PORTLAND, Ore.—Scientists have long been interested in integrating lasers onto silicon to enable on-chip communications using light instead of electrons, which would lead to faster communication and increased bandwidth. But, unfortunately, there is a lattice mismatch between silicon and the traditional III-V materials used to craft semiconductor lasers.
Now, researchers at the University of California at Berkeley claim to have surmounted that obstacle by growing indium-gallium-arsenide nano-pillars vertically. The researchers claim their technique is adaptable to mass production of robust on-chip structures for optical interconnects, field emitters, nonlinear optical signal generation, sensors, solar cells, displays and nano-fluidics.
"The lattice mismatch between silicon and III-V materials, which would ordinarily cause cracking, is mitigated by the small footprint of the nano-pillars," said doctoral candidate Roger Chen, who worked with professor Connie Chang-Hasnain on the project.
The hexagonal indium gallium arsenide (InGaAs) nano-pillars were grown directly on the top layer of a silicon chip at 400 degrees Celsius—so as not to damage any underlying CMOS circuitry. The technique can also make use of standard lithography and metal-organic chemical vapor deposition (MOCVD) techniques, allowing it to be integrated with CMOS lines.
The resulting nano-pillars form into vertical hexagonal structures that taper to a point, providing a cavity inside whose dimensions are matched to the wavelength of light to be produced. The resulting light-trapping optical cavity forces light to circulate up and down in a helical fashion, amplifying the light by virtue an optical feedback mechanism. Lasing in the infrared spectrum of 950 nanometers was demonstrated at room temperature using an external optical pump. Next the researchers plan to develop electrically pumped on-chip lasers, which they have already demonstrated for III-V on silicon LEDs.
Funding for the project was a provided by the Defense Advanced Research Projects Agency (DARPA) and a Department of Defense (DoD).
Growth, crystallography, and optical studies have shown nanoneedles to have a unique, single crystal dislocation-free wurtzite lattice, which can be used for lasing and second harmonic generation. The high crystal quality of III-V nanoneedles grown on silicon is unmatched, making nanoneedles an exciting approach towards developing practical silicon-based optoelectronics for the future. Source: http://light.eecs.berkeley.edu/cch/nanoneedles.html.
I presume they would need to be grown at the place where they are needed...random placement of some lasers on top of the chip will probably not work;-)...some sort of seed for epitaxial growth???...dr Kris
A step added to the long road of opti-tronics?
Or how would you call that?
Instead of electrical current, use light.
In theory that would make faster computers and no FCC regulations needed... that would reduce the cost of electronics at the long term.
Ooops, sorry, optronics ;-)
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.