Astron. Astrophys. 327, 299-308 (1997)

1. Introduction

Radio ultra-compact HII regions (UC HII), H₂ O masers and infrared (IR) sources with IR excess are the trademarks of star forming regions (SFRs). However, although they are all present in the same SFR, they may not be necessarily physically related to the same YSO. In fact, new indications from high resolution observations suggest that each of them may represent a different and independent YSO and that the possible associations between them, that occur in some cases, may simply reflect the fact that each type of emission may overlap in time with the others during the evolution of a YSO.

In particular, there is now increasing evidence, coming from the comparison of arcsec resolution radio and IR observations, that the conditions required to produce strong water maser emission occur in an evolutionary phase prior to the formation of an UC HII region. In fact, when observed in the radio at high resolution and sensitivity, many masers in SFRs do not show any associated continuum source or are not directly related to UC (or more extended) HII regions also present in the same area (Tofani et al. 1995, Hunter et al. 1995, Palla et al. 1995, Felli et al. 1997, Persi et al. 1996, Hofner & Churchwell, 1996, Codella et al. 1997). Similarly, submillimeter continuum observations in the vicinity of H₂ O masers (Jenness et al. 1995) also show that the majority of the sub-mm sources detected are directly related to the water maser and not to the UC HII regions. This further supports the association of H₂ O masers with the earliest manifestation of a YSO, i.e. when it is detectable only as a moderately warm (25 - 50 K) and luminous dust core. Finally, molecular observations in high density tracers like CS, NH₃, CH₃ CN confirm that the masers are found associated with compact and dense molecular blobs, not necessarily coincident with UC HII regions (Turner & Welch, 1984, Wink et al. , 1994, Cesaroni et al. 1994, Codella et al. 1997).

Forster & Caswell (1989), observing with VLA a sample of OH and H₂ O masers associated with 74 star forming regions at arcsec resolution, found that the maser sources occur near a common center but in physically distinct zones and that there was no evidence that the centers were coincident with the continuum peaks. Testi et al. (1994), imaging in the near-IR a sub-sample of these sources, discovered very faint K sources (all with a strong IR excess) within few arcseconds from the H₂ O masers. They concluded that the maser activity is characteristic of the youngest phases of the life of a YSO and that the near IR emission comes from an associated YSO, detectable only in the IR and not in the radio continuum. The IR excess can be due to a hot circumstellar dust envelope around the YSO, free-free and bound-free emission from ionized gas or photospheric stellar radiation (either direct or reflected). In any case, the close association of an IR source with the H₂ O maser is the best proof of a close-by YSO responsible for its excitation.

Near IR emission in SFRs may be also associated with diffuse HII regions. In this case it is usually extended and comes from the contribution of the ionized gas, with possible weaker contributions from reflected starlight and from the near IR tail of the dust emission (Natta & Panagia 1976).

In summary, the fact that UC HII regions, OH and H₂ O masers and near IR sources are found in the same region may merely reflect the formation of a stellar cluster inside a molecular cloud and does not imply an unique YSO responsible for all the emissions. However, in order to establish the true associations between the different types of emission as well as their occurrence during the lifetime of a YSO, high resolution radio and near IR observations are required. Mid infrared observations with similar resolutions are also essential to better define the spectral energy distribution (SED) of each source.

To achieve this goal we have undertaken a project of near and mid-infrared images at arcsec resolution of a selected sample of H₂ O masers from the list of Forster & Caswell (1989). In the present work, the results relative to the near and mid-IR images obtained with similar sub-arcsec resolution of the SFR known as G 35.20-1.74 are discussed. This nomenclature usually refers to the UC HII region; we shall use it in a broader sense referring to the entire SFR-molecular cloud complex. We shall also compare our observations with radio continuum, molecular and sub-mm observations taken from the literature.

G 35.20-1.74 is the most compact (angular diameter 4 , Woodward et al. 1985) of the W48 cluster of HII regions (W48A in the nomenclature of Onello et al. 1994). When observed with even higher resolution ( 1 ) the radio continuum morphology is that of a cometary UC HII region (Wood & Churchwell 1989, hereafter WC). OH and H₂ O masers, discovered from this source with single dish observations (Turner 1979; Genzel & Downes 1977) were found to be located at angular separations 20 to the N-W of the UC HII region (Forster & Caswell 1989, Hofner & Churchwell 1996). In the following we shall use the positions of the two H₂ O masers reported by Hofner & Churchwell (only one maser was reported by Forster & Caswell), obtained with a resolution a factor 3 better than that of Forster & Caswell, and call them HC1 ( (1950) = , = ) and HC2 ( (1950) = ^s, (1950) = ). Previous infrared studies (Dyck & Simon 1977, Zeilik & Lada 1978, Moorwood & Salinari 1981), detected only a single bright source with a luminosity between 1 and 33 of . Given the large diaphragms used ( 20 ) it was not possible to discriminate between the UC HII region and the maser. More recently, a new near-IR image of the region, besides revealing the structure of the diffuse gas around the UC HII region and of the UC HII region itself, also showed a weak source detectable only at K (13.1 mag) close to the HC1 maser (Testi et al. 1994). The UC HII region and the masers are located close to an extended molecular cloud. Vallée & MacLeod (1990) estimated a total mass for the molecular cloud of 1500 from CO observations. The presence of higher density molecular blobs within the molecular cloud at the position of HC1 is suggested by observations in other molecules, such as CS(J=1-0) and NH₃ (Anglada et al. (1996), Churchwell et al. 1990, Churchwell et al. 1992), even though in all cases the resolution is not adequate to separate independent components on the scale size of the UC HII region-HC1 separation. The visual extinction towards the complex has been derived from CO observations (Vallèe & MacLeod 1990) and from the Br /Br ratio (Woodward et al. 1985); both determinations give A_V 23. A distance between 3.2-3.4 kpc was assigned by Bridle & Kesteven (1972); in the following we shall use the more recently adopted value of 3.1 kpc (see e.g. Churchwell et al. 1990).

© European Southern Observatory (ESO) 1997

Online publication: April 8, 1998
helpdesk.link@springer.de