The white dwarf stars were zipping across the Milky Way at more than 1000 kilometers per second—thousands of times faster than a speeding bullet, so fast that they would eventually escape the gravitational clutches of the Galaxy. “They were not like anything we had seen before,” says Boris Gaensicke, an astronomer at the University of Warwick.
Gaensicke and his colleagues suspected these burnt-out embers were fleeing scenes of violence: supernova explosions in which another white dwarf had detonated like an Earth-size hydrogen bomb. In the standard picture of these explosions, known as type Ia supernovae, a nearby giant star lights the fuse. But the extreme speed of the white dwarfs suggested a different scenario, in which the fleeing dwarfs had delivered the sparks, from close orbits around the doomed stars. When they blew up, these partners were flung away like shots from a biblical sling.
The speeds of the three white dwarfs, discovered in a 2018 data set from Europe’s Gaia satellite, were just one clue to this picture. Subsequent ground-based observations found traces of iron and other metals in the stars’ light—elements that might have been implanted by a supernova blast. From the color and brightness of the light, the astronomers could deduce that the stars were hotter and larger than typical white dwarfs, as if they had been puffed up by a sudden energy boost. Most telling, the researchers rewound the stars’ trajectories and found that one hailed from a known site: the remnant of a 90,000-year-old supernova. They are “the very best evidence” for the twin white dwarf scenario, says team leader Ken Shen of the University of California (UC), Berkeley.
The evidence that twin white dwarfs drive most, if not all, type Ia supernovae, which account for about 20% of the supernova blasts in the Milky Way, “is more and more overwhelming,” says Dan Maoz, director of Tel Aviv University’s Wise Observatory, which tracks fast-changing phenomena such as supernovae. He says the classic scenario of a white dwarf paired with a large star such as a red giant “doesn’t happen in nature, or quite rarely.”
Which picture prevails has impacts across astronomy: Type Ia supernovae play a vital role in cosmic chemical manufacturing, forging in their fireballs most of the iron and other metals that pervade the universe. The explosions also serve as “standard candles,” assumed to shine with a predictable brightness. Their brightness as seen from Earth provides a cosmic yardstick, used among other things to discover “dark energy,” the unknown force that is accelerating the expansion of the universe. If type Ia supernovae originate as paired white dwarfs, their brightness might not be as consistent as was thought—and they might be less reliable as standard candles.
Robert Kirshner, a longtime supernova watcher at the Gordon and Betty Moore Foundation, isn’t ready to give up on the classic scenario, but he acknowledges the misgivings about it. “It’s plausible that there is more than one path to a type Ia supernova, but the uniformity of the light output is a tiny bit of a paradox,” he says. Nevertheless, “There’s a nagging doubt: Do we fully understand the nature of these explosions?”
Supernovae come in two basic flavors. The most common are called core-collapse. They occur when a massive star, far larger than the Sun, runs out of fuel. The pressure of the core’s heat can no longer counteract gravity; the outer parts of the star collapse into the core with such force that a rebounding shock wave, or sometimes a renewed fusion burn, blows out the star’s outer layers, leaving a neutron star or black hole behind.
The rest are all type Ia—white dwarf stars somehow reignited into a runaway fusion reaction. The burst of new energy happens too fast for the star to absorb, blowing the entire thing to smithereens in a blast that is brighter and longer than a core-collapse supernova.
White dwarfs might seem unlikely candidates for fireworks. They are the cinders of Sun-like stars that have burned up their hydrogen and helium fuels, leaving—in most cases—carbon and oxygen, heavier elements that can’t fuse in such a low-mass star. White dwarfs shrink to the size of Earth, and only glow from leftover heat. Theoretically, they should cool to black over billions of years.
But if the white dwarf orbits in a binary pair with another star, a more spectacular fate may await it. The classic type Ia scenario was proposed in 1973 by John Whelan and Icko Iben. They mapped out the fading light of supernova SN 1972e for one full year, and realized its brightness could be explained by the decay of about one solar mass worth of radioactive metals. These, they proposed, were forged in a white dwarf that had grown to a size at which the pressure and temperature in its core would be high enough to fuse carbon, causing a thermonuclear blast. Whelan and Iben suggested the white dwarf might grow to that mass threshold if its gravity stole hydrogen gas from a companion star such as a puffed-up red giant, which doesn’t clutch its outer layers too tightly.
Later modeling showed achieving this growth was a tricky balancing act. If a white dwarf gobbles up hydrogen too fast, the hydrogen layer that forms on its surface can get hot enough to blow up prematurely, in a more modest thermonuclear explosion called a classical nova or, if it happens repeatedly, a recurrent nova. If it accumulates too slowly, the white dwarf can tiptoe up to a mass 1.44 times that of the Sun. At that threshold, known as the Chandrasekhar limit, theorists predict the pressure inside will cause electrons and protons to fuse into neutrons, and the white dwarf will quietly collapse into a neutron star.
