AUG. 24-31, 1998
|SPECIAL REPORT CONTENTS|
Silently and efficiently, the new team member toils away in a chemistry lab at the University of California at Santa Barbara. With perfect precision, she lays down an ultrathin layer of an organic substrate. Onto this, she deposits interlocking calcite crystals, atom by atom. The two layers bond into a delicate crystal lattice. Under a microscope, it calls to mind the flawless thin-film layers on a silicon chip.
But there is no clean room, vacuum chamber, or chip gear in this lab, where professors Galen D. Stucky and Daniel E. Morse brainstorm new materials. For that matter, the ''team member'' is no ordinary staff researcher. She's a mollusk--an abalone. And like so many of nature's creations, she has acquired, through millions of years of evolution, an exquisite form of molecular machinery to create her shell--machinery that leaves today's best fabrication tools in the dust.
At Santa Barbara and hundreds of other research centers, scientists and engineers are getting ready for the next great technological revolution--the leap into the world of the very small. Nature has an unrivaled genius for designing and producing tough, versatile materials--from seashells to spider's silk--in self-replicating ''factories,'' and at an atomic scale. Scientists want to clone this ingenuity, to forge new industries of the 21st century. ''Atom by atom--that's how nature designs and builds things,'' muses Cherry Murray, Physical Research Lab director at Lucent Technologies' (LU) Bell Labs. ''If you could influence design at such a scale, you could make any material you ever wanted.''
BYE BYE PC BUGS? That ultimate goal is still mostly a glimmer. But the high-tech landscape is on the brink of change. The development pipelines at many high-tech companies already showcase a whole new breed of miniaturized marvels with capabilities well beyond today's chips. Over the next half-decade, these so-called ''microelectromechanical systems'' (MEMS)--which combine sensors, motors, and digital smarts on a single sliver of silicon--are likely to supplant more expensive components in computer hardware, automobile engines, factory assembly lines, and dozens of other processes and products. The operating software for these devices, now in the process of being written, is expected to suffer neither the bugs nor the bloat of today's PC programs.
Going somewhat further out--probably 15 to 20 years--high-tech visionaries foresee a transition that's far more radical and disruptive. Its quintessence won't be smaller, cheaper, faster electronics--though we will have all that in abundance. The transition scientists speak of involves nothing less than the hijacking of nature's own creative machinery.
In medicine, this spells the ability to repair or replace the body's failing organs. In manufacturing, it means coercing molecules to assemble into useful devices--the same way that crystals and living creatures assemble themselves. The coming wave of miniaturization and molecular electronics--sometimes called ''nanotechnology''--is taking shape at the intersection of chemistry, physics, biology, and electrical engineering. And if it crests as many scientists predict, it will bring a wholesale industrial transformation, more dramatic than the late-20th century flowering of microelectronics.
No one dismisses the enormity of the challenges. Atoms, at room temperature, inhabit a turbulent world ruled by forces we don't fully understand. Today's best theories, scientific instruments, and computer simulations provide only imperfect access to this domain.
Why, then, do so many scientists believe in a Molecular Revolution? Because some of the necessary capabilities are already within reach. As scientists at Bell Labs figure it, the widths of the circuit lines that make up electronic elements on chips will shrink 80%, to just 50 nanometers, by 2010. That's 50 billionths of a meter--the distance of about 300 atoms tethered in a row--and roughly the thickness of the protein layers in the abalone's shell.
In other words, engineers are already plunging deep into nature's hidden preserve. And in life sciences, they've gone even further. Biotechnologists can tailor antibodies that fight cancer. And they can slip new genes into plants or animals, to produce plastics or drugs (page 86). Nano-engineers believe they can build on this base, and write new recipes or scripts that will instruct unruly atoms to form desired materials. The first successes--though not quite at the atomic level--are already visible. Bell Labs boasts of chips that literally pick themselves up off a substrate. And at the University of Rochester, researchers coax polymeric molecules to form hollow cylinders and solid rings. Once perfected, such devices alone stand to bring enormous savings to many industries. Says Venkatesh Narayanamurti, dean of the engineering college at UC-Santa Barbara: ''The Internet is nothing compared to what's coming.''
QUAKE WARNINGS. The shifting economics are already evident with the arrival of MEMS. Early mass-produced applications include tiny automobile accelerometers, which trigger air bags in a crash, and a host of other industrial sensors. But as sketched out by Sandia National Laboratory's MEMS research chief, Paul J. McWhorter, the next wave will be far more dramatic. Deployed in networks, MEMS will sense one another and configure themselves to perform information-processing tasks. In the not-so-distant future, MEMS will steer our planes, monitor our health, and warn us of earthquakes, faulty aircraft parts, or cracks in bridges.
Since most MEMS chips have less circuitry than memory chips or microprocessors, they can be fabricated inexpensively on older chipmaking lines. That signals the first stage of a whole new paradigm for technology evolution, says Bell Labs' Murray. In the past, breakthrough devices carried a premium when they hit the market, and they were deployed sparingly. But with MEMS, Murray says, from the beginning, ''you can manufacture in high volume at low cost. This has the potential to break down the old economic order.''
