Two O2 If*/WN6 stars possibly ejected from the massive young Galactic cluster Westerlund 2
In this paper we report the identification of two new Galactic O2 If*/WN6 stars (WR20aa and WR20c), in the outskirt of the massive young stellar cluster Westerlund 2. The morphological similarity between the near-infrared spectra of the new stars with that of WR20a and WR21a (two of the most massive binaries known to date) is remarkable, indicating that probably they are also very massive stars. New optical spectroscopic observations of WR20aa suggest an intermediate O2 If*/WN6 spectral type. Based on a mosaic made from the 3.6m Spitzer IRAC images of the region including part of the RCW49 complex, we studied the spatial location of the new emission line stars, finding that WR20aa and WR20c are well displaced from the centre of Westerlund 2, being placed at pc (15.7 arcmin) and pc (25.0 arcmin) respectively, for an assumed heliocentric distance of 8 kpc. Also very remarkably, a radius vector connecting both stars would intercept the Westerlund 2 cluster exactly at the place where its stellar density reaches a maximum. We consequently postulate a scenario in which WR20aa and WR20c had a common origin somewhere in the cluster core, being ejected from their birthplace by dynamical interacion with some other very massive objects, perhaps during some earlier stage of the cluster evolution.
keywords:Stars: individual: WR20aa - Stars: individual: WR20c - Stars: Wolf-Rayet (Galaxy): open clusters and associations: individual: Westerlund 2
Westerlund 2 (Wd2, Westerlund (1961), Moffat & Vogt (1975)) is a young massive stellar cluster placed in the core of the Hii region RCW49, which is located in the fourth Galactic quadrant, projected in the area of the Carina constellation. This cluster hosts one of the most massive binaries known: the star WR20a, comprised by two O3 If*/WN6 type stars with absolute masses of 83M and 82M (Bonanos et al., 2004; Rauw et al., 2005). Another massive star of the same spectral-type, WR20b (Shara et al., 1991) appears about 4’ south-east of the cluster core. A second massive WR binary, WR21a (that is located 16’ east of the Wd2 core) has a primary component of spectral type O3 If*/WN6 with an estimated minimum mass of 87M, and an early O type secondary with a minimum mass of 53M (Niemela et al., 2008). However the membership of WR21a to Wd2 is still uncertain. Such WR stars are members of the WNH type, a subset of very luminous and massive Hydrogen core burning WN objects (see definition in Smith & Conti (2008) and references there-in).
|Star||Program ID||Date||Telescope||Instrument||Grism/Disperser||Slit||R||Range (m)|
|WR20a||075.D-0210||2005-06-21||NTT||SofI||GR||0.6 x 290||1000||1.53-2.52|
|072.D-0082||2004-01-29||NTT||EMMI/2.3||GRAT#3||1 x 120||1600||0.39-0.47|
|WR20b||075.D-0210||2005-06-20||NTT||SofI||GR||0.6 x 290||1000||1.53-2.52|
|WR21a||075.D-0210||2005-06-20||NTT||SofI||GR||0.6 x 290||1000||1.53-2.52|
|WR20aa||075.D-0210||2005-06-21||NTT||SofI||GR||0.6 x 290||1000||1.53-2.52|
|GOSSS||2010-06-30||du Pont||Boller & Chivens||1200 lines mm||1x270||2500||0.39-0.55|
|WR20c||075.D-0210||2005-06-21||NTT||SofI||GR||0.6 x 290||1000||1.53-2.52|
|WR25||GOSSS||2008-05-20||du Pont||Boller & Chivens||1200 lines mm||1x270||2500||0.39-0.55|
|HD93129A||GOSSS||2010-06-30||du Pont||Boller & Chivens||1200 lines mm||1x270||2500||0.39-0.55|
With empirical masses exceeding 80 M, WNH stars are probably among the most massive stars known, being in the top-end of the mass distribution in young clusters and OB associations. This assumption is supported by results obtained from several studies of WNH stars in massive clusters such as R145 in 30 Doradus (Schnurr et al., 2009), WR25 (Gamen et al., 2006) and WR22 (Rauw et al., 1996) in the Carina Nebula region, and NGC3603-A1 in NGC3603 (Schnurr et al., 2008).
