A K-band spectral mini-survey of Galactic B[e] stars
A. Liermann, O.Schnurr, M. Kraus, A. Kreplin, M. L. Arias, L. S. Cidale
aa r X i v : . [ a s t r o - ph . S R ] J u l Mon. Not. R. Astron. Soc. , 1–10 (2014) Printed 22 October 2018 (MN L A TEX style file v2.2) A K -band spectral mini-survey of Galactic B[e] stars ⋆ A. Liermann , , O. Schnurr , M. Kraus A. Kreplin M. L. Arias , L. S. Cidale , Max-Planck-Institut f¨ur Radioastronomie, Auf dem H¨ugel 69, 53121 Bonn, Germany Astrophysikalisches Institut Potsdam, An der Sternwarte 16, 14482 Potsdam, Germany Astronomick´y ´ustav, Akademie vˇed ˇCesk´e republiky, Friˇcova 298, 251 65 Ondˇrejov, Czech Republic Departamento de Espectroscop´ıa Estelar, Facultad de Ciencias Astron´omicas y Geof´ısicas, Universidad Nacional de La Plata,Paseo del Bosque s/n, B1900FWA, La Plata, Argentina Instituto de Astrof´ısica de La Plata, CCT La Plata, CONICET-UNLP, Paseo del Bosque s/n, B1900FWA, La Plata, Argentina
Accepted 2014 June 12. Received 2014 June 11; in original form 2014 March 25
ABSTRACT
We present a mini-survey of Galactic B[e] stars mainly undertaken with the Large BinocularTelescope (LBT). B[e] stars show morphological features with hydrogen emission lines and aninfrared excess, attributed to warm circumstellar dust. In general, these features are assumedto arise from dense, non-spherical, disk-forming circumstellar material in which moleculesand dust can condensate. Due to the lack of reliable luminosities, the class of Galactic B[e]stars contains stars at very different stellar evolutionary phases like Herbig AeBe, supergiantsor planetary nebulae.We took near-infrared long-slit K -band spectra for a sample of Galactic B[e] stars with theLBT-Luci I. Prominent spectral features, such as the Brackett γ line and CO band heads areidentified in the spectra. The analysis shows that the stars can be characterized as evolvedobjects. Among others we find one LBV candidate (MWC 314), one supergiant B[e] candidatewith CO (MWC 137) and in two cases (MWC 623 and AS 381) indications for the existenceof a late-type binary companion, complementary to previous studies.For MWC 84, IR spectra were taken at different epochs with LBT-Luci I and the GNIRSspectrograph at the Gemini North telescope. The new data show the disappearance of thecircumstellar CO emission around this star, previously detectable over decades. Also no signsof a recent prominent eruption leading to the formation of new CO disk emission are foundduring 2010 and 2013.
Key words: infrared: stars – stars: winds, outflows – circumstellar matter – stars: emissionline, Be – supergiants.
B[e] stars are enigmatic objects. Their optical spectra show strongBalmer emission lines as well as permitted and forbidden emission ⋆ Based on data acquired using the Large Binocular Telescope (LBT) andGemini Observatory. The LBT is an international collaboration among insti-tutions in Germany, Italy, and the United States. LBT Corporation partnersare LBT Beteiligungsgesellschaft, Germany, representing the Max PlanckSociety, the Astrophysical Institute Potsdam, and Heidelberg University; Is-tituto Nazionale di Astrofisica, Italy; The University of Arizona on behalf ofthe Arizona university system; The Ohio State University, and The ResearchCorporation, on behalf of the University of Notre Dame, University of Min-nesota, and University of Virginia. Gemini Observatory is operated by theAssociation of Universities for Research in Astronomy, Inc., under a coop-erative agreement with the NSF on behalf of the Gemini partnership: theNational Science Foundation (United States), the National Research Coun-cil (Canada), CONICYT (Chile), the Australian Research Council (Aus-tralia), Minist´erio da Ciˇencia, Tecnologia e Inovac¸˜ao (Brazil) and Ministe-rio de Ciencia, Tecnolog´ıa e Innovaci´on Productiva (Argentina). lines of lowly-ionized metals. In addition, B[e] stars display a near-and mid-infrared excess that is attributed to hot and warm circum-stellar dust. However, because the definition of the B[e] class ispurely morphological, it contains objects that are physically verydifferent in terms of their initial mass and evolutionary phase: B[e]supergiants (B[e]SGs), Herbig AeBe (HAeBe) stars, compact plan-etary nebulae, and symbiotic objects (e.g., Lamers et al. 1998). Es-pecially B[e]SGs and HAeBe stars are difficult to distinguish; theyhave dense, cool, and dusty equatorial disks, which are related toeither accretion and disk winds (in the case of HAeBe stars) orequatorially enhanced stellar winds (B[e]SGs). These disks giverise to molecular emission such as the first overtone bands of car-bon monoxide (CO) in the near-infrared which are observed in bothHAeBe stars and B[e]SGs (Bik et al. 2006, McGregor et al. 1988b,Morris et al. 1996). In addition, the position of HAeBe stars in theempirical Hertzsprung-Russell diagram overlaps with that of thelow-luminosity B[e]SGs.So far, comprehensive studies of B[e] stars, focused on thespectral classification and characterization, were mostly based on c (cid:13) A. Liermann et al.
Table 1.
