The Origin of Fe II Emission in Active Galactic Nuclei

We used a very large set of models of broad emission line region (BELR) clouds in active galactic nuclei to investigate the formation of the observed Fe II emission lines. We show that photoionized BELR clouds cannot produce both the observed shape and observed equivalent width of the 2200-2800 AA Fe II UV bump unless there is considerable velocity structure corresponding to a microturbulent velocity parameter upsilon sub(turb) greater than or equal to 100 km s super(-1) for the locally optimally emitting cloud models used here. This could be either microturbulence in gas that is confined by some phenomenon such as MHD waves or a velocity shear such as in the various models of winds flowing off the surfaces of accretion disks. The alternative way that we can find to simultaneously match both the observed shape and equivalent width of the Fe II UV bump is for the Fe II emission to be the result of collisional excitation in a warm, dense gas. Such gas would emit very few lines other than Fe II. However, since the collisionally excited gas would constitute yet another component in an already complicated picture of the BELR, we prefer the model involving turbulence. In either model, the strength of Fe II emission relative to the emission lines of other ions such as Mg II depends as much on other parameters (either upsilon sub(turb) or the surface area of the collisionally excited gas) as it does on the iron abundance. Therefore, the measurement of the iron abundance from the Fe II emission in quasars becomes a more difficult problem.