¿ìè¶ÌÊÓƵ

ON 24 FEBRUARY 1987, astronomers in Chile and New Zealand discovered
the first supernova in almost 400 years that was bright enough to be visible
to the naked eye. Exactly three years on, Supernova 1987A is still producing
surprises.

The supernova lies 170 000 light years away in the Large Magellanic
Cloud, the nearest galaxy to the Milky Way. It is the closest supernova
seen since the invention of the telescope, so astronomers have been able
to study it in unprecedented detail.

Two of the most important discoveries came in the first couple of days
after the supernova was discovered. To their delight, astronomers realised
that they had catalogued and examined the dying star before it exploded.
This meant that, for the first time, they knew exactly what kind of star
had ‘gone supernova’. It was a blue supergiant star, about 20 times the
Sun’s mass and 50 times its diameter. Also for the first time, physicists
detected a stream of neutrinos, emitted by the collapsing core of the exploding
star.

The past three years of study have laid bare some of the details of
the explosion itself, and also illuminated – quite literally – the star’s
neighbourhood in the Large Magellanic Cloud.

Although astronomers saw the star explode in February 1987, it continued
to brighten for a further three months. Theorists suggested that this increasing
brightness was caused by energy released by radioactive nuclei as they decayed.
In particular, the explosion could convert some oxygen nuclei into unstable
nickel-56, which would decay into cobalt-56 over a period of a few weeks.
The cobalt-56 is itself unstable, and decays into stable iron-56 with a
half-life of 77.1 days, which means that it can supply power for many months.

The first direct evidence for this theory came towards the end of 1987,
when satellites picked up gamma rays, characteristic of the nuclei of cobalt-56.
Since then, the expanding gases have become sufficiently transparent for
astronomers to be able to observe spectral lines from cobalt atoms at optical
and infrared wavelengths. The strength of these lines suggests that the
supernova produced a mass of cobalt equal to about 8 per cent of the Sun’s
mass. This is reassuring to astronomers, because it is close to the amount
their theories require in order to explain the peak luminosity of the supernova
in May 1987, when it was 250 million times as bright as the Sun.

Now, for the first time, astronomers from Imperial College, London,
and the Anglo-Australian Observatory in New South Wales have observed directly
the decay of a radioactive substance in another galaxy. Using the large
Anglo-Australian Telescope, they have observed the spectrum of supernova
1987A in the infrared, at a wavelength near 1.5 micrometres. Here, they
can see a spectral line of cobalt that is close to a line produced by iron.
In a series of spectra taken between 255 and 574 days after the explosion,
the cobalt line clearly fades in strength: by day 574, it has dwindled virtually
to nothing.

The spectra also show that the explosion was ‘mixed and lumpy’, according
to Peter Meikle of Imperial College. Each spectral line is spread out in
wavelength by the doppler effect.

If the gas in the explosion was distributed uniformly within the expanding
shell of the supernova, the spectral lines would have a smooth shape. But
astronomers have found that this is not the case. They conclude, therefore,
that gas is clumpy.

Stan Woosley, of the Lick Observatory in California, had predicted just
such a clumpy distribution. His calculations had shown that, in the first
few weeks after the explosion, the energy from the fast-decaying nickel-56
should create a very hot ‘bubble’ in the middle of the expanding gases.

The pressure in the nickel bubble would have been so high that it would
have forced fingers of this hot, dense material outwards through the outer
layers of gas. These ‘nickel bullets’ would have made the explosion clumpy,
and mixed material that was deep inside the star with gases that were originally
nearer the surface.

During the year after the explosion, most of the energy of the supernova
was radiated as light: the energy from the decay of radioactive nickel,
then cobalt, was largely trapped inside the exploding star and served to
heat the gas, making it glow. As a result, the supernova’s output of light
declined with a half-life of 77 days, matching the decay of the cobalt within.
However, once hot bullets of radioactive material had broken through, some
of the energy emerged directly as gamma-rays and X-rays. This high-energy
radiation is now about as intense as the supernova’s output of light.

Because some of the radioactive energy has been emerging directly, the
amount of light from the supernova has fallen below the level astronomers
predicted from the rate it was fading initially. Towards the end of 1988,
however, the amount of light dropped even lower, according to a team at
the South African Astronomical Observatory in Cape Town. The team, which
has been monitoring the supernova’s light on a regular basis, found that
in the first half of 1989, one-third of the expected energy emerged as gamma-rays
and X-rays, one-third emerged as light, and the rest was missing.

The South African team suggests that the culprit is dust. The supernova’s
gases could have cooled so much that some of the atoms may have condensed
out as small solid grains of graphite or silicates. The dust would absorb
some of the supernova’s energy, and re-radiate it at long infrared wavelengths,
around 10 micrometres.

Astronomers, using both the Anglo-Australian Telescope and the Kuiper
Airborne Observatory (a plane carrying a telescope to an altitude of 12
000 metres), have indeed detected radiation at wavelengths around 10 micrometres.
But this apparent confirmation of dust in the supernova gases is far from
watertight because the team in Australia has found that the infrared emission
comes from an extended region in space, and not just from the supernova
itself.

The South African astronomers think they can reconcile the observations.
They suggest that most of the infrared light comes from dust in the supernova
itself, with a smaller amount arising in dust clouds behind the supernova.
These background clouds, warmed by the supernova explosion, produce the
extended region of emission.

During the past year, astronomers have been able to investigate both
the shape of the supernova’s expanding cloud of gas and the clouds of gas
and dust within a couple of light years of the supernova.

