First Light-Driven Nanomachine
A silicon nanobeam uses optical force to do mechanical work in an integrated circuit.
Since the 1980s, researchers have used lasers to stop molecular vibrations, so that the molecules can be observed in their natural environment. Now researchers at Yale University have used the same kind of nanoscale optical force to control an integrated circuit. Their device could form the basis of fast, low-power optical chips, just as transistors are the building blocks of today’s electronic circuits. The new device, a light-driven nanoresonator, could also be used as an extremely sensitive chemical detector. The work is a major landmark in uniting mechanical and optical forces at the nanoscale.
Chips that use light instead of electrons to carry data should be faster and consume less power than traditional integrated circuits. But so far even the fastest optical chips have incorporated electrical elements called modulators. These modulators encode light with data by converting the signal from light into electrons and back again. This extra step makes optical chips complex and drains power. A circuit developed by Yale researchers led by electrical-engineering professor Hong Tang incorporates a modulator that’s driven by light, not electrons.
The Yale group began its work by creating a silicon optical chip. To make the modulator, they etched a small portion of the waveguide, the thin silicon road along which the photons travel, into a 500-nanometer-wide bar. This silicon beam, which is suspended from the chip’s surface so that it can flex, has two functions. It both carries the optical signal and modulates it. Tang and his colleagues sent a light signal through the integrated circuit, then shone laser light onto the nano-optical modulator, causing it to oscillate up and down. These oscillations modulate the speed of the light traveling through the beam.
The Yale team is the first to demonstrate the existence of this optical force on an integrated circuit–and the first to exploit it to make a working device. “The light force can be put to real use,” says Tang. His group has also demonstrated that it can make arrays of hundreds of working resonators on a single chip.
Optical tweezers have been very useful for manipulating free-floating nanoscale objects in solution, but they’re very complex, requiring a high-power laser and an entire benchtop. Although it still requires input from a laser that isn’t yet integrated on the chip, the Yale setup is simpler than that required for optical tweezers.
Described in the journal Nature, the Yale circuit “represents a technical breakthrough,” says Columbia University mechanical-engineering professor James Hone. “It opens up a new way to make opto-mechanical switches that can reroute one optical signal using another.” Hone says that such devices could be the building blocks of optical circuits. Adam Cohen, a professor of chemistry, chemical biology, and physics at Harvard, agrees–as long as making these devices proves compatible with standard semiconductor processing. The traditional approach, which involves converting the optical signal into an electrical one and back again, “slows things down and is more complicated,” Cohen says.
Because the mechanical oscillation of the beam changes the way that light flows through it in a measurable way, the beams could be developed into very sensitive chemical sensors, says Hone. The Yale group has not demonstrated a chemical sensor. In theory, however, arrays of the on-chip silicon oscillators could be decorated with antibodies that bind blood proteins characteristic of diseases such as cancer. If a blood sample placed on the chip contained a small amount of the protein, it would bind to the silicon beam, changing the frequency of its oscillations–and thereby causing a measurable change in the speed of light carried through it. Other nanoscale sensors work on a similar principle, picking up differences in the flow of electrical current through oscillating silicon beams or carbon nanotubes when they bind to molecules of interest. Optical resonators might be even more sensitive, says Hone, because optical devices are “better behaved,” giving clearer signals than electrical devices do.
However, such applications are many years away. The device is still in very early development in Tang’s lab, where his group is refining its mechanical properties.
Keep up with the latest in intelligent machines at EmTech Digital.
The Countdown has begun.
March 25-26, 2019
San Francisco, CA