[4.0] Stellar Mechanics & Evolution (2)

v3.3.0 / chapter 4 of 10 / 01 feb 16 / greg goebel

* This chapter follows up the previous chapter to discuss the explosive end of very large stars as well as other classes of exploding stars, and the very dim substellar "brown dwarf" stars.

Veil Nebula, a supernova remnant



* The evolution of stars more massive than about 8 to 10 Suns is particularly interesting and spectacular. The bigger a star is, the brighter it is, since the cores must produce more energy to resist the pressure of their greater mass. A star with a mass of 20 Suns is 20,000 times brighter than our Sun, and uses up its hydrogen fuel much faster, turning into a red giant in 10 million years instead of 10 billion.

The high core pressures of such massive stars lead to further nuclear fusion processes. At 500 million degrees Kelvin, carbon nuclei fuse into neon, magnesium, and sodium. At a billion degrees Kelvin oxygen nuclei fuse into silicon, sulfur, and phosphorus. At two billion degrees Kelvin, silicon nuclei fuse to create iron. Intermediate stages of fusion and nuclear decay produce other elements, but none heavier than iron.

These nuclear reactions can be summarized as follows:

   C12 + C12 -> Ne20 + He4 (alpha_particle)
   C12 + C12 -> Mg24 + gamma_ray
   C12 + C12 -> Na23 + proton

   O16 + O16 -> S32 + gamma_ray
   O16 + O16 -> P31 + proton
   O16 + O16 -> S31 + neutron
   O16 + O16 -> Si28 + He4 (alpha_particle)

   Si28 + Si28 -> Ni56
   Ni56        -> Co56 + positron + neutrino
   Co56        -> Fe56 + positron + neutrino

The massive star develops a series of concentric shells, with hydrogen fusion at the top shell, helium fusion in the next lower shell, and an iron core accumulating at the bottom. Each stage of fusion burning is shorter than those before it, with helium burning lasting on the order of a million years, carbon burning about 10,000 years, and silicon burning on the order of 10 days. With each stage of fusion reactions, the star grows hotter and more extended. Its large, tenuous outer atmosphere is strongly affected by any internal disturbances, resulting in eruptions of gas and dust that grow more violent as the oversized star ages.

The star will not perform iron fusion. Iron occupies a crucial position in the periodic table. A nucleus is held together by the strong nuclear force, which overcomes the mutual electromagnetic force repulsion of the positively-charged protons in the nucleus. However, the strong nuclear force is short range, acting between adjacent atomic particles while the electromagnetic force is long range, with particle couplings through the entire nucleus. As the number of protons in a nucleus builds, the repulsive electromagnetic forces tend to overbalance the attractive strong force. Iron is at the balance point. For elements above iron, energy is not released by their synthesis; it takes a net input of energy to synthesize them.

Once the massive star builds up an iron core and begins to run out of fuel in the outer shells, it is doomed. Its death is accelerated by the fact that at the increasingly high core temperatures set up by successive stages of fusion processes, neutrino emission increases. Neutrinos flood out of the core unobstructed by the higher layers, robbing the star of energy.

The end comes quickly. The star collapses in on itself, producing nuclear fusion and fission reactions that absorb energy and accelerate the collapse, quickly compressing the stellar core to incredible densities. The core collapse is announced with a huge burst of neutrino emission. The core is squeezed so heavily that it forms into a tiny but ultradense "neutron star". The biggest stars form an infinitely tiny and dense "singularity" that wraps spacetime around itself, becoming a "black hole" in space.

A neutron star or black hole cannot be compressed further, and following the burst of neutrino emission, the infalling matter "rebounds" in a "core bounce", resulting in a tremendous explosion that tears the star apart. This is a Type II supernova, and it can be as bright as a billion Suns for a few weeks, emitting heavily in the ultraviolet. The nova observed by the Chinese in 1054 CE was a Type II supernova and resulted in the Crab Nebula, which has a neutron star at its core.

The destruction of the star spews its contents into space, throwing out carbon, oxygen, sulfur, and silicon. It has been generally believed that explosion itself creates the heavy elements, those above iron, transforming iron nuclei into gold through neutron bombardment, then similarly converting gold into lead, and turning lead into heavier elements all the way up to uranium. However, theoretical work on the synthesis of the elements above iron is sketchy at best, and some astrophysicists have challenged this viewpoint, suggesting that other cosmic events may be at least partly involved. One interesting model has envisioned that the collision of two co-orbiting neutron stars, producing an intense burst of gamma rays, might do the job, the neutron stars being a high-density source of neutrons for the synthesis of heavy elements. The matter remains an open field for ambitious theoreticians out to make a contribution to the field.

