In preceding chapter we discussed the hydrogen-burning, or main-sequence, phase that follows the birth of stars. After stars have exhausted their hydrogen fuel, they are unable to support the weight of the overlying layers with their internal gas pressure. This crisis means that they must renew the battle against gravitational compression through a sequence of contracting, heating, and igniting new sources of nuclear fuel in order to survive. This chapter discusses these advanced stages of stellar evolution, which ultimately lead to the death of stars.
18.1.1. Hydrogen Exhaustion
Throughout hydrogen burning the cores of stars contract very slowly as stars try to maintain hydrostatic equilibrium. This contraction converts gravitational potential energy into thermal energy, raising the core temperature and the star's luminosity. As the star brightens, it moves up from the zero-age main sequence in the H-R diagram, moving toward the point on the diagram which represents the time when the star has exhausted its hydrogen fuel.
The main sequence consists of a broad avenue of stars of different ages that are evolving away from the zero-age main sequence to their respective hydrogen-exhaustion points (Figure 18.1). Unlike what happens in pre-main-sequence contraction, a star's overall radius does not decrease but rather increases. And since the H-R diagram records increases in luminosity and temperature at a star's surface, the diagram does not reflect the core's slow contraction. For a star like the Sun, the radius will roughly double during its main-sequence life as about 12 percent of its hydrogen is depleted.
[Figure 18.1] As hydrogen burning ends, hydrostatic equilibrium shifts in favor of gravity, and the core contracts more rapidly. The contraction releases substantial amounts of gravitational potential energy, further heating the interior. The layers just outside the former energy-generating core are now hot enough to initiate hydrogen burning in a relatively thin shell surrounding an inactive helium-rich core. Although the core continues to contract and heat, it is the burning of hydrogen in the surrounding shell that is responsible for supplying the star's luminosity. Evolution now carries stars along tracks leading to the red-giant region in the H-R diagram.
The red-giant branch in Figure 18.1 is the portion of the evolutionary track that extends steeply upward at the extreme right. When the contracting helium-rich core reaches about 120 million K (Table 18.1), the second major thermonuclear reaction, helium burning, begins. In this reaction, three helium (He4) nuclei fuse to form a carbon (C12) nucleus and two gamma-ray photons. Stars burn helium (and hydrogen in a shell about the helium-burning core) for about 5 to 20 percent of the time they spent burning core hydrogen as main-sequence stars.
[Table 18.1] When helium burning commences, rapid contraction in the core ends. And just as hydrogen burning led to a slow contraction of the energy-generating core, helium burning also causes a slow contraction. While part of the main-sequence, a star's core is some 20 percent of its radius, but when helium burning starts, the star's core is more like 0.1 percent of its radius. The star's structure has thus changed dramatically from what it was as a main-sequence star.
When will contraction raise the core temperature to 120 million K so that helium begins thermonuclear fusion? This event is determined by a star's main-sequence mass. To discuss this point, we must divide stars into two groups for their approach to helium burning and subsequent evolution: low-mass stars, whose masses are less than about 2M., and high-mass stars, whose masses are greater than 2M..
18.1.2. X-Rays from Stars]
X-rays emitted by celestial objects do not reach the surface of the Earth. Instead, they are absorbed in the atmosphere above the about 30 km. Therefore, in order to study celestial X-rays, we must fly rockets or balloons above the absorbing layers.
In 1949, the Sun was demonstrated to be a source of X-rays, by a group of scientists at the U.S. Naval Research Laboratory. They flew a captured German V-2 rocket outfitted with a device to detect X-rays from a large area of the sky. The fact that the quantity of X-rays that the detector saw was largest when it was pointed toward the sun clearly indicated a solar origin for the radiation. During the decade that followed, improved systems were sent into space to continue studying the Sun's X-ray emission. The steady X-ray emission from the Sun has been found to come from the plasma in its 2 million K corona. In addition, solar flares were found to be transient sources of X-rays.
The researchers soon realized that if they could build sensitive enough detectors, they might be able to measure X-rays from other stars, as well. In 1962, the first celestial X-ray source outside the Solar System was discovered by scientists at the American Science and Engineering Company, and was quickly verified by other groups of scientists. The source was dubbed Sco X-1, since it was the first and strongest X-ray source in the constellation Scorpius. Subsequent studies of Sco X-1 have shown it to be a type of object known as a low-mass, X-ray binary system. These systems, of which a number of examples are now known, consist of a pair of stars one of which is a main sequence star and the other is a neutron star in close proximity. In a somewhat oversimplified picture, plasma flows from the larger star and accretes onto the tiny but massive neutron star. The accretion process causes the plasma to be compressed and heated to such high temperatures that it can emit X-rays.
