As the volume of data exchanged around the world grows, so does the demand for more efficient and smaller network components.
These components include so-called modulators that convert information from an electrical form into optical signals. Modulators are really nothing more than fast electrical switches that turn a laser signal on or off at the frequency of the incoming electrical signals.
Data centers house thousands of these modulators. The problem is they’re too large, measuring a few centimeters across.
Six months ago, a working group led by Jürg Leuthold, a professor of photonics and communications at ETH Zurich, proved the technology could be made smaller and more energy efficient. As part of that work, the researchers presented a micromodulator measuring just 10 micrometers across—or 10,000 times smaller than modulators in commercial use.
Leuthold and his colleagues have now taken this to the next level by developing the world’s smallest optical modulator. And this is probably as small as it can get: The component operates at the level of individual atoms.
The footprint was further reduced by a factor of 1,000, if you include the switch together with the light guides. However, the switch itself is even smaller, with a size measured on the atomic scale. The team’s latest development was recently presented in the journal Nano Letters.
In fact, the modulator is significantly smaller than the wavelength of light used in the system. In telecommunications, optical signals are transmitted using laser light with a wavelength of 1.55 micrometers. Normally, an optical device can not be smaller than the wavelength it should process.
“Until recently, even I thought it was impossible for us to undercut this limit,” says Leuthold.
But senior scientist Alexandros Emboras proved the laws of optics wrong by successfully reconfiguring the construction of a modulator. This construction made it possible to penetrate the order of magnitude of individual atoms, even though the researchers were using light with a “standard wavelength.”
How the modulator works
Emboras’s modulator consists of two tiny pads, one made of silver and the other of platinum, on top of an optical waveguide made of silicon. The two pads are arranged alongside each other at a distance of just a few nanometers, with a small bulge on the silver pad protruding into the gap and almost touching the platinum pad.
Light entering from an optical fiber is guided to the entrance of the gap by the optical waveguide. Above the metallic surface, the light turns into a surface plasmon.
A plasmon occurs when light transfers energy to electrons in the outermost atomic layer of the metal surface, causing the electrons to oscillate at the frequency of the incident light. These electron oscillations have a far smaller diameter than the ray of light itself. This allows them to enter the gap and pass through the bottleneck.
On the other side of the gap, the electron oscillations can be converted back into optical signals.
If a voltage is applied to the silver pad, a single silver atom or, at most, a few silver atoms move towards the tip of the point and position themselves at the end of it. This creates a short circuit between the silver and platinum pads, so that electrical current flows between them.
“We have been looking for a solution like this for a long time.”
This closes the loophole for the plasmon; the switch flips and the state changes from “on” to “off” or vice versa. As soon as the voltage falls below a certain threshold again, a silver atom moves back. The gap opens, the plasmon flows, and the switch is “on” again. This process can be repeated millions of times.
Professor Mathieu Luisier, who participated in this study, simulated the system using a high-performance computer. This allowed him to confirm that the short circuit at the tip of the silver point is brought about by a single atom.
As the plasmon has no other options than to pass through the bottleneck either completely or not at all, this produces a truly digital signal—a one or a zero.
“This allows us to create a digital switch, as with a transistor. We have been looking for a solution like this for a long time,” says Leuthold.
Not ready for large-scale production
Although it has the advantage of operating at room temperature, unlike other devices that work using quantum effects at this order of magnitude, the new modulator is relatively slow. Right now it only works for switching frequencies in the megahertz range or below. The researchers want to fine-tune it for frequencies in the gigahertz to terahertz range.
The researchers also want to further improve the lithography method, which was redeveloped by Emboras to build the parts, so that components like this can be produced reliably in future. At present, fabrication is only successful in one out of every six attempts. Nevertheless, the researchers consider this a success, as lithography processes on the atomic scale remain uncharted territory.
Source: ETH Zurich