Influence of the delay line jitter on the SHG FROG reconstruction

Dispersion scan frequency resolved optical gating for consistency check of pulse retrieval

Many methods commonly used to characterize ultrafast laser pulses, such as the frequency-resolved optical gating (FROG) or the dispersion scan (d-scan), face problems when they are used on pulses with a spectrum or phase varying within the laser beam cross section or the acquisition time. The presence of such pulse shape variation leads to discrepancy between the measured FROG trace and its reconstructed counterparts. Nevertheless, it is difficult to reliably discern this shape variation because even the distorted experimental FROG trace can be reasonably reproduced by a realistic pulse shape. In this work, we examine and discern the variation of the pulse shape based on a new method, dispersion-scan FROG (D-FROG), which combines the idea of dispersion scanning with the FROG method. This technique provides a means of careful evaluation of the laser pulse based on a set of FROG traces connected by known dispersion changes. Therefore, this method can disclose seemingly correct pulse retrievals from distorted datasets. The D-FROG method can be used as a simple extension of the FROG technique to provide a consistency check able to identify the shortcomings in the pulse characterization.

Targeted generation of complex temporal pulse profiles

A targeted shaping of complex femtosecond pulse waveforms and their characterization is essential for many spectroscopic applications. A 4f pulse shaper combined with an advanced pulse characterization technique should, in the idealized case, serve this purpose for an arbitrary pulse shape. This is, however, violated in the real experiment by many imperfections and limitations. Although the complex waveform generation has been studied in-depth, the comparison of the effects of various experimental factors on the actual pulse shape has stayed out of focus so far. In this paper, we present an experimental study on the targeted generation and retrieval of complex pulses by using two commonly-used techniques: spatial-light-modulator (SLM)-based 4f pulse shaper and second-harmonic generation frequency-resolved optical gating (FROG) and cross-correlation FROG (XFROG). By combining FROG and XFROG traces, we analyze the pulses with SLM-adjusted complex random phases ranging from simple to very complex waveforms. We demonstrate that the combination of FROG and XFROG ensures highly consistent pulse retrieval, irrespective of the used retrieval algorithm. This enabled us to evaluate the role of various experimental factors on the agreement between the simulated and actual pulse shape. The factors included the SLM pixelation, SLM pixel crosstalk, finite laser focal spot in the pulse shaper, or interference fringes induced by the SLM. In particular, we observe that including the SLM pixelation and crosstalk effect significantly improved the pulse shaping simulation. We demonstrate that the complete simulation can faithfully reproduce the pulse shape. Nevertheless, even in this case, the intensity of individual peaks differs between the retrieved and simulated pulses, typically by 10–20% of the peak value, with the mean standard deviation of 5–9% of the maximum pulse intensity. We discuss the potential sources of remaining discrepancies between the theoretically expected and experimentally retrieved pulse.