Engineers efficiently 'mix' light at the nanoscale

Engineers efficiently ‘mix’ light at the nanoscale
Light emitted from the underside of the cavity. THe dotted outlines represents the orientation of the cadmium selenide nanowire.

The race to make computer components smaller and faster and use less power is pushing the limits of the properties of electrons in a material. Photonic systems could eventually replace electronic ones, but the fundamentals of computation, mixing two inputs into a single output, currently require too much space and power when done with light.

Researchers at the University of Pennsylvania have engineered a nanowire system that could pave the way for this ability, combining two waves to produce a third with a different frequency and using an to amplify the intensity of the output to a usable level.

The study was led by Ritesh Agarwal, professor of materials science and engineering in Penn's School of Engineering and Applied Science, and Ming-Liang Ren, a post-doctoral researcher in his lab. Other members of the Agarwal lab, Wenjing Liu, Carlos O. Aspetti and Liaoxin Sun, contributed to the study.

It was published in Nature Communications.

Current computer systems represent bits of information—the 1's and 0's of binary code—with electricity. Circuit elements, such as transistors, operate on these , producing outputs that are dependent on their inputs.

"Mixing two input signals to get a new output is the basis of computation," Agarwal said. "It's easy to do with electric signals, but it's not easy to do with light, as light waves don't normally interact with one another."

Engineers efficiently ‘mix’ light at the nanoscale
A schematic of the optical cavity.

The difficulty inherent in "mixing" light may seem counterintuitive, given the gamut of colors on TV or computer screen that are produced solely by combinations of red, green and blue pixels. The yellows, oranges and purples those displays make, however, are a trick of perception, not of physics. Red and blue light are simply experienced simultaneously, rather than combined into a single purple wavelength.

So-called "nonlinear" materials are capable of this kind of mixing, but even the best candidates in this category are not yet viable for computational applications due to high power and large volume constraints.

"A nonlinear material, such a , can change the frequency, and thus the color, of light that passes through it," Ren said, "but you need a powerful laser, and, even so, the material needs to be a many micrometers and even up to millimeters thick. That doesn't work for a computer chip."

To reduce the volume of the material and the power of the light needed to do useful signal mixing, the researchers needed a way to amplify the intensity of a light wave as it passed through a cadmium sulfide nanowire.

The researchers achieved this through a clever bit of optical engineering: partially wrapping the nanowire in a silver shell that acts like an echo chamber. Agarwal's group had employed a similar design before in an effort to create photonic devices that could switch on and off very rapidly. This quality relied on a phenomenon known as , but, by changing the polarization of the light as it entered the nanowire, the researchers were able to better confine it to the frequency-altering, nonlinear part of the device: the nanowire core.

"By engineering the structure so that light is mostly contained within the cadmium sulfide rather than at the interface between it and the silver shell, we can maximize the intensity while generating the second harmonic," Ren said.

Like a second harmonic played on a guitar string, this meant doubling the frequency of the . Information in a photonic computer system could be encoded in a wave's frequency, or the number of oscillations it makes in a second. Being able to manipulate that quality in one wave with another allows for the fundamentals of computer logic.

"We want to show we can sum two frequencies of light,"Agarwal said, "so we simplified the experiment. By taking one frequency and adding it to itself, you get double the frequency in the end. Ultimately, we want to be able to tune the light to whatever frequency is needed, which can be done by altering the size of the nanowire and the shell."

Most important, however, was that this frequency mixing was possible on the nanoscale with very high efficiency. The researchers' optical cavity was able to increase the output wave's intensity by more than a thousand times.

"The -changing efficiency of cadmium sulfide is intrinsic to the material, but it depends on the volume of the material the wave passes through," Agarwal said. "By adding the silver shell, we can significantly decrease the volume needed to get a usable signal and push the device size into the nanoscale."

Explore further

Researchers make first all-optical nanowire switch

More information: "Enhanced second-harmonic generation from metal-integrated semiconductor nanowires via highly confined whispering gallery modes." Nature Communications 5, Article number: 5432 DOI: 10.1038/ncomms6432
Journal information: Nature Communications

Citation: Engineers efficiently 'mix' light at the nanoscale (2014, November 13) retrieved 23 July 2019 from
This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no part may be reproduced without the written permission. The content is provided for information purposes only.

Feedback to editors

User comments

Nov 13, 2014
Has anyone thought of trying an optical computer using 3 (or more) different wavelengths of light for computer design (instead of just "white" or whatever and black)?
I remember when the first color tv cameras came out, they used a device in the lenses consisting of coated prisms to separate light into 3 image sets (R, G, B).
In an optical computing system, these 3 colors could be used to represent at 6 different values: B(lack )= 0, R = 1, G = 2, up to R+G+B = 6 . Adding a 4th color could provide at least 8 bit data in a parallel mode of transmission. If the filtering (separating) optics where precise enough, even more values could be assigned to more colors. Think of the processing sophistication and speeds that could be accomplished with all these colors simultaneously streaming through the fibres.

Just a thought.

Nov 14, 2014
Red and blue light are simply experienced simultaneously, rather than combined into a single purple wavelength.

That would be rather impossible in the first place, since there exists no such thing as purple wavelenght of light.

Purple is a specific combination of red and blue in the absence of green. It doesn't exist as a real color.

A better example would be violet, which exists beyond the actual color primaries of a television screen, so it technically cannot even produce that color, but, it can create the appearance of that wavelenght by mimicing the pattern which it would create in the eye and produce a sort of de-saturated violet.

Please sign in to add a comment. Registration is free, and takes less than a minute. Read more