Conference Agenda

Accurate characterization of ultrashort pulses is essential for ultrafast optics. We review the amplitude swing (a-swing) technique, a robust, in-line diagnostic tool specifically developed for high-precision measurements. Based on common-path birefringence-induced modulations, a-swing has evolved into a versatile device for ultrafast field reconstruction. Its capabilities include wide range of spectra and duration pulses, and full vectorial characterization. Thus, a-swing provides a comprehensive, simplified solution for metrology and process control in diverse ultrafast scenarios

Advanced temporal pulse shaping in the picosecond regime has seen remarkable developments over the recent years. An example is the generation of ultrashort laser pulses in a burst-mode supporting terahertz repetition rates (TRR), corresponding to an intraburst pulse spacing of only a few picoseconds, or less. It shows novel characteristics given by tunable spectral narrowbands in millijoule-level near-infrared and microjoule-level terahertz waveforms. A prominent application of TRR bursts is the use as a multi-pulse probe in pump-probe experiments, enabling single-shot analysis of ultrafast dynamics over a broad picosecond time-range. However, temporal characterization in the TRR burst-mode regime is limited, since it requires the resolution of femtosecond-scale features (the shape of individual burst pulses) over the picosecond burst range. In this work, we discuss the potential of a single-shot characterization technique for TRR bursts that is based on individual pulse amplitude and phase retrieval from spectral acquisition only. It is enabled in the case of sufficient similarity of pulses within a burst, reducing the requirement to the identification of peak amplitudes and relative carrier-envelope phases.

Phase-matching-free pulse characterization techniques are particularly important since they provide robust and direct access to ultrashort pulse information over a wide spectral range. In this work, we present a study on the fundamentals of the Plasma-Induced Frequency-Resolved Optical Switching (PI-FROSt) technique, focusing on the dependence of the obtained traces on the spatial characteristics of the system. We investigate this behaviour both numerically and experimentally. Our results show that the PI-FROSt traces strongly depend on the focusing conditions of both pump and probe pulses, as well as on the spatial region selected for signal detection. This spatial dependence is mainly attributed to the interaction between the plasma generated by the pump pulse and the probe pulse, being different for each configuration. Therefore, understanding and optimizing the system is crucial to obtain traces that accurately represent the pulse to be reconstructed, and to avoid distortions arising from the spatial response of the system rather than from the intrinsic properties of the pulse itself.

We have generated computationally plasma-mirror frequency resolved optical gating (PM-FROG) traces and the corresponding spectral electric fields for picosecond laser pulses in order to train an artificial neural network as a tool for reconstructing such pulses. The trained neural network works well on a test dataset, with low reconstruction errors. The neural network also predicts from a PM-FROG image the plasma rise time and the chirp related group delay dispersion and the third order dispersion.

A current hot topic on ultrafast pulse characterization is the diagnosis of ultrafast pulses including train instabilities, due to its significant impact on the laser source performance. Here, we present two approaches to characterize pulse trains with varying temporal profiles, based on the amplitude swing technique. First, we show a single-shot implementation, based on a pair of uniaxial wedges. Additionally, we can extract statistical information about the pulse train from the trace in the case of the scanning implementation. The single-shot solution is most suitable for high-power, low-repetition-rate lasers, while the statistical analysis is preferable for low-power, high-repetition-rate systems.