1 Introduction

Outflows play a fundamental role in star formation. Excess angular momentum in the protostellar system is removed by the outflow, the star-forming core of dust and gas is disrupted by the outflow, possibly limiting the mass of the resulting star, and the interaction of the outflow with the natal molecular cloud may have far-reaching effects on subsequent star formation within the cloud.

Outflows from low mass young stellar objects (YSOs) have been well studied to date; for example the reviews by Bachiller (1996) and Fukui (1993) and references therein. Molecular outflows and high-velocity ionised jets have been observed toward low mass YSOs. The molecular outflows are thought to be driven by the jets, with the momentum from the jet transferred to the molecular material by entrainment (either prompt entrainment by bow shocks at the head of the jet or steady state entrainment by turbulent mixing along the sides of the jet, De Young 1986). There is a strong relationship between the mechanical force and luminosity of the molecular outflow and the bolometric luminosity, particularly the 6cm continuum emission (Cabrit & Bertout 1992) of the central protostar driving the outflow, which suggests a close association between ionised and molecular gas. The driving force behind the bipolar ionised jets is not yet clear. Current theories suggest the jets are wind-driven, possibly by either a disc-collimated wind (Pelletier & Pudritz 1992) or from boundary layer effects in the protostellar accretion disc (Shu et al. 1995).

The overall picture of massive YSO outflows is much less certain. The large distance of most high mass star forming regions makes it difficult to resolve fine detail and the higher degree of turbulence present in massive star forming cores confuses the location of high-velocity emission from outflows. It is important though to determine the properties of massive YSO outflows to place them in the context of the models developed for low mass YSO outflows and to test the predictions of the models against the properties of high mass YSO outflows. Until recently only a few isolated examples of massive stellar outflows were known (e.g. G5.89-0.39, DR 21 and G45.12). Studies by Shepherd & Churchwell (1996a,b) have revealed high velocity CO emission present in 90$\%$ of a survey of massive YSOs and have determined the properties of five outflows by mapping a sample of sources with high velocity emission from the survey.

One of the most important aspects of this recent work on massive YSO outflows has been the confirmation that massive YSO outflows lie at the more extreme end of the scale. The mass contained in massive YSO outflows is at least an order of magnitude higher than that of most YSO outflows. The flow rates of massive YSO outflows are also in general larger than their low mass counterparts. If both outflow phenomena share a common origin, massive YSO outflows thus represent an ideal test of the models developed for low mass YSO outflows. Doubts have been raised whether disc-collimated wind or boundary-layer (so-called X-wind models) models of outflows are capable of driving the large masses in massive YSO outflows (Churchwell 1997), although the massive bipolar outflow associated with G192.16 exhibits properties of both disc-collimated and X-wind models (Shepherd et al. 1998). Further studies of massive YSO ouflows are necessary to determine whether high and low mass protostars share a common outflow mechanism.

The chemistry in massive and low mass YSO outflows is presumably driven by a common process. Outflows drive shocks into the ambient molecular gas and strongly affect the chemistry. The chemistry is altered by both the heating action of the shock which permits reactions with high activation energies to take place and by the injection of disrupted grain mantles (and perhaps grain cores) into the gas phase. The most striking difference between quiescent cloud and shock-driven chemistry is in the abundance of SiO, which is several orders of magnitude higher toward shocked regions (Schilke et al. 1997a). Other species, such as SO, NH₃ and CH₃OH are also observed to have raised abundances toward outflows (Bachiller 1996 and references therein).

Molecular line surveys of low mass YSO outflows (e.g. Blake et al. 1995; Bachiller & Perez-Gutierrez 1997) have been extremely useful in characterising the chemistry of the outflow and in deriving the physical parameters of the gas from the detected molecular lines. However as in the dynamical study of outflows the main effort has focussed on low mass YSO outflows. For this reason we have undertaken a 330-360 GHz molecular line survey of the massive YSO outflow associated with G5.89-0.39. The outflow is compact, the lobes are separated by $\sim\! 6$$^{\prime\prime}$ as traced in SiO emission, and massive with an outflowing mass of $\sim\! 80$ solar masses (Harvey & Forveille 1988; Acord et al. 1997). There is also a dense envelope of dust (Harvey et al. 1994) and molecular gas (Gomez et al. 1991) surrounding the UC HII region, which is the prototypical example of the shell morphology (Wood & Churchwell 1989).

The compactness of the outflow means that both lobes and the envelope gas associated with the UC HII region can be sampled in a single 13$^{\prime\prime}$ HPBW observation. The line profiles can be used to determine whether emission arises predominantly from the envelope gas located close to the UC HII region or from within the accelerated gas of the outflow. Turbulence within the envelope, observed to be fairly ubiquitous towards regions of massive star formation (Plume et al. 1997), will confuse the issue although the line profiles of particular species will indicate species that can be explored further with higher resolution interferometric observations.

The HII region G5.89-0.39 is also postulated to be young from observations of its expansion rate (Acord et al. 1998) and the high dust density in the immediate environment (Harvey et al. 1994). We will be able to contrast the results of the 330-360 GHz survey with other recent molecular line surveys of G34.26 (Macdonald et al. 1996) and Orion-KL (Schilke et al. 1997b), in order to test the hypothesis that the chemistry of these objects can be used as a "chemical clock'' to date the evolution of the molecular gas.

In summary, the aims of this survey are: (1) to characterise the chemistry of the molecular gas associated with G5.89-0.39, (2) to identify species located within the outflow by their high velocity line wings, (3) to derive the physical parameters of the molecular gas and (4) to examine whether hot core or shock-driven chemical models can reproduce the observed column densities. Parts (1) and (2) are presented in this paper, along with a first approach to part (3) by deriving the rotational temperatures and column densities of those species with sufficient detected transitions. A more detailed physical analysis and chemical modelling (parts 3 and 4) will be contained in the forthcoming Paper II.

The observations and data reduction procedure are detailed in the next section. The identified lines, inspection of their profiles for high velocity emission and rotation diagrams are presented in Sect. 3. The implications of the survey are discussed in Sect. 4 with particular reference to shock-driven and hot core chemical models.