HOT AND COOL: BRIDGINGGAPS IN MASSIVE- EVOLUTION ASP Conference Series, Vol. 425, c 2010 C. Leitherer, P. D. Bennett, P. W. Morris, and J. Th. van Loon, eds.

Infrared Interferometry of the Upper Hertzsprung-Russell Diagram

John D. Monnier University of Michigan, Astronomy Department, Ann Arbor, MI 48109, USA

Abstract. The new generation of optical and infrared interferometers have unprecedented angular resolution and sensitivity. I will review recent advances in studies of massive objects in the upper portion of the Hertzsprung-Russell dia- gram, including fundamental parameters of individual red and blue supergiants, first interferometric imaging of interacting binaries, and new views of dust and gas in strong stellar winds. We note that massive stellar evolution is still usually discussed in the context of single star evolution, despite mounting evidence that binarity is commonplace and likely lies at the root of many observed phenomena.

1. Introduction

Massive present many challenges to our understanding of stellar evolution. These challenges lie in three major areas: fundamental stellar parameters, mass transfer in interacting binaries, and mass-loss mechanisms. All three areas are still poorly understood, thus the theory of massive stellar evolution rests on a shaky foundation. One way to make progress is to study individual objects using our best techniques in order to elucidate the physical mechanisms at play. For this, infrared and optical interferometry is potentially powerful to measure and image stars and their surroundings with amazing angular resolution. Consider the following examples:

• A nearby hot star (10 R⊙) at 200 pc is only 0.2 milliarcseconds across. • A nearby cool (massive) star (diameter 5 AU) at 200 pc is 25 milliarcsec- onds across. • The 50 AU dust shell around a close-by Wolf-Rayet star at 1500 pc is 33 milliarcseconds. • The inner wind at 10 AU around Eta (2300 pc) is about 4 milliarc- seconds. You can see from this list that even the closest (and thereby prototypical) mas- sive stars present serious imaging challenges for single telescopes. The required angular resolution demands a 50+m-diameter telescope operating at the diffrac- tion limit—a prospect beyond even our most ambitious projects today. Fortu- nately, this level of angular resolution is routine for existing infrared interfer- ometers and I will be reporting a wide range of important work from today’s facilities in this review. 257 258 Monnier

I will end each section of this review with a provocative question or two suggested by recent interferometry work.

2. Interferometry

This review will not discuss the technical details of optical and infrared inter- ferometry. Please see review in Monnier (2003) for more information on this topic. Here, I did want to list and briefly describe the facilities that will be men- tioned in this review. In order of appearance:

1. COAST (Cambridge Optical Aperture Synthesis Telescope). This facil- ity was based in Cambridge, England, and was a pioneer in developing long-baseline interferometric techniques, most noted for the first aperture synthesis image (Baldwin et al. 1996). The array used four telescopes with baselines typically less than 50 m, and observed binaries and evolved stars. This facility is now closed, and the principals are working on the next generation facility: the Magdalena Ridge Optical Interferometer (MROI, Creech-Eakman et al. 2008).

2. IOTA (Infrared-Optical Telescope Array) was located at Mt. Hopkins, AZ, and was operated by the Smithsonian Astrophysical Observatory with ma- jor contributions from University of Massachusetts at Amherst (Traub et al. 2003). This system observed mainly in the near-infrared (H and K bands) and had a longest baseline of 38 m. First imaging with the three-telescope system was reported in 2004 (Monnier et al.). This facil- ity is now closed.

3. CHARA (Center for High Angular Resolution Astronomy) array is a six- telescope interferometer located on Mt. Wilson, CA. The array was built and is operated by Georgia State University (McAlister et al. 2005; ten Brummelaar et al. 2005). CHARA boasts the largest baselines in the world, maximum of 330 m baselines, and has instrumentation that work from visible wavelengths to K band.

4. ISI (Infrared Spatial Interferometer) is a mid-infrared, heterodyne inter- ferometer also located on Mt. Wilson, CA (Hale et al. 2000). ISI has three telescopes with commissioned baselines up to about 85 m.

