Buckystuff

Graphite is the most stable form of carbon. Its fibers strengthen everything from jet engines to golf clubs and tennis rackets. It also has less technical uses as a lubricant and pencil "lead," and as coal it burns marvelously well. It is made out of atoms arranged in a planar hexagonal pattern, each carbon atom connected to three others, a structure science-fiction writer James Blish once called "polybathroomfloorite." The atoms are kept in place by covalent bonds, which occur between atoms close enough together to share electrons. Neighboring planes are held together by the Van der Waals force, a weak interaction that allows the graphite sheets to slide against one another and flake apart. This is what gives graphite a slick feeling when you rub it with your fingers.

The best known form of carbon is diamond, widely used in industrial processes that require tough materials, and in the mating rituals of Homo sapiens. Each atom in a diamond crystal has four neighbors, giving it a tetrahedral structure. Although diamond seems phenomenally enduring, in truth the crystals are metastable -- those we see around us will, someday in the far future, flake away into graphite.

Graphite and diamond share similar properties. The covalent bonds between the carbon atoms in each are among the strongest in nature. The seeming fragility of graphite comes from the weak bonding between the carbon planes. Diamond is harder because the bonds form a three-dimensional network rather than two-dimensional planes, so no easy way exists to separate its atoms.

However, atoms at the surface of either graphite or diamond don't have enough neighbors to bond with. The leftover electrons form dangling bonds which eagerly latch on to passing atoms or molecules. This is why a monatomic layer of hydrogen usually covers the surface of a diamond; the hydrogen atoms are attracted to the surface atoms and tie off the dangling bonds. Similarly, dangling bonds at the edges of a graphite sheet often connect to hydrogen atoms.

What if we wanted to make a pure carbon system, just for fun? To do that, we would first have to get rid of the dangling bonds. Start with a small sheet of graphite and quickly remove all hydrogen atoms at its edges. Before the dangling bonds have time to attract other atoms, roll the carbon sheet into a ball, letting all of the danglers form covalent bonds. (The process can be simulated by letting a three-year old loose in a room full of shoestrings.) Now you have a ball of carbon. It is not quite graphite, since it has a curved surface. It is certainly not diamond, since most, if not all, of the carbon atoms are bonded to only three other carbons. This new material is a fullerene.

Figure 1. The C₆₀ buckyball. Circles represent atoms, lines show bonds. Each atom bonds to three neighbors. There are twelve pentagonal and twenty hexagonal faces.

Figure 1 shows a C₆₀ molecule, the most famous fullerene. It is a sphere of sixty carbon atoms arranged in a soccer ball shape, with the atoms at the vertices of the "ball." (This is called a truncated icosahedron.) The bonds between the atoms make twenty hexagons, reminiscent of graphite. The new feature is the appearance of twelve pentagons. Why twelve? The great Swiss-born mathematician Leonhardt Euler (1707-1783), who spent a lot of time playing with geometrical figures, proved that any closed surface made of pentagons and hexagons must have exactly twelve pentagons, independent of the number of hexagons.

All closed fullerenes obey Euler's rule, including C₆₀ and its cousin C₇₀, which has seventy atoms and is roughly the shape of a rugby football. Both molecules are reminiscent of the geodesic domes created by the American architect Buckminster Fuller (1895-1983), which led to the unwieldy name of buckminsterfullerene. This was soon shortened to "buckyball" for C₆₀. However, staid journals such as The Physical Review would not accept such a title, so the second half of the original name was chosen. All of the new compounds are now known collectively as "fullerenes," which at least sounds like a chemical.

Buckyball and C₇₀ were first noticed by Richard Smalley and co-workers at Rice University (Kroto et al. 1985; Curl and Smalley 1991). They produced C₆₀ by vaporizing graphite in a jet of helium, which stripped away hydrogen from the edges of the graphite particles in the flame. (Helium is an inert noble gas, which means it will not bind to the dangling bonds.) Some of the stripped graphite curled in on itself, forming proto-fullerenes, and a small portion of these structures settled into buckyballs. The high symmetry of the buckyball made it an interesting novelty (Chung and Sternberg 1993), but its usefulness was limited because it was so difficult to make. To see how this difficulty was overcome, we need to look at the interstellar contribution to fullerene study.