More recently, Crowther et al. (2010) from the spectroscopic analyses of four WN stars located within R136 (the core of 30 Doradus Nebula in the Large Magellanic Cloud - LMC) and three WN stars of HD 97950 (the core of NGC 3603 in the Milky Way), computed initial masses in the range M, and M, for the stars in R 136 and NGC 3603, respectively. They point out that due to their proximity to the Eddington limit, the very high-mass progenitors would possess an emission-line spectrum, at the beginning of their main-sequence evolution, mimicking the spectral appearance of classical WR stars.
The spectral-type O3If*/WN6-A was introduced almost thirty years ago by Walborn (1982) to classify the bright emission line star Sk-67 22 in LMC, which shows an intermediate spectrum between those of HD93129A (O2 If*) and HD93162 (=WR25, O2 If*/WN6). New examples of this intermediate spectral type were mainly found in LMC, always consisting in bright massive objects (see for example, Melnick (1985), Walborn & Blades (1997), Crowther et al. (2010)).
|WR20b (WN6ha)||-7.4 1.5||-11.8 2.1||-5.3 1.0||-7.0 0.8||-18.6 2.1||-12.1 1.9||-74.5 10.6||-33.7 4.5|
|WR20a (O3 If*/WN6+O3 If*/WN6)||-3.0 0.7||-3.3 0.7||-1.7 0.4||-1.3 0.3||-8.0 1.4||-2.0 0.5||-29.9 6.3||-14.5 3.8|
|WR21a (O3 If*/WN6+O)||-1.5 0.3||-2.3 0.5||-2.9 0.7||-1.9 0.5||-10.2 2.4||-1.2 0.5||-20.1 5.2||-5.3 1.3|
|WR20aa (O2If*/WN6)||-3.1 1.0||-3.3 1.0||-0.2 0.1||-1.6 0.4||-6.1 1.3||-1.1 0.4||-9.9 2.1||-4.8 1.2|
|WR20c (O2If*/WN6)||-1.5 0.5||-1.5 0.6||-1.3 0.5||-1.3 0.4||-5.2 1.2||-2.7 0.7||-12.5 2.7||-4.0 1.2|
In this paper we report the identification of two new Galactic O2 If*/WN6 stars (WR20aa and WR20c) in the outskirt of the massive young cluster Wd2. In section 2 we describe the used observational data and the data reduction procedures, in section 3 we present the results and discussion, and in section 4 it is performed a summary of the work.