5. NPOI (Navy Prototype Optical Interferometer) is located in Flagstaff, AZ, and is operated by the Naval Research Lab and the Naval Observatory (Armstrong et al. 1998) NPOI specializes in visible-wavelength observa- tions and has used six telescopes simultaneously for imaging.

6. SUSI (Sydney University Stellar Interferometer) is the only visible-light interferometer in the southern hemisphere (Davis et al. 1999). SUSI uses two telescopes at a time, with baselines up to ∼100 meters. IR Interferometry of the Upper HRD 259

7. VLTI is the Very Large Telescope interferometer (Haguenauer et al. 2008), located on Paranal, Chile, and is famously part of the European South- ern Observatory. The VLTI can be used with any of the four 8-m unit telescopes or any of three 1.8 m auxiliary telescopes. VLT currently has a near-infrared three-beam combiner with spectral capabilities (AMBER) and a mid-infrared two-beam combiner (MIDI). The longest baseline is usually around ∼120 m, although longer baselines are possible with the ATs. 8. PTI was the Palomar Testbed Interferometer (Colavita et al. 1999), located at Palomar Observatory, CA. PTI worked primarily at K band, observing a single baseline with any two of three telescopes in the array. The longest baseline was around 110 m. This facility closed in early 2009.

3. Red Supergiants

Pioneering observations with single telescopes using speckle, rotation-shearing, and aperture masking (e.g., Lynds et al. 1976; Roddier & Roddier 1983; Buscher et al. 1990) have found that the surfaces of red supergiants are mottled with hotspots which vary on timescale of many months (Tuthill et al. 1997). These changing hotspots are now ascribed to the motions large convective elements on the surface (Schwarzschild 1975; Freytag et al. 2002). Using near-simultaneous, multi-wavelength data, the COAST interferome- ter found that the hotspots on Betelguese were markedly more pronounced at wavelengths probing the cooler molecular layers (Young et al. 2000). This led to the idea that the spots were generated by opacity structures (“holes”) in upper molecular layers. While opacity variations in the upper layers may be crucial in the optical regime, more recent work has found unmistakable spot structures in the H band likely resulting from true temperature differences on the continuum photosphere (with IOTA: Haubois et al. 2006, with CHARA: Kiss et al., these proceedings). Another major advance has been the discovery of an optically-thick water layer around cool stars (miras and red supergiants). Observers using IOTA (Per- rin et al. 1999, 2004), Keck Masking (Tuthill et al. 1999b; Woodruff et al. 2009), and ISI (Weiner et al. 2000) have reported very large diameter variations in and out of the near-infrared molecular bands and also at 11.15 µm. Essentially, at wavelengths where the water vapor absorbs infrared light, the measured angu- lar diameter is much larger than in the stellar continuum. Modeling this effect reveals a thick layer of warm molecular gas at about 1.5 R⋆. This idea was first suggested by Tsuji (2000) based on ISO data and water line lists have been used to self-consistently explain the wavelength-dependent diameter data and some line profiles (Ohnaka 2004; Weiner et al. 2003). These major results have yet to be incorporated in modern atmsopheric models. Another interesting result was that the diameter of the extreme red super- giant VY CMa was measured using a combination of IOTA + Keck data, which allowed for the correction of the dust emission. VY CMa has a K-band diam- eter of 21 AU, making it the largest known star (Monnier et al. 2004b). This large size could be affected by the aforementioned water layer, and more work is needed to understand if the continuum size is smaller. 260 Monnier

Open Questions: • Is spot intensity on red supergiant photospheres related to a star’s propensity for peculiar mass ejections seen as detached dust shells (e.g., α Ori, α Sco; Danchi et al. 1994)? • Does an interferometric temperature scale for red supergiants (corrected for molecular blanketing) match up with the recent hotter scale of Levesque et al. (2005)?