X-ray astronomy continued through the 1960s with sounding rocket flights which took instruments above the bulk of the atmosphere for periods of five minutes or so, and with high-altitude balloon flights. A handful of sources of celestial X-rays were identified, and their positions on the sky were ascertained with some accuracy.
The first major satellite for X-ray astronomy was the Uhuru satellite, launched in late 1970. The satellite was launched from a platform off the coast of Kenya on that country's independence day, so it was dubbed Uhuru, the Swahili word for freedom. Over the next several years, a number of additional satellites were launched to study the X-ray sky, culminating in the High Energy Astronomy Observatory, HEAO-1, launched in 1977. Many of the X-ray emitting stars found by these observatories were, in fact, binary star systems where the X-rays were produced in an accretion process.
In 1978, HEAO-2, renamed the Einstein Observatory, was launched with the first true telescope for imaging X-rays from objects outside the Solar System. Its great sensitivity permitted it to observe an enormous array of hither to unknown X-ray emitting objects. The totality of X-ray sources observed since 1950 contain a variety of objects, including supernova remnants, active galaxies, and quasars. We will talk briefly about non-binary stars below.
The Sun has a hot corona because energy produced in its turbulent envelope is deposited in the corona through complex magnetic interactions and heats the plasma to the enormous temperature that is observed. The most common main-sequence stars, cooler than roughly 7000oK or spectral type F5, share the same type of structure, and should have coronae like the Sun. Observations from the Einstein Observatory have verified that stars with convective outer layers do in fact emit X-rays, and the strength of the X-ray emission is related to the strength of magnetic fields of the stars.
The Einstein Observatory found that hot young stars without convective outer regions also emit X-rays, but at a lesser rate than the cool stars. This observation remains something of a mystery. Some clues were found when the ultraviolet emissions from the stars was studied extensively using a satellite known as the International Ultraviolet Explorer. These stars turn out to have extensive stellar winds, in which mass pours out of the stars at a rate more than a million times larger than in the case of the Sun's wind. At least some of the X-ray emission may arise in plasma heated by processes such as shock waves in the wind. However, the fact that X-rays are also emitted by hot young stars with weak winds points to an additional, as yet unknown source as well. It remains for future generations of scientists using new, high-technology X-ray and ultraviolet observatories in space to sort out this enigma.
18.2.1. Helium Flash
Continuing our discussion of evolution away from the main sequence, let us consider first the post-main-sequence evolution of low-mass stars. When hydrogen is exhausted, these stars begin an internal restructuring that eventually ends when they are red giants (Figure 18.2). The result of this action is, first, to decrease their surface temperature but to keep about the same luminosity (Figure 18.1). Second, before helium burning actually begins, these stars will brighten appreciably and follow an evolutionary track into the red-giant region that is basically upward in the H-R diagram. For the Sun, this change will take about 1 billion years; for less massive stars, the time will be even longer.
[Figure 18.2] By the time a star like the Sun reaches the red-giant region, the helium-rich core (containing about 30 percent of the star's mass) has been compressed to a volume about twice that of the Earth, with a density of some 106 g/cm3. At this density, the gas no longer acts as a perfect gas, for the electrons have become degenerate (Section 17.1).
For low-mass stars the presence of degenerate electrons in their cores causes helium to ignite very differently from the way it would in a perfect gas. In a perfect gas, any rise in temperature is followed by an increase in pressure, and the contraction then underway is halted. Under perfect-gas conditions, therefore, the increase in pressure expands the core slightly, cools it, and reduces the thermonuclear burning rate until thermal equilibrium is restored.
But when the gas in a star's core is a degenerate-electron gas, it behaves more like a solid than a gas. The core does not expand with an increase in temperature, and if it cannot expand, it cannot cool. Since the rate of helium burning increases with temperature, it continues to rise as more energy is generated; in turn, this increases the rate at which energy is generated, which pushes the temperature still higher; and a runaway condition follows. In a few hours temperatures leap to hundreds of millions of degrees, causing the generation of as much energy as 100 billion stars of solar luminosity--about that of a whole galaxy. This explosive flash of energy is called a helium flash. Even as huge as this energy generation is, the helium flash is not enough to blow the star apart. Instead, all the energy goes into removing electron degeneracy, and this so alters the structure that the star may now burn helium in a core that acts as a perfect gas.
The helium flash just described is generated in computer models of stars--astronomers have not and never will see helium flashes in real stars. This is because a helium flash occurs deep inside a star, hidden by millions of kilometers of gas. Do we believe that such a process as the helium flash actual takes place just because our computers say so? Galileo long ago point science in the direction of discovering the mathematical laws that relate observable quantities in nature. When these laws are discovered and put together in the larger framework of a theory, they represent the boldest stroke of our imagination for understanding nature. Until such a means of pursuing understanding is shown to be incorrect, then science will continue to believe the predictions of its theories even though we may not be able to observationally validate each step, so long as we can relate to the observable world at some points along the quest.