Log of the LBT observations.Object RA DEC Date (UT) t int [s]Science:MWC 137 06 18 45.52 +15 16 52.25 2010-11-10 900MWC 84 04 19 42.14 +55 59 57.70 2010-11-10 120MWC 314 19 21 33.97 +14 52 56.89 2010-11-13 240AS 381 20 06 39.95 +33 14 28.10 2010-11-13 600MWC 300 18 29 25.69 -06 04 37.29 2011-05-12 200MWC 623 19 56 31.54 +31 06 20.12 2011-05-12 80Standards:HD 44585 06 23 09.15 +15 50 32.33 2010-11-10 1440HD 29371 04 40 49.54 +57 52 46.63 2010-11-10 600HD 185195 19 37 17.84 +15 15 02.41 2010-11-13 900HD 196006 20 33 30.39 +32 54 27.78 2010-11-13 900HD 175644 18 56 47.20 -14 01 40.76 2011-05-12 400HD 191720 20 10 01.65 +36 58 42.47 2011-05-12 400 optical spectra. Over the last two decades new instruments havegiven access to the infrared (IR) spectral range and allowed high-quality spectra with the necessary spectral resolution to be ob-tained. But only now projects are checking systematically how ap-propriate spectral classification based on IR spectra alone is forgeneral application purposes. For example, Oksala et al. (2013)find that based on K -band spectra three distinct groups of stars canbe identified: (1) “regular” B[e]SGs with the expected spectrum ofemission lines including the Pfund series and CO in emission, (2)S Dor-like luminous blue variables (LBVs) with a variety of strongemission lines but lacking the expected circumstellar CO emission,and (3) a group of cool stars, consisting of LBVs in outburst andYellow Hypergiants (YHGs), showing the Pfund series in absorp-tion. The CO bands can be used to obtain an age estimate fora star. As proposed by Kraus (2009), stellar evolution mod-els with rotation predict the enhancement of the carbon iso-tope C on the stellar surface during the core-hydrogen burn-ing of massive stars through mixing. Via mass loss, it is trans-ported into the circumstellar environments and locked into COmolecules. Hence, in evolved stars (i.e., supergiants), the enrichedisotope should become detectable as CO bands. This was con-firmed by detection of CO emission in known, extra-galacticB[e]SGs (Liermann et al. 2010); vice versa the CO absorptioncan be used to distinguish late-type supergiants from dwarf stars(Wallace & Hinkle 1997).We have embarked on a mini-survey of a sample of Galac-tic B[e] stars to investigate their evolutionary status and character-ize their age and circumstellar material from K -band spectra. Wepresent our observations in Sect. 2 with the results of the spectralanalysis following in Sect. 3. The final part of the paper is dedicatedto the discussion and conclusion of our findings (Sects. 4 and 5). Our sample stars (cf. Table 1) were observed with the Large Binoc-ular Telescope (LBT) in Arizona, USA, during 3 runs in November2010, April and May 2011. We used the LBT-Luci I spectrograph(Seifert et al. 2003) to obtain seeing-limited, high-quality, long-slitspectra with the N1.8 camera, a slit width of 0.5 ′′ , and the K -bandfilter (1.93–2.48 µ m) with the 210 zJHK grating tilted to center on λ = 2 . µ m to have best wavelength coverage for the CO bands redwards of 2.29 µ m. The resulting spectra have a spectral resolu-tion of about R = ≈ R =18,000),using the 110.5 l/mm grating, the long camera (0.05 ′′ /pix), andthe 0.10 ′′ slit. We obtained spectra in the K -band (2.28–2.34 µ m)centered on λ = 2 . µ m to cover the first two CO bandheads.Data sets from both telescopes were reduced using standard IRAF routines for long-slit spectroscopy including removal of cos-mic rays and dead pixels, dark frames, division by a normalized flatfield, distortion correction, and wavelength calibration. Sky sub-traction was done by subtracting offset frames per target (AB pat-tern) correspondingly before the spectra extraction.The frames of the telluric standard stars were treated in thesame way. We produced calibration curves as the ratio of the ex-tracted standard star spectra and Kurucz models (corresponding tothe spectral types of the standard stars, Kurucz 1993) scaled to thestars’ 2MASS K -band magnitudes. As the Kurucz models do notreproduce the observed Br γ line in the standard star spectra thatpart of the calibration curve was smoothed by linear interpolationbetween 2.15 and 2.17 µ m. For flux calibration, the extracted sci-ence target spectra were divided by the corresponding calibrationcurves. The flux-calibrated spectra of the sample stars are presented inFig. 1. Most spectra show more or less pronounced telluric residu-als at about 2.316 and 2.370 µ m. We attribute this to the not alwaysperfect match of observing conditions between the science targetand standard star.For all sample stars we clearly detect the Brackett γ line (Br γ , λ µ m) in emission. Additionally, iron (Fe II λ µ m)and a magnesium doublet (Mg II λ µ m) are detected.Three stars, MWC 314, MWC 84, and AS 381, also show featuresof the sodium doublet (Na I λ µ m) and helium (He I λ µ m) in emission. In Table 2 we list the measured equiva-lent widths of these lines for all stars.Remarkably, MWC 314 is the only star whose spectrum showsvery prominently lines of the Pfund series (see Fig. 1, top-mostpanel). Overall its spectral appearance is close to S-Dor like Lu-minous Blue Variables (LBVs) and B[e]SGs (Oksala et al. 2013);a classification of MWC 314 as LBV will be discussed below (seeSect. 4.1). Also, no molecular CO bands, neither in emission nor inabsorption, are present in the spectrum.However, CO bands are detected in the spectra of three stars:MWC 137 (in emission), MWC 623 and AS 381 (in absorption).Details of the spectra are shown in Figs. 3 and 6. MWC 137 dis-plays CO in emission which shows that the star is evolved, i.e.possibly a B[e]SG. IRAF is distributed by the National Optical Astronomy Observatories,which are operated by the Association of Universities for Research inAstronomy, Inc., under cooperative agreement with the National ScienceFoundation. c (cid:13) , 1–10 -band mini-survey of Galactic B[e] stars MWC 314 - B2 F e II H e I F e II M g II H e I B r γ / H e I H e I N a I F e II H e I H I Pfund series -12.6-12.3-12.0-11.7 MWC 137 - B0 F e II H e I F e II F e I ? M g II H e I B r γ / H e I F e II C O ( - ) H e I C O ( - ) C O ( - ) C O ( - ) C O ( - ) C O ( - ) -13.0-12.8-12.6 MWC 300 - B1 Ia F e II H e I F e II F e I ? M g II H e I B r γ / H e I H e I F e II -12.8-12.7-12.6-12.5 MWC 84 - B0-B2 F e II H e I F e II ? F e I ? F e II M g II H e I H e I B r γ / H e I H e I H e I N a I H e I -11.90-11.85-11.80 l og F λ [ e r g s - c m - A o - ] MWC 623 - B4 III + K2 Ib-II F e II M g II H e I H e I B r γ / H e I M g II N a I C O ( - ) C O ( - ) C O ( - ) C O ( - ) C O ( - ) C O ( - ) -12.6-12.4-12.2 EM* AS 381 - B1 + K0 I-II F e II H e I F e II ? F e I ? M g II H e I ? H e I B r γ / H e I H e I N a I C O ( - ) C O ( - ) C O ( - ) C O ( - ) C O ( - ) C O ( - ) -13.3-13.2 2.10 2.15 2.20 2.25 2.30 2.35 2.40 λ [ µ m] Figure 1.
Flux-calibrated K -band spectra of our sample stars observed with LBT-Luci I. Line identifications of prominent emission lines and the band headsof CO and CO are indicated. Please note that the scaling of the flux axis is different for each panel. Spectral types listed for the individual stars are takenfrom the literature as listed in Table 4.
