Nanotechnology Now

Home > Interviews > William E. Buhro on Boron Nanowhiskers - June 2002

William E. Buhro Interview 06.19.2002

William Buhro, Ph.D., clears up some questions regarding boron nanowhiskers

After reading the article "Chemists make first boron nanowhiskers", I was curious to know more about this recent discovery, so I emailed Professor Buhro to ask him for clarification. The following is our "conversation":

As nanowires and nanowhiskers are relatively new terms for our readers [and myself for that matter], I was wondering if you could answer a few questions about them.

1. The article says that nanowhiskers are solid, as opposed to carbon nanotubes which are hollow. Is this correct? [I realize that solid is a relative term, and does not take into account the space between the atoms].

Yes; the nanowires are not hollow like carbon nanotubes are, but fully dense like conventional wires are (but of course nanowires have smaller diameters).

2. So far, boron nanowhiskers are always semiconducting, unless doped?

That is correct. We are trying to dope them now.

3. whiskers = nanowhiskers = nanowires = crystalline nanowires? And all are "dense whiskers", not hollow nanotubes?

This terminology is not perfectly clear. The prefix "nano" implies that the diameter of the 1D structure is in the nanometer regime, rather than larger than a micrometer. The term "fiber" implies nothing specific about the structure of the fiber - it may be a single crystal, crystallographically amorphous (like conventional glass is), polycrystalline, etc. The term "whisker" is generally accepted to mean that the 1D structure is a single crystal, or at least approaches single-crystal character fairly closely. "Tube" of course implies a hollow center. "Nanowire" is perhaps the most troublesome. To some of us the term "nanowire" implies that the electrical-transport measurements have been made, and the 1D structures have been found to be either semiconductors or metallic conductors. To others, the term "nanowire" is used in much the same way as "nanofiber." So you can see, standard definitions have not yet been fully developed. Nanotubes may also be considered to be "nanowires." Nanowires must not necessarily be fully dense as our's are; they could be hollow.

He went on to explain some other aspects of his team's work:

Please explain what you've made -- what are these crystalline nanowires?

They are crystals of elemental boron that have a wiry shape (morphology). Actually, they are "twinned" crystals, which means each nanowire consists of two crystals fused along a planar boundary that contains the wire long axis.

What are the advantages of a Boron-based nanotube over Carbon?

If one imagines designing wires for nanometer-scale circuitry, one would first think of using metallic wires composed of copper, aluminum, or some other metal. However, metallic wires are very susceptible to electromigration failure (thinning and complete breakage) when they possess nanometer-scale diameters. So metallic nanowires are out. Carbon nanotubes have attracted _much_ recent attention for nanoscale electronic devices. However, carbon nanotubes have variable electronic properties that depend on wall orientation (the way the chicken-wire carbon-layer structure is wrapped into a tube). Only 33% of the carbon nanotubes in a mixture exhibit metallic conductivity - the rest are semiconductors - so if one is assembling circuitry in which a carbon nanotube is to function as an interconnect (electrical conductor/current-carrying wire) then one has only a 33% chance of selecting an appropriate nanotube. Ideally nanowire interconnects would be highly conductive irrespective of crystallographic orientation (like metals), possess covalent bonding (like carbon nanotubes) to reduce (or eliminate) susceptibility to electromigration thinning and failure, and be refractory (able to withstand high temperatures). Theoretical studies have suggested that boron nanotubes should be stable, highly electrically conductive irrespective of wall orientation, and boron is a refractory element, so we attempted to make them.

What could nanoscale electronic wires be used for?

Ultimately, in circuitry of a smaller scale than is currently accessible by photolithographic methods.

What does the solid, dense nature of these nanowires mean?

It means that they are not hollow nanotubes like carbon. Indeed they are not hollow at all, but fully dense structures like conventional wires are.

How far is this from boron-based nanotubes?

That is hard to evaluate with any certainty. The challenge of growing boron-based nanotubes may be as simple as reducing the diameters of the wires further. For example, both carbon and gold form dense wire structures at diameters of ~ a micrometer and above, but form nanotubes at smaller diameters of a few to several nanometers. Thus we are trying to gain better synthetic control over the diameters of the boron nanowires we grow so we can investigate the structures that form at smaller diameters.

You mention where your work is incorporated -- what nanoelectronic devices are we talking about?

I don't believe our work is incorporated anywhere, and we have no specific nanoelectronic devices in mind. We are merely doing a basic-science study to support the eventual emergence of nanoelectronics. We intend to study the fundamental conductivity (electrical-transport) properties of nanowires. One can't have nanotechnology before nanoscience.

Please give me your best guess as to when we might see the first functional nanoelectronic devices? What are we likely to see first?

Well, nanoelectronic devices in the strictest sense of the term, I'm not sure and this is outside my area of expertise. The real problem as I see it is the well-known, classical, "tyranny of numbers" problem. Researchers are already capable of assembling simple nanocircuits, but "nanochips" based on such nanocircuitry would require integration and interconnection of very large numbers of such circuits. The numbers of such circuits to be interconnected are so large that the job surely can't be done in a one-at-time, serial process. That is the big advantage of conventional photolithographic fabrication used now - numerous circuits can be fabricated and interconnected in parallel. Photolithography should (may) soon reach its small-size limit (hence the "end" of Moore's Law). But until a comparable parallel fabrication and interconnection methodology emerges for the nanometer scale, "nanochips" and nanoelectronics are mere pipe dreams. Various ideas for such methodologies are percolating, but I know of nothing yet that appears to be a "can't-miss" prospect. Our work is a long way away from addressing this critical challenge.

However, nanoscale electro-optic devices already constitute a widely available, commercial technology. Many or most compact-disk players (and other commerical appliances) contain a quantum-well semiconductor laser. The light-emitting regions of these lasers consist of nanometer-thick semiconductor layers of precise compositions and thicknesses, which are critical to the function of the devices.

Anything else you'd like to add?

The work was supported by the Semiconductor Research Corporation, and was a collaboration between researchers at Washington University, Northwestern University, Zyvex Corporation, and Arizona State University.

If you have a comment, please us.