S. Van Loo

Parsec-scale SiO emission in an infrared dark cloud

We present high-sensitivity 2 × 4a rcmin 2 maps of the J = 2→1 rotational lines of SiO, CO, 13 CO and C 18 O, observed towards the filamentary infrared dark cloud (IRDC) G035.39−00.33. Single-pointing spectra of the SiO J = 2→1 and J = 3→2 lines towards several regions in the filament are also reported. The SiO images reveal that SiO is widespread along the IRDC (size ≥2 pc), showing two different components: one bright and compact arising from three condensations (N, E and S) and the other weak and extended along the filament. While the first component shows broad lines (linewidths of ∼4–7 km s −1 ) in both SiO J = 2→1 and SiO J = 3→2, the second one is only detected in SiO J = 2→1 and has narrow lines (∼0.8 km s −1 ). The maps of CO and its isotopologues show that low-density filaments are intersecting the IRDC and appear to merge towards the densest portion of the cloud. This resembles the molecular structures predicted by flow-driven, shock-induced and magneticallyregulated cloud formation models. As in outflows associated with low-mass star formation, the excitation temperatures and fractional abundances of SiO towards N, E and S increase with velocity from ∼ 6t o 40 Ka nd from∼10 −10 to ≥10 −8 , respectively, over a velocity range of ∼ 7k m s −1 . Since 8 μm and 24 μm sources and/or extended 4.5 μm emission are detected in N, E and S, broad SiO is likely produced in outflows associated with high-mass protostars. The excitation temperatures and fractional abundances of the narrow SiO lines, however, are very low (∼9 K and ∼10 −11 , respectively), and consistent with the processing of interstellar grains by the passage of a shock with vs ∼ 12 km s −1 . This emission could be generated (i) by a large-scale shock, perhaps remnant of the IRDC formation process, (ii) by decelerated or recently processed gas in large-scale outflows driven by 8- and 24-μm sources or (iii) by an undetected and widespread population of lower mass protostars. High-angular-resolution observations are needed to disentangle between these three scenarios.

A multi-wavelength investigation of the non-thermal radio emitting O-star 9 Sgr

We report the results of a multi-wavelength investigation of the O4 V star 9 Sgr (= HD 164794). Our data include observations in the X-ray domain withXMM-Newton, in the radio domain with the VLA as well as optical spectroscopy. 9 Sgr is one of a few presumably single OB stars that display non-thermal radio emission. This phenomenon is attributed to synchrotron emission by relativistic electrons accelerated in strong hydrodynamic shocks in the stellar wind. Given the enormous supply of photospheric UV photons in the wind of 9 Sgr, inverse Compton scattering by these relativistic electrons is a priori expected to generate a non-thermal power law tail in the X-ray spectrum. Our EPIC and RGS spectra of 9 Sgr reveal a more complex situation than expected from this simple theoretical picture. While the bulk of the thermal X-ray emission from 9 Sgr arises most probably in a plasma at temperature3 10 6 K distributed throughout the wind, the nature of the hard emission in the X-ray spectrum is less clear. Assuming a non-thermal origin, our best fitting model yields a photon index of2: 9f or the power law component which would imply a low compression ratio of1:79 for the shocks responsible for the electron acceleration. However, the hard emission can also be explained by a thermal plasma at a temperature2 10 7 K. Our VLA data indicate that the radio emission of 9 Sgr was clearly non-thermal at the time of the XMM-Newton observation. Again, we derive a low compression ratio (1.7) for the shocks that accelerate the electrons responsible for the synchrotron radio emission. Finally, our optical spectra reveal long-term radial velocity variations suggesting that 9 Sgr could be a long-period spectroscopic binary.

Shock-triggered formation of magnetically-dominated clouds

Aims. Our aim is to understand the formation of a magnetically dominated molecular cloud out of an atomic cloud. Methods. A thermally stable warm atomic cloud is initially in static equilibrium with the surrounding hot ionised gas. A shock propagating through the hot medium interacts with the cloud. We follow the dynamical evolution of the cloud with a time-dependent axisymmetric magnetohydrodynamic code. Results. As a fast-mode shock propagates through the cloud, the gas behind it becomes thermally unstable. The β value of the gas also becomes much smaller than the initial value of order unity. These conditions are ideal for magnetohydrodynamic waves to produce high-density clumps embedded in a rarefied warm medium. A slow-mode shock follows the fast-mode shock. Behind this shock a dense shell forms, which subsequently fragments. This is a primary region for the formation of massive stars. Our simulations show that only weak and moderate-strength shocks can form cold clouds which have properties typical of giant molecular clouds.

Can single O stars produce non-thermal radio emission?

We present qualitative models for the non-thermal radio emission of single O stars, in terms of synchrotron emission by wind-embedded shocks. When we include the fact that shocks weaken as they move out with the wind, as predicted by time-dependent hydrodynamical simulations, these models produce a radio spectrum with a positive slope (as function of frequency), in contradiction with the observed negative slope. We conclude that non-thermal radio emission cannot originate from wind-embedded shocks, and is likely to be caused by a wind-colliding shock. A radio light curve analysis of two non-thermal O stars that are generally assumed to be single supports this conclusion.