In any case, the end result of nucleosynthesis processes can be shown by the following table, which lists the abundances of the elements relative to a sample of 10^12 hydrogen atoms:

   element   atomic_number   abundance   element   atomic_number  abundance
   ___________________________________   __________________________________

   H   hydrogen    1 1,000,000,000,000   Tc  technetium    43            >1
   He  helium      2    80,000,000,000   Ru  ruthenium     44            68
   Li  lithium     3             2,000   Rh  rhodium       45            13
   Be  beryllium   4                30   Pd  palladium     46            51
   B   boron       5               900   Ag  silver        47            20
   C   carbon      6       450,000,000   Cd  cadmium       48            63
   N   nitrogen    7        92,000,000   In  indium        49            75
   O   oxygen      8       740,000,000   Sn  tin           50           140
   F   fluorine    9            31,000   Sb  antimony      51            13
   Ne  neon       10       130,000,000   Te  tellurium     52           180
   Na  sodium     11         2,100,000   I   iodine        53            33
   Mg  magnesium  12        40,000,000   Xe  xenon         54           160
   Al  aluminum   13         3,100,000   Ce  cesium        55            14
   Si  silicon    14        37,000,000   Ba  barium        56           160
   P   phosphorus 15           380,000   La  lanthanum     57            17
   Su  sulfur     16        19,000,000   Ce  cerium        58            43 
   Cl  chlorine   17           190,000   Pr  praseodymium  59             6
   Ar  argon      18         3,800,000   Nd  neodymium     60            31
   K   potassium  19           140,000   Pm  promethium    61            <1
   Ca  calcium    20         2,200,000   Sm  samarium      62            10
   Sc  scandium   21             1,300   Eu  europium      63             4
   Ti  titanium   22            89,000   Gd  gadolinium    64            13
   V   vanadium   23            10,000   Tb  terbium       65             2
   Cr  chromium   24           510,000   Dy  dysprosium    66            15
   Mn  manganese  25           350,000   Ho  holmium       67             3
   Fe  iron       26        32,000,000   Er  erbium        68             9
   Co  cobalt     27            83,000   Tm  thulium       69             2
   Ni  nickel     28         1,900,000   Yb  ytterbium     70             8
   Cu  copper     29            19,000   Lu  lutetium      71             2
   Zn  zinc       30            47,000   Hf  hafnium       72             6
   Ga  gallium    31             1,400   Ta  tantalum      73             1
   Ge  germanium  32             4,400   W   tungsten      74             5
   As  arsenic    33               250   Re  rhenium       75             2
   Se  selenium   34             2,300   Os  osmium        76            27
   Br  bromine    35               440   Ir  iridium       77            24
   Kr  krypton    36             1,700   Pt  platinum      78            56
   Rh  rubidium   37               260   Au  gold          79             6
   Sr  strontium  38               880   Hg  mercury       80            19
   Y   yttrium    39               250   Tl  thallium      81             8
   Zr  zirconium  40               400   Pb  lead          82           120
   Nb  niobium    41                26   Bi  bismuth       83             5
   Mo  molybdenum 42                93   Th  thorium       90             1
                                         U   uranium       92             1
   ___________________________________   __________________________________

   Elements 84 through 102 are mostly unstable and have negligible 
   quantities, so only thorium (90) and uranium (92) are listed here.

The irregular patterns of the abundances of the elements reflects the paths of nuclear fusion and decay. Lead, for example, is relatively common for a heavy element partly because it is the stable decay destination for many of the elements above it. Beryllium is uncommon because it is generally consumed in fusion reactions. Notice the somewhat inconsistent pattern where atoms with even atomic numbers are usually more common than those with odd atomic numbers. This is because the direct or indirect building blocks of many fusion reactions are helium nuclei / alpha particles, which have an atomic number of 2.

The materials produced by a Type II supernova explosion expand outward in a shell that, by cosmic standards at least, dissipates quickly, spreading its materials through the galaxy. While they last these remnants are spectacular, their tangled appearance giving little doubt to the most untrained observer that they were produced by an explosive event. The Crab Nebula is one of the youngest and most impressive such supernova remnants in our neighborhood. The Veil Nebula, the product of a Type II supernova explosion that took place 30,000 to 40,000 years ago, is much more diffuse, in the form of a broken ring of streamers.

