2013A&A...550A..37L

Water is an essential molecule in oxygen chemistry and the main constituent of grain icy mantles. The formation of water can be studied through the HDO/H₂O ratio. Thanks to the launch of the Herschel satellite and the advance of sensitive submillimeter receivers on ground telescopes, many H₂O and HDO transitions can now be observed, enabling more accurate studies of the level of water fractionation. Using these new technologies, we aim at revisiting the water fractionation studies toward massive star-forming regions. We present here a detailed study toward G34.26+0.15, a massive star-forming region associated with compact HII regions. We present observations of five HDO lines obtained with the APEX telescope. Two of those transitions are ground-state transitions. Two of the three high-excitation lines were additionally observed at higher angular resolution with the SMA. We analyzed these observations using the 1D radiative transfer code RATRAN and adopting different physical profiles from two different models. Although the inner and outer fractional abundances relative to H₂ can be best constrained to be X^HDO_in(T>100 K)=(5-7)x10^–8(3σ) and X^HDO_out(T≤100K)=(0.3-2)x10^–11(3σ), the line profile of the 893GHz ground transition cannot be well reproduced. This line profile is shown to be very sensitive to the velocity field. To better constrain the velocity field, it is necessary to observe the HDO line at 893 GHz with high angular resolution. The H₂O abundance is deduced from one high-excitation and one ground transition H₂¹⁸O line. The D/H ratios of water are 3.0x10^–4 in the inner region and (1.9-4.9)x10^–4 in the outer region of the core. The HDO fractional abundance in the inner and outer regions are different by more than four orders, which implies that the sublimation is very similar in low- and high-mass protostars. The D/H ratios of water in G34.26+0.15 are close to the value obtained for the same source in a previous study, and similar to those in other high-mass sources, but lower than those in low-mass protostars, suggesting the possibility that the dense and cold pre-collapse phase is shorter for high-mass star-forming regions.