But nanotubes always had their drawbacks: difficult to arrange precisely, they are also hard to wire to the outside world without losing much of their vaunted electrical conductivity.
Now a new carbon jewel has the caught the eyes of nanotech researchers, some of whom are already speculating that it might pick up where nanotubes left off in their bid to be the savior of electronics.
Oh, and by the way, you can make it with Scotch tape.
Called graphene, it is essentially a nanotube unrolled—a single layer of atoms arranged like a honeycomb. The difference may sound cosmetic, but when the goal is manipulating things that are a few atoms thick, going from tube to sheet makes a big difference.
Although graphene, too, faces many obstacles on the road to applications, its combination of exotic physics and high-tech potential is attracting scores of researchers. "For the moment there is at least a big hope that graphene might be the future," says physicist Andre Geim of the University of Manchester in England, who first isolated it in 2004.
The Great Carbon Hype
Today, Intel and other manufacturers stamp out microchips from dinner plate–size silicon wafers. By creating ever more detailed stamps, they cram chips with increasing numbers of the tiny switches known as transistors. But researchers believe that once silicon circuits slim down to 10 nanometers, which the semiconductor industry predicts will occur after 2020, they will start leaking electricity profusely. Already this year Intel and IBM announced that they would begin adding new materials to counteract leaky currents in their upcoming 45-nanometer transistors.
The question is what material comes next. Many have preened for the role. Graphene, like the carbon nanotube, meets the first requirement: it is a snappy conductor of electricity—better than many semiconductors. As in the nanotube, each carbon atom has three neighbors and an unused electron that is free to skitter around, hence conduction.
But nanotubes grow in dense thickets that are hard to separate and place with precision. To create circuits from them, researchers must attach relatively bulky wires that spoil much of their conductivity. "Carbon nanotube integrated electronics was hyped from the start," says nanotube-cum-graphene researcher Walter de Heer of the Georgia Institute of Technology. "Graphene is different."
With graphene, researchers envision stamping out circuits from large wafers, much as they already do with silicon. But perfecting those wafers has proved challenging. Another long-term "if," Geim says, is whether graphene can be carved into small pieces that actually work. But researchers are just learning. "Strictly two-dimensional materials didn't exist until 2004," he says.
Spite Turns Concept into Reality
As a concept, graphene is nothing new. A piece of graphite is simply a stack of graphene layers loosely stuck to each other, like a deck of cards. That is why scraping a pencil point across paper leaves a mark; the layers flake off in chunks and become caught in the paper fibers. And researchers had always thought of nanotubes as rolled up graphene sheets.
But most assumed that isolating a single pristine layer would be impossible. If the strain of being peeled from its neighboring layers did not shred it, they figured, its own heat would crumple it like newspaper on a campfire.
Nevertheless, some were game to try. Physicist Philip Kim of Columbia University began trying to flake off graphene layers in 2002 by dragging a tiny graphite rod with an atomic force microscope, which is like an exquisitely sensitive phonograph needle. In 2003 de Heer at Georgia Tech received a grant from chipmaker Intel to cook up graphene from a mix of silicon and carbon.
Then along came Geim. Creator of a gecko-inspired adhesive that made headlines four years ago, Geim is the type of researcher who likes to try something totally new every few years. Jealous of all the attention that nanotubes received over the years—"It's a beautiful system," he says—he began hunting for graphene in autumn 2002. "Spite is a very important motivation for physicists," he deadpans.
Researchers knew how to expose flawless surfaces from graphite chunks by peeling away thick layers with Scotch tape. Geim decided to push the method to its limit.
He quickly achieved 10-atom-thick layers. "Then you always ask yourself a question," he says: "Let's try nine." In a 2004 Nature paper, he and his colleagues reported success. They had placed a single graphene layer on silicon, connected electrodes to it and measured the amount of charge it carried when they applied different voltages.
"That was of course a big deal," laughs Kim, who had wagered on the needle-dragging method. "We realized we were completely scooped."
Massless Electrons Attract the Masses
Few groups took notice, Geim says, even after his team showed how to make graphene layers using what amounted to a pencil.
The field would not really take off until late the following year, when he and Kim independently confirmed a prediction with roots dating to the 1940s. They applied a magnetic field to graphene and observed that its resistance (the drag on a current) increased in a stepwise pattern called the quantum Hall effect. The effect comes in two varieties, both of which have garnered Nobel Prizes. Graphene, they observed, supports yet a third kind.