Both MWC 623 and AS 381 have been suspected to be bi-nary stars; for MWC 623 Zickgraf & Stahl (1989) find a spectro-scopic binary with two sets of spectral lines (SB2) and for AS 381Miroshnichenko et al. (2002a) find neutral metal lines inferring acool companion. Indeed, the spectra of both stars display CO ab-sorptions which we attribute to a cool late-type companion, respec-tively. We measure the equivalent widths of the first band head (CO 2-0), see Table 2. Following Gonz´alez-Fern´andez et al. (2008, seeEqs. (5) and (6)), the equivalent width can be used to derive boththe effective temperature and a spectral-type of the companion (theindicator G = G = c (cid:13) , 1–10 A. Liermann et al.
Table 2.
Equivalent width measurements (in ˚A ) for prominent emission lines (negative) and absorption lines (positive).star Fe II He I Mg II Br γ Na I CO (2-0) CO (3-1)2.089 µ m 2.112 µ m 2.138/144 µ m 2.166 µ m 2.206/209 µ m 2.293 µ m 2.322 µ mMWC 314 -2.0 -3.6 -3.3 -2.7 -32.1 -5.1 -5.0 - -MWC 137 -0.9 - -0.8 -0.4 -25.6 - - -21.1 -19.1MWC 300 -1.9 - -0.3 -0.5 -8.7 - - - -MWC 84 -0.4 -3.4 -0.3 -0.4 -4.3 -0.1 -0.2 - -MWC 623 -0.9 - -0.8 - -5.8 - - 14.0 11.6AS 381 -0.5 -0.8 -1.2 -1.3 -3.6 -0.6 -0.4 5.5 5.3 temperature of T eff = 4030 ± K. In the case of AS 381, we de-termine a spectral type of the companion of K0 ( G = T eff = 4550 ± K. Errors on the effective temperatureare dominated by those given in Eq. 5 of Gonz´alez-Fern´andez et al.(2008). In both cases one can argue for the detection of weak COabsorptions in the spectra, indicating a slightly evolved giant or su-pergiant companion (luminosity class of II to I).Both MWC 84 and MWC 300 do not show any Pfund linesor CO emission in their spectra. However, MWC 84 shows thestrongest He I and MWC 300 the strongest Fe II and Br γ lines inemission, of all the sample stars. For high-mass stars, stellar evolution models predict a short tran-sitional phase as LBV from the core hydrogen-burning OB starprogenitors to the core helium-burning Wolf-Rayet stars (e.g.Maeder et al. 2008). During the LBV phase the stars undergo ex-treme mass loss events (“outbursts”) after or followed by a ratherstable (“quiescent” or “dormant”) state and are often associatedwith circumstellar nebulae. In addition, significant variability inbrightness and spectral appearance can be detected more readilyin the outburst phase but also are present in the quiescent state.For MWC 314 spectral, photometric, and polarimetric vari-ability was found covering different periods, e.g. Miroshnichenko(1996), Wisniewski et al. (2006), Groh et al. (2007). A detailedspectral analysis by Miroshnichenko et al. (1998), finding photo-spheric lines for the first time, classified the star as B0 super-giant with T eff = 25 , K. Additionally, the authors present in-dications for a non-spherical wind and derive a distance of d =3 . ± . kpc from radial velocity ( RV ) measurements. This makesMWC 314 one of the most luminous stars in the Milky Way with log ( L/L ⊙ ) = 6 . ± . (Miroshnichenko et al. 1998).Later studies find different (effective) temperatures, basedon different indicators and methods, e.g, 26,700 K to 32,000 K(Cidale et al. 2001), 16,200 K (Carmona et al. 2010), 18,000 K(Lobel et al. 2013). At the same time, different studies re-port the rather stable photometric appearance of MWC 314 in UBV RIJHK filter observations with about . mag variabil-ity, e.g. Bergner et al. (1995), Miroshnichenko (1996, covering thetime span from 1954 to 1995). This results in a conflicting situationto determine the spectral type and evolutionary state of MWC 314.The star was first proposed as LBV candidate byMiroshnichenko (1996). In case the above listed tempera-tures reflect a real change over the last 20 years, MWC 314 wouldbe rather similar to other LBVs (see Table 3), thus supporting the classification of MWC 314 as LBV candidate. In Fig. 2, we showMWC 314 in the Hertzsprung-Russell diagram (HRD) amongknown Galactic LBVs and LBV candidates. It is located betweenthe empirical Humphreys-Davidson limit (Humphreys & Davidson1994, solid line in Fig. 2) and the hot temperature LBV minimumlight strip (Clark et al. 2005, dashed line in Fig. 2).In our spectrum of MWC 314 we find pronounced lines ofhydrogen, Br γ and Pfund series, and prominent emission linesof Fe II , Mg II , and Na I . Comparing the spectra of MWC 84,MWC 137, and MWC 314, we find most pronounced Na I and Mg II emission for the latter with remarkable line strengths. Typically,comparable line strengths are only found in late-type (F and G) su-pergiants (Hanson et al. 1996, Wallace & Hinkle 1997). In B-typeemission-line stars these lines are usually (much) weaker, but witha slight trend for an increase towards later types (see Oksala et al.2013). If we consider that the temperature obtained by Lobel et al.(2013) within the past years is the currently most reliable one, thenwe can assign MWC 314 a spectral type of B2 according to thetemperature-spectral type relation for Galactic supergiant stars ofSearle et al. (2008).Previous studies by Muratorio et al. (2008) found indica-tions for a quasi-Keplerian circumstellar disk from double-peakedemission lines. We do not detect any significant IR excess that onewould expect from warm/hot dust in the circumstellar disk; this re-sult is highly indicative for the absence of such dust. We also do notfind any CO emissions or absorptions. This is comparable to otherLBV candidates where the lack of CO emission has been attributedto the circumstellar material having too low a density (Morris et al.1996, Oksala et al. 2013).The presence of a dense and compact gas disk in MWC 314can be concluded from the detection of double peaked [Ca II ] emis-sion together with the non-detection of [O I ] by Aret et al. (in prepa-ration). Aret et al. (2012) show, that [Ca II ] traces regions of higherdensities, i.e. closer to the star. Thus the lack of [O I ] in MWC 314must be attributed to a dense but compact disk. On the other hand,the presence of the strong emission from neutral sodium, whichhas a very low ionization potential, indicates the existence of a sec-ond very dense, but cool region of circumstellar material. To allowfor the emission of Na I , that region should be shielded from di-rect stellar radiation. In a spherically-symmetric stellar wind, toohigh densities would be required. But in such a high density en-vironment the emission of [O I ] can be expected. Hence, it seemsmore logical to assume that the Na I emission arises from a cool cir-cumstellar disk (or ring), which provides ideal shielding conditions(e.g. , Scoville et al. 1983, McGregor et al. 1988a,b). However, at A disk in Keplerian rotation displaying simultaneously a small outflowcomponent, see e.g., Krtiˇcka et al. (2011), Kurf¨urst et al. (2013).c (cid:13) , 1–10 -band mini-survey of Galactic B[e] stars Table 3.
