IGR J18245–2452: The most important neutron star you’ve never heard of

Astronomers have spent thirty years on the theory behind how millisecond pulsars form. Now we know they got it right.

Graham Doskoch
Jan 22 · 5 min read

Neutron stars are known for their astonishing rotational speeds, with most spinning around their axes many times each second. The mechanism behind this is simple: When a fairly massive star several times the radius of the Sun collapses into a dense ball about ten kilometers in diameter, conservation of angular momentum dictates that it must spin quicker.

However, one class of neutron stars can’t be explained this way: millisecond pulsars. These exotic objects spin hundreds of times each second, with the fastest, PSR J1748–2446ad, rotating at over 700 Hertz! Since their discovery in the 1980s, a slightly different evolutionary path has been proposed. After studying dozens of systems, astronomers theorized that millisecond pulsars are very old — old enough that they’ve lost much of their original angular momentum to radiation. However, they’re also in binary systems, and under certain conditions, a companion star can transfer matter — and thus angular momentum — to the pulsar, spinning it back up again.

A plot of the periods and magnetic fields of pulsars. Millisecond pulsars have extremely short periods, and comparatively weak magnetic fields. Image credit: Swinburne University of Technology

During this period of accretion, the system should become an x-ray binary, featuring strong emission from the hot plasma in the neutron star’s accretion disk. There should also be periods where the neutron star behaves like an ordinary radio pulsar, emitting radio waves we can detect on Earth. If we could detect both types of radiation from a single system, it might be the clinching bit of evidence for the spin-up model of millisecond pulsar formation.

In 2013, astronomers discovered just that: a binary system known as IGR J18245–2452 from its x-ray outbursts, and PSR J1824–2452I for its radio emissions. First observed in 2008, it had exhibited both radio pulsations and x-ray outbursts within a short period of time, clear evidence of the sort of transitional stage everyone had been looking for. This was it: a confirmation of the ideas behind thirty years of work on how these strange systems form.

INTEGRAL observations of IGR J18245–2452 from February 2013 (top) and March/April 2013 (bottom). The system is only visible in x-rays in the second period. Image credit: ESA/INTEGRAL/IBIS/Jörn Wilms.

The 2013 outburst

Towards the end of March of 2013, the INTEGRAL and Swift space telescopes detected x-rays from an energetic event coming from the core of the globular cluster M28 (Papitto et al. 2013). It appeared to be an outburst of some kind — judging by the Swift observations, likely a thermonuclear explosion. A number of scenarios can lead to x-ray transients, including novae and certain types of supernovae. Binary systems are often the culprits, where mass can be transferred from one star or compact object to another.

Fig. 7, Papitto et al. Swift data from observations of an outburst show its characteristic exponentially decreasing cooling.

One thermonuclear burst observed by Swift followed a time evolution profile expected for such a detonation: An increase in luminosity for 10 seconds, followed by an exponential decrease with a time constant of 38.9 seconds. This decrease represents the start of post-burst cooling. The other outbursts from the system should have had similar profiles characteristic of x-ray-producing thermonuclear explosions, and indeed later observations of the system have confirmed that this is indeed the case (De Falco et al. 2017), albeit with slightly different rise times and decay constants.

To determine the identity of the transient, now designated IGR J18245–2452, astronomers made follow-up observations using the XMM-Newton telescope. The nature of the outburst would determine how it evolved over time. For instance, supernovae (usually) decrease in brightness over the course of weeks or months. In this case, however, the x-rays were still detected — albeit a bit weaker. More surprisingly, the strength of the emission appeared to be modulated, varying with a period of 3.93 milliseconds.

Such a short period seemed to indicate that a pulsar might be responsible. The team checked databases of known radio pulsars and found one that matched the x-ray source: PSR J1824–2452I, a millisecond pulsar in a binary system. Even after this radio counterpart had been found, however, two questions remained: Were these x-ray pulses new or a long-term process, and how did they relate to the radio emission?

Diving into the archives

A handy tool for observational astronomers is archival images. By looking at observations taken months, years or decades before an event, scientists can — if they’re lucky — peek into the past to see what an object of interest looked like long before it became interesting. Archival data is often of use for teams studying supernovae, as even a previously uninteresting or unnoticed star can tell the story of a supernova’s progenitor.

Fig. 3, Papitto et al. Chandra images from 2008, showing the system in quiescent (top) and active (bottom) states.

In this case, Papitto et al. looked at Chandra observations from 2008, comparing them with new data from April 2013. They found x-ray variability occurring shortly after a period of radio activity by the pulsar, indicating that the system had switched off its radio emissions and started emitting x-rays. This was extremely interesting, because new observations with three sensitive radio telescopes — Green Bank, Parkes, and Westerbork — indicated that the pulsar was no longer active in radio waves. It was possible that the pulsar had been eclipsed and emission was ongoing, and this may indeed have happened at some points, but was not likely to be the main factor behind the apparent quiescence.

A few weeks later, however, the exact opposite happened: the pulsar exited its quiescent radio state and was again picked up by the three radio telescopes. In short, over a period of months, it had oscillated between behaving like an x-ray binary and a normal millisecond pulsar. Finally, x-ray observations had conclusively shown that this sort of bizarre transitional state was possible!

The mechanism

IGR J18245–2452 spends the vast majority of its time in what is known as a “quiescent” state, during which there is comparatively little x-ray activity. The pulsar’s magnetosphere exerts a pressure on the infalling gas, forming a disk at a suitable distance from the surface. Eventually, however, there is enough buildup that an x-ray outburst occurs, lasting for a few months. The outburst decreases the mass accretion rate, and the magnetosphere pushes away much of the transferred gas, allowing radio pulsations to take place once more.

Fig. 2, De Falco et al. Over a period of a few weeks, IGR J18245–2452 underwent a number of individual x-ray outbursts, themselves indicative of a brief period of x-ray activity and radio silence.

It’s expected that the pulsar will eventually be spun-up until its rotational period is on the order of a millisecond or so. It will cease x-ray emissions, and be visible mainly through radio pulses. All of this, however, is far in the future, and during our lifetimes, IGR J18245–2452 will stay in its current transitional state, halfway between an x-ray binary and a millisecond pulsar.

Look Upwards

A blog about the strangest astronomical objects in the universe.

Graham Doskoch

Written by

Astrophysics undergrad, currently looking at the skies through gamma-rays and x-rays. I write about deep sky objects and the future of observational astronomy.

Look Upwards

A blog about the strangest astronomical objects in the universe.