Massive Young Stellar Objects and Outflow in the Infrared-Dark Cloud G79.3+0.3
DDraft version April 30, 2019Typeset using L A TEX twocolumn style in AASTeX62
Massive Young Stellar Objects and Outflow in the Infrared-Dark Cloud G79.3+0.3
Anna S.E. Laws,
1, 2
Joseph L. Hora, and Qizhou Zhang Astrophysics Group, University of Exeter, Stocker Road, Exeter, EX4 4QL, U.K. School of Physics and Astronomy, University of Southampton, Highfield, Southampton SO17 1BJ, UK Center for Astrophysics | Harvard & Smithsonian, 60 Garden Street, Cambridge, MA 02138, USA
ABSTRACTG79.3+0.3 is an infrared-dark cloud (IRDC) in the Cygnus-X complex that is home to massive deeply-embedded Young Stellar Objects (YSOs). We have produced a Submillimeter Array (SMA) 1.3 mm continuumimage and CO line maps of the eastern section of G79.3+0.3 in which we detect five separate YSOs. We haveestimated physical parameters for these five YSOs and others in the vicinity of G79.3+0.3 by fitting existingphotometry from
Spitzer , Herschel , and ground-based telescopes to spectral energy distribution (SED) models.Through these model fits we find that the most massive YSOs seen in the SMA 1.3 mm continuum emissionhave masses in the 5 − M (cid:12) range. One of the SMA sources was observed to power a massive collimated COoutflow extending at least 0.94 pc in both directions from the protostar, with a total mass of 0.83 M (cid:12) and adynamical timescale of 23 kyr. Keywords: stars: formation – stars: massive – ISM: general – ISM: clouds – ISM: structure – ISM: individualobjects: G79.3+0.3 INTRODUCTIONMassive stars are known to form in clusters in dense molec-ular clouds, but there are several aspects of the formation pro-cess that are not well-understood. For example, it is not yetknown which is the dominant massive star formation mech-anism (i.e., monolithic collapse, competitive accretion, andstellar collisions and mergers; e.g. see reviews by Zinnecker& Yorke 2007; McKee & Ostriker 2007; Tan et al. 2014;Motte et al. 2017). In order to determine the critical pro-cesses involved, it is necessary to find and study examplesof massive stars in various stages of formation. This task ismade difficult by the relatively short formation period andlifetime of massive stars, the low numbers of massive starsgiven by the mass function of stellar clusters, and the fact thatthe birthplace of these stars are in the cores of dense molecu-lar clouds, making the massive Young Stellar Objects (YSOs)difficult to observe.With the availability of sensitive infrared surveys fromthe
Spitzer
Space Telescope (Werner et al. 2004), and high-resolution (sub)millimeter interferometers such as the Sub-millimeter Array (SMA), it is possible to probe these deeply-embedded YSOs near the beginning of their formation pro-
Corresponding author: Anna S.E. [email protected] cess. Several investigations have identified and studied theobjects known as infrared dark clouds (IRDCs), which havebeen often found to host sites of massive star formation (e.g.Wang et al. 2006; Rathborne et al. 2006; Cyganowski et al.2008; Peretto & Fuller 2010). The SMA has been used tostudy young massive embedded clusters, revealing informa-tion on the earliest stages of formation and mass accretion(e.g., Zhang et al. 2009; Zhang & Wang 2011; Wang et al.2013, 2014). By combining the IR and submm data, we canstudy the formation of YSOs and the cluster from its earlieststages onward.Cygnus-X is one of the most massive molecular cloud com-plexes in the Galaxy with a total mass of 3x10 M (cid:12) (Schneideret al. 2006), and is one of the nearest massive star-formingcomplexes at a distance of 1 . ± .
08 kpc (Rygl et al. 2012).Along with YSOs, Cygnus-X includes 800 distinct H II re-gions (evidence of widespread massive star formation), sev-eral OB associations, and several Wolf-Rayet and O3 stars(Beerer et al. 2010). The Cygnus OB2 star cluster containshundreds of massive OB and O stars, with a stellar mass of ∼ . × M (cid:12) (Wright et al. 2015). Altogether this makesthe complex an excellent target for observing massive starsand their formation.In this paper, we study a region in the IRDC G79.3+0.3(Carey et al. 1998; Redman et al. 2003, see Figure 1). ThisIRDC is one of the largest in the Cygnus-X region, and lies inthe central subsection of Cygnus-X near DR15, an H II region a r X i v : . [ a s t r o - ph . S R ] A p r Figure 1.
