2003ApJ...594..347D

We use published ROSAT observations of the X-ray Nova V1974 Cygni 1992 to test a model for interstellar dust, consisting of a mixture of carbonaceous grains and silicate grains. The time-dependent X-ray emission from the nova is modeled as the sum of emission from an O-Ne white dwarf plus a thermal plasma, and X-ray scattering is calculated for a dust mixture with a realistic size distribution. Model results are compared with the scattered X-ray halos measured by ROSAT at nine different epochs, including the early period of rising X-ray emission, the ``plateau'' phase of steady emission, and the X-ray decline at late times. We find that the observed X-ray halos appear to be consistent with the halos calculated for the size distribution of Weingartner & Draine, which reproduces the Milky Way extinction with R_V=3.1, provided that the reddening to the nova is E(B-V)~0.20, consistent with E(B-V)~0.19 inferred from the late-time Balmer decrement. The time delay of the scattered halo relative to the direct flux from the nova is clearly detected. Models with smoothly distributed dust give good overall agreement with the observed scattering halo, but tend to produce somewhat more scattering than observed at 200"-300", and insufficient scattering at 50"-100". While an additional population of large grains can increase the scattered intensity at 50"-100", this could also be achieved by having ∼30% of the dust in a cloud at a distance from us equal to ∼95% of the distance to the nova. Such a model also improves agreement with the data at larger angles, and illustrates the sensitivity of X-ray scattering halos to the location of the dust. The observations therefore do not require a population of micron-sized dust grains. Future observations by Chandra and XMM-Newton of X-ray scattering halos around extragalactic point sources can provide more stringent tests of interstellar dust models.