Astron. Astrophys. 325, 725-744 (1997)

1. Introduction

The aim of this work is to study the morphology and dynamics of the molecular gas immediately around a (proto)star in its earliest evolutionary phase.

On an large scale (arc minute angular resolution or pc linear scale) the basic ingredients that occur in a star forming region/molecular cloud complex are already well known: cool extended molecular clouds with hotter, unresolved and denser clumps, molecular outflows, jets observed in H₂, masers (in particular H₂ O), ultracompact (UC) HII regions, cool dust envelopes emitting in the far infrared (FIR) and hotter dusty disks emitting in the near infrared (NIR).

However, given the complexity of these regions, observations with arc minute resolution are totally inadequate to study what occurs around the (proto)star itself, assuming one knows where the (proto)star itself is located. Does the outflow morphology remains the same as a function of distance from the (proto)star? What is the connection between outflows and masers? Is there an accretion disk around the (proto)star? What fraction of the molecular cloud really takes part in the formation of the (proto)star? At which evolutionary stage do the various features (maser, outflow, HII, etc.) occur? Many different star forming events are often present in the same star forming complex and their features are blended when observed with large beams, complicating the interpretation.

Only the comparison between arcsec resolution observations of an isolated (proto)star can disentangle the morphology and dynamics of the molecular gas directly involved in star formation from that of the larger scale surrounding molecular cloud.

With the intention of choosing an object in an early evolutionary phase, we have selected for our study the surroundings of the H₂ O maser source IRAS 20126+4104 (Tofani et al. 1995, hereafter TFTH), which is associated with a strong bipolar molecular outflow, extended in the N-S direction (Wilking et al. 1990; hereafter WBM). In the following we shall use the IRAS name for the source, although, as we shall show, nomenclature itself becomes a problem when we want to use an arcmin beam originated nomenclature, such as IRAS, for arcsec structures. IRAS will undoubtly be able to give the total luminosity of the dust cloud (and hence of the stars) within the beam, but it will not be able to decide between several possible "stellar" candidates separated by few seconds of arc.

On which grounds the choice of IRAS 20126+4104 satisfies our goals?

We believe that H₂ O masers represent an excellent tracer for newly formed massive (proto)stars and, in particular, the maser without associated radio continuum may be in an earlier phase than those at the edge of an UC HII region. Such an idea is based on two observational results:

1. In several cases H₂ O maser spots are not associated with an UC HII region - although the Lyman continuum of the (proto)star is amply sufficient to ionise an HII region - as shown by observations with high spatial resolution (): these do not detect any continuum at 1.3 cm or 3.6 cm towards the H₂ O masers location, down to a level of 0.2 mJy (TFTH), namely the UC HII region (if at all present) is so small and dense (hence young) that it is strongly self absorbed;
2. in several objects the H₂ O maser spots show an excellent positional agreement with hot and dense molecular clumps revealed in transitions like HCN(1-0) (Turner & Welch 1984), CH₃ CN(5-4) (Wink et al. 1994), NH₃ (4,4) (Cesaroni et al. 1994a), or CH₃ CN(6-5) (Cesaroni et al. 1994b), but are well separated from the nearby UC HII regions.

We thus conclude that H₂ O masers are closely related to massive stars, but may be formed even before the development of an UC H II region.

With this in mind, we have used the 30-m telescope to observe several molecular transitions towards a selected sample of 12 H₂ O maser sources that observations with high spatial resolution in the maser line and in the radio continuum (TFTH) have shown to be well separated () from the closest HII region. Our survey (Cesaroni et al. in prep.) was extremely successful, proving that all our targets are associated with molecular clumps and are centred on the H₂ O maser spots, as expected. Results on S235 A-B are reported in Felli et al. (1996). IRAS 20126+4104 was mapped in ¹³ CO(2-1), HCO (1-0), HCN(1-0), CH₃ OH(3-2) and (5-4), C³⁴ S(2-1), (3-2) and (5-4), CH₃ CN(8-7) and (12-11), and CS(3-2) and found to have an unresolved peak at the maser position.

Subsequently, the source was observed with the Plateau de Bure interferometer (PdBI) in molecular lines that can trace the high density molecular gas, such as CH₃ CN, in lines that are sensitive to more extended material, such as , and in the 3.3 mm continuum.

An association between NIR and H₂ O maser sources was searched recently by Testi et al. (1994) in 17 masers from the list of Forster & Caswell (1989), using high resolution NIR images. At least one NIR source was found within from maser components. These newly discovered sources are faint in K-band (12-14 mag) and usually undetected in J and H images, namely they have a large NIR excess implying the presence of a hot dust envelope. Testi et al. (1994) associate the very red objects with the maser spots, and suggest that they may represent the direct manifestation of the (proto)star or of its interaction with the region immediately around it. We present here NIR observations of IRAS 20126+4104 taken with ARNICA (Lisi et al. 1996) in J, H, and K and in the H₂ vibrational line with different resolutions, using the TIRGO and NOT telescopes.

In Sect. 2 we describe the observations and the data reduction techniques. Sects. 3 and 4 are devoted to the description of the results and to their interpretation, while in Sect. 5 we outline a model capable of explaining all the observed features in terms of a disk-outflow system. In Sect. 6 a comparison with similar objects is discussed. Finally, in Sect. 7 the conclusions are drawn.

© European Southern Observatory (ESO) 1997

Online publication: April 28, 1998

helpdesk.link@springer.de