Another way to explode

Curiously, nature has contrived more than one way to destroy a star. Both types of stellar explosion emit comparable amounts of light, so supernova types have been confused from the early days of this subject. SN la come from exploding white dwarfs. But other stars explode by collapsing. The idea that supernovae have their origin in collapsing stars was proposed by Fritz Zwicky and Walter Baade in 1934. As explained by Willy Fowler and Fred Hoylc in I960, stars with eight or more times the mass of the sun do not produce white dwarfs at the end of stellar burning, but have a different way to explode. For massive stars the explosion energy comes from gravity, not from fusion. Although massive stars have different histories, different structures, and a different energy source for the explosion, the light that is emitted is not so different, so it has taken decades to sift out gravity-powered supernovae from their thermonuclear cousins. This is very important if you want to estimate the distance of a star by using its brightness. To get good results you had better compare objects that are the same. If you don't recognize all the various types of supernovae, you will be sure to make errors in the distances.

Massive stars burn their fuel more quickly than low-mass stars. A star with 10 times the mass of the sun has ten times the fuel to burn, but uses its fuel 10,000 times faster to shine 10,000 times more brightly than the sun, so it exhausts its nuclear energy supply one another way to explode 35

thousand times faster. Quantities matter: 10 million years for the duration of a 10 solar mass star is very different from 10 billion years for the sun's lifetime in the same proportion as a ten-dollar bill is different from a penny. Ten million years is short. For a star.

Though they are brief, massive stars are thorough. Massive stars squeeze energy from nuclear fusion from carbon and oxygen into silicon and sulfur and then all the way up to iron Most of the star is still unburned hydrogen, but the interesting stuff is hidden deep within the star's core, where helium and the heavier elements reside. The residue from hydrogen burning is helium, the ashes from burning helium are carbon and oxygen, oxygen burning produces elements near silicon, and the fusion of silicon reaches the dead end of fusion- iron The products of each stage of nuclear fusion surround the iron core like the rings of a tree stump as the core relentlessly continues on its path toward destruction.

At the point where it has accumulated an iron core, a massive star is like a teenager with a credit card. It has a huge outflow, but no source to maintain its balance—for a star, that's the pressure balance against gravity's relentless inward pull. In low-mass stars, quantum mechanics intervenes to keep 1.4 solar masses of cold carbon and oxygen from collapsing, but massive stars employ pressure from hot gas to balance gravitation. As the core shrinks, trading gravitation for heat in the way Lord Kelvin imagined, the core's temperature rises.

In previous burning stages, as when a massive star ignites its carbon, a higher temperature ignites a new fuel whose energy release maintains a new, if limited, period of equilibrium. When the core is iron, this pattern ends, because you don't get any energy from making heavier elements out of iron. The star has tremendous energy flowing out of the corc, much of it in the form of deadbeat neutrinos that have no electric charge and don't bounce off nuclei either, so they stream out freely and contribute nothing to the support of the overlying material. Eventually, the central temperature reachcs 3 billion kelvins at which point the iron nuclei begin to melt back into lighter nuclei.1 This doesn't produce new energy—it costs energy to break up iron. The inevitable then takes place. The core, about 2 solar masses with a radius about half the size of the Earth, loses its pressure support and suddenly slumps inward. Gravity is so strong in the dense small core that this implosion takes only one second as the iron core accelerates inward to about a third of the speed of light. As the inrushing core approaches the density of an atomic nucleus, the strong nuclear force suddenly halts the contraction and the innermost core begins to form a neutron star. This abrupt deceleration, like a train hitting a wall, sends a powerful shock wave back upstream through the imploding star and, with aid of a blast of neutrinos, ejects the outer layers of the star in a rype II supernova (SN II).

