Compact and high-performance, like a Swiss Army knife

SwissFEL, the new large research facility at PSI, has been in regular operation since January 2019. And it turns out that this X-ray free-electron laser really is as compact, high-performance, and versatile as planned. "The Swiss Army knife of FELs" is what PSI researcher Henrik Lemke likes to call this large research facility. In an interview, he talks about the latest research successes.

Henrik Lemke, research group leader at PSI, at the Bernina experiment station of SwissFEL.
(Photo: Paul Scherrer Institute/Markus Fischer)

Mr. Lemke, SwissFEL has been running in regular operation for a good two years now, which means that researchers from PSI and other institutions can conduct experiments. How is it going?

Henrik Lemke: I'm very satisfied. It's going well! As is usual with large research facilities, you don't just flip a switch on day one and operate at full power. Initially, SwissFEL too started up with less energy, longer pulses, and a lower repetition rate. But now we have achieved the values for which the system was designed in almost all parameters.

What values are these and what, for example, does "energy" represent?

By energy we usually mean that of the X-ray laser light. The following applies: The higher the energy, the deeper the radiation can penetrate into materials or even pass through them. The most energetic radiation that we can now achieve at SwissFEL is 12.4 kilo electron volts. With such high energy, one also speaks of "hard" X-rays. However, SwissFEL has two beamlines: Aramis, which was built first, delivers this hard X-ray radiation but can also produce radiation with lower energy, so-called "tender" X-rays. The new Athos beamline, on the other hand, which went into operation in 2020, specialises in "soft" X-ray light of even lower energy. This is a distinctive feature of SwissFEL: We cover everything from hard to soft X-rays without any gaps. And, in contrast to SLS, in the form of very short X-ray light pulses.

We'll come back to the short pulses. Why is the range of energy important?

Because you have to approach different samples and questions with different excitation energies. By using the right energy in each case, you can obtain very important information about materials. We at SwissFEL can therefore say: No matter what sample researchers bring to us, we have the right X-ray light ready for almost everything.

You are a scientist at the SwissFEL experiment station Bernina, which specialises in materials research. What have you been investigating there lately?

Two scientific papers, on very different experiments, just got published. One of those experiments we did end of January 2019; it was the first user experiment on Bernina. We examined one of the most universal materials: water.


Yes. The research group that came up with this idea is from the Swedish capital Stockholm. It was about the specific heat of water at low temperatures. Water has strong so-called anomalies. In everyday life we see this in the fact that the solid state is not heavier than the liquid, but lighter: That's why ice floats on liquid water. These anomalies are not yet fully understood, and that is the detail this experiment probed. In addition, of course, water is ubiquitous – which is why knowing the properties of water really well and with precision is relevant for many research areas.

Wasn't this study somewhat atypical for your experiment station?

It was in fact not the kind of experiment we do every day. But we have great flexibility at Bernina to implement different experiments. In this case we were able to install a vacuum chamber provided by the Swedish group, which was a round thing about 30 centimetres in diameter. Water droplets were injected into it. While these evaporate in the vacuum, the evaporative cooling chills the shrinking droplets. Because we used completely clean water without crystalisation nuclei, each drop remains liquid well below zero degrees – it is in the so-called no man's land of water. We now first warmed up the drop minimally with a short infrared laser pulse and then measured it with the X-ray light of SwissFEL. From this we get a diffuse scattering image and can see what mean atomic arrangement prevails immediately after this warming – that is, what structure the water has.

And what did you find out?

With this method we determined the specific heat capacity, the ability of water to store heat. This heat capacity showed a sudden change in the range of minus 44 degrees Celsius – with the water still liquid, as I said. This indicates that a structural change is taking place here. This enables other molecular vibrations; that is, from this temperature on, the water molecules have new possibilities to vibrate. The Swedish research group had already observed such a structural change in previous experiments using other facilities. What is new is that we have also seen this structural transition in the thermal properties and can therefore describe it better.

So, a complete success?

Yes, because with this experiment we also succeeded in making a technically demanding extension of a known measurement. That certainly helped to establish SwissFEL in the scientific community.

And the second experiment you mentioned earlier?

