Shen thought of a way to check: If there are two white dwarfs rotating around each other, and one explodes as a supernova, nothing will be left to hold onto the other. Like a swinging lasso that’s suddenly released, it should fly away as a “hypervelocity” white dwarf.

If the D6 theory is correct, hypervelocity white dwarfs should be common. If it’s wrong, there should be none.

The opportunity to test the scenario arrived in 2018, when the European Space Agency’s Gaia space telescope released a massive new census of objects in the Milky Way. On the day of the release, Shen and his team stayed up all night analyzing the data. They found three fast-moving white dwarfs. Not many, and not none. This was troubling.

Simulating Supernovas

Around this time, multiple teams set to work on computer simulations to test the D6 hypothesis.

Shen and colleagues published simulations in 2021 that played out the aftermath of a D6 detonation. The radioactive nickel-56 nuclei should disintegrate into additional particles, which will then spend months decaying and interacting in the region around the supernova. (Most of our earthly manganese, nickel and cobalt, and a large fraction of our iron, probably originated in reactions such as these.) To capture the tumult, Shen and company simplified the math: They assumed the supernova is perfectly spherical and then simulated the physics along a single line radiating outward from the center.

Strikingly, this “one-dimensional” simulation yielded the correct luminosity curve. “There was no way I would have seen that coming,” Bildsten marveled. “They’re showing they can get a supernova to fall on the Phillips relation, so that’s pretty exciting.”

To verify that a detonation can happen in the first place, though, two other groups were busy developing sophisticated supercomputer simulations of the D6 scenario in three dimensions.

One of these teams recently showed that the D6 scenario can indeed trigger a supernova. The researchers, led by Ruediger Pakmor at the Max Planck Institute for Astrophysics in Garching, Germany, simulated a primary white dwarf with a thick helium outer layer. As the star sucked even more helium from its companion, its outer layer ignited. The explosion traveled quickly around the white dwarf, sending a shock wave deep inside the core that detonated the carbon and oxygen.

But Pakmor’s simulations also produced a strange result. The shock wave traveling through the primary white dwarf sometimes smacked into the companion dwarf hard enough to trigger a supernova in that star as well. This happened in the simulations when the companion’s mass was less than 70% of our sun’s mass, as is usually the case with white dwarfs.

If both white dwarfs often go supernova together, this could explain why fewer hypervelocity white dwarfs are seen. But astronomers have met the news of Pakmor’s double-supernova simulations with caution. “I’m not convinced it happens,” Shen said, “but that’s a really interesting possibility.”