Neutrinos are produced copiously just outside the nasccnt neutron star, about 100 kilometers from the centcr of the collapsc. In models for SN II, the explosions of massive stars, most of the energy of the collapse comes out as neutrinos, about 1% goes into the motion of the exploding star, and only about 1/10,000 of the energy goes into the display of light that makes us pay attention to an exploding star. Although they have no clectric chargc, and nearly no mass, neutrinos carry energy, and this hail of energetic neutrinos plays a decisive role in making the rest of the star explode. Computer models of exploding stars (often done at weapons labs like Los Alamos or Livemnore, which have a professional interest in physical situations where the sudden release of energy blows things apart) show that the hot gas outside a forming neutron star could well be one place where new elements arc synthesized right up to the end of the periodic table. Even though it costs energy to make iron into gold, the region just outside the nasccnt neutron star is made of iron and there is lots of energy from the powerful shock wave ripping through the star. Supernovae turn iron into gold, gold into lead (oops!), and lead into uranium. Elements beyond iron are rare in nature because they are made in very special environments.

Massive stars also blast off their thick unburned and partially burned outer layers as part of the supernova explosion. So core-collapse supernovae from massive stars will eject hydrogen if the star still has its outermost layers, and large amounts of oxygen and other middleweight elements in any ease. Massive stars that exploded more than 5 billion years ago arc the source of oxygen atoms that we're breathing right now.

In the 1930s, Fritz Zwicky and Walter Baade started the modern study of supernovae. They worked as a team, with Zwicky at Cal-tech in Pasadena, California, and Baade just a mile up Lake Avenue at the Santa Barbara Street offices of the Carnegie Institution's Mount Wilson Observatory. Baade and Zwicky coined the name "supernovae" to distinguish them from ordinary novae. Novae are explosions on the surface of a white dwarf that are 10,000 times dimmer and do not destroy the white dwarf. Supernovae, rarely seen in our galaxy, but more frequently when you search large volumes that contain many galaxies, arc much more violent. Although Baade and Zwicky discussed the nugget of this idea in a legitimate scientific setting, a meeting of the American Physical Society, the most vivid early form of publication was a cartoon in the Los Angeles Times on 19 January 1934.2

This is one of Zwicky's remarkable insights, perhaps second only to his discovery of dark matter in galaxy clusters. Zwicky's impact on astronomy has grown over time as supernovae and dark matter have bubbled to the top of the astronomical stew. Fritz died in 1974, and a Ph.D. takes about five years, so five generations of astronomers have grown up knowing the legend but not the person. For those of us who actually met Fritz, as I did in early-morning encounters in the second sub-basement of Caltech's Robinson Lab, time has begun to erode and soften the memory of his abrasive personality. Somewhat. What remains are the ideas without the person: in Zwicky's case this has made it easier to admire his work.

Looking back, we see Zwicky and Baade bravely attributing the energy in supernova explosions to a wild idea: the gravitational collapse to neutron stars. And, in the years after this insight, Zwicky built the first telescope at Palomar Mountain, the 18-inch Schmidt, to follow up this idea, backing his talk with action. Since we now knowr that some supernovae are, in fact, powered by gravitation and do indeed leave neutron stars, we credit Fritz with another daring insight.

Truth is more complex than legend. Zwicky, working at Cal-tech, had begun the systematic study of supernovae to check his theory of collapse to a neutron star. Fritz discovered one supernova in 1936 and six in 1937. All of the supernovae that Zwicky ;ind

Be Scicntific with OL' DOC DABBLE.

Figure 3.1 Be Scientific with OT Doc Dabble. Zwicky's compact 1934 publication of a wild speculation for the origin of supernovae in the gravitational collapse of scars to form neutron scars, "little spheres 14 miles thick.1* This is now thought to be the mechanism for type II Supernovae, though, in 1934. Zwicky was talking about type I super novae. Courtesy of the Associated Press

Figure 3.1 Be Scientific with OT Doc Dabble. Zwicky's compact 1934 publication of a wild speculation for the origin of supernovae in the gravitational collapse of scars to form neutron scars, "little spheres 14 miles thick.1* This is now thought to be the mechanism for type II Supernovae, though, in 1934. Zwicky was talking about type I super novae. Courtesy of the Associated Press

Baade studied in those years showed very similar spectra and light curves. It wasn't until 1940 that Rudolph Minkowski, also working at Mount Wilson, observed the spectrum of a supernova that was completely different. At that point, supernovae were, quite sensibly, split into two types: type I, the original type, and type II, the new kind.5 The legendary insight that supernovae make neutron stars was the inspiration for Zwicky's own observational work on supernovae in 1936 and 1937. But, as luck would have it, all of those were supernovae of type la—the type that does not form neutron stars. Sometimes a good story is better than the facts. Or, as the newspaperman says in The Man Who Shot Liberty Valance, "When the legend becomes fact, print the legend."