Comparison of stellar parameters for LBV stars.star T eff [K] log L/L ⊙ referenceMWC 314 (1986/91) 25 000 6.10 (1)MWC 314 (1997/98) 32 000 - (2)MWC 314 (2007) 16 200 - (3)MWC 314 (2009-2011) 18 000 5.84 (13) η Car 25 000 6.57 (4) η Car 9 400 6.70 (11) η Car 35 300 6.74 (12)AG Car (min 1985-1990) 22 800 6.17 (5)AG Car (min 2000-2001) 17 000 6.17 (5)AG Car (max 2002-2003) 14 000 6.00 (6)FMM 362 11 300 6.25 (7)Pistol star 11 800 6.20 (7)AFGL 2298 (min 2006) 10 300 6.30 (14)AFGL 2298 (max 2001) 15 000 6.10 (8) ζ Sco 19 500 6.02 (9)P Cyg (min 1980-2000) 18 200 5.70 (9)HR Car 17 900 5.70 (10)(1) Miroshnichenko et al. (1998), (2) Cidale et al. (2001),(3) Carmona et al. (2010), (4) Humphreys & Davidson (1994),(5) Groh et al. (2011), (6) Groh et al. (2009b), (7) Najarro et al. (2009),(8) Clark et al. (2003), (9) van Genderen (2001), (10) Groh et al. (2009a),(11) Groh et al. (2012), (12) Hillier et al. (2001), (13) Lobel et al. (2013),(14) Clark et al. (2009) this point we cannot exclude the option that the Na I emission arisesfrom the environment of the companion.Oksala et al. (2013) present the K -band spectrum ofLHA 120-S 127 and S Dor which look remarkably similar toMWC 314, finding disk tracers and no CO emission. The authorsstate that CO molecules should form between the locations tracedby the Ca and O emitting material and dust particles, and concludethat the disk of S 127 might not be a continuous disk but rather con-sisting of rings/shells. This cannot be explained with the model ofa continuous B[e] wind but rather requires an LBV eruption sce-nario. Hence, for MWC 314 we can assume a similar scenario, i.e.,the presence of at least two different rings, a hot and compact oneclose to the star from which the [CaII] lines arise, and very coolone at much larger distance, where the NaI lines are excited. Theintermediate region must be of very low density due to the lack ofboth [OI] and CO band emission.A study by Marston & McCollum (2008) found a bipolar neb-ula around MWC 314 that is similar in morphology to the onearound η Car. Their results are not conclusive whether or not thatnebula was ejected in a past LBV outburst or not.Recently, it has been discussed for a large fraction of theknown LBV stars, that bipolar nebulae might be linked to a pos-sible binary nature of these stars. For MWC 314 the binary sta-tus is still under debate. Early reports of photometric variations(Miroshnichenko 1996) were interpreted as being more consistentwith the pulsations of a slightly evolved supergiant (or LBV candi-date) rather than being attributed to a binary companion. However,Muratorio et al. (2008) find indications for binarity based on RV variations with an orbital period of P = 30 . d. More recent studiescontinue the ambiguity of the binary status, e.g. Rossi et al. (2011)cannot detect any periodic RV variations. However, Lobel et al.(2013) find variations with a period of P = 60 . d (and an orbitaleccentricity of e = 0 . ), inferring a massive, possibly evolvedsupergiant companion to MWC 314. Our spectrum does not showany CO absorption indicative for cool companions. Thus K and M- eff /K) l og ( L / L ) η Car P Cyg AG Car HR Car AFGL 2298 ζ Sco Pistol star FMM 362
MWC 314
Figure 2.
HRD for known Galactic LBV stars (circles) and LBVcandidates (squares). The empirical Humphreys-Davidson limit(Humphreys & Davidson 1994, solid line) and the hot LBV minimumlight strip (Clark et al. 2005, dashed line) are indicated for illustration. ForMWC 314, we show a range of fundamental parameters listed in Table 3;for temperatures with no simultaneous luminosity determinations we adopt log (
L/L ⊙ ) = 6 . from Miroshnichenko et al. (1998). MWC 137 C O ( - ) C O ( - ) C O ( - ) C O ( - ) C O ( - ) C O ( - ) -13.0-12.92.25 2.30 2.35 2.40 λ [ µ m] l og F λ [ e r g s - c m - A o - ] Figure 3.
Detail of the spectrum of MWC 137 (solid line) showing the COemission. The over-plotted model (dotted line) combines CO and Pfundemission added to the observed continuum. type companions can be excluded for MWC 314. However, a mas-sive companion as suggested by Lobel et al. (2013) cannot be ruledout. Further high-resolution spectroscopy and/or spatially-resolvedimaging is necessary to confirm or contradict the binary suspicion. CO emission
This star is a typical example for the difficulty to distinguishB[e]SG and HAeBe stars. Based on the preferred distance estimate,the star has been considered to be a young unevolved object, e.g.with a low luminosity around log (
L/L ⊙ ) = 4 . , assuming d =1 . kpc (Hillenbrand et al. 1992). Also, Testi et al. (1997) reportthe detection of an embedded cluster around MWC 137 support- c (cid:13) , 1–10 A. Liermann et al.
MWC 84
Gemini/GNIRS 2013 Gemini/GNIRS 2011 (nflux + c) LBT/Luci I 2010 (nflux + c) C O ( - ) C O ( - ) C O ( - ) λ [ µ m] no r m a li ze d f l ux [ a . u . ] Figure 4.
Comparison of the spectra of MWC 84 taken at the LBT andGemini; spectra are shifted by a constant for better viewing. Over the cov-ered time span of 3 years no CO emission is detected. eff /K) l og ( L / L ) MWC 137 MWC 623 AS 381 MWC 84 MWC 300 MWC 314
Figure 5.
