Skip to main content
. 2019 Jan 28;10:458. doi: 10.1038/s41467-019-08367-y

Fig. 2.

Fig. 2

Experimental setup. The MOPA generates a train of 18.4-MHz-repetition-rate, 250-fs pulses centred at 1030 nm. The output pulses are nonlinearly compressed to < 40 fs via spectral broadening upon multiple-pass self-phase modulation in fused silica (FS) followed by a chirped-mirror (CM) compressor, and coherently enhanced in a femtosecond enhancement cavity. HHG is driven in a 25 × 32-µm2-radius focus between two curved mirrors, one of which is pierced for on-axis XUV output coupling. IC: input coupler, PM: pierced mirror. Intracavity pulse diagnostics: autocorrelator (AC), optical spectrum analyser and power meter. The output-coupled XUV can either be sent to an XUV spectrometer or to the PES experiment. A 96%-NIR-transmission beam splitter (BS) can be used to separate the NIR radiation from the co-propagating XUV beam. Alternatively, the BS can be exchanged with an Au mirror, resulting in 1011 W/cm2 NIR intensity on the sample for laser-dressed PES. In the ultra-high-vacuum PES chamber (<10−9 mbar) the XUV beam is focused onto the tungsten (110) crystal with a 250-mm radius-of-curvature multilayer (scandium-silicon) double mirror. The inner segment of the mirror can be delayed with a piezoelectric-ceramic stage. A moveable on-axis aluminium filter can optionally block the NIR otherwise impinging on the inner mirror segment. Fine-tuning the NIR power is achieved with a variable aperture. The electrons are detected by a time-of-flight spectrometer equipped with a retarding grid. Two operation modes are available, the drift mode without any manipulation of the electron trajectory and the low-angular-dispersion mode with capture in a larger solid angle (see Methods)