It is unclear just how big stars can get. There is no consensus on any theoretical limit to their size, but a statistical analysis of the size of stars in young star clusters gives a curve that goes to zero somewhere between 120 and 200 Suns.

* Supernovas are rare events in any single galaxy. Supernovas are often observed in other galaxies, but intergalactic distances limit the amount of information astronomers can obtain from the event.

In the 1980s, astronomers were treated to a supernova at a relatively close distance that was observed by a wide range of instruments. On 23 February 1987, a supernova was spotted in the "Large Magellanic Cloud", another relatively small irregular satellite galaxy of the Milky Way. The supernova was associated with a star known before the explosion as "Sanduleak -69 degrees 202", or "Sanduleak" for short, after the astronomer Nicholas Sanduleak (1933:1990), who cataloged it in the 1960s.

The star named Sanduleak was born about ten million years ago as a massive star of about 18 Suns. As a main sequence star, it was about 40,000 times brighter than our own Sun. Of course, as its death grew near it had left the main sequence and become a supergiant, but puzzlingly before its death it was a hot blue supergiant, not a red supergiant as theory and observations of extragalactic supernovas indicated. Theorists suggest that the discrepancy may have been due to the different chemical composition of the molecular clouds of the Large Magellanic Cloud.

The expected burst of neutrinos from the supernova explosion, which was designated "SN 1987A", was observed in underground neutrino observatories on Earth, confirming theoretical models and implying the creation of a neutron star in Sanduleak's core. The light from the explosion was seen a few hours later by alert amateur and professional astronomers in the Southern Hemisphere, where the LMC can be observed.

Supernova 1987A

Orbiting observatories, such as the International Ultraviolet Explorer, the Japanese X-ray observatory satellite Ginga, the US Solar Maximum Mission satellite, and instruments on the Soviet Mir space station, were brought to bear on the event. Balloons with observational packages were flown from Australia and Antarctica, and NASA's jet-borne Kuiper Airborne Observatory infrared telescope performed observations as well. The observations verified the synthesis of new elements in the exploding star and the breakdown of unstable isotopes.

As of last report, no neutron star has been detected in the heart of the expanding cloud of gases. The event is by no means over, however. In 2003, the Hubble Space Telescope took a spectacular image of the shockwave expanding rapidly outward at the supernova, showing it to be brightly illuminated by energetic collisions with gas clouds in its path. The resulting "hot spots" formed a ring a light-year across that resembled pearls on a necklace. The first such hot spot was observed in 1996, but they have since multiplied. The region inside the ring was also glowing, mostly due to the breakdown of short-lived radioactive isotopes produced by the blast, particularly titanium-55. The ring is expected to become even brighter, possibly illuminating regions of space around it to permit detailed observations.

Supernova 1987A / pearl necklace

* Incidentally, the closest star likely to become a Type II supernova in the near (by astronomical terms) future is the red giant Betelgeuse, 427 light-years away. At that range the strong neutrino emissions at the beginning of the event could possibly cause some chromosomal damage to organisms on the surface of the Earth -- a testimony to the intensity of the neutrino flux, since neutrinos react so reluctantly with matter. The ultraviolet burst that would follow would be dangerous to spacefarers, but the ozone layer would protect the Earth.

The explosion of Betelgeuse would not threaten life on our planet, though it would give a fairly impressive lightshow -- shining at magnitude -12 for about three months, making it about as bright as the quarter Moon, or much brighter than any star or planet in the night sky. If Betelgeuse were only a tenth as far away, we might be in big trouble.



* The periodic eruptive variables known as "novas" and "dwarf novas" and the explosive events known as "Type I supernovas" have no direct relationship to Type II supernovas. All three of these phenomena are associated with binary systems in which a white dwarf is in a close orbit with a conventional star.

The conventional star in the binary system grows in diameter as it ages, and if it grows beyond a certain diameter defined by the "Roche limit", at the "Lagrangian midpoint" where the gravity between the two stars is balanced, the white dwarf will strip off material, mostly hydrogen, from the conventional star. The outer atmospheres of both stars will be distorted into teardrop-shaped "Roche Lobes" that meet at the Roche limit, with the conventional star losing mass and the white dwarf gaining it. The drag caused by the mass transfer will also cause the two stars to fall closer together, increasing the mass transfer rate. However, once they attain equal mass, the star that has lost mass, now being smaller, gradually moves out into a wider orbit as dictated by the conservation of angular momentum.