|ID||Niv 4058||Heii 4686||H||Nv 4604||Niii 4634-40-42|
|WR25||-2.5 0.1||-15.6 0.7||-2.7 0.1||0.3 0.1||-0.4 0.1|
|WR20a||-2.3 0.1||-14.0 0.6||—||0.4 0.1||-0.2 0.1|
|WR20aa||-2.4 0.1||-9.6 0.5||0.3 0.1||0.7 0.1||-0.2 0.1|
|HD93129A||-1.3 0.1||-3.4 0.4||1.3 0.1||0.5 0.1||-0.1 0.1|
2 Observations and data reduction
2.1 Optical and Near-Infrared photometry of WR20aa and WR20c
Coordinates and photometric parameters for the new O2 If*/WN6 stars are shown in
Table 1. The near-IR photometric data are from the Two-Micron All Sky Survey
(2MASS, Skrutskie et al. (2006)), and were retrieved using the
The stars were named WR20aa and WR20c, following the common practice consisting
in giving the WR Galactic stars
numbers according to their RA, with further additions
between integers following van der Hucht (2001, 2006). Searching in
2.2 Spectroscopic observations
|WR20aa (O2 If*/WN6)||1.2||1.5||4.7||-6.5||-6.5|
|WR20c (O2 If*/WN6)||2.6||2.9||9.0||-6.1||-6.1|
|WR20a (O3 If*/WN6+O3 If*/WN6)||1.6||1.9||5.9||-7.0||-6.3|
|Early-O members||1.2 - 1.6||1.5 - 1.9||4.7 - 5.9||—||—|
|WR21a (O3 If*/WN6+O)||1.4||1.7||5.3||-7.2||-6.7|
The raw frames were reduced following the NIR reduction procedures presented by Roman-Lopes (2009), and briefly described here. The two-dimensional frames were sky-subtracted for each pair of images taken at the two nod positions A and B, followed by division of the resultant image by a master flat. The multiple exposures were combined, followed by one-dimensional extraction of the spectra. Thereafter, wavelength calibration was obtained using the IDENTIFY/DISPCOR IRAF tasks applied to a set of OH sky line spectra (each with about 35 sky lines in the range 15500Å -23000Å ). The typical error (1-) for this calibration process is estimated as 20Å which corresponds to half of the mean FWHM of the OH lines in the mentioned spectral range. Telluric atmospheric corrections were done using H- and K-band spectra of B type stars obtained before or after the target observation. The photospheric absorption lines present in the high signal-to-noise telluric spectra, were subtracted from a careful fitting (through the use of Voigt and Lorentz profiles) to the hydrogen and helium absorption lines (He absorption lines are sometimes seen at 1.70m and 2.11m in the earliest B-type stars), and respective adjacent continuum.
Optical spectroscopic CCD observations were performed with the Boller & Chivens spectrograph attached to the 2.5 m du Pont telescope at Las Campanas Observatory (Chile) in June 2010. The spectrum was obtained in the framework of the Galactic O-stars Spectroscopic Survey (GOSSS, Maíz Apellániz et al. (2010)), and therefore using its instrumental setup. A spectrum of HD 93129A was also obtained in June 2010, and a spectrum of HD 93162 (WR25) in May 2008, during the first GOSSS Southern campaign. Additionally, two optical spectra of WR20a were retrieved from the ESO Archive under the program 072.D-0082 (PI Rauw). These spectra were obtained with EMMI instrument attached to the ESO NTT at La Silla (Chile) in February 2004, and they belong to the same dataset analyzed and published by Rauw et al. (2005). Data were reduced and normalized using ONEDSPEC IRAF routines. A journal of the optical spectroscopic observations is presented in Table 2.
3 Results and Discussion
3.1 NIR spectra of WR20aa and WR20c
In Figure 1 we present the telluric corrected (continuum normalized) H- and K-band spectra of the WR20aa and WR20c stars, together with those of WR20a, WR20b and WR21a. The strongest features are the blends of Hi and Heii emission lines at 1.736m and 2.167m, as well as the Hei+Niii and Heii emission lines at 2.115m and 2.189m, respectively. A list with the main emission lines and corresponding equivalent line-widths is presented in Table 3.
The similarity between the NIR spectra of the known O2 If*/WN6 stars with that of WR20aa and WR20c (Figure 1) is remarkable, indicating that the later may also belong to the O2 If*/WN6 spectral type. This assumption is reinforced by the close morphological match of the WR20aa H- and K-band spectra with that for WR20a, one of the most massive binaries known to date (O3 If*/WN6+O3 If*/WN6) for which masses of 83M and 82M were derived (Bonanos et al., 2004; Rauw et al., 2004), and by the match of the WR20c H- and K-band spectra with that of WR21a, another extremely massive binary system (O3 If*/WN6 + early O), for which Niemela et al. (2008) estimated minimum masses of 87M and 53M, respectively.
