Conventional accelerators have to be hundreds of metres or longer to accelerate particles to energies in the GeV range or higher. Laser-produced plasmas could form the basis of next-generation "table-top" accelerators because they can support electric fields that are thousands of times stronger than those produced in traditional accelerators.

In the "laser wakefield" accelerator approach exploited by Krushelnick and colleagues at Imperial, the Ecole Polytechnique in Paris, the Rutherford Appleton Laboratory (RAL), the University of California at Los Angeles (UCLA)and AWE Aldermaston, the radiation pressure of an intense laser pulse is used to displace the electrons in a laser-produced plasma, leaving a large electric field in its wake. It is this field that accelerates the electrons. By using higher laser intensities than before it has been possible to accelerate electrons to higher energies.

However, the team has shown that at laser intensities above 1020 watts per square centimetre, the acceleration mechanism changes from the wakefield mechanism to one in which the laser directly accelerates the electrons. Computer simulations of the experiment show that the acceleration at these intensities is due to the radiation pressure of the laser "digging out" a hollow channel, which contains virtually no electrons, in the low-density plasma.

"One of the implications of these results is that simply turning up the laser power is not sufficient to produce really good quality electron beams," says team member Stuart Mangles of Imperial. "Both theoretical and experimental work is underway to work out how we can make high intensity electron acceleration experiments push to even higher energies and better beam qualities."

The experiments rely on the Vulcan Petawatt laser at RAL, which is currently the world's highest intensity laser, and simulations carried out with the Osiris computer code. The team now plans to push to even higher electron energies.