SLAC Today logo

Need Ultrashort X-ray Pulses? Try Aluminum Foil

Slotted foil: The open "V" (bottom) allows one bunch of electrons through; the double-slot "V"s (middle and top) create two narrow bunches. The width of the V determines the bunch width (single bunch) or timing between bunches (double-slot). (Image courtesy Paul Emma.)

Last week, users on the Atomic, Molecular and Optical science instrument at the Linac Coherent Light Source experimented with a novel piece of technology that aims to make the world's quickest X-ray pulses even quicker.

Some of nature's most thrilling mysteries—such as the processes behind the forming and breaking of chemical bonds—occur in a matter of femtoseconds, or millionths of one billionth of one second. But capturing images of these events requires a camera with a flash and shutter speed on the same fleeting time scale. The LCLS was built, in part, to provide pulses of X-rays bright enough and quick enough to do just that.

"The reaction time of atoms can be less than 10 femtoseconds, which you can't see with longer pulses," said SLAC accelerator physicist Paul Emma.

LCLS pulses were designed to last about 100 femtoseconds, roughly the time it takes light to travel the width of a human hair. The lengths of the X-ray pulses are dictated by the size of electron bunches accelerating down the linac. After the electrons reach a given energy, they travel through a series of magnets that cause them to wiggle and release radiation in the form of X-ray photons. The pulses can be reduced to 60-70 femtoseconds by packing the electrons in each bunch closer together, but even these ultrashort bursts are too long for the quickest of atomic activities.

"Our goal was to generate pulses short enough to see these movements," Emma said. In 2003, three years before construction had even begun on the LCLS, Emma and his colleagues came up with a simple way to shorten the electron pulses to as little as one femtosecond: slice the beam.

Shortly after being injected into the linac, electron bunches reach a three-sided detour known as a bunch compressor, where they are compacted into an even shorter bunch. At one point along this detour, each bunch is tilted so that the electrons are spread out perpendicularly to their direction of travel, instead of single file.

(Image - electrons traveling through a bunch compressor)
The foil (yellow) slices each electron bunch (pink) while it is diverted into an upright shape inside a bunch compressor. The green triangles represent magnets that redirect the electrons' path of travel. (Image courtesy Paul Emma.)

Emma and his team suggested inserting a thin piece of slotted aluminum foil at exactly this point, creating a barrier with a narrow door in the middle. Only the electrons with a direct path through the door contribute to the laser pulse, while the electrons that hit the foil are scattered. The tapered slot is 2.2 millimeters wide at the top and narrows to about 220 microns at the bottom. Sliding the foil vertically through the beam dictates the number of electrons allowed to pass through the foil without scattering; the fewer electrons, the shorter the resulting X-ray pulse. The team predicted that, at its narrowest point, the slotted foil could effectively cut a 100-femtosecond pulse down to about one femtosecond.

Through computer simulations that took into account many other parameters of the free-electron laser, the group showed that, in theory, the slotted foil worked. They published their results in a 2004 Physical Review Letters paper.

But it wasn't until last week that the idea was tested experimentally. A group of SLAC physicists, including Clive Field, Mark Petree and David Kharakh, built and installed a prototype foil in early June; user experiments led by Reinhard Kienberger from the Technische Universitaet Munich, together with data collected previously by LCLS staff scientists, could determine whether the foil is, in fact, producing one-femtosecond X-ray pulses.

"We know we can get pulses down to one femtosecond, and maybe even in the sub-femtosecond region," said Michael Meyer, a scientist from the European XFEL in Hamburg, who is also involved in the research. "What we don't have is a way to measure them." The researchers address this challenge by striking neon atoms with photons and measuring the shift in kinetic energy, which is a direct reflection of the pulse duration.

It will take months for the group to comb through the thousands of snapshots taken, but so far what they have seen seems to agree with Emma's computer-simulated data, Meyer said. One goal of the analysis is to establish a relationship between the slot width and the resulting pulse width so that the technique can be applied to studies looking at specific atomic and molecular behaviors. In addition to allowing much better time resolution, the short pulses provide an advantage in that they don't destroy the very atoms they are designed to image, as do the longer, more intense pulses.

Future studies will also look at a double slot configuration. While the single slot is a hollow "V" etched into the foil, the double slot design will have either 120-micron or 250-micron-wide slots along the edges of the "V." The separation between the slots will serve as a tunable time delay between two ultra-short pulses, which will be useful for femtosecond pump-probe experiments.

Emma also envisions "building a new type of foil with more features," which include a smaller point at the tapered end of the slot, as well as a setup that allows the slotted foil to move horizontally across an electron bunch rather than vertically, to investigate whether particular sections of the bunch lase differently. 

—Lauren Rugani
SLAC Today, July 27, 2010