Conti, Hanson & Morris (1995) showed that some O4 If stars (like HD 16691 and HD 190429) can be erroneously classified as WN stars if observed only in the K-band. In fact, their K-band emission features are albeit morphologically identical to those seen in the spectra presented in Figure 1. However, as can be seen in Blum et al. (1997), at least in the case of HD 190429 this is not true for the H-band. Indeed, the H-band spectrum of HD 190429 is completely different of those shown in Figure 1, being virtually featureless in the mentioned spectral range. Also Hanson et al. (2005) present the H-band spectrum of Cyg OB2 #7 (O3 If*) which shows several H and He lines in absorption. On the other hand, as can be noticed from Figure 1 the H-band spectra of WR20aa and WR20c are not featureless, with all H and He relevant lines clearly seen in emission. In this sense, H- and K-band spectra of heavily reddened massive star candidates may became useful tools to discriminate between the OIf* and OIf*/WN spectral types.
3.2 The optical spectrum confirms: WR20aa is an O2 If*/WN6 star
The spectral morphology derived from NIR spectrograms of one of the new sources (WR20aa), is confirmed from new optical spectroscopic observations. In Figure 2, we present the optical spectrum of WR20aa, together with GOSSS spectra of HD 93129A (O2 If*) and WR25 (O2.5 If*/WN6 + O) obtained with the same instrumentation and setup, along with the archival ESO/EMMI spectrum of WR20a (O3 If*/WN6 + O3 If*/WN6) (Crowther & Walborn, 2011). Table 4 lists the equivalent widths (EW) for the main diagnostic lines used for spectral classification of WR20aa, HD 93129A, WR20a, and WR25.
The blue spectrum of WR20aa is dominated by Heii 4686Å and Niv 4058Å emission lines, together with less pronounced emission lines of Niii 4634-40-42Å and Siiv 4089Å . Numerous absorption lines of Hi and Heii, and also some diffuse interstellar bands (DIBs) at 4428Å 4726Å 4765Å and 4865Å are present. A special mention deserve the strong Nv 4604-20Å absorptions, which are characteristic of the earliest O-type (O2-3) and WNha stars. These absorption lines show weak P-Cyg profile structure. The strength of the Niv 4058Å emission line relative to the Niii 4634-40-42Å emission lines indicates that the spectral type of WR20aa is at least as early as O2, according to the criteria described in Walborn et al. (2002). In addition, the presence of a weak P-Cyg profile in the H-beta line of WR20aa, compared to the same line in HD 93129A, which shows no emission, and the P-Cyg profile observed in WR25, is indicative of an intermediate spectral type between OIf* and WNha. In this way, Crowther & Dessart (1998) suggest that genuine WN6 stars are those with EW of He ii 4686Å lower than -12Å . The measurement of He ii 4686Å line in WR20aa (-9.6Å see Table 4) also indicates an intermediate spectral type. On the other hand, no detection of He i absorptions is also indicative of a spectral type earlier than the O3. This picture is seen in the direct comparison with the spectra of HD 93129A and WR20a, which shows that the blue spectrum of WR20aa resembles better to the later (as it was already noticed from the NIR regime). It is also noticeable from Table 4 that the emission lines of WR20aa are closer in strength to those from WR20a, being the lines from the later slightly broader (this difference is perhaps related to the binary nature of WR20a). Consequently, we classify the spectrum of WR20aa as O2If*/WN6 (Crowther & Walborn, 2011).
A spectral classification for WR20c using optical MK criteria is not possible in this moment due to the lack of a good signal-to-noise optical spectrum of this star. From inspection of our NIR H- and K-band spectra, we find that WR20c is spectroscopically very similar to WR20aa, consequently we propose that WR20c should also be classified as O2 If*/WN6.
