2007ApJ...666..806N

This is the second paper studying the QSOs in the Spitzer QUEST sample. Previously we presented new PAH measurements and argued that most of the observed far-infrared (FIR) radiation is due to star-forming activity. Here we present spectral energy distributions (SEDs) by supplementing our data with optical, NIR, and FIR observations. We define two subgroups, of ``weak FIR'' and ``strong FIR'' QSOs, and a third group of FIR nondetections. Assuming a starburst origin for the FIR, we obtain ``intrinsic'' active galactic nucleus (AGN) SEDs by subtracting a starburst template from the mean SEDs. The resulting SEDs are remarkably similar for all groups. They show three distinct peaks corresponding to two silicate emission features and a 3 µm bump, which we interpret as the signature of the hottest AGN dust. They also display drops beyond ∼20 µm that we interpret as the signature of the minimum temperature (∼200 K) dust. This component must be optically thin to explain the silicate emission and the slope of the long-wavelength continuum. We discuss the merits of an alternative model in which most of the FIR emission is due to AGN heating. Such models are unlikely to explain the properties of our QSOs, but they cannot be ruled out for more luminous objects. We also find correlations between the luminosity at 5100 Å and two infrared starburst indicators: L(60 µm) and L(PAH 7.7 µm). The correlation of L(5100 Å) with L(60 µm) can be used to measure the relative growth rates and lifetimes of the black hole and the new stars.