Mechanism of hollow-core-fiber infrared-supercontinuum compression with bulk material

We numerically investigate the pulse compression mechanism in the infrared spectral range based on the successive action of nonlinear pulse propagation in a hollow-core fiber followed by linear propagation through bulk material. We found an excellent agreement of simulated pulse properties with experimental results at 1.8 $\mu$m in the two-optical-cycle regime close to the Fourier limit. In particular, the spectral phase asymmetry attributable to self-steepening combined with self-phase modulation is a necessary prerequisite for subsequent compensation by the phase introduced by glass material in the anomalous dispersion regime. The excellent agreement of the model enabled simulating pressure and wavelength tunability of sub-two cycles in the range from 1.5 to 4 $\mu$m with this cost-efficient and robust approach.

Figure 1

Experimental setup. Idler pulses from a high-energy OPA are propagated in a hollow-core fiber filled with 1.4 bar argon. The broadband pulses are subsequently compressed by linear propagation through a fused silica (FS) plate of 3 mm thickness.

Figure 2

Comparison between experimental (red circles) and theoretical (solid blue line) results. (a) Power spectrum for propagation in argon at pressure of 1.4 bar, (c) spectral phase of the pulse before and (d) after compensation with 3-mm FS. The temporal intensity profile of the compressed pulse is shown in (b).

Figure 3

Pulse compression for the spectrum of Fig. 2 by adding different amounts of FS. The color coding for the spectral phases in (a) corresponds to the temporal intensities in (b). Comparing the green with the black dashed plot demonstrates the effect of uncompensated TOD components.

Figure 4

Spectral evolution as a function of propagation distance for the simple case considering only $n_2$ in (a) and for the full model in (b). The latter clearly resembles an asymmetry because of self-steepening, which tends to promote the higher-frequency region.

Figure 5

Comparison of the three numerical models with experimental measurements. (a,b) Time profile of (a) the uncompressed and (b) the compressed pulses. (c) Broadened power spectrum and (d) its associated phase when the pulse is compressed.

Figure 6

Pressure dependence of (a) minimal pulse duration achievable by compression with optimized FS thickness in each case and (b) ratio between main and satellite pulse intensities.

Figure 7

(a) Temporal profile and (b) residual spectral phase of the pulse for a pressure of 2.6 bar without (dashed red line) and with (blue line) the best compression leading to a FWHM duration of 1.25 cycles.

Figure 8

(a) Temporal duration (in cycle units) at different wavelengths of pulses compressed with an optimal FS thickness. The best FS thicknesses are displayed in (b). Note that the CaF${}_2$ exit window is taken into account.