July 11, 2007
Modern redux of classic experiment illuminates nanoscience of 'plasmonics'
By David Orenstein
Look at a shiny piece of metal, like a butter knife, and you would be reasonable to conclude that light cannot move through it. Under certain circumstances, however, you would be wrong.
The science of plasmonics describes how metals can essentially transmit and manipulate light waves at length scales much smaller than their wavelengths. Now, by redoing a classic optics experiment with plasmonics, engineers at Stanford have made key insights into the nature and the practical limits of this up-and-coming nanoscale information technology.
"The optical properties of metals keep surprising us," says Mark Brongersma, a Stanford assistant professor of materials science and engineering, who co-authored the report on the experiment published online July 1 in the journal Nature Nanotechnology. "In my mind, it is now becoming apparent that subwavelength control over light afforded by nanometallic structures (plasmonics) will impact many areas of technology."
Plasmonics has great potential because it embodies the best of two technological worlds: the tiny scale of electronics and the immense data-carrying capacity of photonics. Application areas could include improved information processing, inexpensive solar power and high-resolution imaging. Before these can be realized, however, engineers need to understand more about the fundamental properties of the emerging field.
To illuminate the burgeoning science, Brongersma and former student Rashid Zia decided to put a plasmonic spin on Young's double-slit experiment. In 1805, scientist Thomas Young showed that light acts as a wave by shining it through a screen with two slits. The pattern formed on the other side of the screen showed tell-tale signs of wave phenomena, such as diffraction and interference.
The Stanford team of researchers expected that plasmonic "waves" called surface plasmon-polaritons (SPPs) would move along a metal surface in the same way that light waves move in air or glass. SPPs, after all, are generated when, under the right conditions, light strikes a metal. The electric field of the light jiggles the electrons in the metal to correspond to the light's frequency, setting off density waves of electrons. The process is analogous to how the vibrations of the larynx jiggle molecules in the air into density waves experienced as sound.
Brongersma and Zia, now an assistant professor at Brown University, recreated Young's experiment by connecting two squares of gold foil with two gold strips 48 billionths of a meter thick and two millionths of a meter wide. The strips, which guide SPPs down their length, act as the slits. To see what pattern the SPPs would make when they left the slits, the researchers used a special microscope called a photon scanning tunneling microscope.
What the pair saw was a pattern very similar to what Young saw with light, marking the first time researchers have actually observed diffraction of SPPs. This discovery shows that SPPs share many of the properties of the light waves that generate them, but only under certain conditions.
"It turns out to be dangerous to 'blindly' assume that SPP always behave like light waves," Brongersma says.
SPPs also have their limitations, Brongersma says. What he and Zia discovered along the way is that some shapes of metals will not carry SPPs at the nanoscale because they experience a "diffraction limit." The gold strips the pair employed in the experiment, for example, had to be millionths of a meter wide (microscale) rather than billionths of a meter (nanoscale). This is discouraging, because nanoscale strip shapes are the form of the wiring commonly found in computer chips.
The realization that some geometries of metal won't work at the small scales of modern electronics is a sobering but useful "heads up" for engineers who would seek to make use of plasmonics in computer chips, Brongersma says. They'll have to adapt to geometries in which light can propagate between two parallel metal surfaces, such as nanoscale coaxial cables, even when the distance between the metal surfaces is much smaller than the wavelength of light.
Yet he remains optimistic that engineers will do so, perhaps because plasmonic devices may provide the only practical solution to link micrometer scale photonic structures to nanoscale electronic components, thereby bringing greater amounts of data into environments where it can be processed very fast.
"In 5-10 years, I think we will see plasmonic components appear in integrated circuits," he says. "They will be able to significantly increase the synergy between electronic and conventional photonic components on a chip."
David Orenstein is the communications and public relations manager at the Stanford School of Engineering.