MULTI-WAVELENGTH SEDs OF HERSCHEL-SELECTED GALAXIES IN THE COSMOS FIELD

We combine Herschel Photodetector Array Camera and Spectrometer and Spectral and Photometric Imaging Receiver maps of the full 2 deg(2) Cosmic Evolution Survey (COSMOS) field with existing multi-wavelength data to obtain template and model-independent optical-to-far-infrared spectral energy distributions (SEDs) for 4218 Herschel-selected sources with log(L-IR/L-circle dot) = 9.4-13.6 and z = 0.02-3.54. Median SEDs are created by binning the optical to far-infrared (FIR) bands available in COSMOS as a function of infrared luminosity. Herschel probes rest-frame wavelengths where the bulk of the infrared radiation is emitted, allowing us to more accurately determine fundamental dust properties of our sample of infrared luminous galaxies. We find that the SED peak wavelength (lambda(peak)) decreases and the dust mass (M-dust) increases with increasing total infrared luminosity (LIR). In the lowest infrared luminosity galaxies (log(L-IR/L-circle dot) = 10.0-11.5), we see evidence of polycyclic aromatic hydrocarbon (PAH) features (lambda similar to 7-9 mu m), while in the highest infrared luminosity galaxies (L-IR > 10(12) L-circle dot) we see an increasing contribution of hot dust and/or power-law emission, consistent with the presence of heating from an active galactic nucleus (AGN). We study the relationship between stellar mass and star formation rate of our sample of infrared luminous galaxies and find no evidence that Herschel-selected galaxies follow the SFR/M-* "main sequence" as previously determined from studies of optically selected, star-forming galaxies. Finally, we compare the mid-infrared to FIR properties of our infrared luminous galaxies using the previously defined diagnostic, IR8 L-IR/L-8, and find that galaxies with L-IR greater than or similar to 10(11.3) L-circle dot tend to systematically lie above (x3-5) the IR8 "infrared main sequence," suggesting either suppressed PAH emission or an increasing contribution from AGN heating.