7. Summary and conclusions
We have performed a 330-360 GHz molecular line survey of the halo gas surrounding the hot core associated with G34.26+0.15. The species detected are mainly simple diatomic or triatomic molecules, with the exceptions of formaldehyde and methanol. We have extended the chemical model of Paper II to predict the column densities of a general line of sight drawn through the model cloud and to predict beam-averaged column densities sampled by a gaussian beam. A comparison of single line of sight and beam-averaged models reveals that for most molecules the predicted column densities do not greatly differ between the two models, apart from a small increase in the column densities of the beam-averaged model. This is because the H2 column density is slightly larger in the beam-averaged model than in the "pencil-beam" single line of sight model.
For certain molecules (CH3OH and SO2) the addition of beam-averaging has significant effects due to the influence of the core chemistry upon the beam-averaged column density. The observational evidence is against this; both SO2 and CH3OH have much narrower linewidths in the halo than in the core and a rotation diagram for CH3OH yields a much lower temperature than that of the core, suggesting that the emission is mainly from the halo. Further observations with larger angular distances from the core and more detailed modelling of mantle evaporation processes are needed to confirm the origins of SO2 and CH3OH in the halo.
A comparison of our models with the observations shows that different species predict different ages for the cloud. The column densities of CO, C2H, CN, H2CO and HCN are reasonably consistent with a halo age of 106 years whereas the column densities of CH3OH, SO2 and HCO+ are inconsistent with halo ages greater than 104 years. We have investigated the propyne (CH3CCH) column density in the halo observed by Hatchell et al. (1998a) and find that neither model can reproduce the column density of propyne for halo ages of greater than 105 years. The high observed column densities of Hatchell et al. (1998a) suggest that propyne chemistry is poorly understood. Further observations and modelling are of the utmost importance in identifying the possible formation routes of CH3CCH.
There are a number of approaches that may resolve the problems outlined above. The model does not take into account the evaporation of dust grain ice mantles in the inner layers of the halo, where the structural model and the CH3OH rotation temperature predict that the temperature is high enough to evaporate CO. The addition of more detailed ice evaporation processes to the model will bring the predictions of the CO abundance better into line with that observed and may provide a better fit to the observed column densities of species such as CH3OH and H2CO which are both thought to form on grain surfaces by hydrogen reactions with CO ice. As noted in Paper II and van Dishoeck & Blake (1998), the adoption of stratified grain mantle desorption within the cloud (i.e. grain mantle composition may vary radially through the cloud) leading to the concept of different species evaporating at different distances from the cloud centre could help to improve the radial chemical model of G34.26+0.15. Dust continuum observations are needed to provide information about the temperature of grain mantles in the halo of G34.26+0.15.
The initial abundance of sulphur in the chemical model may be too high by a factor of 10, as indicated by recent studies of hot core chemistry (Hatchell et al. 1998b). Reducing the value to that suggested reverses the prediction of SO2 that the halo chemistry is young, instead lending weight to the supposition that the halo is older than 105 years. A reduction in the sulphur abundance would also bring the model predictions of CS better into line with those observed. We conclude that, based on current evidence, this is the most likely age for the halo given the agreement between CO, C2H, H2CO and HCN. However we note that if grain mantle evaporation is important in the inner layers of the halo then these timescales are likely to change as a substantial fraction of the CO and H2CO in the halo may be directly evaporated from the grains rather than slowly built up by gas-phase chemistry.
An outstanding problem in the study of cold gas such as the halo of G34.26 is that many of the detected molecules are linear, implying that their transitions are spaced widely in frequency. The column densities of the species detected in the survey of the halo are thus almost all lower limits (with the exception of methanol). Multifrequency studies are needed to constrain the column densities of the observed linear species, perhaps by extending the survey into the 210-280 GHz range. Studies at many positions are needed to constrain the abundance profiles, these can be obtained efficiently (at least for low frequency lines) by using the new generation of array receivers in a similar manner to the surveys of the giant molecular clouds M17 and Cepheus A by Bergin et al. (1997). Optical depth effects not accounted for in our model can also be dealt with by coupling the chemical model with a radiative transfer model in a similar manner to that pursued in models of low mass star forming cores. A valuable first step in this direction has been achieved by Doty & Neufeld (1997) wherein the UMIST chemical model has been combined with the radiative transfer model of Doty (1997). However the number of species modelled is not large (they calculate line profiles of only CO and water isotopes) and there is scope for enhancing our knowledge of the cloud chemistry by matching the line profiles generated by model abundances to molecular line observations, using a less restrictive radiative transfer model such as that developed by Phillips & Little (1998).
© European Southern Observatory (ESO) 1999
Online publication: February 23, 1999