4. Blue Supergiants

Not surprisingly, blue supergiants are much smaller in physical size than red su- pergiants; they are nearly impossible to resolve with even today’s most powerful interferometers. While there are few results to report, we note that the NPOI optical interferometer will soon debut longer baselines, and the CHARA inter- ferometer recently commissioned two new visible instruments (VEGA, Mourard et al. 2008, and PAVO, Ireland et al. 2008). At visible wavelengths, CHARA will have approximately 0.2 mas angular resolution sufficient to resolve a handful of blue supergiants. Here we report mainly one interesting result. The nearby blue supergiant Deneb was resolved by the CHARA interferometer by Aufdenberg et al. (2006). They found the object was slightly elongated, implying a rotation rate of 65% of breakup. This model suggests the pole is 10% hotter than the equator, imply- ing that the apparent and temperature is viewing-angle dependent. These issues lead to an open question: Open Question: • How important is rapid rotation to consider for Blue Supergiants (e.g., latitude-dependence of winds, true location on H-R dia- gram)?

5. Transitioning Supergiants from Red to Blue

While finding nearby examples of red or blue supergiants is difficult enough, we have to search even farther away to find the rare specimen of a transitioning object. While difficult to study, these objects are critical to understanding this short-lived phase of stellar evolution which may be associated with large mass- loss episodes. IRC +10420 is a so-called yellow , currently showing a spectrum of A star (Oudmaijer et al. 2006). This object used to be an F spectral type, having increased in temperature by 2200 K in the last 30 . The contracting central object is surrounded by a thick envelope seen in detail by Hubble and ground-based observatories (e.g., Humphreys et al. 1997). The VLT Interferometer was able to observe this object using a near- infrared spectrograph to isolate the Br-γ emission line (de Wit et al. 2008). They found the emission line region was more resolved than the continuum, and found some indirect evidence for asymmetry on the milli-arcsecond scale. IR Interferometry of the Upper HRD 261

Open Questions: • What happens to a star and its mass loss during a transition from red to blue supergiant? • Is there any relation to the onset of asymmetry seen when AGB stars become pre-planetary nebulae?

6. Masses of Massive Binaries

Precise masses are hard to measure for massive stars and the number of resolved binaries systems are relatively few. Most recently, there was a study of the σ Sco (B1III+B1IV) system (North et al. 2007a) and the nearest Wolf-Rayet γ Vel binary (WR+O, North et al. 2007b). VLTI observations of Gamma Vel by Millour et al. (2007) see evidence for the colliding wind region in their data. There have been reports from VLTI, SUSI, CHARA, NPOI, and IOTA. Measurements of masses in young stellar binaries offer additional constraints on star formation and pre-main sequence stellar models. There have been a number of exciting papers on the eccentric system θ Ori C (IOTA: Kraus et al. 2009; NPOI: Patience et al. 2008), which has implications regarding the source of the runaway BN object nearby (Tan 2008). Although initial papers fueled some controversy regarding the binary eccentricity and period, recent data from VLTI-AMBER appears to have resolved the issues and the orbital parameters might now be considered “well established” (Kraus et al. 2009). Open Questions: • What is the underlying binary fraction for young OB stars and how does this impact our interpretation of the H-R diagram and the importance of binarity? • Binarity is probably more the rule than the exception, so why does “single-star evolution” still dominate our discussions?

7. Imaging Interacting Binaries

As previously mentioned, current infrared interferometers lack the resolution to resolve most hot (blue) supergiants. However, there are a class of interacting (Algol-type) systems that can be resolved and even imaged for the first time. Spica (B1III+B2V) is a short-period binary with period of 4.01 days. It was originally observed by the intensity interferometer (Herbison-Evans et al. 1971). New data from the SUSI interferometer (visible light) and the CHARA interferometer (H and K bands) have been combined with a detailed spectrum to generate a new model of the system, incorporating gravity darkening of the stellar components (Aufdenberg et al. 2009). Another exciting result is the sequence of images of the famous eclipsing and interacting system Beta Lyrae (see Fig. 1). This edge-on system contains a B6-8 II donor star with a more massive B star gainer surrounded by a thick disk. Zhao et al. (2008) report imaging of this system using CHARA with MIRC combiner (Monnier et al. 2006), detecting the donor to be partially filling its Roche lobe 262 Monnier and detecting the elongated disk around the gainer. Additionally, new images in the H-α line have been made using NPOI (Schmitt et al. 2009).