3.3 The new O2 If*/WN6 stars and the Westerlund 2 cluster: are they related?
In Figure 3 it is shown a mosaic made from the 3.6m Spitzer IRAC images (Churchwell et al., 2004) of the region towards the RCW49 complex. There we identify WR20a, WR20b and WR21a together with WR20aa and WR20c. It is interesting to notice that the three already known WR stars are well displaced from the core of the Wd2 cluster, which harbors at least a dozen of early O-type stars (Rauw et al., 2007). The closest of them (WR20a), is located at about 0.6 arcmin from the cluster centre, with the others being placed at core distances of about 3.7 arcmin and 15.6 arcmin, respectively. The situation for WR20aa and WR20c is similar to that of WR21a, with the stars localized at angular distances (from the cluster centre) of 15.7 arcmin and 25 arcmin, well beyond the Wd2 cluster core. The presence of such massive stars well isolated in the outskirt of the RCW49 complex, naturally leave us to ask: how did they arrive there?
Interstellar extinction and absolute visual magnitudes
The assumption of a possible physical connexion between WR20aa and WR20c with the Wd2 cluster population, in principle can be tested by comparing the (B-V) colors of these new stars with those of the presently accepted cluster members (WR20a and the 12 known early O-type members), which we do in Table 5. For the sake of completeness, we also included here the colors of WR20b and WR21a, the other two known WR stars in the vicinity of Wd2. The (B-V) colors of WR20aa and WR20c were computed from the B- and V-band photometry shown in Table 1, while the values for the other stars were taken from Rauw et al. (2007) (WR20a and the early-O stars), Nazé, Rauw & Manfroid (2008) (WR20b), and Niemela et al. (2008) (WR21a). In the case of the early O-stars, we indicate the interval of (B-V) colors shown in Table 1 of Rauw et al. (2007). The corresponding visual extinctions were computed considering intrinsic colors (B-V)= -0.3 magnitudes, and the canonical value for the ratio of total to selective extinction R=3.1. Finally, absolute visual magnitudes were estimated assuming a distance of about 8 Kpc (Moffat, Shara & Potter, 1991; Rauw et al., 2007).
In case of WR20aa we notice that its derived (B-V) color and visual extinction are compatible with those
for the other massive stars in Wd2, while
a very red color index (B-V)=2.6 was obtained for WR20c, which corresponds to about
twice of the visual extinction inferred for the early-O stars of Wd2.
The higher interstellar extinction derived for WR20c is probably due to the presence of a foreground molecular cloud as can
be inferred from an inspection of the 2MASS J-band source density map
On the other hand, the absolute visual magnitudes derived for WR20aa and WR20c are similar to that for WR20b, which is the only one in the area of RCW49, not known to be a binary. Indeed, if we correct the absolute visual magnitudes of WR20a (O3 If*/WN6+O3 If*/WN6) and WR21a (O3 If*/WN6+ early-O) in order to take into account the binary companion contribution (by 0.7 and 0.5 magnitudes, respectively), we can see that WR20aa is probably as luminous as the other OIf*/WN or WN6ha stars in Wd2, with WR20c being a bit less luminous than them. This is not surprising if one considers that stars of same specific spectral types may span a range in absolute magnitudes (see Table 3 of Crowther & Walborn (2011)).
Two runaway stars?
It is an observational fact that very massive stars preferentially are found in the core of their parental clusters, normally forming binary or multiple systems. However, there are a certain number of very young and massive stars that are found well isolated in the field. Two canonical scenarios try to explain the origin of these stars. The first known as the binary-supernova scenario, considers the disruption of a short-period binary system from the asymmetric supernova explosion of one of the binary components, while the second, named dynamical ejection scenario, involves dynamical three- or multiple-body encounters in dense stellar systems.
