failed Supernova
This illustration shows the final stages in the life of a massive star that fails to explode as a supernova but instead implodes under gravity to form a black hole. From left to right: the massive star has evolved to a red supergiant, the envelope of the star is ejected and expands, producing a cold, red transient source surrounding the newly formed black hole. Some residual material might fall onto the black hole, as illustrated by the stream and the disk, potentially powering visible light and infrared radiation years after the collapse.
NASA / ESA / P. Jeffries (STScI)

Still, it’s too early to definitively rule out alternatives. The infrared glow could instead emanate from dust enshrouding a surviving star, perhaps one that just experienced an extreme outburst.

The team now hopes to follow up with more Hubble and Spitzer images to keep an eye on the infrared radiation. If it starts to brighten, then we might be seeing a surviving star emerging from an expanding cocoon of dust. But if the infrared glow continues to fade, then we’re more likely seeing stellar remnants feeding the black hole. If there’s not too much debris in the way, detection with the Chandra X-ray Observatory could confirm the latter explanation.

“The longer we don't see any rebrightening at the location of the progenitor star,” Adams says, “the more implausible alternative explanations become.”

The find, if confirmed, is an important one, says Philipp Mösta (University of California, Berkeley), because it explains how massive black holes, like the ones detected by LIGO last year, came to be. "This will be an exciting target to follow in the next couple of years."


Image of Anthony Barreiro

Anthony Barreiro

June 1, 2017 at 6:29 pm

How confident are we that supernovae are more numerous than massive stars that collapse without exploding? Seven is a pretty small sample. It's a lot easier to see a supernova than to notice that a single star in another galaxy isn't there any more.

You must be logged in to post a comment.

Image of John-Murrell


August 19, 2020 at 6:31 am

My view is we have no idea ! I am not sure if the long term surveys that are operating look for stars that have disappeared. I am pretty sure the Gaia Transient Alerts will not show this as if they triggered on every star that the satellite did not get a measurement for there would b lots of false alerts. When GAIA DR3 photometry is published there a worthwhile project would be to look at the photometry for stars that no longer exist. That would increase the number of stars monitored but over a relatively short baseline. One would have to be certain that DR3 contains stars that 'no longer exist', that would require a consultation with those in the data preparation consortium who understand what would happen to the data for a star where the readings had stopped part way through the calibration process. NGC6946-BH1 was both too long ago and also probably too faint to be seen by Gaia.

You must be logged in to post a comment.

You must be logged in to post a comment.