Infrared emission of young HII regions: a Herschel/Hi-GAL study

Context. Investigating the relationship between radio and infrared emission of Hii regions may help shed light on the nature of the ionizing stars and the formation mechanism of early-type stars in general. Aims: We have taken advantage of recent unbiased surveys of the Galactic plane such as Herschel/Hi-GAL and VLA/CORNISH to study a bona fide sample of young Hii regions located in the Galactic longitude range 10°-65° by comparing the mid- and far-IR continuum emission to the radio free-free emission at 5 GHz. Methods: We have identified the Hi-GAL counterparts of 230 CORNISH Hii regions and reconstructed the spectral energy distributions of 204 of these by complementing the Hi-GAL fluxes with ancillary data at longer and shorter wavelengths. Using literature data, we obtained a kinematical distance estimate for 200 Hii regions with Hi-GAL counterparts and determined their luminosities by integrating the emission of the corresponding spectral energy distributions. We have also estimated the mass of the associated molecular clumps from the (sub)millimeter flux densities. Results: Our main finding is that for ~1/3 of the Hii regions the Lyman continuum luminosity appears to be greater than the value expected for a zero-age main-sequence star with the same bolometric luminosity. This result indicates that a considerable fraction of young, embedded early-type stars presents a "Lyman excess" possibly due to UV photons emitted from shocked material infalling onto the star itself and/or a circumstellar disk. Finally, by comparing the bolometric and Lyman continuum luminosities with the mass of the associated clump, we derive a star formation efficiency of 5%. Conclusions: The results obtained suggest that accretion may still be present during the early stages of the evolution of Hii regions, with important effects on the production of ionizing photons and thus on the circumstellar environment. More reliable numerical models describing the accretion process onto massive stars are required to shed light on the origin of the observed Lyman excess.