Quasar feedback: more bang for your buck

Abstract

We propose a 'two-stage' model for the effects of feedback from a bright quasar on the cold gas in a galaxy. It is difficult for winds or other forms of feedback from near the accretion disc to directly impact (let alone blow out of the galaxy) dense molecular clouds at ∼kpc. However, if such feedback can drive a weak wind or outflow in the hot, diffuse interstellar medium (a relatively 'easy' task), then in the wake of such an outflow passing over a cold cloud, a combination of instabilities and simple pressure gradients will drive the cloud material to effectively expand in the direction perpendicular to the incident outflow. This shredding/expansion (and the corresponding decrease in density) may alone be enough to substantially suppress star formation in the host. Moreover, such expansion, by even a relatively small factor, dramatically increases the effective cross-section of the cloud material and makes it much more susceptible to both ionization and momentum coupling from absorption of the incident quasar radiation field. We show that even a moderate effect of this nature can dramatically alter the ability of clouds at large radii to be fully ionized and driven into a secondary outflow by radiation pressure. Since the amount of momentum and volume which can be ionized by observed quasar radiation field is more than sufficient to affect the entire cold gas supply once it has been altered in this manner (and the 'initial' feedback need only initiate a moderate wind in the low-density hot gas), this reduces by an order of magnitude the required energy budget for feedback to affect a host galaxy. Instead of ∼5 per cent of the radiated energy (∼100 per cent momentum) needed if the initial feedback must directly heat or 'blow out' the galactic gas, if only ∼0.5 per cent of the luminosity (∼10 per cent momentum) can couple to drive the initial hot outflow, this mechanism could be efficient. This amounts to hot gas outflow rates from near the accretion disc of only ∼5–10 per cent of the black hole accretion rate.

Files

Additional details