21-4 The most intense radiation bursts in the universe may be caused by the formation of black holes

As we have seen, bursts of X-rays from binary systems such as Cygnus X-1 lead us to the remarkable conclusion that black holes are real, and that certain stars can evolve into black holes. Even more remarkable is that astronomers may actually be able to observe the moment that a black hole forms from a dying star. The key to this comes from gamma-ray bursts, mysterious objects that emit the most powerful bursts of high-energy radiation ever measured. To understand the connection between these objects and black holes, we begin by looking at how gamma-ray bursts were discovered and how astronomers struggled to determine their true nature.

A Gamma-Ray Mystery

Gamma-ray bursts were discovered in the late 1960s by the orbiting Vela satellites, whose detectors noticed flashes of gamma rays coming from random parts of the sky at random intervals. This discovery was an unexpected consequence of the Cold War. The Vela satellites were originally placed in orbit by the United States to look for high-energy photons coming from above-ground tests of nuclear weapons by the Soviet Union, tests that had been banned by treaty since 1963. (No such tests were ever detected.) With a new generation of gamma-ray telescopes currently in orbit (see Section 6-7), new gamma-ray bursts are being found at a rate of about one per day.

Gamma-ray bursts fall into two types. Long-duration gamma-ray bursts, which are more common, last from about 2 to about 1000 seconds before fading to invisibility. The less common short-duration gamma-ray bursts last from a few hundredths of a second to about 2 seconds, and tend to emit photons of shorter wavelength and hence higher energy. Unlike X-ray bursts (see Section 20-12), gamma-ray bursts of both types appear to emit only one burst in their entire history.

What are gamma-ray bursts, and how far away are they? These questions plagued astronomers for almost 30 years. One important clue is that gamma-ray bursts are seen with roughly equal probability in all parts of the sky, as Figure 21-12a shows. This suggests that they are not in the disk of our Galaxy, because then most gamma-ray bursts would be found in the plane of the Milky Way (Figure 21-12b). One idea to explain their uniformity on the sky was that gamma-ray bursts are relatively close to us and lie in a spherical halo surrounding the Milky Way. Alternatively, gamma-ray bursts could be strewn throughout space like galaxies, with some of them billions of light-years away.

Figure 21-12: Gamma-Ray Bursts (a) This map shows the locations of 2704 gamma-ray bursts detected by the Compton Gamma Ray Observatory (see Section 6-7). The entire celestial sphere is mapped onto an oval. The colors in the order of the rainbow indicate the total amount of energy detected from each burst; bright bursts appear in red and weak bursts appear in violet. (b) This map shows the same sky as in part (a) but at visible wavelengths. Comparing part (b) with part (a) shows that unlike X-ray bursts, which originate in the disk of the Milky Way Galaxy, gamma-ray bursts are seen in all parts of the sky.

(a: NASA)

Long-Duration Gamma-Ray Bursts and Supernovae

Figure 21-13: R I V U X G
The Host Galaxy of a Gamma-Ray Burst This false-color image was made on February 8, 1999, 16 days after a gamma-ray burst was observed at this location. The host galaxy of the gamma-ray burst has a very blue color (not shown in this image), indicating the presence of many recently formed stars. The gamma-ray burst may have been produced when one of the most massive of these stars became a supernova.

(Andrew Fruchter, STScI; and NASA)

For many years there was no convincing way to decide between these competing models, which placed the bursts either quite close or very far away. This state of affairs changed dramatically in 1997, thanks to the Italian-Dutch BeppoSAX satellite. Unlike previous gamma-ray satellites, which could determine the position of a burst only to within a few degrees, BeppoSAX could pin down its position to within an arcminute. This made it far easier for astronomers using optical telescopes to search for a counterpart to the burst. Just 21 hours after BeppoSAX observed a long-duration gamma-ray burst on February 28, 1997, astronomers located a visible-light afterglow of the burst. Since then, astronomers using large visible-light and infrared telescopes have detected the afterglow of many gamma-ray bursts and found that they occur within galaxies (Figure 21-13). As of this writing, the most distant gamma-ray burst yet seen lies within a galaxy some 13 billion (1.3 × 10¹⁰) ly away. (In Chapter 23 we will see how astronomers determine the distances to remote galaxies.) To be seen at such immense distances a gamma-ray burst must release a truly stupendous amount of energy in the form of radiation.

