On the Efficiency of Grain Alignment in Dark CloudsWe dedicate this paper to the memory of Edward M. Purcell and Lyman Spitzer, Jr., two pioneers in the quantitative study of the interstellar medium.

Autor: Lazarian, A., Goodman, Alyssa A., Myers, Philip C.
Zdroj: The Astrophysical Journal; November 1997, Vol. 490 Issue: 1 p273-280, 8p
Abstrakt: A quantitative analysis of grain alignment in the filamentary dark cloud L1755 in Ophiuchus is presented. We show that the observed decrease of the polarization-to-extinction ratio for the inner parts of this quiescent dark cloud can be explained as a result of the decrease of the efficiency of grain alignment. We make quantitative estimates of grain alignment efficiency for six mechanisms involving grains with either thermal or suprathermal rotation, interacting with either magnetic field or gaseous flow. We also make semiquantitative estimates of grain alignment by radiative torques. We show that in conditions typical of dark cloud interiors, all known major mechanisms of grain alignment fail. All the studied mechanisms predict polarization at least an order of magnitude below the currently detectable levels of ~1%. On the contrary, in the dark cloud environments where Av< 1, the grain alignment can be much more efficient. There the alignment of suprathermally rotating grains with superparamagnetic inclusions, and possibly also radiative torques, account for observed polarization. These results apply to L1755, which we model in detail, and probably also to B216 and other similar dark clouds. Our study suggests an explanation for the difference in results obtained through polarimetry of background starlight and polarized thermal emission from the dust itself. We conjecture that the emission polarimetry selectively reveals aligned grains in the environment far from thermodynamic equilibrium, as opposed to starlight polarization studies that probe the alignment of grains all the way along the line of sight, including the interiors of dark quiescent clouds, where no alignment is possible.
Databáze: Supplemental Index