When it comes to chip designs for the next century, it looks like conventional methods may not be able to keep the pace.
Experiments financed by the Office of Naval Research (ONR) and the Army Research Office (ARO) are examining the ability of silicon wafers to conduct photons (commonly called light waves) in the hope of someday creating a ultra-fast computer chip that operates at the speed of light - or about 100,000 times faster than current semiconductors.
The research, conducted by scientists at the Quantum Device Laboratory at the University of North Carolina, Charlotte, with silicon provided by a small, New York-based R&D house, NanoDynamics Inc., so far has found that if electrical voltage is sent through the substrate, visible light is created that "shines" from the silicon.
"We believe that a giant step has been taken in silicon technology to include photons," said Raphael Tsu, a professor of electrical engineering at UNC-Charlotte. "The integration of electronic and photonic capability on a single silicon chip is a very real possibility."
The research is being conducted by Tsu and Qi Zhang, a doctoral fellow, and could have dramatic applications in the computer industry.
Presently, computers and other electronics products process information through electrical current, with bits of data transmitted via electrons - subatomic particles with a charge of negative electricity, and the primary carriers of electricity in solids.
The research, if confirmed in subsequent testing, could also have an impact on the networking industry, for fiber optics employ photons to carry information between two points. But at each point, there must be semiconductors of compound silicon to transform the data from photons into electrons.
Currently, electronic and photonic semiconductors cannot be built on the same chip. But with light-emitting silicon, electronic and photonic devices could be built, conceivably, on the same chip. That would simplify the "transformation" process, said Zhang.
The news that the research had been completed was greeted enthusiastically in the semiconductor community, for scientists have published thousands of research papers on the topic of photons.
"Hell yes, it's a big deal - if it can be done," said Jon Peddie, editor in chief of the Peddie Report, an industry insider newsletter. But, he cautioned, questions do indeed remain.
"How do you get photons - as energetic as they are - through a baked hunk of opaque sand? Maybe they don't go through but are generated in some way, like a laser, or some other quantum jump."
Research being conducted by Tsu may help computer science bypass the limitations of today's integrated circuits, according to a paper, published by a commercial research scientist named C. G. Wang at NanoDynamics Inc., called "A Perspective on Integrated Circuit."
As the chip industry struggles with the limits of conventional silicon processing power, it is believed by many scientists that a sub-0.1 micron chip will be hard to attain. So Tsu and other researchers have developed an interesting approach to the problem of chip design. They have constructed lattices made of complex, silicon and silicon oxide molecules, in a crystalline form, to develop a new kind of chip surface altogether (the chip design offers a "low energy strain," compared to conventional chips). Electric transmission switches can be designed using these so-called "superlattice barriers" through which ballistic electrons, or waves, could tunnel.
Although the technology is very complex, the breakthrough can be compared, in a sense, to the development of fiber optic cable for transmitting telephone and data messages across long distances. By sending information over fiber optic wires, messages can be delivered much more swiftly than they could be over copper-based wires. In this case, though, information is being sent across a tiny sliver of silicon inside a PC, rather than a network.
The final word on how the researchers accomplished all of this, however, is going to have to wait.
Tsu has submitted a proposed paper to a peer reviewed scientific journal, where he hopes to lay out the entire story of the photon process, hard as it may be for laypeople to comprehend. "It may be a bit too techie even for the digirati," concluded Peddie.