The relationship between gamma-ray bursts and galaxies strongly suggests that a burst is a cataclysmic event involving one of a galaxy’s stars. Observations of a relatively bright (and, hence, relatively close) long-duration gamma-ray burst have confirmed this idea. On March 29, 2003, the orbiting gamma-ray telescope HETE-2 (short for High Energy Transient Explorer) detected this burst—denoted GRB 030329 for the date it was observed—in the constellation Leo. (GRB 030329 was so intense that its high-energy photons temporarily ionized part of Earth’s atmosphere—the only event of its kind ever caused by an object beyond our own Galaxy.) Within 90 minutes a visible-light afterglow of GRB 030329 was found by astronomers in Australia and Japan. The afterglow was bright enough that astronomers in the United States and Chile were able to measure its detailed spectrum, which turned out to be that of a Type Ic supernova at a distance of some 2.65 billion (2.65 × 10⁹) ly. We saw in Section 20-9 that a Type Ic supernova results when a massive star undergoes core collapse after having first shed the hydrogen and helium from its outer layers (see Figure 20-18c). Subsequent observations with the Hubble Space Telescope revealed an expanding shell of gas at the position of the gamma-ray burst. This shell is just what we would expect from a supernova explosion. Another long-duration gamma-ray burst with an associated Type Ic supernova was seen in 2006.

Beamed Radiation

A long-duration gamma-ray burst cannot simply be an ordinary Type Ic core-collapse supernova. Such supernovae release about 10⁴⁶ joules of energy, of which only about 0.03% is released as electromagnetic radiation. (Most of the remaining energy goes into neutrinos, while some goes into accelerating the debris that expands away from the supernova.) This radiation streams outward from the supernova in all directions equally, like light from the Sun. If we assume that a gamma-ray burst also emits light equally in all directions, the inverse-square law that relates brightness, luminosity, and distance (see Section 17-2) tells us that the most energetic bursts would have to emit about 3 × 10⁴⁷ joules of radiation in less than a minute. It would take 100,000 ordinary supernovae going off simultaneously to release this much radiation!

This dilemma can be resolved if we imagine a type of supernova that emits most of its radiation in narrow beams. If such a beam happened to be aimed toward Earth, we would detect a far more intense burst of radiation than we would from an ordinary supernova.

Collapsars and the Birth of a Black Hole

One theoretical model that would produce beamed radiation invokes a special type of supernova called a collapsar (also called a hypernova). These objects are thought to result from progenitor stars that are very massive (more than about 30 M_⊙), have lost their outer layers of hydrogen and helium, and are rotating rapidly. The core of such a star is too massive to be a white dwarf or a neutron star, so when thermonuclear reactions cease, the core will become a black hole. Figure 21-14 depicts what happens when thermonuclear reactions cease in the core of such a star. The black hole forms before the material outside the core has a chance to contract very much (Figure 21-14a). As a result, the black hole quickly forms an accretion disk from the surrounding stellar material as shown in Figure 21-14b. This process is aided because the progenitor star was rotating rapidly, which makes it easier for the infalling material to form a rotating disk around the black hole. Some of the infalling material does not fall into the black hole, but is ejected in powerful, back-to-back jets along the rotation axis of the accretion disk. (Much the same mechanism acts in close binary star systems in which one member is a black hole, as we described in Section 21-3.)

Figure 21-14: The Collapsar Model of a Long-Duration Gamma-ray Burst These illustrations show the final few seconds in the life of a massive, rapidly rotating supergiant star that has lost its outer layers of hydrogen and helium. After the jets have subsided, what remains is a Type Ic supernova with a black hole at its center.

(a, c: NASA/Sky Works Digital; b: NASA and A. Field, STScI)

The jets are so energetic that they reach and break through the surface of the star within 5 to 10 seconds after being formed, as Figure 21-14c shows. (It helps that the star has already lost much of its outer layers, which makes it easier for the jets to break through what remains of the star.) As they travel outward, the jets produce powerful shock waves that blow the star to pieces.

If one of the jets from a collapsar happens to be aimed toward Earth, we see an intense burst of gamma rays. These are produced as the relativistic particles in the jet slow down and convert their kinetic energy (energy of motion) into radiation. The burst is short-lived because the jets have only a brief existence: It takes the black hole only about 20 seconds to accrete the entire inner core of the star, after which there is no longer an accretion disk to produce the jets. We do not see a burst if the jets are aimed away from Earth. Hence, for each gamma-ray burst that we observe, there may be hundreds or thousands of collapsars that go undetected.

The collapsar model cannot explain short–duration gamma–ray bursts, whose bursts of less than 2 seconds are far shorter than those of a collapsar’s jets. A number of models for short–duration bursts have been proposed, such as the energy released by the merger of two neutron stars or a particularly intense magnetar burst.

The newest tool in the search for answers about gamma–ray bursts is a NASA satellite called Swift, which was launched in 2004, and has observed over 800 gamma–ray bursts. In addition to sensitive gamma–ray detectors to identify bursts, Swift carries telescopes that are sensitive to X–rays, ultraviolet light, and visible light. Thus, instead of having to wait for ground-based telescopes to make follow-up observations of gamma–ray bursts, Swift makes these observations itself as swiftly as possible (hence the spacecraft’s name). Observations from Swift may help us better understand the nature of long-duration bursts, and provide important clues about the still enigmatic short–duration bursts.