Spitzer Mid - Infrared Imaging of Nearby Ultraluminous Infrared

We have observed 14 nearby (z < 0.16) Ultraluminous Infrared Galaxies (ULIRGs) with Spitzer at 3.6—24µm. The underlying host galaxies are well-detected, in addition to the luminous nuclear cores. While the spatial resolution of Spitzer is poor, the great sensitivity of the data reveals the underlying galaxy merger remnant, and provides the first look at off-nuclear mid-infrared activity. ULIRGs as a class are characterized by L ir > 10 12 L ⊙ , equivalent to the lumi-nosity of classical optically selected quasars. Based on extensive ground and space-based imaging, nearly all are known to be advanced merger remnants resulting from collisions of gas-rich galaxies. Dynamics of the collision result in a rearrangement of the gas, dust, and stellar content of the galaxies. The rapid inflow of gas deep into the merger core results in a powerful burst of star formation , and/or the fueling (in at least one-quarter) of an active nucleus. It has been postulated that ULIRGs are the progenitors of both QSOs and elliptical galaxies. Extensive imaging campaigns in the UV, optical, near and mid-IR during the '90s revealed the complex composite nature of the ULIRGs (Surace et al. At optical/UV wavelengths the emission is dominated by the old stellar population, as well as actively star-forming " super star clusters " located both in the nuclear regions and along the tidal features. At infrared wavelengths the emission is increasingly due to thermal dust emission arising in a compact core typically less than 100pc in diameter. Ground-based mid-infrared observations constrained a large fraction of the IRAS 12 and 25µm emission to an unresolved core. As part of the IRAC GTO program, we observed 14 nearby ULIRGs selected from the well-studied Bright Galaxy and Warm Galaxy Samples of Sanders et al. (1988, 1988b). The spatial resolution of Spitzer is relatively poor (≈2 ′′) at IRAC wavelengths compared to the known optical compact structure in these galaxies. However, it can resolve the galaxy bodies and extended tidal features, and separate the known double nuclei in some systems. While ground-based observations (e.g. from Keck) offer several times higher spatial resolution, they lack the sensitivity to detect anything beyond the high surface brightness nuclei. The extended, and hence potentially resolvable, emission is too faint to be readily 1