TROY, N.Y. A chemical-vapor deposition technique has been applied to carbon nanotubes to give them unusual electronic properties, according to researchers here at Rensselaer Polytechnic Institute. The treated nanotubes could be used by chip makers to interconnect single-electron transistors with high-efficiency wires. The group aims to build a nanotube architecture that will exhibit near-superconducting speeds at room temperature, plus the ability to pack devices tighter and control quantum effects.
"Nanotubes have fascinating electrical properties, but for them to be used in real devices, you need a process like ours to make really big architectures from these very small building blocks," said assistant professor Ganapathiraman Ramanath, co-developer of the nanotube architecture with associate professor Pulickel Ajayan, both in the materials science department at Rensselaer Polytechnic Institute.
Since their discovery in 1991, carbon nanotubes have held the promise of providing superconducting-class conductivity at room temperature for interconnecting single-electron transistors. Carbon nanotubes have also demonstrated quantum effects when doped with impurities in a manner similar to the way silicon becomes a semiconductor when doped with impurities. Similarly, nanotubes can be doped with impurities to create quantum dots.
"Nanotubes can conduct electricity much faster than metals you get 'ballistic' transport of electrons, so if you connect transistors with nanotubes you can increase both the speed and the density of devices by several orders of magnitude. Our ability to grow nanotubes in any orientation will facilitate [building] those kinds of devices," said Ramanath.
Until now, no single process was known that would precisely grow carbon nanotubes at any predetermined position on a chip. Other researchers have demonstrated nanotubes growing in one direction, or in clumps of random orientations, but the Rensselaer Polytechnic Institute researchers claim to have the first semiconductor process to handle all cases.
"Our process is the world's first to grow nanotubes in any direction that we want vertical, horizontal or at an oblique orientation and this can be predetermined before the nanotubes begin to grow. We can make them grow where we want them to, and we can exclude them from places where we don't want them to grow. This is the first step toward making complex chips out of molecular-sized units," Ramanath said.
Carbon nanotubes were tamed by these researchers using standard chemical-vapor deposition (CVD) processes familiar to all chip makers. The basic mechanism is selective growth processes that are initiated at any point on a chip where silicon dioxide has been deposited. The nanotubes grow perpendicularly to the edge of the silicon dioxide region.
VLSI semiconducting processes have mastered the ability to deposit silicon dioxide in very fine and detailed patterns, so this method is a way to obtain a lot of control over the growth and positioning of carbon nanotubes. Gas phase delivery of a metal catalyst, essential for nanotube growth, ensures that the team's CVD growth process is flexible and scalable. By merely changing the exact metal catalyst used, different kinds of nanotubes can be grown.
"Our process is easily scalable, because you can grow millions of nanotubes in millions of different places at the same time, all with a single process. You don't need a different process for growing them vertically and another process for growing them horizontally. Our process can also be used to grow other types of molecular wires in a similar fashion such as boron nitride or carbon nitride. Our research opens up new avenues for other researchers using those other materials, because now we can begin to build architectures out of nanotubes," Ramanath said.
CVD, the same standard process used by semiconductor manufacturers to make integrated circuits, grows the nanotubes at any orientation but always perpendicularly to the silicon dioxide. By sculpting the underlying silicon wafer, before patterning the silicon-dioxide with photolithographic masks, chip designers can easily arrange for there to be a "seed" at every site where a nanotube should grow.
"We can grow nanotubes in any direction that we want, in a very controllable fashion, and the way we achieve this is by a very selective growth process. We chisel the shape of the silicon dioxide template topography, which is a common technique used in standard semiconductor fabrication. You lay the oxide down on the silicon, then you pattern it into the shape that correctly orients the nanotubes when they grow perpendicular to the surface," Ramanath said.
In the near term, the group plans to commercialize its method in bulk materials, since the process solves one of the big problems for material scientists attempting to reinforce polymers with nanotubes. Carbon nanotubes are stronger than steel and can act as reinforcement when mixed into a liquid-polymer solution. Unfortunately, only a small percentage can be added, because the tiny structures tend to clump together when mixed.
"Nanotubes are one of the strongest materials known to man, so in order to add strength, people add carbon nanotubes and mix them into polymers. But you cannot add beyond a few weight percent before you get mixing problems," Ramanath said.
The answer? The researchers used thin-film technologies to grow a "field" of loosely packed, vertically oriented carbon nanotubes all over the surface of a silicon substrate. Then they poured a liquid polymer into the spaces in between the nanotubes to create a bulk composite material that is as strong as steel, but thinner than paper. Such designer materials may be capable of realizing some future designs, such as engine-less spacecraft powered by giant solar sails that also act as solar cells.
"With our process, what you can do is grow the nanotubes first on the substrate, then you have a skeleton around which you can pour your polymer. In this way you can get a very high density of carbon nanotubes around which there is polymer. Nanotubes act as mechanical reinforcements, but also their wonderful electronic properties may be able to play an active roll as well," Ramanath said.
For the long term, the researchers are organizing their efforts in order to more quickly characterize the technology and the composite materials that they can build with it. By taking inventory of the capabilities of the materials, they hope to pave the way for both bulk composite materials and the future molecular computer age. The single-electron "transistors" quantum dots of molecular computers could theoretically be realized with an architecture based on carbon and other types of nanotubes.
"We are currently characterizing these composite materials, trying to measure the properties of their highly aligned structures, both their electrical-transport and mechanical properties. So far we have only demonstrated that we can grow bunches of nanotubes where we want them, so we next want to demonstrate that we can grow a single nanotube where we want it," Ramanath said "Only then will we be able to build single-electron devices."
An audio recording of reporter R. Colin Johnson's full interview with Ganapathiraman Ramanath can be found online at AmpCast.com/RColinJohnson.