Circumstellar Disk Systems in the LMC, SMC, and ...

THE NATURE AND EVOLUTION OF DISKS AROUND HOT STARS ASP Conference Series, Vol. 337, 2005 Richard Ignace and Kenneth G. Gayley

Circumstellar Disk Systems in the LMC, SMC, and Milky Way Galaxy John P. Wisniewski1 , Karen S. Bjorkman1 , Antonio M. Magalh˜aes2 , Jon E. Bjorkman1 , & Alex C. Carciofi1,2 1 Department

of Physics & Astronomy, University of Toledo, of S˜ ao Paulo

2 IAG-University

Abstract. Classical Be stars are rapidly rotating near main-sequence B-type stars which have gaseous circumstellar disks. Our understanding of these stars is incomplete, especially as regards two fundamental questions: 1) Is the Be phenomenon an evolutionary effect? and 2) What role does metallicity play in the formation of Be circumstellar disks? Initial attempts to address these questions have used 2-color diagram photometric techniques to determine the frequency of candidate Be stars in clusters of various ages and metallicities. These techniques classify all B type objects with excess Hα emission as classical Be stars, without accounting for the likelihood that other B-type stars, such as post main-sequence B[e] stars, YSOs, and B-supergiants, may also exhibit Hα emission and therefore contaminate the sample. We discuss our efforts to quantify the reliability of 2-color diagrams to identify classical Be stars.

1.

2-Color Diagram Technique

Grebel, Richtler, & de Boer (1992), Keller et al. (1999), and others have demonstrated how one can use a 2-color diagram (2-CD) technique to identify candidate Be stars. In such a diagram, normal blue main-sequence stars and supergiants associate in one area (e.g., (B-V) ∼ 0 and (R-Hα) ∼ -5.7 in Figure 1a), while red main-sequence and supergiants associate in another (e.g., 0.4 < (B-V) < 2 and (R-Hα) ∼ -5.7 in Figure 1a). Blue stars which show excess Hα emission, e.g. (R-Hα) > -5.5, are designated as candidate Be stars. We have observed 16 LMC, SMC, and Milky Way clusters of various ages, and used the 2-CD technique to identify 528 candidate Be stars. Four clusters with ages less than 10 Myr were found to have significant numbers of candidate Be stars. This conflicts with the suggestion of Fabregat & Torrejon (2000) that the Be star-disk systems develop in the second half of a star’s main-sequence lifetime, e.g. ∼ 10 Myr for B0 stars. 2.

2-CD: Detecting Field Be Stars

As a secondary test of the technique, we observed 2 LMC fields containing 12 candidate Herbig Ae/Be stars initially identified by their large-scale photometric variability (Lamers, Beaulieu, & de Wit 1999; de Wit, Beaulieu, & Lamers 2002). We detected 75% of these objects as excess Hα emitters, supporting their 333

334

Wisniewski et al.

classification as young pre-main-sequence objects. These results demonstrate that not all objects detected as Hα emitters via the 2-CD technique, especially those located outside of cluster environments where one can make a reasonable assumption of stellar ages, should automatically be considered classical Be stars.

Figure 1. Left panel : Sample 2-CD of the SMC cluster NGC 371. Right panel : 2-CD of a LMC field which contains candidate Herbig Ae/Be stars.

3.

Intrinsic Polarization

In an effort to better identify the classical Be stars in clusters, we have obtained UBVRI imaging polarimetry of 12 LMC/SMC clusters that have had candidate Be star populations identified via the 2-CD technique. Since classical Be stars are known to exhibit a characteristic intrinsic polarization signature, we can use these observations to improve the selection of true classical Be star-disk systems in these clusters. Details of the technique will be discussed in papers to be published elsewhere. Here we merely present some examples from our ongoing work. Sample intrinsic polarization signatures from our dataset are shown in Figure 2 and partially summarized in Table 1. We have divided our data into four categories: 1) Disks w/ B.J.: objects with polarization Balmer jumps(B.J.), e.g. definitive classical Be stars; 2) Disks w/o B.J.: objects whose polarization appears to be due to pure electron scattering, which is expected from weak classical Be star-disk systems; 3) Not Inconsistent: objects whose polarization, to within 2-sigma, are not inconsistent with that expected from Be stars (this definition includes unpolarized objects, which could be face-on Be star-disk systems); and 4) Unlikely: objects whose polarization deviates by more than 2-sigma with that expected from classical Be stars, hence are considered unlikely to be classical Be stars.

