Towards external injection in laser wakefield acceleration

In laser wakefield acceleration (LWA) a plasma wave is driven by a high intensity ultra short laser pulse and the longitudinal electric fields in the plasma wave are used to accelerate electron bunches. Electrons with an appropriate kinetic energy, injected on the right phase of the plasma wave, get trapped by the plasma wave and are accelerated. This thesis investigates experimentally the feasibility of LWA with injected electron bunches produced by a radio frequency photogun. A laser system was developed which is able to focus 3 TW pulses on a spot with a 1/e2 radius of 40 µm and a shot-to-shot position stability of 4 µm. Accelerating distances exceeding the Rayleigh length of the laser are achieved by preforming the plasma density profile to obtain a collimated propagation of the laser pulse through the plasma (guided laser pulse). The laser pulses are guided over centimeter scale distances through a capillary discharge plasma with an on-axis electron density of ~1×1024 m-3. The guiding properties of the plasma channel were investigated. It is found that a second discharge current pulse through the plasma, ~1 µs after the primary discharge, improves the guiding properties considerably. The transmittance is higher (>90%), the guided laser spot is more cylindrically symmetric at the exit of the plasma channel and the time-window for guiding becomes approximately 10 times longer (~600 ns). An RF-photogun had been purpose-built as an injector of electrons into the plasma channel. Different properties of the RF-photogun and the electron bunches produced were measured to determine the optimal settings for LWA. For an electron bunch with 10 pC charge and 3.7 MeV kinetic energy, the energy spread is 0.5% and the transverse emittance is 1.9 µm. After focusing the electron bunch at the entrance of the plasma channel by a pulsed solenoid lens, the focal spot has an RMS radius (standard deviation) of 40 µm with a shot-to-shot position stability of 5 µm. The RMS length of this electron bunch, derived from simulations, is 400 fs at focus. The arrival time jitter between laser pulse and electron bunch at the entrance of the plasma channel was inferred from earlier work to be around 150 fs in the present setup. This implies consistent temporal overlap between the laser wakefield and the injected electron bunch. The shot-to-shot stability and focal spot of the laser pulse and electron bunch at focus shows that there is always good overlap in transverse direction between the injected electron bunch, the laser pulse and the plasma channel. Due to technical difficulties, the energy of the electrons from the RF-photogun was limited to 3.7 MeV. With this energy, the injector can serve for one particular version of laser wakefield acceleration, i.e. injection ahead of the laser pulse. Using the actually measured electron bunch parameters and simulating the injection of a 3.7 MeV electron bunch of 10 pC in front of a 25 TW laser pulse with a waist of 30 µm in a plasma with a density of 0.7×1024 m-3, the maximum accelerated charge was found to be 1.2 pC with a kinetic energy of ~900 MeV and an energy spread of ~5%. These results show that laser wakefield acceleration of electrons injected by an RF photogun is feasible