A near-IR view of the G79.3+0.3 and its surroundings. The region in G79.3+0.3 observed with the SMA is directly in the center ofthis image, highlighted with a white square that is enlarged within the inset. The inset shows the more immediate surroundings of the region ofthe IRDC imaged with the SMA, spanning about 3.7 × × µ m (blue), 8 µ m(green), and 24 µ m (red) from the Spitzer
Space Telescope. that has been previously targeted as a candidate for high-massstar formation (Rivera-Gálvez et al. 2015; Schneider et al.2016). Star formation is especially promising in DR15 withits molecular pillar and prominent bright envelope (as seenin Figure 1), though the IRDC is likely not interacting withthe pillar (Rivera-Gálvez et al. 2015). Located close to theIRDC is the luminous blue variable (LBV) candidate starG79.29+0.46, seen as the blue star in the middle of the circularred nebula to the right of center in Figure 1. Studies by Umanaet al. (2011) with the EVLA and observations of NH (1, 1)and (2, 2) emission by Rizzo et al. (2014) showed evidencethat the LBV is interacting with the IRDC. Following Rizzo et al. (2014), we assume that the IRDC G79.3+0.3 is locatedat the same distance as Cygnus-X at 1.4 kpc.A total of 226 young stellar sources have previously beenseen in the DR15 area, and the Class 0 and I objects aremostly intermediate- to high-mass stars (Rivera-Gálvez et al.2015). G79.3+0.3 has a molecular mass of 803 M (cid:12) , whichis average for the sample of 45 massive star-forming IRDCsgiven in Ragan et al. (2012), but high compared with theother Cygnus-X IRDCs (Calahan et al., in prep.) thus makingthe cloud a likely location of massive YSOs in Cygnus-X.G79.3+0.3 is much closer than most IRDCs, at a distance of1.4 kpc compared with the mean distance to classical IRDCsof > O masers were also detected in massive IRDC clumps(Wang et al. 2006, 2012, 2014), indicative of star formingactivities in these clouds. Thanks to the physical connectionswith accretion, outflow energetics provide valuable insightsinto the accretion process of protostars.In this paper we present the results of our study of the YSOsin G79.3+0.3, including new high-resolution SMA interfer-ometric observations of a portion of the G79.3+0.3 IRDC.In Section 2 we provide the details of the new observations,and also describe the photometry we used from the
Spitzer and
Herschel missions and from ground-based near-IR sur-veys. Section 3 gives the results of our SED modelling ofYSOs. The analysis of the outflow detected in SMA source1 is described in Section 4, along with discussion of formerstudies of these objects. We summarize the main findings inSection 5. OBSERVATIONS AND DATA REDUCTION2.1.
SMA
The Submillimeter Array (Ho et al. 2004) is an interfer-ometer consisting of eight 6 m-diameter antennas operatingat millimeter and sub-millimeter wavelengths located nearthe summit of Maunakea, Hawaii. The SMA observationsof G79.3+0.3 were carried out on July 8th 2012 and August3rd 2012 using the Compact and Subcompact array config-urations respectively. The array used the 230 GHz receiverstuned to an LO frequency of 224.92 GHz. With an IF fre-quency of 4 to 8 GHz the observations covered sky frequen-cies of 216 . − . . − . The Submillimeter Array is a joint project between the Smithsonian As-trophysical Observatory and the Academia Sinica Institute of Astronomy andAstrophysics and is funded by the Smithsonian Institution and the AcademiaSinica.