The evolution of two close binary stars is necessarily a complicated subject, the results differing depending on the types of the two stars involved, and in fact the problem is in some cases more complicated than astrophysicists can handle at the present time. In case of concern here -- a white dwarf in close orbit with a conventional star -- the effects are understood and very spectacular. The mass will flow in a spiral into a spinning "accretion disk" around the white dwarf, heated by friction to temperatures to produce X-rays, and will fall out of the inner edge of the accretion disk to accumulate on the surface of the white dwarf.

Once the hydrogen builds up on the surface of the white dwarf to a certain level, a fusion reaction begins that releases energy and increases the rate of fusion reactions, leading to a "runaway" process and a huge explosion that increases the brightness of the white dwarf to hundreds or thousands of times that of our Sun. This is a nova. A shell of material with a mass of about a ten-thousandth or hundred-thousandth that of the Sun is also ejected. The binary system fades to normal a few months after this explosion, and material begins to build up again, leading eventually to another explosion. The periods are irregular, but unsurprisingly the longer the period the bigger the explosion, since more material has accumulated.

recurring nova RS Ophiuchi

Dwarf novas are similar binary systems, but the periodic explosion occurs in the accretion disk, not on the white dwarf, due to some instability in the accretion disk that is not well understood. Dwarf novas are about one or two orders of magnitude fainter than true novas.

* Type I supernovas also occur in such binary systems, but they go to the other extreme in brightness, producing the light of billions of Suns. They occur because, as discussed in detail later, there is a limit of 1.44 Suns to how massive a white dwarf can be. Above this limit, gravity forces the white dwarf to collapse in on itself.

Now suppose a large white dwarf, not far below the mass limit, is accumulating mass in a binary system as does a nova star. If the white dwarf does not contain large quantities of carbon, once it reaches a mass of 1.44 Suns, it will collapse in on itself, becoming a superdense neutron star.

This is a more or less nonviolent process, at least in comparison with what happens if the white dwarf contains large quantities of carbon. Once such a carbon-rich white dwarf approaches -- but does not actually reach -- the mass limit of 1.44 Suns, carbon in the core of the white dwarf begins to fuse into iron, cobalt, and nickel. The burning expands outward from the core in a subsonic "deflagration wave", gradually losing energy and then only synthesizing lighter elements, such as magnesium, silicon, and sulfur. Once the deflagration wave travels about 70% of the distance from the core of the white dwarf to its surface, however, the pressure of the layers of the star above the wave is no longer adequate to confine the burning process. It then goes into a supersonic "detonation mode" that effectively blasts the white dwarf apart in a brilliant explosion within seconds.

This is a Type I supernova, and the explosion is even brighter than that of a Type II supernova. Tycho's nova of 1572 and Kepler's nova of 1604 were both Type I supernovas. What was believed to be the surviving companion star of Tycho's nova was discovered in 2004. It was a star on the red giant path of stellar evolution, cast off at high proper motion when its white dwarf companion blew itself up; it showed signs of having been disrupted in the past few centuries.

Interestingly, the expanding fireball retains the structure of the original thermonuclear combustion process, with the heavier elements in the center and the lighter elements outside. Another interesting feature is that the brilliance of the event is not due to the explosion itself, at least not directly. The Type I supernova converts about 40% of the white dwarf's mass into nickel-56, which is an unstable isotope that decays into cobalt-56 with a half-life of 6.1 days. Cobalt-56 is also unstable, decaying into iron-56 with a half-life of 77.1 days. The breakdown of these two isotopes emits floods of gamma rays, which heat up the rest of the material being blasted out from the wrecked white dwarf to about 10,000 degrees Kelvin. The Type I supernova reaches its maximum brightness about 20 days after the destruction of the white dwarf.

All Type I supernovas have, roughly, the same brightness. This is because a Type I supernova occurs when the mass of the white dwarf approaches the 1.44 solar mass threshold, and so Type I supernovas amount to "bombs" of a limited range of size. This predictable brightness allows them to be used as reliable (as such things go) standard candles for intergalactic distances.

The two types of supernovas can be distinguished. Type I supernovas fade in brightness more rapidly than Type II supernovas, and since hydrogen has been generally exhausted on the white dwarf, the spectra of Type I supernovas exhibits little or no hydrogen. Type I supernovas with strong silicon spectral lines are subclassified as "Type Ia" supernovas, while those with strong helium lines are "Type Ib" supernovas.

