1: \documentclass[11pt,twoside]{article}
2:
3: %%% PREAMBLE MATTER
4:
5: \usepackage{asp2006}
6: \usepackage{epsf}
7: \usepackage{psfig}
8: \usepackage{lscape}
9:
10: \markboth{}{} %%% Fill in authors' names and short running title
11:
12: \pagestyle{myheadings}
13: \setcounter{equation}{0}
14: \setcounter{figure}{0}
15: \setcounter{footnote}{0}
16: \setcounter{section}{0}
17: \setcounter{table}{0}
18:
19: %%% MAIN PART OF DOCUMENT
20:
21: \begin{document}
22: \title{Were All Massive Stars Born in OB Associations and Clusters?}
23: \author{You-Hua Chu and Robert A. Gruendl}
24: \affil{Astronomy Department, University of Illinois,
25: 1002 West Green Street, Urbana, Illinois 61801, USA}
26:
27:
28: \begin{abstract} %%% Abstract to run on from here.
29: It has been commonly conjectured that all massive ($>$ 10 $M_\odot$)
30: stars are born in OB associations or clusters. Many O and B stars in
31: the Galaxy or the Magellanic Clouds appear to exist in isolation,
32: however. While some of these field OB stars have been ejected from
33: their birthplaces, some are too far away from massive star forming
34: regions to be runaways. Can massive stars form in isolation? The
35: {\it Spitzer} survey of the Large Magellanic Cloud (aka SAGE) provides
36: a unique opportunity for us to investigate and characterize the formation
37: sites of massive stars for an entire galaxy. We have identified all
38: massive young stellar objects (YSOs) in the Large Magellanic Cloud.
39: We find that $\sim$85\% of the massive YSOs are in giant molecular
40: clouds and $\sim$65\% are in OB associations. Only $\sim$7\% of the
41: massive YSOs are neither in OB associations nor in giant molecular clouds.
42: This fraction of isolated massive stars in the Large Magellanic Cloud
43: is comparable to the 5--10\% found in the Galaxy.
44: \end{abstract}
45:
46: %%% MAIN BODY OF TEXT GOES HERE. CONSULT "INSTRUCTIONS FOR AUTHORS USING
47: %%% LATEX2E MARKUP", SECTIONS 2.3-2.6 FOR HELP WITH EQUATIONS, FIGURES,
48: %%% AND TABLES.
49:
50: \section{Introduction}
51: Inventories of massive O-type stars in the Galaxy show that
52: $\sim$70\% of them are associated with OB associations or clusters,
53: and that more than 1/3 of the remaining 30\% are runaway OB stars
54: \citep{Gies87,Metal98,Metal04}. These results imply that less
55: than 20\% of Galactic O-type stars were formed in isolation.
56:
57: Recently, the origin of 43 field stars from the \citet{Metal98}
58: sample of 227 O stars with $V <$ 8 has been investigated: deep
59: images show that 5 of them are in small clusters \citep{witetal04},
60: and analysis of their location and space velocities suggests
61: that 22 are likely runaways \citep{witetal05}.
62: Only 10--20 of the sample of 227 bright Galactic O stars cannot
63: be assigned to any OB associations or clusters, indicating that
64: 5--10\% of the Galactic O stars are truly isolated (Zinnecker
65: \& Yorke 2007).
66:
67: The formation of isolated massive stars is of great interest.
68: While studies of Galactic O stars have yielded important results
69: on the fraction of O stars formed in isolation, it is not known
70: under what interstellar conditions isolated massive stars are formed
71: or why field stars have steeper initial mass functions (IMFs) than
72: OB associations \citep{Metal95}.
73: To answer these questions, we need to examine young massive stars
74: before their energy feedback has significantly altered the ambient
75: interstellar conditions and dispersed the natal clouds.
76: It is also necessary to inventory young massive stars in an entire
77: galaxy to gain statistical insights.
78: Recent {\it Spitzer Space Telescope} observations of the Large
79: Magellanic Cloud (LMC) provide an ideal dataset to study young
80: massive stars.
81:
82: \section{Spitzer Sample of Massive YSOs in the LMC}
83:
84: The LMC is at a distance of 50 kpc (distance modulus = 18.5),
85: where 1$''$ corresponds to 0.25 pc. Its nearly face-on
86: orientation allows a clear view of the entire galaxy with
87: little confusion and extinction along the line-of-sight.
88: {\it Spitzer} observations are able to resolve individual stars
89: in the mid-IR for the first time \citep{Jetal05,Cetal05}.
