Magnetars are some of the most extreme objects we know about, with magnetic fields so strong that chemistry becomes impossible in their vicinity. They're neutron stars with a superfluid interior that includes charged particles, so it's easy to understand how a magnetic dynamo is maintained to support that magnetic field. But it's a little harder to fully understand what starts the dynamo off in the first place.
The leading idea, which benefits from its simplicity, is that the magnetar inherits its magnetic field from the star that exploded in a supernova to create it. The original magnetic field, when crushed down to match the tiny size of the resulting neutron star, would provide a massive kick to start the magnetar off. There's just one problem with this idea: we haven't spotted any of the highly magnetized precursor stars that this hypothesis requires.
It turns out that we have been observing one for years. It just looked like something completely different, and it took a more careful analysis, published today in Science, to understand what we've been observing.
Not what it seems
HD 45166 on some levels appears to be a relatively straightforward binary star system, composed of a normal star and a hot Wolf-Rayet star co-orbiting at a short distance, with the light showing a 1.6-day periodicity, presumably due to the orbit.
But even at that level of understanding, a few things seemed weird about the Wolf-Rayet portion of the system. These are typically hot, massive stars that are helium-rich, having ejected most of their hydrogen through violent eruptions. But the one in HD 45166 is only four times as massive as the Sun—half the mass of the smallest example we've seen anywhere else. It also has lots of carbon, oxygen, and nitrogen present, which is rare, and their spectral lines have some unusual features.
The star's axis of rotation also appeared to be oriented in the direction of its orbit, which is also a bit difficult to arrange. So, there were several things that were difficult to explain about the system even prior to the new observations. In some ways, the updated data makes the system easier to understand; in others it makes it worse.
The key finding was that the spectrum of light from the star indicated that the 1.6-day periodicity is probably from a regular physical pulsing by the normal star in the HD 45166 system. That finding caused the researchers to examine other periodic changes to the light from HD 45166. The most likely orbital signal suggests that the orbit takes roughly 8,200 days—a rather radical difference from 1.6 days. It places the stars much farther apart and means neither of them is likely to orbit in the direction of its axis of rotation.
The larger separation in turn calls for a revision in the masses that were estimated based on their orbital interactions. The new estimate cuts the mass of the Wolf-Rayet star down by half, making it only twice the mass of the Sun.