Millimeter interferometry of W3 IRS5: A Trapezium in the making

Context. Although most young massive stars appear to be part of multiple systems, it is poorly understood how this multiplicity influences the formation of massive stars. The high-mass star-forming region W3 IRS5 is a prime example of a young massive cluster where the cluster center is resolved into multiple subsources at cm and infrared wavelengths, a potential proto-Trapezium system. Aims. We investigate the protostellar content in the 1.4 mm continuum down to subarcsecond scales and study the compact outflow components, also tracing the outflows back to their driving sources via the shocktracing SiO and SO2 emission. Methods. The region W3 IRS5 was mapped with the PdBI at 1.4 mm and 3.4 mm in the AB configurations, tuning the receivers to observe the molecular transitions SO2 (222,20–221,21), SO2 (83,5–92,8), SiO (2–1), and SiO (5–4). Results. In the continuum we detect five sources, one of them for the first time, while counterparts were detected in the NIR, MIR or at radio wavelengths for the remaining four sources. Three of the detected sources are within the inner 2100 AU, where the protostellar number density exceeds 106 protostars pc−3 assuming spherical symmetry. Lower limits for the circumstellar masses of the detected sources were calculated, ranging from ∼0.3 to ∼40 M⊙ although they were strongly affected by the spatial filtering of the interferometer, losing up to ∼90% of the single-dish flux. However, the projected separations of the sources ranging between ∼ 750 and ∼ 4700 AU indicate a multiple, Trapezium-like system. We disentangled the compact outflow component of W3 IRS5, detecting five molecular outflows in SiO, two of them nearly in the line of sight direction, which allowed us to see the collapsing protostars in the NIR through the cavities carved by the outflows. The SO2 velocity structure indicates a rotating, bound system, and we find tentative signatures of converging flows as predicted by the gravoturbulent star formation and converging flow theories. Conclusions. The obtained data strongly indicate that the clustered environment has a major influence on the formation of high-mass stars; however, our data do not clearly allow us to distinguish whether the ongoing star-forming process follows a monolithic collapse or a competitive accretion mechanism.