In the early 1980s Wolfgang Kratschmer of the Max Planck Institute for Nuclear Physics and Donald Huffman of the University of Arizona were studying interstellar dust. Since carbon is the element that most commonly forms molecules, they began with the logical assumption that most interstellar dust is made up of carbon particles. Not content to wait for the starship Enterprise to bring them samples to prove this theory, they vaporized carbon in the laboratory and condensed it in every way they could. They then compared the way laboratory dust interacted with light to how interstellar dust scattered starlight. They made these comparisons by studying the spectrum of the light, which shows how its intensity varies as a function of its frequency. Every element and chemical compound has a unique spectrum, so matching a laboratory spectra with an astronomical one makes it possible to identify the composition of interstellar dust.

In 1983 Kratschmer and Huffman determined the absorption spectra of carbon dust that they created in a helium atmosphere (Curl 1991). They found that in the far-ultraviolet region, each spectrum had two peaks which looked like the humps of a camel. It didn't match the interstellar spectra they observed, but when they read about the discovery of the buckyball they wondered if this might be the cause of the humps. However, if fullerenes were the source, then most (if not all) of the dust they produced had to be pure buckyballs. At the time, that didn't seem likely.

In 1989 Kratschmer and Huffman redid their experiments, this time focusing on the infrared spectrum of the dust. They discovered their "camel" samples had spectra that matched what theory predicted for the C₆₀ molecules (Kratschmer et al., 1990). Continuing with their experiments, they found that when they dissolved fullerene dust in benzene and evaporated the solvent, the remaining buckyballs solidified into a new form of carbon, a crystal made up of stacked C₆₀ balls. Although they hadn't determined the composition of interstellar dust, in a marvelous example of serendipity they found a new way to isolate fullerene molecules in large quantities, far beyond previous expectations.

Fullerenes are thought to form in any flame which produces soot. Anytime you hold your hand above a candle you collect buckyballs. Given the prevalence of dust in the universe, fullerenes are probably the most common form of pure carbon. Buckyballs with anywhere from 24 to 600 carbon atoms are known to exist. The exact shape of the ball depends on the number of atoms in it. But all contain an even number of atoms, since closed odd-numbered structures are prohibited by Euler's theorem. If there happens to be an odd number in a proto-fullerene, then either one atom is stripped from the molecule as the fullerene forms or else the dangling bond attracts another atom.

Fullerene molecules form natural cages. Metal atoms trapped inside a forming buckyball will remain there until the molecule is destroyed. It has been suggested that some of the radioactive particles from the Chernobyl accident were carried into the atmosphere inside bucky-cages (Kroto 1988). Fullerenes can also trap other fullerenes, making bucky "onions." Spiraled bucky "seashells" can also form.

An ordinary carbon atom in a fullerene molecule forms three bonds, just as in graphite. However, each carbon atom has the potential to make four bonds, as in diamond. This makes each fullerene a potential chemistry laboratory. Organic molecules might bind to the surface of a buckyball, which could then act as a catalyst, a substance which speeds up the interactions between two molecules by holding them together on a surface.

Inorganic molecules can also bind to fullerenes. For example, a fluorine atom could attach to a carbon atom. Adding fluorine to each of the carbons on a buckyball would make a double-layered molecule, C₆₀F₆₀. The bonds in this compound are strong enough that it should be stable under high pressure, high temperature, and in corrosive environments. Since the molecule is round and difficult to compress, it might serve as a ball bearing in a nanomachine. A powder of C₆₀F₆₀ balls could serve as a "dry" (non-oily) lubricant, probably superior to graphite. Some of the most interesting properties of fullerenes occur when metal atoms are added to the solid structure (Hebard, 1992). Such materials can be used to make superconductors, which allow electric current to pass freely, without any resistance. Although solid C₆₀ is an insulator, adding potassium to make KC₆₀ turns the substance into a conductor. Adding more potassium, enough to change the composition to K₃C₆₀, produces a metal which superconducts at temperatures equal to or less than 18 Kelvin (over 400 degrees below zero, Fahrenheit). Superconductors are potentially a boon for the electrical power industry because a large share of the electricity generated in the world is lost traveling the wires from the generators to factories, houses and schools. Unfortunately, superconductors exist only at very low temperatures. The current record holding "high-temperature" (high Tc) superconductor, a complicated compound of mercury, barium, calcium, copper and oxygen, becomes superconducting at temperatures below of 133 Kelvin, which is 220 below zero degrees Fahrenheit (Schilling et al. 1993). These cuprate (copper-oxygen) superconductors were discovered nearly a decade ago but have yet to find technological applications. Their crystal structures are complex and their brittleness makes it difficult to use them in manufacturing wires and electronic components. Metallic fullerenes have simpler crystal structures than the cuprates, so fullerene superconductors might be easier to use in constructing loss-free electrical lines. The 18 Kelvin temperature where K3C₆₀ becomes superconducting is rather cold, but the transition temperature can be increased by replacing potassium with another atom. Rubidium fulleride, Rb₃C₆₀, superconducts up to 30 Kelvin. Adding small amounts of thallium moves the transition temperature to 43 Kelvin, the highest superconducting temperature found outside of the high-Tc cuprates and their cousins. The goal in superconductor research is to find an easily workable material which superconducts at or above 77 Kelvin (-320 Fahrenheit), the boiling temperature of nitrogen. The technology of making and storing liquid nitrogen is well developed. It can be produced in large quantities every day, and kept in insulated bottles.1 Many superconducting devices -- superconducting computers, for example -- could easily work at liquid nitrogen temperatures. Imagine the morning routine this could lead to: go into your office, pour a cup of coffee, and pour liquid nitrogen into your computer before starting to work. You would have to be careful what you poured into each container!

Fullerenes don't have to be shaped in balls or onions. Suppose we take a sheet of graphite, longer than it is wide, and roll it up to make a tube. Then we crimp the ends to tie off the dangling bonds, or attach hydrogen atoms to the ends. This new molecule is, logically, called a buckytube (Ross 1991). A typical buckytube is helical; a microscopic insect following a chain of carbon atoms in it would trace out a path that spiraled around the tube. A single-layer tube can be as small as 1 nanometer in diameter (Iijima and Ichihashi 1993; Bethune et al. 1993).2 The tubes can be conducting or insulating depending on the pitch of the spiral (Hamada et al. 1992).

Richard Smalley conjectures that buckytubes are self-healing (Ross 1991). If one is damaged, the dangling bonds should quickly rejoin, restoring the tube to its original strength. Buckytubes are essentially single molecules, so they should be strong, perhaps even more so than graphite fibers. If they can be made in macroscopic lengths, they may eventually replace the graphite fibers which strengthen objects such as jet fighters, golf clubs, and tennis rackets.

A meters-long buckytube would be much like the Sinclair monofilament science-fiction writer Larry Niven introduced in several of his Known Space stories -- impossible to see with the naked eye, immensely strong, and able to cut through almost anything. Smalley has suggested that the diamond fibers used to build the geosynchronous space elevator in Arthur C. Clarke's science-fiction novel The Fountains of Paradise could be replaced by buckytubes. If the tubes are stronger than diamond, this would reduce the weight of the elevator, making it more likely that we really could construct such a device.

Figure 2. Open-ended buckytube (Pederson and Broughton, 1992). Dark circles are carbon atoms, light circles are hydrogen atoms. Unlike most natural buckytubes, this one is unspiraled.

Buckytubes have other unique properties. My colleagues, Mark Pederson and Jeremy Broughton of the Naval Research Laboratory, have shown that an uncapped buckytube with its dangling bonds tied off by hydrogen atoms (see Figure 2) will attract polar molecules such as hydrogen fluoride and hold them inside the tube (Pederson and Broughton 1992). These "suckytubes" could be used to deliver small quantities of needed medicines directly to sites of infection. Indeed, they could serve as hypodermic needles. Broughton has noted that when a buckytube is placed inside a larger diameter tube it tends to remain inside. However, an electric field forces the smaller tube out of the larger. If the smaller tube contains a supply of medicine, it can be "injected" by this small hypodermic. Perhaps a third nanotube, inside the other two, can be used to push the medicine into the cell.

One could also use this mechanism to design a piston for a nanoscale motor. The inner tube is connected to a fullerene "crank shaft." Switching the electric field on and off moves the inner tube back and forth, causing the crank shaft to rotate.

Thomas Ebbeson and others at NEC, the Japanese-owned electronics firm, have shown that buckytubes can be formed with a tube of lead inside. The insulating tube together with its metallic core forms a microscopic wire, which suggests nanoscale electronic devices could be built using these buckywires to connect components.

Currently, small-scale devices are produced by lithography, where an electron beam traces a circuit onto the surface of a silicon wafer. However, the resulting circuits, confined to the silicon surface, are mostly two-dimensional. Suppose an engineer wants to connect component A to component C, but not to component B which sits between them? In two dimensions the connection must go around B without crossing other wires, which makes circuit planning a difficult art. Using buckywires adds a third dimension to the circuit construction and makes design much easier. Buckytubes with the correct pitch to their spiral can conduct electricity, so we might not even need a metallic core to produce circuit components. For example, a nano-scale coaxial cable could be constructed by placing one conducting bucky-tube insid e of another conducting tube with a slightly larger radius.

