The Evolution of Starburst Galaxies to Active Galactic Nuclei

A model is proposed for active galactic nuclei in which a massive central star cluster builds up and feeds a central black hole as a result of mass loss during post-main-sequence stellar evolution. The starting point for the specific model treated here is a coeval stellar cluster of 4 x 10^9^ M_sun_ within the central ~10 pc, possibly formed as the result of a starburst in the central region of the galaxy. It is argued that such a concentration of stars could arise from the massive starburst triggered by the collision of gas-rich spiral galaxies and the accumulation of ~10^10^ M_sun_ in the central regions as observed in starburst galaxies such as Arp 220. The starburst observations are a major motivation for our consideration of this model since the extreme conditions, only previously hypothesized, are now observed. Within ~10^8^ yr, mass loss from the evolving upper main-sequence stars results in the accumulation of a central mass ~1.5 x 10^9^ M_sun_. Accretion at a rate of 10 M_sun_ yr^-1^ (with 10% conversion efficiency) results in a luminosity of ~10^13^ L_sun_ averaged over the first 10^8^ yr. This luminosity, radiated at X-ray and UV wavelengths from the inner accretion disk of the black hole, irradiates and ionizes the mass-loss envelopes of red giant stars in the continually evolving stellar population. These ionized stellar envelopes are proposed as the source of the broad emission lines seen in active galaxy nuclei (Scoville and Norman). The velocities of more than ~3000 km s^-1^ are due to orbital motion in the deep potential well of the cluster and black hole. The model predicts that the overall luminosity attributed to accretion should decay approximately as t^-1^ approaches t^-1.5^. The observed density and luminosity evolution of quasars at high Z could be due to the greater frequency of collisions between gas-rich galaxies in the early universe and the short lifetime of the high accretion-luminosity phase.