The FWHM of the primary beam response of the 6 m an-tennas is approximately 55 (cid:48)(cid:48) . The observations employed atotal of 10 pointings separated by half of the FWHM of theprimary beam. We used MWC349A and QSO 2007+404 asthe time dependent gain calibrators. The spectral bandpasswas calibrated using 3C279. The flux calibration was carriedout through observations of a known flux source Titan. Theuncertainty of the absolute flux calibration is about 15%. Thesystem temperatures during the observations were from 120to 160 K.We calibrated the data using the MIR software (MillimeterInterferometer Reduction) following the procedure outlinedat the SMA data reduction website. The calibrations includesystemic temperature correction, bandpass, time dependentgain variations, and flux calibration. The calibrated visibili-ties were then exported to MIRIAD and CASA for imaging.We separated the continuum and line emission in visibilitiesand then Fourier transformed and cleaned. We used tclean task in CASA with Briggs weighting and a robust parameterof 0.5. The resulting 1.3 mm continuum image has a 1 σ rmsnoise level of 4 mJy beam − . The 1 σ rms in the 1 km s − channels is 140 mJy beam − . The spatial resolution of thecontinuum image is approximately 2 . (cid:48)(cid:48) Infrared Photometry
We used four photometry catalogs in our study of Cygnus-X. The first contained photometry from the following infraredsurveys: 2MASS (Skrutskie et al. 2006), UKIDSS (Dye et al.2006), and
Spitzer (Hora et al. 2011). The catalogs containover three million objects in the region, of which we classi-fied 30,902 as YSOs (Classes 0, I and II) using the methodsdescribed by Gutermuth et al. (2008). There are 28 objectsclassified as YSOs in the field observed by the SMA. For theJ, H, and K bands, we used data from the two ground-basedsurveys UKIDSS and 2MASS. The UKIDSS survey has betterresolution than 2MASS ( ∼ (cid:48)(cid:48) vs. 4 (cid:48)(cid:48) , respectively) but is sat-urated for objects brighter than 11 mag, so we used 2MASSdata for recorded magnitudes below 11 mag and UKIDSSdata otherwise. The IRAC images have FWHM resolutionsof 1 . (cid:48)(cid:48)
66, 1 . (cid:48)(cid:48)
72, 1 . (cid:48)(cid:48)
88, and 1 . (cid:48)(cid:48)
98 for the 3.6, 4.5, 5.8, and 8 µ mbands, respectively (Fazio et al. 2004). The MIPS instrumenthas a 24 µ m resolution of 6 (cid:48)(cid:48) (Rieke et al. 2004).We utilized Herschel data from three separate catalogs,one containing photometry for 579 YSOs in Cygnus-X de-tected by both
Spitzer and
Herschel /SPIRE at 250, 350, and500 µ m (Kirk 2014). We also downloaded data from twocatalogs containing photometry for 698 YSOs in Cygnus-Xobserved with the PACS 70 and 160 µ m bands (Marton et al.2017). The spatial resolution of Herschel is approximately ∼ cqi/mircook.html Table 1.
Fluxes and masses for the 1.3 mm sourcesSMA R.A. Dec. Flux density MassID (J2000.0) (mJy) (M (cid:12) )1 20:32:22.1 40:20:17.1 65 . ± . .
362 20:32:22.0 40:20:09.7 128 ± .
603 20:32:21.4 40:20:14.1 150 ± .
394 20:32:28.6 40:19:41.6 61 . ± . .
215 20:32:23.0 40:19:22.7 39 . ± . . (cid:48)(cid:48) , 13 (cid:48)(cid:48) , 18 (cid:48)(cid:48) , 25 (cid:48)(cid:48) , and 36 (cid:48)(cid:48) for the 70, 160, 250, 350, and500 µ m bands, respectively. RESULTS3.1.
SMA Continuum
A comparison of G79.3+0.3 at different wavelengths isshown in Figure 2, with the new SMA 1.3 mm continuumimage shown in the lower right panel. Five distinct continuumpeaks (labeled SMA objects 1 through 5) can be seen, andeach object’s position and flux is given in Table 1. Theposition of SMA sources 1, 4, and 5 are indicated by crossesoverlaid on the
Spitzer and
Herschel images in Figure 2. Twoof these objects (labeled 2 and 3) show extended emissioncoinciding with molecular outflows. We will explore outflowsidentified in the CO emission in Section 3.3.The continuum flux measurements can be used to calculatethe mass of the gas surrounding the objects following themethod in Motte et al. (2007, eqn.1). We assumed valuesof dust emissivity κ . = .
01 cm g − , T dust = ±
30% relativeto each other when dust temperatures vary from 15 to 25 K,or by ±
50% for the 10 to 20 K temperature range.Using the assumed distance to G79.3+0.3 of 1 . ± .