The type la story with degenerate white dwarfs and crenelated nuclear burning flames is complicated, but the mechanism for type II with a core collapsc, bounce, and emerging shock wave seems downright baroque. How can we test whether massive stars really do all the things that the computers at Los Alamos and IJvermore predict? There's no way to do a controlled test of a supernova out in the desert near Las Vegas. Astronomy is an observational science, which means we need patience, good luck, and many lines of evidence to test our ideas.

Massive stars that become SN II mature and explode so quickly they could erupt right in the cloud of gas and dust where they formed. The galaxy nearest to our own, the Large Magellanic Cloud (LMC), has many patches of lively star formation, including the giant 30 Doradus region where hot young stars make the surrounding gas glow by ripping off their electrons. The LMC is pan of our galaxy's entourage—it is a satellite of the Milky Way but only easily visible from Earth's southern hemisphere. The brightest stars in the LMC have the luminosity we expect from stars around 20 times the mass of the sun Back in the 1960s, Nick Sanduleak of Case Western Reserve University compiled a catalog of the bright stars in the LMC, One of them is not there anymore.

That star, Sanduleak -69 202, was last seen shining brightly in late 1986—as a massive blue supcrgiant in the LMC. But that star exploded 165,000 years ago, and emissions from the supernova explosion reached the Earth at 7:36 Universal Time on Monday,

23 February 1987. That was supernova 1987A." Neutrinos, nearly massless particles w ith no electric charge, erupted from the forming neutron star in SN 1987A, arrived and flashed through the Earth, which is transparent to neutrinos, before anyone saw the star start to brighten.

At the Carnegie Institution's Las Campanas Observatory in the north of Chile, around 2 a.m. (5 hours Universal Time on Tuesday,

24 February), telescopc operator Oscar Duhalde took a break at the 40-inch Swope telescope, leaving the astronomers in the data room, and going downstairs to heat water for his nightly coffee. While the kettle warmed on the hot plate, he stepped out for a glance at the sky. It was a wonderful clear night, the kind when astronomers can measure the brightness of stars without fear of clouds confusing the data, the kind of night astronomers call "photometric." Looking to the South, Oscar saw the large fuzzy patch of the LMC. Right near 30 Doradus, a patch of star formation in the LMC, Oscar saw something new, a bright star he'd never seen before. Neither had anyone else.

He knew this was worth mentioning to the observers, Barry Madore and Robert Jedrzejewski, but when he came into the control room, they were just reaching the punch line of an off-color joke. By the rime they explained what was so funny about Italians by converting idiomatic English into Chilean Spanish, Oscar had forgotten about the new star. Barry turned up Echo & the Bunnymen on the sound system and they all got back to work.

Ian Shelton, a young Canadian astronomer working at the University of Toronto's telescope, also on Las Campanas, came into the control room at 4 a.m., a little like Tycho seeking confirmation from the "country people who by chance were traveling past in carriages." Ian had discovered a big solid dot on his photographic plate of the LMC, near 30 Doradus. There wasn't any star there on his earlier plate of the same place. He went outside and saw it with his eyes, but he still wanted confirmation of this nova in the LMC,

"Oh yes," said Oscar. "I saw it. Two hours ago. Near 30 Dorado. I saw it."

"A nova?" Barry, an expert on the distance to the LMC, thought for a moment, doing the inverse square computation in his head. "No," he said, "that would be a supernova."

This event was SN 1987A, the brightest supernova seen since 1604.

