Infrared and radio observations of the earliest stages of massive star formation

Abstract

It is assumed that objects which will eventually evolve into high-mass main sequence stars can initially appear as low- to intermediate-mass young stellar objects that are still accreting mass. By initiating a Spitzer mid-infrared spectroscopic survey of intermediate mass young stellar objects, selected from massive star-forming clumps and infrared dark clouds (IRDCs), an evolutionary sequence was established. A group of particularly young stellar objects (YSOs) was found showing several ice and silicate absorption features. A second group was identified: these objects represent more evolved YSOs that already emit a significant amount of UV radiation and feature shock-excited gas driven by proto-stellar outflows. Importantly, the observed extended emission of cationic polycyclic aromatic hydrocarbons (PAHs) and the analysis of molecular hydrogen indicates the creation of a very young photon dissociation region while the expected radio emission from the associated HII region is still undetectable. An automated photometric pipeline was developed to detect cold dense cores and extract their far-infrared fluxes in Herschel bolometer maps while taking into account the complicated background and additional instrumental effects. By studying the fragmentation in the high-mass part of the Herschel EPoS program a typical point source separation of ~ 0.5 pc was found throughout the sample. The detected sources are in an evolutionary stage where they are embedded in the dust clumps associated with the parental cloud. This typical source separation is not retained in later stages, such as in young stellar clusters. Furthermore, a comprehensive case study is presented in which an isolated IRDC region is used to analyze the conditions of massive star formation in the absence of strong external effects. Two point sources found in this region are candidates for evolving into high-mass main-sequence stars. The filamentary star-forming cloud is in free collapse while the embedded molecular clumps are gravitationally bound. The observed dust temperature structure shows that the dark cloud is not isothermal. The locations of temperature minima are in good spatial agreement with the column density peaks. Together with the structure of molecular clumps identified by radio line observations these results confirm that the dust temperatures are lowest in the densest parts of the dark cloud.