90: We have used our deep IRAC and MIPS observations of seven
91: LMC H\,II complexes from {\it Spitzer} Cycle 1 and the SAGE
92: survey of the entire LMC from Cycle 2 \citep{Metal06}
93: to identify massive young stellar objects (YSOs) in the LMC.
94:
95: We have retrieved the above mentioned {\it Spitzer} observations
96: of the LMC, and carried out photometry for all point sources in
97: the four IRAC bands and three MIPS bands. The MIPS 70 and 160 $\mu$m
98: data are less useful because of their poor angular resolution.
99: The SAGE survey of the LMC contains observations made in two epochs.
100: Photometric measurements were made for each epoch separately,
101: compared to reject transients and spurious sources, and combined
102: to improve signal-to-noise ratios.
103: To check our method of photometric measurements, we have retrieved
104: the {\it Spitzer} Wide-area Infrared Extragalactic (SWIRE) Legacy
105: Survey \citep{Letal04} data and made our measurements;
106: we find general consistency between our and their results.
107:
108: \begin{figure}
109: \centerline{
110: \psfig{file=fig1.eps,height=2.5in,angle=0}
111: }
112: %\vspace*{-0.25cm}
113: \caption{Color-magnitude diagrams of all point sources in the
114: LMC (left panel) and the SWIRE survey (right panel). The massive
115: YSO candidates in the LMC are located in the upper right corner
116: of the CMD; the AGB/post-AGB stars are marked by ``+'' and the
117: YSOs by filled circles.
118: }
119: \end{figure}
120:
121: The photometric measurements in IRAC bands and MIPS 24 $\mu$m band
122: were used to assemble various diagnostic color-color and
123: color-magnitude diagrams (CCDs and CMDs). Among these, the most
124: informative is the [8.0] vs [4.5]$-$[8.0] CMD.
125: In Figure 1 we present such CMDs for all point sources in the
126: LMC (left panel) and in the SWIRE survey (right panel)
127: for comparison.
128: The SWIRE CMD shows a vertical branch of stars of which the
129: upper part bends to the right owing to the artifact that the
130: brightest stars are saturated in the 4.5 $\mu$m band, as well
131: as a concentration of galaxies.
132: The CMD of the LMC is much more complex, as it contains normal
133: stars with zero color, evolved stars with IR excess, YSOs with
134: IR excess, and background galaxies.
135:
136: To select YSO candidates in the LMC, we first adopt the galaxy
137: discriminator suggested by \citet{Hetal06} -- the tilted dashed
138: line in Figure 1. As shown in the SWIRE CMD, this discriminator
139: is indeed an effective upper boundary for most background galaxies.
140: To exclude normal stars and evolved stars, we adopt a color cutoff
141: of [4.5]$-$[8.0] $\ge$ 2.0.
142: In the LMC CMD, we have overplotted the expected locations of
143: 3000 $L_\odot$ asymptotic giant branch (AGB) and post-AGB stars
144: from \citet{Gr06}; it is evident that our discriminator does
145: exclude most of the normal and AGB/post-AGB stars.
146: Finally, we adopt the [8.0] $\le$ 8 criterion to select
147: massive YSOs with masses greater than $\sim$10 $M_\odot$
148: \citep{Cetal05}.
149: As show in Figure 1, the massive YSO candidates are in the upper
150: right corner of the CMD; however, some AGB and post-AGB stars
151: are still expected to occupy this region.
152:
153: \begin{figure}
154: \centerline{
155: \psfig{file=fig2.eps,width=5.15in,angle=0}
156: }
157: \caption{Determining the nature of the YSO candidate 044927.3$-$691206.3.
158: Images at multiple wavelengths are displayed simultaneously using ds9.
159: The left-most panel in the top row displays an H$\alpha$ image with
160: NANTEN CO contours to show the large-scale environment; the other
161: panels are matched close-up images from the Digitized Sky Survey red
162: (dss-r), MCELS H$\alpha$, 2MASS $J$ and $K$, IRAC 3.6, 4.5, 5.8, and 8.0
163: $\mu$m, and MIPS 24 and 70 $\mu$m bands. Plots of the YSO candidate's
164: SED and location in the [8.0] vs [4.5]$-$[8.0] CMD are displayed to the
165: right. These images and plots are examined simultaneously to assess
166: the nature of the object.
167: }
168: \end{figure}
169:
170: To determine the true nature of the massive YSO candidates, it
171: is necessary to examine their images and spectral energy
172: distributions (SEDs) carefully.