Figure 1. This figure shows CHARA-MIRC imaging results for the inter- acting binary system β Lyrae from Zhao et al. (2008). The is shown along with the image reconstructions for five epochs. The high resolution of CHARA (∼0.5 milliarcseconds) allows the components to be cleanly resolved for the first time and for the first-ever astrometric orbit determination.

Open Question: • How is mass transferred in interacting systems and how does this affect stellar evolution?

8. Circumstellar Material

We can learn about stellar evolution through investigation of circumstellar ma- terial, which act as a kind of fossil of previous important mass-loss episodes. The material also can tell use about underlying geometry, for instance when viewing a disk or torus of dust. This latter property was critical for new VLTI-MIDI measurements of the dusty extragalactic RSG WOH G64 in the Large Magel- 6 lanic Cloud, which had previous luminosity estimates higher than 5 × 10 L⊙ in violation of the Humphreys-Davidson Limit. Ohnaka et al. (2008) resolved this dust shell and created a self-consistent model of the interferometry and spec- tral energy distribution that allowed a more accurate estimate of the luminosity, bringing it more into line with expectations. In addition, there has been much work on B[e] stars. This heterogeneous group of objects often involve a massive B star with significant infrared excess of unknown origin. Some are associated with hard x-ray bursts suggesting a neutron star or black hole companion. Until now, the circumstellar envelopes have not been resolved spatially. Thureau et al. (2009) resolved the dust shell around CI Cam using IOTA and PTI interferometers, finding it to be very elongated in a ring-like structure. IR Interferometry of the Upper HRD 263

In addition, Domiciano de Souza et al. (2007) found evidence that CPD −57 2874 also has an asymmetric shell. These asymmetries point to a binary origin for the dust, although detailed scenarios for producing dust in this system are still sketchy. Open Question: • How do B[e] supergiants fit into our picture of massive stellar evolution and what is origin of the observed dust shells?

9. Dusty Wolf-Rayet Stars

Long ago, Usov (1991) and Williams et al. (1990) uncovered the mechanism for how colliding winds in Wolf-Rayet + OB star binaries can produce dust. The colliding winds compress gas enough for dust production and the dust is sprayed out in the wind-wind interface. This dust can form a beautiful spiral because of the motion of the underlying binary (e.g., Tuthill et al. 1999a). In fact, all dusty Wolf-Rayets that have been well-resolved show spiral struc- ture (Monnier et al. 2007) associated with binarity and many others show non- thermal radio emission (Dougherty & Williams 2000), leading to the idea that colliding winds are a necessary ingredient for forming dusty around these hot stars. Monnier et al. (1999) pointed out that since most late WC stars show dust we can then surmise that most WC stars are binaries. Recent VLTI measure- ments (Rajagopal et al. 2007) lack the UV coverage to detect spiral structure, but confirm the overall size scale first measured by single telescopes. Tuthill et al. (2006) recently uncovered more spirals of dust in the dusty WCs amongst the Quintuplet cluster. Open Question: • How can our current picture of massive stellar evolution explain the high binary fraction for late-type WC Wolf-Rayets? • Could it be that Wolf-Rayets are indeed formed via binary in- teractions after all, and not primarily due to wind instabilities?

10. Eta Carina

Eta Carina is typically in a category by itself. Considering entire conferences are held on this object, my review here only points the reader to the tour de force effort of Weigelt et al. (2007). Following previous work by VLTI (van Boekel et al. 2003; Kervella 2007), this paper constructs an impressive dust and wind model to explain a wide range of observations, including near-infrared interferometer measurements in and out of multiple spectral lines. This work confirms the central source to be elongated on the few AU scale at K band, consistent with increased mass loss at the hot pole of a rapid rotator, with no obvious sign of the putative “companion.” By resolving the IR emission lines as a function of wavelength across the line, an impressive confirmation of an optically-thick wind model (within a region of 10 × 10 AU at 2300 pc) was presented. 264 Monnier

I highly recommend this paper as a model of what can be learned by combin- ing interferometry and modelling together, even for complicated and challenging targets. Open Question: • Can one see changes in the inner environment of Eta Car follow- ing the recent periastron passage of companion, perhaps to even detect the companion directly?

