Last Friday we celebrated the one-year anniversary of an event that those of us who were involved will never forget. The Virgo gravitational-wave detector had joined the two LIGO instruments on August 1, 2017, and the three detectors had since then been patiently listening out together for gravitational wave sounds coming from anywhere in the Universe. On August 17, the deep quiet was interrupted by a squeal, a chirp lasting much longer and going to a much higher pitch than the GW150914 chirp that had launched the field of gravitational wave observational astronomy two years earlier. We named it, prosaically, GW170817.
This one-minute-long squeal was followed by an incredible explosion that radiated intense gamma-rays, X-rays, light, radio waves — right across the whole electromagnetic spectrum. What came first was a burst of gamma-rays, just 2 seconds after the end of the squeal. Then it began brightening up at other wavelengths. The explosion itself did not register in LIGO and Virgo, because as it rushed out in all directions it was too smooth to generate gravitational waves. But astronomers at their telescopes saw it: a kilonova, a new type of cosmic explosion.
You may like to think of the chirp as the sound of a fuse, the fuse that triggered a giant explosion. The chirping sound was produced by two neutron stars orbiting one another at nearly the speed of light. The pitch of this gravitational-wave sound is at twice the frequency their orbit. Their orbital motion stirs up gravity near them, and this creates gravitational waves. These waves take away energy, and this friction brings the stars closer together. The binary that became GW170817 had probably formed billions of years earlier, with widely separated stars. At first the weak gravitational waves carried away energy so slowly that the system hardly changed from one millennium to the next. But eventually, inevitably, the stars drew so close to one another that the waves had frequencies of around 20-30 Hz, and they entered the observation frequency band of LIGO and Virgo. The emitted waves were very strong by then, and so the system was evolving quickly, shrinking noticeably with every orbit, the gravitational wave pitch rising rapidly. Within another minute or so, the stars made contact, triggering the kilonova. I mentioned in my post of June 9 last year that we were hoping to detect this kind of event with neutron stars, and ten weeks later it happened.
You can listen to the reconstructed sound of this death-spiral, this fizzling fuse, in the video here, although you need patience, because for most of the one minute its pitch is too low for normal hearing. LIGO and Virgo can go to deeper pitches than the human ear. But the end is worth waiting for, because of what it says about the stars: as the pitch approaches about 1 kHz (two octaves above middle C), the stars are orbiting one another 500 times each second!
What blew up after the chirp ended?
Previous gravitational wave detections had also recorded chirps, but these fuses had finished without igniting explosions. That is because they were the gravitational wave sounds made by orbiting black holes, which are strongly curved but empty space. So when they came together, they formed a bigger black hole, but it was still just empty space. There was nothing to explode. But last year, the things that came together were neutron stars, and they smacked together at something like 20% of the speed of light. Perfect for creating an enormous splash.
Neutron stars are not ordinary stars like our Sun. They are stars formed from the remains of a normal star that has run out of its nuclear fuel, so it can no longer maintain its original size against the pull of gravity. Astronomers know about two kinds of such “remnant” stars: white dwarfs and neutron stars. Our Sun will form a white dwarf in the end (some 5 billion years from now). It will shrink down to the size of the Earth after slowly boiling away maybe half of its gas (and incinerating the Earth in the process). Much bigger stars than the Sun end much more dramatically in what we call supernova explosions, and these usually leave behind a neutron star, which is typically a little heavier than our Sun but is only the size of a big city, about 20 km across. Neutron stars are sometimes described as “dead” stars, or as “cinders” left by the supernova, but in reality they are balls of tightly packed explosive, just waiting for something else to set them off.
What kind of explosive material is a neutron star made of?
Some of the densest things we normally deal with in everyday life are made up of atoms that fit very tightly together, packed right up against one another: we call them crystals, like diamond. But a diamond the size of a city (what an image!) will not have anything like the mass of the Sun; it will only have roughly the mass of a city. So if you want to pack all 1057 of the Sun’s atoms into a ball the size of a city, there will be no room for the atoms to be ordinary atoms. Instead, they get squashed together until their nuclei bump together and merge — the neutron star becomes basically one big atomic nucleus, an element with atomic weight 1057.
Now, a nucleus is unstable if its atomic weight is more than 240 or so, i.e. above uranium. So a nucleus with atomic weight 1057 is not something that would normally stay together for more than a fraction of a second, let alone millions of years. It is intrinsically very unstable, very unhappy being so big and dense. It only stays together as a star because of the tremendous force of gravity holding it in.
