Materials – Nanotechnology

Welcome to the world of nanotechnology, a place where small things, namely nanometer-sized particles, can have a big impact. A nanometer is one millionth of a millimeter. That's approximately one fifty-thousandth the diameter of a hair. Too small to bother with? Not according to Berndt Samsinger of Capital-Stage, a Hamburg, Germany-based investment company that specializes in nanotechnology. Says Samsinger: "The impact of this new field in coming years will be greater than that of biotechnology and the Internet combined during the last decade."

Market hype aside, Gentzkow's particles are relatively simple. Their layered silicates have a structure similar to that of puff pastry and are, for example, used in such mundane products and processes as cat litter and paper production. The layers can be separated with sodium or calcium ions, and when treated with organic ions can also be expanded in such a manner that they detach from one another when incorporated into plastics. This results in individual, tiny silicate plates. If they are added to a polymer at a ratio of up to five percent, the mixture inherits the properties of both substances. In other words, it becomes transparent and strong. It is also inexpensive to produce and can be manufactured in large quantities. Experts predict that in about two years the amazing material will be mass produced and used as a plastic-coated lens for very bright and temperature-stable light-emitting diodes.

But the plastic from Erlangen has one big drawback: it looks utterly unexceptional. When you hear the word "nano," which means "dwarf" in Greek, you are more apt to think of miniature submarines that prowl through the bloodstream and annihilate cancer cells, or of miniature robots made of a handful of atoms that cooperate and reproduce themselves—as described in Michael Crichton's new novel "Prey," for example. "But that's pure science fiction, and it's very doubtful whether there will ever be applications of that sort," says Rainer Nies, who wrote a study titled "Impact of Materials" at Siemens CT in Erlangen (see article ? Invisible Revolutions). Pioneering innovations? They will probably be the absolute exception, says Nies, who studied physics. Instead, many small innovations will gradually appear in completely ordinary products—but the net result will probably be just as revolutionary.

Chip Structure at the Limit. The manufacturers of microchips are depending on nanotechnology for their very survival. Moore's Law, which predicts that the number of transistors per unit area of chip will double every 18 months, will hold true until approximately 2010. But what happens when chip structures supposedly drop below 100 nm? That's the question that is occupying Dr. Lothar Risch, who conducts research on nanoelectronics at Infineon in Munich. Risch's projects reach far into the future. He estimates that components now being manufactured in his lab as individual pieces will not be used to produce marketable products for at least ten years.

Risch builds field-effect transistors, which are the smallest units of any chip. Risch's FETs have a gate length of a mere ten nanometers. The gate acts like a valve that controls the electric current in a silicon channel that is only two nanometers thick. However, when the layers are that thin, the electrons begin to tunnel through the gate as if it were not even there. Risch's team therefore manufactured the prototype of a double-gate transistor tilted 90 °, in which two gate electrodes sandwich the silicon channel, thereby making it possible for them to control the current much more effectively.

The next step is a quantum-dot memory module in which an insulator with an edge length of 20 nm is placed between the gate and the silicon channel. Less power is needed here for saving and deleting—a quantum-dot memory of this kind is so sensitive that even a single additional electron in the quantum dot shifts the characteristic curve of the transistor noticeably. Quantum dots have made quite an impression on the research community and scientists hope to use them in supercomputers or in lasers for ultra-fast fiber-optic links.

"Everything that's possible with silicon is also possible with nanotubes," says Hönlein. His team can deliberately grow clusters of the tubes on silicon—which enables connections to be created between the layers of a microchip. In the future, the conductors could also consist of nanotubes, as could diodes and transistors. And there's more: If you place one nanotube directly on top of another one and apply an electric field, they bend and stick to each other until a voltage pulse separates them again—a tiny switch that could also be used as a data storage device. Individual samples already exist in research labs, but a reproducible manufacturing method is still a long way off.

Nanotubes on Display. Nanotubes are expensive—as much as 500 €/g. But they're likely to drop to just a few euros if, as announced, Japanese companies begin mass production this year. Korea's Samsung has announced its intention to market its first nanotube displays in 2003. Electrons can be shot at a phosphor from the ends of the tubes by applying an electric field—as is the case with conventional cathode ray tubes. The difference is that the surface is totally flat and there is no wear and tear. A nine-inch diagonal prototype that displays images in all their glorious color already exists.

"If transistors made of nanotubes one day became as good as those made of silicon, my work would be superfluous," Risch admits. But since no one can say for sure whether nanotubes will make it possible to squeeze 100 million transistors onto a chip, he is likely to have work for years to come. Nanotubes hold similarly untapped—but uncertain—potential in other fields of research. For instance, they might be used as an admixture for particularly hard materials, or as a hydrogen storage medium for fuel cells. However, all such potential fields are already dominated by established technologies. Whether nanotubes will be able to offer viable alternatives is anyone's guess.

At Siemens in Erlangen, Dr. Joachim Wecker and his team are investigating magnetic multilayers that are only a few atomic layers thick for use in future memory chips. Such components are expected to hit the market in 2004. Data bits in these MRAMs, are not stored in capacitors but in miniature magnets. Their polarity is reversed by a weak electrical pulse, and their memory content is read out electrically. The big benefit here is that once stored, data bits can be retained for any length of time. The PC memory modules used today must be refreshed many times per second and therefore need more power. A type of storage that retained its memory would also dramatically shorten the boot-up process.

Nevertheless, some fundamental questions remain unanswered. For instance, Wecker's team is still trying to determine if there is a lower limit to the size of magnetic structures. Calculations indicate that structures below 25 nm are not possible because at that point ambient heat can nullify the magnetization of the mini-magnets and make stored data unreadable. Nevertheless, Wecker is optimistic that he will be able to lower this value by a few nanometers.

Wecker's goal is to develop components that can be used in Siemens products. At the top of the wish list, therefore, are tiny magnetic-field sensors for imaging processes in medicine or for use as sensors in automobiles. Another project deals with reconfigurable logic chips in which tiny sandwich magnets can be linked to form arbitrary logic gates through alteration of the magnetization direction. The millions of transistors in today's microprocessors are hard-wired, which means it's not possible to change the circuits to accommodate new tasks. "Many products would profit if you could change the hardware later on," says Wecker. An audio processor could become a video processor, for example. Processor and memory could be combined on one chip whose resources would adjust to fit each job.

Power for such frugal chips could be supplied by the new organic solar cells that Dr. Jens Hauch is developing at Siemens CT in Erlangen. In these cells, light is converted to electricity by a polymer semiconductor. This synthetic is full of buckyballs—nanoscale soccer balls made of 60 or 70 carbon atoms. The cells' energy yield is still a meager 3.5 %, but Hauch is optimistic that his team will be able to manage 10 %. Such nanoscale power plants would not only be flexible but also much less expensive than today's silicon solar cells, which cost between five and ten euros per watt of output. "We're counting on less than 1 €/W," says Hauch.

Bernd Müller