Bruce T. Draine

Princeton University Observatory

The wavelength dependences of interstellar extinction and polarization, supplemented by the spectrum of infrared emission from dust heated by starlight and observed elemental abundances, strongly constrain dust models. I will discuss models that appear to be consistent with those constraints. These models reproduce the observed IR emission from the diffuse ISM of the Milky Way, and the integrated emission from other star-forming galaxies. If the grains are assumed to be partially-aligned spheroids, the models can reproduce the observed extinction and polarization of starlight. For such models we predict the far-infrared and submm polarization; the spectrum of the far-infrared and submm polarization is model-dependent, allowing such models to be tested by Planck.

The composition, structure, and size distribution of interstellar grains is the result of injection of dust from stellar outflows into the interstellar medium, followed by destruction, growth, coagulation, and photoprocessing of interstellar grains. The balance among these poorly-understood processes is responsible for the mix of solid material present in the ISM. I will argue that the bulk of present-day interstellar dust in the Milky Way was "grown" in the ISM -- most dust is NOT "stardust". Grain growth in the ISM can also account for the large masses of dust seen even in high-z objects such as J114816 at z=6.4 -- there is no need to invoke extraordinary production of dust in supernovae.

If time permits, I will discuss very puzzling observations of interstellar dust entering the heliosphere today, which, if correct, may have dramatic consequences on our understanding of the (local) interstellar medium.