1994A&A...288..903C

We present VLA C- and D-array measurements of NH₃(4,4) towards a sample of four galactic ultracompact HII regions: G9.62+0.19, G10.47+0.03, G29.96-0.02, and G31.41+0.31. Our data shows that the ammonia emission originates from optically thick clumps with kinetic temperature of 50-200K and size ∼0.1pc. The hot ammonia clumps are situated close to the compact HII regions seen in the radio continuum but in three out of four cases (G10.47+0.03 is the exception) are not absolutely coincident with them. Moreover, there is good positional coincidence between the water masers observed in these regions and the hot ammonia clumps. This suggests to us that the ultimate energy source responsible for the hot ammonia is not the ionizing star of the compact HII region but is a separate young massive star. We have made estimates of the luminosity required for such an embedded object and derive values of order 10⁴ to 10⁶ solar luminosities although with very large uncertainties. We have also made estimates of the masses of the hot ammonia clumps under the assumption that the virial theorem is applicable. In this way, we obtain masses of typically 100M_☉, densities of ∼10⁷cm^–3 and NH₃ abundances relative to H₂ of >10^–6. Such high ammonia abundances are probably caused by evaporation of grain mantles due to the high temperature in these cores. The dynamical state of these ammonia clumps is not clear but we detect a blue-shifted absorption line towards the ultracompact HII region G10.47+0.03B which suggests that in this case, the predominant motion is expansion. We conclude in general that such clumps are likely to be the sites of massive star formation in an early evolutionary phase prior to the development of an ultracompact HII region.