Now, a third team has shown that optical-bit storage is inching closer to practicality--in the form of a crystal that promises to stop light. Using an yttrium-silica crystal seeded with praseodymium, an exotic metal, the team slowed light to the relative snail's pace of 45 meters per second.
Team leaders Philip R. Hemmer, a physicist at the Air Force Research Laboratory in Bedford, Mass., and Massachusetts Institute of Technology researcher Selim M. Shahriar, hope to go further. By fine-tuning the crystal's recipe, Shahriar says he expects to corral light "for as long as a second in the fairly near future." A crystal with that power could work in an all-optical telecom switch, or for transient data storage in optical or quantum computers. Many key processes in chemistry, physics, and biology--notably protein folding--take place in such fleeting instants that scientists can only guess at what's happening. To freeze the action, they need an ultrafast strobe light. Ahmed H. Zewail of California Institute of Technology won a 1999 Nobel prize for developing an X-ray strobe that emitted pulses lasting just a few hundred femtoseconds, or quadrillionths of a second. (It's hard to grasp how ephemeral this is, but if one femtosecond were stretched to a full second, that second would last 30 million years.)
Protein folding is faster still. So the quest for speedier strobes continues. Last year, a team from Austria, Germany, and Canada set a new record for brevity: an X-ray pulse lasting a mere 1.8 femtoseconds. The flash (the purple stream in the photo) was produced by shooting an infrared laser into a jet of ionized helium erupting from a tiny metal tube at Vienna Institute of Technology's Photonics Institute.
Even faster strobes could be coming. The researchers are confident that they will soon move down into the attosecond realm. An attosecond is 1,000th of a femtosecond. Such flashes might enable scientists to observe electrons in atoms. Certain wavelengths of infrared light can penetrate about one inch of human tissue. That's enough to open the door to some exciting vistas in medical imaging: pictures sharper than ultrasound and without the radiation of X-rays. Naomi J. Halas, an electrical engineer at Rice University, heads a team that is developing tiny, infrared-sensitive particles, which she calls nanoshells, to use as implantable drug pumps and instant tests for allergies or viruses.
Like seashells, nanoshells are composite structures. Glass spheres, 40 to 60 nanometers in diameter, are covered with 3 to 20 nm of gold. The image shows the progressive growth of the gold layer. By varying the core size and metal thickness, researchers can tune the particles to respond to specific frequencies. In one application, chemicals sensitive to allergens are attached to the gold. After the particles are ingested, any reactions can be spotted by drawing a small blood sample and testing it with various reagents.
Halas figures doctors would do most testing in their offices, ending the wait for results when tests are sent to a lab.
Another idea is to embed the shells in a plastic, along with a drug. The spheres absorb infrared energy, heat up, and trigger the plastic to release the drug. Similarly, researchers at Osaka University in Japan use infrared energy to recharge a battery for a pacemaker or insulin pump. -- By decade's end, light will no longer be able to print chips with thinner lines. That's because all conventional lenses have a positive index of refraction, meaning they bend light in only one direction. This limits their sharpness. Last year, John B. Pendry, a physicist at London's Imperial College, speculated that a negative-index lens was possible and that it could create lines far thinner than those imagined by chipmakers. Now, researchers at the University of California at San Diego have built such a lens for microwaves--but not light. Pendry remains upbeat about the prospect of a negative-index lens for light.
-- Computing with rainbows? University of Rochester researchers did it at CLEO. But only an optics engineer is likely to understand how. Here's the idea: A single pulse of light with many colors, each representing bits of data, is split in two. One half passes through a special filter. When the pulses are recombined, then split again, a spectrometer can read the results--in this case, the location of a desired citation in a 50-document database. This is no big deal, but the team thinks the principle could speed searches through astronomically huge databases.