Telecom, she says, will be an important testing ground. At Lucent, Siemens, and Tellabs (TLAB), engineers want to shrink whole hunks of the next-generation telephone and data networks onto MEMS, possibly beginning in as little as two years. The arena they have targeted is a new mode of high-speed optical transmission called wave division multiplexing, in which a single beam of light is split into multiple colors, or channels, and zipped through fiber. Lucent's Bell Labs already has shown how today's pricey prototypes could be replaced by tiny microscopic mirrors sculpted onto MEMS. Combine that with satellite technology, and long-distance communications costs drop to zero. ''This changes how meetings are conducted, how banking is done, how information is transported, everything,'' Murray says.
Wireless MEMS offer similar promises. At the University of California at Los Angeles, a team of 50 researchers is working on single-chip MEMS radios that could replace the $500 cards used in today's wireless data networks. In about five years, says William J. Kaiser, chairman of UCLA's electrical engineering department, all PCs and palmtop computers will come with radio MEMS. ''And they'll be embedded in the ceiling in your office cafeteria, your hotel room, your airplane,'' he predicts, ''all of them seamlessly linked to the Internet.''
Around the same time, says California Institute of Technology physics professor Michael L. Roukes, vast numbers of wireless MEMS could be deployed as seismic and metal stress sensors, in patient-monitoring systems in hospitals, and as communications systems on satellites. ''This is absolutely not science fiction,'' Roukes insists. ''It's here and now.''
In computing, disk storage capacity could be increased a hundred-fold by a MEMS-based instrument called an atomic force microscope (AFM). Such ''probe'' microscopes, invented at IBM (IBM) and Stanford University, produce images of atoms by dragging a superfine needle over a surface (above). But these new probes also enable data storage on a near-atomic scale, since they can nudge atoms from one position to another. It's a painstaking process, today. But researchers can speed it up by clustering hundreds of microscope tips on the same silicon device.
Health care is another beckoning frontier for MEMS--and its economics closely mirror those of the computer industry. Lawrence Livermore National Laboratory has developed a MEMS alternative to today's multithousand-dollar DNA sequencers. Its parts can be produced for less than $100. And the same notebook-size box can also include a miniature blood analyzer. Blood analyzer chips could continuously monitor an out patient's blood and radio the doctor at the first sign of a crisis.
Sturdier, more aerodynamic airplanes are another big goal. By studding the back and wings of a plane with thousands of MEMS-size flaps, engineers can alter the lift and drag to minutely control the plane. In theory, planes with MEMS wouldn't rely so much on wear-prone rudders, wing flaps, or tail elevators. The concept has been proven in wind tunnels. And the military's Defense Advanced Research Projects Agency (DARPA), which is seeding MEMS research to the tune of $47 million a year, is now test-flying one-seventh-scale planes.
DISTANT DREAMS. DARPA is interested in military applications, such as radio MEMS that can assemble themselves into networks in a battle or crisis situation. ''You drop them off the back of a truck, they find each other, and establish communication links,'' says Albert P. Pisano, director of DARPA's MEMS program. In addition, DARPA is spending millions on so-called microbot planes that would be outfitted with MEMS sensors to detect biochemical weapons or relay images of enemy positions.
But in business environments as well, such devices can spell huge efficiency gains. Office networks wouldn't need to be configured manually--they could assemble themselves right out of the box. Laptop computers and cell phones with built-in radio MEMS and GPS circuits will always know exactly where they are, and how to connect to the nearest Internet backbone. MEMS also can be assembled into more complex robotic organisms. Kristofer S.J. Pister, an electrical engineer at UCLA, and two Vanderbilt University mechanical engineers are among those who are developing insect-like robots, which one day will be able to handle assembly jobs in electronics factories, for example, or hunt for survivors in collapsed buildings.
To be sure, the vision of nanotechnology pur-ists--teaching smart devices to assemble themselves from the ground up, atom by atom--is still a distant dream. As Caltech's Roukes puts it, ''The only true nanotechnologist today is Mother Nature.'' But slowly, humans are learning to mimic her handiwork. For example, by vaporizing carbon in vacuum chambers, Rice University physicist Richard E. Smalley and his colleagues create atomically perfect carbon nanotubes that don't exist in nature.
NO RETURN. The tubes are chemically stable, about 100 times tougher than steel, and scientists have just begun to explore their possible applications in industry. This summer, a Dutch research team turned one of Smalley's nanotubes into the world's first single-molecule transistor functioning at room temperature. Just 4 to 5 atoms in diameter, the circuit shattered a size barrier that ordinary silicon devices can't hope to cross, and it offered the first physical proof that atom-scale electronics are feasible. IBM has also demonstrated carbon nanotube transistors.
How far off is a commercial device? Smalley admits that going from one experimental carbon transistor to one trillion of them on a chip is a staggering challenge. ''We have good days and bad days,'' he says.
That phrase perfectly expresses the zeitgeist of nanotechnology at the turn of the millennium. But few scientists seem inclined to turn back. The results will be a big surprise to economists who believe that industry already has reaped all the easy benefits of the Information Revolution. The revolution has barely begun.
Updated Aug. 13, 1998 by bwwebmaster
Copyright 1998, Bloomberg L.P.