In order to investigate whether massive stars form similarly to their low-mass counterparts, we have used the standard envelope plus disc geometry successfully applied to low-mass protostars to model the near-IR to sub-millimetre SED and several mid-
IR images of the embedded massive star IRAS20126+4104. We have used a Monte Carlo radiative transfer dust code to model the continuum absorption, emission and scattering through two azimuthally symmetric dust geometries, the first consisting of a rotationally flattened envelope with outflow cavities, and the second which also includes a flared accretion disc. Our results show that the envelope plus disc model reproduces the observed SED and images more accurately than the model without a disc, although the latter model more closely reproduces the morphology of the mid-IR emission within a radius of 1.1 or ~1800au. We have put forward several possible causes of this discontinuity, including inner truncation of the disc due to stellar irradiation, or precession of the outflow cavity. Our best fitting envelope plus disc model has a disc radius of 9200 au. We find that it is unlikely that the outer regions of such a disc would be in hydrostatic or centrifugal equilibrium, however we calculate that the temperatures within the disc would keep it stable to fragmentation.
We have conducted a search for ionized gas at 3.6 cm, using the Very Large Array, towards 31 Galactic intermediate- and high-mass clumps detected in previous millimeter continuum observations. In the 10 observed fields, 35 HII regions are identified,
of which 20 are newly discovered. Many of the HII regions are multiply peaked indicating the presence of a cluster of massive stars. We find that the ionized gas tends to be associated towards the millimeter clumps; of the 31 millimeter clumps observed, 9 of these appear to be physically related to ionized gas, and a further 6 have ionized gas emission within 1. For clumps with associated ionized gas, the combined mass of the ionizing massive stars is compared to the clump masses to provide an estimate of the instantaneous star formation efficiency. These values range from a few percent to 25%, and have an average of 7 +/- 8%. We also find a correlation between the clump mass and the mass of the ionizing massive stars within it, which is consistent with a power law. This result is comparable to the prediction of star formation by competitive accretion that a power law relationship exists between the mass of the most massive star in a cluster and the total mass of the remaining stars.