11. Conclusion

The recent advances in sensitivity with optical and infrared interferometry allow a range of phenomena to be investigated with unprecedented angular resolution. This review has meant to show concrete results across the entire upper part of the HR diagram, impacting many of the most uncertain areas of stellar evolution: binarity, rotation, mass loss. Please contact me (monnier (at) umich.edu) if you want to discuss the feasiblity of any observations using today’s facilities. I will be happy to encourage you! Acknowledgments. I wish to acknowledge all the fine papers that I have missed or overlooked in preparing this short review. JDM acknowledges support from the National Science Foundation (NSF AST-0352723).

References

Armstrong, J. T., et al. 1998, ApJ, 496, 550 Aufdenberg, J. P., et al. 2009, American Astronomical Society Meeting Abstracts, 213, #410.19 Aufdenberg, J. P., Morrison, N. D., Hauschildt, P. H., & Adelman, S. J. 2006, Astro- physics in the Far Ultraviolet: Five Years of Discovery with FUSE, 348, 124 Baldwin, J. E., and 15 colleagues 1996, A&A306, L13 ten Brummelaar, T. A., et al. 2005, ApJ, 628, 453 Buscher, D. F., Baldwin, J. E., Warner, P. J., & Haniff, C. A. 1990, MNRAS245, 7P Colavita, M. M., et al. 1999, ApJ, 510, 505 Creech-Eakman, M. J., and 33 colleagues 2008, SPIE 7013 Danchi, W. C., Bester, M., Degiacomi, C. G., Greenhill, L. J., & Townes, C. H. 1994, AJ, 107, 1469 Davis, J., Tango, W. J., Booth, A. J., ten Brummelaar, T. A., Minard, R. A., &Owens, S. M. 1999, MNRAS, 303, 773 Domiciano de Souza, A., et al. 2007, A&A, 464, 81 de Wit, W. J., Oudmaijer, R. D., Groenewegen, M. A. T., Hoare, M. G., & Malbet, F. 2008, A&A, 480, 149 Dougherty, S. M., & Williams, P. M. 2000, MNRAS, 319, 1005 Freytag, B., Steffen, M., Dorch, B. 2002, Astro. Nach. 323, 213 Hale, D. D. S., et al. 2000, ApJ, 537, 998 Haguenauer, P., et al. 2008, SPIE, 7013 Haubois, X., Perrin, G., Lacour, S., Schuller, P. A., Monnier, J. D., Berger, J.-P., Ridgway, S. T., Millan-Gabet, R., Pedretti, E., & Traub, W. A. 2006, SF2A 471 Herbison-Evans, D., Hanbury Brown, R., Davis, J., & Allen, L. R. 1971, MNRAS, 151, 161 Humphreys, R. M., et al. 1997, AJ, 114, 2778 IR Interferometry of the Upper HRD 265

