杏吧原创

ON 5 March 1979, a sudden and unexpected blast of gamma rays hit the Solar
System. The dozen or so satellites that were sensitive to these wavelengths a
rude jolt to their detectors, many going off the scale.

It lasted only a fifth of a second. But this energy wave still holds the
record for the most intense burst of cosmic gamma rays ever seen. Less violent
bursts followed from the same place in the sky, while sporadic bursts were
detected from a few other sources (see 鈥淏olts from the blue鈥�). Astronomers
were baffled. The gamma rays could not be coming from explosive events
that destroyed the source, because the bursts kept repeating themselves. So what
could provide such sudden and dramatic outbursts?

鈥淲e have been thinking about it almost every day for the past six years,鈥�
says theorist Robert Duncan of the University of Texas at Austin. With his
collaborator Christopher Thompson of the University of North Carolina at Chapel
Hill he published a paper in 1992 suggesting a possible explanation for the 1979
event and other outbursts from similar 鈥渟oft gamma-ray repeaters鈥� (SGRs). It was
an extraordinary idea: the researchers proposed the existence of a new kind of
star, called a magnetar. Unimaginably dense, this star would have a solid crust
covering an exotic liquid core. More importantly, it would bear huge magnetic
fields whose motion would heat up the surface to such an extent that it would
crack under the appalling strain. The upshot: starquakes that blast the cosmos
with gamma rays.

It sounds extraordinary, but in the past few months astronomers have
discovered that Duncan and Thompson could well be right. 鈥淭he door is opened to
a whole new population of objects, previously rumoured but never seen,鈥� says
Chryssa Kouveliotou of the Universities Space Research Association at NASA鈥檚
Marshall Space Flight Center. Moreover, the bizarre new stars could explain
other astronomical mysteries, unresolved until now.

Duncan and Thompson based their idea for magnetars on the more
familiar鈥攖hough still extraordinary鈥攔elatives of these objects,
neutron stars. Born in the aftermath of a supernova explosion, neutron stars
lead dramatic lives from the start (see 鈥淟ife after death鈥�). Ten thousand
billion times denser than lead, they are liquid on the inside. And at only
around 10 kilometres across, they would be dwarfed by the Earth. Yet the most
intriguing fact of all about neutron stars is that they have a solid crust, with
a surface made of iron鈥攁 far cry from the tenuous, shifting plasma that
envelops ordinary stars like the Sun.

There are good reasons to connect SGRs to neutron stars. The multiple
satellite detections of the 5 March SGR event enabled astronomers to reconstruct
the direction of the incoming wave of gamma rays. They found that the source
appeared to lie in the cloud left over from a supernova. And wherever such
remnants appear, the chances are that a neutron star will be lurking there
too.

By the mid-1990s, advances in satellite instrumentation had allowed
astronomers to pinpoint the position of another SGR. These advances enabled
detectors to spot the persistent X-ray glow that SGRs produce between bursts.
Sure enough, a supernova remnant was found there too.

But there was a problem. Usually, isolated neutron stars formed out of
relatively recent supernovae are radio pulsars. They spin very rapidly, and have
magnetic fields which, although enormous by terrestrial standards, are at least
a hundred times weaker than those proposed for magnetars. The energy from the
rapid spinning sends charged particles spiralling out along magnetic field lines
that protrude from the star鈥檚 surface. Many of these field lines have both ends
embedded in the star, so the particles are trapped. But at the poles, the field
lines are like bundles of straws extending out into space. Here, the particles
stream out into the cosmos, producing beams of radio waves. As the poles spin
across our line of sight, we see these beams as regular radio blips, like the
periodic flashes of a lighthouse beacon.

But SGRs have quite different properties: they give off both bursts and
steady glows of gamma rays and X-rays. Moreover, the 5 March 1979 burst was
followed by a surge of lower-energy radiation that shows a periodic ripple,
suggesting that the source rotates every eight seconds. The problem is that
young radio pulsars typically spin much faster, completing each rotation in a
fraction of a second.