Theory predicts that most of the energy of a core-collapse supernova comes streaming out as nearly massless, chargeless neutrinos. One of the most interesting observations of SN 1987A was nor made with a telescope, but with an underground tank of water designed to find out if protons are immortal, or just very long-lived. The experimenters had hoped to detect flashes of light produced by rhe death of protons inside the tank and measure a finite lifetime for the proton. This would have been quite interesting, since the physical world we see around us is made of protons. It would have shown that matter is evanescent—just a phase that nature is going through. The decay of protons was predicted by interesting theories of particle physics called grand unified theories that unite the strong and weak and electromagnetic forces in a single conceptual framework. The theorists were so persuasive that experimenters excavated a chamber in a salt mine and built a giant tank containing 6000 tons of ultrapure water to confirm those predictions. They didn't. Just before the Department of Energy cut off their funding, a blast of neutrinos emitted from the star's collapsing core flashed through their detector. This was the sharp yelp of a neutron star being born deep in rhe heart of SN 1987A.

When the supernova was discovered, by its optical emission, the report came to Brian Marsden, the person behind the Central Bureau for Astronomical Telegrams. His office is about 200 feet from mine, but I did not hear about SN 1987A from him. Craig Wheeler called me from Texas. A Texas graduate student was in Toronto, where everybody was talking about Ian Shelton's discovery. The student called Craig, and Craig called me.

"Bob, there's a supernova in rhe Large Magellanic Cloud."

"Ha, ha, ha, Craig Wheeler! Fool me once, shame on you; fool me twice, shame on me."

Nine years earlier, Craig masterminded a practical joke, sending a fraudulent urgent telegram to me in a remote village in Italy. "Return at once! Bright supernova in M51," the fake message said. I was in the midst of complicated airline ticket changes when Craig and his co-conspiritors took pity and let me in on the joke. Had I forgotten? No!

Craig starred to fill me in on the details. 1 cut him off.

"Craig, tell me all this later. Maybe we can observe this puppy with IDE. Hang up and I'll see if we can get NASA going on this."

IUE was the International Ultraviolet Explorer, a nimble little satellite that could observe at ultraviolet wavelengths where the Earth's atmosphere is opaque. I had sent in a "Target of Opportu nity" proposal to observe any bright supernova that came along. Since this was the brightest in 383 years, I was pretty sure they would approve the request and aim the satellite at SN 1987A. Hut T didn't want to waste any time. If we acted fast, we might see ultraviolet light from the hot expanding surface of the star right after the powerful shock wave from the star's core blasts through.

I was looking up the telephone number for the Goddard Space Flight Center when my phone rang. It was Yoji Kondo, the IUF Project Scientist at Goddard. Yoji was courteous, but wildly excited. His mood was catching.

"Bob, good morning."

"Good morning, Yoji." I bowed slightly toward the receiver.

"Perhaps you have heard about the supernova in the Large Magellanic Cloud."

"Yes, I have just been speaking with Craig Wheeler who informed me of this event."

"We thought you might be interested in making observations," Yoji said.

"Yes, I think that would be of interest. "

"They have already begun."


My partner in this work, George Sonnebom, a NASA scientist at Goddard Space Flight Center, was at the IUF control console to make these prompt observations of SN 1987A with the IUE. Our data showed the outer layers of the star being blasted off at 30,000 kilometers per second, 1/10 the speed of light. Over the next weeks, the supernova cooled and faded from our view in the ultraviolet, but we still saw two bright hot stars at the site of the explosion. This was puzzling. Sanduleak -69 202 was known to have one close blue neighbor. Perhaps both Sanduleak's star and its dimmer neighbor had survived and that's what we were seeing with IUF. Perhaps the star that exploded was yet another star in that cruwded neighborhood of the Large Magellanic Cloud.

For a few weeks in 1987, I wasn't sure whether Sanduleak -69 202 had really been vaporized and I said so in public places. Stan Woosley, a supernova theorist at the University of California, Santa Cruz, wasn't persuaded. The match between his models and the observations was far too good. Stan said, "If it wasn't Sanduleak -69 202, the star that exploded was exactly like it." Luckily, I did not publish my mistaken conclusion, though I talked about it enough to richly deserve a roast crow, with stuffing and cranberry dressing. Careful measurement of old data showed that there had been not one but two additional hot blue stars there all along, hidden in the glare of Sanduleak -69 202. The IUE was seeing those other two stars. Star 202 had, in fact, disappeared. Nick Sanduleak was fond of showing a Cleveland newspaper headline drawing the conclusion, "Sanduleak Explodes!" This case of mistaken identity didn't do any permanent harm to human understanding, especially because our observational "fact" didn't convince Stan Woosley that his models were wrong. Rut this was an experience I did not want to repeat.^

