Diagnosis and application of laser wakefield accelerators

Abstract

This thesis is concerned with experimental diagnostic methods of the laser-plasma interaction in laser wakefield accelerators, and how high energy photon beams from such accelerators may be exploited in a potential application.

Raman scattered light from the 20 TW Astra laser pulse was found to have shorter wavelength than expected given the interferometrically-measured plasma density, due to the relativistic motion of plasma electrons. Simulations indicated that the scattering helps to shape the pulse, removing laser energy not trapped in the wake. The signature of Raman scatter from the 200 TW Astra-Gemini pulse was instead the generation of extended plasma filaments, the direction of which indicated weaker relativistic effects. Altering the focal spot position or quality generated more filaments and reduced electron beam charge, linking Raman scattering, pulse focussing, and accelerator performance.

When a low intensity (10¹⁰ W/cm²) probe pulse crossed a high intensity (10¹⁹ W/cm²) drive pulse of equal frequency, the probe was observed to be amplified in intensity by up to 10⁵ over a distance of tens of microns. No frequency shift of the probe was observed, and the effect required the polarisations of the pulses to be parallel. In simulations the pulses created a ponderomotive grating, trapping electrons which were able to scatter the drive into the probe in the superradiant regime of Raman amplification.

Laser wakefield-driven x-ray beams were used to perform a microtomographic scan of a human femoral trabecular bone sample. Under optimal conditions x-ray beams containing 1.3 ± 0.5 × 10⁹ photons were produced on 97% of laser shots, with critical energy 33 ± 12 keV and source size of 2-3 μm. Image resolution was 36 ± 7 μm and after 180 image acquisitions the 3D resolution of the tomogram was up to 50 μm. The average photon flux between 10 and 100 keV was 5.9 ± 2.4 × 10⁵ ph/s/mrad² which is comparable to microfocus sources capable of few-micron source sizes.