The economics and hurdles of nanoelectronics are explored

The rush toward ever-smaller transistors in the burgeoning field that has come to be known as nanoelectronics is driven more by economics than by functionality, according to Marc Van Rossum, senior scientific and strategy advisor at the Inter-university Micro Electronics Center (IMEC), a European electronics research and development center in Leuven, Belgium.

Van Rossum, who heads IMEC's nanotechnology efforts, will be the featured speaker at a luncheon scheduled for Nov. 13, during ASME's International Mechanical Engineering Congress and Exposition in New York City.

Though the promise of increased performance by nano-size devices provides a strong commercial reason for doing nanoelectronics research, "by itself it wouldn't be sufficient if there weren't the promise of an impact on prices. The industry is essentially looking at its margins," Van Rossum said.

Smaller devices mean lower costs because as the size of transistors shrink, the number of such devices that can be packed onto a single chip increases, as does the number of chips that can be squeezed onto a wafer.

And while the lure of lower costs for nano-scale processors has led to numerous nanoelectronic successes pertaining to nanotubes, significant hurdles must still be surmounted before nanoelectronic circuits become a commercial reality, Van Rossum said. Key among these hurdles are the "design crisis," the "global interconnect crisis," and what could be termed the "gate insulator crisis."

Marc Van Rossum

The design crisis revolves around the fact that even at today's microelectronic scale, chips are becoming so complex that software developers are falling behind. "The latest processors in PCs are used well below their potential because writing software that would explore them to their full capacity has become a tremendous challenge," he said.

This problem will only grow larger as electronic devices shrink to nano-sizes — generally accepted to be below 100 nanometers. The design crisis would be "a strong reason for the industry to pause, and say, 'Let's wait a minute and see how our software colleagues are going to cope with this,' " Van Rossum said. But with the price of nanoelectronic circuitry being the driving force, full utilization of a chip's capacity takes a back seat to cost. Yet how long the balance between cost and function will hold is open to question.

"At some point, and it may happen even before the nanotechnology era, there will be at least a pause just because we will no longer be able to manage the program side of the chips," Van Rossum said.

On the hardware side of the equation, the "global interconnect crisis" is also a child of the complexity of minuscule circuitry. In this crisis, the wiring between transistors on a chip "is becoming almost unmanageable," Van Rossum said.

"To bring some order to the chaos of all those wires, it has become clear that, in the future, a chip will work with hierarchies of wires," Van Rossum

explained. "We will first have layers of wires that connect transistors on the local scale — small islands of transistors — and then those islands will be connected by different wiring schemes."

The gate insulator problem results from the fact that silicon dioxide, currently used to separate the body of a transistor from the electrodes that control the current, becomes too thin to insulate at about 50 nanometers. Hence, a new insulating material must be found, Van Rossum said.

However, there is another side to the nanoelectronics revolution, where cost and full functionality are not as important as the ability to accomplish what was previously impossible. Because nanotechnology operates at the molecular scale, it enables the merging of solid-state electronics with chemical and biological elements to create completely new systems. Such is the case with biological sensors, a primary area of investigation at IMEC.

In this area, Van Rossum's group is aiming at the creation of protein chips for medical diagnostics. "They could be used for diagnosis of diseases, but more broadly we could also use them as health monitoring systems," he said. Such a system, for example, could be used to monitor the foot and mouth disease virus.

The merging of electronics and biology is currently being applied in the creation of DNA chips for gene analysis and therapy. These semiconductor structures contain a layer of DNA that interacts with other biological agents. The interaction is then detected at the semiconductor level. Replacing the DNA on the chips with proteins could have much wider applications because, unlike DNA, the body uses proteins for all of its metabolic functions.

Using protein molecules instead of DNA molecules "would make the chip sensitive to a number of other agents in the body, like blood and body fluids, and could be used for automating biomedical analysis," Van Rossum said. The final aim would be to have fully automated tests that could be run virtually instantaneously in a doctor's office.

But, before this can be done, a way must be found to stabilize proteins, which are more flexible than DNA molecules and tend to fold in on themselves, Van Rossum explained. He said, "We must devise schemes in which we combine proteins with some ligands that will immobilize them in certain shapes."

The IMEC nanotechnology scientist thinks protein chip technology will be available within five years. But that is only half the battle.

"These kinds of devices or techniques need approval by public authorities," Van Rossum said. "In the United States, the FDA is watching very carefully."

— Victor D. Chase

back to news & features