Trapping and acceleration of nonideal injected electron bunches in laser wakefield accelerators

Summary form only given. The standard regime for the laser wakefield accelerator (LWFA) usually requires external injection of MeV electrons. Ideally, the injected electron bunch should be injected into the proper phase of the accelerating wake, have a bunch length that is small compared with the plasma wavelength, and a low emittance and energy spread. There are several optical injection concepts that are theoretically capable of producing such precisely-timed (phased), high quality electron bunches, but the requirements are severe. This paper reports simulation studies of several nonideal injection schemes that demonstrate strong phase bunching and good accelerated beam quality in a channel-guided laser wakefield accelerator. A plasma channel is used to confine the laser pulse over a distance of many Rayleigh lengths. For the case of mono-energetic, unphased (long bunch) injection, the accelerated electrons have a large energy spread when the injection energy W/sub 0/ is high. However, there is an optimum range of injection energies for which the LWFA can trap a significant fraction of the injected pulse while producing an ultrashort, high-quality accelerated pulse. These favorable results are due to a combination of pruning of particles at unfavorable phases, rapid acceleration, and strong phase bunching. Simulation results for the trapping fraction, final average energy and energy spread, and the final bunch length agree well with the predictions of a simple Hamiltonian model using an ideal sinusoidal wake moving at the group velocity of the laser pulse. This long bunch case may apply to injection with a conventional or photocathode RF gun. The case of a channel-guided LWFA with a broad injected energy spread has also been simulated. Although the trapping fraction is generally much smaller than in the monoenergetic case, some simulations exhibit final accelerated bunches with remarkably small energy spread. These results suggest that relatively poor quality inj- ction pulses may still be useful in LWFA demonstration experiments. The implications for planned LWFA experiments at NRL using various optical injection schemes will be discussed.