THE ASTROPHYSICAL JOURNAL, 486:L59L62, 1997 September 1
©1997. The American Astronomical Society. All rights reserved. Printed in U.S.A.

---

Up:

A New Molecular Hydrogen Outflow in Serpens

T. M. HERBST, ¹ S. V. W. BECKWITH, ¹ AND M. ROBBERTO ^{1, 2}

Received 1997 April 7; accepted 1997 June 10

ABSTRACT

     We report the discovery of a new molecular hydrogen outflow in the Serpens molecular cloud. Narrowband filter images taken in the 2.12 m v = 10 S(1) transition of H₂ and adjacent continuum reveal a series of bright knots of pure line emission apparently emerging to the north-northwest from the embedded source SMM-3 and passing close to the visible star CK-8. Low-resolution H- and K-band spectra of the region show more than a dozen distinct H₂ transitions, whose strength ratios point to shock heating with T_exc 2000 K. Echelle spectra of the S(1) transition with 20 km s^-1 resolution reveal unusual kinematics: the line center velocity increases linearly with distance to the north-northwest from SMM-3 until the bright knots of emission, at which point the velocity begins dropping to a fraction of its maximum value. The molecular hydrogen emission likely arises in limb-brightened bow shocks as a jet from SMM-3 encounters the ambient molecular cloud. This scenario is strengthened by recent HCO⁺ and SiO submillimeter observations of SMM-3, which show an apparent outflow corresponding to the H₂ structures.

Subject headings: infrared: ISM: lines and bandsinfrared: starsISM: jets and outflowsshock wavesstars: mass lossstars: premain-sequence

CONTENTS

• 1. INTRODUCTION
• 2. OBSERVATIONS AND RESULTS
• 3. DISCUSSION
  • 3.1. Excitation Mechanism, Temperature, and Mass of H₂
  • 3.2. Origin of the H₂ Emission
• 4. THE JET SOURCE
• ACKNOWLEDGMENTS
• REFERENCES
• FIGURES
• TABLES
• REFERENCES TO THIS ARTICLE

FOOTNOTES

     ¹ Max-Planck Institut für Astronomie, Königstuhl 17, 69117 Heidelberg, Germany.

     ² Osservatorio Astronomico di Torino, Strada Osservatorio 20, I-10025 Pino Torinese, Italy.

§1. INTRODUCTION

     Outflows from young stellar objects are central to the star formation process, since they both catalyze molecular cloud collapse by carrying away angular momentum and prevent further protostar formation by injecting energy and momentum into the ambient medium. These outflows have revealed themselves primarily by their interaction with the surrounding cloud, either through visible wavelength spectral lines in Herbig-Haro objects, or through millimeter wavelength transitions of CO in large-scale flows.

     The recent availability of large-format infrared detector arrays has added a new dimension to the search for stellar outflows: the impact of a wind on the ambient cloud will produce shock fronts that glow brightly in near-IR atomic and molecular lines. For example, in a classical bow shock, forbidden emission of Fe II will be associated with the strong, dissociative shocks near the wind terminus, while vibrational-rotational transitions of molecular hydrogen will trace the slower, oblique shocks near the flow periphery.

     In this Letter, we report the discovery, through wide-field near-IR imaging, of a new molecular hydrogen outflow in the Serpens molecular cloud. The H₂ likely arises in bow shocks produced by an outflow from the heavily embedded source, SMM-3. The following section describes the near-IR imaging and spectroscopy of the region, while § 3 discusses the nature, excitation, and source of the molecular hydrogen emission.

§2. OBSERVATIONS AND RESULTS

     We discovered the jet while observing the Serpens molecular cloud during poor weather conditions on 1996 June 9, using the MAGIC IR camera (Herbst et al. 1993) mounted at the Cassegrain focus of the Calar Alto 2.2 m telescope. Figure 1 shows images of the region taken 2 days later in the 2.12 m v = 10 S(1) H₂ line filter and the adjacent continuum. A number of previously unreported molecular hydrogen structures appear in the S(1) filter image, including a prominent filament extending approximately 10 to the north-northwest of the visible star CK-8 (Churchwell & Koorneef 1986). A significantly fainter source lies 10 to the south-southeast. The continuum image shows no corresponding structures.

Fig. 1

     We performed follow-up observations on 1996 August 25 using a similar filter set in the UKIRT facility IR camera, IRCAM3. These measurements took advantage of the newly commissioned tip-tilt secondary built by the MPIA (Glindemann et al. 1997). Figure 2 shows that the higher spatial resolution of this instrument combination has resolved the filament into at least six distinct knots of emission. The structure appears slightly misaligned or displaced to the southwest of CK-8. Table 1 lists the coordinates of all the knots of H₂ emission in Figures 1 and 2.

Fig. 2

     Figure 2c locates the embedded source SMM-3 (Casali, Eiroa, & Duncan 1993, hereafter CED), based on new 3.4 and 2.7 mm continuum observations (Hogerheijde et al. 1997) with the millimeter array of the Owens Valley Radio Observatory. Note that the original position quoted in CED is approximately 5 west of this location. Also plotted in Figure 2c are the HCO⁺ and SiO contours from Hogerheijde et al. 1997. The presence of a millimeter continuum source and aligned H₂, HCO⁺, and SiO structures strongly suggest an outflow from SMM-3 at P.A. = -20°.

     We acquired spectra of the molecular filament at low resolution in the H and K photometric bands and at high resolution in the v = 10 S(1) line on 1996 July 17, using the UKIRT facility spectrometer, CGS4. A slit orientation of P.A. = -20° included all the line emission as well as CK-8 and the star labeled in Figure 2, allowing a convenient test of slit placement and spatial scale.

     Extended emission from at least a dozen distinct transitions of H₂ appears in the low-resolution spectra. Gaussian fits to the high-resolution S(1) line profiles produce the kinematics shown in Figure 3. The velocity shift of the H₂ line center increases linearly with distance to the north-northwest from SMM-3 until the bright knots of emission are encountered, at which point the velocity begins dropping to some fraction of its extremum value. Far from SMM-3, the line center velocity approaches that of the Serpens cloud. Although not plotted, the deconvolved FWHM of the transition remains a relatively constant 44 km s^-1.

Fig. 3

§3. DISCUSSION

§3.1. Excitation Mechanism, Temperature, and Mass of H₂

     The ratio of the v = 10 S(1) to the v = 21 S(1) line is typically 5 : 1 or more in shock-heated gas and approximately 2 : 1 for UV excitation. This ratio is 820 along the entire filament. A stronger discriminant against UV fluorescence is based on the absence of higher excitation H₂ transitions in the H band (Herbst et al. 1996, hereafter Paper I). There is no hint of these diagnostic lines, demonstrating that shock heating powers the H₂.

     Assuming local thermodynamic equilibrium, the logarithm of the column density per degenerate sublevel, _v,J, is equal to the upper state energy difference, _v,J (expressed in kelvins), divided by the excitation temperature T_exc (e.g., Paper I). Figure 4a plots ln (_v,J) versus _v,J relative to the v = 10 S(1) transition in the vicinity of knots BC. A single line with inverse slope T_exc = 2000 K matches all the H- and K-band measurements, although there is some hint of a flattening with greater _v,J. Such curvature implies that the higher excitation lines trace warmer regions of the shock, an effect that has been observed in other objects (e.g., the Orion Nebula; see Beck, Lacy, & Geballe 1979; Beckwith et al. 1983).

Fig. 4

     Similar plots of ln (_v,J) versus _v,J along the entire length of the filament produce good fits characterized by a small range of excitation temperatures. The calculations assume A_V = 3, the extinction to the Serpens molecular cloud (Eiroa & Casali 1992, hereafter EC92). This assumption is supported by the fact that higher or lower extinction values introduced a wavelength dependence of ln (), disrupting the straight line fits. Figure 4b shows the derived T_exc as a function of position along the slit. The excitation temperature shows an approximately 30 K arcsec^-1 increase with increasing distance from SMM-3.

     We use T_exc to estimate the partition function and hence derive the total column density and mass of molecular hydrogen (e.g., Paper I). The flux and emitting area of the six H₂ knots in Figure 2 imply a total mass of about 5 × 10^-7 M of molecular gas at the assumed 310 pc distance of the Serpens cloud (Chavarriá et al. 1988).

     Hartigan, Raymond, & Hartmann (1987) note that a typical bow shock should produce emission lines with profile full width at zero intensity (FWZI) equal to the actual bow shock velocity, independent of the viewing geometry. Although the S(1) line profiles may not trace the full velocity width of the shock, the estimated FWZI 70 km s^-1 of the S(1) profiles coupled with the measured velocities, implies an inclination of 60°70°. The jet then has a total extent of 7000 AU for d = 310 pc and can form in approximately 800 yr.

§3.2. Origin of the H₂ Emission

     The bright pairs of knots (BC and DE) likely represent the limb-brightened collars of bow shocks (e.g., Smith 1991). These bow shocks arise when the outflow from SMM-3, also traced by the HCO⁺ and SiO (Figure 2c), encounters the ambient molecular cloud. This interpretation is supported by the presence of 1.644 m [Fe II] emission in the region of the knots. The caps or working surfaces where the jet encounters the ambient medium should be the sites of strong, dissociative shocks and hence [Fe II] emission. Despite the relatively coarse spatial sampling of CGS4, the [Fe II] appears strongest at locations slightly farther from SMM-3 than the H₂ knot pairs, consistent with the bow shock scenario.

     The kinematics of the molecular hydrogen near SMM-3 also supports the jet interpretation. To the north-northwest, the S(1) transition shows a linear increase in velocity with distance, a behavior often seen in molecular outflows. The corresponding increase in shock excitation temperature (Fig. 4) suggests that the velocity increase is real and is not, for example, a projection effect (Shu et al. 1991). A decrease in flow velocity with time can produce this behavior. The apparent deceleration 12 north-northwest of SMM-3 may result from a decrease in shock velocity as the wind encounters dense parts of the ambient cloud. The fact that this deceleration zone corresponds to the brightest emission supports this view. Note that the molecular hydrogen outflow powering Burnham's Nebula displays similar velocity and brightness variations (Herbst, Robberto, & Beckwith 1997).

     The H₂ to the south-southeast is also somewhat blueshifted with respect to SMM-3, inconsistent with the simple jet picture presented above. This gas may represent a mix between emission from the ambient cloud and jet. Notice that the line center velocities approach that of the Serpens cloud wherever the emission is faint (-10, 5, and 25 in Fig. 3).

§4. THE JET SOURCE

     Is SMM-3, in fact, the source of the molecular jet? The close alignment of the star CK-8 with the H₂ knots suggests that it, and not SMM-3, may be responsible for the outflow. To investigate this possibility, we observed the jet region at 520 m using MAX, the new MPIA mid-IR-camera, on UKIRT in 1996 August. We readily detected SVS-20 and a red (N - K 4) point source whose position corresponds within 05 of CK-8. No evidence of SMM-3 appeared at any wavelength. The 113 m spectral energy distribution of CK-8 shows an IR excess indicative of a circumstellar disk. Such disks play a central role in the collimation of stellar jets. Additional evidence for CK-8 as the source of the molecular jet comes from the spatial, kinematic, and temperature symmetry of the H₂ knots with respect to the star: SMM-3 is not so obviously central (Figs. 2, 3, and 4). Finally, both CK-8 (EC92) and the molecular hydrogen (§ 3.1) appear to suffer 3 mag of visual extinction, suggesting they lie at a similar distance along the line of sight into the cloud.

     Despite these arguments, we believe that the submillimeter source, SMM-3, produces the molecular hydrogen jet. Like most outflow sources, SMM-3 is a heavily embedded object with associated submillimeter line and continuum emission. The 8001300 m spectral index of SMM-3 implies substantial optical depth at these wavelengths (CED), explaining our nondetection of the source with the MAX camera. Evidence of high-velocity CO gas at the location of SMM-3 (White, Casali & Eiroa 1995), coupled with the morphology of the HCO⁺ and SiO (Fig. 2c; Hogerheijde et al. 1997), strongly suggests an outflow aligned with the H₂ jet. A line joining H₂ knots AE with knot F passes within 03 of the revised SMM-3 position and is displaced 1 to the southwest of CK-8. The kinematics of the H₂ also corresponds well with that of the submillimeter tracers in the jet zone (Fig. 3).

     The general increase of extinction to the southeast in the Serpens cloud (CED) can obscure most of the receding part of the H₂. Increased densities in this region would be a prerequisite to producing detectable SiO emission. Finally, the immediate environment of SMM-3 shows a local maximum in T_exc (Fig. 4), and there is no brightness or line width enhancement of the molecular hydrogen near CK-8, an effect observed near more obvious jet sources (e.g., Paper I).

ACKNOWLEDGMENTS

     It is a pleasure to acknowledge fruitful discussions with Mordecai-Mark MacLow, Reinhard Mundt, and Michael Smith. Michiel Hogerheijde kindly provided the HCO⁺ and SiO maps prior to publication. This research has made use of the SIMBAD database, operated at CDS, Strasbourg, France.

REFERENCES

• Beck, S. C., Lacy, J. H., & Geballe, T. R. 1979, ApJ, 234, 213 First citation in article
• Beckwith, S. V. W., Evans, N. J., III, Gatley, I., Gull, G., & Russel, R. W. 1983, ApJ, 264, 152 First citation in article | NASA ADS
• Casali, M. M., Eiroa, C., & Duncan, W. D. 1993, A&A, 275, 195 (CED) First citation in article | NASA ADS
• Chavarriá, K. C., DeLara, E., Finkenzeller, U., Mendoza, E. E., & Ocegueda, J. 1988, A&A, 197, 151 First citation in article | NASA ADS
• Churchwell, E., & Koorneef, J. 1986, ApJ, 300, 729 First citation in article | NASA ADS
• Eiroa, C., & Casali, M. M. 1992, A&A, 262, 468 (EC92) First citation in article | NASA ADS
• Glindemann, A., Pitz, E., Hippler, S., Rohloff, R.-R., & Wagner, K. 1997, in preparation First citation in article
• Hartigan, P., Raymond, J., & Hartmann, L. 1987, ApJ, 316, 323 First citation in article | NASA ADS
• Harvey, P. M., Wilking, B. A., & Joy, M. 1984, ApJ, 278, 156 First citation in article | NASA ADS
• Herbst, T. M., Beckwith, S. V. W., Birk, C., Hippler, S., McCaughrean, M. J., Mannucci, F., & Wolf, J. 1993, Proc. SPIE, 605 First citation in article
• Herbst, T. M., Beckwith, S. V. W., Glindemann, A., Tacconi-Garman, L. E., Kroker, H., & Krabbe, A. 1996, AJ, 111, 2403 (Paper I) First citation in article | NASA ADS
• Herbst, T. M., Robberto, M., & Beckwith, S. V. W. 1997, AJ, in press First citation in article
• Hogerheijde, M. R., Blake, G. A., van Dishoeck, E. F., & van Langevelde, H. J. 1997, in preparation First citation in article
• Shu, F., Ruden, S. P., Lada, C. J., & Lizano, S. 1991, ApJ, 370, L31 First citation in article | NASA ADS
• Smith, M. D. 1991, MNRAS, 252, 378 First citation in article | NASA ADS
• White, G. J., Casali, M. M., & Eiroa, C. 1995, A&A, 298, 594 First citation in article | NASA ADS

FIGURES

Full image (27kb)

Full image (46kb)

Full image (24kb)

Full image (35kb)

TABLES

TABLE 1
COORDINATES OF NEW MOLECULAR HYDROGEN STRUCTURES IN FIGURES 1 AND 2

Source	R.A. a(1950)	Decl.
CK-8 b...	18 27 26.8	1 12 4
A...	-3.0	2.6
B...	-2.6	5.4
C...	-3.6	6.3
D...	-4.6	8.6
E...	-3.4	9.6
F...	2.6	-10.7
S1...	18 27 24.7	1 11 48
S2...	18 27 24.3	1 11 15
S3...	18 27 25.7	1 11 40
S4...	18 27 20.4	1 12 51
S5...	18 27 18.1	1 12 10
S6...	18 27 18.4	1 13 11
S7...	18 27 15.6	1 13 62
S8...	18 27 16.4	1 14 16
S9...	18 27 16.7	1 14 48
S10...	18 27 17.7	1 14 45
S11...	18 27 21.3	1 13 51
S12...	18 27 26.6	1 15 28
FIRS 1 c...	18 27 16.6	1 13 39

     NOTE. Units of right ascension are hours, minutes, and seconds, and units of declination are degrees, arcminutes, and arcseconds.