Like the ball-shaped fullerenes, ordinary buckytubes are composed of five and six-member rings. At the Fall 1992 meeting of the Materials Research Society in Boston, Shin-ichi Sawada and Atsushi Oshiyama of NEC showed calculations supporting the idea that seven-member rings may also be stable (Sawada and Oshiyama, 1992). If this is confirmed experimentally, a whole new array of shapes can be made. The simple spheres and tubes of the known bucky molecules all have "positive curvature." Seven-fold rings can have "negative curvature," which allows structures to be open, like the mouth of a trumpet for example. Sawada has shown how tubes can be joined together at angles to form "buckypipes" and other three dimensional structures. Such fullerenes could become the scaffolding for nanostructures, and nanomachines could use the scaffolds as workbenches.

In recent months we have learned a lot about how fullerenes form (Curl, 1993). Sophisticated quantum mechanical calculations can now model fullerenes containing up to 1000 atoms (Lu and Yang, 1993). This new knowledge should help us better understand these molecules. If we learn to produce and shape fullerenes economically, they will become the structural material of the future. Macroscopic length buckytubes will replace graphite fibers in composite materials. Nanoscale buckytubes and fullerenes will be the structural material for nanomachines and small scale devices. Metal-filled fullerenes may be used to construct electronic devices only a few nanometers apart. Bucky-pistons and bucky-injectors might be the basis for the dream of nanotechnology. We don't know if these things are possible, but in a few years we will have some answers to the many questions about this newest and most fascinating form of carbon.

Acknowledgments

I've had a lot of help with this article from coworkers at the Naval Research Laboratory who showed me preprints, talked about their current research and looked over the manuscript. The literature search was vastly simplified by the bibliography from bucky, the electronic repository for fullerene abstracts. Also, the electronic-mail mewsletter Physics News Update, prepared by Phillip Schewe of the American Institute of Physics, was an invaluable source of information about new discoveries in the fullerenes.

This article is a work of the U.S. Government, and is not subject to U.S. copyright. It originally appeared under the title "Buckystuff" in Mindsparks: The Magazine of Science and Science Fiction, Vol. 2, no. 1, 1994.

References

Bethune, D.S., Klang, C.H., de Vries, M.S., Gorman, G., Savoy, R., Vazquez, J., and Beyers, R. 1993. "Cobalt-catalysed growth of carbon nanotubes with single-atomic-layer walls." Nature 363: 605-607.

Chung F., and Sternberg, S. 1993. "Mathematics and the Buckyball." American Scientist 81: 56-71.

Curl, R.F. 1993. "Collapse and growth." Nature 363: 14.

Curl, R.F., and Smalley, R.E. 1991. "Fullerenes." Scientific American 265: 54-63. A general overview of fullerenes.

Hebard, A.F. 1992. "Superconductivity in Doped Fullerenes." Physics Today 45: 26-32.

Iijima, S. and Ichihashi T. 1993. "Single-shell carbon nanotubes of 1-nm diameter." Nature 363: 603-605.

Kratschmer, W., Lamb, L.D., Fostiropoulos, K., and Huffman, D.R. 1990. "Solid C₆₀: a new form of carbon." Nature 347: 354-358.

Kroto, H. 1988. "Space, Stars, C₆₀, and Soot." Science 242: 1139-1945.

Kroto, H.W., Leath, J.R., O Brien, S.C., Curl, R.F., and Smalley, R.E.. 1985. "C₆₀: Buckminsterfullerene." Nature 318: 162-163.

Lu, J.P., and Yang, W. 1993. "The Shape of Bucky Onions. Preprint.

Pederson, M.R., and Broughton, J.Q. 1992. "Nanocapillarity in Fullerene Tubules." Physical Review Letters 69: 2689-2692. Ross, P.E. 1991. "Buckytubes." Scientific American 265: 24.

Schilling, A., Cantonii, R., Guo J.D., and Ott, H.R. 1993. "Superconductivity above 130 K in the Hg-Ba-Ca-Cu-O system." Nature 363: 56-58.

Sawada, S. and Oshiyama, A. 1992. "Carbon Nanotubes: Electronic Band Structure." Presented at the Fall, 1992, Meeting of the Materials Research Society.

Footnotes

1. A common sight in a chemistry or physics laboratory is a row of large insulated dewars of liquid nitrogen. The dewars remind me of the Daleks on the Dr. Who television series, especially when they arrive by freight elevator with no humans aboard. (Return to text)

2. One nanometer, abbreviated nm, is one billionth (10^-9) of a meter. A typical atom has a radius of between 0.1 and 0.2 nm. For comparison, the C₆₀ buckyball is 1.4 nm in diameter.(Return to text)

Back to my my Home Page.

This page hosted by Get your own Free Home Page