Later this year physicists will be celebrating the centenary of Paul Dirac's birth. One of the most influential scientists of the 20th century, Dirac combined quantum mechanics and special relativity to explain the strange magnetic or "spin" properties of the electron. What Dirac could not have foreseen, however, is how the spin of the electron could change the field of microelectronics.
Indeed, the spin of the electron has attracted renewed interest recently because it promises a wide variety of new devices that combine logic, storage and sensor applications. Moreover, these "spintronic" devices might lead to quantum computers and quantum communication based on electronic solid-state devices, thus changing the perspective of information technology in the 21st century.
Since the 1970s conventional electronic microprocessors have operated by shuttling packets of electronic charge along ever-smaller semiconductor channels. Although this trend will continue for the next few years, experts predict that silicon technology is beginning to approach fundamental limits. By 2008, for example, the width of the "gate electrodes" in a silicon microprocessor will be just 45 nanometres across, which will place severe demands on the materials and manufacturing techniques used in the semiconductor industry. Indeed, the cost of implementing a new production line for such devices is predicted to reach $33bn.
Although successors to silicon technology have been discussed, most of them rely on a complete set of new materials, new handling and processing techniques, and altered circuit design, among other developments. These new technologies include single-electron transistors and molecular-electronic devices based on organic materials or carbon nanotubes (see Carbon nanotubes roll on Physics World June 2000 pp31-36).
But the ability to exploit the spin degree of freedom in semiconductors promises new logic devices with enhanced functionality, higher speeds and reduced power consumption. Crucially, these devices could be fabricated with many of the tools already used in the electronics industry, thereby speeding up their development. The challenge for manufacturers is to combine the technology in the semiconductor industry with the completely different techniques used in the magnetic-recording industry to produce devices on the nanometre scale.
In the April issue of Physics World, Dirk Grundler of the University of Hamburg describes how the recent developments in spin transport and spin injection may herald a new era of semiconductor spintronics that could potentially transform the microelectronics industry.