173: Our approach is illustrated in Figure 2, where ds9 is used to
174: display a large-scale H$\alpha$ image overplotted with CO
175: contours, and close-up images in Digitized Sky Survey red (dss2r),
176: H$\alpha$, 2MASS $J$ and $K$, IRAC 3.6, 4.5, 5.8, and 8.0 $\mu$m,
177: and MIPS 24 and 70 $\mu$m, alongside plots showing the YSO
178: candidate's SED and location in the [8.0] vs [4.5]$-$[8.0] CMD.
179: The H$\alpha$ images are from the Magellanic Cloud Emission Line
180: Survey (MCELS, Smith et al.\ 1999), and the CO contours are from
181: the NANTEN survey \citep{Fetal99,Fetal01}.
182: By examining these images and plots, we can easily identify
183: foreground stars, AGB/post-AGB stars, and bright galaxies.
184: After eliminating these contaminating sources, we are left with
185: 234 probable (high-confidence) and 14 possible ($\sim$50\% probability)
186: massive YSOs. Details of the {\it Spitzer} photometric measurements
187: and selection of massive YSOs will be reported by Gruendl \& Chu
188: (2008, in preparation).
189:
190: Massive stars are often born in complex environments and the
191: {\it Spitzer} resolution is inadequate to show the detail.
192: The 2MASS images are also of limited usefulness. We have
193: thus used the Infrared SidePort Imager (ISPI) on the 4m
194: Blanco telescope at Cerro Tololo Inter-American Observatory
195: to obtain deeper $J$ and $K$ images with higher angular resolution.
196: The necessity of such high-resolution images is illustrated in
197: Figure 3, where close-up images of the object shown in Figure 2
198: are displayed. The near-IR counterpart of this source appear
199: unresolved in the 2MASS $J$ and $K$ images, but is resolved into
200: a diffuse source and two point sources in the ISPI $J$ and $K$
201: images, with one source being significantly brighter in $K$ than
202: in $J$. This example fully demonstrates the danger in blindly
203: modeling the SED of a YSO in the LMC for its physical properties
204: without using high-resolution images to examine the multiplicity
205: of the YSO.
206:
207:
208: \begin{figure}
209: \centerline{
210: \psfig{file=fig3.eps,width=4.5in,angle=0}
211: }
212: %\vspace*{-0.25cm}
213: \caption{Near-IR images of the massive YSO candidate
214: 044927.3$-$691206.3 in the LMC.
215: The YSO is located at the center of the field-of-view.
216: The two left panels display the 2MASS $J$ and $K$ images,
217: and the middle two panels ISPI $J$ and $K$ images.
218: The central region is enlarged and shown in the two
219: right panels. The diffuse emission and the two point sources
220: resolved by the ISPI images appear as a point source in
221: 2MASS and {\it Spitzer} IRAC images in Figure 2.
222: }
223: \end{figure}
224:
225: \section{Environments of Massive YSOs}
226:
227: \begin{figure}
228: \centerline{
229: \psfig{file=fig4a.eps,width=2.4in,angle=0}
230: \psfig{file=fig4b.eps,width=2.4in,angle=0} }
231: \centerline{
232: \psfig{file=fig4c.eps,width=2.4in,angle=0}
233: \psfig{file=fig4d.eps,width=2.4in,angle=0}
234: }
235: %\vspace*{-0.25cm}
236: \caption{Massive YSOs in or near OB associations surrounded
237: by superbubbles. The H\,II complexes in the top row are
238: N44 (left) and N206 (right), and the bottom row N144 (left)
239: and N51 (right). The YSOs are plotted as filled circles
240: in these MCELS H$\alpha$ images overplotted with CO contours
241: from the NANTEN survey.
242: }
243: \end{figure}
244:
245: On a global scale, massive star formation is associated with
246: gravitational instability.
247: The gravitational instability of the LMC has been analyzed
248: by \citet{Yetal07} considering the gas disk only and considering
249: both the collisional gas and the collisionless stars in the disk.
250: They find that only 62\% of the {\it Spitzer} sample of
251: massive YSOs fall in gravitationally unstable regions for
252: the former, and 85\% for the latter.
253: The most visible difference between these two cases is along the
254: star-forming stellar bar of the LMC, which is largely stable for
255: the gas-only disk and unstable in the gas+star disk.
256: It is thus important that the contribution of stars is included in
257: the calculation of gravitational instability.
