Stars form when a cloud of interstellar gas and dust collapses under its own gravity. The denser the gas gets, the higher its temperature, up to the point when thermonuclear reactions ignite in the star's core. Radiation pressure pushes away the remaining gas surrounding the star and after a while it stops growing. Astronomers have determined through physical models that this process can create stars no more massive than 8 Suns. Larger stars, however, have been observed. This presents a puzzle. University of Toronto astronomers Peter Martin and Alana Rivera-Ingraham (now at the Institut de Recherche en Astrophysique et Planétologie in Toulouse, France) think they have come close to solving it.
The researchers used far-infrared images of areas containing clusters of very hot, young stars of the spectral types O and B obtained by ESA's Hershel Space Observatory. Herschel is blind to the visible and ultraviolet radiation emitted by these stars, but it can detect infrared and submillimetre radiation of interstellar dust from which these young stars have recently formed. While analyzing the images, the astronomers noticed that the densest areas in which new stars are likely to be formed are often surrounded by already existing, older stars. Each of them also exerts radiation pressure on the interstellar gas. If these stars happen to be favourably placed, they will blow the gas and dust into one common area. A new star that forms there will try to blow the gas away, but the other stars will blow it right back. Thus the new star in the centre can continue to grow beyond the theoretical limit of 8 solar masses.
This infrared image from Herschel shows a giant cloud of dust and gas called Westerhout 3 (W3). It is located 6,500 light-years away from us and spans about 230 light-years. Only interstellar dust is visible. Stars are not, since they do not emit significant amounts of far infrared radiation.
Sources: http://bit.ly/10xfVog, http://bit.ly/Z2Xwhr
Image credit: ESA/PACS Array SPIRE consortium, A. Rivera-Ingraham and P. G. Martin, University of Toronto, HOBYS Key Programme (F. Motte)