18.2.2 Open-Cluster H-R Diagrams
Astronomers have found observational evidence in H-R diagrams for open clusters that supports the evolution of low-mass stars as just described. As individual stars in a cluster evolve, their rate of aging is determined by their luminosity, which in turn depends on their mass. The more massive ones age faster than do stars of lower mass, as illustrated in Figure 18.3. It is from the relative aging of a cluster's members that astronomers can find the age of the cluster. The critical time when a star reaches the point at which it exhausts its core hydrogen depends on its mass, which in turn is related to its luminosity and hence its age. From stellar models, ages can be placed along the vertical axis of an H-R diagram that correspond to the luminosity of the turnoff point for the least massive star that has reached hydrogen exhaustion, such as that shown in Figure 18.4.
[Figure 18.3]
[Figure 18.4] In Figure 18.4, the youngest cluster is NGC 2362, whose age is estimated to be less than 2 million years, since none of its massive blue stars have yet crossed the Hertzsprung gap. In the next youngest cluster, h and Persei, some of the stars that were originally massive blue ones have evolved across the diagram and are now red giants. Next youngest is the Pleiades, followed, in order, by M41, M11, the Hyades, NGC 752, M67, and NGC 188. The difference between the two oldest clusters, M67 and NGC 188, is that in M67 subgiant and giant stars are more luminous than comparable ones in NGC 188. This is so because stars leaving M67's main sequence are more massive and therefore younger than those in NGC 188. Subgiant branches in M67 and NGC 188 strongly resemble evolutionary tracks for 1.25 and 1M. stars, respectively. Estimated ages are 5 billion years for M67 and about 10 billion years for NGC 188.
18.2.3 Horizontal-Branch Stars
One conspicuous difference between an H-R diagram for open clusters and one for globular clusters is the horizontal branch in that for globular clusters (Figure 18.4). This region of stars runs almost horizontally across the diagram from the red-giant branch to the very blue stars at a visual absolute magnitude of about +1. What does the horizontal branch represent in the evolution of stars?
To answer that question, first remember that the helium flash in solar-mass stars reduces a stars' central temperature over a very short time interval, say, about 10,000 years. After the flash, the drop in core and the hydrogen-burning shell temperatures reduces the star's luminosity while its surface temperature increases. This causes the star to move down and to the left of its original position in the red-giant region and onto the horizontal branch. From stellar models we are led to believe that horizontal-branch stars are stars between about 1 and 0.5M. that are burning helium.
The horizontal branch is a relatively stable phase in a star's life. Yet there is a period when such stars can become temporarily unstable and pulsate as RR Lyrae variables (Section 13.5). Even though the comparison in Figure 18.5 suggests reasonable agreement between H-R diagrams for stars in the globular cluster M92 and a theoretical cluster made up of stellar models of different masses, all aspects of horizontal-branch stars are still not known.
[Figure 18.5] As the time approaches when helium will be exhausted at the low-mass star's center, the battle resumes between gravity and gas pressure. The conversion of helium to carbon causes a star's core to contract slowly and heat, and the star moves upward from the horizontal branch in the H-R diagram along a track that carries them into the giant region a second time. For these stars, the outer layers are bound so loosely by gravity that large amounts of matter may be expelled as a result of stellar winds (Section 16.1). As stated earlier, observations show that luminous red giants and supergiants are indeed losing significant amounts of matter. Most important for low-mass stars, however, is that mass loss reduces the compression of these stars, consequently, compression can not sufficiently heat the core to start carbon burning, the next possible stage in nuclear burning, as shown in Table 18.1. This end to the sequence of nuclear burning processes is the beginning of the end for low-mass stars.
18.3.1. Stars That Become Planetary Nebulae
While low-mass stars evolve toward the red-giant branch the second time after their horizontal-branch phase, burning helium and hydrogen in two separate shells, they pulsate slightly. Eventually their outer layers separate from the core and expand outward at several tens of kilometers per second to form the shell of a planetary nebula. The ejected shell holds most of the original outer layers that missed thermonuclear burning because they were never hot enough (Figure 18.6). Near the red-giant branch the luminosity and temperature of the star's original surface are what is plotted in the H-R diagram. But as the ejected shell expands, the former core is exposed so that its luminosity and temperature become the quantities to plot. Because the former core is hot, the star's position is now on the blue side of the diagram. Thus, after the outburst, the tracks in the H-R diagram along which the star evolves very rapidly cross the top of the diagram from the low-temperature to the high-temperature side. Later, by the time the central star moves into the white-dwarf region, the expanding nebula has mixed with the interstellar medium and is no longer recognizable.