The time scale for the binary-supernova scenario involves values larger than the expected age (1-2 Myrs) for WN6ha stars (Crowther et al., 2010) although lower mass counterparts could reasonably be rather older (approx. 2.5 Myr). On the other hand, the multiple ejection of massive stars from dense massive clusters is not a new idea (see discussion in Zinnecker & Yorke (2007)), and recent discoveries of very massive runaway stars (e.g. 30 Dor #16) highlighted the importance of this scenario in the evolution of massive clusters (Evans et al., 2010; Gvaramadze, Kroupa & Pflamm-Altenburg, 2010). For more details on the two models see for example the discussion in Gvaramadze & Gualandris (2010) and references there-in.
We searched for some observational constrain that could give support to the scenario in which WR20aa and WR20c were ejected from the Wd2 cluster centre following the dynamical ejection model. We found the surprising result that a line connecting WR20aa and WR20c (see Figure 3) intercepts the cluster core exactly where the star surface density hits a maximum. Also interesting, a vector conecting WR21a and the Wd2 core (see Figure 3) forms almost a right angle with the line connecting WR20aa and WR20c. This apparent configuration strongly resembles the dynamical scenario proposed by Gualandris et al. (2004) for the origin of AE Aur, Col and the very eccentric binary Ori, as runaway stars. From N-body simulations the authors conclude that these stars were ejected from their parental cluster as result of binary-binary interactions occurred in the Trapezium cluster. In this sense, it is remarkable to point out that the massive binary system WR21a is known to present a high eccentricity (, Niemela et al. (2008)).
The fine geometrical alignment between WR20aa, WR20c and the Wd2 core, suggests that these stars possibly had a common origin somewhere in the central part of the cluster, being ejected from their birthplace (in a timescale not greater than their own age of no more than 2 Myrs (Crowther et al., 2010), with minimum recession velocities (projected into the sky plane) of 18 km s and 28 km s, respectively. It is also interesting that Gvaramadze & Gualandris (2010) performed numerical simulations of a dynamical encounter between single massive stars and the very massive binary WR20a, founding that in a fly-by encounter the average recession velocity attained by a 70-80 M star could be quite moderate (less than 30 km s), being not formally classified as runaway.
Taking into account that the O2 If*/WN6 type stars are very rare and the perfect alignment of the WR20aa-cluster core-WR20c system, we postulate that both stars probably were formed somewhere in the core of the Wd2 cluster, being ejected from their birthplace by dynamical interaction with some other very massive star, in some previous stage of the cluster evolution. Of course, further observational (radial velocities, proper motions), and theoretical dynamical studies are still necessary to properly confirm our assumption. However, it is our opinion that the impressive alignment seen between such rare stars and the Wd2 cluster core, is in fact a strong clue favoring our hypothesis. In this sense, we speculate that WR20aa and WR20c may be two of the most interesting Galactic runways known to date.
In this work we report the detection of two new Galactic O2 If*/WN6 stars (WR20aa and WR20c) in the outskirt of the massive young cluster Wd2.
The similarity between the NIR spectra of WR20aa and WR20c with those of WR20a and WR21a (two of the most massive binaries known to date), is remarkable, suggesting that they could be members of the O2 If*/WN6 group. Indeed, the optical spectral morphology indicates that WR20aa presents an intermediate O2 If*/WN6 spectral type, based in the intensities of the emission lines of Heii 4686Å , Niv 4058Å , Niii 4634-40-42Å , the N v absorptions at 4604-20Å , and the P-Cyg profile observed in H. As an optical spectrum of WR20c is not yet available, we propose for it a spectral type similar to that of WR20a, based on the similarity of their NIR spectra.
From the analysis of the spatial distribution of the new O2 If*/WN6 stars through the mosaic made from the 3.6m Spitzer IRAC images of the region towards the RCW49 complex, we found the very interesting result that WR20aa and WR20c are placed at angular distances of 15.7 arcmin ( pc) and 25 arcmin ( pc) from the Wd2 core, respectively. The stars are well isolated in the outskirt of the RCW49 complex with the radius vector connecting them intercepting the cluster core exactly where the star density is maximum.