Researchers were floored. It was as if someone had slapped a piece of tar paper onto an open phone book and ripped out a page, says nano researcher Philip Collins of the University of California, Irvine. "You would expect the page to be junk," he says, but instead, "you can still get the phone numbers off of it." Those who study nanotubes and the quantum Hall effect alike have flocked to graphene, says Pablo Jarillo-Herrero, a researcher in Kim's laboratory and one of the flock.
Even particle physicists are excited, says materials theorist Antonio Castro Neto of Boston University, one of those who predicted the new quantum Hall effect. The key to graphene's abilities, including its high conductivity, comes from something that particle physicists are very familiar with. Most particles, including the electron, have mass. Like billiard balls, they do not move until given an energy boost by some impetus (such as a pool cue). The more energy they get, the faster they fly.
In contrast are photons, particles of light, which have no mass and move ceaselessly at a constant speed—the speed of light. What Geim and Kim confirmed is graphene's electrons effectively lose their mass. No matter how much energy they carry, they scurry at one four-hundredth the speed of light.
To describe particles produced in accelerators or elsewhere in the cosmos moving at close to light speed, physicists have to bring in special relativity, Einstein's theory of motion. The same goes for graphene.
Nanotubes share this property; they are actually better conductors than graphene because they force electrons to zip along a straight line. Graphene's flatness, however, makes its relativistic behavior more pronounced and easier to probe, Kim says.
Hyper Powerful or Overhyped?
One consequence of graphene's massless electrons, Geim says, is that it conducts electricity equally well if it contains many electrons, a handful or only one. Electrons typically bounce off of surrounding contaminants, which slows or stops them. When traveling through silicon, for example, they make it an average of 10 nanometers before scattering.
But graphene's electrons expand, in a sense, to cover large swaths, effectively riding over impurities like the tires of a monster truck over potholes. As a result, they go 100 times farther than silicon's—nearly a micrometer.
"It's so highly conductive," Geim says, "immediately the first thought, and the current thought, are for application to electronics."
The material that can replace silicon for the least cost will ultimately rule the day, but at scales of tens of atoms, graphene might perform as well as silicon or better, Kim says.
Of course, there are a few potential snags in the vision of nanometer-size circuits stamped onto Frisbee-size graphene wafers. Graphene is already "ridiculously" hyped, says Collins of U.C. Irvine. When a material is only three years old, he says, "it's kind of hard to start comparing that to something like silicon."
One problem is that Frisbee. Georgia Tech's de Heer reported some success with his method of cooking silicon carbide so that carbon atoms bubble to the top in one or a few atomic layers. But so far he has not observed the quantum Hall effect, suggesting that something crucial separates his graphene from the kind made with adhesive tape.
Then there is the challenge of how to toggle currents on and off in graphene. Silicon and other semiconductors transmit electrons of certain energies but block those of other energies. Nanotubes share this property, referred to as a band gap, but graphene does not.
The Light on the Graphene Horizon
Researchers have some ideas how to get over that hurdle. Geim's group reported in March that it had dialed a current up and down by shaping single-layer graphene into a very narrow hourglass. The presence of an electron in the waist of the hourglass blocked other electrons from passing—at room temperature, no less.
In another approach, Kim's group, working with IBM researchers, sliced graphene into 10-nanometer-wide ribbons, team member Jarillo-Herrero reported at the March meeting of the American Physical Society. Electrons carry energy only in specific amounts, or levels, and according to the team, electrons confined in graphene strips required larger doses of energy to reach the next level, creating a kind of band gap.
Etching these small shapes raises yet other problems. The carving process does not always work, although Geim expects that graphene can piggyback on improvements in silicon etching. It also leaves graphene with ragged edges, because the electrons that made chemical bonds with the excised carbon atoms remain like loose threads, interfering with passing electrons and dragging down the ribbon's current. Kim says graphene enthusiasts will have to find some way to mend the frayed ends.
"The nanotube solves that problem rather nicely by rolling up on itself," Collins says. "The ideal thing would be if we could come up with some structure that combines the two." Or if researchers learn to more precisely manipulate nanotubes, they would regain much of their attractiveness, says Jarillo-Herrero.
At least a decade probably remains before graphene will have to walk the walk. Silicon will have no real competitors, Geim says, until it reaches 10 nanometers. "The death of silicon was announced many, many times before," he says. "It's not easy to stop a train, if you try. And silicon is much bigger than that." For now, graphene is the little engine that might.