Since the development of chirped pulse amplification, there has been a significant increase in high peak intensity many-cycle laser pulses. Such pulses are of great interest for driving a variety of laser-plasma interactions, with many of these effects depending greatly on the driving intensity of the laser field. These effects include phenomena such as relativistic high harmonic generation with solid density targets and electron acceleration to relativistic speeds using laser wakefield acceleration.
These interactions scale with the peak intensity of the laser pulse, meaning higher peak intensities are desired. Since the pulse duration for many of these systems are $\approx 10x$ longer than their single cycle limit, a factor of $10$ increase in peak intensity could be achieved by creating a pulse with identical energy but with a single cycle pulse duration.
The current amplifier technologies makes the creation of high energy few-cycle laser pulses a very difficult task. Instead of focusing on maintaining a few-cycle laser pulse during amplification, one aspect of this thesis focuses on the application of self-phase modulation as a high efficiency method to compress a many-cycle laser pulse after amplification, enabling the creation of high energy few-cycle laser pulses.
As the single-cycle limit is approached, the bandwidth required to support the pulse increases substantially. This can cause issues as a pulse propagates through material to an experiment, as material dispersion is able to substantially alter the temporal profile of the pulse, substantially reducing the peak intensity. In addition to material dispersion, there can be a nonlinear coupling between self-phase modulation and the material dispersion which can cause a substantially larger than expected change in the peak intensity of the pulse, which can not be pre-compensated for like material dispersion. In this thesis, I examine the effect of this coupling for a variety of laser intensities and discuss methods of mitigating the undesired decrease of the peak intensities.
Since a few-cycle laser pulse is easily able to have the peak intensity reduced due to material propagation, proper temporal profile characterization is required to ensure the desired pulse duration is actually obtained. In this thesis, I discuss two machine learning based methods that utilize self-phase modulation to characterize the phase and temporal profile of the laser.
