The reason laser light works so well in everything from cd players to surgery is that it's "coherent"--that is, ordinarily separate photons of light merge to make one powerful light wave that can be aimed with terrific precision. Getting matter to do the same thing is another story. Although quantum mechanics tells us that atoms are really waves, they're waves that don't like to merge with one another except at temperatures mere billionths of a degree above absolute zero. Which is why, for the past several years, mit physicist Wolfgang Ketterle and his students have been trying to make sodium atoms colder and colder, by using magnetic fields and laser beams to slow the atoms to a near stop. Three years ago, Ketterle's group merged a bunch of atoms into a single meta-atom (he prefers to call it a giant meta-wave). A year later they used fluctuating magnetic fields to shape the meta-wave into a beam much like a laser's, but they had no way to prove it to their scientific colleagues. Proof came a few months later, when they generated two separate meta-waves, made them overlap, and snapped a photograph. The resulting image showed alternating light and dark stripes--a so-called "interference pattern," which can only be created by coherent waves. In January 1997, Ketterle announced the first atom laser. The potential of atom lasers is huge, but Ketterle winces when you ask him about specific applications. Under duress, he'll admit that they might someday be used to build microscopic mechanical structures atom by atom, or to create incredibly fine circuitry on electronic chips, but "such applications are at least a decade away," he says. "The real significance of this work is that we're approaching the ultimate control over the motion of atoms. Learning to perfect that control is our challenge over the next several years."
FINALISTS
Slip SlidingArgonne National Laboratory's Carbon Film CoatingInnovator: Ali Erdemir
Friction is eating away car engines, hip replacements, and the joints that move the mirrors of the Hubble Space Telescope. But if Ali Erdemir has his way, this constant grinding could be brought to a near standstill. Last July, Erdemir, a materials scientist at Argonne National Laboratory in Illinois, finished developing a thin carbon coating that sticks to pretty much any plastic, ceramic, or metal and makes them slicker than any oil could. Erdemir claims that these "near frictionless" carbon films, as he calls them, have 40 times less friction than Teflon, and 100 times less than lubricated steel.
All carbon films are hard and smooth because almost all the molecular bonds at the surface get filled when the material cools. When something slides across them there's no hook for a chemical or physical reaction. But with his new carbon film, Erdemir went so far as to satisfy bonds that are about four hundred-millionths of an inch below the surface, but which also contribute to friction. The trick is in the way he deposits the carbon film. In a vacuum chamber, he takes a carbon-rich gas such as methane and adds heat, which turns it into a plasma, or a highly charged cloud of ions and electrons. When the plasma settles on the surface, the energetic particles react with the surface molecules, filling up the available bonds. "It's a very special plasma composition and very elaborate control of that plasma that seem to be the key to the ultralow friction of these films," he says.
So far, Erdemir has licensed the technology to two automotive firms. One wants to make turbochargers that last longer, and the other is trying to make diesel engines that don't need lubricants based on sulfur, which the epa is phasing out. An aerospace firm has also expressed interest in using it, to coat the moving parts in satellites and telescopes.
Sound PowerMacroSonix's Resonant Sound TechnologyInnovator: Tim Lucas
Imagine a compressor in your refrigerator with no pistons, crankshafts, or lubricated bearings. Instead, all the work is done by sound waves bouncing around in an empty cavity.
When this idea first began bouncing around Tim Lucas's head ten years ago, his fellow physicists told him it would never work. Sound waves, they pointed out, can store only a relatively small amount of energy before turning into jagged shock waves that dissipate any added energy as heat. At least that's what happens when a wave travels through the open air, or through a cylindrical "wave guide." Undaunted, Lucas experimented and found that by shaping the sound chamber, or resonator, into something like a cone or a bulb, he could keep shock waves from forming. "Most of the research had been done in a simple cylindrical tube, and it turns out that's the one resonator guaranteed to give you a shock wave," says Lucas. "There's an infinite family of resonators that can give you non-shocked waves." In his technology, which he calls Resonant Macrosonic Synthesis, sound waves store thousands of times more energy than previously thought possible.
