Exploring the limits of integrated photonic devices Prof. Ting-Yi Gu's group

On the third floor of DuPont Hall at the University of Delaware, electrical engineers are analyzing sophisticated centimeter-sized computer chips on a large optical table surrounded by oscilloscopes, lenses and lasers. The researchers are busy collecting data on how these chips convert light waves into electrical signals, with the goal of figuring out how to make the next batch of chips they make faster, more energy efficient or with higher computing power.

This is the lab of Ting-Yi Gu, an assistant professor in the Department of Electrical and Computer Engineering, where researchers are pushing the limits of the field of integrated photonic devices. Using a high-risk, high-reward research strategy, Ting-Yi Gu's group has made progress in developing new chip designs and applying unique materials to a wide range of optical communication, sensing and computing applications [1].

Prof. Ting-Yi Gu

Simply put, photonic devices are devices that can generate, control, or detect light, and photonic integrated circuits can use light to perform more complex functions such as data analysis. Ting-Yi Gu started working in this field when she was a graduate student, focusing on improving photonic integrated circuits through fundamental research, with a focus on developing new chip designs and investigating how to incorporate materials from other applications into photonic devices.

Ting-Yi Gu says of her team's research strategy, "It's unlikely for me and my students that we'll read a paper and revise something to show a slightly better edge. Instead, we try to find something more revolutionary by fundamentally changing the way we do our research. It's a riskier approach, but it's more fun to explore that approach than trying to repeat what someone else has already done or make some incremental progress."

Two examples of how Ting-Yi Gu's research strategy has led to advances in photonics can be found in two papers from her team in early 2022, one published in Nature Communications and the other in Advanced Materials.

01A milestone for photonic chips

In 2019, Tingyi Gu and her graduate student Zi Wang developed an on-chip conversion optical design principle for robust wavefront control on an integrated photonic platform [2], which can be used for complex processes related to other fields such as quantum optics.

Now, the latest Nature Communications paper [3] from Ting-Yi Gu's group shows how advanced computing capabilities can be integrated directly onto these photonic chips. "In 2019, our devices have very simple components, such as the Fourier transform. Now, the integrated metasystem has nearly 1,000 preprogrammed elements that can handle uncertainties across the spectral domain, which is a milestone for integrated photonic processors versus electronic processors," said Ting-Yi Gu.

Zi Wang, now a postdoctoral fellow at the National Institute of Standards and Technology, said extending their original design while making it compatible with the fabrication process was the most challenging part of this recent paper. "The structure was designed using a gradient back-propagation method, which consumed a lot of time and computational resources in our original design. But I found that our structure has a special symmetry, and by using the symmetry in the mathematical calculations, the calculations became much easier."

Members of Ting-Yi Gu's lab

As a result of this insight, the researchers found they could use light diffraction to perform complex calculations and data analysis. Ting-Yi Gu explains, "Since each programmable element is much smaller than a conventional chip, more elements can fit on the same chip area."

The paper is an example of how a new design approach can help researchers use existing fabrication methods to create chips that are more powerful than current state-of-the-art technology. "Integrated photonic circuits have much more potential than just the same way they've been used and studied for decades, and even now there are limits to the circuits we can break," she said.

02Making new optical memories

A second paper, published in the journal Advanced Materials [4], shows how Gu's lab took inspiration from materials used in other applications to evaluate whether they could be used in photonic memories, which rely on light rather than magnetism to store information.

These platforms for rewritable memory storage, called optical memristor devices, have the potential to reduce overall energy consumption, but currently rely on slow processes associated with changes in the phase or physical state of the material (most commonly solids, liquids, and gases). Phase shifting is how optical devices are stored, but here the phase shifting process requires a transition between an amorphous phase (a phase without much structure, like a pile of sand grains) and a crystalline phase (a highly structured phase, like a close-up shot of a snowflake).

Fabricating a phase shifter for photonic integrated circuits that is both compact and controllable remains a challenge because the materials currently available for optical devices can only change phase very slowly at very high temperatures.

In this paper, the group investigated indium selenide (In2Se3), a material commonly used in electronic devices but not yet widely used in optical applications, to see if they could create optical storage by switching between different crystalline phases rather than between crystalline and amorphous phases.

In this study, first author Tiantian Li, a former postdoctoral fellow at the University of Delaware and now an associate professor at Xi'an University of Posts and Telecommunications, first discovered that the phase transition mechanism of indium selenide differs from initial theories based on simulation results. The researchers then used these theoretical insights to phase shift between different crystalline states to create optical memories using short nanosecond light pulses.

Commenting on the impact of this work, Dr. Tiantian Li said, "Optical phase change materials have attracted a lot of interest due to their promising applications in optical computing. The high power consumption of phase change materials affects the computational speed of neural networks, and our material is expected to break this bottleneck."

Photonic chips in Ting-Yi Gu's lab

In addition to their applications in photonics, the two papers demonstrate the importance of creativity and the unique sources of inspiration for this field. "We try to use other resources and combine knowledge from different fields," said Tingyi Gu. "In the Nature Communications paper, we were inspired by people who do image classification for machine learning, which we brought to our integrated photonic platform, and in the Advanced Materials paper, we were inspired by chemists who study phase change mechanisms."

03The future of photonics

Researchers in Ting-Yi Gu's lab are focused on developing new chip designs and investigating how materials from other applications can be integrated into photonic devices. The work has three phases: simulation, in which different chip designs are evaluated; fabrication, in which the chips are fabricated at the University of Delaware's nanofabrication facility; and testing, in which the chips are evaluated to see how their performance compares to that predicted by the simulations.

For Yahui Xiao, a doctoral student working on photonic crystals, doing this type of research requires knowledge from fundamental physics to fabrication, and it provided a rewarding graduate school experience, especially since she is looking forward to doing this "hybrid" research in optical engineering and nanophotonics.

I want to gain insight into current technologies by understanding the underlying physics," says Yahui Xiao. At the University of Delaware, we have nanofabrication facilities, and these are fabrication skills that we can use as we move toward industry, because we can do the whole fabrication process."

Dun Mao, a doctoral student working on the indium selenide project, said that although this area of research is challenging, it is encouraging when they are able to make breakthroughs and get good results. "The most exciting part is when we observe some interesting phenomena from the experiments that can make the devices faster or more efficient," he said.

Ting-Yi Gu added that while there are many unanswered research questions in photonics that their team can address, the work in her lab is always driven by the interest and passion of her students. "We try to take a higher-risk approach in the lab, sometimes good, sometimes not, but I think the students learn a lot."

Both Mao and Xiao say that Professor Ting-Yi Gu's support has been instrumental in their current success at the graduate level, and Xiao adds that having Professor Ting-Yi Gu as a female mentor has been an added incentive for her in a field that is typically dominated by men. "I can learn a lot from her, and she has been very successful in this field."

Reference links：

[1]https://www.udel.edu/udaily/2022/december/tingyi-gu-electrical-computer-engineering-photonics-optics/

[2]https://www.nature.com/articles/s41467-019-11578-y

[3]https://www.nature.com/articles/s41467-022-29856-7

[4]https://onlinelibrary.wiley.com/doi/abs/10.1002/adma.202108261