Investigation of temporal compression of few-cycle pulses from an ultrabroadband, multi-mJ optical parametric amplifier

Summary form only given. High-energy, ultrashort pulses in the few-cycle regime are of great interest as drivers of intense attosecond-pulse sources [1]. Noncollinear optical parametric chirped pulse amplification (OPCPA) is a promising technique to reach an ultrabroad amplification bandwidth and high pulse energies, simultaneously. Recently, in our laboratory a 3-stage OPCPA system delivering pulses with a bandwidth ranging from 670-1400 nm and a pulse energy of 1.8 mJ at 3 kHz repetition rate has been demonstrated [2]. This ultrabroadband amplified spectrum supports 4.3 fs pulses, with a central wavelength around 1 μτη. In this study we investigate the possibility of pulse compression by double-angle chirped mirrors [3].In order to compress the pulse to its transform limit, the chromatic dispersion introduced by the gain media and other dispersive components along the optical path needs to be compensated. In our system we take into account 3 mm FS, 2 mm of BaF2 (of a cross-polarized wave generation stage added to the setup described in [2]), 6 mm LBO, 2 mm BBO and several metres of air. We have calculated the resulting group delay (GD) and group delay dispersion (GDD) according to the Sellmeier equations. The resulting GD curve with opposite sign represents the target curve for the mirror design of the compressor, since the GD of the system and the double-angle chirped-mirror compressor add up linearly. In the double-angle chirped-mirror technique one multilayer design allows for the compensation of GDD oscillations by using reflections under two distinct angles of incidence. In this sense the mirrors have to be used in “pairs” in order to obtain the lowest possible oscillations. The result of the design in terms of the GDD for one such pair, i.e. one mirror used under 5^o and one used under 21^o angle of incidence, is shown in the central panel of Fig. 1. For such a broadband chirped mirror it is common- even after the compensation using the double-angle technique to have relatively high residual GDD oscillations. However, by analyzing the corresponding GD, which is obtained by numerically integrating the GDD curve once with respect to frequency, we can see that these large oscillations translate into relatively small ripples on the GD curve. This is shown on the left in Fig. 1 where the GD curves for four and five pairs are compared with the target curve. The good match between the design and the target together with the small resulting GD oscillation theoretically allow for a compression with this device to close to the transform limit of the amplified spectrum. This can be seen in Fig. 1 (right), where the compression of a supergaussian spectrum (670 - 1400 nm) to 4.6 fs has been calculated using the GD of 5 pairs. The experimental implementation of the compressor is currently under way, however the full characterization of such an ultrabroad spectrum with close to single-cycle pulse duration is far from straightforward. While in the past a second harmonic (SHG) FROG (frequency-resolved optical gating) was used, in the future we plan to implement a Transient-Grating (TG) FROG device [4], which in our case will overcome the significant bandwidth limitations and phase-matching constraints of the SHG-FROG.