But if the hydrogen is added at just the right rate, a white dwarf, especially one rich in carbon and oxygen, can respond more dramatically. Just short of the Chandrasekhar limit, at about 1.4 solar masses, the density and temperature of the core shoot up. Flares of carbon fusion break out. After smoldering for 100 or more years, a runaway reaction detonates the star and blows it apart in a matter of seconds. The resulting fireball, 5 billion times as bright as the Sun, forges a suite of metals from chromium to nickel in the periodic table. The radioactive decay of nickel to cobalt and then iron powers a brilliant afterglow that peaks in a couple of weeks and fades over years. And because, in this picture, every type Ia explodes with the same mass, they should all have the same peak brightness.
That scenario satisfied astronomers for decades. But they have yet to find definitive evidence for it. After a supernova disperses enough to be transparent, astronomers routinely search for a surviving red giant companion but have never found one.
Models also suggest a flash of blue light should appear, hours after the supernova begins, as the expanding fireball slams into the tenuous hydrogen atmosphere of the red giant and heats it enough to glow in the ultraviolet. But when astronomers spotted a supernova in the nearby Pinwheel galaxy within hours of its ignition in 2011, they saw no blue flash. “SN 2011fe was a paradigm changer,” says Stan Woosley, a supernova expert at UC Santa Cruz.
A second predicted signal—weaker but more persistent—has also been elusive. Astronomers would expect some hydrogen from the red giant to be swept along with the other explosion debris, creating an absorption line in the spectrum of the supernova’s light as the remnant cools. But a 2019 study of 227 type Ia supernova remnants found no hint of this hydrogen.
The numbers are against the classic scenario, Maoz says. In a galaxy like the Milky Way, a type Ia supernova occurs once every few hundred years. If they all originate from a white dwarf and a red giant, the galaxy would have to host some 10,000 of these pairs—and there’s little evidence of them. Because it’s hard to see binary pairs as separate stars at such great distances, astronomers have looked for indirect evidence, such as the recurrent novae triggered when hydrogen from a red giant spills quickly onto a white dwarf companion, or the soft x-ray glow that can result from a steady stream of hydrogen. If the Milky Way harbored 10,000 white dwarf-red giant pairs, we should see many novae and soft x-ray sources, but only a handful have been found, Maoz says. “We’re missing 99.9% of the systems.”
The theoretical underpinnings of the classic scenario have problems, too. A key assumption is that the explosion is so violent that the entire star is blown apart. But theorists have had a hard time modeling how the smoldering fusion flares would burgeon into an all-consuming detonation. Modeling that process requires simulating nuclear reactions with centimeter resolution across an object the size of Earth for 100 years or more. Until recently, computers couldn’t handle that challenge—and now that they can, Shen says, only some models predict the transition. “The state of the field is not conclusive yet.”
Enter another way to blow up a white dwarf. Astronomers know that two white dwarfs orbiting each other—a so-called double degenerate—can gradually spiral inward as their rapid orbits throw out gravitational waves, which carry away energy. When they merge, there would be easily enough heat and pressure to kick-start carbon fusion in the combined stars.
Models suggest a problem with this mechanism: The burning would not start in the core. It might kick off in the hot zone where the two are actively colliding, leading to a lopsided and incomplete explosion that would leave much of the stars’ mass unburned. Or it could ignite close to the surface, only blowing off the outer layers of the merged star. “Getting a detonation in a double degenerate is by no means trivial,” Woosley says.
Yet Maoz says the odds are irresistible: White dwarf binaries are thought to be fairly common. In a 2018 study, Maoz and his colleagues looked for the wobble of white dwarfs being tugged by a partner, combining survey results from the Sloan Digital Sky Survey and Europe’s Very Large Telescope. They concluded that about half a billion white dwarf binaries have merged in the Milky Way since its formation. If just one-sixth of those mergers led to a type Ia supernova, that would be one supernova every 200 years—roughly what is observed in the Milky Way. Another team, using Gaia data, came to a similar conclusion.
Maoz is undeterred by the problem of getting a complete, symmetrical blast. “Just because we don’t know how it happens doesn’t mean nature hasn’t found a way.” In fact, many believe nature has found a way to blow up one member of a white dwarf pair—but without a merger. White dwarfs can have some leftover helium in their atmospheres after the core stops burning. When an orbiting pair is on the cusp of merging, the larger of the two stars can rapidly steal helium from the smaller one to form a dense helium layer on its surface. The helium layer can act as a kind of blasting cap, exploding in a small thermonuclear blast and sending a shock wave into the star that can ignite the core.
This scenario is called D6, for dynamically driven double-degenerate double detonation. The idea was first developed in 2010 by James Guillochon, then a researcher at UC Santa Cruz, and his colleagues. It leaves the smaller white dwarf battered but thrown free, like those Shen’s group found in the Gaia data. D6 was originally thought to require a hefty amount of helium, making it a rare event, but more recent modeling suggests just a few percent of a solar mass could be enough, Woosley says.
