ACTIONS 1999INTERACTIONS 1999
Filming a Drama on a Very Small Stage
by Jim Hannon
 |
Jim Hannon received his Ph.D. in physics from the University of Pennsylvania in 1994, was a Humboldt fellow at the KFA in Julich, Germany then a post-doctoral researcher at Sandia National Laboratory. He joined the faculty of Carnegie Mellon as assistant professor in 1998. |
The stage is very small — 1 micron, or about one thirtieth of the thickness of a human hair. The actors are even smaller — individual atoms that move around on the surface of a material during the growth of a crystal. Under-standing how the plot unfolds, and hopefully directing the individual players, is key to the development of new materials for electronics, chemical sensors and magnetic applications. Measuring and eventually controlling the complex motion of atoms at surfaces is a challenge, but offers big payoffs for both science and technology. In many technological applications, nearly ‘perfect’ materials are required, with as many atoms as possible residing in their proper lattice positions. It is impractical to build up a material by placing the individual atoms, one by one, in their assigned places. A better approach is to make the atoms go where you want them to all by them-selves. To accomplish this goal, we need to understand the basic motivation of the characters, that is, how and why atoms move at surfaces.
Describing atomic motion at surfaces is also of fundamental scientific interest. How atoms move is closely related to how atoms bond to each other. As an atom ‘hops’ from site to site on a surface, bonds are continually broken and reformed. The rate at which atoms hop is a measure of the bond strength. Surface bonds are especially interesting because the atoms involved do not have their normal complement of neighbors. Surface bonds tell us about how atoms interact in situations where the local geometry is unfavorable, and provide detailed information on the electronic environment near the surface. This information can be compared with the results of detailed quantum mechanical calculations to give us a clearer understanding of the bonding process in general. Determining how atoms move at surfaces and how atomic motion is related to surface growth is the major focus of my research at Carnegie Mellon.
 |
 |
| Figure 1. LEEM image of the (111) surface of silicon at approximately 850 C. The light areas correspond to the low-temperature structure (7x7) and the black regions indicate the high-temperature (1x1) structure. The slanted lines in the image show the locations of steps. The field of view is about 4 microns. | Figure 2. Dark-field LEEM images (2.6 eV) of one-layer deep holes on Si(001) at 840 C. The time interval between images is 50 s. The size of each image is roughly 10 x 6 microns. Eventually all the holes are filled in by diffusing atoms that come from the steps. |
Measurements of atomic motion at surfaces are difficult for a number of reasons. For many materials, atoms move extremely rapidly at room temperature, and few microscopy techniques produce an image of surface quickly enough to capture the atomic motion. This is because in many microscopy techniques images are generated by scanning a probe over the surface. Scanning a large area takes time, and while the probe is situated over one part of the surface, something interesting might be happening somewhere else. A new technique, low-energy electron microscopy (LEEM), images all points on the surface simultaneously. A LEEM microscope is similar in many ways to an optical microscope. When you look at something with your eyes, you see an image from the light scattered from the surface of a material. The same principle is used in LEEM, except that scattered electrons are used to form the image instead of light. Bright areas of the image correspond to high reflectivity regions on the sample, and dark regions correspond to low reflectivity. For example, Figure 1 shows an image of the (111) surface of silicon at about 850C. Right around this temperature the atomic arrangement of the surface atoms changes. The low-temperature structure (called "7 x 7") reflects more electrons than the high-temperature structure (called "1 x 1"). In a LEEM image, the low-temperature phase appears white and the high-temperature phase looks black. When the sample temperature is lowered below about 850C, the white areas grow at the expense of the black regions. With LEEM, the motion of the boundaries between the two structures can be followed in real time, yielding important information on how the change of phase takes place on the atomic level. The speed at which the domain boundary moves tells us how quickly atoms diffuse on the silicon surface.
Atomic motion can also be studied by watching unfavorable surface structures decay with time. Figure 2 shows a sequence of LEEM images of the (100) surface of silicon. The white ellipses in the image are holes in the surface that are one atomic layer deep. The atomic arrangement at the bottoms of the holes is the same as at the surface, but is rotated by 90 degrees. This rotation leads to increased electron reflectivity in certain directions, making the holes appear white in the LEEM image. A surface with holes is energetically unfavorable because atoms at the edges of the holes have fewer neighbors with which to bond. As the sequence of images shows, the surface evolves so as to fill in the holes, with the small holes filling in faster than the large ones. The holes are filled in by atoms diffusing on the surface, and the rate at which the holes fill in is governed by the diffusion rate of the atoms. In fact, a very accurate determination of the diffusion constant (atomic hop rate) is possible using this method.
LEEM movies of each of these silicon surfaces can be viewed at www.andrew.cmu.edu/user/jhannon/research/leem.html. As these examples illustrate, LEEM "movies" can give new insight into the motion of atoms at surfaces. By measuring how surfaces evolve in time with LEEM, and comparing the observations with models of atomic motion, a clearer picture of the complicated drama unfolding at surfaces is developing. In the last 10 years, as experimental and theoretical tools to study surfaces have matured, our appreciation of the complexity of surface atomic motion has increased. These tiny actors continue to surprise us. Sometimes, if you put your ear very close to the surface, you can even hear a diffusing actor say: "I came to Barium, Cesium, not to Praseodymium."
