Abstract

Parallel to the rising of a number of scientific and industrial applications of ultrashort laser pulses, their precise temporal characterization has become a major issue. Recently, the new pulse measurement technique of dispersion-scan (d-scan) was introduced, which is based on measuring the spectrum of a nonlinear signal as a function of varying amounts of known dispersion added to the pulse [1]. A reconstruction algorithm then enables extracting the amplitude and/or the spectral phase of the pulse from the measured two-dimensional (nonlinear spectrum vs. dispersion) trace. The first implementation of the technique used chirped mirrors and a pair of wedges that introduce varying amounts of dispersion depending on how much wedge is inserted or removed into the beam path in a scanning sequence. This implementation is very well suited for very short pulses (i.e., with broad bandwidths and high sensitivity to dispersion) being capable of measuring even single-cycle pulses [2-4]. On the other hand, longer pulses (i.e., with narrower spectra) present lower sensitivity to dispersion and the wedge setup is not enough to perform a complete d-scan, so a more complex (and expensive) calibrated dispersion scanning module is needed, e.g., a prism/grating/grism compressor or an acousto-optical programmable dispersive filter (AOPDF).

© 2017 IEEE