258:
259: For the 234 probable and 14 possible massive YSOs in the LMC,
260: we mark the YSO positions in H$\alpha$ images overplotted with
261: CO contours to examine their stellar and interstellar environments.
262: We compare the YSO location with the OB associations compiled by
263: \citet{LH70}, H\,II regions cataloged by \citet{H56} and \citet{DEM},
264: and giant molecular clouds identified from NANTEN survey by
265: \citet{Fetal99}. The results are compiled in a table for further
266: statistical analysis.
267:
268: Among the 248 probable and possible massive YSOs in the LMC,
269: $\sim$65\% are in OB associations, most of which are surrounded
270: by superbubbles or bright
271: H\,II regions. Figure 4 shows examples of massive YSOs in or near
272: OB associations in superbubbles. It is evident that star formation
273: continues into the molecular clouds near OB associations, and
274: most likely the expansion of the superbubble plays a significant
275: role in the propagation of star formation. Low-level star formation
276: exists within the superbubble where no molecular material is detected
277: by the NANTEN survey; these are probably pc-sized dust globules
278: compressed by the photoionized surface layer (i.e., photo-implosion
279: or globule-squeezing), as illustrated by the YSOs in the superbubble
280: N51D \citep{Cetal05}.
281:
282: About 85\% of the massive YSOs in the LMC are formed in giant
283: molecular clouds.
284: Interestingly, some massive YSOs are located within giant molecular
285: clouds but are also on the peripheries of supernova remnants (SNRs).
286: Figure 5 shows two examples. While it is possible that the expansion
287: of the SNR triggered the star formation, it is also possible that
288: the star formation was triggered by the bubble blown by the
289: supernova progenitor's fast wind during the main sequence phase and
290: the subsequent SNR is largely confined within the bubble.
291: To distinguish between these two possibilities, velocities of the
292: SNR shell and the superposed molecular cloud need to be compared.
293: Bubble expansion velocities are expected to be small, $\le$20 km~s$^{-1}$,
294: while SNR shocks are expected to be $>$100 km~s$^{-1}$.
295: An absence of high velocities in the molecular material would suggest
296: that the expanding wind-blown bubble was responsible for triggering
297: the star formation.
298:
299:
300: \begin{figure}
301: \centerline{
302: \psfig{file=fig5a.eps,width=2.5in,angle=0}
303: \psfig{file=fig5b.eps,width=2.5in,angle=0} }
304: \caption{Massive YSOs associated with SNR 0513$-$69.2 (left)
305: and SNR DEM\,L256 (right). The YSOs are plotted as filled
306: circles in the MCELS H$\alpha$ images overplotted with CO contours
307: from the NANTEN survey. The SNR is the small ($\sim$3$'$ across)
308: shell near the field center.
309: }
310: \end{figure}
311:
312: \begin{figure}
313: \centerline{
314: \psfig{file=fig6.eps,width=5in,angle=0}
315: }
316: %\vspace*{-0.25cm}
317: \caption{Massive YSOs that are not in OB associations or
318: crowded environments. The massive YSOs are plotted as filled
319: circles in MCELS H$\alpha$ images overplotted with CO
320: contours from the NANTEN survey. The field-of-view of each
321: panel is 15$'$ $\times$ 15$'$, or 225 pc $\times$ 225 pc.
322: }
323: \end{figure}
324:
325: %\section{Isolated Massive YSOs}
326:
327: The massive YSOs that are not in OB associations may be
328: associated with giant molecular clouds, as shown in the
329: top row of Figure 6.
330: It is very likely that the stellar energy feedback of these
331: first-generation massive stars will trigger further star formation,
332: resulting in an OB association or a cluster.
333: Only massive YSOs formed in an environment without a large
334: reservoir of molecular gas may become truly isolated massive stars,
335: for example, the YSOs shown in the bottom row of Figure 6.
336: Among the massive YSOs in the LMC, only $\sim$7\% are neither
337: in OB associations nor associated with giant molcular clouds.
338: These will evolve into isolated massive stars.
339: This fraction is consistent with the finding for Galactic
340: O-type stars -- only 5-10\% are truly isolated.
341: It should be noted that the isolated massive YSOs are in general
342: less luminous, and hence less massive, than those in active
343: star forming regions.
344: Detailed comparisons of YSO properties for different environments
345: are still underway.
346:
347:
348: \acknowledgements %%% Text of acknowledgments runs on after this command.
349: This research is supported by NASA grants JPL 1264494 and 1290956.
