Stability, beam profile, spectra and pulse duration of compressed pulse and dispersive wave

Temporal characterization of ultrafast pulse. Temporal characterization of pump pulses and pulses output from MPC and HCF compression stages was based on a home-built all-reflective split-mirror SHG-FROG set-up, based on a 10-μm-thick BBO crystal cut for type I phase-matching. Pulse characterization of dispersive-wave pulses at 250 nm and 400 nm was carried out using a home-built SD-FROG set-up. The optical path of SD-FROG is similar to SHG-FROG, with the main difference being 50-μm-thick fused silica replaced BBO crystal and used aluminum mirrors with high reflectivity for UV pulses. The raw traces were spectrally corrected using the frequency marginal and retrieved using a commercial software. The small FROG errors and excellent agreement between retrieved and measured spectra (Figs. 2, 3 and 6) confirm the reliability of the pulse measurement process.Measurement of laser beam profile. The beam profiles of the compressed pulses and the dispersive waves are both measured by a CCD camera (BGS-USB3-SP932U, Ophir-Spiricon). It should be pointed out that the measurements of dispersive-waves profiles are first filtered by the bandpass filters (Pelham Research Optical for 200-260 nm, Thorlabs for 260-400 nm, and LBTEK for >400 nm). The far-filed images are measured at the focus point of the lens with a focal length of 1 m.Measurements of laser spectrum. At a repetition rate of 25 kHz, we measure the pulse spectrum in the wavelength range of 200 nm to 300 nm using a UV spectrometer (Maya2000-Pro, 185–400 nm, Ocean Optics), in the wavelength range of 300 nm to 1000 nm using a ultraviolet-to-visible spectrometer (PG2000-Pro-Ex, 200–1100 nm, Ideaoptics), and in the wavelength range of 1000 nm to 1700 nm using a near-infrared spectrometer (NIR17S, 900–1700 nm, Ideaoptics). All three spectrometers have been calibrated and can be used to measure pulse spectra in a wide spectral range of 200-1700 nm. Splice the spectra measured by these three spectrometers, we finally obtained the full spectra of the HCF output pulses when the dispersive wave was generated at the wavelength of 250 nm, as shown in Figs. 5 and 6. Maya2000-Pro and HR4000CG (200–1100 nm, Ocean Optics) were used to measure the dispersive-wavs spectra at 100 kHz. It should be noted that some spectral measurements, dispersive waves with a wavelength range of >300 nm at 25 kHz and >230 nm at 100 kHz are first reflected by a fused silica to protect the spectrometer, and then the reflected light from the front surface is coupled to the fibre-coupled spectrometer connected with an integrating sphere.

超快脉冲的时间表征。针对抽运脉冲以及多通池（MPC）、空心光纤（HCF）压缩级输出的脉冲，我们采用自研搭建的全反射分镜式二次谐波产生-频率分辨光学快门（Second Harmonic Generation-Frequency Resolved Optical Gating, SHG-FROG）装置进行表征，该装置基于一块厚度为10 μm、切制为I类相位匹配的偏硼酸钡（BBO）晶体。针对波长250 nm与400 nm处的色散波脉冲，我们采用自研搭建的光谱色散-频率分辨光学门（SD-FROG）装置完成脉冲表征。SD-FROG的光路与SHG-FROG相似，主要区别在于以厚度50 μm的熔融石英替换了BBO晶体，并针对紫外脉冲采用了高反射率铝镜。我们利用频谱边际对原始FROG轨迹进行光谱校正，并通过商用软件完成脉冲重构。较低的FROG误差以及重构光谱与实测光谱间的极佳吻合度（见图2、图3与图6）验证了本次脉冲测量流程的可靠性。
激光光束轮廓测量。压缩脉冲与色散波的光束轮廓均通过电荷耦合器件（Charge-Coupled Device, CCD）相机（BGS-USB3-SP932U，Ophir-Spiricon）完成测量。需说明的是，色散波的轮廓测量需先经过带通滤光片滤波：200~260 nm波段采用Pelham Research Optical产品，260~400 nm波段采用Thorlabs产品，大于400 nm波段采用LBTEK产品。远场光斑图像在焦距为1 m的透镜焦点位置采集。
激光光谱测量。在重复频率为25 kHz的条件下，我们分别采用三款光谱仪完成不同波段的脉冲光谱测量：波长200~300 nm波段采用紫外光谱仪（Maya2000-Pro，185~400 nm，Ocean Optics），300~1000 nm波段采用紫外-可见光谱仪（PG2000-Pro-Ex，200~1100 nm，Ideaoptics），1000~1700 nm波段采用近红外光谱仪（NIR17S，900~1700 nm，Ideaoptics）。上述三款光谱仪均经过校准，可覆盖200~1700 nm的宽光谱范围进行脉冲光谱测量。将三台光谱仪测得的光谱进行拼接后，我们最终得到了当色散波在250 nm波长处产生时HCF输出脉冲的完整光谱，如图5与图6所示。在重复频率为100 kHz的条件下，我们采用Maya2000-Pro与HR4000CG（200~1100 nm，Ocean Optics）测量色散波光谱。需注意的是，部分光谱测量场景中：25 kHz下波长大于300 nm的色散波、100 kHz下波长大于230 nm的色散波，需先通过熔融石英片反射以保护光谱仪，随后将前表面反射光耦合至带有积分球的光纤耦合光谱仪中。