The brightest high mass X-ray binaries and the Eddington threshold

Recent results have shown that the energetic output of the populations of X-ray binaries we see in nearby star forming galaxies (predominantly high mass X-ray binaries, hereafter HMXBs) is dominated by the very brightest of those sources, with for example > 80% of their radiative emission emerging from the small minority of objects with X-ray luminosities in excess of 10³⁸ erg/s. This is important because recent theoretical studies have suggested that HMXBs may have played an important early role in shaping our cosmos, with their mechanical and radiative feedback shaping their local environments, a process which would have been particularly important in young, low mass galaxies at high redshift. Hence, these effects would have come primarily from the brightest HMXBs present at that time.

However, to understand the potential impact of these objects, we first need to understand the physics of accretion for these luminous members of the HMXB family. Our group has led the recent good progress in understanding the most luminous objects in this class, the ultraluminous X-ray sources (ULXs), that are now thought predominantly to host stellar-mass black holes that accrete in a new super-Eddington ultraluminous accretion state. In contrast, the physics of objects accreting at around the Eddington limit, and the composition of the luminous X-ray source population in the corresponding 10³⁸ - 3x10³⁹ erg/s ‘Eddington threshold’ regime, remain relatively uncertain.

A visualisation of an accreting black hole high mass X-ray binary system. Credit: Image created by Tom Russell (ICRAR) using the software created by Rob Hynes (Louisiana State University).

At Durham we are working towards developing a good understanding of the physics of this Eddington threshold regime, and the objects underlying it. We are taking a two-pronged approach to this problem. Firstly, a population of predominantly transient objects in our own Galaxy can reach this regime; we are analysing data from these objects that are taken by the current generation of space-based missions (e.g. XMM-Newton, Suzaku, Swift). These data have the critical advantage compared to past missions of detecting X-ray photons with energies below 3 keV, where the majority of ionising photons originate. Between them they can reveal much of the physics for accretion at the Eddington threshold, from systems harbouring either black holes or neutron stars.

The second approach is to utilise the much larger sample of objects available by observing other, nearby galaxies. We have developed a new catalogue of 300 candidate ULXs and 800 less luminous X-ray sources associated with nearby galaxies, based on XMM-Newton source detection catalogues. This provides a large dataset that is ripe for exploitation, in terms of both studies of individual objects and the population as a whole. By combining this data with multi-wavelength data (primarily HST) we can identify the best HMXB candidates and so begin to build a composite view of this population and its properties, and from there how it contributes to feedback at high redshift.

Staff involved with this project at Durham are Tim Roberts and Chris Done.