That was more typical for us. The sample was a solid, albeit in powder form: it was crystalline nanoparticles. We invited researchers from Rennes in France to join us for this study. Together we investigated how the material – a titanium oxide – can be switched from a semiconductor state into an electrically conductive state by laser pulses – and back again with other laser pulses. We already knew that this was possible in principle. With the help of SwissFEL, however, we were now able to see how this change in state propagates within the granules: The laser heats the material selectively, this heat leads to an expansion of the material, and this in turn triggers a shock wave that travels through the nanoparticle.

What follows from this insight?

These materials might one day be used for computer memories, microswitches, or the like, where speed is of the essence. Our measurements have now shown: It is not only the heat – which spreads rather slowly – that causes the change of state; instead, it is mainly this shock wave. A shock wave, in turn, is an acoustic pulse that propagates through the material at the speed of sound. It's a bit like a guitar string being struck. Although our experiment was basic research, it is precisely here that it is easy to imagine how the knowledge we have acquired might help in the development of new applications.

Getting back to SwissFEL: At the beginning we talked about the energy of the X-ray pulses, this value of 12 kilo electron volts. What other key figures are relevant, and what is the status?

The so-called repetition rate is also relevant, how many laser pulses per second we can get SwissFEL to produce. Since in most experiments one pulse delivers exactly one data point, we naturally get more data in less time if we increase the repetition rate. So if you don't want to measure for days, a high repetition rate is advantageous. In the early summer of 2020, after the first limited operations phase that PSI went into during the first wave of the coronavirus pandemic, we dared for the first time to go to 100 pulses per second. That worked, and we've been measuring with it ever since. This is the maximum repetition rate for which SwissFEL is designed.

And the length of each pulse?

That's another very important metric. Generally we can say: that the shorter the pulses, the better. At first, our experiment station wasn't perfect: There was still a problem with the timing diagnostics. While PSI was in limited operations in 2020, we had the opportunity to set up a new diagnostic system. Thanks to this component, we can now routinely achieve a time resolution of around 50 femtoseconds. Fortunately, this was not yet needed for the experiments I mentioned earlier. But an important goal of SwissFEL has always been that we want to investigate ultrafast processes here. And that is only possible with the high temporal resolution we now have.

Is that as far as you can push it?

We hope to push it even further. With the pulse lengths of SwissFEL and the stimulating lasers, we want at least to get down to ten femtoseconds. For the other two parameters mentioned, we have already achieved the values for which SwissFEL is designed. But we may be able to achieve even better figures in the coming years with optimisations. This would be remarkable, since SwissFEL is a rather compact facility in comparison to other X-ray free-electron lasers worldwide.

You call a building 740 metres long "compact"?

Other FELs are actually even bigger – and therefore more expensive. This makes it easier for them to achieve high energies in the X-ray light. But in spite of that, we are not doing badly in a global comparison. With our high-precision arrangement of undulators, for example, we make up for a lot and achieve key figures that position us right on the heels of the others. We must not forget: FELs are always national or even multinational efforts. As a comparatively small country, we got SwissFEL up and running on our own, and we set it up in the forest of Würenlingen in such a way that it hardly disturbs nature. SwissFEL is, so to speak, the Swiss Army knife of X-ray free-electron lasers. Compact, flexible, and high-performance for its size.

Interview: Paul Scherrer Institute/Laura Hennemann

Further information

Research and tinkering – SwissFEL in 2019 – a text from 29 August 2019


Dr. Henrik Lemke
Head of the SwissFEL Bernina Research Group
Paul Scherrer Institute, Forschungsstrasse 111, 5232 Villigen PSI, Switzerland
Telephone: +41 56 310 49 82, e-mail: [German, English]

Original publication

Enhancement and maximum in the isobaric specific-heat capacity measurements of deeply supercooled water using ultrafast calorimetry
H. Pathak et al.
PNAS, 9. February 2021
DOI: 10.1073/pnas.2018379118(link is external)

Strain wave pathway to semiconductor-to-metal transition revealed by time-resolved X-ray powder diffraction
C. Mariette et al.
Nature Communications, 23. February 2021
DOI: 10.1038/s41467-021-21316-y


PSI provides image and/or video material free of charge for media coverage of the content of the above text. Use of this material for other purposes is not permitted. This also  includes the transfer of the image and video material into databases as well as sale by third parties.