STM image of the surface of a silicon crystal, courtesy of Dr. B. Swartzentruber, Sandia National Labs
Not very long ago, this was a truism of modern science: "We know atoms exist, but we can never see them." That was before 1982. In that year, Gerd Binnig and Heinrich Rohrer published their first work on a new device known as the Scanning Tunneling Microscope, or STM for short. In 1986, they were awarded the Nobel Prize in Physics for their work.
This article will describe this breakthrough that has at last allowed us to "see" atoms. I have kept the term "see" in quotes, because the microscope in question does not operate by bouncing light off a sample and creating a magnified image. Rather, the STM operates by taking advantage of quantum tunneling to measure the shape of the electron "cloud" at the surface of a material. That is, it makes a contour map of the surface, where the contours follow lines of constant electron density. Some people will argue that this is not really "seeing," but there can be little doubt that the peaks and valleys on many STM images represent individual atoms and the spaces between them.
How does this amazing device work? The key is in quantum
tunneling. Tunneling is an essential feature of quantum
mechanics, that we can understand in terms of Heisenberg's uncertainty principle. In one
form, the uncertainty principle says
We can interpret this to mean that the energy of a particle can increase for a short period of time, even if the energy does not come from the particle's surroundings. That is, for short periods of time, energy is not conserved. The implications of this notion are astounding. In the case of tunneling, the particle briefly "borrows" an extra bit of energy from nowhere, uses it to travel through a barrier and then gives the energy back after it is done. Once the process is done, the particle has its original energy back, but it now is on the other side of the barrier.
Here is an analogy. A marble is placed in a cereal bowl and given enough of a push to roll half way up the side. If there is little friction, the ball can roll back and forth for a while, but it will never have enough energy to roll all the way up and over the wall. It cannot overcome the barrier. However, for a subatomic particle, the situation is very different. The particle's energy is tiny, and so are the barriers it faces. There is a reasonable chance that a subatomic particle can pass over a barrier and escape. Once it is out, it doesn't need the energy anymore.
This is how the scanning tunneling microscope works. In the STM, a very sharp metal tip (usually tungsten) is held near the surface of a conducting sample. A small voltage (usually 1.5 to 3 V) is placed between the tip and the sample. Now, the electrons in the tip would like to flow to the sample. If the tip touches the sample, current will flow easily; if the tip is far away, the electrons cannot overcome the barrier and "jump" the gap. However, if the distance is just right, the electrons can tunnel out of the tip and get into the sample. This process is illustrated in the diagrams.
What makes this process truly wonderful (besides the fact that it can happen at all) is the tunneling current depends very sensitively on the size of the gap. If we move the tip laterally over the sample, the current changes strongly in response to variations in the surface height. This is known as constant height mode. What is actually even better is constant current mode. In this operation, the tip is moved laterally, and as it moves its height is adjusted in order to keep the tunneling current constant. As this process occurs, the height adjustments are recorded. The tip literally follows the contours of the sample.
It would be a shame to leave you with the impression that STM is a scientific curiosity with no purpose beyond making pretty pictures. On the contrary, STM is a powerful tool for learning new and useful information about the physics and chemistry of surfaces. One example is in the understanding of how films of materials grow. Many important technologies (semiconductors and magnetic storage come to mind) depend on devices that are made of thin films of material (silicon and magnetic alloys in those two examples). The precision of the devices depends crucially on the precision with which the films can be grown. STM has helped to solve many problems in our understanding of film growth.
The figure to the left shows an example (courtesy of Dr. J. Stroscio of the National Institute for Standards and Technology). In these images we do not see the individual atoms laterally, but we do see single atom changes in height. The brightness represents the height of the surface, brighter areas are higher than dark. Thus, the film on the left (a) has grown many small islands, and there are at least three separate layers showing. On the other hand, the film on the right (b) is growing in a "layer-by-layer" mode. This is desirable because the resulting film will be flatter and have fewer defects. Furthermore, if it is necessary to grow another film on top, the next film will have a better surface to start from, and there may be less intermixing. What allowed the improvement? The only difference was the temperature during film growth. Film (a) was grown at 488 K, whereas film (b) was grown at 573 K. Films grown at lower temperatures are even worse than (a), and growth above 573 K causes other problems. Thus, an optimum temperature was obtained, along with a clear understanding of the reason.
Further study links:
And those guys at IBM really go all out ...
Further Study Questions: