User Highlight

by Peter Bennett, Arizona State University

At the Center for Functional Nanomaterials, we are working to build useful nanoscale objects. This requires that we understand and learn to control their growth. Just seeing structures at the nanoscale is difficult. Watching them interact in real time at high temperature is even more challenging.

No tool is better suited for this than a low-energy electron microscope (LEEM). Unlike in a conventional transmission electron microscope, electrons in a LEEM do not penetrate a sample; rather, they are reflected, gently probing the surface.

Peter Bennett (right) works with Peter Sutter at the CFN's LEEM.

We've started working with an Elmitec-V LEEM, the first tool installed at the CFN. It's capable of 5 nanometers lateral resolution at sample temperatures over 1000°C.

Phase contrast provides very high sensitivity to vertical features such as single atom-high steps on the surface.

In the case of silicon(100), successive crystal layers have two alternate structures, leading to a strong black/white contrast between alternate layers. The resulting zebra-like pattern provides a clear indicator for the configuration (and motion) of single atomic steps on the surface. Other tools available in the CFN are atomic force microscopes and transmission electron microscopes. These instruments provide powerful complementary information about a sample.

In one experiment, we are looking at the motion of tiny droplets of liquid metal, in the form of a platinum-silicide (PtSi) alloy on a clean silicon surface. We began by studying the growth and thermal stability of long and thin PtSi nanowires, which could help link active electronic elements on a silicon chip. Such nanowires are smaller and more highly conductive than anything that can be made with conventional electron beam lithography, which is key to making more compact microelectronic devices.

Following the growth study, we checked to see that these PtSi nanowires would survive the higher temperatures required in a fabrication process. The nanowires remained unchanged up to 1000°C, at which point they collapsed into compact structures, then abruptly melted, forming perfectly spherical droplets of liquid metal. We were then astonished to see these droplets begin to move, all in the same direction! After ruling out gravity, strain or mutual attraction as a driving mechanism, we realized that the droplets were moving towards regions of higher temperature.

Image sequence showing a "roadway" built up by multiple passes of droplets (at 1100°C). Droplets of interest are numbered. Elapsed times are 0, 5, 10 and 30 seconds for panels a-d, respectively.

This phenomenon of "thermo migration" is, in fact, well known in bulk materials and is important in such events as failure of containment vessels in a radiation environment or metals mixing during alloy formation. (I might note a sociological analogue whereby senior citizens from Minnesota migrate towards Arizona in January.) But the phenomenon of thermo migration has rarely been observed on clean surfaces and never in real time.

Our study provides insight into fundamental aspects of the motion of sub-micron liquid droplets on a well-defined surface, which will be useful to help control the formation of nanometer-sized metal islands on semiconductors. Such metallic dots are used as seeds for the catalytic growth of semiconducting nanowires, widely studied for futuristic electronic circuits. Looking ahead, we plan to formulate a model for the drag force exerted by a single atom step. We also are eager to explore these effects in other material systems.

------

> Next article: Working with Nanomaterials at Brookhaven