One key feature of the D6 scenario is that the exploding white dwarf can be well below the critical mass, because the spark comes from a shock wave and not from gravitational pressure. A less massive exploding star will produce less nickel and be less bright.
Recent studies of the metallic elements supernovae forge suggest the low-mass type Ia may be the norm. According to models, the production of manganese in type Ia supernovae is particularly sensitive to density in the white dwarf ’s core: If the star is close to the 1.4–solar mass threshold, its high-density core produces lots of manganese; if the star is lighter—as is likely in the D6 mechanism—it produces one-tenth as much.
As a result, manganese abundances derived from the light of stars today can hint at the masses of the ancient supernovae that seeded them with heavier elements. “Manganese provides an indirect way to probe previous generations of type Ia supernovae that went off in that galaxy,” says Ashley Ruiter of the University of New South Wales, Canberra.
In a pioneering study from 2013, researchers led by Ivo Seitenzahl, then at the Julius Maximilian University of Würzburg, compared manganese abundance in the Sun with models of how much manganese would be produced by supernovae of different masses. They found that only half of the supernovae that exploded in the solar neighborhood in the past needed to be high mass to explain the Sun’s manganese content. “This was the first of a new wave of results,” says Maria Bergemann of the Max Planck Institute for Astronomy. This year, she and her colleagues reported looking at manganese in 42 stars across the Milky Way and concluded that the abundances suggest 75% of the galaxy’s type Ia supernovae were low mass.
The implications of undersize type Ia supernovae extend far beyond the elements in the present-day universe. They also raise questions about the explosions’ long-standing role as “standard candles” for probing cosmic history.
It’s a real golden age for supernovae because we’re finding so many.
In 1998, researchers compared a few dozen distant type Ia supernovae with closer ones and found that they were dimmer than they should have been. They concluded that the universe’s expansion is accelerating, driven by some unknown dark energy—a discovery for which they were awarded the 2011 Nobel Prize in Physics. Supernova distances are also at the heart of a dispute over the value of the expansion rate itself, known as the Hubble constant. In the nearby universe the expansion rate is measured using standard candles such as type Ia supernovae; in the distant, early universe, it is derived from clues such as the cosmic microwave background, the echo of the big bang. When the effect of dark energy is taken out, the two values should agree—but they don’t.
Could a not-so-standard candle jeopardize those discoveries? “It means something, but not that dark energy goes away,” Woosley says. Dark energy has been confirmed using other methods, so he’s not worried about that. But he thinks cosmologists will run into trouble as they put their theories to more rigorous tests that require more precise standard candles. “Supernovae could be less useful for precision cosmology,” he says.
Astronomers already knew the peak brightness of type Ia supernovae isn’t perfectly consistent. To cope, they have worked out an empirical formula, known as the Phillips relation, that links peak brightness to the rate at which the light fades: Flashes that decay slowly are overall brighter than those that fade quickly. But more than 30% of type Ia supernovae stray far from the Phillips relation. Perhaps low-mass D6 explosions can explain these oddballs, Shen says. For now, those who wield the cosmic yardstick will need to “throw away anything that looks weird,” Gaensicke says, and hope for the best.
Andy Howell, a supernova watcher at Las Cumbres Observatory, thinks type Ia supernovae could still be reliable tools for cosmology if astronomers could separate the different varieties of type Ia that are now lumped together. “If we knew there were two populations, we could make the measurements even better,” he says.
So far, astronomers can’t say how many of their favorite cosmic explosions are sparked by white dwarf pairs rather than a giant and a dwarf. “It’s too early to say with certainty what that fraction is,” Ruiter says. But the coming years could bring more clarity.
Survey telescopes that scan the skies nightly or even hourly are catching more and more supernovae. The current frontrunner, the Zwicky Transient Facility in California, spots about 30 supernovae per night. Its output will be dwarfed in 2022 with the opening of the Vera C. Rubin Observatory, an 8.4-meter survey telescope in Chile that is expected to find thousands of supernovae nightly. Other telescopes able to obtain spectra from thousands of objects simultaneously will enable astronomers to study the explosions for the features—the blue flash, the hydrogen absorption lines—that could betray the involvement of a giant star.
Shen and Gaensicke hope the next data release from Gaia will contain more high-speed white dwarfs fleeing from D6 explosions. And the Laser Interferometer Space Antenna, an orbiting gravitational wave detector due for launch in 2034, will be able to sense white dwarf pairs as they spiral in toward merger, giving astronomers a better idea of how common they really are. “It’s a real golden age for supernovae because we’re finding so many,” Howell says. “We’ve now finally got the tools to see them in new ways.”
*Correction, 10 June, 3 p.m.: This story has been updated to clarify James Guillochon’s affiliation.