Researchers across Europe including Professor John Costello from Dublin City University’s School of Physical Sciences, have dramatically enhanced the achievable time resolution at free-electron lasers with a new technique. 

A laser that operates at X-ray wavelengths became the goal of physicists and engineers almost as soon as the first laser was invented in 1960. The vision was that one could combine the unique properties of X-rays with the unique features of lasers to build a tool that could revolutionise areas like molecular medicine.

The goal was achieved about a decade ago when the first X-ray free-electron laser (XFEL) was announced at Stanford. The vision, single molecule sensitive imaging for revolutionary developments in health and medicine, is still a work in progress.

XFELs produce pulses that are typically a few femtoseconds (million-billionth of a second) in duration at wavelengths of 10 picometers or so. These are exactly the time and space scales of atoms in molecules, ideal for making molecular movies.

Auger electrons as cancer therapeutic

But there is a big problem – called Auger or secondary electrons. They are created by the X-ray pulses and can damage the biomolecule before you get a chance to properly image it. Work on Auger electrons as a cancer therapeutic is ongoing globally.

To get around the molecular damage problem, physicists and engineers have been working on making the XFEL X-ray pulses so short in time that the molecule is photographed before it suffers radiation damage.

To make it work, one needs to detect how many ‘femtoseconds’ elapse, after the X-ray pulse impinges on the biomolecule, before the secondary electrons are emitted (a time delay that is specific to each atom/molecule).

In that way one can figure out strategies to minimise (or even eliminate) radiation damage in molecular radiobiology. However, as is often the case in research, to every solution there is a further problem.

Miniscule timescales

On these miniscule timescales (one millionth of a billionth of a second), it is extremely difficult to synchronise the X-ray pulse (or flash) that illuminates the molecule with a separate external laser that is used for detection of the secondary electrons.

That timing jitter between the X-ray and external laser pulses can be hundreds of times bigger than the delay you want to measure, thereby blurring the measurement and molecular images. 

Now, a large international research team involving collaborators from the Max Planck Institute for Structural Dynamics (MPSD) and DESY in Hamburg, Dublin City University, and many other institutions spread across seven countries has found a method to get around this problem at XFELs and demonstrated its efficacy by measuring a fundamental Auger decay process in neon gas.

To circumvent the jitter problem, the research team came up with a pioneering, highly precise approach and used it to chart the time evolution of Auger decay.

The technique, dubbed self-referenced attosecond streaking, is based on mapping the electrons in thousands of images and deducing when they were emitted based on global trends in the data.

Neon gas sample

For the first application of this method, the team used a simple neon gas sample, where the decay timings have been inferred from quantum theory and other experiments.

After exposing both types of emitted electrons (initial and Auger) to a ‘streaking’ laser pulse, the researchers determined their final kinetic energy in each of tens of thousands of individual measurements. Crucially, in each measurement, the Auger electrons always interact with the streaking laser pulse slightly later than the photoelectrons.  

This constant phase difference forms the foundation of the technique. By combining so many individual observations, the team was able to construct a detailed map of the physical process, and thereby determine the characteristic time delay between the initial (photo-) and Auger- electron emission. 

X-ray ionisation

Lead author Dan Haynes, a doctoral student at the MPSD, said: “Self-referenced streaking enabled us to measure the delay between X-ray ionisation and Auger emission in neon gas with sub-femtosecond precision, even though the timing jitter during the experiment was in the hundred-femtosecond range.

"It’s like trying to photograph the end of a race when the camera shutter might activate at any moment in the final ten seconds.”

DCU collaborator Prof John Costello said: “The trick lies in being able to run the race lots of times knowing that the two horses that you bet on will always make it to the finish line with the same delay between them. That means you have a calibration scale to measure the resulting timing jitter and compensate for it.” 

In addition, the measurements revealed that the photoionisation and the subsequent relaxation and Auger decay must be treated as a single unified quantum process rather than a two-step process in the theoretical description of Auger decay. In previous time-resolved studies, Auger decay had been modelled in a semi-classical manner.

However, under the conditions present in these measurements at LCLS, and at XFELs generally, this model was found to be inadequate.

Instead, Andrey Kazansky and Nikolay Kabachnik, the collaborating theorists on the project, applied a fully quantum-mechanical model to determine the fundamental Auger decay lifetime from the experimentally observed delay between ionisation and Auger emission. 

The researchers are hopeful that self-referenced streaking will have a broader impact in the field of ultrafast science. Essentially, the technique enables traditional attosecond streaking spectroscopy, previously restricted to tabletop sources, to be extended to XFELs worldwide as they approach the attosecond frontier.

In this way, self-referenced streaking may facilitate a new class of experiments benefitting from the flexibility and extreme intensity of XFELs without compromising on time resolution. 

You can read the full paper here.

(Main image: Artistic depiction of the experiment. The inherent delay between the emission of the two types of electron leads to a characteristic ellipse in the analysed data. In principle, the position of individual data points around the ellipse can be read like the hands of a clock to reveal the precise timing of the dynamical processes. Image: © Daniel Haynes/ Jörg Harms)