[Figure 18.6] It seems likely that the Sun will eject its outer layers as a planetary nebula some 6 billion years from now. And most, if not all, of the disk-population stars with main-sequence masses between 1 and 4M., as well as some low-mass spheroidal-population stars, also seem likely to become planetary nebulae after being red giants and before becoming white dwarfs.
18.3.2 White Dwarfs
By the time a low-mass star has exhausted its nuclear fuel supply, gravitational contraction has compressed it to extremely high densities, and the electrons in its interior are once again degenerate. Electron degeneracy prevents further contraction, and the star's only source of energy is its store of thermal energy, which supplies the luminosity, cooling its interior. The star is now a white dwarf.
Mathematical modeling of white dwarfs leads to a surprising result: The larger its mass, the smaller will be its radius. For example, a white dwarf whose mass is 0.4M. has a radius equal to about 1.5 percent of the Sun's radius, or about 10,000 km. But a white dwarf whose mass is 0.8M. has a radius that is about 1 percent of the Sun's radius, or about 7000 km (about the size of Earth). A theoretical limit called the Chandrasekhar limit after S. Chandrasekhar is reached for a white dwarf of mass 1.4M. in that it has a radius of zero. We can infer from this that a star whose mass at the time is less than 1.4M. can evolve to become a white dwarf, but one whose mass is greater than that limit cannot. For stars whose main-sequence masses exceed 1.4M., they must lose mass during their lives if they are ever to become white dwarfs.
A star of solar mass enters the white-dwarf stage as a small, hot, blue object. It radiates energy away into space, thereby depleting its internal energy and consequently lowering its temperature. The white dwarf cools, changing from blue to white to yellow and eventually to red. Since such a star can no longer contract, its evolutionary tracks roughly follow lines of constant radius. Evolutionary tracks of white dwarfs in the H-R diagram are nearly parallel to but well to the left and below the main sequence (Figure 18.7).
[Figure 18.7] Most white dwarfs that have been discovered, several thousand in all, are relatively close to the Sun. Their very low luminosity makes them too faint to be seen at greater distances even by large telescopes. Of the 50 or so nearest stars, 3 are white dwarfs--the companions of Sirius and Procyon and van Maanen's star (Figure 13.10). The conservative estimate of Table 20.1 places their total number in the Galaxy at around 35 billion. Most of them are presumably descendants of the oldest disk-population stars.
After billions of years, a white dwarf's thermal energy will be exhausted. As the star cools, its rate of cooling slows, making its approach to its final nonluminous state as a black dwarf quite long. The Galaxy is probably not old enough for very many white dwarfs to have cooled sufficiently to become black dwarfs. This approach to obscurity, though, is definitely a one-way track, from which nothing can save a white dwarf.
Our knowledge of the evolution of low-mass stars is not so complete that we can state precisely what evolutionary track a particular star, such as the Sun, will follow. What has been described is a general evolution for all low-mass stars. If the Sun does indeed proceed along this general evolutionary track, then Figure 18.8 traces out its entire evolution from formation in the interstellar medium to its final state as a white dwarf.
[Figure 18.8] 18.3.3. Sirius B: A White Dwarf
In the constellation Canis Major, the Great Dog, we find the brightest star in the sky, Sirius (Alpha Canis Majoris). Sirius, also called the Dog Star, is one of the three members of the winter triangle of very bright stars, the other two being Procyon (Alpha Canis Minoris) and Betelgeuse (Alpha Orionis). Sirius B, the faint companion of Sirius, was one of the first white dwarf stars discovered. Procyon also has a white dwarf companion. Table 18.2 compares the physical properties of Sirius B with those of the Sun, our stellar yardstick, and the Earth. Sirius B is about 10,000 times less luminous than Sirius and is very difficult to photograph (Figure 18.9).
[Biography - Subrahmanyan Chandrasekhar]
Just as it does for low-mass stars, the day comes in the lives of high-mass stars (those greater than about 2M.), such as Regulus (B7 V), when they also exhaust the hydrogen fuel in their cores. The evolutionary track in the H-R diagram followed by high-mass stars, such as Betelgeuse (M1 I) and Antares (M2 I), is described in this and the next section. Evolution for high-mass stars is very different in many respects from what we have just learned for low-mass stars.
18.4.1 Helium Burning In High-Mass Stars
With hydrogen gone in the cores of massive main-sequence stars, the core shrinks drastically, driving up the density and temperature and exaggerating the differences between a star's central regions and its outer layers.
