Power/Performance Bits: June 15

Low-loss photonic IC; MEMS for x-ray modulation; superconductors and AI.


Low-loss photonic IC
Researchers at EPFL built a photonic integrated circuit with ultra-low loss.

The team focused on silicon nitride (Si3N4), which has orders of magnitude lower optical loss compared to silicon. It is used in low-loss applications such as narrow-linewidth lasers, photonic delay lines, and nonlinear photonics.

In applying the material to photonic ICs, they took advantage of the photonic Damascene process developed at EPFL. Using this process, the team made integrated circuits of optical losses of only 1 dB/m, a record value for any nonlinear integrated photonic material.

According to the researchers, such low loss significantly reduces the power budget for building chip-scale optical frequency combs, which are used in applications like coherent optical transceivers, low-noise microwave synthesizers, LiDAR, neuromorphic computing, and optical atomic clocks. The team used the new technology to develop meter-long waveguides on 5×5 mm2 chips and high-quality-factor microresonators. They also report high fabrication yield.

“These chip devices have already been used for parametric optical amplifiers, narrow-linewidth lasers and chip-scale frequency combs,” said Dr. Junqiu Liu, who led the fabrication at EPFL’s Center of MicroNanoTechnology (CMi). “We are also looking forward to seeing our technology being used for emerging applications such as coherent LiDAR, photonic neural networks, and quantum computing.”

MEMS for x-ray modulation
Researchers from Argonne National Laboratory developed a MEMS device to modulate x-rays. The new x-ray optics-on-a-chip device measures about 250 micrometers and weighs just 3 micrograms and performed 100 to 1,000 times faster than conventional x-ray optics, which tend to be bulky.

“Our new ultrafast optics-on-a-chip is poised to enable x-ray research and applications that could have a broad impact on understanding fast-evolving chemical, material and biological processes,” said research team leader Jin Wang from the U.S Department of Energy’s Argonne National Laboratory. “This could aid in the development of more efficient solar cells and batteries, advanced computer storage materials and devices, and more effective drugs for fighting diseases.”

The team investigated MEMS because they are used to manipulate light for high-speed communication, said Wang. “We wanted to find out if MEMS-based photonic devices can perform similar functions for x-rays as they do with visible or infrared light.”

They found that the extremely small size and weight of their MEMS-based shutter allows it to oscillate at speeds equivalent to about one million revolutions per minute (rpm). The researchers leveraged this high speed and the MEMS material’s x-ray diffractive property to create an extremely fast x-ray shutter.

The device was demonstrated using the x-ray source at Argonne’s Advanced Photon Source synchrotron. “Although we demonstrated the device in a large x-ray synchrotron facility, when fully developed, it could be used with conventional x-ray generators found in scientific labs or hospitals,” added Wang. “The same technology could also be used to develop other devices such as precise dosage delivery systems for radiation therapy or fast x-ray scanners for non-destructive diagnostics.”

The team is working to make the devices more versatile and robust so that they can be used continuously over long periods of time. They are also integrating the peripheral systems used with the chip-based MEMS devices into a deployable stand-alone instrument.

Superconductors and AI
Researchers at the National Institute of Standards and Technology (NIST) propose a way to achieve large-scale, general AI by integrating photonic components with superconducting electronics rather than semiconducting electronics.

“We argue that by operating at low temperature and using superconducting electronic circuits, single-photon detectors, and silicon light sources, we will open a path toward rich computational functionality and scalable fabrication,” said author Jeffrey Shainline of NIST. “What surprised me most was that optoelectronic integration may be much easier when working at low temperatures and using superconductors than when working at room temperatures and using semiconductors.”

They point out that the ability of superconducting photon detectors to pick up just a single photon, compared to the about 1,000 photons required for semiconducting photon detectors.

While the superconducting electronics need to be kept cold, near 4 Kelvin, and thus aren’t useful for room-temperature applications, the researchers believe advanced computing systems could make use of the concept. Ultimately, they think that it could be used in designing an AI hardware system on the scale of the human brain, or larger.

The researchers plan to explore more complex integration with other superconducting electronic circuits as well as demonstrate all the components that comprise artificial cognitive systems, including synapses and neurons.

Leave a Reply

(Note: This name will be displayed publicly)