Overview of Device Physics Research

Katsumi Murase
Device Physics Laboratory

Today's high-technology society can be said to be based on semiconductor devices. Device physicists and engineers have been pursuing higher speeds and more sophisticated functions. This will be the guiding principle for the development of semiconductor devices for the foreseeable future. However, when we think of the limitedness of the earth, environmental friendliness should also be part of this guiding principle.
With this in mind, we are carrying out research for developing innovative electron devices, such as ultimately-low-energy-consuming single-electron devices. We are also doing research for controlling surface atom arrangement on a wafer scale and creating novel materials, which will lead to the invention of future devices based on an entirely new operating principle.
The single-electron device is, as its name indicates, a device whose operation can be controlled by a single electron. By employing single-electron devices for LSIs, we can reduce the amount of energy consumed to about 1/10,000-1/100,000 that of conventional LSIs. We succeeded in demonstrating room-temperature operation of a silicon (Si) single-electron transistor in 1994. Since then, we have been making efforts to integrate devices as well as to create more sophisticated functions using a simpler device structure. Establishing nanofabrication technology for Si single-electron devices has also been one of our main research targets.
In fiscal year of 1999, we achieved important and interesting results in four research areas. The first is a silicon single-electron inverter. The inverter is the most basic element in logic circuits. Our fabrication of the first Si single-electron inverter by means of a novel method called V-PADOX is an important milestone in the development of single-electron logic LSIs.
The second achievement is a new type of cross-linked resist for small pattern roughness. The sidewalls of a lithographically defined resist pattern have roughness caused by the molecular structure of resist. This roughness, while not large enough to affect the characteristics of conventional devices, must be reduced for devices that need precise size control of the order of a nanometer, single-electron devices for instance. We found that the roughness can be effectively reduced by cross-linking polymers of resist.
Thermal oxidation of Si is of central importance in Si device technology. It has been used for more than 30 years in the semiconductor industry. Nevertheless, the microscopic mechanism of thermal oxidation has remained unclear. We have created a universal theory for thermal oxidation of silicon that can describe various aspects of thermal oxidation quantitatively.
The ultimate processing technology may employ atom manipulation. For such a technology to be applicable to device fabrication, however, artificial atom manipulation should be done on the whole surface of a wafer. To achieve atom manipulation, we first need to have a deep understanding of atom dynamics. The knowledge of step dynamics observed on ultra-flat Si surfaces will make a key contribution to establishing a practical atom manipulation technology. The successful observation of steps resulted from the combination of our unique surface modification technique and a specially designed scanning electron microscope.