Abstract

A nonlinear phaseshift can be sensed by a laser pulse during Second Harmonic Generation (SHG) in a nonlinear crystal [1]. The induced nonlinear index n₂(∆k) is due to a cascade of two nonlinear phenomena ( SHG and difference frequency mixing). The sign and amplitude of n₂(∆k) can be controlled by the phase mismatch ∆k = k(2ω) − 2k(ω) between the fundamental −k(ω)- and the harmonic −k(2ω)- wave vectors propagating in the crystal. This phenomenon makes its possible to compensate for spatial and temporal nonlinear phase shift occurring in the optical elements used in amplification chains. In femtosecond range, the influence of cascading phenomena in the spatial domain has been measured by z-scan technique [2] or by direct compensation for positive Kerr phase shift with negative phases generated by SHG in a nonlinear crystal [3]. While these measurements are demonstrative, they are not quantitative. For instance, one may wonder about the limitations imposed by nonlinear temporal and transverse effect discussed previously by different authors [4, 6] and so far experimentally overlooked. Therefore, for an implementation in a real laser amplification chain, one needs to measure accurately both spectral and spatial phase modulation induced by n₂(∆k) during SHG. Hereafter, in a pump probe experiment, we demonstrate that spectral interferometry as well as Shack-Hartmann wavefront analyzer are very well suited techniques for measuring precisely the spectral phase shift or the spatial distortions produced by n₂(∆k) during SHG of femtosecond pulses. Contrary to what is generally admitted, our experiment allows us to demonstrate that at ∆k=0 a spectral and spatial distortion of a femtosecond laser pulse occurs and can be explained.

© 2002 Optical Society of America