Cosmic-ray acceleration in a superbubble

Cosmic-rays are very energetic particles that pervade the Universe. Discovered one century ago, their origin is still unclear. In the Galaxy, the most promising sources are the powerful magnetized shock waves triggered by supernovae, which mark the cataclysmic end of a star: the thermonuclear disruption of a white dwarf (as for SN 1006) or the core-collapse of a massive star (as for RX J1713-3946). Massive stars are born in groups, called OB associations. And as they live fast and die young, they are likely to spend their whole life into clusters. Their strong stellar winds, together with their subsequent explosions, profoundly alter their environment: they carve huge, hot and tenous cavities around them, called superbubbles. It is naturally expected that these objects play a key role in the acceleration of cosmic rays in the Galaxy. And measurements of the composition of the cosmic rays indeed point to OB associations and their superbubbles as the likely source of the material of a substantial fraction of them. But the particles that we collect on the Earth don't tell where they come from: being charged particles, they wandered a lot in the Galactic magnetic field before reaching us. We can spot freshly accelerated particles next to their sources thanks to the high-energy radiation they emit. But, until recently, observations of superbubbles were very scarce.

The situation changed dramatically in the past years with the rise of a new generation of instruments. The Fermi satellite is one of them. Launched by NASA in 2008, this international mission observes the gamma-ray sky between the energies of about 100 MeV to 100 GeV with unprecedented sensitivity. It monitors all the sky, and, of particular interest for us today, a massive star forming region in the direction of Cygnus. Analyzing the data from high-energy instruments is a difficult task, especially for extended sources. After modelling and carefully removing a number of backgound diffuse emissions, scientists found a significant excess of high-energy emission from this part of the sky. They were stricken to see that the emission is coincident with the cavities formed around the numerous clusters of massive stars: it nicely follows the ionization fronts eroding the surrounding molecular clouds, as if in a "cocoon". Various objects can produce gamma-ray radiation (individual stars, pulsars and their nebulae), but after checking all possibilities the collaborators concluded that such small sources don't make a good match to the extended emission. The most likely explanation is that the cocoon is filled with cosmic-rays, that slowly diffuse inside the superbubble, and radiate some of their energy in the process (they could be either electrons or protons). The energy distribution of the observed photons is notably hard, meaning that the fraction of particles with the highest energies is higher than usual. Because it's all the more easy for particles to escape their accelerator when they are energetic, this hardness points to freshly accelerated particles. So here we have it: the Fermi observation provides the first strong direct evidence for the long-advocated hypothesis that OB associations host cosmic-ray factories.

The two main clusters in the cocoon are so young (4 to 6 million years for Cygnus OB2, 6 million years for NGC6910) that they might or not have experienced supernovae yet. But the winds of massives stars themselves can produce similar shock waves. And they generate lots of magnetic turbulence in the bubble, which traps particles in the cocoon, where they can get re-accelerated by other shocks and/or by magnetic waves. Existing models of acceleration into superbubbles predict spectra with a shape similar to that infered, which gives credance to the scenario. However it is too early for a detailed comparison of models with the cocoon spectrum: more work is needed on both fronts. Theoreticians will refine their models, and observers will have a closer look a other regions. The Arches, Quintuplet, and Sgr B2 clusters are good targets towards the Galactic centre, as are Westerlund 1 and Westerlund 2. These objects have already been detected at even higher energies, by one of the other prominent new instruments, the H.E.S.S. telescope.