JWST shocks the world with colliding neutron star discovery

Where do the heaviest elements in the Universe come from? If you were like most astrophysicists during the 20th century, you might’ve said from supernova explosions: stellar cataclysms that occur either within the cores of massive stars or from stellar corpses (white dwarfs) that undergo destructive, energy-releasing events that trigger a rapid succession of nuclear fusion reactions. Unfortunately, a comprehensive study of these classes of events — including both type II (core-collapse) and type Ia (exploding white dwarf) supernovae — showed that, although they do produce large sets of fusion reactions, they really only produce elements up to about zirconium (element #40) on the periodic table.

Beyond that, or for more than half of the known elements that exist, a different set of processes are required. While the slow neutron capture process (s-process) can occur within evolved, Sun-like stars, accounting for large fractions of certain elements, such as niobium, tin, barium, and lead, the majority of heavy elements require another process to explain their observed abundances.

In 2017, a new candidate rose to the top of everyone’s mind: colliding neutron stars. A kilonova event, seen in both gravitational waves and conventional (electromagnetic radiation) forms of light, indicated a potential origin for most of the heavy elements: gold, platinum, radium, uranium and more. The lack of any similar events since that first, especially in gravitational waves, threw doubt on this scenario. But a remarkable JWST observation showed us that colliding neutron stars may be far more common than we thought. Here are the current thoughts on how the heaviest elements of all are produced.

The ultimate event for multi-messenger astronomy would be a merger of either two white dwarfs or two neutron stars that was close enough. If such an event occurred in near-enough proximity to Earth, neutrinos, light, and gravitational waves could all be detected. Although we’ve detected kilonovae events in both gravitational waves and electromagnetic radiation before, it’s only happened once: in the famous 2017 event.

The scientific story behind the formation of heavy elements begins with two types of observations:

• gravitational wave observations, which allow us to detect the “ripples” in spacetime created from the inspiral and merger of two massive, compact, co-orbiting objects,
• and gamma-ray burst observations, which allow us to detect the highest-energy photons in all the Universe, produced only in the most energetic cataclysms of all.

On the gravitational wave side, we have observatories like LIGO, Virgo, and KAGRA, which are particularly sensitive to gravitational waves produced by very compact, dense, massive objects that orbit very closely to one another: tight orbits involving either a neutron star-neutron star, a neutron star-stellar mass black hole (i.e., black holes with less than around ~100 solar masses), or two stellar mass black holes.

When the current generation of gravitational wave detectors (advanced LIGO) turned on in 2015, black hole-black hole inspirals and mergers were seen almost immediately. Neutron stars would take longer to show up, with the first such detection arriving on August 17, 2017. This didn’t only show up with gravitational wave signals, but also — almost simultaneously — from another source: a gamma-ray burst detector.

In the final moments of merging, two neutron stars don’t merely emit gravitational waves, but a catastrophic explosion that echoes across the electromagnetic spectrum. Whether it forms a stable neutron star or a black hole (like the 2019 merger), or a neutron star that then turns into a black hole (like the 2017 merger), will depend on factors like the total mass of the predecessor neutron stars and their combined spin. Copious quantities of heavy elements are produced in these events.

Light, contrary to what our eyes perceive, comes in an enormous variety of wavelengths:

• gamma-ray light, produced at the highest energies,
• X-ray light, which is still high-energy but not as high as gamma-rays,
• ultraviolet light, which can be produced by conventionally hot objects, such as stars,
• visible light, which is the small portion of the spectrum we can see,
• infrared light, which human eyes cannot see but which our bodies can “feel” as heat,
• microwave light, which is longer-wavelength than infrared light and corresponds to objects radiating at just a few degrees above absolute zero,
• and radio light, which is the lowest-energy light of all.

A few of these wavelengths, like visible, radio, and a few sections of the infrared spectrum, are indeed visible from the ground. But for many wavelengths, including gamma-ray and X-ray light, you have to go to space to observe them.