The "Type Ic" supernovas lack hydrogen and helium lines and are associated with what are known as "Wolf-Rayet" stars. As mentioned earlier, as the energy production of a star increases as it ages, the result is greater radiation pressure and a stronger "stellar wind". For stars of 30 solar masses or more, the result is that once the star goes into the red giant phase the hydrogen envelope and then the helium envelope is quickly blown away, leaving only a core composed of metals. In fact, for stars of 50 solar masses or more, the star never actually becomes a red giant at all. In either case, the result is a Wolf-Rayet star, which has little or no hydrogen and helium in its spectrum and has lost much, even a half or more, of its mass.

* Incidentally, although the Type I supernova model described above is widely accepted, there is some debate over its validity. A minority of astronomers suspect that Type Ia supernovas are caused by the collision of two white dwarfs in a binary system. The main reason for this belief is the lack of hydrogen in the spectral lines -- since the normal companion star should be rich in hydrogen, while white dwarfs are lacking. The discovery of the companion of Tycho's nova has thrown some cold water on that idea.

Even among those who accept the popular model, there are questions about the details. For example, the mass accretion rate onto the white dwarf can't be too fast or too slow, since various mechanisms such as stellar winds or thermonuclear flashes would otherwise tend to eject the mass buildup into space. However, the mass flow rate between a white dwarf star and a normal star companion will change over time, so it is plausible that such binary systems will pass through an era, an "evolutionary bottleneck", where the mass flow is at the precisely the right level.

Another interesting consideration is the possibility that the process of accumulating mass onto the white dwarf transfers angular momentum to it, causing it to "spin up" and not detonate even if it exceeds the Chandrasekhar limit -- at least until it slows down again and centripetal force no longer cancels out the mass surplus. Theorists also agree that the current models of the white dwarf detonation process involve "a lot of hand-waving", and so work continues on more rigorous models.

There is, incidentally, a candidate for a Type Ia supernova closer to Earth than Betelgeuse. The binary system "IK Pegasis", only 150 light-years away, consists of a high-mass white dwarf in orbit around a old star on the track towards becoming a red giant. The white dwarf is not gaining mass from its partner just yet, but it is likely to happen sometime in the near future ("near" in cosmic terms at least). An explosion at that range could seriously damage the Earth's ozone layer.

* In January 2002, the Hubble Space Telescope observed a novalike event that didn't fit into any known categories. The star "V838 Monocerotis (V838 Mon)", ballooned into a cool supergiant and brightened by a factor of about 600,000 for a time. For about 40 days it was the most luminous star in the Milky Way.

eruptive variable V838 Monocerotis

V838 is an unstable star that has shed multiple clouds of dust, which now form a multilayered cocoon around the star. The Hubble observed these shells as the light from the outburst propagated outward, illuminating layers of the cocoon in sequence, allowing astronomers to learn more about the structure of this unusual star system.



* Supernovas are spectacular enough, but they are now seen as the underlying cause of an even more spectacular phenomenon, the "gamma-ray bursters (GRBs)". Cosmic gamma-ray bursts were spotted in the late 1960s by the US "Vela" nuclear test detection satellites. The Velas were launched to detect radiation emitted by weapons tests, but they had a secondary mission to observe astronomical events that might be confused with weapons tests; they did indeed pick up occasional bursts of gamma rays that weren't matched to any verified weapons tests. Although the sensors on the Vela satellites had low angular resolution, in 1973 researchers at the US Los Alamos National Laboratory in New Mexico were able to use the data from the satellites to determine that the bursts came from deep space.

Astronomers believed that once better gamma-ray detectors were put in orbit, they would be able to quickly pin down the locations of the GRBs -- after all, that's what happened with X-ray sources. However, when such improved detectors were sent into space in the 1970s, optical searches of the regions where the bursts originated showed nothing of interest. The sensors were not accurate enough to pinpoint the location of the bursts for detailed inspection.

Further information on the burst sources proved hard to obtain and led to more questions than answers. The first question posed by the GRBs was: were they local to our own Galaxy, or did they occur in the distant reaches of the Universe? The second question was: what mechanism caused the bursts? If they did occur in the distant Universe, the mechanism must be producing enormous amounts of energy.