Considering the dynamical ejection model, we propose a scenario in which WR20aa and WR20c had a common origin somewhere in the central part of Wd2, being ejected from their birthplace in a timescale not greater than their own age, which is probably no more than 1-2 Myrs. Taking into account the rarity of such massive stars, and considering the perfect geometrical alignment observed between WR20aa, the cluster core, and WR20c, we believe that WR20aa and WR20c may represent one of the most interesting cases supporting the ejection of massive stars, produced by dynamical interaction with other massive companions.
We would like to thank the anonymous referee by the careful reading of the manuscript. Her/his comments were valuable to improve the clarity and presentation of the paper. We also would like to thank Dr. Nolan Walborn for stimulating discussions concerning the classification of intermediate OIf/WN6 stars. This work was partially supported by the ALMA-CONICYT Fund, under the project number 31060004, A New Astronomer for the Astrophysics Group - Universidad de La Serena, by the Department of Physics of the Universidad de La Serena and by the Research Direction (DIULS) of the Universidad de La Serena. This research has made use of the NASA/ IPAC Infrared Science Archive, which is operated by the Jet Propulsion Laboratory, California Institute of Technology, under contract with the National Aeronautics and Space Administration. This publication makes use of data products from the Two Micron All Sky Survey, which is a joint project of the University of Massachusetts and the Infrared Processing and Analysis Center/California Institute of Technology, funded by the National Aeronautics and Space Administration and the National Science Foundation. Also, this research has made use of the SIMBAD database, operated at CDS, Strasbourg, France. Partially based on observations made with ESO Telescopes at the La Silla Observatory under programs IDs 072.D-0082(A), 075.D-0210(A). ARL acknowledge financial support from DIULS through Project CD11103. RHB acknowledge financial support from DIULS through Regular Project PR10101.
- pagerange: Two O2 If*/WN6 stars possibly ejected from the massive young Galactic cluster Westerlund 2–LABEL:lastpage
- pubyear: 2010
- Blum, R. D., Ramond, T. M., Conti, P. S., Figer, D. F., Sellgren, K. 1997, AJ, 113, 1855
- Bonanos, A. Z., Stanek, K. Z., Udalski, A., Wyrzykowski, L., Å»ebruÅ, K., Kubiak, M., SzymaÅski, M. K., Szewczyk, O., PietrzyÅski, G., SoszyÅski, I. 2004, ApJ, 611L, 33B
- Conti, P.S., Hanson, M.M., Morris, P.W. 1995, ApJ, 445, L35
- Churchwell E., B. A. Whitney , B. L. Babler , R. Indebetouw , M. R. Meade , Christer Watson et al. 2004, ApJS, 154, 322
- Crowther, P. A., Dessart, L. 1998, MNRAS, 296, 622
- Crowther, P. A.,Schnurr, O., Hirschi, R., Yusof, N., Parker, R. J., Goodwin, S. P., Kassim, H. A. 2010, MNRAS, 408, 731
- Crowther P.A., Walborn N.R., 2011, MNRAS, in press (2011arXiv1105.4757C)
