2013ApJ...771..112A

We use mid-infrared Spitzer spectroscopy and far-infrared Herschel photometry for a sample of 20 main sequence O9-B2 stars in the Small Magellanic Cloud (SMC) with strong 24 µm excesses to investigate the origin of the mid-IR emission. Either debris disks around the stars or illuminated patches of dense interstellar medium (ISM) can cause such mid-IR emission. In a companion paper, Paper I, we use optical spectroscopy to show that it is unlikely for any of these sources to be classical Be stars or Herbig Ae/Be stars. We focus our analysis on debris disks and cirrus hot spots. The local, prototype objects for these models are the debris disk around Vega and the heated dust cloud surrounding the stars in the Pleiades, also known as a cirrus hot spot. These two cases predict different dust masses, radii, origins, and structures, but the cleanest classification tools are lost by the poor physical resolution at the distance of the SMC. We also consider transition disks, which would have observable properties similar to debris disks. We begin classification by measuring angular extent in the highest resolution mid-IR images available. We find 3 out of 20 stars to be significantly extended, establishing them as cirrus hot spots. We then fit the IR spectral energy distributions to determine dust temperatures and masses. Analysis yields minimum grain sizes, thermal equilibrium distances, and the resultant dust mass estimates. We find the dust masses in the SMC stars to be larger than for any known debris disks. The difference in inferred properties is driven by the SMC stars being hotter and more luminous than known debris disk hosts and not in any directly observed dust properties, so this evidence against the debris disk hypothesis is circumstantial. Finally, we created a local comparison sample of bright mid-IR OB stars in the Milky Way (MW) by cross-matching the Wide-field Infrared Survey Explorer (WISE) and Hipparcos catalogs. We find that of the thousands of nearby ( ≤ 1 kpc) hot stars in the MW that show a mid-IR excess, only a small fraction (few percent) match the high mid-IR luminosities of the SMC stars. All such local stars in the appropriate luminosity range that can be unambiguously classified are young stars with optical emission lines or are spatially resolved by WISE with sizes too large to be plausible debris disk candidates. We conclude that the very strong mid-IR flux excesses are most likely explained as cirrus hot spots, although we cannot rigorously rule out that a small fraction of the sample is made up of debris disks or transition disks. We present suggestive evidence that bow-shock heating around runaway stars may be a contributing mechanism to the interstellar emission. These sources, interpreted as cirrus hot spots, offer a new localized probe of diffuse interstellar dust in a low metallicity environment.