Shock-triggered formation of magnetically dominated clouds – II. Weak shock–cloud interaction in three dimensions

To understand the formation of a magnetically dominated molecular cloud from an atomic cloud, we study the interaction of a weak, radiative shock with a magnetized cloud. The thermally stable warm atomic cloud is initially in static equilibrium with the surrounding hot ionized gas. A shock propagating through the hot medium then interacts with the cloud. We follow the dynamical evolution of the shocked cloud with a time-dependent ideal magnetohydrodynamic code. By performing the simulations in 3D, we investigate the effect of different magnetic field orientations including parallel, perpendicular and oblique to the shock normal. We find that the angle between the shock normal and the magnetic field must be small to produce clouds with properties similar to observed molecular clouds.

Non-thermal radio emission from O-type stars III. Is Cygnus OB2 No. 9 a wind-colliding binary?

The star Cyg OB2 No. 9 is a well-known non-thermal radio emitter. Recent theoretical work suggests that all such O-stars should be in a binary or a multiple system. However, there is no spectroscopic evidence of a binary component. Re-analysis of radio observations from the VLA of this system over 25 years has revealed that the non-thermal emission varies with a period of 2.35 ± 0.02 yr. This is interpreted as a strong suggestion of a binary system, with the non-thermal emission arising in a wind-collision region. We derived some preliminary orbital parameters for this putative binary and revised the mass-loss rate of the primary star downward from previous estimates.

Non-thermal radio emission from single hot stars

We present a theoretical model for the non-thermal radio emission from single hot stars, in terms of synchrotron radiation from electrons accelerated in wind-embedded shocks. The model is described by five independent parameters each with a straightforward physical interpretation. Applying the model to a high-quality observation of Cyg OB2 No. 9 (O5 If), we obtain meaningful constraints on most parameters. The most important result is that the outer boundary of the synchrotron emission region must lie between 500 and 2200 stellar radii. This means that shocks must persist up to that distance. We also find that relatively weak shocks (with a compression ratio <3) are needed to produce the observed radio spectrum. These results are compatible with current hydrodynamical predictions. Most of our models also show a relativistic electron fraction that increases outwards. This points to an increasing efficiency of the acceleration mechanism, perhaps due to multiple acceleration, or an increase in the strength of the shocks. Implications of our results for non-thermal X-ray emission are discussed.

Sputtering in oblique C-type shocks

We present the first results for the sputtering of grain mantles and cores obtained with selfconsistent multifluid hydromagnetic models of oblique C-type shocks propagating through dusty media. The threshold shock speed for mantle sputtering is about 10 km s −1 and is independent of density. The mantles are completely vapourized in shocks with speeds of 20–25 km s −1 . At such shock speeds, core sputtering commences and gas-phase SiO forms. Core destruction is not total in any C-type shock because grains are not completely destroyed in shocks with speeds near the minimum speeds at which J-type shocks appear. Due to the density dependence of the critical shock speed for this transition, higher gas-phase SiO fractional abundances are produced behind shocks propagating in lower density gas. For shock speeds near the threshold speeds for both core and mantle sputtering, sputtering is much greater for shock velocities at smaller angles relative to the upstream magnetic field. At higher shock speeds, the angular variation is still present but less pronounced.

Time-dependent simulations of steady C-type shocks

Using a time-dependent multifluid, magnetohydrodynamic code, we calculated the structure of steady perpendicular and oblique C-type shocks in dusty plasmas. We included relevant processes to describe mass transfer between the different fluids, radiative cooling by emission lines and grain charging, and studied the effect of single- and multiple-sized grains on the shock structure. Our models are the first of oblique fast-mode molecular shocks in which such a rigorous treatment of the dust grain dynamics has been combined with a self-consistent calculation of the thermal and ionization structures including appropriate microphysics. At low densities, the grains do not play any significant role in the shock dynamics. At high densities, the ionization fraction is sufficiently low that dust grains are important charge and current carriers and, thus, determine the shock structure. We find that the magnetic field in the shock front has a significant rotation out of the initial upstream plane. This is most pronounced for single-sized grains and small angles of the shock normal with the magnetic field. Our results are similar to previous studies of steady C-type shocks showing that our method is efficient, rigorous and robust. Unlike the method employed in the previous most detailed treatment of dust in steady oblique fast-mode shocks, ours allow a reliable calculation even when chemical or other conditions deviate from local statistical equilibrium. We are also able to model transient phenomena.

The effect of ambipolar resistivity on the formation of dense cores

Aims. We aim to understand the formation of dense cores by magnetosonic waves in regions where the thermal to magnetic pressure ratio is small. Because of the low-ionisation fraction in molecular clouds, neutral and charged particles are weakly coupled. Ambipolar diffusion then plays an important role in the formation process. Methods. A quiescent, uniform plasma is perturbed by a fast-mode wave. Using 2D numerical simulations, we follow the evolution of the fast-mode wave. The simulations are done with a multifluid, adaptive mesh refinement MHD code. Results. Initial perturbations with wavelengths that are 2 orders of magnitude larger than the dissipation length are strongly affected by the ion-neutral drift. Only in situations where there are large variations in the magnetic field corresponding to a highly turbulent gas can fast-mode waves generate dense cores. This means that, in most cores, no substructure can be produced. However, Core D of TMC-1 is an exception to this case. Due to its atypically high ionisation fraction, waves with wavelengths up to 3 orders of magnitude greater than the dissipation length can be present. Such waves are only weakly affected by ambipolar diffusion and can produce dense substructure without large wave-amplitudes. Our results also explain the observed transition from Alfvenic turbulent motion on large scales to subsonic motions at the level of dense cores.