08 kpc,the largest project separation between SMA objects is 0.76 pcbetween objects 3 and 4, and the smallest separation is 0.069pc between 1 and 3. Objects 1, 4 and 5 form a triangle withroughly equal side lengths. The separations are: 1–4, 0.70pc; 4–5, 0.58 pc; and 1–5, 0.38 pc, all with uncertainties onthe order of ± SED Modeling
We used SEDFitter v1.0 (Robitaille et al. 2007) to fitmodel YSO SEDs to each object’s photometry and return https://github.com/astrofrog/sedfitter an estimate of its physical parameters, including mass andluminosity. It is necessary to find the object parameters bycomparing observations with models as direct observationalmethods cannot distinguish features such as the separate cen-tral object, disc, and envelope. For objects with multiplemodel fits, we selected the best fit as that with the lowestchi-square value and derived object parameters from that fit.The two constraints fed into the SEDFitter are the distanceto the object 1 . ≤ d ≤ . . ≤ A v ≤
100 mag. We created a custom filter toincorporate the SMA photometry with SEDFitter followingthe documentation (Robitaille 2013). The atmospheric trans-mission at the SMA is 100% in the frequency ranges coveredby both upper and lower sidebands (Welch 1988), so the fil-ter uses a simple function with 100% transmittance in thesefrequency ranges and 0% outside.Most YSOs in the
Spitzer catalog have no detection at thelonger PACS and/or SPIRE wavelengths, constraining theseobjects to have a flux below a certain sensitivity limit atthese wavelengths. We estimated these limits in Cygnus-Xusing a histogram of the number of objects measured withvarious fluxes. The upper limits for undetected sources areas follows: 223 and 827 mJy for the PACS 70 and 160 µ mbands respectively; and 3000, 6000, and 4000 mJy for theSPIRE 250, 350, and 500 µ m bands respectively. We similarlyestimated a sensitivity limit for the SMA observations byassuming that the limit was 10% of the flux of the faintestdistinct object. For SMA object 5 at flux 39.0 mJy, this gavea sensitivity limit of 3.90 mJy.We used SEDFitter to fit all of the YSOs identified in theCygnus-X region using color-color and color-magnitude rela-tions (Beerer et al. 2010); the fluxes for the objects are givenin Table 2. The SEDFitter results for the 28 YSOs within theregion observed by the SMA are given in Table 3. The SEDfitting results for the whole of Cygnus-X will be discussed ina later paper.The SEDs of SMA objects 1, 4, and 5 are shown in Fig-ure 3. The data points shown are from the near-IR, Spitzer , Herschel , and SMA. Where the source is not detected in the
Herschel
SPIRE bands, upper limits were used, indicated bythe downward-pointed triangle symbols. Where the sourcewas not separately resolved in the
Herschel bands (such asSMA sources 1, 2, and 3), an upper limit based on the fluxfor all sources combined was used in the fitting process. ForSMA source 1, we used the flux values given in the MIPS 24 µ m catalog as an upper limit due to multiple sources present.The models have discontinuities in flux at wavelengths above200 µ m (e.g., see the left panel of Figure 3) since the signal-to-noise ratio in the models becomes poor above 100 µ mand the median uncertainty in the model fit is above 20% atmillimeter wavelengths (Robitaille et al. 2006). Two of theSMA sources each have only one valid fit, likely because the Figure 2.
Views of G79.3+0.3 as seen in various infrared wavelengths and by the SMA. The same 3.4 (cid:48) × (cid:48) region is shown in each panel. Inthe lower right panel the SMA image is shown with each of the continuum sources labeled. The positions of these sources are given in Table1. The SMA objects 1, 4, and 5 (marked with the crosses in the five other panels) are each clearly visible in every image up to and includingthe PACS 160 µ mimage. The SMA sources 2 and 3 are not detected at wavelengths shorter than 70 µ m. At 70 and 160 µ m, evidence for SMAsources 2 and 3 is present but they are not fully resolved because of the lower spatial resolution at these wavelengths. ( m) F ( e r g s / c m / s ) J203222.10+402017.1Model: 3012453_9Best fit = 39.809 A V = 62.9 Scale = 0.11 ( m) F ( e r g s / c m / s ) J203228.61+401941.6Model: 3011049_8Best fit = 93.195 A V = 24.8 Scale = 0.16 ( m) F ( e r g s / c m / s ) J203223.04+401922.7Model: 3019153_10Best fit = 90.566 A V = 7.8 Scale = 0.11 Figure 3.
SEDs and the best model fit for SMA objects 1 (left), 4 (middle), and 5 (right). Round markers with error bars are flux measurementsand their uncertainties, and downward-pointing triangles are upper limits for flux. Parameters derived from these SEDs are given in Table 3.
Table 2.