For stars above 2M. in Figure 18.1, evolution away from the main sequence is quite rapid, almost horizontal, across the H-R diagram toward the red-giant region. These stars expand their radii by a factor of about 10, which constitutes a hundredfold increase in their surface area. Still the temperature is reduced enough to keep the luminosity roughly constant. The region in the diagram across which these stars are evolving is the Hertzsprung gap. Very few stars have been found in this gap, so the transition across the gap must be rapid--occurring in 50 million years to less than 1 million years.
Just when will contraction raise the temperature in the core to 120 million K so that helium burning may begin? Again this is decided by the star's mass. For stars between 2 and 9M. the time comes when the star has become a red giant. When helium burning begins (without a flash, as in low-mass stars, because the density is lower), the star's evolution upward along the red-giant branch stops. Following this, a star, still burning helium in its core and hydrogen in a concentric shell, first contracts its radius and moves to the high-temperature side of the red-giant branch and then back to the low-temperature side a second time, as shown in Figure 18.1.
The path these stars follow in the H-R diagram leads some across the Cepheid instability strip (Section 13.5 and Figure 18.1). A Cepheid variable pulsates only in its outermost layers, where small rhythmic expansions and contractions cause the cyclic variation in its brightness. Once pulsation begins, it continues for a brief time until evolution carries the star toward higher temperatures in the H-R diagram. Classical Cepheids appear to be a spiral-arm component of the stars in the disk of the Galaxy (whose masses range from 3 to 9M.) and that are in their helium-burning phase.
Stars more massive than 9M. begin and complete helium burning on the blue side of the H-R diagram before they cross the Hertzsprung gap leading to the red-giant branch.
18.4.2 Carbon Burning
After helium burning ceases, the parts of a star that have burned helium are rich in carbon and oxygen (Figure 18.10). The next step in thermonuclear evolution is carbon burning, which a star more massive than about 25M. initiates even before crossing the Hertzsprung gap. Carbon burning, which takes place at temperatures above 600 million K, is the fusion of two carbon nuclei to produce primarily neon (Ne20), sodium (Na23), or magnesium (Mg24).
[Figure 18.10] By the time helium is exhausted in the cores of 3 to 9M. stars, they have returned to the red-giant branch a second time (Figure 18.1). The core is slowly contracting and heating. The increased density of an already-dense core squeezes electrons and ions closer together, imposing degeneracy on the electrons.
Once the central temperature in the core has reached about 600 million K, carbon nuclei initiate the thermonuclear fusion reactions of carbon burning. Carbon burning in an electron-degenerate gas can start explosively, just as helium burning does. Thus the core can experience a carbon flash, which removes the degeneracy and allows carbon burning to restore thermal equilibrium.
Stellar models suggest, however, that carbon burning in some stars may start so explosively as to blow the star completely apart and is one type of supernova outburst, as suggested in Table 18.3, where we summarize evolution of all masses of stars. Astronomers are not yet absolutely certain that this explosion actually occurs. This is so because extensive mass loss can apparently prevent these stars from ever reaching the carbon-burning phase and instead they end up as carbon-rich white dwarfs, such as those found in some clusters.
[Table 18.3] For stars with main-sequence masses greater than about 9M., carbon burning begins in a gaseous core that still obeys the perfect-gas law. This is so because the temperature is so high in these massive stars that radiation pressure is an important part of the total pressure, inhibiting the occurrence of electron degeneracy. Thus there is no carbon flash, just as there is no helium flash for any of the high-mass stars.
18.4.3 Oxygen, Neon, and Silicon Burning
In stars more massive than 9M. further contraction after carbon burning makes their cores hot enough to initiate the next round of thermonuclear reactions, known as oxygen, neon, and silicon burning, which take place at 1 to 3 billion K (Table 18.1). Nuclear burning for these massive stars ends with the nuclear reactions that produce the nuclei near iron in the periodic table. Because iron and those elements near it are the nuclei most resistant to any type of structural change, thermonuclear burning beyond them does not release energy but takes up thermal energy from its surroundings. From here on a star cannot go through the alternate contracting, heating, and nuclear-burning sequence that has sustained it against gravity through most of its existence.
The evolution of very massive stars (15M. and up) is the most difficult range of stellar mass about which to have confidence in our stellar models (Table 18.1). Nevertheless, several important conclusions can be made from these stellar models. First, stars more massive than 9M. initiate helium burning as blue supergiants, and those more massive than 25M. also burn carbon while located on the blue side of the H-R diagram. While most massive stars probably become red supergiants, the vigorous mass loss by the very rare 50 to 100M. O stars may prevent them from ever crossing the H-R diagram and becoming red supergiants. That is their entire life span is spent as extremely hot, blue, luminous stars. This is a point that is still under investigation since it bears on our understanding of the most recent nearby supernova outburst, supernova 1987A in the Large Magellanic Cloud--a companion galaxy of our Galaxy that is visible only in the southern hemisphere skies.
