2008AJ....136..919B

The goal of this paper is to present the results of a preliminary analysis of the extended infrared (IR) emission by dust in the interstellar medium (ISM) of the Large Magellanic Cloud (LMC). We combine Spitzer Surveying the Agents of Galaxy Evolution (SAGE) and Infrared Astronomical Satellite (IRAS) data and correlate the infrared emission with gas tracers of H I, CO, and H_α. We present a global analysis of the infrared emission as well as detailed modeling of the spectral energy distribution (SED) of a few selected regions. Extended emission by dust associated with the neutral, molecular, and diffuse ionized phases of the ISM is detected at all IR bands from 3.6 µm to 160 µm. The relative abundance of the various dust species appears quite similar to that in the Milky Way (MW) in all the regions we have modeled. We construct maps of the temperature of large dust grains. The temperature map shows variations in the range 12.1-34.7 K, with a systematic gradient from the inner to outer regions, tracing the general distribution of massive stars and individual H II regions as well as showing warmer dust in the stellar bar. This map is used to derive the far-infrared (FIR) optical depth of large dust grains. We find two main departures in the LMC with respect to expectations based on the MW: (1) excess mid-infrared (MIR) emission near 70 µm, referred to as the 70 µm excess, and (2) departures from linear correlation between the FIR optical depth and the gas column density, which we refer to as FIR excess emission. The 70 µm excess increases gradually from the MW to the LMC to the Small Magellanic Cloud (SMC), suggesting evolution with decreasing metallicity. The excess is associated with the neutral and diffuse ionized gas, with the strongest excess region located in a loop structure next to 30 Dor. We show that the 70 µm excess can be explained by a modification of the size distribution of very small grains with respect to that in the MW, and a corresponding mass increase of ≃13% of the total dust mass in selected regions. The most likely explanation is that the 70 µm excess is due to the production of large very small grains (VSG) through erosion of larger grains in the diffuse medium. This FIR excess could be due to intrinsic variations of the dust/gas ratio, which would then vary from 4.6 to 2.3 times lower than the MW values across the LMC, but X_CO values derived from the IR emission would then be about three times lower than those derived from the Virial analysis of the CO data. We also investigate the possibility that the FIR excess is associated with an additional gas component undetected in the available gas tracers. Assuming a constant dust abundance in all ISM phases, the additional gas component would have twice the known H I mass. We show that it is plausible that the FIR excess is due to cold atomic gas that is optically thick in the 21 cm line, while the contribution by a pure H₂ phase with no CO emission remains a possible explanation.