2006ApJ...648..452S

Using Chandra observations we have measured the energy-resolved dust-scattered X-ray halo around the low-mass X-ray binary GX 5-1, which shows signs of both singly and multiply scattered X-rays. We compared the observed X-ray halo at various energies to predictions from a range of dust models. These fits used both smoothly distributed dust and dust in clumped clouds, with CO and 21 cm observations helping to determine the positions of the clouds along the line of sight. We found that the BARE-GR-B model of Zubko and coworkers generally led to the best results, although inadequacies in both the overall model and the data limit our conclusions. We did find that the composite dust models of Zubko and coworkers, especially the ``no carbon'' models, gave uniformly poor results. Although models using cloud positions and densities derived naively from CO and 21 cm data gave generally poor results, plausible adjustments to the distance of the largest cloud and the mass of a cloud in the expanding 3 kpc arm led to significantly improved fits. We suggest that combining X-ray halo, CO, and 21 cm observations will be a fruitful method to improve our understanding of both the gas and dust phases of the interstellar medium.