Searching for massive protostars

Introduction

Massive protostars are hard to find because the Kelvin-Helmholtz timescale is shorter than the free-fall timescale for massive stars which means that massive stars begin burning hydrogen before they have finished collapsing. Also massive star-forming regions are typically more distant than their commonly-observed low-mass counterparts and massive stars form in clusters making confusion a severe problem.

Ultracompact HII regions and hot molecular cores are well-known phases in the early evolution of massive stars, but they are phases where a central compact object has already formed and is pouring energy into its surroundings. What about stages prior to this point? Can we detect and identify massive pre-protostellar cores?

A JCMT-BIMA-JCMT survey

An initial SCUBA survey by Thompson et al. (2002) of ultracompact HII regions (UCHIIs) from the Wood & Churchwell (1989) and Kurtz et al. (1994) radio surveys revealed a number of interesting sources with extended dust emission and/or multiple dust peaks within the SCUBA field of view. We have since followed up these observations with a survey of 10 UCHIIs using the 10-element BIMA interferometer at 3 mm (roughly 10-arcsec resolution). We employed a variety of tracers to simultaneously probe the quiescent and highly-excited gas.

The BIMA survey was spectacularly successful: observations of N2H+ 1-0 in all 10 sources resulted in the detection of approximately 50 cores, the majority of which were not known of previously. More importantly, most of these cores have no known infrared (IRAS or MSX) or radio sources within them, nor other signs of massive star formation such as water and methanol masers. Even more intriguing is the fact that a number of these cores are weak in C18O 1-0 emission, suggesting that CO molecules may be freezing out, and therefore that the core may be collapsing (Bergin & Langer 1997; Rawlings et al. 1992).

Such discoveries had us back at the JCMT for further followup observations to confirm our depletion interpretation (although we are unable to say as yet), search for outflows associated with some of the warmer cores we found in our BIMA results and look for kinematic signatures of collapse.

New outflows

Thus far we only have CO data on G35.20-1.74 but it was obviously a good place to start as we have already discovered a new outflow and evidence for CO outflow associated with a known maser source. Figure 1 shows high-velocity red (yellow contours) and blue shifted (blue contours) CO 3-2 emission in G35.20-1.74. The large open square marks the position of the UCHII region, the small open square the position of the methanol and water masers and the filled circle one of the submillimetre peaks from our SCUBA maps. The compact outflow at the west edge of the image is centred on a molecular peak bright in N2H+, C18O and CH3CN indicating the likely presence of a hot-core-like source. The maser positions also appear to be at the centre of a second CO outflow.

The promise of wide-field imaging

Some of the fields we observed (such as G29.96-0.02) were so rich in new cores that we decided to extend the mapping to include a larger area to search for further sources. Figure 2 shows a comparison of the BIMA and SCUBA images for G29.96-0.02. The agreement between the 850-micron dust emission and the 1-0 N2H+ emission is amazingly good, even out to the east where a new ridge of dust emission was revealed by our SCUBA scan-map (marked `East core' in Figure 2). No radio or infrared sources are observed towards this core; neither are any masers known. We estimate a mass of ~8500 Msun for this core.

The BIMA N2H+ data are superimposed on the MSX 8-micron image of the same region. It is interesting to note that the MSX data show no signs of point sources within any of the newly-discovered molecular/dust cores. What is even more interesting is that the eastern dust ridge shows up towards a region which is essentially dark at 8 microns. This is a very good indicator that the high dust column density we measure at this position with SCUBA is due to *cold* dust and that here may be one site where the next generation of massive stars will form.

Summary and future work

Our BIMA results have shown that N2H+ is an excellent tracer of the dense gas in regions of massive star formation. It is such a pity that RxA3 cannot tune to the 3-2 line and that it is difficult to get the 4-3 line with RxB3 as they would undoubtedly be excellent transitions for constraining the properties of the gas traced by the 1-0 line. We are working on determining whether or not we are seeing signs of depletion and are continuing the JCMT study to look for outflows and collapse signatures.

References

Bergin E.A., Langer W.D., 1997, ApJ, 486, 316
Kurtz S.E., Churchwell E., Wood D.O.S., 1994, ApJS, 91, 659
Rawlings J.M.C., Hartquist T.W., Menten K.M., Williams D.A., 1992, MNRAS, 255, 471
Thompson M.A., Hatchell J., Macdonald G.H., Millar T.J., 2002, in preparation
Wood D.O.S., Churchwell E., 1989, ApJS, 69, 831

Click here for printable version.