Fluxes for 28 YSOs in the SMA-observed Region of G79.3+0.3
R.A. Decl. Flux Density and Uncertainty (Jy)Name (J2000.0) (J2000.0)
J J err
H H err
K K err 3.6 3.6 errJ203222.10+402017.1 308.092102 40.338074 · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · ·
160 160 err 250 250 err 350 350 err 500 500 err 1300 1300 err · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · ·· · · · · · · · · · · · · · · · · · · · · · · · · · · · · ·· · · · · · · · · · · · · · · · · · · · · · · · · · · · · ·· · · · · · · · · · · · · · · · · · · · · · · · · · · · · ·· · · · · · · · · · · · · · · · · · · · · · · · · · · · · ·· · · · · · · · · · · · · · · · · · · · · · · · · · · · · ·· · · · · · · · · · · · · · · · · · · · · · · · · · · · · ·
Note—Table 2 is published in its entirety in the machine-readable format. A portion is shown here for guidance regarding its form and content.
Table 3.
SED-derived parameters for the 28 YSOs in the SMA-observed Region of G79.3+0.3R.A. Dec SMA
Spitzer
YSO Mass Luminosity Extinction ( A v )(degrees) (degrees) ID ID Class ( M (cid:12) ) ( L (cid:12) ) (mag)308.092102 40.338074 1 J203222.10+402017.1 1 5 . ± .
73 1543 ±
470 62 . ± . . ± .
55 815 ±
244 24 . ± . . ± .
20 37 . ± . . ± . . ± .
27 268 ±
80 91 . ± . . ± .
99 29 . ± . . ± . . ± .
97 34 . ± . . ± . . ± .
97 96 . ± . . ± . . ± .
95 122 ±
37 54 . ± . . ± .
94 67 . ± . . ± . . ± .
92 15 . ± . . ± . . ± .
90 32 . ± . . ± . . ± .
72 9 . ± .
88 100 ± . ± .
67 25 . ± . . ± . . ± .
66 24 . ± . . ± . . ± .
63 4 . ± .
37 14 . ± . . ± .
41 9 . ± .
82 14 . ± . . ± .
41 9 . ± .
82 16 . ± . . ± .
30 5 . ± .
79 38 . ± . . ± .
221 3 . ± .
03 29 . ± . . ± .
173 3 . ± .
18 22 . ± . . ± .
141 0 . ± .
241 5 . ± . . ± .
112 1 . ± .
39 23 . ± . . ± .
064 0 . ± .
216 11 . ± . . ± .
052 1 . ± .
38 30 . ± . . ± .
047 0 . ± .
207 19 . ± . . ± .
046 0 . ± .
075 9 . ± . . ± .
044 1 . ± .
37 6 . ± . . ± .
032 0 . ± .
034 6 . ± . input fluxes were too strongly constrained (Robitaille 2008).The χ value for SMA object 1 is large, indicating a poor fit.These are a few reasons to only use the SED-derived parame-ters as first estimates, along with further reasons discussed inSection 4.2. A selection of the G79.3+0.3 object parametersis given in Table 3 with the uncertainty in each parameter setto ±
30% following Saral et al. (2017).Figure 4 shows the spatial distribution of YSOs acrossG79.3+0.3 and its surrounding area, with the symbols show-ing their YSO class and mass. Within G79.3+0.3, the moremassive objects tend to be near the center of the cloud andlocated close to each other. These more massive objects arealmost entirely Class I objects, whereas the Class II objectsare lower mass and are located on the outskirts of the coveredarea. The distribution is similar in the IRDC region seen onthe right side of Figure 4, which is unsurprising given that thetwo regions are connected behind the foreground warm dustemission (Redman et al. 2003). 3.3.
Molecular outflows seen in CO To identify molecular outflows in G79.3+0.3, we examinedthe CO J=2-1 emission in various velocity bins across the re-gion. The dominant CO emission feature in the images is thenarrow linear structure at a position angle (PA) of 100 . ± . ◦ .This structure is seen in the blueshifted velocities from − − up to −
50 km s − , and in the redshifted velocitiesfrom 11 km s − up to 45 km s − . This high-velocity COemission traces a protostellar outflow from a young protostarin the region. Following Zhang et al. (2005), we integratethe blueshifted and redshifted CO emission to produce theoutflow image shown in Figure 5. We estimate outflow pa-rameters including its mass, energy, and momentum using the CO emission. The mass of the gas can be estimated fromthe column density, ¯ N (CO). Following the formula in Garden h m s s s s m s α (J2000) ◦ ' '' ' '' ' '' ' '' ' '' δ ( J ) SMAClass IClass II0.8 0.4 0.0 0.4 0.8log(Mass ( M fl )) Figure 4.
