PORTLAND, Ore.—The world's first "truly nanoscale" silicon waveguide for integrated on-chip optical communications was recently claimed by the U.S. Department of Energy's Lawrence Berkeley National Laboratory. Based on its invention of a new type of quasi-particle called a hybrid plasmon polariton, Berkeley Lab's HPP sidesteps optical losses plaguing earlier attempts at silicon photonics with a new operating mode that makes the best of photonic and plasmonic systems—combining high quantum confinement with low signal loss, thus opening the door to the nanoscale on-chip lasers, quantum computing, and single-photon all-optical switches.
Created in the lab of Xiang Zhang, principal investigator with Berkeley Lab’s Materials Sciences Division and director of the University of California at Berkeley’s Nano-scale Science and Engineering Center, the breakthrough was also facilitated by doctoral candidates Volker Sorger and Ziliang Ye, who claim that HPPs will enable a new era of nanoscale waveguides for intra-chip optical communication, signal modulation, on-chip lasers and bio-medical sensing.
Quasi-particles called surface plasmon polaritons (SPPs) were already known to be created by directing waves of light across a metal surface to generate electronic surface waves—called plasmons—that then interact with photons. Unfortunately, SPPs suffer significant signal losses when propagating through metal. The Berkeley Lab researchers solved this problem by adding a low-k dielectric layer between the metal and the semiconductor to form a metal-oxide-semiconductor architecture that redistributes incoming light waves into the low dielectric gap where optical losses are less.
3-D image overlap of the deep sub-wavelength HPP mode signal (red spot) that indicates the waveguide's potential to create strong light-matter-interaction for compact and highly functional photonic components. (Photo courtesy of Zhang group).
The resulting HPPs propagate more freely, providing designers with a best-of-both-world's scenario where nanoscale waveguides can be cast on standard CMOS chips with optical properties rivaling exotic III-V compounds. The researchers estimate that it will take two to five years to commercialize the technique.
Funding for the project was provided by the National Science Foundation’s Nano-Scale Science and Engineering Center.
"hybrid plasmon polariton"? Someone's been watching too much Star Trek. By diverting the INCOMING light back into the dielectric, you cannot generate plasmons...so, it's ultimately a dielectric waveguide with a fancy name...if I got it right, then ....yawn
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.