* It wasn't until the 1990s that real progress was made on the matter. In 1991, NASA launched the "Compton Gamma Ray Observatory" satellite. One of the instruments on board Compton was the "Burst & Transient Source Experiment (BATSE)", which could detect gamma-ray bursts and locate their positions in the sky with reasonable accuracy. Within a year, BATSE determined that GRBs occur about once a day and are randomly distributed over the entire sky. If they were events occurring in our own Galaxy, they would cluster in the plane of the Milky Way. Even if they were associated with the galactic halo, they would still be preferentially distributed towards the galactic center -- in just the same way that an observer in city suburbs will notice more bright lights towards downtown than towards the city limits -- unless the halo was truly enormous.

Even if that was the case, nearby galaxies would be expected to have similar halos, but they did not show up as "hot spots" of faint gamma-ray bursts. To many astronomers, this finding implied that the GRBs originated in the distant Universe, but that led to the problem of finding a mechanism that could generate so much energy.

By the late 1990s the local hypothesis for GRBs had been ruled out. The first clue came from the Italian-Dutch "BeppoSAX" satellite, launched in 1996, designed to pin down the location of bursters. In 1997 and 1998, BeppoSAX managed to pin down the locations of several GRBs quickly enough to for inspection by other instruments, including large ground-based telescopes. Spectroscopic analysis showed that the GRBs were clearly occurring in the distant Universe, and images suggested a possible link between the bursters and supernovas.

On 23 January 1999, the Compton GRO satellite detected a burster and sent out an alert to tip off ground-based observatories. The event, designated "GRB 990123", was one of the brightest GRBs observed. It was so bright that its visible-light component could have been picked up by binoculars. The burst rapidly faded out, but not before it had been carefully inspected by the world's most powerful telescopes. Analysis of its redshift showed it to be about 9 billion light-years away -- how a redshift reveals cosmic distance is discussed later.

At 9 billion light-years, the gamma-ray energy released by the burster was the equivalent of converting the entire mass of a star 1.3 times the mass of our Sun completely into gamma radiation. At visual wavelengths, if the burster had gone off in our own Galaxy 2,000 light years away, it would have shined twice as bright in our night sky as the Sun does during the day. Since that time, several other satellites have been launched to spot GRBs, and many more have been found, including events more powerful than the 1999 burst.

* GRB light curves vary all over the map, no two being alike. There may be multiple peaks; or a "precursor event" leading up to the burst; or the light curve may be very noisy and chaotic. They can be classified as "long duration" -- the majority of events, lasting more than two seconds -- or "short duration" -- lasting less than two seconds.

It is thought that the long-duration GRBs were caused by supernovas. Initially, the supernovas were thought to be much more powerful than ordinary supernovas and were called "hypernovas", but that idea wasn't universally accepted, primarily because nobody could figure out how such energy levels could be produced. If a narrow beam of gamma rays were emitted during the collapse, a supernova only slightly more powerful than average could account for the observed energies, though theorists found it difficult to figure out how such a narrow beam would be produced. One initial explanation was that the beam was focused by a "gravitational lens", caused by the distortion of space by a large galaxy between Earth and the GRB, but statistical analysis later showed that lensing is too rare to account for the phenomenon.

Studies of the few such long-duration GRBs that were close enough for reasonably detailed examination showed they were derived from Type Ic supernovas, or exploding Wolf-Rayet stars. Computer models showed that the collapse of a Wolf-Rayet star into a black hole would produce tightly coiled electromagnetic fields that would produce high-energy jets, burning through the upper layers of the star to produce a gamma-ray burst. The subsequent dissipation of the jets would result in an afterglow. The probable mechanism for the short-duration GRBs does not involve supernovas and is discussed later.

Incidentally, theorists suggest that if the Earth were hit by a gamma-ray beam from a burster not too far away, it would incinerate half the surface of our planet. Since short-duration GRBs produce much less total energy than long-duration GRBs, long-duration GRBs were seen as the real threat. Fortunately, it appears that long-duration GRBs are usually born from big first-generation stars, low in "metals" content, with the bursts increasingly damped as the "metals" content rises. Since big, short-lived stars in our own Galaxy are several generations past the first and so rich in "metals", GRBs are no longer seen as a much of a menace to Earth.

The narrowness of the beam also suggested that only one in 500 GRBs are seen from Earth, meaning they are a fairly common phenomenon in the Universe. This means that astronomers might be able to observe "orphan afterglows", exactly like those following a GRB, but not associated with a gamma-ray burst.

