Stellar collision confirms theoretical predictions about the periodic table.
Gold, platinum, uranium and many of the rare-earth elements that are crucial to today’s high-tech gadgets are generated during the formation of black holes, astronomers have said. The collision of two small but dense stars simultaneously solved several cosmic mysteries, researchers announced at a press conference in Washington DC on 16 October. More than 30 papers have been published so far in five journals — Physical Review Letters, Science, Nature, Nature Astronomy and Astrophysical Journal Letters.
Astronomers watched as two neutron stars — small but very dense objects formed after the collapse of stars bigger than the Sun — collided and merged, forming a black hole, in a galaxy 40 million parsecs (130 million light years) away, according to two dozen researchers interviewed by Nature’s News team.
The collision generated the strongest and longest-lasting gravitational-wave signal ever seen on Earth. And the visible-light signal generated during the collision closely matches predictions made in recent years by theoretical astrophysicists, who hold that many elements of the periodic table that are heavier than iron are formed as a result of such stellar collisions.
Neutron-star mergers are also thought to trigger previously mysterious short γ-ray bursts, a hypothesis that now also seems to have been confirmed.
Astronomers have good reasons to believe that they are looking at the same source of both the gravitational waves and the short γ-ray bursts, says Cole Miller, an astronomer at the University of Maryland in College Park, who was not involved in the research but who has seen some of the papers ahead of their publication.
The event was detected on Earth on 17 August, and triggered weeks of febrile, round-the-clock activity on all 7 continents, as more than 70 teams of researchers scrambled to observe the aftermath.
The collision was felt first as a space-time tremor by the Laser Interferometer Gravitational-wave Observatory (LIGO) in the United States and by its Italy-based counterpart Virgo, and seen seconds afterwards as a smattering of high-energy photons by NASA’s Fermi Gamma-ray Space Telescope.
Alerted by the LIGO–Virgo team, astronomers then raced to find and study what was seen as a bright object in the sky using telescopes big and small, famous and obscure, on land and in orbit, and spanning the spectrum of electromagnetic radiation, from radio waves to X-rays.
Cody Messick was at his home at 08:41 local time (12:41 ut) on 17 August when he first found out about the event. “I remember standing on my stairs and looking at my phone, thinking: ‘Wow!” he says. Messick, who is a physicist at Pennsylvania State University in University Park, belongs to a small team of LIGO first-responders who receive frequent automated alerts from the two interferometers, which are based in Livingston, Louisiana, and Hanford, Washington. Normally, LIGO’s algorithms flag a potential signal in real time only if both interferometers detect it. Messick was surprised, because the message on his smartphone mentioned a strong signal — but one seen only at the Hanford site.
Messick quickly got on a conference call with his team leader, Chad Hanna, also at Pennsylvania State, and other colleagues. Together, they examined the data online. The Hanford signal looked like a textbook example of the waveform of the gravitational waves emitted by two compact objects, each slightly more massive than the Sun, as they spiral into each other, he says. In particular, the waves lasted much longer — about 100 seconds — and had a higher pitch than the signals from the much more massive black-hole mergers that LIGO had previously detected.
When they looked at the data stream coming from Livingston, the LIGO researchers found a similar signal there as well, but one with a loud, spurious glitch towards the end. It was that anomaly that had caused the real-time-analysis software to ignore the signal, says David Shoemaker, a physicist at the Massachusetts Institute of Technology in Cambridge who is LIGO’s spokesperson.
Meanwhile, researchers received another alert: Fermi had detected a short γ-ray burst that had occurred 1.7 seconds after the gravitational waves had ended. Called GRB170817A, it was unusually faint for such a burst.
In Italy, another technical glitch had suspended the continuous stream of data normally sent out by Virgo. So it took another 40 minutes for researchers to realize that they, too, had a signal — albeit a faint one. It transpired that the waves had travelled close to one of the interferometer’s four blind spots, says Jo van den Brand, a physicist at the Vrije Universiteit Amsterdam and spokesperson for the Virgo Collaboration.
By 13:21 ut, 40 minutes after the event, the LIGO–Virgo team had decided to notify its roughly 70 follow-up partners — teams of astronomers on standby to look for related events using conventional telescopes.
Four and a half hours later, the team sent a second, much more useful alert. The timing of Virgo’s feeble signal had been sufficient for the LIGO-Virgo team to identify the source of the waves. It pointed to a region of the sky spanning an angle of just a few degrees, in the southern sky. They called the event GW170817, after the date it was detected.
Virgo had joined LIGO’s observation campaign only on 1 August, after a five-year shutdown for upgrades. And just three days before the event’s detection, on 14 August, LIGO and Virgo had made their first joint detection. It enabled them to rehearse the more precise identification of the patch of sky of interest. The event on 17 August enabled them to narrow it down even further. And the estimated distance was ten times closer to Earth than in the previous events. They could tell this because of how loud and persistent the waves were: it was the strongest signal LIGO had ever sensed. After the fact, Hanna’s team was able to extract a signal that lasted a full six minutes.
Together, the alerts from LIGO–Virgo and Fermi sent astronomers into a frenzied rush. Each team wanted to be first to spot the fireworks produced by a neutron-star merger. It was daytime on most of the world’s land mass, so teams began to formulate strategies for their nocturnal observations. They knew that, at that time of the year, the region to search was not far from the Sun. That left a window of observation of a couple of hours after dusk, before the region of sky would set below the horizon.
“We had a complicated, choreographed dance of telescopes that night,” says Iair Arcavi, an astrophysicist at the University of California, Santa Barbara, whose team made non-stop observations using the Las Cumbres Observatory, a worldwide network of robotic telescopes. It began by activating a number of telescopes in Chile.