Ireland, M. J., et al. 2008, SPIE, 7013 Kervella, P. 2007, A&A, 464, 1045 Kraus, S., et al. 2007, A&A, 466, 649 Kraus, S., et al. 2009, arXiv:0902.0365 Levesque, E. M., Massey, P., Olsen, K. A. G., Plez, B., Josselin, E., Maeder, A., Meynet, G. 2005, ApJ, 628, 973 Lynds, C. R., Worden, S. P., & Harvey, J. W. 1976, ApJ, 207, 174 McAlister, H. A., et al. 2005, ApJ, 628, 439 Millour, F., et al. 2007, A&A, 464, 107 Monnier, J. D. 2003, Reports on Progress in Physics 66, 789 Monnier, J. D., and 28 colleagues 2004a, ApJ, 602, L57 Monnier, J. D., et al. 2004b, ApJ, 605, 436 Monnier, J. D., et al. 2006, SPIE, 6268 Monnier, J. D., Tuthill, P. G., & Danchi, W. C. 1999, ApJ, 525, L97 Monnier, J. D., Tuthill, P. G., Danchi, W. C., Murphy, N., & Harries, T. J. 2007, ApJ, 655, 1033 Mourard, D., et al. 2008, SPIE, 7013 North, J. R., Davis, J., Tuthill, P. G., Tango, W. J., & Robertson, J. G. 2007a, MNRAS, 380, 1276 North, J. R., Tuthill, P. G., Tango, W. J., & Davis, J. 2007b, MNRAS, 377, 415 Oudmaijer, R. D., Davies, B., Dawson, F., Lockett, O., Mottram, J. C., Patel, M., & Groenewegen, M. A. T. 2006, Stars with the B[e] Phenomenon, 355, 181 Ohnaka, K. 2004, A&A, 424, 1011 Ohnaka, K., Driebe, T., Hofmann, K.-H., Weigelt, G., & Wittkowski, M. 2008, A&A, 484, 371 Patience, J., Zavala, R. T., Prato, L., Franz, O., Wasserman, L., Tycner, C., Hutter, D. J., & Hummel, C. A. 2008, ApJ, 674, L97 Perrin, G., Coud´e du Foresto, V., Ridgway, S. T., Mennesson, B., Ruilier, C., Mariotti, J.-M., Traub, W. A., & Lacasse, M. G. 1999, A&A, 345, 221 Perrin, G., Ridgway, S. T., Coud´e du Foresto, V., Mennesson, B., Traub, W. A., & Lacasse, M. G. 2004, A&A, 418, 675 Rajagopal, J., et al. 2007, ApJ, 671, 2017 Roddier, C., & Roddier, F. 1983, ApJ, 270, L23 Schmitt, H. R., et al. 2009, ApJ, 691, 984 Schwarzschild, M. 1975, ApJ, 195, 137 Tan, J. C. 2008, arXiv:0807.3771 Thureau, N. D., Monnier, J. D., Traub, W. A., Millan-Gabet, R., Pedretti, E., Berger, J.-P., Garcia, M. R., Schloerb, F. P., & Tannirkulam, A.-K. 2009, MNRAS, 398, 1309 Traub, W. A., and 16 colleagues 2003, SPIE 4838, 45 Tsuji, T. 2000, ApJ, 538, 801 Tuthill, P. G., Haniff, C. A., & Baldwin, J. E. 1997, MNRAS, 285, 529 Tuthill, P. G., Monnier, J. D., & Danchi, W. C. 1999a, Nat, 398, 487 Tuthill, P. G., Monnier, J. D., & Danchi, W. C. 1999b, Working on the Fringe: Optical and IR Interferometry from Ground and Space, 194, 188 Tuthill, P., Monnier, J., Tanner, A., Figer, D., Ghez, A., & Danchi, W. 2006, Science, 313, 935 Usov, V. V. 1991, MNRAS, 252, 49 van Boekel, R., et al. 2003, A&A, 410, L37 Weigelt, G., et al. 2007, A&A, 464, 87 Weiner, J., Danchi, W. C., Hale, D. D. S., McMahon, J., Townes, C. H., Monnier, J. D., & Tuthill, P. G. 2000, ApJ, 544, 1097 Weiner, J., Hale, D. D. S., & Townes, C. H. 2003, ApJ, 589, 976 Williams, P. M., van der Hucht, K. A., Pollock, A. M. T., Florkowski, D. R., van der Woerd, H., & Wamsteker, W. M. 1990, MNRAS, 243, 662 266 Monnier

Woodruff, H. C., et al. 2009, ApJ, 691, 1328 Young, J. S., and 11 colleagues 2000, MNRAS, 315, 635 Zhao, M., et al. 2008, ApJ, 684, L95