The discovery of a youthful neutron star with an extreme magnetic field blurs the once clear divide between magnetars and pulsars.
Radio pulsars are rapidly spinning, strongly magnetized neutron stars that produce pulses of X-rays and γ-rays as well as radio waves. A combination of data from two X-ray satellites recently led to the discovery of a very young pulsar with an unusually high magnetic field, as reported by Eric Gotthelf and collaborators1 in Astrophysical Journal Letters. This X-ray pulsar is associated with the supernova remnant Kesteven 75. Although neutron stars are thought to be created in supernova explosions, there are surprisingly few clear examples of this. Gotthelf et al. suggest that the unusual X-ray pulsar may be a missing link between different classes of pulsars, and that it could provide important clues to how particles are created and radiate in the magnetospheres surrounding neutron stars. It may turn out that many of the neutron stars in our Galaxy are born with properties similar to this pulsar rather than to the bulk of previously known neutron stars.
Neutron stars are extremely dense objects: they have masses similar to the Sun but diameters of only about 20 km. Most neutron stars are detected as 'ordinary' radio pulsars with magnetic fields of 1012 gauss — one trillion times as strong as the Earth's magnetic field — and spin periods between 16 milliseconds and 8.5 seconds. The rapid rotation combines with the strong magnetic field to produce electric forces, like the alternator in a car, that generate particles moving at the speed of light. These relativistic particles radiate intense electromagnetic waves directed along the magnetic poles, which appear to an observer as pulses of radiation as the star rotates, like the light from a lighthouse. The magnetic field also slows the pulsar down through magnetic braking, an effect caused by the radiation carrying away the star's angular momentum. Ordinary pulsars stay radio 'loud' for about 10 million years, the time it takes for them to slow down to a spin rate at which particle creation stops.
Over the past five years, considerable interest has focused on objects that appear to be even more highly magnetized than typical radio pulsars. These are the so-called magnetars with magnetic fields that range from a few times 1013 to nearly 1015 gauss. A few have been identified in X-ray and γ-ray observations (Fig. 1). They spin down much more rapidly than radio pulsars, on timescales of about 10,000 years. Two subclasses of magnetars have been identified: soft-γ-ray repeaters (SGRs) and anomalous X-ray pulsars (AXPs).
The first three SGRs were discovered in 1979: two in the plane of our Galaxy, the third as a spectacular burst of γ-rays from an object associated with a supernova remnant in a satellite galaxy of the Milky Way. The γ-ray burst was followed by a fainter oscillation indicative of a slowly spinning object, now interpreted as a rotating neutron star that has undergone severe magnetic braking. Since 1979, one more SGR has been found, also in the plane of the Milky Way2.
The AXPs are anomalous in that their spin properties differ from the much larger class of X-ray pulsars found in binary systems. The AXPs also differ from the SGRs in that they do not show the same 'bursting' behaviour. Binary X-ray pulsars tend to spin faster as they gravitationally attract material from their companion star. At present, none of the magnetars is known to be in a binary system, and their evolution is towards longer spin periods — more like that of radio pulsars. Unlike radio pulsars, however, AXPs appear to radiate more X-ray energy than they need to lose as they slow down. It is thought that some of this anomalous X-ray emission comes from decay of the strong magnetic field.
When first identified, magnetars appeared to be a minority class of neutron star, with properties distinct from those of most radio pulsars3,4. But the short lifetimes of magnetars means they are probably underrepresented in observed samples, suggesting that their descendants litter our Galaxy. Nonetheless, there is no consensus about the conditions that determine whether a supernova explosion will leave behind an ordinary radio pulsar or a magnetar. One possibility is that the spin rate of the neutron star at birth governs the strength of additional magnetic fields generated through any dynamo effects5.
It has also been suggested that magnetars differ from radio pulsars, as far as processes in the magnetosphere are concerned. In a regular pulsar, some of the relativistic particles are electron–positron pairs generated from high-energy photons. The pairs generate, in turn, more high-energy photons, thus creating a cascade. The cascade subsides when a pulsar's spin slows down. But these cascades may also be suppressed in magnetic fields larger than the so-called critical quantum field of 4.4×1013 gauss (ref. 6). In such intense fields a competing process — photon splitting — can degrade the photon energies to below the requirement for pair production. For a short while, the magnetars and known radio pulsars were thought to fall on opposite sides of this divide.
X-ray emission from the new pulsar found by Gotthelf et al., PSR J1846-0258, is not unlike that seen from other radio pulsars, suggesting that it originates from electron– positron pairs, despite having a magnetic field greater than the critical quantum field. There are now several radio pulsars that are close to this limit7 (Fig. 1). Such objects seem to indicate that photon-splitting is not a dominant process in the magnetospheres of neutron stars. But our calculations of magnetic field strengths make several assumptions that may not hold for all neutron stars, so there may be systematic errors in the estimated field strengths. Despite this caveat, discoveries such as PSR J1846-0258 blur the division between pulsars and magnetars. It seems entirely possible that they represent extremes of the neutron-star population, depending on their original spin and magnetic field strength. So, for neutron stars, there may well be unity in their diversity.