Mystery of the Missing Star

The spectacular supernova of 1987 left a hole astronomers have tried to plug for nine years.

By Adam Frank|Sunday, December 01, 1996
On February 23, 1987, an unprepossessing star called Sanduleak - 69°202, which had sat without incident in the southern skies for eons, suddenly blew itself to bits in the most spectacular supernova to be seen in nearly three centuries. SN1987A, as Sanduleak was now designated, was the closest supernova to our solar system ever to be observed with modern telescopes, and it offered astronomers an unparalleled view of the death of a massive star. It also gave them dramatic, unequivocal confirmation of what by then had become the classic theory of supernova explosions. There was only one problem. Something fundamental was missing from SN1987A. Nine years later, it’s still missing.

According to the classic theory, there should be a neutron star where Sanduleak used to be. Neutron stars are dead stars, stellar cinders made of neutrons squeezed through the bars of their atomic cages and thus able to achieve extraordinary densities--a neutron star just 10 miles across contains as much matter as our sun. Physicists don’t know very much about how these odd beasts behave, but they know that the stars usually appear in the sky as pulsars, rotating stars that project beams of intense radio waves into space, like cosmic lighthouses. In the case of SN1987A, however, no pulsars have been detected.

The only alternative suggested by classic theory is that Sanduleak, rather than forming a neutron star, collapsed into a black hole, but astronomers don’t put much stock in this option. For one thing, conventional wisdom says that Sanduleak was too small to turn into a black hole. And besides, black holes tend to swallow everything in their vicinity, supernovas included. If there was a black hole where Sanduleak used to be, we would never have seen the supernova that produced it in the first place. Yet we did, and its hollow remnant haunts us still.

When conventional wisdom falls so far short of explaining what astronomers see in the sky, it often means that it’s time for some radical new idea. Such ideas usually come from the younger, hungrier researchers in astrophysics, the ones who are struggling for job, tenure, and reputation and who, by virtue of the youthful combination of inexperience and exuberance, are apt to see old problems in a new light. SN1987A is no exception, at least not in its demand for novel explanations. What is surprising, however, is that those explanations are coming from two of astronomy’s most respected--and most senior--scientists.

Between them, Hans Bethe and Gerry Brown have 120 years of experience in physics research, not to mention a Nobel Prize--Bethe won it in 1967 for a theory articulating the physics of nuclear reactions in stars. At 90, he is now a professor of physics at Cornell in Ithaca, New York. We had a conference a few years back, says Brown--a mere tot at 70 and a professor of physics at the State University of New York at Stony Brook--and someone pointed out that I published my first paper before many of the speakers were born, and Hans had published his first paper before I was born. Drawing on their vast experience, and with the help of some fresh insight, Brown and Bethe have offered an ingenious solution to the puzzle of SN1987A. If they are correct, they will have rewritten completely the physics of neutron stars.

Brown and Bethe believe that the failure to find a neutron star in SN1987A is not atypical at all. They cite a study in which about half of all known supernova remnants were shown to lack conclusive evidence of neutron stars. If all supernovas produce neutron stars, says Brown, then why do so many supernova remnants [the giant smoke rings left over from the blast] lack evidence for neutron stars at their centers? The answer, he and Bethe believe, is that there is indeed a black hole at the center of SN1987A--a small one, formed in a fundamentally different way than classical theory suggests.

To all outward appearances, Sanduleak’s behavior was precisely what you’d expect of any ordinary, medium-size dying star--that is, one with only about 20 times the mass of our sun. Having run out of fuel for the nuclear fusion reactions in its core, it succumbed to its own gravity and collapsed. In half a second the core, which consisted of silicon, carbon, oxygen, and other heavy elements, was squeezed so tightly that the individual nuclei, the clusters of protons and neutrons at the heart of each atom, merged. Protons and electrons began to turn into neutrons, and the core became one big nucleus--a proto-neutron star.

The core continued to collapse, compressing itself to a state as much as ten times denser than an atomic nucleus, or 10,000 trillion times more compact than water. (I don’t know how he figured this out, says Brown, but Hans once calculated that a teaspoon of neutron star material weighs as much as all the buildings in Manhattan.) Once the core crossed this threshold of nuclear density, the strong nuclear force--the same force that binds protons and neutrons to the atomic nucleus--began to repel with a vengeance, bringing the collapse to an abrupt halt. Once you pass nuclear densities, says Brown, the core acts like a dense rubber ball. The harder you squeeze it, the harder it pushes back. The hydrogen and helium gases that made up the bulk of the star’s outer layers weren’t massive enough to overcome the repulsive strong force. Instead of falling into the core, they bounced off and went hurtling into space at great speed. This was the supernova explosion.