- Evans, C. J., Walborn, N. R., Crowther, P. A., HÃ©nault-Brunet, V., Massa, D., Taylor, W. D., Howarth, I. D., Sana, H., Lennon, D. J., van Loon, J. Th. 2010, ApJ, 715, L74
- Gamen, R., Gosset, E., Morrell, N., Niemela, V., Sana, H., Nazé, Y., Rauw, G., Barbá, R., Solivella, G. 2006, A&A, 460, 777
- Gvaramadze, V. V., Kroupa, P., Pflamm-Altenburg, J. 2010, A&A, 519, 33
- Gvaramadze, V. V., Gualandris, A. 2010, MNRAS, 410, 304
- Gualandris, A., Portegies Z., Simon, E., Peter P. 2004, MNRAS, 350, 615
- Hadfield, L. J., van Dyk, S. D., Morris, P. W., Smith, J. D., Marston, A. P., Peterson, D. E. 2007, MNRAS, 376, 248
- Hanson, M. M., Howarth, I. D., Conti, P. S. 1997, ApJ, 489, 698
- Hanson, M. M., Kudritzki, R.-P., Kenworthy, M. A., Puls, J., Tokunaga, A. T. 2005, ApJS, 161, 154
- Knödlseder, J. 2000, A&A, 360, 539
- Maíz Apellániz, J., Sota, A., Walborn, N. R., Alfaro, E. J., Barbá, R. H., Morrell, N. I., Gamen, R. C., Arias, J. I., Penadés Ordaz, M. 2010, arXiv:1010.5680
- Massey, P., Puls, J., Pauldrach, A. W. A., Bresolin, F.; Kudritzki, R. P.; Simon, T. 2005, ApJ, 627, 477
- Melnick, J. 1985 A&A 153, 235
- Moffat, A. F. J., Vogt, N. 1975, A&AS, 20, 125
- Moffat, A. F. J., Shara, M. M., Potter, M. 1991, AJ, 102, 642
- Moorwood, A., Cuby, J.G., Lidman, C. 1998, Msngr, 91, 9
- Nazé, G. Rauw, and J. Manfroid 2008, A&A, 483, 171
- Niemela V. S., Gamen R. C., Barbá R. H., Fernández Lajús E., Benaglia P., Solivella G. R., Reig P., Coe M. J., 2008, MNRAS,389, 1447
- Rauw, G., Vreux, J.-M., Gosset, E. 1996, RMxAC, 5, 108
- Rauw G., De Becker M., NazÃ©, Y., Crowther, P. A., Gosset E., Sana H., van der Hucht K. A., Vreux, J.-M., Williams, P. M. 2004 A&A, 420L, 9
- Rauw, G., Crowther, P. A., De Becker, M., Gosset, E., NazÃ©, Y., Sana, H., van der Hucht, K. A., Vreux, J.-M., Williams, P. M. 2005, A&A, 432, 985
- Rauw, G., Manfroid, J., Gosset, E., NazÃ©, Y., Sana, H., De Becker, M., Foellmi, C., Moffat, A. F. J. 2007, A&A, 463, 981
- Roman-Lopes, A. 2009, MNRAS, 398, 1368
- Shara, M. M., Smith, L. F., Potter, M., Moffat, A. F. J. 1991, AJ, 102, 642
- Skrutskie, M. F., Cutri, R. M., Stiening, R., Weinberg, M. D., Schneider, S., Carpenter, J. M., Beichman, C., Capps, R., Chester, T., Elias, J., and 21 coauthors 2006, AJ, 131, 1163
- Smith, N. & Conti, P. S. 2008, ApJ, 679, 1467
- Stephenson, C.B. and Sanduleak, N. 1977, ApJS, 33, 459
- Schnurr, O., Casoli, J., ChenÃ©, A.-N., Moffat, A. F. J., St-Louis, N. 2008, MNRAS, 389, 38
- Schnurr, O., Moffat, A.F.J., Villar-Sbaffi, A., St-Louis, N., Morrell, N. 2009, MNRAS, 395, 823
- van der Hucht, K. A. 2001, NewAR, 45, 135
- van der Hucht, K. A. 2006, A&A, 458, 453
- Walborn, N. R. 1982, ApJ, 254, L15
- Walborn, N.R. & Blades, J. C. 1997, ApJS, 112, 457
- Walborn, N.R., Howarth, I.D., Lennon, D.J., Massey, P., Oey, M.S., Moffat, A.F. J., Skalkowski, G., Morrell, N.I., Drissen, L., Parker, J. Wm. 2002, AJ, 123, 2754
- Westerlund, B. 1961, Arkiv för Astronomii, 2, 419
- Zinnecker, H., Yorke, H. W. 2007, ARA&A, 45, 481