350:
351: %%% THE BIBLIOGRAPHY
352: %%%
353: %%% CONSULT SECTION 3 OF "INSTRUCTIONS FOR AUTHORS" FOR HOW TO USE NATBIB.
354: %%% AUTHORS ARE ENCOURAGED TO USE EITHER THE "THEBIBLIOGRAPY" ENVIRONMENT
355: %%% BY UNCOMMENTING (DELETING THE "%" SYMBOL) THE COMMANDS BELOW, OR BY
356: %%% USING THE BIBTEX ENVIRONMENT. TO FIND OUT WHICH IS APPLICABLE TO YOUR
357: %%% CONTRIBUTION, CONSULT THE VOLUME EDITORS FOR YOUR PROCEEDINGS.
358: %%%
359:
360: \begin{thebibliography}{}
361: \bibitem[Chu et al.(2005)]{Cetal05}
362: Chu, Y.-H., et al.\ 2005, \apjl, 634, L189
363: \bibitem[Davies et al.(1976)]{DEM}
364: Davies, R.~D., Elliott, K.~H., \& Meaburn, J.\ 1976, \memras, 81, 89
365: \bibitem[de Wit et al.(2004)]{witetal04}
366: de Wit, W.~J., Testi, L., Palla, F., Vanzi, L., \& Zinnecker,
367: H.\ 2004, A\&A, 425, 937
368: \bibitem[de Wit et al.(2005)]{witetal05}
369: de Wit, W.~J., Testi, L., Palla, F., \& Zinnecker, H.\ 2005, A\&A, 437, 247
370: \bibitem[Fukui et al.(1999)]{Fetal99}
371: Fukui, Y., et al.\ 1999, \pasj, 51, 745
372: \bibitem[Fukui et al.(2001)]{Fetal01}
373: Fukui, Y., Mizuno, N., Yamaguchi, R., Mizuno, A., \& Onishi, T.\ 2001,
374: \pasj, 53, L41
375: \bibitem[Gies(1987)]{Gies87} Gies, D.~R.\ 1987, ApJS, 64, 545
376: \bibitem[Groenewegen(2006)]{Gr06}
377: Groenewegen, M.~A.~T.\ 2006, \aap, 448, 181
378: \bibitem[Harvey et al.(2006)]{Hetal06}
379: Harvey, P.~M., et al.\ 2006, \apj, 644, 307
380: \bibitem[Henize(1956)]{H56}
381: Henize, K.~G.\ 1956, \apjs, 2, 315
382: \bibitem[Jones et al.(2005)]{Jetal05}
383: Jones, T.~J., Woodward, C.~E., Boyer, M.~L., Gehrz, R.~D.,
384: \& Polomski, E.\ 2005, \apj, 620, 731
385: \bibitem[Lonsdale et al.(2004)]{Letal04}
386: Lonsdale, C., et al.\ 2004, \apjs, 154, 54
387: \bibitem[Lucke \& Hodge(1970)]{LH70}
388: Lucke, P.~B., \& Hodge, P.~W.\ 1970, \aj, 75, 171
389: \bibitem[Ma{\'{\i}}z-Apell{\'a}niz et al.(2004)]{Metal04}
390: Ma{\'{\i}}z-Apell{\'a}niz, J., Walborn, N.~R., Galu{\'e}, H.~{\'A}., \&
391: Wei, L.~H.\ 2004, ApJS, 151, 103
392: \bibitem[Mason et al.(1998)]{Metal98}
393: Mason, B.~D., et al.\ 1998, AJ, 115, 821
394: \bibitem[Massey et al.(1995)]{Metal95}
395: Massey, P., Lang, C.~C., Degioia-Eastwood, K., \& Garmany, C.~D.\ 1995,
396: ApJ, 438, 188
397: \bibitem[Meixner et al.(2006)]{Metal06}
398: Meixner, M., et al.\ 2006, AJ, 132, 2268
399: \bibitem[Smith et al.(1999)]{Setal99}
400: Smith, C.\ R., et al.\ 1999, in IAU Symposium 190, 28
401: \bibitem[Yang et al.(2007)]{Yetal07}
402: Yang, C.-C., Gruendl, R.~A., Chu, Y.-H., Mac Low, M.-M., \&
403: Fukui, Y.\ 2007, ApJ, in press
404: \bibitem[Zinnecker \& Yorke(2007)]{ZY07}
405: Zinnecker, H., \& Yorke, H.~W.\ 2007, \araa, 45, 481
406: \end{thebibliography}
407:
408: \end{document}
409: