This is Chapter 1 of my Ph.D. thesis, entitiled Fast Gas in High Mass Star Forming Regions, October 1997. For more information please contact me via email at acord [@] imagiware [.] com.

Copyright 1997, Jerry M. Acord, Jr. All rights reserved.

Chapter 1. Introduction and Overview

Massive stars are among the primary driving forces in the evolution of galaxies. The distribution and number of hot, high mass stars determines the visible structure seen in spiral galaxies like our own Milky Way. Massive stars emit most of their radiation in the ultraviolet, producing large regions of ionized hydrogen (HII regions). They have strong stellar winds which deposit momentum and mechanical energy into the surrounding interstellar medium (ISM) which, over their lifetime, are comparable to those in supernova explosions. They have a significant impact on the chemistry of their environment by providing energy for endothermic gas phase reactions, by heating interstellar dust and releasing volatile dust grain mantles into the gas phase, and by destroying dust close to the star. They largely determine the nature of a galaxy's ISM, which controls the rate of formation and the composition of the next generation of stars. Extremely massive stars undergo supernova explosions which generate much of the diffuse Galactic gamma- and X-radiation, enrich the interstellar medium with heavy elements, produce shocks which accelerate cosmic rays, destroy dust, and propel gas and dust to large distances above the Galactic disk. The explosions may also trigger the gravitational collapse of nearby molecular clouds, which leads to the next generation of star formation.

Once a molecular cloud, or some part of it, manages to become gravitationally unstable, three major phases of star formation have been predicted theoretically. The earliest is the isothermal collapse phase in which nonhomologous collapse proceeds at the free-fall rate (i.e., gas closer to the center falls toward the center faster than gas located further away). This results in the formation of a small, dense core. This phase continues until a small, star-like, optically thick, hydrostatic core (a ``protostar'') is formed containing a few tenths of a solar mass inside a free-falling outer envelope of diameter about 0.1 pc (0.3 light years). At this point the core is optically thick: photons originating within the core are more likely to be reabsorbed than to escape from the core. Since the core is optically thick, it cannot cool efficiently and its temperature increases, temporarily halting collapse of the core because of the increase in central pressure. At this point the accretion phase begins. During this phase the core grows in mass via accretion from the surrounding protostellar envelope which continues to collapse at the free-fall rate. Due to accretion, the core slowly contracts towards temperatures and densities necessary to sustain thermonuclear burning of hydrogen. The luminosity of the core is produced both by accretion shocks and possibly by some low temperature thermonuclear reactions. As the core grows in mass, it finally reaches the densities and temperatures needed to support hydrogen nuclear burning. Ultimately, accretion is halted by processes not yet fully understood and the star settles down with a temperature and luminosity appropriate for its mass (it becomes a ``main sequence'' star).

During the accretion phase, infalling matter may prevent the central star from forming a detectable HII region, but as the accretion stops, stars more massive than about 10 solar masses (spectral type B2) produce enough ultraviolet photons to form detectable HII regions. The ionizing photons and particle wind produced by the stars create hot, ionized centers that are completely surrounded by the cool molecular cloud. This is the ultracompact (UC) HII phase of massive star evolution. UC HII regions are smaller, denser, brighter, and younger than classical HII regions such as M17 and Orion. It has recently been discovered that several UC HII regions are apparently at the center of luminous, bipolar molecular outflows. Such outflows are thought to be driven by the accretion process and are integrally associated with pre-main sequence evolution. The issue for UC HII regions is whether the outflows are still being driven or are relics of an earlier accretion phase. If they are still being driven, the implication is that the stellar core is probably still accreting matter.

UC HII regions are among the brightest objects in the Galaxy at wavelengths near 100 microns. The warm circumstellar dust spectrum resembles a blackbody of temperature around 30 K and the ionized gas emits thermal bremsstrahlung emission at radio wavelengths. To generate strong infrared emission and block visible and ultraviolet radiation, the objects must have an outer layer of cool molecular gas and dust at temperatures of 20 to 30 K and densities greater than or equal to 10⁶ cm^-3. Thermal emission from the dust makes these regions bright far-infrared (FIR) sources.

UC HII regions were first discovered in single-dish radio observations of emission nebulae in the neighborhood of other young stars in the Galactic plane. Radio and microwave interferometry can resolve their structure. Through radio continuum surveys at centimeter wavelengths with the Very Large Array (VLA) radio interferometer at high angular resolution, Wood and Churchwell (1989), Garay et al. (1993), and Kurtz, Churchwell, and Wood (1994) have found a range of UC HII region morphologies -- spherical, core-halo, shell, cometary, and irregular. All are coincident with bright IR point sources found by the Infrared Astronomical Satellite (IRAS).

The warm, dense molecular gas associated with a sample of roughly a hundred UC HII regions have been observed in several molecular lines (Churchwell, Walmsley, & Cesaroni 1990). The high detection rates in the molecular lines strongly supports the premise that the regions are surrounded by natal molecular gas. Moreover, H₂O, OH, and CH₃OH masers are also commonly found towards UC HII regions and are clearly tracers of star formation activity.

Bipolar molecular outflows are associated with young stellar objects of all masses and luminosities. Although most observed to date are associated with low mass star formation, recent observations have detected outflows in the vicinity of UC HII regions. It is believed that during the early stages of star formation the newly formed star generates a fast, well collimated bipolar outflow that sweeps up the ambient gas in its vicinity, forming two cavities oriented in opposite directions. The interaction of the high velocity jets from young stars with the surrounding ambient gas produces strong shock waves. Outflows display a range of velocities, with those from low mass stars averaging roughly 5 km/s (Stahler 1994), although flow velocities of over 100 km/s have been observed. In systems observed at high spatial resolution with optically thin molecular line tracers, it is possible to determine the total mass of moving gas. Generally, these masses either equal or surpass those of the driving stars. It is not thought, therefore, that a molecular outflow consists of gas lost from these stars. Rather, it must be composed of ambient cloud gas that has been set in motion by an as yet unkown mechanism.

Bipolar outflows were first detected as high velocity wings (V - V_systemic > 10 km/s) in the CO radio emission lines from the Orion star forming region (Zuckerman et al. 1976; Kwan and Scoville 1976). Since then many sources of bipolar outflows have been detected and confirmed through spatial mapping, either with single-dish or interferometric observations involving many lines (for a review see Bachiller and Gómez-González 1992). While CO is the most sensitive tracer of the mass in a molecular outflow, its chemistry is relatively insensitive to non-dissociative (T < 10⁴ K) shocks (Mitchell 1984) which are the predominant type of shock in these outflows (Masson and Chernin 1993). On the other hand, SiO and SO abundances may be enhanced at the temperatures produced in this kind of shock (Mitchell 1984, Hartquist et al. 1980, Pineau des Forêts et al. 1993).

Sinuous knotty strands of optical emission, known as Herbig-Haro jets, are also seen extending in a bipolar fashion from embedded, young, low-mass stars. Spectroscopic studies have established that the knots, which are traditionally known as Herbig-Haro objects, represent gas that has been shock-heated to about 10⁴K; the pre-shock jet speed is typically 300 km/s (for a review see Stahler 1994). In many cases, the brightest Herbig-Haro objects are found at the jet ends; these terminal regions have been successfully modelled as bowshocks presumably created by fast jet material impacting molecular cloud gas (Raga, Bohm, & Solf 1986).

Recently Chernin and collaborators (Chernin et al. 1994a) studied the properties of the outflow created by a protostellar jet in a dense molecular cloud using three-dimensional modelling. They showed that this type of outflow produces extremely high velocity clumps of post-shock gas which resemble features seen in some outflows (Tafalla et al. 1994). Chernin and co-workers (Chernin and Masson 1992; Chernin et al. 1994a) have also shown that gas in the molecular cloud, stirred and dragged forward by the jet, develops a cocoon around the central jet. Depending upon the speed of the jet and the densities of the jet and ambient material, either the cocoon or the bowshock at the jet head may dominate the outflow's appearance (Chernin et al. 1994a).

In a survey of CO emission towards a sample of nine molecular outflows, Margulis and Snell (1989) found that in four cases, the highest velocity emission observed might be produced in a region distinct from that responsible for lower velocity wing emission which arises from the swept-up outflowing molecular gas. They argued further that the very high velocity emission may arise from gas entrained in the driving winds or jets of these outflows. The extremely high velocity outflows are apparently common in regions of moderate to high mass star formation (Choi et al. 1993; Shepherd & Churchwell 1996a, 1996b).

In a recent CO survey towards 94 UC HII regions Shepherd and Churchwell (1996a) report that over 90% of their sample display excess line wing emission, possibly due to outflows. They mapped twelve sources (Shepherd & Churchwell 1996a, 1996b), and showed that the CO emission arises in outflows in seven of these regions; in four cases, the UC HII regions or IRAS sources are, within the errors, on the axes of the flows and reasonably centered between the red and blue shifted lobes. This would seem to imply that the flows originate from, and are driven by, the central stars of the UC HII regions.

In fact, the large masses, energies, and momentum transfer in these flows probably indicate that they are driven by massive stars since they are at least an order of magnitude larger than those typically found for low mass stars. However, with an angular resolution of only 60", these data do not prove unambiguously that the flows are driven by the star that powers the UC HII region. In the immediate vicinity of UC HII regions, an entire cluster of emerging stars are generally present (Kurtz, Churchwell, and Wood 1994).

Only the most massive stars are detectable at radio wavelengths via their ionizing radiation, and the youngest ones of these may not be detected if mass accretion rates are high enough to absorb ionizing radiation close to the star so that no UC HII can form. In fact, Shepherd & Churchwell (1996b) have found several outflows where the UC HII region is either well off the flow axis or is very asymmetrically located relative to the positions of the red and blue shifted lobes.

If the flows are not driven by the central star of the UC HII regions, then what is the source of the flows? The large masses and energies associated with the flows strongly suggest that the central engine is probably a massive star (Shepherd & Churchwell 1996b). If it does not reside in the UC HII region and it is massive, the driver of the flows is probably a massive protostar (at an earlier evolutionary stage than the star that powers the UC HII region) that has not been able to form a detectable HII region yet because of rapid mass accretion. Such an object might be detectable in the infrared due to a warm dust cocoon or massive accretion disk and/or molecular line emission (e.g. high excitation transitions of CH₃CN, NH₃, or CS) from warm, dense gas accumulated around the protostar. They are unlikely to be detectable in the UV or soft X-rays because of high extinction.

Indeed, the most widely accepted theory for the formation of molecular outflows and jets from young stellar objects postulates that accretion disks and magnetic fields are the origins for these phenomena. According to the magnetocentrifugally driven outflow model of Shu et al. (1994), a stellar magnetic field which partially threads the inner boundary of the accretion disk around a protostar can lead to the formation of a magnetic fan or ``X'' point in the disk, with some field lines terminating on the stellar surface while others remain open. Material flowing along the field lines from this point in the disk can not only accrete onto the forming star, but may also be whipped out into space along the open field lines like beads on a wire, forming jets and sweeping up ambient gas into two oppositely oriented lobes --- a molecular outflow.

Thermal emission from the ground vibrational state (v=0) of the SiO molecule may be a good tracer of shocked gas in molecular outflows. From observations of a sample of molecular clouds containing outflows, Martín-Pintado et al. (1992) found small abundances of SiO in the ambient gas, regardless of temperature. However, the SiO abundance in the outflow component was observed to be typically two orders of magnitude higher than in the ambient cloud. At the highest velocities in the fast L1448 outflow, this enhancement increased to about 10⁵.

The low abundance of SiO in dark clouds can be understood by assuming that most of the SiO is depleted onto dust grains, while the large enhancements associated with bipolar outflows can be explained by assuming that the SiO is formed in shocked layers, in which shocks have partially destroyed dust grains or dust grain mantles, ejecting atomic silicon into the gas phase. Once in the gas phase in the post-shock region, ion-molecule reactions or shock chemistry (Bachiller and Gómez-González 1992) may convert atomic silicon into SiO. In evolved outflows (kinematic timescales > a few 10⁴ years), the SiO emission is quite weak or even undetected; this may be due to the depletion of SiO onto dust grains. However, theoretical modelling by Mackay (1995, 1996) suggests that the fractional abundance of SiO may gradually increase over a million year timescale in these regions through reaction of silicon hydrides with molecular hydrogen and oxygen.

Millimeter wavelength SO emission is often found in quiescent molecular cores (Chernin et al. 1994b) where the density is high enough (> 10⁵ cm^-3) for collisional excitation. However, the spatial distribution of SO emission can be quite different from that of tracers of the dense gas such as CS and C₃H₂ (Swade 1989), indicating either unusual excitation or strong abundance variations. Like SiO, SO emission has also been found in protostellar outflows from massive star forming regions such as Orion-IRc2 and Ceph-A (Martín-Pintado et al. 1992) where the SiO and SO abundances are enhanced by orders of magnitude over that in quiescent material. Observations of the NGC 2071 outflow show unusually strong high velocity SO and SiO emission and that the spatial and velocity distributions of SiO and SO emission are quite different from that of CO (Chernin and Masson 1992). The differences in SiO and SO relative to CO are likely due to abundance enhancements rather than excitation differences. Chernin, Masson and Fuller (1994b) observed SiO v=0, J=5 -- 4 and SO J_N=6₅ -- 5₄ emission towards nine regions containing protostellar molecular outflows. They found that the SiO and SO line shapes were similar, but were quite different from the CO profiles.

By studying UC HII regions, an early phase of massive star evolution, we learn more about one of the primary evolutionary forces in galaxies. UC HII regions are among the youngest objects in the Galaxy and are still deeply embedded in the molecular clouds where they formed. This allows us to study their environment by observing molecular transitions in the centimeter and millimeter wavelength range. Observations of a selection of molecular lines has helped researchers to understand the dynamics of UC HII regions and to determine physical parameters such as density, temperature, mass, column density, and composition in the dense molecular cloud cores where the exciting stars of the HII regions form (e.g. Cesaroni et al. 1994, 1992, 1991; Churchwell et al. 1992, 1990).

We present here a three-part study of fast gas in high mass star forming regions. The first part (Chapter 2) is a multi-transition survey of SiO emission towards UC HII regions to search for possible molecular outflows. Chapter 3 presents an in-depth look at the source G5.89-0.39, a young massive star embedded in its natal molecular cloud, associated with one of the most extraordinary molecular outflows known. In combination with ammonia and carbon monoxide observations, we determine the physical properties of the outflow. Finally, in Chapter 4 we determine the expansion rate of the ionized shell around G5.89-0.39 from VLA observations that span 5.3 years. The inferred dynamical age makes this one of the youngest nebulae associated with high mass star formation known. Combined with a model for the nebular emission and H radio recombination line spectra, we determine the distance to the source.

References

• Bachiller, R., & Gómez-González, J. 1992, Astron. & Astroph. Rev., 3, 257
• Bachiller, R., Cernicharo, J., Martín-Pintado, J., Tafalla, M., & Lazareff, B. 1990, Astron. & Astroph., 231, 174
• Cesaroni, R., Churchwell, E., Hofner, P., Walmsley, C. M., & Kurtz, S. 1994, Astron. & Astroph., 288, 903
• Cesaroni, R., Walmsley, C. M., & Churchwell, E. 1992, Astron. & Astroph., 256, 618
• Cesaroni, R., Walmsley, C. M., Kömpe, C., & Churchwell, E. 1991, Astron. & Astroph., 252, 287
• Chernin, L. M., Masson, C. R., Gouveia Dal Pino, E. M., & Benz, W. 1994a, Astroph. J., 426, 204
• Chernin, L. M., Masson, C. R., & Fuller, G. 1994b, Astroph. J., 436, 741
• Chernin, L. M., & Masson, C. R. 1992, Astroph. J., 396, L35
• Choi, M., Evans, N. J., & Jaffe, D. T. 1994, Astroph. J., 417, 624
• Churchwell, E., Walmsley, C. M., & Wood, D. O. S. 1992, Astron. & Astroph., 253, 541
• Churchwell, E., Walmsley, C. M., & Cesaroni, R. 1990, Astron. & Astroph. Supp., 83, 119
• Garay, G., Rodríguez, L. F., Moran, J. M., & Churchwell, E. 1993, Astroph. J., 418, 368
• Hartquist, T. W., Oppenheimer, M., & Dalgarno, A. 1980, Astroph. J., 236, 182
• Kurtz, S., Churchwell, E., & Wood, D. O. S. 1994, Astroph. J. Supp., 91, 659
• Kwan, J., & Scoville, N. Z. 1976, Astroph. J., 210, L39.
• Margulis, M., & Snell, R. L. 1989, Astroph. J., 343, 779
• Martín-Pintado, J., Bachiller, R., & Fuente, A. 1992, Astron. & Astroph. , 254, 315
• Masson, C. R., & Chernin, L. M. 1993, Astroph. J., 414, 230
• Mitchell, G. F. 1984, Astroph. J., 287, 665
• Pineau des Forêts, G., Roueff, E., Schilke, P., & Flower, D. R. 1993, Mon. Not. Royal Astron. Soc., 262, 915
• Raga, A. C., Bohm, K. H., & Solf, J. 1986, Astron. J., 92, 119
• Shepherd D. S., & Churchwell E. 1996a, Astroph. J., 457, 267
• Shepherd D. S., & Churchwell E. 1996b, Astroph. J., 472, 225
• Shu, F., Najita, J., Ostriker, E., Wilkin, F., Ruden, S., & Lizano, S. 1994, Astroph. J., 429, 781
• Stahler, S. W. 1994, Pub. Astron. Soc. Pacific, 106, 337
• Tafalla, M., Bachiller, R., & Wright, M. C. H. 1994, Astroph. J., 432, L127
• Wood, D. O. S., & Churchwell, E. 1989, Astroph. J. Supp., 69, 831
• Zuckerman, B., Kuiper, T. B. H., & Kuiper, E. N. R. 1976, Astroph. J., 209, L137