Imaging of detached shells around the carbon stars R Scl and U Ant through scattered stellar light^*

¹ Stockholm Observatory, 133 36 Saltsjöbaden, Sweden
² Nordic Optical Telescope, Apartado 474, 38700 Sta. Cruz de La Palma, Spain
³ CTIO, Casilla 603, La Serena, Chile
⁴ Uppsala Astronomical Observatory, Box 515, 751 20 Uppsala, Sweden

Corresponding author: D. González Delgado, delgado@astro.su.se

Received: 20 December 2000
Accepted: 26 March 2001

Abstract

We present the first optical images of scattered light from large, detached gas/dust shells around two carbon stars, R Scl and U Ant, obtained in narrow band filters centred on the resonance lines of neutral K and Na, and in a Strömgren b filter (only U Ant). They confirm results obtained in CO radio line observations, but also reveal new and interesting structures. Towards R Scl the scattering appears optically thick in both the K and Na filters, and both images outline almost perfectly circular disks with essentially uniform intensity out to a sharp outer radius of ≈21″. These disks are larger -by about a factor of two -than the radius of the detached shell which has been marginally resolved in CO radio line data. In U Ant the scattering in the K filter appears to be, at least partially, optically thin, and the image is consistent with scattering in a geometrically thin (≈3″) shell (radius ≈43″) with an overall spherical symmetry. The size of this shell agrees very well with that of the detached shell seen in CO radio line emission. The scattering in the Na filter appears more optically thick, and the image suggests the presence of at least one, possibly two, shells inside the 43″ shell. There is no evidence for such a multiple-shell structure in the CO data, but this can be due to considerably lower masses for these inner shells. Weak scattering appears also in a shell which is located outside the 43″ shell. The present data do not allow us to conclusively identify the scattering agent, but we argue that most of the emission in the K and Na filter images is to due to resonance line scattering, and that there is also a weaker contribution from dust scattering in the U Ant data. Awaiting new observational data, our interpretation must be regarded as tentative.