The first person to see the event may have been Charles Kilpatrick, an astronomer at the University of California, Santa Cruz. He was part of a team that was scanning the sky with the more modest means of the single one-metre Swope Telescope in Chile. Like his competitors, Kilpatrick was closely watching the exposures one by one as they came out, comparing them with archival images of the same patch of sky. By the ninth exposure, he saw something very conspicuous in a galaxy called NGC 4993. “It looked exactly like a point source in this image that wasn’t in the reference image,” Kilpatrick says. The team named it SSS17a.
At least two other groups say they spotted the bright dot independently. They and other teams also made sure that there were no other plausible candidates within the search region. GW170817, GRB170817A and SSS17a really seemed to be three different messengers from the same source.
LIGO and Virgo lacked a sufficiently detailed signal of the final instants of the collision to be certain that the objects were neutron stars, Shoemaker says. From gravitational-wave data alone, they could have been two unusually small black holes. But the presence of visible light strongly suggested that at least one of the objects in the merger was a neutron star, he and other researchers say.
The group at the University of California, Santa Cruz, was also the first to measure the optical spectrum of SSS17a. On the first night, the dot was bright blue, says astronomer Ryan Foley, who led that effort. NASA’s Swift telescope also detected blue, as well as ultraviolet, light. But during the next few nights of observation, those colours faded away, and the object became more red, according to multiple teams.
Colliding neutron stars should spread debris — a mix of neutrons, but also some protons — in three ways, says Brian Metzger, a theoretical astrophysicist at Columbia University in New York City. First, they fling matter out from their outer layers during the final orbits. Then some matter gets squeezed out in the actual collision. Finally, as the two stars begin to collapse into a black hole, it forms an accretion disk of matter, some of which flies out instead of falling in.
Over the past decade or so, astrophysicists had come to believe that this was the most plausible mechanism to explain the abundance of the heavier elements of the periodic table1. The theory held that, overall, about 2% of the combined mass of the stars would escape the fate of the rest. Within one second of the collision, this material would have expanded to become a cloud tens of thousands of kilometres across, but still about as dense as the Sun. In this cauldron, protons and neutrons would immediately clump together to form neutron-heavy nuclei, which would then begin to decay radioactively. This radioactivity would keep the cloud glowing hot for several days, even as it reached the size of the Solar System. Within a million years, it would spread across an entire galaxy.
Metzger says that the switch from blue to red was just what he expected to see. His models suggest that nuclei in this early cloud would reach the masses of many of the elements beyond iron, although not the heaviest ones. This chemical composition would cause the cloud to glow blue.
But the real smoking gun for this model, the signatures of the formation of the heaviest elements, would be a cloud that glowed in the red and infrared. These would be elements forged in a separate wave of the explosion, probably the one coming from the accretion disk, says Metzger.
“We had predicted exactly what kind of red,” says Daniel Kasen, a theoretical astrophysicist at the University of California in Berkeley. Jennifer Barnes, another theorist then in Kasen’s team who is now at Columbia University, had run the supercomputer simulations that predicted the experimental signatures in 20132. “I had just finished my PhD thesis predicting what these things would look like,” she says.
Eleonora Troja, an astronomer at NASA Goddard Space Flight Center in Greenbelt, Maryland, was part of one of the first teams to use the Hubble Space Telescope to view the event. “The spectra were phenomenal,” she adds, and almost indistinguishable from the theoretical predictions. “You could clearly see the fingerprints of the metals that had formed.”
But Troja and other observers were also puzzled, because they couldn't find any signal in the X-ray and radio regions of the spectrum. These would be expected during the formation of a black hole, which is thought to shoot jets of out of its poles at close to the speed of light. Nine days later, Troja’s team was the first to find the X-rays.
Alessandra Corsi, an astronomer at Texas Tech University in Lubbock, and her collaborators kept looking for radio emissions using the Very Large Array in New Mexico. Day after day, the dishes recorded nothing. “It turned out we had to wait 16 very long days in order to see the first radio glow,” she says.
The late onset of the radio and X-ray signals, together with the weakness of the initial γ-rays, suggest that the jets were pointed away from the line of sight to Earth. Gamma-ray bursts that happen to be pointed in the right direction can look very bright even from billions of parsecs away.
After a few weeks, most observatories had to stop looking at the object, because that part of the sky had got too close to the Sun. But radio telescopes are still tracking it to this day, Corsi says. More discoveries might yet be made.
“The idea that all this stuff has happened, it’s too much. It is just hard to process,” says Daniel Holz at the University of Chicago in Illinois. “It’s unreasonable that we have done so much with just one event of its kind.”
“All our hopes and dreams have basically come true,” says Jocelyn Read, an astrophysicist at California State University, Fullerton. “All this time we have been saying, look at this amazing thing we are going to be able to see. And it is still hard to believe when it actually happens.”
Metzger, B. D. et al. Mon. Not. R. Astron. Soc. 406, 2650–2662 (2010).
Kasen, D., Badnell, N. R. & Barnes, J. Astrophys. J. 774, 25 (2013).
Additional reporting by Alexandra Witze. Also see the related News & Views.
Cosmic furnaceA simulation of the merger of two neutron stars, leading to the formation of a black hole. About 2% of the stars' mass gets ejected at high speed, producing radioactive, heavy atoms. Credit: W. Kastaun/T. Kawamura/B. Giacomazzo/R. Ciolfi/A. Endrizzi via Caltech
Related links in Nature Research
About this article
Cite this article
Castelvecchi, D. Colliding stars spark rush to solve cosmic mysteries. Nature 550, 309–310 (2017). https://doi.org/10.1038/550309a