Solid-State Attosecond Measurements

Physicists at the Max Planck Institute fur Quantum Optics in Munich have previously looked at the behavior of electrons in atoms in the gas phase over a timescale of a hundred attoseconds (1 as=10^-18 second) or so. Now the scientists, led by Ferenc Krausz, have, in collaboration with their colleagues from Bielefeld, Hamburg, Vienna and San Sebastian, made a measurement of electron motion in a solid-state environment over a comparable timescale. The specific measurement-observing the difference in arrival times of electrons flying out of an atom struck by laser light-represents the sharpest time-resolution ever achieved in a condensed-matter experiment.

To bring about this feat, a near infrared (NIR) laser pulse consisting of only a few well-chosen cycles is sent through a column of neon, producing a number of secondary beams of shorter wavelength. One of these beams, at extreme ultraviolet (XUV) wavelengths, appears in very truncated bursts lasting only 300 as. Next, the XUV pulse is directed at a tungsten target where atoms lying close to surface can be ionized. Actually the ultraviolet light tends to liberate an out-lying (delocalized) electron from the atom as well as an inner-lying (localized) electron. These two electrons can proceed through the crystal and toward a detector where, depending on the time of their arrival, they can be told apart.

This identification process is enhanced in a clever way. Traveling co-linearly with the XUV pulse (and coherently linked to it) is part of the original NIR laser beam. The NIR intensity was carefully chosen so that it would not do the work of ionization (that task being assigned to the XUV light) but would be strong enough to accelerate the ionized electrons as they sprang out of the sample surface. The arrival of the NIR pulse with its well-controlled electric field was staged so that the first of the two electrons to appear (the faster-moving outer electron) would receive a boost in speed from the electric field of the NIR radiation, while the second electron (the slower inner-shell electron) received less of a boost. In other words the NIR light acted like a atomic-sized accelerator, speeding up the electrons, but in differing amounts. This accentuated the difference in the arrival times of the two electrons, making it easier to tell them apart.

The net result was an ability to measure the time delay of the two electrons coming across the top few layers of the solid-state sample. The measured interval, 110 attoseconds with an accuracy of 70 attoseconds, constitutes the unprecedented "attosecond" measurement. One of the researchers, Adrian Cavalieri (adrian.cavalieri@mpq.mpg.de) says that monitoring electron motions in a crystal with this level of precision is the first step in developing a much faster style of electronics, maybe even at a petahertz (10^15 Hz) rate. First comes measurement at 100-attosecond levels, later comes control of electron activity. (Cavalieri et al., Nature 25 October 2007; http://www.attoworld.de/)