Electron excitement in three dimensions.

Electron excitement in three dimensions

Just as light waves can be thought of a streams of particles (photons), particles such as electrons can be thought of as waves. And, like overlapping waves in which crests and troughs cancel or reinforce one another, waves associated with electrons create interference patterns.

That wave-like property of electrons is the basis for a recently proposed technique for determining the three-dimensional atomic structure of surfaces. The technique, called photoelectron holography, produces the electron-generated equivalent of the visible-light holograms now seen so often as a security feature on credit cards and in three-dimensional displays.

In photoelectron holography, a burst of finely tuned X-rays illuminates a small patch on a crystal surface. The X-rays excite atoms in the illuminated area, forcing them to release electrons. Some electrons travel directly to a nearby detector while others first rebound from neighboring atoms before completing their journey. Because the direct-flight and scattered electrons travel slightly different distances to reach the detector, the waves associated with the electrons produce an interference pattern. In the Sept. 19 PHYSICAL REVIEW LETTERS, John J. Barton of the IBM Thomas J. Watson Research Center in Yorktown Heights, N.Y., shows theoretically that if such a hologram can be obtained and measured, then this two-dimensional interference pattern can be mathematically converted into a three-dimensional image showing the locations of atoms on the crystal surface.

Unlike the scanning tunneling microscope, a widely used tool for tracing a surface's atomic bumps and hollows, photoelectron holography offers the possibility of not only showing atomic locations but also identifying the atoms. By tuning the illuminating X-rays to a particular wavelength, researchers force only atoms of a certain element to respond, producing a characteristic "signature" that identifies the emitting atoms and possibly the scattering atoms.

Because individual atoms yield only one electron, a photoelectron hologram is actually the sum of many events. "Because it takes many thousands of photoelectron events to create the hologram, we're really looking at the average [atomic position] over the illuminated area of the crystal," Barton says. The technique would work best with a single-crystal sample or a composite sample made up of a thin film laid down on a crystal surface. If the material is too disordered, then no clear picture of the surface can be reconstructed from the hologram.

Although photoelectron interference is a farily well-known phenomenon, until Barton's work no one had thought it useful to measure the full interference pattern in two dimensions. No one understood that the pattern was actually a hologram. "My paper shows that if you can make the measurements, then you can determine the structure," Barton says. "The real question comes down to: Is it feasible to make these measurements?" Barton and his colleagues are now attempting to build a suitable detector to test the idea.