After that there is some disagreement as to what in fact happened to the star formerly known as Sanduleak. The core should have been left behind as a neutron star, but it is, of course, nowhere to be seen. Instead of turning into a cosmic rubber ball, the core could have continued to collapse, becoming a black hole, except that classic theory requires a tremendously heavy torrent of infalling matter from the outer layers to overcome the strong nuclear force--far more than puny Sanduleak could have mustered.

Brown and Bethe offer a third option: they think that the rubber of the cosmic hard rubber ball is much softer than most astrophysicists think. The term that astrophysicists use to refer to the hardness or softness of the core is its equation of state. A softer equation of state means that something doesn’t push back as hard when you squeeze it, says Stan Woosley, a supernova expert at the University of California at Santa Cruz. In other words, a softer equation of state would mean that when collapsing, even a moderately massive star, such as Sanduleak, can overcome the strong nuclear force and turn itself into a black hole.

Brown and Bethe, to be sure, are not observers but theorists, and so it is somewhat misleading to couch their insights in terms of observational data. Their ideas derive from an intuition about the way nature behaves on its most fundamental level, the kind of feeling, or hunch--almost a personal aesthetic--that is every bit as important for the good theorist as the ability to solve equations. To turn this hunch into a theory, however, Brown and Bethe needed to come up with a physical mechanism that would soften the equation of state. And they needed to explain another annoying observation that belied their assertion that a black hole formed at SN1987A: during the supernova explosion, scientists detected a burst of neutrinos lasting about ten seconds emanating from Sanduleak. Neutrinos are ghostly subatomic particles that weigh almost nothing. They are produced in prodigious numbers when protons and electrons merge into neutrons during the core’s collapse. Since neutrinos have little if any mass, they can easily escape the gravitational clutches of a neutron star, but--and here’s the rub--not a black hole’s. A black hole would swallow the neutrinos, says Woosley. So if one was going to form in SN1987A, it would have to wait at least ten seconds to let the neutrinos we saw escape. To make their idea work, Brown and Bethe needed to find some trick hidden deep in the unknown structure of matter that could soften up a neutron star--but one that would wait precisely ten seconds before doing so.

Every January, Brown and Bethe meet in Pasadena, at Caltech. There the two friends spend a month working together on problems involving supernovas and nuclear physics. The way we usually work, says Brown, I cook dinner for Hans and then give him the problems I want him to solve. Then he puts off thinking about them until he can take a bath.

I do my best thinking in the bath, confirms Bethe. We get together the next morning and then we discuss the problem.

In this manner the two physicists began in 1993 to work on the problem of soft neutron stars. While they were walking in the hills above the campus one afternoon, Brown mentioned a remarkable idea, which later became a cornerstone of their work on SN1987A. There was this theory of kaon condensates two kids from Harvard had come up with, recalls Brown. I told Hans about it, and he quickly pictured how it could work in a collapsing star.

These two kids were David Kaplan and Ann Nelson, a husband-and- wife team of physicists now at the University of Washington, and their theory described how a particular chain of subatomic transformations might occur deep within a neutron star. The idea rested on the quantum mechanical notion that particles do not have stable identities and that they can, with enough energy, spontaneously transform themselves from one type of particle to another. What we suggested, says Kaplan, was that when ordinary matter is squeezed to the densities you find in a neutron star, electrons in the star can be transformed into particles called kaons.

Kaons are exotic, very heavy particles with many unusual qualities, and one that is particularly relevant to the soft equation of state that Brown and Bethe were seeking: two or more kaons can occupy the same energy level at the same time. This is to say, in physicist’s jargon, that kaons do not obey a law of quantum mechanics known as the exclusion principle. Electrons, on the other hand, do: the more electrons you cram together, the higher their energy levels go. Thus in a collapsing stellar core, the more electrons you had occupying the same space, the higher the energy and the greater the resistance that would be generated against infalling gas--making the core harder. Kaons, by contrast, could coexist peacefully in the core, all occupying the lowest energy level without exerting any resistance at all against infalling matter. They would, in fact, form a Bose-Einstein condensate, a form of matter long predicted but created in a laboratory for the first time just last year. And the appearance of a kaon condensate would have an immediate and dramatic effect. Kaon condensates don’t bounce around, says Kaplan. The weird thing about them is they have essentially zero temperature. They are ice cold and don’t help support the neutron star at all. The core, in a word, would be soft.

It’s a lot to ask of an electron, a very light particle, to make the leap to the very heavy kaon. To physicists, the transformation Kaplan and Nelson proposed was as startling as having your dog suddenly transform itself into your mailman. Turning something so light into something so heavy would require tremendous energy. How would it take place? Kaplan and Nelson found the answer in Einstein’s famous legacy--that mass and energy are interchangeable.