That’s why it was so fascinating, in 2017, when NASA’s Fermi satellite, which was a gamma-ray telescope launched back in 2008, reported an “alert” that was almost perfectly coincident in time with the arriving gravitational wave signals from GW170817. In fact, when the time series were made public, it came to light that the gamma-ray signature arrived just 1.7 seconds after the gravitational wave signal ended.

In short order, the signals were localized on the sky, and found to correspond to the same location: a nearby galaxy just 140 million light-years away. The afterglow was found, observed with dozens of ground-based and space-based telescopes, and an extraordinary amount of data was collected.

On August 17, 2017, the Laser Interferometer Gravitational-wave Observatory detected gravitational waves from a neutron star collision. Within 12 hours, observatories had identified the source of the event within the rather mundane galaxy NGC 4993, shown in this Hubble Space Telescope image, and located an associated stellar cataclysm called a kilonova (box), caused by the collision of two neutron stars. Note that a kilonova is only one possible origin of gamma-ray bursts, and cannot account for all of them. Inset: Hubble observed the kilonova fade (in optical light) over the course of six days.

What we found was remarkable: enormous quantities of the heaviest elements of all, including gold, platinum, iridium, barium, uranium, and many others. The closeness of this event, coupled with its nature — two colliding neutron stars — enabled us to estimate how much of a great number of element species were produced. Given that:

• the burst of gamma-rays produced in this explosion was very short (i.e., less than two seconds),
• that 30% of all gamma-ray bursts are short-period bursts like this one,
• and that we were able to associate the gravitational wave event of a neutron star-neutron star collision with this associated heavy element-producing kilonova,

scientists determined that it was reasonable to conclude that the overwhelming majority of the heaviest elements in our Universe were indeed produced by neutron star-neutron star collisions.

But then the mystery deepened.

Additional gravitational wave signals that showed neutron star-neutron star mergers were detected, but there were no kilonovae associated with them: no gamma-ray signals, no detected remnant, no sign of heavy element production. Other gamma-ray bursts were seen, sure, and some of them were identified with kilonova remnants, but the rate at which they were seen to occur was far below the initial rate suggested by that one 2017 event. It was as though the Universe had pranked us: showing a nearby, relatively rare event when our gravitational wave detectors first turned on, leading us to overestimate the production rate of heavy elements by neutron star-neutron star mergers.

Advanced LIGO’s range for black hole-black hole mergers (purple) is far, far greater than its range for neutron star-neutron star mergers (yellow), owing to the mass dependence of the signal amplitude. A difference by a factor of ~10 in range corresponds to a difference of a factor of ~1000 for volume, so that even though the number density of low-mass black holes far outstrips the higher-mass ones, LIGO and Virgo are more sensitive out to greater distances for higher-mass systems.

What could the resolution be? Was there some other source of heavy element production that we hadn’t considered? Or were there additional sources of kilonovae — triggered by neutron star-neutron star mergers — that we had perhaps been overlooking?

Both of those options are reasonable lines of thinking to consider, especially when we consider an important aspect of gravitational wave data: the largest-magnitude ripples are produced by the highest-mass objects orbiting at the shortest, closest distances. The current generation of gravitational wave detectors that we possess at present — LIGO, Virgo, KAGRA, etc. — can detect the mergers of massive black holes (of around ~100 solar masses each) out to distances of around 10 billion light-years. However, lower-mass black holes, of more modest masses (like ~10 solar masses each) might only be detectable out to a couple of billion light-years, and neutron stars, which cap out at a maximum of around 2.1 solar masses (based on what we’ve seen so far), can only be seen out to a few hundred million light-years.

Remember that we live in a three-dimensional Universe, so a factor of ten in distance corresponds to a factor of a thousand in terms of volume. Furthermore, the farther away we look, the farther back in time we’re looking as well. If:

• neutron star-neutron star mergers were more common in the past,
• and we’re only capable of probing a tiny fraction of the late-time volume of the Universe for those mergers,

then we’re just going to miss most of them until we build new science facilities with better detection capabilities, such as LIGO II.