The gases from the supernova are not expanding spherically, according
to a team from the Harvard-Smithsonian Centre for Astrophysics in Cambridge,
Massachusetts. They have used the 4-metre telescope at the Cerro Tololo
Interamerican Observatory in Chile to produce images of the supernova in
unprecedented detail. The team finds that the gas cloud from the supernova
is appreciably egg-shaped, with the gases travelling about 50 per cent more
rapidly in a north-south direction than they are east-west.

The Harvard-Smithsonian team found a ‘mystery spot’ only two months
after the supernova’s explosion. The spot was only 17 light-days to the
south of the supernova. Astronomers in Australia confirmed that the spot
was real, but it subsequently disappeared. Last year, however, the American
team announced they had found another spot, in the same direction relative
to the supernova, but further out. Intriguingly, the new spot was at the
right distance to be a reappearance of the original spot, if it had left
the supernova at the time of the explosion and had travelled outwards at
one-third the speed of light.

The mystery spot, as its name suggests, has perplexed astronomers. A
blob of hot gas that has been ejected by the supernova should not stay together.
The most likely explanation is that the spots consist of matter distributed
at several different distances from the supernova that is being lit up by
energy emerging from the supernova.

Astronomers are now using the flash of light from the supernova’s explosion
to explore the clouds of gas and dust that lay around the star before it
exploded. By now, these clouds would be emitting spectral lines as the electrons
recombine with the ionised atoms.

A team from NASA’s Goddard Space Flight Center and the Carnegie Institute
in Washington DC, led by Arlin Crotts, has been at the forefront of these
investigations. They have used an electronic light detector, known as a
charge-coupled device, on telescopes at the Las Campanas Observatory in
Chile, with filters that can isolate the light that is emitted from different
ions.

Although astronomers would expect clouds near the supernova to consist
of a mixture of gas and dust, Crott’s team finds that the pictures they
obtain in ordinary light – reflected from dust – are different from the
images in the light from hydrogen ions or from oxygen ions. The dust clouds
reflect two rings of light, slightly off-centre from the position of the
supernova. Crotts says that these shapes indicate that the dust clouds form
two sheets that lie behind the supernova. Each is about half a light year
away from the supernova at its closest point, and is angled to our line
of sight.

Crotts points out that these sheets of dust could well account for the
extended nature of the infrared source that observers have seen at a wavelength
of 10 micrometres.

Several teams of astronomers are now keeping an eye on this region close
to the supernova. As time goes by, the burst of radiation will sweep through
the clouds here, illuminating successive cross sections. By putting the
pieces together, the researchers hope to build up a picture of the gases
the star lost before it exploded, and of the interstellar clouds in the
vicinity.

Already, the supernova’s light is providing a view of the cloud structures
hundreds of light years away. Light that is reflected from clouds in front
of the supernova appears as large circular rings. Photographs of these so-called
‘light echoes’ have revealed two curtains of gas and dust in the Large Magellanic
Cloud, about 400 and 900 light years in front of the supernova.

But the most intriguing observation about the supernova has still to
be confirmed. A year ago, a team of American astronomers, working at Cerro
Tololo in Chile, found pulses of light coming from the supernova at a rate
of 1968 pulses per second. They interpreted these as light from a rapidly-spinning
neutron star, or pulsar, that had formed from the collapsed core of the
supernova. In neutron stars, which are only a few tens of kilometres across,
matter has been compressed so that the protons and electrons of atoms become
neutrons.

This result has caused immense problems for theorists and other observers.
The theorists find it very hard to build models for neutron stars that could
spin that fast, without flying apart. And many other astronomers have failed
to pick up pulses from the supernova. Now, the claim of a fast-spinning
pulsar in the supernova has been retracted (see This Week, this issue).

According to the latest measurements of the output of infrared from
the supernova, at wavelength of 5 to 20 micrometres, the pulsar may be showing
its presence indirectly. Astronomers at the European Southern Observatory
in Chile have found that the supernova’s brightness at these wavelengths
has remained steady for the past four months, indicating a new source of
energy. Theorists had expected that another isotope of cobalt, cobalt-57,
should be providing some energy at this stage: it is produced in smaller
amounts than cobalt-56, but has a longer half-life.

But the European team says that the supernova would need to have 20
times the quantity of cobalt-57 that theorists calculate. It is more likely,
they say, to be the energy from the hidden pulsar.

* * *

THE CHEMICAL FACTORY IN THE SUPERNOVA

ASTRONOMERS have found dozens of different molecules in cold interstellar
clouds, and in clouds of gas ejected by red giant stars. But Supernova 1987A
has provided the first example of molecules made in the hot environment
of gases from a supernova explosion.

Infrared astronomers discovered characteristic emission from carbon
monoxide about 200 days after the explosion, when the gas was at about 3000
kelvin. According to Alex Dalgarno of the Harvard-Smithsonian Centre for
Astrophysics, the most likely reactions to produce carbon monoxide are the
direct combination of carbon and oxygen atoms in the gaseous state, or the
reaction of a negatively charged oxygen atom with a neutral carbon atom.
Another possibility is that hydrogen molecules (H₂) may react
with oxygen atoms to form hydroxyl (OH), which in turn reacts with a carbon
atom to yield carbon monoxide together with a free hydrogen atom.

Dalgarno calculates that the major way that carbon monoxide is destroyed
in a supernova is by reaction with a positively charged helium ion, which
strips an electron from the carbon monoxide molecule and causes it to split
up.

According to these calculations, the supernova should produce smaller
quantities of other molecules, including carbon monosulphide, silicon monoxide,
and silicon monosulphide. These have yet to be detected.