March 9, 2004
Stanford team invents scheme for stopping light on chip
By Esther Landhuis
A first-ever blueprint for capturing light on a chip may boost prospects for smaller, faster optical computers in the future.
Several years ago, physicists showed they could stop light in gaseous systems; the motion of photons -- light particles -- could be converted into the motion of electrons. However, these systems require ultracold temperatures and other extreme conditions that make their implementation on silicon chips nearly impossible.
Using computer simulations, Stanford scientists propose a light-halting strategy based on fundamentally different principles. Rather than using electron states, their scheme slows a light pulse by sending it through an array of semiconductor rods that trap the light's electromagnetic energy. The researchers say this approach can be realized on-chip at room temperature, paving the way for a revolution in information technology if the chips could be mass-produced.
"This work is opening up possibilities of manipulating light in ways you could never have dreamed of before," says Shanhui Fan, assistant professor of electrical engineering. A paper describing this system ? devised by Fan and applied physics doctoral student Mehmet Fatih Yanik ? appeared in the Feb. 27 issue of Physical Review Letters.
A key feature of the new strategy is that it keeps intact the photons, each of which carries a signature set of quantum mechanical properties. Fan says the ability to preserve these signatures on a microchip would clear a major hurdle in quantum computing, which promises processing capacities exponentially greater than that of existing computers. "What we propose can almost be envisioned as quantum memory," he says.
The challenge for now, though, is to use the scheme to build an actual device. Producing arrays of semiconductor rods with the high precision required for optical communication will be very difficult. However, Yanik doesn't think it will demand new technology. "It's just going to require us to use existing techniques in a state-of-the-art manner," he says.
Fan is confident their strategy will work because he says that in photonics, a field that marries light with electronics, theoretical calculations can predict natural behavior with uncanny accuracy. He says these calculations provide, in essence, "the cleanest experiment."
A serendipitous start
Yanik came up with a preliminary sketch of the optical scheme for stopping light somewhat fortuitously. Last April, he had been designing arrays of semiconductor rods for a different research project when Fan posed a seemingly unrelated question: Is it possible to stop light without using the unwieldy atomic setups developed previously?
Within a few days, Yanik realized that the semiconductor rod arrangement he had been toying with for the other project could be adapted for an optical strategy to stop light. He drew up a scheme and showed it to Fan. In several months, they were able to tweak it into a working model.
"Good things usually happen like this," Yanik says. "You strive for other things, and in the middle of nowhere you come up with something quite cool."
Yanik and Fan bolstered their paper-and-pencil theory with large-scale computations performed at the Pittsburgh Supercomputing Center. Funded by a National Science Foundation grant, the computations required 2,500 processors running in parallel for several hours. A typical desktop computer would have taken about a year to complete this task.
Eventually, the Stanford team hopes to control light pulses on a chip about the width of a human hair -- a tenth of a millimeter. To begin addressing whether this would be technologically feasible, Yanik is trying to first demonstrate the optical storage effect for lower-energy light waves, where the array structures are much larger and easier to make. To do this, he is collaborating with scientists at Palo Alto-based Agilent Technologies and Thomas Lee, a Stanford associate professor of electrical engineering.
"It looks hard -- damn hard -- but not impossible," says Lee. "There are no violations of fundamental laws of nature that need to occur."
Philip Hemmer, an associate professor of electrical engineering at Texas A&M University, points out that the Stanford scheme, if applied to quantum memory, might not be able to stop light pulses for the length of time required by many applications. Nevertheless, Hemmer says, "it is a novel approach that is likely to have some commercial applications and will stimulate other researchers in the field to develop even better devices."
Esther Landhuis is a science-writing intern at Stanford News Service.
This release was written by science-writing intern Esther Landhuis. A photo of Fan is available at http://newsphotos.stanford.edu . Slug: Â?stoplight_fan.jpgÂ?