This photograph, taken at LIGO Livingston in Louisiana, shows a perspective of looking “down” one of the two perpendicular 4 kilometer arms through which laser light is reflected many times and brought together to construct an interference pattern that’s sensitive to the presence of gravitational waves. With a detector that has arms ten times as long, we’d be sensitive to such events at much greater distances, as well as to orbiting objects with periods that take up to 10 times longer, allowing us to fill in the gap between what LIGO and LISA are sensitive to. The NSF’s “LIGO II” would allow us to close that gap: either partially or wholly.

It was long assumed that the two classes of gamma-ray bursts — the short-period ones (lasting less than two seconds) and the long-period ones (lasting longer than two seconds) — would have different sources and means of production. After all, there are many different classes of supernovae for the different “light-curves” they display, with a variety of origins for these different classes of stellar cataclysm. When GW170817 occurred, it seemed to solve the mystery for the short-period class: these were kilonovae, or merging neutron stars, as was definitively shown for that particular event. Given the similar features found in other short-period gamma-ray bursts, it was reasonable to assume that they were also merging neutron stars.

But what about the long-period gamma-ray bursts? It’s not like we’re splitting hairs when it comes to the two different classes, where short-period bursts typically last 1-2 seconds and long-period bursts last more like 3-5 seconds. No; short-period bursts very often have sub-second durations, and long-period bursts often last for several minutes up to a few hours. Additionally, only about 30% of all gamma-ray bursts are short-period bursts, while 70% of them fall firmly into the long-period category. So what, then, are these long-period gamma-ray bursts, and could they potentially contribute to the substantial amounts of heavy elements that we observe?

These graphs show the energetic production of photons, over time and sorted by energy, for the recent bright gamma-ray burst GRB 230307A: the first bright GRB to occur in the JWST era. Although this was clearly a long-period burst, with a duration well over 2 seconds, its nature was not anticipated.

That specific question, in particular, was of great interest to astrophysicists heading into the JWST era. Long-period gamma-ray bursts come in an array of durations and brightnesses, but almost all of them are located at cosmological distances: from extremely far away. There are a number of candidate ideas for what these long-period gamma-ray bursts could be as far as their underlying nature goes, including:

• merging neutron stars (the same as short-period bursts),
• various classes of supernovae, including superluminous supernovae,
• stellar cataclysms like tidal disruption events,
• and potential collisions of other celestial objects (such as white dwarfs) with neutron stars.

Most scientists working in the field favored some class of supernova, as it was difficult to picture how such a long-duration burst of high-energy particles could arise from the other events. However, many who disagreed pointed out that the environments around these objects could lead to an extended duration for gamma-ray emissions regardless of the source.

But that’s where JWST comes in. Due to its unprecedented capabilities — including its long-wavelength capabilities that could perform spectroscopy, identifying the characteristic absorption and/or emission signatures unique to specific heavy elements — JWST would be able to search for heavy element production from the remnants of these long-period gamma-ray bursts. If a gamma-ray observatory, such as Fermi or Swift, could localize the burst, then JWST would be able to find what elements were present in its remnant.

This image shows a photometric JWST image of a host galaxy for a neutron star-neutron star merger, along with the location of the remnant of GRB 230307A, shown at the top left. The source is faint and barely detectable at bluer (shorter-wavelength) colors, but appears bright farther into the infrared. Although this photometric data is fascinating, the spectroscopic data is needed to understand its origin.

This is precisely what happened with the exceptionally bright gamma-ray burst GRB 230307A: which was first observed on March 07, 2023. In fact, it was observed by both Swift and Fermi, and it was well into the long-period category of gamma-ray bursts, lasting for more than three minutes, or around 200 seconds all told. It was also an excessively bright burst: the second most energetic gamma-ray burst observed since they were first observed more than 50 years ago. Realizing this was a “target of opportunity,” researchers swiftly proposed (and were granted) two windows to perform both imaging and spectroscopy of this remnant: 29 days and 61 days after the initial burst was detected.