On that exciting Tuesday in February 1987, I had recently moved to Harvard from the University of Michigan. At Michigan, several people in the Physics Department were part of the Irvine-Michigan-Brookhaven experiment to find the decay of protons. Since they hadn't found the lifetime of the proton, I thought it was my duty to call them up to alert them to a possible neutrino blast from the supernova in the LMC. I called the Michigan Physics Department. It was a strange encounter; everybody I called was in Moriond, France, at a ski resort for a very important conference on cosmology and particle physics. Undoubtedly they were studying the effects of powder snow on the gravitational descent of physicists. After 20 minutes of finding nobody home, I just left a message.

"Not the lifetime of the proton, but the supernova of a lifetime— look for the neutrinos."

Fortunately, the neutrinos had also left a message on their data-recording equipment at the mine. The team found a flash of neutrinos that had entered the tank (after passing through the Earth!) in the hours before the optical discovery of the supernova. A similar detector in Japan, which had been looking for neutrinos emitted by nuclear reactions in the center of the sun, saw the same event. Having two independent measurements gives you confidence that you are observing something real, not noise in the equipment.

John Bahcall was visiting Harvard from the Institute for Advanced Study in Princeton. Tie came to my office, looking for con versation about Supernovae and for a razor. I le had arrived that morning without shaving and he wanted to clean up before the physics colloquium that afternoon. John was going to bring us up to date on the puzzling measurements of neutrinos from the sun, which showed only about one-third of the amount predicted. John was doing the predicting, and he wanted to convince us that the discrepancy was real and not the result of something he might have forgotten. Like his razor.

I keep a razor in my desk, since I sometimes arrive on overnight flights from observing in Chile in an unkempt state. John used it. Clean-shaven and clear-headed, John started thinking about SN 1987A, the optical observations, and the neutrino signal. By the end of the day, after talking with his friends in the physics department, John sent a letter to Nature, the science journal that believes it is the world's most prestigious, using the timing of the neutrino arrival to place a stringent limit on the mass of the neutrino, tetter than any limit from terrestrial laboratory work in 1987. In 1999, measurements of solar neutrinos emitted from the sun's core, and from atmospheric neutrinos, both detected by giant underground water tanks, now indicate that the mass of the neutrino is not quite zero, a very important fact for particle physics, and one small source of dark matter for cosmology.

The explosion of supernova 1987A in the Large Magellanic-Cloud was the best opportunity in four centuries to study the collapse of a massive star. Underground detectors in Ohio and in Japan were jolted by a sudden spike of neutrinos, signaling the birth of a neutron star in the center of the dying star. Since we know that a spinning neutron star lies at the center of the Crab Nebula, a supernova recorded by the emperor's astrologers in the Sung dynasty of China on 4 July 1054, it was natural to think that SN 1987A might have one too, so eager research groups began to look for the telltale flashes from a dense spinning nugget at the center of the cataclysm. Sure enough, in 1989 a group led by Jerry Kristian at the Carnegie Observatories and including Rich Muller, Carl Pen-nypacker, and Saul Perlmutter from Lawrence Berkeley Lab reported seeing the pulses at 37G—the firm signature of the youngest neutron star ever seen.^1 If you would bet your house at 5o, you should probably be willing to bet your life at 37o, but nobody takes statistics that seriously. Plus, there are ways to go wrong that statistics don't include.

I was invited to give a talk in April 1989 at the National Academy of Sciences in Washington, D.C. The academicians often invite people too junior to be elected members of the Academy, but who are working on interesting new developments to amuse them at their annual meeting. It was the first time I had been to that temple of science. I was amazed by how old the academicians were. Scientific research must be good for your longevity. (Now that I have been a member of this geriatric organization for a few years, the antiquity of its memters still makes me feel like a kid—maybe that is the secret of the members' vitality.) Descending the stairway to the talk, Frank Press, the President of the Academy and the father of Bill Press, one of my astronomy colleagues at Harvard, confided that he was especially interested in hearing more details about the amazing neutron star in the center of SN 1987A. According to Kristian et al.'s report published in Nature, the neutron star was spinning at a rate of 1968.629 times per second, compared to the Crab pulsar's leisurely 33 times a second. The investigators said they were further analyzing the data with tantalizing hints that the pulsar might be in an 8-hour orbit around an unseen companion, perhaps a planet. This was wild and exciting stuff.