Massive stars evolve through their nuclear-burning sequence to a structure in which they possess a very small, extremely dense core with shells like those of an onion that have different chemical compositions. In them the successive thermonuclear reactions are taking place simultaneously: silicon burning at the center with neon, oxygen, carbon, helium, and hydrogen burning in successive concentric shells outward, as depicted in Figure 18.11. Surrounding these shells is a highly distended hydrogen-rich envelope. The star is very large, very bright, and quite red or possibly it is distinctly blue depending on its mass. Thermonuclear burning is a viable energy source for shorter and shorter periods as synthesis proceeds toward heavier elements. This is true in part because energy carried away by neutrinos becomes an increasingly greater fraction of the total energy with each successive burning stage beyond carbon burning, thereby reducing the energy available to supply the luminosity. For example, a 25M. star spends about 5 to 10 million years in hydrogen burning, 0.5 to 1.0 million years in helium burning, 500 to 1000 years in carbon burning, 6 to 12 months in oxygen burning, and a mere day or so in silicon burning.
[Figure 18.11]
For massive stars, once silicon burning commences, their end is near for these stars have exhausted all available nuclear fuels that have sustained them throughout their lives against the compressional forces of gravity. We believe that the result of nuclear fuel exhaustion is a supernova outburst with the principal question being what is the actual physical mechanism for this catastrophe?
18.5.1 Supernova Outbursts
For massive stars, the build-up of an inert, iron-rich core signals an impending death that is unbelievably violent. Supernovae outbursts are probably these violent events, predicted by astronomers' computer calculations of stellar models. All the details were still somewhat uncertain, in part because no star that later underwent a supernova outburst had been carefully studied before the outburst, until February of 1987 when supernova 1987A occurred (Figure 18.12). The progenitor of supernova 1987A, a blue supergiant called Sanduleak -69o 202, had not been investigated in great detail, but at least much was known about this star so that many questions could be answered. And as far as study is concerned, supernova 1987A attracted a great deal of interest by astronomers, so that it is undoubtedly the most studied supernova outburst in the history of observational astronomy. But let us return to the question of why the outburst takes place.
[Figure 18.12] In some stars, nuclear burning of carbon, oxygen, and silicon, if they occur in degenerate or nearly degenerate matter, can lead to a violent explosion, analogous to the helium flash in low-mass stars only larger. For stars, whose masses are between 8 and 12M., capture of degenerate electrons by carbon, oxygen, silicon, or iron nuclei lowers the pressure in the core that supports the weight of outer layers. This leads to a collapse of the matter in the core and its compression to approximately nuclear densities (1014 g/cm3). The matter that constituted the outer layers rebounds off the compressed core and it is eventually ejected into space as the observed supernova outburst. For stars with masses greater than 12M., temperatures at the end of nuclear burning can eventually reach 5 billion K at which point very high-energy, gamma-rays photons breakup iron nuclei causing the catastrophic collapse of the core. In turn, this is followed by the compression of core matter and the expulsion of the outer layers by a shock wave created when infalling material rebounds off a rigid core, as sketched in Figure 18.13.
[Figure 18.13] With the explosion of the star, gravitational potential energy is released both as radiant energy and as kinetic energy of ejected matter, liberating as much energy in seconds as the Sun expends in 108 years. As the outer layers of the star are blown off in the catastrophic explosion, a great flux of high-energy radiation and high-energy particles is released along with an avalanche of neutrinos. What is left is a rapidly spinning compressed remnant of the former star.
Although it might seem at first that what might remain of a star after a supernova outburst should be less than 1.4M. and thus be a white dwarf, this is not the case. Even as far back as the 1930s a few astronomers felt that the stellar remains were likely to be too massive to be a white dwarf. We now believe that all that remains of massive stars that have undergone a supernova outburst is one or the other of two of the strangest objects in all the cosmos. If the stellar core remnant is less than 3M., the collapse will halt at approximately the density of atomic nuclei with a radius on the order of several tens of kilometers, producing an object known as a neutron star. However, if the stellar core remnant is much greater in mass, nothing can stop the collapse and the end result is a black hole (Chapter 19). In the 1930s, neutron stars were predicted theoretically, but essentially forgotten about in the interim for lack of any evidence for their existence. Before considering neutron stars in detail, let us digress to see what revived interest in them.