The spatial distribution of YSOs across G79.3+0.3 and its surrounding area. The region covered by our SMA data is outlined in bluenear the upper left of the image. Class I YSOs are plotted as diamonds, Class II YSOs as triangles, and the symbols are shaded according totheir mass derived from the SED fitting using the color scale shown below the image, with darker blue being higher mass. Each of the threeSMA objects is highlighted with a white square. The background image is the 8 µ m Spitzer image, and the dark black regions are IRDCs. et al. (1991), we find the column density from the equation:¯ N ( CO ) = . · ·
12 exp ( . / T ex )· ( T ex + . ) exp (− . / T ex ) · ∫ T B τ d v ( − exp (− τ )) . (1)The integral is simplified by placing ¯ N into the M gas equationto obtain: M gas = . · − ·
12 exp ( . / T ex ) · ( T ex + . ) exp (− . / T ex ) · ¯ τ − exp (− ¯ τ ) · (cid:20) θ arcsec (cid:21) ∫ T B d v M (cid:12) . (2)This method uses estimates of the mean optical depth ¯ τ ofCO, and the excitation temperature T ex . We assume that COis optically thin, and a typical T ex of 20K.The total outflow mass can be used to derive the outflowmomentum P = M ( ∆ v , ¯ τ ) | ¯v | and energy E = M ( ∆ v , ¯ τ ) ¯ v for mean optical depth ¯ τ . We can combine these results withthe dynamical timescale of the outflow t dyn to find: the rate of outflow mass (cid:219) M = M / t dyn ; the mechanical force F = P / t dyn ;and the outflow luminosity L = E / t dyn (Zhang et al. 2005).The dynamical timescale can in turn be found by taking theratio of the maximum extent of the outflow and the maximumspeed of the gas. In order to compare with the studies in theliterature, we did not correct for the inclination angle of theoutflow when estimating outflow parameters.The CO emission that tracks the major outflow stemmingfrom SMA object 3 is shown in Figure 5. The outflow pa-rameters are given in Table 4. We estimated the spatial extentof the outflow using trigonometry, finding projected lengths L R = .
43 pc and L B = .
94 pc for the red- and blueshiftedhalves of the outflow, respectively. The actual extent of theoutflow is likely larger, given that it is inclined towards us atsome unknown angle. DISCUSSION4.1.
SMA Continuum
The SMA objects 1, 4, and 5 can clearly be seen withthe infrared instruments featured in Figure 2. It is possi-ble that all three closely-placed SMA objects (1 through 3)
Figure 5. H and CO tracers of the large outflow from SMA object 3, plotted on the 24 µ m Spitzer image. The CO emission is integratedover 7 . . − for the redshifted emission (red contours), and − . − . − for the blueshifted emission (blue contours).The contours are spaced at 7.8 Jy km s − intervals. Green contours show the H emission from Makin & Froebrich (2018) as described inSection 4.3. The black markers labeled 1 through 5 show the locations of the five sources in our SMA 1.3 mm continuum image. The origin ofthe H and CO is consistent with SMA object 3 being the source of the outflow. have significant emission in the
Herschel bands, but weretoo poorly-resolved to be distinguished as individual objects.There is some indication that Source 1 is slightly extended inthe direction of Object 3 at 70 and 160 µ m. The dominantemission is still from Source 1 at 160 µ m. However, at the1.3 mm band Objects 2 and 3 are each more than twice asbright as Object 1, suggesting that it is likely that Objects 2and 3 are deeply embedded in the cloud, and are at an earlierstage of evolution.Our SMA continuum map is consistent with a previousobservation of G79.3+0.3 by Redman et al. (2003), who ob-served a portion of the IRDC at 3 mm using BIMA andresolved our SMA Objects 1 through 3 (their sources C, B,and A) and seem to have also detected SMA Object 5. Theirimage shows blueshifted HCO + (1–0) line emission that isconsistent with a strong outflow from Object 3, matching theoutflow that is traced with the CO line emission from theSMA. The BIMA and SMA observations agree that Object 1is the most massive star in G79.3+0.3. Redman et al. (2003)conclude that our SMA Object 1 will likely evolve into a Bstar on the main sequence and is too young to have disruptedthe IRDC and triggered further star formation, but that thiscould happen over the next 10 years.Motte et al. (2007, Object S37) reported a total mass of45 M (cid:12) for SMA Objects 1, 2 and 3 based on the MOMBOdata at the 1.2 mm band and a resolution of 11 (cid:48)(cid:48) . The SMAobservations have sufficient spatial resolution to distinguish the compact cores. Therefore, the mass estimates in this paperare consistent with that reported in Motte et al. (2007).SMA Objects 4, 5, and the grouping of 1 through 3 areroughly equidistant (the distances are: 1–4, 0.70 pc; 4–5,0.58 pc; and 1–5, 0.38 pc). This is similar to the massiveIRDC 18223 reported by Beuther et al. (2015), which hastwelve cores regularly spaced at 0 . ± .
