8/28/97
CONTACT: David F. Salisbury, News Service (650) 725-1944;
e-mail: david.salisbury@stanford.edu
Electrical devices that work one electron at a time are the ultimate in microelectronics.
Such devices could provide the smallest, most sensitive electrical circuitry conceivable and operate on minimal amounts of power, if they can be made to work. For this reason, there is a considerable worldwide interest in this new area, termed mesoscopic electronic devices.
Researchers are already exploring ways to use these devices to make neural networks, cellular automata, quantum computers and high precision measurement systems. But so far no one has managed to come up with a design for a single-electron device with the electrical characteristics suitable for building actual circuitry.
Now, writing in the Aug. 25 issue of Applied Physics Letters, Wayne H. Richardson, a research associate in Stanford's Ginzton Laboratory, proposes a new design for a single-electron device with a key characteristic that has been missing from previous single-electron designs: the ability to amplify electrical signals in a controllable fashion over a significant range of operating voltages.
Richardson's design looks deceptively traditional. It consists of a gallium arsenide junction between a region where the dominant carriers of electrical current are electrons (n-type) and a region where the dominant carriers are holes microscopic regions in the material that act as if they are positively charged particles (p-type). The gallium arsenide is doped with different kinds of impurities to produce these two conditions.
But in this case the material has been so heavily doped that it becomes what solid-state physicists call degenerate. In this condition it has electrical properties similar to those of a metal.
The electrical property of a material is determined by the behavior of its electrons. In individual atoms, electrons travel around nuclei in a series of distinct patterns called orbitals. Those closer to the nucleus are at lower energy levels than those further away, so electrons fill the closer orbitals first. When millions of atoms are joined together in a repeatable pattern to form a material, these orbitals tend to join together to form a series of bands. In solids, most of the electrons occupy the lower energy bands, where they remain bound to individual atoms. The electrical properties of a material are determined mainly by electrons in the higher energy bands, where they can move freely.
In metals, the bands overlap and are only partially filled so that it is easy for electrons to jump into the upper energy bands. That is why metals conduct electricity so readily. In insulators, however, there is a large gap between the lower and upper bands. Very few electrons can make the jump, so the lower bands tend to be fully populated while the upper bands are nearly empty. As a result, the material is a poor electrical conductor.
As their name implies, semiconductors are an intermediate case. It takes more energy for electrons to jump up into the upper bands than it does for metals, but the gap is small enough so that material can be switched back and forth between conductive and non-conductive conditions. As a result, its electrical characteristics are highly sensitive to the amount and type of impurities in the material. In a degenerate semiconductor, an impurity that provides extra electrons has been added in a large enough concentration so that electrons are present in the upper bands, forcing it to act much like a metal.
Richardson has shown that when a degenerate p