And so the theorists were presented with a challenge. If SGR bursts were
caused by neutron stars, it was a type that had never been seen before. During
the 1980s, many exotic theories were proposed to explain the SGRs, but none
seemed to provide a robust and credible explanation of all the facts.

Enter Duncan and Thompson, both then at Princeton University, who became
interested in what causes the magnetic field in pulsars. The two researchers
were intrigued by the fact that all young radio pulsars seem to have magnetic
fields of about 10¹² gauss鈥攁bout a thousand billion times more powerful
than the Earth鈥檚 magnetic field. Impressive as this is, the researchers
discovered that if a neutron star is born with a fast enough spin, the field
could be increased to a hundred or a thousand times this value by a mechanism
known as dynamo action. During this process, magnetic field lines are dragged
along and twisted by hot moving liquid in the star鈥檚 interior. The field is
unable to detach itself, because charged particles in the liquid act as a sort
of electromagnetic glue. And as the field writhes within the star, electric
currents are produced which generate more magnetic flux, steadily building up
the field.

Thus was born the idea of a 鈥渕agnetar鈥�. 鈥淲e began to wonder what such a
strongly magnetised neutron star would look like,鈥� says Duncan.

For one thing, such a star should rotate slowly. Though starting off with a
rapid spin, it would be very efficiently slowed down. The magnetic field axis
and the rotation axis tend not to be perfectly aligned in rapidly spinning,
highly magnetised neutron stars. So as the star rotates, the magnetic field
effectively changes direction, causing the emission of electromagnetic waves.
These waves carry away rotational energy, causing the star to spin down. This
happens to pulsars too but their magnetic field is so much weaker that they
can鈥檛 lose their spins as dramatically (see Diagram).

So young radio pulsars are still spinning fast enough to emit substantial
stores of rotational energy as radio beams. But magnetars are rotating too
slowly for this, except for a fleeting period right after their birth. The
powerhouse of a magnetar is not rotation but the magnetic field itself.
Moreover, a young magnetar would be very hot, because of frictional heat
generated by mobile material, redistributed by the powerful magnetic field.

Duncan and Thompson realised that these characteristics were starting to ring
bells. Could the magnetars that they had conjured up be soft gamma-ray
repeaters? The frictional heat of a young magnetar could account for the steady
X-ray glows in SGRs. And a magnetar鈥檚 slow rotation would be consistent with the
eight-second modulation in the 5 March burst. There remained one problem: the
cause of the gamma-ray outbursts themselves.

But again their calculations of what a magnetar would be like came up trumps.
The motion of the intense magnetic field through the crust would heat up the
surface to millions of degrees. The outer shell of the star would be placed
under unbearable strain, until something had to give. With a violent shake, the
crust would relieve the tension by cracking apart, in the stellar equivalent of
an earthquake. The surge of magnetic energy released in this way would shake the
stellar surface and heat the tenuous particle atmosphere above it, causing the
emission of copious amounts of soft gamma rays. After this short-lived tantrum,
the magnetar would calm down, presumably only to flare up again at some later
date.

Sudden burst

One aspect of the 5 March 1979 event doesn鈥檛 quite fit with this picture: why
was it some 10 000 times brighter than normal SGR bursts? But this could be
explained, said Duncan and Thompson, by a process called magnetic reconnection,
in which twisted, broken field lines suddenly fuse back together, releasing an
enormous burst of magnetic energy. A similar process causes solar flares on the
Sun.

Though this all fitted neatly with observations of SGR bursts, many
astronomers remained sceptical about whether magnetars could actually exist.
What the theory really needed was an observational nail. This is where
Kouveliotou and her colleagues come in. Between October 1996 and November 1997,
an old soft gamma-ray repeater called SGR 1806-20 became active again.
Kouveliotou and her co-workers made crucial observations of the source with a
satellite called the Rossi X-ray Timing Explorer. Their findings, reported in
May in Nature finally promoted magnetars from theoretical possibility
to observational fact.

The team discovered that the source pulses every 7.47 seconds, and that it is
slowing down at a measurable rate. These two measurements allowed the
researchers to deduce that they were seeing a slowly rotating, young neutron
star鈥攍ess than 10 000 years old. The rate at which it is slowing means the
star must have a magnetic field about a hundred times greater than that of a
radio pulsar. The first ever magnetar had been found, and with it the key to the
SGR mystery.

鈥淥ur results force us to see neutron stars and even galaxies in a new
light,鈥� says Kouveliotou. Magnetars spin too slowly to generate radio waves from
their rotational energy, which could explain why some supernova remnants do not
contain radio pulsars. Moreover, magnetars may solve the mystery of a group of
objects known as the anomalous X-ray pulsars. These stars resemble SGRs, but
they have never been observed to emit gamma-ray bursts. Kouveliotou believes
they could be old magnetars that have outlived their youthful bursting period.
After about 10 000 years鈥攖he blink of an eye by astronomical
standards鈥攖he magnetic field should have decayed so much that the seismic
activity and gamma-ray bursts will cease. This is consistent with the fact that
all SGRs appear to be associated with young supernova remnants.

Duncan estimates that there may be as many as a hundred million magnetars in
our Galaxy, most of which, like extinct volcanoes, have long since ceased to be
active. As Kouveliotou remarks: 鈥淢agnetars may be one of the most common types
of object in our Galaxy, but few will ever be observed because their active
lifetimes are so short.鈥�

And what of the dead population? 鈥淭hey are now slowly rotating dark stars,
quietly drifting though the Galaxy, difficult to detect,鈥� says Duncan. In space,
as on Earth, a peaceful retirement may await even the wildest of stars.

GAMMA-RAY bursts came in two distinct types. The first has very energetic or
鈥渉ard鈥� gamma rays. Roughly one burst of this type is observed per day, always
from different directions in the sky. Nobody knows exactly what causes the
bursts, but they come from the far reaches of the Universe, possibly from an
extreme type of supernova, or colliding neutron stars or black holes (鈥淕od鈥檚
Firecrackers鈥�, New 杏吧原创, 31 May 1997, p 28).

The other type of burst鈥攖he soft gamma-ray repeater of our
story鈥攊s much rarer. The outbursts come from just a few places in the sky,
although more than one burst is seen from each source, hence the name
鈥渞epeater鈥�. Three SGRs were discovered in 1979, a fourth in 1997 and a fifth
just two months ago. Their gamma rays are less energetic than those of the other
type鈥攈ence the name 鈥渟oft鈥�. But do not be deceived into thinking that SGR
bursts are feeble: in one second they radiate as much as the Sun does in a
year.

A NEUTRON star is the hot, compact cinder that remains when an ordinary star,
perhaps ten times heavier than the Sun, explodes in a supernova. It is
incredibly dense because it is made of nuclear matter, whereas ordinary material
comprises atoms and molecules. An atom is made chiefly of empty space. Neutron
stars achieve their extreme densities by eliminating this space. As an ageing
star collapses, it generates enormous pressure which destroys atoms as nuclei
are rammed together, and protons and electrons combine to make neutrons.

The inside of the resulting neutron star is a weird and varied world. 鈥淚n the
deep interiors of neutron stars, our knowledge breaks down,鈥� says theorist
Gordon Baym, of the University of Illinois at Urbana-Champaign. Here, the
density is too great even for neutron matter; instead, physicists believe that
more exotic forms of nuclear matter may exist, such as fields of quark-antiquark
pairs called pions, which in normal matter are thought to bind particles
together in the atomic nucleus.

Nearer to the surface, the density鈥攚hile still extremely
high鈥攍essens progressively. It is here that the bulk of the neutrons
reside, in a superfluid quantum sea, interspersed with protons and electrons.
This fluid of neutrons is so dense that a single drop would weigh several
million tonnes. In the outermost regions of the star, the lower pressures and
temperatures allow the formation of nuclei, which solidify into the iron-clad
crust.