Abstract

Since the original prediction and demonstration of attosecond pulses, attosecond physics has entrenched itself in the ultrafast sciences, and promises to advance a wide range of scientific disciplines. It has the potential to provide key developments and insights in several research areas, such as atomic physics, quantum chemistry, biology and medicine. At present, engaging in this novel field of research is rather prohibitive, due to the high costs of cutting-edge technology and a steep learning curve. After all, playing with attosecond pulses is tantamount to playing with the shortest events ever made by man! Nonetheless, these are just typical growing pains of a new and exciting research area, and will eventually subside to make attosecond science accessible to a broad research community. In the meanwhile, as this promising field is taking its baby steps, it is the responsibility of those working at the cutting edge to propose novel experiments, and develop the tools and models that will be used in the future, as the field matures. Attosecond science comprises two frontiers: (i) the generation and characterization of increasingly intense, energetic, short and isolated attosecond pulses; and (ii) the design of experiments to probe physical systems on the attosecond time scale, the holy grail being the attosecond pump-attosecond probe time-resolved spectroscopic measurement. The second frontier offers a deeper understanding of the temporal behavior of the microcosm, but relies on advancements made in the first one. At present, both of these frontiers heavily rely on the attosecond streaking technique, which consists in energy-resolving photoelectrons ejected by an attosecond extreme ultraviolet pulse, in the presence of a phase-stabilized and temporally synchronized near-infrared field. Although it was originally devised as a means to characterize attosecond pulses, this measurement technique has even produced new discoveries in atomic and solid-state physics, due to pioneering experiments by M. Drescher, A. Cavalieri, G. Sansone, M. Schultze, and others, and has inspired novel theories of laser-dressed photoionization by V. S. Yakovlev, A. Scrinzi, O. Smirnova, M. Y. Ivanov and others. In the first part of this thesis, I focus on new methods I developed for the analysis of attosecond streaking measurements. One of these methods, based on a formalism I devised based on electron trajectories in a laser field, can directly recover the chirp of an attosecond pulse from a set of streaked photoelectron spectra. Next, I describe a robust optimization algorithm, based on a formalism due to M. Kitzler et al., that can completely recover the temporal profile of an attosecond pulse with an arbitrary shape. This optimization algorithm was used to characterize the field of 80 as pulses, the shortest on record, and to uncover a delay of 20 as between the photoemissions from the 2s and 2p sub-shells of neon; both experiments were performed here at the Max Planck Institut fuer Quantenoptik in 2008 and 2010, respectively. Moreover, during the course of this work, it was established by V. S. Yakovlev et al. that the attosecond streaking technique generally measures a quantity that is related to the photoelectron wave packet, not the attosecond light pulse. Only when the energy-resolved dipole response, given by the bound-free transition matrix elements, is nearly constant can we take the electron wave packet as a replica of the attosecond pulse. In light of this finding, I show that the attosecond streaking technique provides a means to measure and even time-resolve the energy-dependent phase of transition dipole matrix elements. Finally, I consider the laser-dressed scattering of an attosecond photoelectron wave packet. I show that the scattering of a photoelectron, emitted by an attosecond pulse from a localized state in a spatially extended system, can be influenced by a near-infrared laser field. Measuring the photoelectron spectrum reveals an interference pattern which is affected by the intensity of the near-infrared field. To describe these effects, I introduce a model based on classical trajectories that quantitatively predicts laser-dressed photoelectron spectra for such a spatially-extended system.