Spectroscopy revealed the copious production of several very heavy elements, including tellurium, tungsten, and selenium. They were all highly ionized, having lost two electrons, indicating an incredibly hot, energetic environment around them. While the tungsten and selenium signatures were only of modest statistical significance, the tellurium was undeniable and extremely strong. These are heavy elements, among the heaviest elements of all, being produced by a long-period gamma-ray burst.

Since the only place known to create these elements and be consistent with such a gamma-ray signal is neutron star-neutron star mergers, it stands to reason that we should now consider that long-period gamma-ray bursts are also caused by neutron star-neutron star mergers: the same events that trigger short-period gamma-ray bursts.

The spectroscopic signatures of several heavy elements, including tellurium, tungsten, and selenium within the spectrum of the remnant of long-period gamma-ray burst GRB 230307A strongly points to a neutron star-neutron star merger as its origin. This turns our ideas about gamma-ray bursts and their progenitors on its head, as both long-period and short-period bursts may owe their origins to neutron star mergers.

So that might be the solution to the heavy element mystery: that not just short-period gamma-ray bursts, but potentially all gamma-ray bursts, are produced by neutron star-neutron star mergers. Sure, LIGO (along with the rest of our suite of gravitational wave detectors) has also seen neutron star-neutron star mergers that don’t:

• emit gamma-rays,
• cause kilonovae,
• or leave any detectable remnant behind,

but that’s precisely what’s expected for the merger of higher-mass neutron stars. In particular, if the combined mass of the neutron stars that merge is more than about 2.75-3.0 solar masses, there will be no explosion, no emission of gamma-rays, and no remnant except one: a black hole, formed from the entire contents of both progenitor neutron stars.

There is another mechanism, however, that could produce some of these heavy elements as well, even though it wouldn’t be associated with a gamma-ray burst: giant flares around magnetars. In addition to kilonovae, it was recently shown that a giant flare from magnetar SGR 1806-20 likely struck the crust of a neutron star, leading to the production of some of these heavy elements. These magnetars could be responsible for at least 1-10% (and possibly more) of all the heavy elements produced in a Milky Way-like galaxy, and possibly more. This is a process that NASA’s planned upcoming COSI mission, the Compton Spectrometer and Imager, should be ideal for investigating and teaching us more about.

This illustration and light-curve shows the scenario, and the observed event, of a magnetar giant flare coming from the neutron star SGR 1806-20. Surprisingly to many, it shows strong, direct evidence for r-process (rapid neutron capture process) elements produced within it, including direct evidence for elements heavier than zirconium.

But for right now, the leading scenario for producing the overwhelming majority of heavy elements in the Universe is kilonovae: merging neutron stars that don’t directly become a black hole, but whose ejecta return enriched material back to the Universe. And instead of having two different origins for the two main classes of gamma-ray bursts, we’re looking instead at a unified scenario, where neutron star-neutron star mergers drive most, or even all, of the gamma-ray bursts we observe. As JWST has shown us, this merger/kilonova scenario does indeed apply to short-period and long-period bursts both, and may ultimately be responsible for 100% of them.

That doesn’t mean the science is “over” on this front, however, not by any means. How often do magnetar giant flares occur, and are there other (weaker) flares that can also create an abundance of these heavy elements? How significantly do they contribute to our heavy element total? Are there other sources of gamma-ray bursts besides neutron star-neutron star mergers, and do they produce heavy elements as well? Are the abundances of neutron star-neutron star mergers throughout cosmic history sufficient to account for all of the heavy elements we observe? And what about the supernova question; are there classes of supernovae that can indeed produce elements above element #40 on the periodic table?

These are questions we’ll be asking in the years to come, hopefully with improved observatories and new facilities at our disposal to show humanity the answer. After all, the search for the solution to the ultimate question — of “Where did we come from?” — demands that we keep moving forward in our quest for knowledge.