2010MNRAS.405.2439O

In order to understand the formation mechanism of the discs around Be stars, it is imperative to have a good overview of both the differences and similarities between normal B stars and the Be stars. In this paper, we address binarity of the stars. In particular, we investigate a previous report that there may be a large population of subarcsec companions to Be stars. These had hitherto not been detected due to a combination of a limited dynamic range and spatial resolution in previous studies. We present the first systematic, comparative imaging study of the binary properties of matched samples of B and Be stars observed using the same equipment. We obtained high angular resolution (0.07-0.1arcsec) K-band adaptive optics data of 40 B stars and 39 Be stars. The separations that can be probed range from 0.1 to 8arcsec (corresponding to 20-1000au), and magnitude differences up to 10mag can in principle be covered. After excluding a few visual binaries that are located in regions of high stellar density, we detect 11 binaries out of 37 Be targets (corresponding to a binary fraction of 30±8 per cent) and 10 binaries out of 36 B targets (29±8 per cent). Further tests demonstrate that the B and Be binary systems are not only similar in frequency but also remarkably similar in terms of binary separations, flux differences and mass ratios. The minimum physical separations probed in this study are of order 20au, which, combined with the similar binary fractions, indicates that any hypotheses invoking binary companions as responsible for the formation of a disc need the companions to be closer than 20au. Close companions are known to affect the circumstellar discs of Be stars, but as not all Be stars have been found to be close binaries, the data suggest that binarity cannot be responsible for the Be phenomenon in all Be stars. Finally, the similarities of the binary parameters themselves also shed light on the Be formation mechanism. They strongly suggest that the initial conditions that gave rise to B and Be star formation must, to all intents and purposes, be similar. This in turn indicates that the Be phenomenon is not the result of a different star formation mechanism.