If you think of mass and energy as money in a bank account, the electron is very poor, and the heavy kaon is very dear indeed. But that’s the valuation only here, on Earth. In the core of a neutron star, the situation is much different. There the potential exists for very light (cheap) kaons to pop into being. The difference arises because in dense nuclear matter a kaon would exert an attractive force on the nearby neutrons, and attraction is a form of energy. Thus the energy needed to create the kaon in the first place is lowered, and the event becomes a possibility. And in quantum mechanics, possibility equals actuality--if there’s potential for an electron to turn into a light kaon, it will happen. Deep inside the nuclear matter, says Kaplan, the kaons weigh less, so the electrons can more easily make the leap.

Although kaons made on Earth in particle accelerators exist for only 10 billionths of a second, kaons in a neutron star would be held stable by their attraction to the neutrons. The kaon is such an evanescent and unstable particle on Earth, says Kaplan. For this rare, exotic, and barely understood form of matter to play such a major role in a neutron star was quite a surprise. For a physicist it’s really delightful.

Brown and Bethe quickly recognized the kaon condensate theory to be the mechanism that just might soften the neutron star’s equation of state. Along with other collaborators, the two physicists hammered out the details of how the condensates might trigger the collapse of the neutron star into a black hole. The theory was both powerful and elegant. But to explain SN1987A there was still the problem of timing. What mechanism could keep the kaon condensate from forming long enough to let the ten-second burst of neutrinos escape?

As it turns out, the neutrinos themselves provide the answer through the exclusion principle. Neutrinos are an inevitable by-product of the compressing core of a dying star. As the core is squeezed, protons and electrons undergo the quantum mechanical quick change and merge into neutrons, each time creating a neutrino as well. Electrons that turn into kaons also release neutrinos, but by the time the core is dense enough to allow kaons to form, the neutron star is already facing a neutrino overpopulation crisis. Since neutrinos obey the exclusion principle, there’s a limit to how many you can cram together into the core. Once that limit is reached, the production of new neutrinos comes to a halt. Since you can’t make a kaon without releasing a neutrino, the production of kaons comes to a stop as well.

Because of the exclusion principle, says Brown, you can only make new kaons at the rate that old neutrinos leave. This means you have to open up holes for the new neutrinos or new kaons won’t be allowed to form.

Once the old neutrinos leave en masse, says Kaplan, the electrons will turn into kaons, the kaons will turn into a condensate, and within a millisecond the young neutron star goes screaming into oblivion. What’s needed, though, is some way to get rid of the neutrinos.

In an ordinary star, neutrinos fly right through matter, but in the superdense core they must bull their way out, wiggling slowly from the center toward the edge. Adam Burrows [a supernova expert at the University of Arizona in Tucson] did a fine paper some years back that calculated the time it takes for neutrinos to diffuse out of the young neutron star, says Brown. We used his result in our calculations. The time worked out to be, surprisingly enough, about ten seconds. With the ten-second timing delay out of the way, Brown and Bethe could now explain why there should be a black hole at the center of SN1987A. They also estimated that there should be 50 times more black holes in the galaxy than previously estimated.

Despite the neat fit, not everybody is convinced. The absence of evidence for a neutron star in SN1987A is certainly not evidence of absence, says Stan Woosley. My prejudice is that there is an as-yet- undetected neutron star in SN1987A.

And Brown, for one, concedes that the theory he and Bethe have conceived is still very speculative. Nobody understands what matter is like at nuclear densities, and here we are, cheerfully talking about densities five or ten times greater. Nobody believes us at those densities, and we really don’t believe ourselves. But in our scenario, the formation of a kaon condensate really has a qualitative effect. It’s a new idea, so we are allowed some poetic license.

Brown, of course, is being modest. His and Bethe’s theory about the physics of neutron stars may have flaws, but it is one of the most complete and plausible theories around. For all the uncertainty, their work may be only the first fruit of a larger symbiosis between nuclear physics and astrophysics that is just now getting under way. For instance, new machines called heavy ion colliders can now smash whole nuclei rather than just individual particles, and they are giving physicists their first experimental glimpse of matter at nuclear densities. We are entering a new era, says Brown. Soon we may get to learn directly about nuclear matter at three, four times nuclear density. This information will in turn help astrophysicists understand the extreme states of stellar evolution. All this activity just adds fuel to Brown and Bethe’s enthusiasm.

Dr. Brown got me into supernovas 20 years ago, says Bethe. Now here I am, 90 years old, and I am still working on them. I am just fascinated. Brown seconds the sentiment. I am more excited than I ever have been, he says. After all these years, I still just want to understand the universe. It looks like the retirement parties will have to wait.
Comment on this article