man in lab with laser

Stephan Heinrich, a coauthor on the new study, works on the laser setup used in the experiments. [Image: Thorsten Naeser, Max-Planck-Institut für Quantenoptik]

In recent years, technologies to generate extreme-ultraviolet (XUV) light pulses with widths on the order of attoseconds (10–18 seconds) have opened a viewport into the unspeakably fast dynamics of atom-scale electron motion, via time- and angle-resolved photoelectron spectroscopy (PES). But while they peer into inconceivably short timescales, the PES experiments themselves are often glacially slow to carry out.

That’s because of the need to carefully control the flux of photoelectrons generated by the pulse. Too few photoelectrons generated per pulse will hammer down the signal-to-noise ratio, requiring days or weeks of lab time to accumulate a large enough data set. But too many will crowd the electrons to the point where Coulomb repulsion between them will gum up the exquisitely precise angular measurements that PES requires—a phenomenon called the space-charge effect.

Now, researchers in Germany have devised a laser setup that they believe could get around these problems—and could shorten PES data acquisition times by as much as a thousand-fold (Nat. Commun., doi: 10.1038/s41467-019-08367-y). Specifically, the team has put together fiber laser technology, nonlinear pulse compression and cavity enhancement to create a source that can pump out attosecond pulses at a blazing repetition rate of 18.4 million pulses per second—and with the right energy and photon flux to produce copious photoelectrons for a good signal, but still avoid the space-charge effect.

The researchers think that the work could have implications not only for studies of atom-scale electron dynamics, but for XUV frequency-comb spectroscopy, nanoplasmonics and other areas.

Fast-wiggling electrons, slow experiments

Attosecond PES relies on XUV sources using the nonlinear phenomenon of high-harmonic generation (HHG). An intense pulsed laser field in the visible/near-infrared is fired into a gas, which both ionizes the electrons and causes them to “wiggle” with so-called ponderomotive energy in the oscillating laser field. When these fast-wiggling electrons crash back into the ground state, they emit pulses of attosecond light that include ultraviolet wavelengths (see “Sources and Science of Attosecond Light,” OPN, May 2015).

One factor holding back these ultrafast experiments lies in the combination of the relatively low, kHz-scale repetition rates of their driving lasers, and the need to avoid the space-charge effect. The latter effect, which can distort or degrade the precise measurements of electron kinetic energy and angular trajectory on which the PES experiments depend, requires that experimenters throttle back the output of HHG photons to the point where Coulomb repulsion effects are manageable. That requirement, coupled with the low repetition rates of the laser sources, means that it can take hours or tens of hours of collection time to amass sufficient data in a single experimental run.

These long experimental timescales are more than just an inconvenience. They are long enough also to allow laser-instability and sample-contamination issues to creep into the experiment, creating “severe technological challenges,” according to the authors of the new study.

Finding the right balance

The German research team—led by Ioachim Pupeza at the Max-Planck-Institut für Quantenoptik, and including scientists involved in the MEGAS Project, a collaboration between Max-Planck, Ludwig-Maximilians-Universität München and two of the Fraunhofer Institutes—sought to get around those problems. To do so, the researchers looked for a laser setup that could boost the repetition rate substantially to speed up PES experiments, with photon energies and fluxes just high enough to create copious photoelectrons without running afoul of the space-charge effect.

diagram of experimental setup

The team’s experimental setup for HHG creation of attosecond pulses at an 18.4-MHz repetition rate. A seed laser at near-infrared wavelengths, after pulse compression, amplification and cavity enhancement (red line), hits an argon gas cell to drive high-harmonic generation of XUV attosecond pulses (blue line), which fire upon a tungsten target and enable angle-resolved photoelectron spectroscopy via a time-of-flight detector setup. [Image: T. Saule et al., Nat. Commun., doi: 10.1038/s41467-019-08367-y; CC-BY 4.0]

The setup begins with a Ti:sapphire master oscillator that acts as a seed laser, churning out a train of 300-µW, 1030-nm pulses at a rate of 73.6 million pulses per second. After a preamp and pulse-picking stage, the seed pulses pass through three chirped-pulse amplification stages in a Yb-doped fiber amplifier, to create an 18.4-MHz train of 5.4-µJ, 250-fs pulses. A next step of nonlinear spectral broadening and chirp removal compresses the pulses to sub-40-fs length. Finally, the pulse train enters a bowtie-shaped enhancement cavity for a 35-fold power boost, and strikes an argon gas chamber for HHG of XUV attosecond light pulses.

A PES test

The team tested out this XUV light source in a PES experiment, firing the pulses from the HHG source in a 10-µm-diameter spot on a tungsten target, and measuring the photoelectron spectra in a time-of-flight (ToF) spectrometer setup. The ToF setup allows both electron momentum (or spatial distribution) and kinetic energy to be picked up at the same time, and can, according to the researchers, support duty cycles of up to 100 percent even at the light source’s high, 18.4-MHz pulse repetition rate.

The team found that each of the XUV attosecond pulses generated in the experiments contained around 105 photons, sufficient to create roughly 104 photoelectrons per pulse and to keep space-charge distortions at manageable levels of a few tens of meV. Further, the high pulse repetition rate meant that the target released some 1010 photoelectrons every second.

That high electron flux over repeated pulses, coupled with the ToF spectrometer setup, points to “a count rate improvement between two and three orders of magnitude over state-of-the-art attosecond photoelectron spectroscopy experiments under identical space charge conditions,” according to the team. And, the researchers believe, it could reduce the required measurement time in some electron-dynamics experiments by the same thousand-fold factor.

“This advance is of considerable significance for research on condensed-matter systems,” team leader Pupeza noted in a press release accompanying the research. “It also opens up new opportunities for the investigation of local electric fields in nanostructures, which are of great interest for applications in future information processing with lightwaves.”