HRD with our sample stars (stellar parameters listed in Ta-ble 4) and single star stellar evolution models for different initial masses byEkstr¨om et al. (2012). The models account for the effects of stellar rotation. ing the young nature of this object. However, Esteban & Fernandez(1998) derived T eff = 30 , K and log (
L/L ⊙ ) = 5 . based ona re-assessment of the distance to a lower limit of d = 6 kpc, andconcluded that the star should be classified as B[e]SG. In additionthe authors argue that the small photometric variations but stablespectral line profiles found for MWC 137, indicate the star being aB[e]SG rather than a HAeBe star (see also Zickgraf 1992). More-over, Esteban & Fernandez (1998) find the associated ring nebulaS 226 around MWC 137 to be isolated and not attached to any large-scale star-forming region as might be expected for an HAeBe star.In contrast these authors remark the spectroscopical resemblanceof the nebula to those observed around LBVs or Wolf-Rayet stars(Esteban & Fernandez 1998, and references therein). Marston & McCollum (2008) confirm the ring nebula and inaddition find a bipolar structure in their narrow-band images. How-ever, the nebula material seems chemically unprocessed whichmight indicate that the circumstellar material either is swept-upinterstellar matter or was ejected during an early stellar-evolutionphase.MWC 137 was listed by Miroshnichenko (2007) as FS CMacandidate star which implies that it is a binary system. How-ever, neither Baines et al. (2006) nor Wheelwright et al. (2010)found evidence for the binary status of MWC 137 from theirspectro-astrometric observations (a method sensitive down toabout 100 mas binary separation, e.g., Bailey 1998). HIPPAR - COS data suggest that MWC 137 is an astrometric binary(Makarov & Kaplan 2005), possibly with a white dwarf or subd-warf companion (Lanning & L´epine 2006).In our spectrum (see Fig. 1), we do not find indications fora cool companion in terms of CO absorption or a set of emis-sion or absorption lines accountable to a hot main-sequence com-panion; instead, a closer look at our spectrum shows prominentCO bandheads in emission, both of CO and CO. In Fig. 3 weover-plot a model combining CO and Pfund emission added to theobserved continuum. The model was computed using the codesof Kraus et al. (2000), Kraus (2009) and Oksala et al. (2013); pa-rameters of the Pfund and CO emitting regions as obtained byOksala et al. (2013). The excellent agreement of the model and theobservations confirms the results by Oksala et al. (2013). In partic-ular the presence of clearly detectable emission from CO impliesthat the circumstellar material is enriched in the C isotope. Thisexcludes a pre-main sequence nature according to the stellar evo-lution models by Ekstr¨om et al. (2012) and MWC 137 has to beclassified as evolved, post-main sequence object.
In two further stars of our sample, MWC 300 and MWC 84, we donot detect CO emission in their spectra. However, in the followingwe summarize the evidences for their classification as B[e]SG:
MWC 300
According to Appenzeller (1977) this star is classi-fied as B1 Ia. Miroshnichenko et al. (2004) detect photosphericlines and their spectral analysis indicates the supergiant status ofthe star, with T eff = 20 , K. From comparison of equivalent-width measurements they derive a luminosity of log (
L/L ⊙ ) =5 . ± . assuming a distance of d = 1 . ± . kpc. RV varia-tions (Miroshnichenko et al. 2004) and a clear signal from spectro-astrometry (Takami et al. 2003) suggest that MWC 300 is a bi-nary. From near-infrared interferometry, Wang et al. (2012) find ev-idence for a dusty circumstellar (maybe even circumbinary) diskand a binary companion at a projected separation of about 7.9 AU.No extended circumstellar material has been detected in H α imag-ing by Marston & McCollum (2008).Our spectrum shows a clear infrared excess that we attributeto the circumstellar/circumbinary dust, but we do not detect COemission. As for MWC 314, this might indicate that the material’sdensity is too low to give rise to molecular emission. In addition, wedon’t find evidence for a cool companion by the lack of CO absorp-tion expected for late-type stars. As Miroshnichenko et al. (2004)suggests a companion of about 6 M ⊙ , i.e. a B-type star, we scannedthe spectrum for the expected hydrogen absorption lines with noresult. This might indicate a supergiant companion for which thespectral lines are less pronounced, such that basically only a con-tribution to the continuum flux might be considered. Interestingly, c (cid:13) , 1–10 -band mini-survey of Galactic B[e] stars models of Wang et al. (2012), fitted to their interferometric data,suggest the two binary components to be about similar in effectivetemperature and to have a brightness ratio of about 2.2. MWC 84
Also known as Cl Cam, MWC 84 had a spectacular out-burst in March 1998, detected from γ -rays to radio emission (e.g.Clark et al. 1999, Robinson et al. 2002). Hynes et al. (2002) classi-fied the star as spectral type B0-B2, finding its emission-line spec-trum typical for a B[e]SG. The authors refer to the star as an atyp-ical high-mass X-ray binary but speculations about the existenceand nature of the compact companion are ongoing. Hjellming et al.(1998) describe the star as X-ray binary/ X-ray transient radiosource, and detect a slow, decelerating shell in radio emission.However, the nature of this emission is also still under debate (e.g.Rupen et al. 2003). In addition, a faint circumstellar shell has beendetected by Marston & McCollum (2008), but the authors find noclear evidence to determine whether this is stellar ejecta or the illu-minated local interstellar material.The analysis of spectral line profiles, the derived extinction,and interferometric observations suggest that MWC 84 has an equa-torial disk wind with a dust-free high-temperature zone close tothe star, and that it is viewed almost pole-on (e.g. Thureau et al.2009, Hynes et al. 2002, Miroshnichenko et al. 2002b). The lat-ter authors revised the distance estimate based on RV measure-ments and interstellar Na I D-lines, to about 2.2 kpc, placing thestar in the Perseus arm; with that distance, an upper-limit luminos-ity of log
L/L ⊙ < is derived. Other studies find both smaller( d = 1 . − . kpc by Barsukova et al. 2006, Clark et al. 2000, ≈ kpc) and larger (e.g. d ≈ kpc Robinson et al. 2002) dis-tances, hampering reliable estimates of the luminosity.Clark et al. (1999) present J, H and K -band spectra obtainedone month after the outburst that are rich in hydrogen, heliumand iron emission lines and clearly show lines of the Pfund se-ries and the presence of CO emission in the K -band. The authorsargue that the molecular emission likely arises due to collisionalexcitation from regions shielded from the stellar radiation, whichrequires high densities. In addition, Clark et al. (2000, covering1998 to 1999) and Thureau et al. (2009, data for 2004 and 2005)present UBV RIJHK photometry showing that the star is slightlybrighter than in pre-outburst state, but rather stable over the longtime range of seven years. Also near-infrared interferometric data(Thureau et al. 2009, covering 1998 to 2006) have supported thescenario of a stable circumstellar disk in the last decade.Our spectra (see Fig. 4) do not show any indication of Pfundlines or CO emission or absorption, neither in 2010 (LBT-Luci I)nor in 2011 or 2013 (both Gemini-GNIRS). We attribute this dis-appearance of Pfund and CO emission to the loss/dilution of thehigh-density circumstellar material and the return of MWC 84 toits pre-outburst stage.