Circumstellar Disk Systems

335

Figure 2. Examples from the polarization survey. Left panel : A candidate Be star in NGC 371 that shows a polarization Balmer jump and flat position angle, indicating the object is definitely a classical Be star. Center panel : A candidate Be star whose intrinsic polarization appears to be a result of pure electron scattering, indicating the object is not inconsistent with a Be star. Right panel : A candidate Be star whose intrinsic polarization and position angle are not consistent, to within 2-sigma, with that expected from a classical Be star. (adapted from Wisniewski et al (2004))

Table 1. Preliminary summary (for one cluster) of the intrinsic polarization of candidate Be stars identified via the 2-CD technique. The detection rate denotes the number of candidates identified via the 2-CD technique for which we have some follow-up polarization measurements. Cluster NGC 371

4.

Detection Rate 74/130 (57%)

Disks w/ B.J. 11/74 (15%)

Disks w/o B.J. 6/74 (8%)

Not Inconsistent 42/74 (57%)

Unlikely 15/74 (20%)

Near-IR Spectroscopy

We are also using 0.8-2.5 µm spectroscopy to examine the nature of candidate Be stars identified via the 2-CD technique in several Galactic clusters which we were unable to observe with our imaging polarimeter. Near-IR spectroscopy can probe the inner circumstellar disk region of Be stars; hence, these data complement the diagnostic capabilities provided by our polarimetric observations. While some candidate Be stars exhibit an expected rich array of near-IR emission features, other candidates show no evidence of H Paschen or Brackett emission lines (see Figure 3) and/or narrow line widths. It is unclear whether these latter objects are classical Be star-disk systems which have temporarily

336

Wisniewski et al.

lost their disks or whether the 2-CD technique misclassified such objects as candidate Be stars.

Figure 3.

5.

Part of the near-IR spectrum of a candidate Be star in NGC 2186.

Summary

While the 2-CD technique discussed here provides an excellent estimate of the number of excess Hα emitters, our follow-up polarimetric and spectroscopic studies strongly suggest that a significant number of these objects are not classical Be stars. Further analysis of these data will allow us to quantify the number of “contaminants” which are interspersed with true classical Be star-disk systems. This will allow for a more accurate determination of the roles age and metallicity may play in Be disk formation. Acknowledgments. We acknowledge support for this work from NASA LTSA (KSB), NASA GSRP (JPW), Sigma Xi GIAR (JPW), and the Brazilian agencies FAPESP (AMM, ACC) and CNPq (AMM). KSB is a Cottrell Scholar of the Research Corporation and gratefully acknowledges their support. JPW thanks the conference organizers for providing travel support to attend this conference. References de Wit, W.J., Beaulieu, J.P. , & Lamers, H.J.G.L.M. 2002, A&A, 395, 829 Fabregat, J. & Torrejon, J.M. 2000, A&A, 357, 451 Grebel, E.K., Richtler, T., & de Boer, K.S. 1992, A&A, 254, L5 Keller, S.C., Wood, P.R., & Bessell, M.S. 1999, A&AS, 134, 489 Lamers, H.J.G.L.M., Beaulieu, J.P., & de Wit, W.J. 1999, A&A, 341, 827 Wisniewski, J.P., Bjorkman, K.S., & Magalh˜aes, A.M. 2004, Astronomical Polarimetry 2004, ASP Conf. Ser., in press