Uncrossing the Wires

A recent study by Prof. Ernesto Joselevich and his team, published in Science, features perfectly aligned horizontal arrays of thin, millimeter-long nanowires. We spoke with Joselevich recently to find out why he and others in the field are excited by this advance:

WSW: Your lab has produced a fair amount of innovative research in recent years. Why is this one special?

EJ: First of all, we succeeded in growing very long horizontal nanowires with exquisite control over their orientation. Because of the numerous potential uses for semiconductor nanowires, there is a lot of competition to create better ones with more efficient processes. We even managed to demonstrate control over the orientation of the atoms within the nanowires.

WSW: What is unique in your method?

EJ: Basically, we provided the nanowires with guidelines to direct their growth - something like stakes for vines. Rather than a tangled mess, we can grow them in neat, straight lines. Our nanowires are made of gallium nitride deposited on an artificial sapphire surface. This sort of deposition is used today to create the thin semiconductor films in blue LEDs or the violet lasers used in Blu-ray discs. This technology was developed in the 1970s, but it took another 30 years to figure out how to reduce the defects in the crystal enough to make them useful.

More recently, gallium nitride nanowires have been grown vertically. The problem here is that the process of harvesting them and creating arrays is messy and inefficient, leading to relatively short and poorly aligned nanowires. With our method, you can grow the nanowires horizontally, in ready-made arrays.

WSW: How does the method work?

EJ: Rather than producing the wires on a smooth surface, we created stepped and grooved surfaces for them to grow along. We found that if we cut the sapphire in certain directions and heated it to around 1,500° C, the surface wrinkled into nice parallel steps and grooves. The nanowires were then grown in the grooves in a vapor-liquid-solid process. We found that we could control the properties of the nanowires by controlling the planes of the grooves. That is, we could grow them in different directions with respect to the matrix of the substrate crystal. This, in turn, led to different orientations of the atoms within the nanowires themselves.

WSW: What was surprising about this research?

EJ: It is not a given that one can grow one crystalline material on top of another crystal (sapphire) without introducing defects. We had previously grown nanotubes on uneven sapphire surfaces, but carbon nanotubes hardly react with the sapphire. In the case of the gallium nitride wires, one would expect to see much more influence of the substrate on the crystal structure.

In fact, our nanowires are surprisingly relaxed - they show none of the stress found in the semiconductor films of this material. We think the reason is that the long, thin wires can easily shrink or swell sideways to fit into the surface features, something that a two-dimensional film can't do. So our nanowires turn out to have few defects and excellent electrical and optical properties.

WSW: What about applications?

EJ: There is a long list of potential applications. In fact, we know how to use nanowires; the need already exists. And since, control of structure and miniaturization go hand in hand in the semiconductor industry, this method could well become standard within the decade.