For both MWC 623 and AS 381, it has been suggested that thestars are in binary systems with a cool companion. Our finding ofprominent CO absorption bands supports this position and we de-termined the spectral types for the companions from the CO bandhead equivalent widths, see Sect. 3 and Table 2.
MWC 623
Based on the detection of a set of early-type (emission)and late-type (absorption) optical lines, Zickgraf & Stahl (1989)conclude that MWC 623 is a spectroscopic binary (SB2) of type B2+K2. However, they find no indication for RV variations, at-tributing this to a binary period longer than covered by their rangeof observational data of 2 years. Zickgraf (2001) reclassifies thespectral types of the binary components to B4 III+K2 II-Ib but noperiodic RV variations are found even in long-time observations(Zickgraf 2001, Polster et al. 2012). This might indicate a pole-onorientation of the system, e.g. Miroshnichenko (2006).Based on the K star’s luminosity class, Zickgraf (2001) derivesa spectroscopic distance of d = 2 . kpc. Our derived spectral typeof the cool component is a slightly cooler K4 I-II star, but with thesame range for the luminosity class. We take MWC 623’s 2MASSmagnitude K S = 5 . mag and assume that the K -band emissionis dominated by the cool companion. With the distance and red-dening as listed in Table 4, and bolometric corrections determinedfollowing Levesque et al. (2005), we derive the companion’s lumi-nosity as log ( L/L ⊙ ) = 3 . . Assuming that MWC 623 is a phys-ical long-period binary and not a chance superposition detected inthe spectrum, each component in all likelihood has evolved likea single star. From comparison with the stellar-evolution track of7 M ⊙ initial mass (see Fig. 7), we find an age difference of 1.8 Myrof the components. This seems negligible compared to the total ageof the system of about 50 Myr. AS 381
The star was first listed in the H α surveys byMerrill & Burwell (1950) and Henize (1976) as emission-linestar; Th´e et al. (1994) classify it as Be star with IR excess.Miroshnichenko et al. (2002a) detected absorption lines of neu-tral metals in their near-IR spectra and CO absorption bands,concluding that the star is a B1 + K binary system. The authorsalso derive a lower limit for the orbital period of P ≈ d.They estimate the distance to be d = 4 ± kpc deriving lumi-nosities of log ( L/L ⊙ ) = 4 . and log ( L/L ⊙ ) = 3 . for theB and K component, respectively. In addition, the authors claimto find signs of ongoing mass transfer between the binary com-ponents. Miroshnichenko (2007) lists an effective temperature of log ( T eff / K) = 4 . for the B star and an interstellar reddening of E ( B − V ) = 2 . mag.As described in Sect. 3, we derive a spectral type of K0 for thecompanion from the equivalent width measured for the first CObandhead. However, the determination of the luminosity class is notthat unambiguous, as a clear detection of CO is not confirmed.Assuming that the flux in the infrared-wavelength range isdominated by the cool companion, we use the mean K -band mag-nitude of those listed by Miroshnichenko et al. (2002a, c.f. Table 2), K = 6 . mag, to derive the companion’s luminosity. Includingthe distance and reddening as listed in Table 4, and the bolometriccorrections for the K -band according to Levesque et al. (2005), weobtain log ( L/L ⊙ ) = 3 . ± . . Within the errors, this result isrobust enough to allow up to 25% of the K -band flux to be con-tributed from the B-type companion.According to the stellar evolution tracks in Fig. 7, we deter-mine lower-limit initial masses of about 16 M ⊙ and 10 M ⊙ , andages of about 13 Myr and 25 Myr for the B-type and K-type star,respectively. Fig. 7 also suggests that the B component of AS 381,although more massive, seems to be less evolved than the K com-panion. Given the short lower-limit orbital period of the system,it seems possible to consider that mass transfer might have hap-pened at some stage of the evolution of this binary star. In that casesingle-star stellar evolution models are of course not adequate formass and age determination. c (cid:13) , 1–10 A. Liermann et al.
Table 4.
Stellar parameters for the sample stars.star SpT T eff log ( L/L ⊙ ) d E ( B − V ) M ini age a reference[K] [kpc] [mag] [ M ⊙ ] [Myr]MWC 314 B2 ± ±
40 6 (1),(8)MWC 137 B0 30000 5.37 >
25 8.0-8.4 (2)MWC 300 B1 Ia 20000 5.1 ± ± ±
20 9.5-10.3 (7)MWC 84 B0-2 20000 ± < ± (5), (6)MWC 623-a B4 III 17200 ± ± ± (3)MWC 623-b K2 Ib-II 4300 ±
200 3.5 ± K4 I-II 4030 ±
100 3.6 7 52 this workAS 381-a B1 19000 4.9 4 ± ± ± ∼ (4)AS 381-b K 3.6 4 ± K0 I-II 4550 ±
100 3.94 ± this workParameters given in italic are from this work, others according to the indicated references: (1) Carmona et al. (2010, distance and E ( B − V ) ),Miroshnichenko et al. (1998, luminosity), (2) Esteban & Fernandez (1998), (3) Zickgraf (2001), (4) Miroshnichenko et al. (2002a), (5) Hynes et al. (2002,spectral type), (6) Miroshnichenko et al. (2002b), (7) Miroshnichenko et al. (2004), (8) Lobel et al. (2013, T eff and luminosity) a Age estimates derived by comparison with single star evolution models (Ekstr¨om et al. 2012), see Fig. 5 and discussion of individual stars for further details.
MWC 623 C O ( - ) C O ( - ) C O ( - ) C O ( - ) C O ( - ) C O ( - ) -12.7-12.6-12.5-12.4 l og F λ [ e r g s - c m - A o - ] EM* AS 381 C O ( - ) C O ( - ) C O ( - ) ? C O ( - ) C O ( - ) ? C O ( - ) -13.30-13.252.25 2.30 2.35 2.40 λ [ µ m] l og F λ [ e r g s - c m - A o - ] Figure 6.
Detail of the spectra of MWC 623 and AS 381 showing the COabsorption attributed to their cool companions.