18 pc and peakseparations varying between 0.19 pc and 0.70 pc, as well as toIRDC G28.34 (Zhang et al. 2009) with five regularly spacedcores. The similarity between the inter-core separations inIRDC 18223 and the inter-stellar separations in G79.3+0.3suggests that our five SMA objects were originally part ofthe same filamentary structure. This can be expanded uponfurther by considering SMA Objects 1 through 3, which areroughly equidistant with a separation of around 0.069 pc, andso were likely formed from a single massive gas structure thatfurther fragmented into the three objects. The separationsbetween our SMA objects are comparable to the Jeans lengthin G79.3+0.3 of 0.50 pc, assuming values of gas density andkinetic temperature from Carey et al. (1998, Table 3). Thissimilarity between inter-star distance and Jeans length is asexpected and has been seen in many other clouds (e.g. Teixeiraet al. 2006; Feng et al. 2016).4.2.
SEDs and Parameters
The SEDs of the three SMA objects are reasonable for afirst estimate of the object parameters, since the inclusion of0 λ F λ ( e r g s c m ‐ s ‐ ) Wavelength (μm)
SMA Object 1
PAH PAHSilicate CO ice Figure 6.
The spectrum of SMA object 1, adapted from Segura-Coxet al. (2011, Object CYGXS37). This spectrum was measured using
Spitzer ’s InfraRed Spectrograph (IRS). There are emission featuresdue to Polycyclic Aromatic Hydrocarbons (PAHs) at ∼ µ m,and absorption features due to H O ice at 6 µ m, silicates at ∼ µ m,and CO ice at ∼ µ m. SMA data gives model SEDs consistent with their previousclassifications based on
Spitzer data.Segura-Cox et al. (2011) reduced a spectrum for SMAObject 1 from data obtained in
Spitzer project ID 50045(Fazio et al. 2008), shown in Figure 6. This covers the5 µ m ≤ λ ≤ µ m range in much higher detail than ourSED in Figure 3. Both the spectrum and the best-fit SEDshow silicate absorption at 10 µ m, which is consistent withthe object being a Class I YSO. Notably, the spectrum showsfeatures usually seen in massive YSOs, including strong CO ice absorption at 15.4 µ m and H O ice absorption near 6 µ m(An et al. 2009). The 15.4 µ m absorption is caused by mixingof CO and CH OH, which are abundant in massive YSOs(An et al. 2011). There was some weak CH OH line emis-sion visible in our SMA spectra across G79.3+0.3, but thesignal-to-noise ratio was too poor to map it across the region.It is important to consider the SED-derived parameters asfirst estimates, rather than the true physical values of theobjects. We are using large uncertainties of ± M (cid:12) limit and are still activelyaccreting mass. Objects 1 and 4 also have high extinctionvalues compared with the expected values for Cygnus-X of 5 ≤ A v ≤
10, which indicates that the objects are deeplyembedded in the IRDC. On the other hand, Object 5 has verylow extinction and therefore might be in the foreground of thedark cloud.The area surrounding G79.3+0.3 shows many Class II ob-jects of varying mass, few of which appear significantly clus-tered. There is a patch of closely- placed Class II objects justbelow our IRDC at approximately 20 h m s , 40 ◦ (cid:48) (cid:48)(cid:48) (J2000.0). These objects could be part of the same cloud ifit has been obscured by foreground emission, as in the caseof the connection between G79.3+0.3 and the nearby IRDCseen in Figure 4. For the most part, the Class II objects arelower-mass and more spatially distributed than the Class Iobjects in this region of Cygnus-X.4.3. Outflow from SMA Object 3
