The NGC 6334 region, located at a distance of 1.74 kpc from the Sun (Neckel 1978), is a complex of radio and infrared sources (Shaver & Goss 1970; Haynes et al. 1978; McBreen et al. 1979). One of its components, NGC 6334 I(N), is also the strongest source in the NH3(J,K) = (1,1) transition (e.g., Schwartz et al. 1978). It is also a strong continuum source at 400, 1000 and 1300 µm (Cheung et al. 1978; Gezari 1982; Loughran et al. 1986) but it is below the detection limit at wavelengths smaller than 137 µm. It has an H2O maser source, CH3OH masers (Moran & Rodríguez 1980; Kogan & Slysh 1998; Walsh et al. 1998), six reddened sources at the JHK bands (Tapia et al. 1996), a bipolar molecular outflow in SiO and eight knots of H2 emission (Megeath & Tieftrunk 1999). However, no HII region associated to this source was detected.
The source NGC6334 I, located approximately 2´ south of NGC 6334 I(N), has H2O, OH and CH3OH masers (Moran & Rodríguez 1980; Gaume & Mutel 1987; Forster & Caswell 1989; Menten & Batrla 1989), an ultra-compact HII region (Rodríguez et al. 1982; DePree et al. 1995) and a cluster of stars detected at the JHK bands (Tapia et al. 1996). It was studied with high angular resolution in the NH3(J,K) = (1,1) transition by Jackson et al. (1988) (hereafter, JHH88), who interpreted the observed structure as a circumstellar molecular disk rotating around a 30 O star. On the other hand, CO and H2 observations suggest that the structure arises from a molecular bipolar outflow (Bachiller & Cernicharo 1990; Persi et al. 1996). Due to these characteristics, it is probable that this region is in a more advanced stage of stellar formation than NGC 6334 I(N).
Both regions have been studied in several transitions of 12CO, 13CO and CS (Kraemer & Jackson 1999), the meta-stable transitions of the ammonia molecule, including NH3(J,K) = (1,1), (2,2), (3,3) and (6,6) [e.g., Schwartz et al. 1978; Forster et al. 1987; Vilas-Boas et al. 1988; Jackson et al. 1988; Kuiper et al. 1995; Kraemer & Jackson 1995, 1999], and the non-meta-stable transition NH3(J,K) = (2,1) (Kuiper et al. 1995). The physical parameters of these regions were always derived from the NH3(J,K) = (1,1) and (2,2) observations, assuming LTE conditions. In fact, the intensity anomalies in the nuclear quadrupole hyperfine structure of the NH3(J,K) = (1,1) spectrum are usually neglected, although they are observed in many warm molecular clouds (Stutzki et al. 1982; Batrla et al. 1983; Stutzki et al. 1984).
Matsakis et al. (1977) interpreted the anomalous spectrum as the consequence of non-thermal population in the hyperfine states, induced by selective trapping in the hyperfine transitions of NH3(J,K) = (2,1) (1,1). This effect is relevant only when the width of the hyperfine lines lie between 0.3 and 0.6 km s-1. Due to this restriction, they assumed that the molecular cloud is formed by clumps, each one producing a narrow line spectrum, so that the observed spectrum would be the superposition of individual clump spectra. Based on this model, Stutzki & Winnewisser (1985) (hereafter, SW85) elaborated a numerical algorithm to calculate the brightness temperature ratios between the hyperfine satellites and the main line in the NH3(J,K) = (1,1) spectrum, and also the brightness temperature ratio between the (2,2) to the (1,1) main lines, as a function of the NH3 column density, H2 density and kinetic temperature of each clump.
Gaume et al. (1996) questioned this model, based on observations of the NH3(J,K) = (1,1) absorption spectra towards the continuum source DR21. Their argument was based on an anti-correlation in the degree of LTE departure of the inner and outer hyperfine components in different regions of the cloud. They proposed instead, inflows and outflows of matter to explain the observations, but did not present any quantitative calculation to support their claim.
In this work, we used our high quality observations of the NGC 6334 region in the NH3(J,K) = (1,1) transition, which allowed the precise determination of the intensity anomalies in the hyperfine structure, to obtain with good reliability its physical parameters assuming non-LTE conditions. The data allowed us also to separate in the spectra the contribution of at least three different sources: NGC 6334 I, NGC 6334 I(N)w and NGC 6334 I(N)e.
© European Southern Observatory (ESO) 2000
Online publication: October 2, 2000