To get a feeling for how strong gravity is on a neutron star, imagine trying to stand on one. Your body would weigh several trillion kilograms — except that it wouldn’t, because that much weight would squash you, and gravity would just merge the few atoms that are you into the vast pool of nuclei already there. In fact, the whole neutron star is pretty close to being pulled in too much by gravity — it is just on the edge of collapsing into a black hole. It manages to resist this pull because the nuclear matter hates being compressed so tightly. The matter pushes back, and this just barely manages to keep the star in balance.
So how do neutron stars explode?
Given this tricky balance, neutron stars are okay if they are left alone, or if a bit of stuff just gradually falls onto them. But smashing two of them together at close to the speed of light is a different story. That will knock a lot of the stuff off of the surface; and once that material is freed from its confinement, it goes wild. All the energy that was put into compressing it in the first place is now available to help it escape the neutron star’s gravity.
The result is a huge kilonova explosion, a cloud of matter expanding away from the site of the collision initially at close to the speed of light, incredibly hot material, full of twisted magnetic fields, giving off radiation across the spectrum. It is a display to entertain astronomers of every persuasion. When GW170817 happened, it is said that something like 10% of all the world’s professional astronomers rushed to their telescopes.
But in fact only part of the neutron star stuff exploded away. When the stars collided, their combined gravity actually got stronger, because there was now nearly twice as much mass. This immense gravity squashed the two stars into one, holding onto most of the original matter. And although a single neutron star can generate enough pressure inside itself to resist collapse, a neutron star with twice the normal mass probably can’t. So the final merged remnant in GW170817 most likely collapsed to a single black hole.
The gamma-ray burst that happened just 2 seconds after the stars merged would have been triggered by this collapse. A beam of this radiation pierced through the already expanding cloud of escaping matter and announced the explosion to us. The material that blew off was the lucky stuff: the rest of our two neutron stars has fallen into the black hole that the stars themselves created, disappearing forever.
Why did everyone get so excited by a couple of neutron stars colliding, so far away?
Normally astronomers don’t know about it when an explosion like this happens. They may chance across evidence of it long afterwards, but by then it is too late. So when the gamma-ray and gravitational-wave astronomers quickly spread the word that this explosion was just starting, astronomers rushed to monitor it. In fact some of them are still watching it, using telescopes in the radio and X-ray parts of the spectrum.
One of the reasons for wanting to watch it happen is that kilonovas like this are rare, occurring maybe once every 100,000 years in any one galaxy. The gravitational waves from GW170817 told us that the explosion had happened in a galaxy 120 million light-years away. This may seem like an enormous distance, but for LIGO and Virgo this was exceptionally close: they could have seen it even if it had been three times more distant. And for astronomers this is also pretty close: even modest-sized professional telescopes could study it.
But the interest in GW170817 is not just its rarity. What is even more interesting is that neutron-star mergers have played a surprisingly important role in enabling life on Earth, and even for human evolution. This is because of what the stuff that is blown off in the kilonova turns into. It starts out, remember, as part of one big nucleus. As it expands away, its nuclear instability reasserts itself, and it breaks up into chunks. This is nuclear fission on a grand scale, and the kilonova event is nothing less than a gigantic atomic bomb, something like a 1035 megaton explosion.
The initial chunks are still huge, so they keep breaking up into smaller and smaller chunks, until the chunks are small enough to be stable atomic nuclei again. At this point the nuclear fission stops. The expanding cloud ends up made up of stable atoms, but its composition is strongly biased toward the heavier elements of the Periodic Table.
Now here is the important bit: astronomers have come to the conclusion that this is the main way that these heavy elements are made throughout the Universe. Most of the ordinary matter in the Universe (and in our Sun) is just hydrogen and helium, the two lightest elements. So essentially everything else has to have been made by stars. Astronomers’ observations of GW170817 have confirmed that the expanding cloud of gas was indeed heavily loaded with elements from the upper half of the Periodic Table.
GW170817 and the origin of life
You may remember the newspaper stories about GW170817 when it was made public in October last year; they enjoyed talking about how all the gold and platinum in our jewelry must have been made long long ago in another similar explosion.
Now, although the extraordinary history of the gold in our rings is truly awe-inspiring, this actually misses the most important point. At least two of these heavy elements are essential to human life: molybdenum and iodine. As shown in the figure, iodine is produced almost exclusively from neutron-star binary mergers, and molybdenum (involved among other things in producing the body’s energy source ATP) comes from neutron stars and low-mass stars in roughly equal amounts.