We conducted a mini-survey of Galactic B[e] stars with the LBT-Luci I and Gemini-GNIRS to characterize their near-infrared K -band spectra. The most dominant emission line feature detected isthe Br γ line that is present in all our sample stars. In many cases,iron and magnesium emission can be identified, while helium andsodium emission lines are less common. In addition, the detectionand analysis of molecular lines like from carbon monoxide is a veryhelpful and powerful tool for the classification of the stars and theircircumstellar material, or for the confirmation of a cool companionrespectively.Summarizing the results of our B[e] survey:(i) MWC 314 shows a spectrum rich in lines of hydrogen, in-cluding pronounced lines of the Pfund series. Several lines of eff /K) l og ( L / L ) MWC 623-a AS 381-a MWC 623-b AS 381-b Z A M S Figure 7.
HRD for MWC 623 and AS 381 and their cool companions; singlestar stellar evolution models for different initial masses by Ekstr¨om et al.(2012) accounting for the effects of stellar rotation. sodium, magnesium, and iron are present as well. Only a weak IRexcess is detected and CO emission is lacking at all. The K -bandspectrum of MWC 314 strongly resembles the ones of LBV starsand candidates, e.g. S Dor and LHA 120-S 127. Tracers for circum-stellar material indicate the presence of a non-continuous gas disk,i.e. rings, around MWC 314. We determine a spectral type of B2and an age of about 6 Myr; with a lower-limit initial mass of 40 M ⊙ it’s the most luminous and most massive star in our sample.(ii) We detect CO bands in emission in the spectrum ofMWC 137 indicating an evolved nature of the star. However, theclassification of MWC 137 as Galactic supergiant B[e] star has tobe confirmed with more data.(iii) MWC 84 shows prominent He I line emissions. After anoutburst event in 1998, circumstellar material was clearly and sta-ble detected over decades, e.g. Pfund lines and CO emission. Weobserved the star at different epochs with LBT-Luci I and Gemini- c (cid:13) , 1–10 -band mini-survey of Galactic B[e] stars GNIRS and no signs of a recent prominent eruption was found dur-ing 2010 and 2013. Moreover, the observations reveal the disap-pearance of the circumstellar Pfund lines and the CO emission.(iv) MWC 300 is a B[e] supergiant candidate in a binary system.Its spectrum does not show CO emission or absorption at the timeof observation.(v) CO absorption bands are found in the spectra of MWC 623and AS 381. Both stars have been suspected to be binary sys-tems previously. Attributing the observed CO absorption to coolcompanions, we find a spectral classification of B4 II + K4 I-II(MWC 623) and B1 + K0 I-II (AS 381) and derive the fundamentalstellar parameters of the companions.All sample stars are slightly evolved (ages of 6 to 50 Myr)with progenitor masses in the intermediate to high-mass range (7to 40 M ⊙ ). In the cases of binary stars and candidates, episodesof mass transfer might have to be considered in the evolution ofthe components, thus hampering the exact characterization of thesystems. Future observations, for example radial velocity varia-tions, pronounced variability, and higher spectral resolution mightbe needed to ultimately confirm the binary companions.Particularly, MWC 84 and MWC 300 might be considered asbinary supergiant B[e] candidates in a quiescent transition phase,showing very low densities in their circumstellar envelopes to sus-tain molecular lines. Unfortunately, some supergiant B[e] stars can-not be distinguished as unambiguously as expected (Oksala et al.2013) and further observation are needed to clarify the evolution-ary status of these stars. ACKNOWLEDGMENTS
The authors thank Jochen Heidt for useful discussions about plan-ning the observations and valuable help in preparing the observ-ing scripts; Steve Allenson and the LBTO team for the supportduring the observing runs at the LBT. We thank our referee A.Miroshnichenko for valuable comments that helped to improve thismanuscript.AL and AK receive(d) financial support from the Max-Planck-Institut f¨ur Radioastronomie. MK acknowledges financial supportfrom GA ˇCR under grant number 14-21373S. The Astronomical In-stitute Ondˇrejov is supported by the project RVO:67985815. MK,MLA, and LSC acknowledge financial support for InternationalCooperation of the Czech Republic (MˇSMT, 7AMB14AR017)and Argentina (Mincyt-Meys ARC/13/12 and CONICET 14/003).MLA and LSC acknowledge financial support from the Agen-cia de Promoci´on Cient´ıfica y Tecnol´ogica (Pr´estamo BID, PICT2011/0885), from CONICET (PIP 0300), and the Programa de In-centivos G11/109 of the Universidad Nacional de La Plata, Ar-gentina.This publication makes use of data products from the TwoMicron All Sky Survey, which is a joint project of the Univer-sity of Massachusetts and the Infrared Processing and AnalysisCenter/California Institute of Technology, funded by the NationalAeronautics and Space Administration and the National ScienceFoundation.