However, I disappointed Frank Press. I mentioned the pulsar data, but I didn't say too much about it in my talk, because unlike the neutrinos seen in Ohio and in Japan it didn't seem to have the converging lines of independent evidence that make a scientific result secure. At the risk of seeming a dull fellow, I thought it was better to emphasize things that were interesting and true at the expense of things that were just interesting. The pulsar, though reported with great precision in Nature, was seen on only one night, 18January 1989- On other nights, the same equipment and the same analysis failed to detect this amazing object. In general, if something is real, the evidence gets stronger over time. In this case there was always the possibility that the expanding clouds of debris might have allowed only a brief peek at a real phenomenon. Still, when others tried to measure the pulsar, they came up blank. This was a had sign. If something is real, another team with an equivalent technique ought to be able measure the same thing. Having more than one group measure anything important is more than fust a good idea. It makes the case.

During 1989, this mystery deepened. Was there some flaw with the original observation, even though the statistics of the initial measurement had seemed so clear-cut? Eventually, the group that had made the measurement got to the bottom of their own problem. The signal that had seemed so certainly the signature of a spinning neutron star in the center of supernova 1987A was, alas, generated in the circuitry of the television camera used to guide the telescope during the data-taking. On the night when that team made the "discovery," the TV camera was on while they were taking supernova data, but as dawn approached, they turned it off while taking calibration data to avoid damaging the sensitive TV camera. So the signal was in the supernova measurements but not in the calibration data they used to check for spurious noise. Ouch! There are many ways to go wrong. The wonderful thing about science is that eventually nature tells you when you are fooling yourself. Real objects can be measured again or measured by somebody else—false signals will eventually be weeded out.

So, is there or is there not a neutron star at the center of supernova 1987A? We still don't know. Even though the neutrino signal was just what was predicted from a forming neutron star, there isn't yet any clear evidence for one in the supernova debris.7 One possibility is that some of the inner debris fell back on the neutron star and pushed it over the upper limit for those objects (somewhere around 3 solar masses). In that case, gravity would win decisively and the stellar core would collapse to become a black hole. A black hole is a region of space where gravitation is so strong that not even light can escape. Even so, invisible objects can have visible effects—and a black hole could have material in orbit around it that we could measure.

The site of the SN 1987A explosion can still be studied over a decade later with the Hubble Space Telescope (HST). My research team, the Supernova INtensive Study (SINS), has been observing SN 1987A since the launch of HST. The bright 20 solar mass star,

Figure 3.2. Supernova I987A. Space telescope image of the site of SN 1987A, seer 10 years later The exploded star itself is the doc in the center of the bright inner ring, heated by the decay of radioactive elements produced in the explosion The inner ring is gas lost from the pre-supemova star, excited and still glowing from the light of the outburst. This ring was the source of the emission seen by the Internationa) Ultraviolet Explorer satellite in 1987-88. Courtesy of P Challis and the SINS collaboration. Harvard-Smithsonian Center for Astrophys-ics/NASA/STSd (Also see Color insert)

Sanduleak -69 202, is definitely absent. At its site, glowing remains of the explosion are visible. It is difficult Lo study SN 1987A from the ground, because light from the two enduring neighbor stars (the same ones that caused me so much grief in 1987) slops into the light from the supernova. The nearby stars are 100 rimes as bright as the supernova is today, and from the ground, atmospheric blurring smears them into big patches of light that obscure the supernova itself. Debris from the exploded star is now 10 million times fainter than it was in 1987 when Oscar Duhalde saw it with his naked eye in Chile. It is still glowing because the explosion produced fresh elements, some in radioactive forms that continue to excite the debris. The present source of energy for SN 1987A is the decay of

Was this article helpful?

0 0

Post a comment