This is the second paper in our series on high-resolution infrared observations of ultracompact HII regions (UCHII s). The first paper (Feldt et al. 1998, hereafter Paper I) concentrated on the object G45.45+0.06. This time G5.89+0.39 (hereafter denoted as G5.89) was chosen for two reasons. Firstly, it met our selection criterion which demands that a nearby bright star exists to serve as a wavefront sensor for the adaptive optics system. Secondly, it is one of the best-studied and yet still disputed UCHII s.
G5.89 was classified as a shell-like UCHII by Wood & Churchwell (1989, hereafter WC89). Its kinematic distance of 2.6 kpc from the sun was determined by Downes et al. (1980) and little disputed since. Although recent estimates by Acord et al. (1998) put this distance close to the possible upper limit, we will use it throughout the paper.
Apart from the VLA observations by WC89 which show the shell-like morphology, Gomez et al. (1991) observed the source with the same instrument in D configuration at 1.3 cm. To them, the source appears essentially unresolved, with a weak extension towards the southeast. Mid-infrared (MIR) observations by Ball et al. (1992) showed that the MIR emission closely resembles the radio structure, apart from the southern half of the shell, where no MIR emission was detected. They regarded this phenomenon to be due to an off-centre location of the ionizing star. Models of the source by Churchwell et al. (1990, hereafter CWW) and recently by Faison et al. (1998) showed that the spectral energy distribution (SED) of the source can be fitted by emission from a spherically symmetric cocoon of dust surrounding a central zero-age main-sequence (ZAMS) star of spectral type O6. To fit near-infrared (NIR) part of the spectrum, a dust-free cavity inside the cocoon was required. The radial intensity profiles of their models already predicted the presence of a ring of emission at all wavelengths. A similar model was proposed by Harvey et al. (1994). However, the latter authors regard their own spherical model as highly unlikely due to the obvious non-spherical morphology of the source.
Table 1. Summary of observations
G5.89 is also known to drive one of the most energetic outflows in the Galaxy. A lot of discussion has been going on about the orientation of this outflow and whether there are multiple outflows and driving sources (Harvey & Forveille 1988; Zijlstra et al. 1990; Cesaroni et al. 1991; Acord et al. 1997).
In this paper, we are going to present the result of a multi-wavelength imaging campaign of the source. High-resolution observations at 1.6 µm , 2.2 µm , and 3.5 µm taken with ESO's adaptive optics (AO) system ADONIS are complemented by MIR data with arcsecond resolution taken at 10.6 µm , 11.7 µm , 12.8 µm , and 21 µm . These observations show that the shell-like structure of the source is preserved throughout this whole wavelength range. Additionally, large-scale narrow-band images were obtained in the and Br lines. These data were used to determine the extinction towards the source as well as to get a large-scale overview of the region. They indicate the presence of a molecular cloud in front of G5.89 which is obscuring about half the source from view even at infrared wavelengths. The presence of this cloud is confirmed by our continuum map taken at 1.3 mm using the SEST.
We have collected infrared data of arcsecond resolution over such a wide range in wavelengths on this source for the first time. The NIR images even are of a resolution of 0:004 (full-width half-maximum (FWHM) of the point-spread function (PSF) in ) thanks to the adaptive optics. We use these data to build a consistent model of the source. We will argue that the spherical dust shell model is essentially correct and present a detailed fit not only to the SED, but also to the intensity distribution and various properties of the source. This model is consistent with all our observations and explains G5.89 as an almost classical Strömgren sphere. We interpret the visible break in spherical symmetry as channel openings, through which an outflow is escaping in north-south direction. This outflow is detected in our data as emission features. From data on a larger scale, we can show not only the presence of a large cloud of cold dust at the rim of which G5.89 is located, but also identify several stellar sources which belong to the complex. Thus, like in Paper I, we are probably looking at a young cluster in formation. Due to the smaller distance compared to the 6.3 kpc of G45.45-0.06, we are now able to look at the single source G5.89 inside this cluster and describe it in detail. This is another hint that classical UCHII s with spherical configurations exist, but that they are usually part of young stellar clusters which will eventually form a larger ionized region.
© European Southern Observatory (ESO) 1999
Online publication: May 6, 1999