* The puzzle of GRBs was underlined on 17 August 2017, when the two LIGO gravitational-wave observatories in the US and the Virgo gravitational-wave observatory in Italy picked up a gravitational disturbance, and were able to identify the general direction of the event. NASA's Fermi gamma-ray space telescope then spotted a long-duration GRB, enduring for about a hundred seconds, in that region of sky. Optical telescopes got in on the act, and helped confirm that the event was due to a pair of binary neutron stars falling in on each other. Observations also confirmed that at least some heavy elements were generated in such cataclysmic events.



* Although astronomers have achieved a remarkable knowledge of the stars, new discoveries continue to be made. One of the most recent and significant has been the discovery of the substars known as "brown dwarfs".

There is a certain critical threshold of stellar mass required to initiate hydrogen fusion. A cosmic object smaller than about 7% the mass of our own Sun, or equivalently smaller than about 75 times the mass of the planet Jupiter, cannot initiate hydrogen fusion. Deuterium fusion does occur if the object is larger than about 13 times the mass of Jupiter, but since deuterium is much less abundant than hydrogen, all of the deuterium will be burned up in a few million years.

The idea that there must be cosmic objects bigger than any planet we know about and smaller than a true star seems fairly obvious, but the first person to pay serious attention to the idea was Shiv Kumar of the University of Virginia, who speculated in 1963 that such objects may be common, with some of them likely dating back to the formation of the first stars and galaxies. At first, these objects were called "black stars" or "infrared stars", but in 1975 astrophysicist Jill C. Tarter (born 1944) named them "brown dwarfs", and the name stuck. In reality, a "brown" dwarf actually glows dull red, from the slow dissipation of heat left over from its formation, but the term "red dwarf" was already in use for low-mass true stars.

Astronomers hunted for brown dwarfs through most of the 1980s, but a brown dwarf wasn't nailed down for certain until 1995. Now brown dwarfs are a lively "cottage industry" for astronomers and astrophysicists as they try to determine how common they are, their range of size, and their life histories.

* The hunt for brown dwarfs was troublesome for the simple reason that they are very faint. The reddish light emitted by a brown dwarf does not stand out, and grows steadily redder and fainter over time as the brown dwarf cools. Most of their emission is in the infrared region of the spectrum, which is difficult to observe from the ground. Brown dwarfs also have small diameters compared to true stars, giving them less surface from which to emit light. Interestingly, though brown dwarfs are by definition much heavier than Jupiter, the diameter of a brown dwarf is actually close to that of Jupiter, since the core of the brown dwarf is in a superdense "degenerate" state.

One approach that astronomers hoped to use to catch a brown dwarf was to observe visible stars in our cosmic neighborhood to see if they had any just-barely-visible companion stars. One possible brown dwarf detected in this way in 1984 turned out to be an observational glitch; similar follow-up observations in the near term gave ambiguous results.

Another approach was to measure the Doppler shift of a star to detect a wobble that indicates the presence of a hidden companion. This approach has been successfully used to discover planets around distant stars. Early surveys along such lines from the late 1980s into the mid-1990s came up zeroes, however. Given the apparent scarcity of brown dwarfs, astronomers began to speak of the "brown dwarf desert".

Other astronomers were taking a third approach. As mentioned, brown dwarfs start out relatively bright and then grow dimmer over time. This implied that one of the best places to look for a brown dwarfs would be in a young open star cluster, before they had time to fade. The brown dwarfs would still be the dimmest, coolest objects in the star cluster. Studies of the bright Pleiades cluster and the Taurus cluster, both well known to amateur astronomers, turned up a number of candidates. However, once again, no brown dwarfs were found. Astronomers began to wonder if the "brown dwarf desert" was barren indeed.

* Meanwhile, on the assumption that brown dwarfs did exist, astronomers had been considering the problem of how to figure out if a dim cool object was actually a brown dwarf and not a red dwarf. In 1992, a group of Spanish astrophysicists working at the Astrophysics Institute in the Canary Islands came up with what became known as the "lithium test". At stellar core temperatures slightly below those at which hydrogen fusion begins, a lithium-7 breakdown process takes place, with an energetic proton reacting with lithium-7 to break down into two helium-4 nuclei.

Even in very low mass stars with very long lifetimes, the lithium-7 breakdown process proceeds rapidly, by stellar standards at least, with all traces of lithium-7 destroyed within 100 million years or so. However, a brown dwarf below about 60 times the mass of Jupiter will never reach the temperatures needed to initiate the breakdown process, and so lithium-7 will persist indefinitely. If lithium-7 is present in a cool star or starlike object, its spectral line is easy to spot in the infrared.

