Astron. Astrophys. 347, 524-531 (1999)

5. Discussion

Our primary goal was to study the role of non-radial pulsations in the occurrence of circumstellar material around B stars. None of the 163 pulsators has a Be-like IR excess flux due to circumstellar gas, despite the fact that our original sample includes several stars with rotation velocities and line-profile variability comparable to those of e.g. Cen. Some examples are the Cep stars Vir (Smith 1985a,b), Sco (De Mey et al. 1997), Sco (Aerts et al. 1998), Sco (Telting & Schrijvers 1998a), and Cen (Telting & Schrijvers 1998b). In the sample of 44 pulsators detected by IRAS, 8 have rotation velocities higher than Cen. This null result seems to indicate that non-radial pulsations cannot be the dominant cause of the presence of a disk around Be stars, even when the pulsation occurs in combination with rapid rotation.

On the other hand, our study did result in the detection of two pulsating B-type stars with an unexpected behaviour in the infrared. For both stars, the observed IR flux is not caused by free-free emission of ionised gas, as is the case for Be stars, but is instead caused by thermal dust emission.

HD 42927 and Lup are the only confirmed regular pulsators that show evidence for circumstellar dust. Another interesting star in this respect is HD 41814, which is also a new slowly pulsating B star discovered from the Hipparcos mission (Waelkens et al. 1998). This star was, however, not further studied by us because it only has an accurate IRAS detection at m (see e.g. Backman & Paresce 1993, who list the star as a possible Vega-type object).

At present, we have no further observational evidence that allows us to derive more details about the dust around HD 42927 and Lup. As far as we know, both stars were not observed by ISO. For Lup, we studied the 6 spectra taken by IUE to check if accreting gas, responsible for e.g. the inverse P-Cygni profiles in UV resonance lines of Pic-like stars (see e.g. Grady et al. 1996), is present around the star. We did not find any trace of such accreting circumstellar gas. To our surprise, there are no observations available in the IUE database for HD 42927. The absence of H emission for both pulsators indicates that there is not much ionised material in the circumstellar environment of the stars. Zaal et al. (1995) have shown that low-density discs around B stars can give rise to HI IR recombination lines in emission, without the presence of visible emission in H. Such low-density discs, however, do not give rise to an IR exces at the wavelengths observed by IRAS. Therefore, the IR excess of the two stars studied here is caused by thermal dust emission.

It is interesting to find early-B main-sequence stars with an IR-excess in terms of the search for Vega-type stars. Backman & Paresce (1993) give an overview of the Vega phenomenon and list the results of surveys of such stars. Out of the 60 BSC main-sequence stars with Vega-like IR excesses in the PSC, 17 (i.e. 28%) have spectral type B. Mannings & Barlow (1998) have recently compiled a list of new Vega-like stars based on main-sequence objects given in the FSC. 29% (21 stars) of their new sample are B-type stars. The Vega-phenomenon therefore seems quite common among B stars. Yet all the Vega stars that have been analysed in detail are cooler. A dedicated study of a hot Vega-type star seems warranted.

Different working definitions of a "Vega-like" candidate are used. If we look at the condition for "accepted sources" given recently by Mannings & Barlow (1998), then HD 42927 and Lup are new Vega-like stars. They are the first ones that are regular intrinsic variables due to stellar pulsation. HD 42927 was not found before by others, probably because it is not listed in the PSC and/or because it was classified as a star of luminosity class III. Our photometry, however, indicates that we are dealing with a main-sequence star rather than with a luminosity class III object.

So far, it is not possible to discriminate between a disk-like or a spherical dust distribution around both stars. Because of the low rotation velocity there is no a priori reason to assume that the dust is distributed in a a-spherically symmetric way. On the other hand, if the dust is really Vega-like, i.e. a reprocessed indirect remnant of the star-formation process, then a disk-like shape is plausible. The nature of the dust should be studied in more detail to evaluate this question. A way to do this is by IR-imaging at m, although the low flux levels will be a challenge to perform such observations in a successful way.

Since we found only two stars among all the B-type pulsators which show evidence for the presence of dust at a large distance from the star, we conclude that the dusty circumstellar environments of Lup and of HD 42927 are not related to their pulsational behaviour.

© European Southern Observatory (ESO) 1999

Online publication: June 30, 1999
helpdesk.link@springer.de