18.5.2. Supernova 1987 A
No more exciting and scientifically significant event has occurred over the last decade in science than supernova 1987A, as it is known. Photographs taken on the night of February 23, 1987, of the Large Magellanic Cloud, a companion galaxy to our own Galaxy, at Canada's southern hemisphere observatory at La Silla, Chile, and at the Siding Springs Observatory in Australia, revealed a 6th-magnitude object where only 12th-magnitude blue supergiants stars had been observed before. We believe that the progenitor of supernova 1987A is a typical blue supergiant of spectral type B3, known as Sanduleak -69o 202--a catalog designation (Figure 18.14). Spectra of Sanduleak -69o 202 taken in 1977 do not suggest anything unusual was happening in the outer layers of the star prior to undergoing the supernova outburst. This is not surprising since the real changes were occurring deep inside in a relatively tiny portion of the star's radius.
The light curve for supernova 1987A was from the beginning odd for supernovae of either Type I or Type II. Most supernovae reach maximum brightness very rapidly and then slowly fade as shown in Figure 18.15. Supernova 1987A, however, initially brightened to a small fraction of a typical supernovae's brightness, quickly faded, and began to brighten again, reaching maximum brightness in late May. Spectroscopic evidence showed that supernova 1987A was a Type II supernova, one resulting from the explosion of a massive star. The outer layers were blown off with velocities near 20,000 km/s (Figure 18.15). Peculiarities in the light curve and the spectra of supernova 1987A may possibly be account for by the fact that it was a blue supergiant, a more dense and compact star, rather than a red supergiant before the outburst. Thus more energy from the initial explosion is taken up in blowing off the outer layers of a blue supergiant than is the case for a red supergiant.
Sanduleak -69o 202's position in the H-R diagram is consistent with its having had a main-sequence mass of about 20M.. Of course at the time of the outburst the mass could have been more like 15M. or less due to mass loss from stellar winds. There is even the suspicion that vigorous stellar winds during the star's red supergiant phase blowing off so much mass had made the remaining star into the blue supergiant we observed before the outburst took place. We may know more about this point over the next several years if the expanding supernova shock wave overtakes and collides with the expanding mass-loss material and causes it to glow.
In addition to the visible sighting of supernova 1987A, the supernova was also sighted at other wavelengths. For example, within hours radio telescopes had detected the supernova, and within a half-a-day the International Ultraviolet Explorer satellite had detected rapidly fading ultraviolet radiation. The Japanese Ginga satellite and the Kvant astrophysics module aboard the Soviet Mir space station detected hard X-rays that continued to increase until about October at which time they leveled off. The Solar Maximum Mission satellite detected gamma rays from the supernova in August and October, while a ground-launched balloon in Australia in November detected both X-rays and gamma rays. These observations are consistent with the long-term behavior of the light curve. After the first 50 or so days, the light curve should be controlled by the radioactive decay of cobalt 56. This is according to our stellar models, which indicate that large amounts of radioactive nickel 56 should be produced in the collapse and rebound phase of the outburst. Nickel 56 has a half-life of 6.1 days and decays into cobalt 56, which in turn has a half-life of 77 days. Gamma rays emitted by the relatively long-lived cobalt as it decays into a stable isotope of iron powered the supernova after the first few weeks.
Probably the most exciting discovery about supernova 1987A was the detection of neutrinos created in the explosion deep inside the exploding star. About 2:36 EST on February 23, 1987, neutrino detectors in Ohio and Japan simultaneously recorded 19 neutrino events, 11 in Japan lasting 12.2 s and 8 in Ohio lasting 5.5 s. Although these two facilities were not designed to detect neutrinos from supernovae, they did make the detection extending the usefulness of these facilities into a brand new experimental science. These neutrinos are the first direct confirmation of supernova outburst theory. Astronomers were elated not only at the detection of the neutrinos, but also how closely the characteristics of these neutrinos agree with theoretical predictions. The derived temperature for the supernova outburst is about 50 billion Kelvins. Estimates are that these neutrinos and the others emitted by the supernova carry away about 1053 ergs of energy, which is about what the entire Galaxy emits in ten years.
It is obvious that we have not learned all that this spectacular event can tell us. So the intense study of supernova 1987A will continue for several more years yet in hopes that it still has more tell us about supernovae.
18.5.3 Pulsars, Cosmic Lighthouses
In 1967, a strange object in the constellation Vulpecula was discovered to be emitting pulses of radio radiation with a very short period between pulses of 1.337 seconds. Later the object was named a pulsar, an acronym for pulsating radio star. Within a few weeks after the discovery, three more pulsars were found, and the count now stands at over 300. The periods between pulses range from a few milliseconds (0.001 second) to nearly a thousand seconds, the majority having periods of a few seconds.