At first, sky surveys that looked for cool objects with lithium lines went nowhere, but another inspection of the Pleiades cluster discovered a dim glowing object that had lithium lines in its infrared spectrum. At last, astronomers seemed to be closing on a brown dwarf. The discovery of this candidate brown dwarf, which was named "PPl 15" -- the 15th object listed in the Palomar Pleiades survey -- was publicly announced in 1995. Incidentally, later work showed that PPl 15 is actually a pair of brown dwarfs in orbit with each other, with the remarkably short orbital period of six days.

* As often happens in science, parallel lines of investigation began to pay off at about the same time. The Canary Islands group also performed a search of the Pleiades, and discovered two objects that were even fainter than PPl 15, named "Teide 1" and "Calar 3" after Spanish observatories. Analysis showed that the mass of each was under 60 Jupiter masses, and further observations confirmed the presence of the lithium-7 spectral line.

Another group, a collaboration between astronomers from the California Institute of Technology and Johns Hopkins University, conducted a survey using the older strategy of observing individual stars to see if faint brown dwarf companions could be spotted. They examined nearby red dwarf stars using the 1.5 meter telescope at Mount Palomar in California, with the telescope fitted with a device that masked out the light of the target star in hopes of revealing faint companions.

The group performed such a survey in 1993, and repeated it in 1994. Since nearby stars as a rule have fast proper motions across the sky, any object associated with any of these stars moves along with them, while faint background stars remain where they are. The group found that the red dwarf "Gliese 229A" had a companion that was a thousand times fainter than its parent star. The infrared spectrum of the object, which was of course designated "Gliese 229B", revealed strong methane lines. Methane, or any other molecule for that matter, is never found in true stars since it breaks down at high temperatures, and the presence of methane confirmed that Gliese 229B was a brown dwarf.

Both the groups announced their discoveries in 1995. The astronomical community had been cautious about brown dwarfs because of the past history of false alarms, but the number and detail of the new discoveries proved persuasive.

   brown dwarfs compared

   object        type               mass    radius  temperature  age

   JUPITER       gas giant planet   1       1       100          4.5
   GLIESE 229B   brown dwarf        30:40   0.9     1,000        2-4
   TEIDE 1       brown dwarf        55      2.1     2,600        0.12 
   GLIESE 229A   red dwarf          300     3.5     3,400        2-4
   SUN           yellow dwarf       1,000   9.7     5,800        4.5

   Mass and radius are given as multiples of Jupiter mass and radius.
   Temperature is in degrees Kelvin.  
   Age is in billions of years.

* After 1995, discoveries of brown dwarfs accelerated. Searches using wide-area infrared observations to spot isolated "field" brown dwarfs began to pile up increasing numbers of candidates, suggesting that the "brown dwarf desert" was not as barren as it had seemed. Over a thousand brown dwarfs have now been identified; astronomers estimate that there might be about 100,000 million brown dwarfs in the Milky Way -- though given their relatively small size, they only made up a tiny fraction of the mass of the Galaxy.

brown dwarf

As the number of brown dwarf candidates increased, astronomers began to investigate the variation in their masses to see if they formed a continuous range with planets, or if there was a "mass gap" between the masses of the biggest planets and the smallest brown dwarfs that hinted at different formation processes. Surveys indicated that brown dwarfs seemed to have masses down to the theoretical lower mass limit, 13 times that of Jupiter, and that no such mass gap existed.

Enough brown dwarfs have been found to permit a classification scheme:

The brown dwarf classes form a continuum into M-type red dwarfs -- though since red dwarfs burn hydrogen, and do it for a long time, they have to be judged as true stars.

* The burst of initial discoveries of brown dwarfs is now over. Brown dwarfs are now well-recognized citizens of our Galaxy. Astronomers are no longer struggling to confirm the existence of such objects, and are instead methodically searching them out and revealing their secrets.

Observations performed in 2000 by the orbiting US NASA Chandra X-ray observatory -- named after Subrahmanyan Chandrasekhar, who was known to his colleagues simply as "Dr. Chandra" -- have shown that brown dwarfs are capable of producing X-ray flares, implying they have strong magnetic fields. That's puzzling, since a magnetic field should imply continuous X-ray emission at a lower level, which hasn't been observed. Brown dwarfs clearly have many surprises in store for astronomers.