University of Minnesota
School of Physics & Astronomy
Home > About > News > News >

General News

A New Way to Die? Simulations Show Ancient Supermassive Stars Could Supernova


If supermassive stars—55,500 times the mass of our Sun, or solar masses—did exist when the Universe was very young, they could have died in a special kind of supernova. But if the stars were slightly more massive—56,000 solar masses—they would have become supermassive black holes, say astrophysicists at the University of California, Santa Cruz (UCSC) and the University of Minnesota. Their findings were recently published in Astrophysical Journal (ApJ).

The researchers came to this conclusion after running a number of supercomputer simulations at the Department of Energy’s National Energy Research Scientific Computing Center (NERSC) and the Minnesota Supercomputing Institute. They relied extensively on CASTRO, a compressible astrophysics code developed at the Lawrence Berkeley National Laboratory’s (Berkeley Lab’s) Center for Computational Sciences and Engineering (CCSE).

As of now, no one has ever detected such a supermassive star. But, some scientists believe that they could exist because supermassive black holes—millions to billions of solar masses, or bigger—have been observed. In fact, these colossal black holes make up the core of nearly every galaxy in the cosmos. Despite their ubiquity, astrophysicists still do not know how they formed. However, they hope that supercomputer simulations might help solve this long-standing mystery by providing clues about how and where to look for supermassive stars.

“We found that there is a narrow window where supermassive stars could explode completely instead of becoming a supermassive black hole—no one has ever found this mechanism before,” says Ke-Jung Chen, a postdoctoral researcher at UCSC and lead author of the ApJ paper. “Without NERSC resources, it would have taken us a lot longer to reach this result. From a user perspective, the facility is run very efficiently and it is an extremely convenient place to do science.”

Supermassive Mysteries

There are currently two competing theories about how supermassive black holes form. The small seed theory predicts that a 100 to 1000 solar mass star collapses to form a black hole, which then grows to be supermassive over time by merging with other structures and sucking in surrounding gas and dust. Meanwhile, the big seed theory postulates that an enormous cloud of cosmic gas and dust collapses to form a very unstable supermassive star—millions to billions of solar masses. This star would only burn for a few million years before collapsing under its own weight to become a supermassive black hole.

Because these gigantic black holes have been spotted as far away as 12 billion light years—when the universe was only about one billion years old—many scientists tend to lean toward the big seed theory. These researchers question whether one billion years would really be enough time for the Universe to cool after the Big Bang; for galaxies and stars to form; for stars to ignite nuclear fusion, burn and become black holes; and for those black holes to accrete enough mass to become supermassive.

If supermassive stars did exist in the early Universe, scientists suspect that they probably behaved very differently than nearby stars. After all, the Universe’s first stars (Population III) formed in a relatively pristine environment—from clouds made primarily of hydrogen, and maybe a little helium. All other chemical elements were forged later on by future generations of stars. To gain some insights about how these Population III stars would have behaved, researchers rely on supercomputer simulations.

The Simulations: What’s Going On?

To model the supermassive star’s life, Chen and his colleagues used a one-dimensional stellar evolution code called KEPLER. This code takes into account key processes like nuclear burning and stellar convection, and relevant for massive stars, photo-disintegration of elements, electron-positron pair production, and special relativistic effects. The team also included general relativistic effects, which is important for stars above 1,000 solar masses.

They found that 55,500 solar mass stars live about 1.69 million years before becoming unstable due to general relativistic effects and then start to collapse. As the star collapses, it begins to rapidly produce heavy elements like oxygen, neon, magnesium, and silicon from helium in its core. This process releases more energy than the binding energy of the star, halting the collapse and causing a massive explosion: a supernova.

CASTRO—a multidimensional compressible astrophysics code developed at Berkeley Lab by scientists Ann Almgren and John Bell—was used to model the star’s death. These simulations show that once collapse is reversed, Rayleigh-Taylor instabilities mix heavy elements produced in the star’s final moments throughout the star itself. The researchers say that this mixing should create a distinct observational signature that could be detected by upcoming near-infrared experiments like the European Space Agency’s Euclid and NASA’s Wide-Field Infrared Survey Telescope.

According to Chen, this supernova is unique because it is triggered by relativistic instability, not pair-production instability. When the team compared chemical elements produced in each case they found that a supernova triggered by pair instability produces a lot of Nickel-56 and other iron-group elements, while supermassive stars do not.

Depending on the intensity of the supernovae, some supermassive stars could enrich their entire host galaxy and even some nearby galaxies with elements ranging from carbon to silicon when they explode. In some cases, supernova may even trigger a burst of star formation in its host galaxy, which would make it visually distinct from other young galaxies.

“My work involves studying the supernovae of very massive stars with new physical processes beyond hydrodynamics, so I’ve collaborated with Ann Almgren to adapt CASTRO for many different projects over the years,” says Chen. “Before I run my simulations, I typically think about the physics I need to solve a particular problem. I then work with Ann to develop some code and incorporate it into CASTRO. It is a very efficient system.”

To visualize his data, Chen used an open source tool called VisIt, which was architected by Hank Childs, formerly a staff scientist at Berkeley Lab. “Most of the time I did my own visualizations, but when there were things that I needed to modify or customize I would shoot Hank an email and that was very helpful.”

This research was germinated at the University of Minnesota, Twin Cities. Ken Chen received much support from the School of Physics and Astronomy led by Ron Poling and Joe Kapusta and from the Minnesota Supercomputing Institute led by Tom Jones and Jorge Viñals.

More information at