|dc.description.abstract||The advancement in laser technology in the past few decades have enabled consistent generation of pulses in the femtosecond (fs, x10^-15 s) timescale. The strong electric fields produced by such pulses are comparable to those experienced by bound electrons within atoms and molecules, leading to highly non-linear interactions. One of the most probable such interaction is that of strong-field excitation where the target is left in an excited state. These excited states have been shown to influence other strong-field phenomena and exhibit unique properties that are useful for further applications such as, generation of coherent extreme-ultraviolet radiation, and lasercooling of noble gases. Therefore, a comprehensive understanding of the fundamental excitation process and how excitation rates are affected is necessary in order to tailor conditions for a desired outcome. So far, there have been only a few experimental studies on excitation yields due to the unique experimental arrangements required for observation of these states. In contrast, there have been more theoretical studies which have yet to be experimentally confirmed. The aforementioned experiments have been undertaken with laser pulses with a duration of 30 fs or more centred at a wavelength of 800 nm which contain many optical cycles. But, numerical calculations predict that excitation yields scale differently as the pulse duration reduces such that it contains only a few optical cycles. This has yet to be experimentally confirmed since there has not been any experimental studies on excitation yields from few-cycle pulses. Furthermore, the use of few-cycle pulses enables precise control over the electric field experienced by the atom which may influence the excitation process.
In this dissertation, we experimentally investigate excitation yields of argon interacting with multi- and few-cycle pulses centred at 800 nm and compare them to solutions of the time-dependent Schrodinger equation (TDSE). The first investigation explores the effect of changing the intensity spanning between 50-300 TW/cm2. By directly detecting excited states surviving the flight time to the particle detector, we show that excitation rates exhibit a step-wise increase within the intensity range which correspond to the absorption of 13 and 14 photons with linearly polarized multi-cycle pulses. These were predicted theoretically but were thought to be washed out due to volume-averaging inevitable in the experiment. Analysis of the numerical predictions reveal that these enhancements are mainly due to excitation into low-lying states, specifically the 5g and 6h states for 13- and 14-photon absorption, respectively. These increases are not observed with few-cycle pulses where the offset between the peak of the pulse envelope and the peak of the central electric field cycle, known as the carrier-envelope phase (CEP), was not locked. This is in excellent agreement with TDSE predictions. Population of low-lying states are largely preferred with few-cycle pulses and these enhancements are less pronounced, to the point where they do not persist after volume-averaging.
The second investigation explores excitation with elliptically polarized laser fields of varying ellipticities at select intensities with both multi- and few-cycle pulses. In all cases, excitation rates decrease quicker with increasing ellipticity than that of Ar+ but slower than predicted with the strong-field approximation as well as Ar2+. This indicates a different mechanism than the tunneling-plus-rescattering model proposed for the formation of Ar2+ through non-sequential double ionization. No anomalous peaks at non-zero ellipticity are observed in the experiment for 30 fs and 6 fs pulses at an intensity of 270 TW/cm2 and 200 TW/cm2, respectively, nor were they predicted by TDSE results. At a lower intensity, where previously published results from semi-classical modeling predict anomalous distributions, no obvious deviations from a normal distribution is observed. However, low statistics at this intensity limits any confident conclusions for a peak at very small, non-zero ellipticity values. Lastly, analysis of TDSE results reveals an anomalous distribution for excitation out of the pm= +-1 initial ground state orbitals. Further experiments are required for solid conclusions as well as good agreement between TDSE results and experiments.
The last investigation explores the role of the CEP of a few-cycle pulse. For the first time, we show that excitation rates are highly dependent on both the peak intensity and CEP of the pulse. At a single intensity, TDSE calculations predict up to a 55% variation in excitation rates. Furthermore, the CEP dependent trends can vary significantly with small changes in the intensity, leading to a significant variation in the optimum CEP for maximum excitation yields. In the experiment, volume averaging reduces the maximum observable variation in the CEP dependent yields to 7%. Furthermore, they are still highly dependent on the exact in situ peak intensity of the experimental pulse with many peak intensities resulting in a variation below 5%. This places tight restrictions on conditions which allow successful observation of the variation in yields with varying CEP. Despite the inability to precisely determine the in situ experimental intensity, the agreement with the numerical predictions is very good which serves to validate the theoretical predictions.
The results from these studies reveal that the population of excited states are dependent on the intensity, polarization, and, in the case of few-cycle pulses, the CEP. If the intensity can be precisely controlled, selective excitation to the 5g and 6h states can be achieved with up to a 60% likelihood with the use of multi-cycle pulses. This is reduced with volume-averaging but these states still remain the most populated states. Knowing this, excitation to the metastable state can be increased through direct stimulation via additional radiation. Further studies to determine the precise efficiency of the process is required in order to evaluate it as a suitable replacement for current metastable generation techniques.||