The outflow parameters we derived have similar magni-tudes to the sample of 39 objects with outflows in Zhang et al.(2005).The total mass of the outflow is 0.83 M (cid:12) , consistent withthe sample of six outflows given in Lee et al. (2002, Table3), which cover a range 0.01 – 1.00 M (cid:12) . It is also twiceas massive as any of the six nearby outflows in DR21 foundby Hawley (2012, Table 4), despite those six outflows beingejected from much more massive YSOs. The outflow couldbe so massive and energetic because there is no other outflowdetected nearby that could disrupt the flow of material orotherwise interfere with the YSO’s accretion process.Other NIR images in the region surrounding G79.3+0.3have revealed H emission corresponding to the outflow fromSMA Object 3, that lies further out from the star than the CO emission (Makin & Froebrich 2018; Davis et al. 2010,“Catalogue of Molecular Hydrogen Emission-Line Objects(MHOs) in Outflows from Young Stars" object MHO 3597).The MHO catalog identifies the source of the outflow asJ203222.10+402017.06 (our SMA object 1) (Kryukova et al.2014), but our SMA observations reveal the true source to bethe newly-identified SMA Object 3 (Figure 5). The H emis-sion extends just beyond the furthest extent of the blueshiftedlobe, and about the same distance in the redshifted direction.The SMA field observed ends about halfway to the H lobein the redshifted direction, so we cannot tell if the redshifted CO emission extends a similar distance.
Table 4 . Estimates of the Parameters for SMA Object 3 OutflowParameter Redshifted lobe Blueshifted lobeMass
Table 4 continued Table 4 (continued)
Parameter Redshifted lobe Blueshifted lobe( M (cid:12) ) 0.37 0.46Momentum( M (cid:12) km s − ) 10.5 14.0Energy( M (cid:12) (km s − ) ) 163 221Extent(pc) 0.43 0.94Dynamical timescale(kyr) 10.5 23.0Mass rate(10 − · M (cid:12) s − ) 11.2 6.34(10 − · M (cid:12) yr − ) 3.53 2.00Mechanical Force(10 − · M (cid:12) km s − ) 3.17 1.93Luminosity(10 − · M (cid:12) (km s − ) s − ) 4.91 3.05(10 · ergs) 9.76 6.06 For all we can find out about the outflow from Object 3, wecannot determine much about the central object itself. Object3 has no previous detections in our catalogs, and so we couldnot fit a model SED and estimate the physical properties ofthis object. CONCLUSIONSWe have produced a 1.3 mm continuum image of the IRDCG79.3+0.3 using the SMA. The image shows that this regionof the IRDC contains five YSOs in an early stage of formation,one of which has a massive collimated CO outflow. Theregular spacing of the objects hints at the fragmentation scaleof the cloud being ∼ .
76 pc.We have estimated the properties of the YSOs using modelSED fitting. In all cases, the slopes of the model SEDsare consistent with the objects’ YSO classifications based on color-color and color-magnitude diagrams using the near- andmid-IR data. The model SEDs for the three SMA objects 1,4, and 5 returned masses in the range of 4 – 6 M (cid:12) . This isconsistent with their being massive YSOs, and in fact they arethe most massive and luminous of the YSOs we identified inthis region of G79.3+0.3.We have identified an enormous CO outflow from SMAobject 3. The outflow extent is at least 0.43 pc and 0.94 pc inthe redshifted and blueshifted lobes respectively. The pres-ence of this outflow is consistent with Object 3 being a pro-tostar, and is also supported by the outflow parameters beingof similar magnitude to parameters of massive outflows fromother studies. The total mass of the outflow is 0.37 and0.46 M (cid:12) and total momentum is 10.5 and 14.0 M (cid:12) km s − inthe red and blue lobes, respectively.ACKNOWLEDGEMENTSThe authors wish to recognize and acknowledge the verysignificant cultural role and reverence that the summit of Mau-nakea has always had within the indigenous Hawaiian com-munity. We are most fortunate to have the opportunity toconduct observations from this mountain.This work is based in part on observations made with the Spitzer
Space Telescope, which is operated by the Jet Propul-sion Laboratory, California Institute of Technology under acontract with NASA. Support for this work was provided byNASA through an award issued by JPL/Caltech.