REFERENCES
Appenzeller I., 1977, A&A, 61, 21Aret A., Kraus M., Muratore M. F., Borges Fernandes M., 2012,MNRAS, 423, 284 Bailey J., 1998, MNRAS, 301, 161Baines D., Oudmaijer R. D., Porter J. M., Pozzo M., 2006, MN-RAS, 367, 737Barsukova E. A., Borisov N. V., Burenkov A. N., Goranskii V. P.,Klochkova V. G., Metlova N. V., 2006, Astronomy Reports, 50,664Bergner Y. K., Miroshnichenko A. S., Yudin R. V., Kuratov K. S.,Mukanov D. B., Shejkina T. A., 1995, A&AS, 112, 221Bik A., Kaper L., Waters L. B. F. M., 2006, A&A, 455, 561Carmona A., van den Ancker M. E., Audard M., Henning T., Se-tiawan J., Rodmann J., 2010, A&A, 517, A67Cidale L., Zorec J., Tringaniello L., 2001, A&A, 368, 160Clark J. S., Crowther P. A., Larionov V. M., Steele I. A., RitchieB. W., Arkharov A. A., 2009, A&A, 507, 1555Clark J. S., Larionov V. M., Arkharov A., 2005, A&A, 435, 239Clark J. S., Larionov V. M., Crowther P. A., Egan M. P., ArkharovA., 2003, A&A, 403, 653Clark J. S. et al., 2000, A&A, 356, 50Clark J. S., Steele I. A., Fender R. P., Coe M. J., 1999, A&A, 348,888Ekstr¨om S. et al., 2012, A&A, 537, A146Esteban C., Fernandez M., 1998, MNRAS, 298, 185Gonz´alez-Fern´andez C., Cabrera-Lavers A., Hammersley P. L.,Garz´on F., 2008, A&A, 479, 131Groh J. H. et al., 2009a, ApJ, 705, L25Groh J. H., Damineli A., Jablonski F., 2007, A&A, 465, 993Groh J. H., Hillier D. J., Damineli A., 2011, ApJ, 736, 46Groh J. H., Hillier D. J., Damineli A., Whitelock P. A., Marang F.,Rossi C., 2009b, ApJ, 698, 1698Groh J. H., Hillier D. J., Madura T. I., Weigelt G., 2012, MNRAS,423, 1623Hanson M. M., Conti P. S., Rieke M. J., 1996, ApJS, 107, 281Henize K. G., 1976, ApJS, 30, 491Hillenbrand L. A., Strom S. E., Vrba F. J., Keene J., 1992, ApJ,397, 613Hillier D. J., Davidson K., Ishibashi K., Gull T., 2001, ApJ, 553,837Hjellming R. M., Mioduszewski A. J., Ueda Y., Ishida M., InoueH., Dotani T., Lewin W. H. G., Greiner J., 1998, IAU Circ., 6872,1Humphreys R. M., Davidson K., 1994, PASP, 106, 1025Hynes R. I. et al., 2002, A&A, 392, 991Kraus M., 2009, A&A, 494, 253Kraus M., Kr¨ugel E., Thum C., Geballe T. R., 2000, A&A, 362,158Krtiˇcka J., Owocki S. P., Meynet G., 2011, A&A, 527, A84Kurf¨urst P., Feldmeier A., Krtiˇcka J., 2013, ArXiv e-printsKurucz R. L., 1993, VizieR Online Data Catalog, 6039, 0Lamers H. J. G. L. M., Zickgraf F.-J., de Winter D., Houziaux L.,Zorec J., 1998, A&A, 340, 117Lanning H. H., L´epine S., 2006, PASP, 118, 1639Levesque E. M., Massey P., Olsen K. A. G., Plez B., Josselin E.,Maeder A., Meynet G., 2005, ApJ, 628, 973Liermann A., Kraus M., Schnurr O., Fernandes M. B., 2010, MN-RAS, 408, L6Lobel A. et al., 2013, in Astronomical Society of the Pacific Con-ference Series, Vol. 465, Astronomical Society of the PacificConference Series, Drissen L., Rubert C., St-Louis N., MoffatA. F. J., eds., p. 358Maeder A., Meynet G., Ekstr¨om S., Hirschi R., Georgy C., 2008,in IAU Symposium, Vol. 250, IAU Symposium, pp. 3–16Makarov V. V., Kaplan G. H., 2005, AJ, 129, 2420 c (cid:13) , 1–10 A. Liermann et al.
Marston A. P., McCollum B., 2008, A&A, 477, 193McGregor P. J., Hyland A. R., Hillier D. J., 1988a, ApJ, 324, 1071McGregor P. J., Hyland A. R., Hillier D. J., 1988b, ApJ, 334, 639Merrill P. W., Burwell C. G., 1950, ApJ, 112, 72Miroshnichenko A. S., 1996, A&A, 312, 941Miroshnichenko A. S., 2006, in Astronomical Society of thePacific Conference Series, Vol. 355, Stars with the B[e] Phe-nomenon, Kraus M., Miroshnichenko A. S., eds., p. 13Miroshnichenko A. S., 2007, ApJ, 667, 497Miroshnichenko A. S. et al., 2002a, A&A, 383, 171Miroshnichenko A. S., Fremat Y., Houziaux L., Andrillat Y.,Chentsov E. L., Klochkova V. G., 1998, A&AS, 131, 469Miroshnichenko A. S., Klochkova V. G., Bjorkman K. S., PanchukV. E., 2002b, A&A, 390, 627Miroshnichenko A. S. et al., 2004, A&A, 417, 731Morris P. W., Eenens P. R. J., Hanson M. M., Conti P. S., BlumR. D., 1996, ApJ, 470, 597Muratorio G., Rossi C., Friedjung M., 2008, A&A, 487, 637Najarro F., Figer D. F., Hillier D. J., Geballe T. R., Kudritzki R. P.,2009, ApJ, 691, 1816Oksala M. E., Kraus M., Cidale L. S., Muratore M. F., BorgesFernandes M., 2013, A&A, 558, A17Polster J., Korˇc´akov´a D., Votruba V., ˇSkoda P., ˇSlechta M.,Kuˇcerov´a B., Kub´at J., 2012, A&A, 542, A57Robinson E. L., Ivans I. I., Welsh W. F., 2002, ApJ, 565, 1169Rossi C., Frasca A., Marilli E., Friedjung M., Muratorio G., 2011,in IAU Symposium, Vol. 272, IAU Symposium, Neiner C., WadeG., Meynet G., Peters G., eds., pp. 422–423Rupen M. P., Mioduszewski A. J., Hjellming R. M., 2003, in NewViews on Microquasars, Durouchoux P., Fuchs Y., Rodriguez J.,eds., p. 221Scoville N., Kleinmann S. G., Hall D. N. B., Ridgway S. T., 1983,ApJ, 275, 201Searle S. C., Prinja R. K., Massa D., Ryans R., 2008, A&A, 481,777Seifert W. et al., 2003, in Society of Photo-Optical Instrumenta-tion Engineers (SPIE) Conference Series, Vol. 4841, Society ofPhoto-Optical Instrumentation Engineers (SPIE) Conference Se-ries, Iye M., Moorwood A. F. M., eds., pp. 962–973Takami M., Bailey J., Chrysostomou A., 2003, A&A, 397, 675Testi L., Palla F., Prusti T., Natta A., Maltagliati S., 1997, A&A,320, 159Th´e P. S., de Winter D., Perez M. R., 1994, A&AS, 104, 315Thureau N. D. et al., 2009, MNRAS, 398, 1309van Genderen A. M., 2001, A&A, 366, 508Wallace L., Hinkle K., 1997, ApJS, 111, 445Wang Y. et al., 2012, A&A, 545, L10Wheelwright H. E., Oudmaijer R. D., Goodwin S. P., 2010, MN-RAS, 401, 1199Wisniewski J. P., Babler B. L., Bjorkman K. S., Kurchakov A. V.,Meade M. R., Miroshnichenko A. S., 2006, PASP, 118, 820Zickgraf F.-J., 1992, in Astronomical Society of the Pacific Con-ference Series, Vol. 22, Nonisotropic and Variable Outflows fromStars, Drissen L., Leitherer C., Nota A., eds., p. 75Zickgraf F.-J., 2001, A&A, 375, 122Zickgraf F.-J., Stahl O., 1989, A&A, 223, 165This paper has been typeset from a TEX/ L A TEX file prepared by theauthor. c (cid:13)000