Developing detectors to transform science with light (part 1)

“Our hope is that our detector developments help in solving some of the grand challenges in areas such as energy, health or the development of new medicines: not only the general advancement of science, but that we enable something that gives a tangible benefit for society,” says Bernd Schmitt.

In fact, the detector group at PSI, which Bernd has led since 2008, have already realised this goal. Their detectors have enabled numerous discoveries with practical impact at synchrotrons and X-ray free-electron lasers throughout the world, ranging from Nobel Prize winning protein structures of biomedical significance to insights into chemical processes or materials.

Detectors effectively determine what is possible with X-ray light sources. And it is from the experimental needs of scientists at PSI that these transformative technologies start. “In the beginning, there’s typically a clear need for a detector that is unmet by existing technology,” says Anna Bergamaschi, who joined the detector group in 2005 and since 2021 leads the group, ‘Detector Science and Characterisation’.

Bernd explains how their solutions go on to advance X-ray science not just at PSI, but at light sources around the world: “Once new detectors are in use here and work well, people come and measure with them. They go somewhere else, and pressure increases at other synchrotrons to have something similar.”

From particle physics to synchrotrons

Bernd Schmitt first joined PSI in 1999, where he focused on the development of hybrid pixel X-ray detectors – a radically different type of detector first developed for the Swiss Light Source SLS. These detectors are essentially an array of hundreds of thousands, or even millions of individual detectors, each one consisting of a silicon sensor sandwiched to readout electronics.

This breakthrough came from their colleagues in the High Energy Physics group at PSI, who were developing the technology to trace particles spraying out from collisions in the Large Hadron Collider at CERN (see November 2022 Scientific Highlight about the upgrade of the barrel pixel detector).

At the time, X-ray detector scientists at PSI were looking for solutions to take advantage of the high brightness at protein crystallography beamlines. From this technology, they developed the first single photon counting detectors dedicated to photon science. These could accurately record the intensity of arriving photons across a high dynamic range with the sensitivity to detect single photons. These qualities together with higher speed opened the door to science with light that could previously only be dreamed of: first in the field of protein crystallography, and soon after, in all fields. With the success of the first of these detectors, Pilatus, the company Dectris was founded as a PSI spin-off.

New detectors, new experimental possibilities

Today, single-photon detectors as developed by PSI and marketed by Dectris are found in synchrotrons throughout the world - a fact that Anna, Bernd and their teams are understandably proud of.

Bernd reflects: “Single photon counting detectors made such a difference, that the CCD detectors that were used at the time couldn’t compete. New types of experiments became possible and, in the end, everyone expected to be able to do these, not just at SLS, but everywhere.”

From their first hybrid pixel detectors, their portfolio has blossomed into a diverse array, all named after Swiss mountains and always driven by the experimental needs.

“There are so many possibilities and we are so limited in our manpower that we have to select carefully the right development,” Bernd says. “It must be needed, it must be useful, it must have a big effect on the experimental usage at the endstation and we must be able to do it. So, there are many boundary conditions that we must consider.”

From synchrotrons to free electron lasers

Depending on the complexity of a project, Bernd points out, a new detector development can easily take five to ten years. Given this, Bernd, Anna and their teams must anticipate future requirements. In the mid-2000s, with the rise of X-ray free electron lasers - and first discussions of one day building a Swiss one, the detector scientists knew they would need to gain experience in XFEL detector development.

In 2007, nine years before SwissFEL would switch on its light, they entered the field of XFEL detector development with two projects: first the Gotthard detector for one of the world’s first X-ray free electron lasers, FLASH at DESY, and then the detector AGIPD for the planned European XFEL in Hamburg. AGIPD was a highly ambitious project, which took 10 years and involved collaborators from University of Bonn, University of Hamburg, European XFEL and DESY in addition to the PSI team.

XFELs brought new detector requirements. The single photon counting detectors that were so successful in synchrotrons were not appropriate for free electron lasers, where, due to the pulsed light, bunches of photons arrive within a few femtoseconds. Detectors were needed that still had the sensitivity to detect single photons whilst offering the dynamic range to cope with ten thousand photons arriving simultaneously at one pixel.

Their solution to this problem was a concept called dynamic gain switching, which was used both in Gotthard and AGIPD. Depending on the number of incoming photons, the gain-level automatically adapts, switching from a high-gain to detect single photons to lower gain when there are more photons – thus giving both the sensitivity and large dynamic range required.

A new detector for a new facility

Around 2012, with the SwissFEL project underway, the detector group made an important choice. Rather than adapting one of these detectors, they decided to embark on a new and ambitious project: Jungfrau.

 “We thought about using AGIPD for SwissFEL. But it didn’t quite fit,” remembers Bernd. “The pixel size was one problem. The energy range also didn’t match. AGIPD had higher noise than we wanted and couldn’t go as low in energy as needed for SwissFEL. Then in other ways, it would’ve been overkill, because certain aspects just weren’t needed for SwissFEL.”

He reflects: “We built a new detector, but Jungfrau’s success stems from all our prior developments. Jungfrau is more closely related in chip design to Gotthard, but our experience with AGIPD – arguably one of our most complex projects yet - made it what it is.”

Perfectly tailor-made for the hard X-ray beamline of SwissFEL, Aramis, Jungfrau provides single photon resolution down to a lower energy than their previous XFEL detectors together with a linear dynamic range over four orders of magnitude – achieved through the automatic gain switching principle they first developed for Gotthard and AGIPD.

The ubiquitous Jungfrau

Today, Jungfrau is not only found at SwissFEL, but at XFELs throughout the world. “They have them at LCLS at SLAC, at the XFEL PAL in Pohang, now at the EU XFEL – they’re really everywhere,” says Anna proudly.

Of course, with around one hundred modules – equating to approximately fifty million pixels, and the detector group on-site to make them to specification and at cost, SwissFEL is better equipped with these detectors than any other XFEL in the world. “Every other place in the world would be happy to have half as many!” Bernd adds.

Jungfrau was developed for the hard X-ray beamline of SwissFEL, Aramis. But what about the soft X-ray beamline Athos? Read part II to see how the detector evolution continues.

Text: Paul Scherrer Institute / Miriam Arrell