Lighting up computers.
Within the next few years, computer chips will process data so rapidly that the wires carrying information between them will have difficulty keeping up. One possible solution is to transport the data using laser beams that blink on and off billions of times a second. Ling Liao, of Intel’s Photonics Technology Lab, is bringing us closer to that goal by figuring out how to build key optical components from the same silicon used to make the rest of a computer chip. That would make integration simpler and bring down costs.
But silicon is normally a poor optical material; it’s particularly hard to get silicon to modulate light–that is, to induce the blinking effect that encodes data. More exotic materials can make light fluctuate in response to a changing electric field, but silicon resists that approach. Liao found that if she put a thin layer of silicon dioxide between two slabs of silicon, she could quickly make electrons accumulate on both silicon surfaces, and remove them quickly as well. That alters any light waves that pass through the silicon, making them either dimmer or brighter. By building up and then releasing the charges using an alternating voltage, she can modulate light at 10 gigabits per second, and shes working on ever faster speeds. It may be a decade before her device finds its way into your computer, but eventually optical connections may be able to handle speeds hundreds of times faster than those seen in today’s PCs.
Making nanowires get in line.
Song Jin has developed a simple way to align nano-wires, allowing researchers to easily incorporate them into useful devices such as biosensors. Jin compresses nanowires afloat on water until they’re oriented in the same direction. He then transfers the aligned nanowires to another surface, on which he deposits electrodes using conventional lithography (see “The Future of Nanoelectronics”).
To illustrate his method, Jin created an array of nanowire-based transistors. Each cluster is a group of electrodes converging on a common electrode in the center. Hairlike nanowires bridge the gaps between the electrodes.
By outfitting the nanowires with molecules that bind to specific viruses or chemicals, researchers have demonstrated the value of his method for sensing applications: when a virus fastens onto a nanowire prepared this way, the wire’s conductivity changes, just as applying a voltage changes the conductivity of a semiconductor in a transistor. Each nanowire is sensitive enough to register a lone virus particle.
The world's smallest soldering iron.
What can you do with a nano-sized soldering iron? William King has a few ideas.
As a graduate student, King was invited to join the Millipede project at IBM’s Zürich Laboratory. The project’s experimental data storage device uses a cantilevered silicon bar with a tip just 20 nanometers wide to melt tiny divots into plastic, “writing” data at 50 times the density of today’s best hard drives.
By modeling the choppy flow of heat through the narrow silicon bars, King helped IBM engineers control it precisely–and so use a single tip to read, write, and even erase data. Since joining the Georgia Tech faculty in 2002, King has incorporated similar tips into nano soldering irons for making items as diverse as electronic circuits and tissue-engineering scaffolds. By cycling the tips to 1,200 °C and back to room temperature millions of times a second, King releases the “solder”–be it a semiconductor, an insulator, or a biocompatible polymer–far more precisely and quickly than researchers who deposit materials by lowering and raising silicon tips.
Stretchable electronic skin.
Bioengineers who hope to help paralyzed patients by melding electronics with nerve or brain tissue face a materials challenge: living tissue and microelectronics could hardly be more different. Most tissues are supple, while the semiconductors and metals used in electronics are brittle and stiff. As a result, the implanted electronics can irritate and damage surrounding tissue. It is precisely this material difference that Stéphanie Lacour is trying to bridge.
As a postdoctoral researcher at Princeton University, Lacour fabricated thin gold strips on elastic rubber substrates that could be stretched like a rubber band without losing electrical conductivity. The Princeton group, led by electrical-engineering professor Sigurd Wagner, then used these strips as the foundation of the first stretchable integrated circuit. Connecting small, rigid islands of conventional semiconductors with the gold strips, the researchers built simple electronic devices that still worked after repeated stretchings. While these circuits consisted of just a few transistors, they demonstrated a way in which engineers might make everything from electronic “skin” for robots to extremely flexible displays.
But it’s the potential applications in biology and medicine that are, Lacour says, “really thrilling.” Now a research project manager at the University of Cambridge in England, she is heading an effort to create implants that surgeons could use to repair nerves severed in an injury.
