Breaking the limit: Super-Eddington accretion onto black holes and neutron stars

Abstract

With the recent discoveries of massive and highly luminous quasars at high redshifts ($z\sim7$; e.g. Mortlock et al. 2011), the question of how black holes (BHs) grow in the early Universe has been cast in a new light. In order to grow BHs with $M_{\rm BH} > 10^9$ M$_{\odot}$ by less than a billion years after the Big Bang, mass accretion onto the low-mass seed BHs needs to have been very rapid (Volonteri & Rees, 2005). Indeed, for any stellar remnant seed, the rate required would need to exceed the Eddington limit. This is the point at which the outward force produced by radiation pressure is equal to the gravitational attraction experienced by the in-falling matter. In principle, this implies that there is a maximum luminosity an object of mass $M$ can emit; assuming spherical accretion and that the opacity is dominated by Thompson scattering, this Eddington luminosity is $L_{\rm{E}} = 1.38 \times 10^{38} (M/M_{\odot})$ erg s$^{-1}$. In reality, it is known that this limit can be violated, due to non-spherical geometry or various kinds of instabilities. Nevertheless, the Eddington limit remains an important reference point, and many of the details of how accretion proceeds above this limit remain unclear. Understanding how this so-called super-Eddington accretion occurs is of clear cosmological importance, since it potentially governs the growth of the first supermassive black holes (SMBHs) and the impact this growth would have had on their host galaxies (`feedback') and the epoch of reionization, as well as improving our understanding of accretion physics more generally.