Lucas, who started his own company, MacroSonix Corp. in Richmond, Virginia, to develop rms, has licensed it to one company (he won't say which) for refrigerator compressors--the part that compresses and circulates the coolant. The coolant passing through the cavity would be compressed when it encounters the high-pressure portion of the wave. Other applications might include cooling computer chips; "micronizaton," which is the pulverizing of particles down to microscopic size; and filtering out particles from factory exhaust (the sound waves would cause the particles to clump together). "We believe rms is a new primary technology, something that functions at a fundamental level of physics," says Lucas.
Quantum ToyMIT's Molecular Quantum ComputerInnovator: Neil Gershenfeld
For 30 years, engineers have been very clever about making computer chips ever smaller and faster, but sooner or later they're bound to run smack into the limitations of physics. No matter what, transistors cannot be smaller than a single atom and electric signals cannot travel faster than light. In quantum physics, however, particles can do all sorts of impossible things, such as exist in two places at once, or point up and down at the same time. This suggests that a quantum computer could conceivably make a great many calculations at once. But so far scientists haven't been able to demonstrate one.
A few years ago, mit physicists Neil Gershenfeld and Isaac Chuang (now a physicist at ibm's Almaden Research Center in California) began brainstorming for a way to overcome a major obstacle: how to peek at the particles that make up a quantum computer without disturbing them. Key to doing so, Gershenfeld thought, was nuclear magnetic resonance imaging, in which radio pulses probe the structure of molecules. "Then I had jury duty in Cambridge," he says. "I was stuck for a day in the jury room and all I had with me was the bible of nmr. That's when I finally understood nmr well enough to connect the classical to the quantum." Gershenfeld learned that as an atom receives an nmr radio pulse, it responds with a pulse of its own in a way that reveals its spin or some other quantum mechanical characteristic. In a sense, Gershenfeld discovered how to turn an ordinary molecule with its many atoms, into a quantum computer.
Gershenfeld and Chuang worked out how to send a series of radio pulses to make the atoms in complex molecules perform simple mathematical operations. By early 1997, the two physicists had coaxed a vial of a common organic compound to add one plus one. Last fall, they got their quantum computer to pluck one name out of a list of four in an even smaller fraction of time than a conventional computer could. "It's a toy problem," Gershenfeld concedes, "but it's the first demonstration of a calculation that's faster on a quantum system."
Memory MoleculeNanogen's DNA Optical Storage MediaInnovator: Michael Heller
The first 331/3 rpm vinyl record, introduced in 1948, was about 12 inches wide and held half an hour of music. Today a 5-inch compact disc can hold more than an hour of digitally perfect sound. In the future, the same amount of music may require a disc no bigger than a dime, thanks to nature's own information storage molecule, DNA.
Right now, digital information in a cd is stored as millions of tiny nonreflective dots, each four hundred-thousandths of an inch across. When your cd player shines a laser light on the disc, the light can either reflect or not reflect. On or off. That's one bit of information. But the dots can't be made much smaller than they are now, otherwise the laser won't detect them. "We'd reached a fundamental limit for optical storage," says Michael Heller, a physical biochemist at Nanogen in San Diego. Heller had been working on integrating synthetic dna molecules into computer chips, for diagnostic testing of disease. "We sort of flipped the whole thing around," he says. "We started wondering if we could take these biological molecules, take the principles that allow you to have molecular information storage, and use dna to develop high-density optical memory."
Heller used synthetic dna as a molecular support structure to hold light-responsive molecules called chromophores, which fluoresce when you shine a laser on them. The absence of light-emitting chromophores on a particular spot constitutes a binary "zero." If he plants thousands of chromophore-bearing dna strands in that same spot and shines the laser light on them, he reads a binary "one." To cram more than one bit onto a single spot, Heller puts down different types of chromophores that each respond to a different frequency of laser. In all, he has identified about 100 chromophore types and has been able to store 256 bits of information on one spot.
Heller hasn't put his dna molecules into a disc yet, but he has shown that they work. Theoretically, he could increase the capacity of cds or cd-roms a thousandfold. Says Heller: "People are beginning to say, 'If nanotech is going to be based on self-assembly and mimics of biological systems, dna sounds like it's going to be part of this.'"
