Molecular Counterparts of Ultracompact HiiRegions with Extended Envelopes

Abstract

We have carried out 13CO J = 1-0, CS, and C34S J = 2-1 and J = 3-2 line observations of molecular clouds associated with 16 ultracompact (UC) H II regions with extended envelopes. The molecular clouds are the ones that give birth to rich stellar clusters and/or very massive (O7-O4) stars. Our data show that the clouds are very clumpy and of irregular morphology. They usually have much larger masses, velocity dispersions, and fractions of dense gas than molecular clouds that form early B or late O stars. This is compatible with earlier findings that more massive stars form in more massive cores. The IR luminosity-to-mass ratio has a mean value of 9 L☉/M☉ and is little correlated with the cloud mass. Most molecular clouds have star formation efficiencies of 1%-2%. We find size-line width and size-density relations in the forms of Δv ∝ D0.4 and n(H2) ∝ D-1.2. 13CO cores are in general associated with compact H II regions regardless of the presence of UC H II regions therein. In contrast, CS cores are preferentially associated with compact H II regions that contain UC H II regions. As with the fact that the compact H II regions containing UC H II regions are more compact than those not associated with UC H II regions, these indicate that the former may be in an earlier evolutionary phase than the latter. The diffuse extended envelopes of H II regions often develop in the direction of decreasing molecular gas density. Based on detailed comparison of molecular line data with radio continuum and recombination line data, the extended ionized envelopes are likely the results of champagne flows in at least 10 sources in our sample. Together these results appear to support a published suggestion that the extended emission around UC H II regions can be naturally understood by combining the champagne flow model with the hierarchical structure of molecular clouds, taking into account various inclinations and low resolutions of our data. In addition, the blister model seems to be still applicable to most H II regions, even though massive stars usually form in the interiors rather than on the surfaces of molecular clouds. This is possible because massive star-forming clouds have hierarchical structure and irregular morphology.