At the back of her mind, says Lacour, is the goal of creating electronic skin that could cover prosthetic limbs. Eventually, the electronics could be directly connected to a person’s nerves, providing mental control over the prosthetic and, through a network of sensors, “feelings” in the limb. Any application that requires an electronic interface with the nervous system could use stretchable electrodes, says Barclay Morrison, a professor of biomedical engineering at Columbia University. For example, neuroengineers are developing micro- electrode arrays that neurosurgeons have begun implanting in quadriplegic patients to allow them to control computer cursors or robotic arms with their minds (see “Implanting Hope”). But conventional metal electrodes are 100 million times stiffer than the brain tissue. “You’re implanting really rigid needles into the brain,” Morrison says. Lacour’s electrodes much more closely match the elasticity of brain tissue, potentially reducing the chance of damage.
Morrison has begun using Lacour’s stretchable metal electrodes in experiments to study brain injuries. The stretching of brain tissue during an accident can set off a chain of cellular events leading to the death of neurons days after the accident.
Morrison is re-creating the injuries by violently stretching thin slices of brain tissue. Lacour’s elastic electrodes can stretch with the tissue, recording in real time the changes in the electrical activity of the neurons.
Still, says Princeton’s Wagner, the field of stretchable microelectronics is very much in its infancy. It will be at least a decade, he predicts, before the technology is ready for use in consumer products like flexible displays.
But for now, bioengineers are just happy to have a way to bridge the material gap between tissue and electronics. A material that can stretch to twice its size and still be conductive is “unheard of,” Morrison says. “It’s incredible.”
The floppy screen.
Plastic semiconductors have already found their way into a number of electronic gadgets, providing displays that are bright and easy to read even in direct sunlight. But efforts to make larger plastic displays–say, a computer monitor that rolls up and slides into your bag-have hit a roadblock: the electronic circuitry that drives their colorful pixels consumes too much power, and existing materials don’t print well onto large sheets of plastic.
Working at Bell Laboratories in Murray Hill, NJ, Ashok Maliakal has homed in on this problem. One major challenge, he says, was to improve the circuits’ gate dielectric-the insulating layer that enables their transistors to switch properly from “on” to “off.”
The titanium dioxide boosts the material’s insulating properties, while the polymer makes it easy to print. Prototype circuits made with the material operate at one-third the voltage of those made with the polymer alone. That could mean displays that consume a lot less power. If Maliakal can incorporate his material into a mass-printing process, a new generation of flexible displays will be ready to roll.
Modeling the flows of light.
To understand the ambitions of Marin Soljacic, think of what the first semiconductor transistors did for the speed and power of computer circuitry–and then think photons instead of electrons.
By calculating the behavior of light in structures called photonic crystals, Soljacic is paving the way for devices that can process information at ultrafast speeds using light alone. The crystals’ microscopic honeycombs shut out light of any but a few particular wavelengths; as a consequence, they could offer a way to guide data-carrying light beams within microprocessors.
Soljacic has also shown how photonic crystals can enable light beams to interact and to control one another, so that photonic devices can carry out logic operations.
The all-optical computers predicted in the 1980s haven’t materialized, he says, because the effects on which photonic interactions depend are usually so weak that they appear only with beams too intense to be of practical use in microchip devices. But Soljacic has shown how photonic crystals can be designed to dramatically boost these optical effects so that they will occur even with very low-power light beams.
Ultimately, he thinks, it should be possible to use a single photon to control another in a photonic-crystal device. This would reduce energy consumption and heat production enough to make optical information processing practical–and it raises the possibility of all-optical quantum computers, which would expand computer power beyond recognition.
Cleaning up with nanoparticles.
With some 300,000 hazardous-waste sites scattered across the United States, cleaning up contaminated soil and groundwater is a daunting challenge. Chemical engineer Michael Wong is taking on toxic waste with tiny particles that can break down organic pollutants more quickly, and perhaps less expensively, than existing technologies.
Each particle measures just four nanometers across and consists of a gold granule spotted with palladium. The nanoparticles’ high surface-area-to-volume ratio enables them to split chemicals faster than larger particles could, but their real advantage lies in their unique combination of metals. Palladium on its own does an adequate job of breaking down toxic chemicals such as trichloroethylene (TCE), an industrial degreaser that is linked to cancer and contaminates 60 percent of sites overseen by the U.S. Environmental Protection Agency’s Superfund project. But spotting the palladium on gold has a synergistic effect: the combination particles catalyze the removal of chlorine atoms from TCE molecules 100 times as fast as palladium particles alone.
Wong is already developing ways to incorporate his nanoparticles into filters for treating contaminated groundwater. To keep the particles in place, he has designed a method of growing them directly on the inner walls of hollow fiber tubes. He plans to test the system in the field this fall.