Context. Despite its major role in the evolution of the interstellar medium, the formation of high-mass stars (M ≥ 10 M_⊙) remains poorly understood. Two types of massive star cluster precursors, the so-called massive dense cores (MDCs), have been observed, which differ in terms of their mid-infrared brightness. The origin of this difference has not yet been established and may be the result of evolution, density, geometry differences, or a combination of these.
Aims: We compare several molecular tracers of physical conditions (hot cores, shocks) observed in a sample of mid-IR weakly emitting MDCs with previous results obtained in a sample of exclusively mid-IR bright MDCs. We attempt to understand the differences between these two types of object.
Methods: We present single-dish observations of HDO, H_218O, SO2, and CH3OH lines at λ = 1.3-3.5 mm. We study line profiles and estimate abundances of these molecules, and use a partial correlation method to search for trends in the results.
Results: The detection rates of thermal emission lines are found to be very similar for both mid-IR quiet and bright objects. The abundances of H2O, HDO (10-13 to 10-9 in the cold outer envelopes), SO2 and CH3OH differ from source to source but independently of their mid-IR flux. In contrast, the methanol class I maser emission, a tracer of outflow shocks, is found to be strongly anti-correlated with the 12 μm source brightnesses.
Conclusions: The enhancement of the methanol maser emission in mid-IR quiet MDCs may be indicative of a more embedded nature. Since total masses are similar between the two samples, we suggest that the matter distribution is spherical around mid-IR quiet sources but flattened around mid-IR bright ones. In contrast, water emission is associated with objects containing a hot molecular core, irrespective of their mid-IR brightness. These results indicate that the mid-IR brightness of MDCs is an indicator of their evolutionary stage.
