Berkeley Lab Science Beat Magazine banner Berkeley Lab logo
July 14, 2004
The Flash Dance of Attosecond X-Rays
Chemistry starts with the movement of electrons, a motion that takes place in a matter of attoseconds — a timescale almost too small to comprehend. An attosecond is one billionth of a billionth of a second, and there are more attoseconds in a minute than there have been minutes in the history of the universe.
A marriage of photons and relativistic electrons will result in record-smashing x-ray pulses, each just 100 billionths of a billionth of a second long.

With a flash of light to stimulate an electron and attosecond x-ray flashes to follow its activities, scientists could directly observe such phenomena as an atom becoming ionized, or the bonding of two or more atoms into molecules. Sound like a technology for the far-distant future? Not according to a pair of researchers at Lawrence Berkeley National Laboratory.

Alexander Zholents and William Fawley, physicists at Berkeley Lab's Center for Beam Physics in the Accelerator and Fusion Research Division, have an idea for creating intense bursts of x-rays in pulse lengths of about 100 attoseconds. If you picture Niels Bohr's classic 1913 model of a hydrogen atom, it takes about 100 attoseconds for the electron to orbit the proton.

"What made our idea possible is the amazing work that has been done with lasers over the past five years," says Zholents. "Lasers can now provide the intense, few-cycle pulses that we can use to effectively slice attosecond soft x-rays from an ultra-relativistic electron beam."

Says Fawley, "Ours is a multistage process in which the first stage is the key. We use femtosecond pulses of light from an optical laser" — a femtosecond is one quadrillionth of a second — "to produce energy modulations in an electron beam. From these modulations, we can select energy peaks to interact with femtosecond pulses of coherent x-rays through a series of magnetic devices to ultimately yield attosecond x-rays."

The process, called seeded attosecond radiation, is the result of putting photons and electrons in harmony. Sending a relativistic beam of electrons through a magnetic wiggler or undulator oscillates the motion of the speeding electrons, causing them to lose energy in the form of light emission. Under the right conditions, sending a pulsed beam of photons through the wiggler or undulator at the same time as the electron beam will modulate the energy loss (light emission) of the oscillating electrons.

This makes it possible to select and spatially separate electrons at a desired energy peak from the rest of the electron beam. These electrons can then be collected and bunched into successively shorter pulse lengths as the process is repeated.

Under the proposed scheme of Zholents and Fawley, femtosecond x-ray photons and relativistic electrons will first be sent through a wiggler. Selected electron bunches will subsequently be sent through a long undulator, along with more femtosecond x-ray pulses. The ultrashort electron bunches that emerge from the undulator will then be used to generate attosecond x-rays.

"We start the process with 2-picosecond electron bunch lengths" — a picosecond is one trillionth of a second — "from which we generate 2-nanometer wavelength x-rays" — a nanometer is one billionth of a meter — "at about 100-femtosecond pulse lengths. We reduce this to 2-nanometer wavelengths at about 10 femtoseconds, and finally get about 1-nanometer wavelength x-rays at 100 attoseconds," says Zholents. "Attosecond pulses at other wavelengths could be produced with this technique, and we believe that pulse lengths shorter than 100 attoseconds are also possible."

A critical issue for the study of chemical reactions is whether a beam of laser light used to trigger the reaction — called the optical pump — can be synchronized with the x-ray probe pulses. Says Fawley, "In principle, the synchronization between the pump and probe pulses in our proposed technique is perfect, because both pulses are originated by the same source."

William Fawley, left, and Alexander Zholents designed the method of extracting 100-attosecond pulses from a free-electron laser. (Photo Roy Kaltschmidt)

While the proposal by Zholents and Fawley could be realized through an independent facility, the most cost-effective means of achieving it would be as an add-on to a facility such as Berkeley Lab's proposed LUX (linac-based ultrafast x-ray source). LUX would be a recirculating superconducting linear accelerator optimized for the production of substantial fluxes of low energy or "soft" x-rays at pulse lengths ranging from 10 to 200 femtoseconds.

Says Zholents, "Our technique could be incorporated into the LUX proposal with no significant change or improvement required for the way that facility is designed to operate. We purposely chose the same electron beam and x-ray light parameters as are currently planned for LUX." Zholents and Fawley used the computational power of the National Energy Research Scientific Computing Center (NERSC) to run rapid simulations that enabled them to refine and interactively improve their design.

Even from the perspective of a researcher at the forefront of the scientific drive to measure the smallest space or the shortest time, Zholents finds the attosecond time scale tough to truly comprehend.

"When I think about the shortness of 100 attoseconds, I try to compare two objects," he says. "I know that it takes about one second for light to cover the distance between the earth and the moon, and I know this is a tremendous speed, because I can see how far away the moon is. I also know how thin a human hair is — and yet, in 100 attoseconds, light will pass only one one-thousandth of the way through it!"

Additional information