The time interval between pulses is very constant--for some pulsars about one part in 10 million. A clock having this accuracy would gain or lose no more than one second in a year. Because the duration of each pulse is very short, the emitting region responsible for the pulses is quite small. If the emitting region were large, that is spread in distance along the line of sight, pulses of one wavelength of radiation leaving at the same time from various points along the line of sight should arrive on Earth at different times, consequently blurring the pulse. For example, if the observed pulse duration is 0.1 ms, the size of the emitting region cannot be larger than the distance light can travel in that interval of time, about 30 km (20 mi), and the pulse still remain sharp. Astronomers had thought up to that time that white dwarfs were the smallest, most compact stars, being some 5000 to 10,000 km in radius. Therefore, what could this object be? In the late 1960s the dormant prediction of neutron stars was revived and astronomers now believe the pulsar to be a neutron star.
Electrons in interstellar space affect the velocities of radio waves: the longer the wavelength, the slower the wave's velocity. Radio waves of different wavelengths that are bunched together in the form of a sharp pulse are therefore spread out along the way, arriving on Earth at different times; shorter wavelengths arrive before longer ones do. We can estimate the distance to a pulsar from this delay time. Calculations of this kind indicate that most pulsars are relatively nearby Galactic objects, at an average distance of about 3000 ly most of them being in or near the Galactic plane.
Some pulsars have been found to pulse not only in radio wavelengths but also in X-rays, some in gamma rays, and some in visible wavelengths. Of all the pulsars discovered, most having been found through their radio-wavelength pulsing, only four have been conclusively identified with a visible object. Of these, three are located in our Galaxy: one in the Crab Nebula, a second in the Vela supernova remnant, and the third in the constellation Vulpecula. The fourth is in a neighboring galaxy, the Large Magellanic Cloud.
Not long after their discovery, a number of pulsars were found to be slowing their rate of pulsing by a few tens of millionths of a second per day. Thus given enough time, the pulsing should cease entirely. For this reason the pulse phenomenon is apparently only something that happens early in the life of a neutron star, so that old neutron stars do not distinguish themselves by being pulsars and so go undiscovered. Also the Crab pulsar, as well as others, is not at the geometric center of the surrounding nebula. If the supernova outburst was not symmetrical, the neutron star may have been forced out in one direction and the expanding nebulous shell in the opposite direction, an example of Newton's third law. If enough time has elapsed since the supernova outburst, the neutron star may no longer even lie inside the nebula, and this, along with the cessation of pulsing, may in part explain why no pulsar has been found associated with most supernova remnants.
18.5.4 Neutron Stars
Astronomers now agree that pulsars are rapidly rotating neutron stars, one form for the stellar core that remains after a supernova outburst. During the collapse of a star's core, leading to a supernova outburst, electrons and protons are jammed together to form neutrons. Neutrons crowded together in the core of a star are subject to the exclusion principle (Section 17.1), just as are electrons, so they can become degenerate also. As such, they can exert a pressure outward that balances the weight of the overlying layers if it is not too great. Thus if the small, dense object left after the supernova outburst is between about 1.4 and 3M., pressure produced by degenerate neutrons can balance the weight of matter and stop any further collapse. Such an object is a neutron star and is, astronomers believe, permanently stable.
Conserving whatever rotational motion the pre-supernova star had, the neutron star spins rapidly because it is so much smaller than the original star. Only a body 10 to 30 km in diameter with a density approaching nuclear densities (approximately 1014 g/cm3) could survive the disruptive force of such rapid rotation. Also, if it conserves its original magnetic field of even a few tens of gauss, the enormous reduction in size amplifies the magnetic field to incredible intensities. The magnetic field in neutron stars may run as high as 1013 G. By comparison, the Earth's magnetic field intensity is 0.5 G and the largest magnetic field produced in a laboratory is about 300,000 G.
The neutron star's high rotational speed provides a reservoir of energy that can power a continuous flow of charged particles streaming from the star's magnetic poles, as shown in Figure 18.16. Such a flow of charged particles emits coherent electromagnetic waves in a highly directed cone of radiation that spins with the neutron star. When this searchlight beam sweeps across our line of sight, we see a pulse of radiation from the neutron star every fraction of a second or so, or we witness a pulsar. And if not, the neutron star is not otherwise observable, since it is so very small.
[Figure 18.16] The magnetic field should slow the pulse rate by ultimately converting the neutron star's rotational energy to electromagnetic radiation. This prediction has been confirmed by observations for over 90 pulsars that show a decrease in the pulse rate. Pulsars with the shortest periods are thus apparently the youngest ones. The average age of the remaining pulsars is about 2 million years; the oldest pulsars are about 10 million years old. Our failure to detect optical pulsations in all but four pulsars may mean that light flashes are a transient phenomenon, occurring only during a pulsar's extreme youth.
In the neutron star, one might think that we have the strangest of all astronomical bodies. But alas, the black hole is even more strange and at greater odds with common experience. Let us move to the next chapter and begin the discussion of the black hole.