And then there are thorium and uranium, which may have played a key role in the evolution of life itself. That is because the gradual decay of the long-lived uranium and thorium isotopes in the Earth’s mantel provides roughly half of the heat that, in flowing outwards through the solid Earth, drives volcanism and plate tectonics. In particular, the movements of the continents seem to need the extra heat generated by these ultra-heavy elements. These geological phenomena have played an important role in the evolution of life on Earth, by challenging species with a constantly changing environment that drives continued evolution. It does not seem unreasonable to speculate that, if the Earth had been geologically quiet, some well-adapted distant ancestor of ours might just never have needed to change, and evolution might have slowed dramatically or even stopped.
The presence in our planet of all the elements in the heaviest half of the Periodic Table tells us that, sometime before the Sun formed 5 billion years ago, a neutron-star merger — driven by the loss of orbital energy to gravitational waves — happened in our own Milky Way galaxy, triggering its version the GW170817 kilonova. Its expanding cloud of heavy elements polluted a cloud of hydrogen and helium gas, and that cloud later collapsed to form the Sun. This gas cloud had also evidently been polluted with the lighter elements of the Periodic Table that had been formed in other much more frequent nearby events, such as in supernova explosions or in the gas blown away at the end of the life of stars like our Sun. The expanding cloud of a supernova or of the kilonova may even have triggered the collapse of the polluted gas cloud, which led to the formation of the Sun, and hence of the planets, and eventually of living things.
We are, as has often been remarked, stardust. But there might just be one particular long-ago gravitational-wave-driven neutron star merger, probably very like GW170817, to which we owe our very lives.
I mentioned here but did not explain how the gravitational wave signal told us how far away the source was. This is a whole story in itself, and I will come back to it in a future post. It led to a milestone in gravitational-wave astronomy when we were able to use GW170817 to make a measurement of the expansion rate of the Universe, the Hubble constant.
6 thoughts on “The merger of two neutron stars, one year on: GW170817”
Hi Professor Schutz,
I am really enjoying your blog. Thank you so much for it.
You almost certainly won’t remember this, but many years ago, around 1987 if memory serves, you came into the Dixons store in Queen Street, Cardiff, where I was working, and we had a chat about what would become LIGO. I was so inspired by that conversation with you that I have been avidly following gravitational-wave research ever since, and I am so incredibly excited about all the recent discoveries. How lucky we are to live in such an era of discovery…and thank you for your part in making it all happen. I actually gave a lecture to my colleagues recently about LIGO (I do a monthly after-work astronomy-related lecture which I call “Tales from the Cosmic Zoo”) and they were amazed.
Well, thanks again, I love the blog, and let’s look forward, after LIGO’s hiatus, to yet more amazing discoveries!
With my very best wishes
Thank you Andy for you kind words. You are right that I don’t recall the chat we had, but I am really pleased that you do and that it planted the interest in gravitational waves in your mind. Do keep up your good work of telling people about astronomy!
Thanks…I will certainly try!
I have a couple of questions about GW170817, actually…concerning the timeline of events.
1. Was the black hole produced instantaneously during the merger (as logic would seem to dictate)? If not, how long after the moment of merging? I realise this may be a difficult question to answer…..is it something that could be detected in some way?
2. How long after the black hole was formed did it generate its jets? As I understand it, the jets are formed from material surrounding the black hole being funnelled towards the poles and ejected/focussed by magnetic fields, so presumably this would take a while after the hole was formed.
Any help you could give me with this timeline would be much appreciated. It’s for an article I’m writing for Astronomy Ireland magazine.
Many thanks, and it really is great to talk to you again after all these years!
All the best
These are hard questions to answer and many people would like to know the answers! A guess is that the gamma burst that happened 1.7 sec after the merger was triggered by the formation of the black hole. Some numerical simulations suggest that the magnetic field is too chaotic to channel a jet until the black hole forms, at which point it settles into a strongly polar form. Although 1.7 seconds seems like a short time, remember that the dynamical timescale, say one orbital time just before merger, was about 2 ms. So the meta stable merged neutron star had to support itself for about 1000 dynamical times before collapsing. Our measure of its mass is very uncertain but it could have been only slightly above the maximum mass of a non-rotating neutron star. Then rotation or magnetic fields could have given the slight extra support needed. But rotation decreased as angular momentum was taken away by the ejected matter and by the magnetic field anchored in that matter. And maybe the magnetic field support decreased too over time. So the picture is plausible but we don’t have good enough measurements to be sure!
Many thanks. I thought they would be difficult questions to answer, so I do appreciate it. That has really helped me with my article. I might ask some more questions about this subject soon if you don’t mind. Thank you so much…and keep up the superb work with your blog!
Everything is very open with a very clear explanation of the challenges. It was really informative. Your site is useful. Thank you for sharing!|