kayla nguyen-038_edit.jpg

The EMPAD, x-ray detector plus electron microscope invented at Cornell, needed an algorithm to translate its images. Kayla Nguyen was on it.
Dave Burbank
Dave Burbank

kayla nguyen-034_edit.jpg

“The interface between the electron beam and the detector created large sets of data, but unfortunately, the innovative nature of our work meant there was no way to understand this data yet.”
Dave Burbank
Dave Burbank

kayla nguyen-043_edit.jpg

Nguyen’s algorithms analyze data to translate diffraction patterns, condensing and reconstructing them into a single-image of all the patterns.
Dave Burbank
Dave Burbank

kayla nguyen-028_edit.jpg

“With the EMPAD, art conservationists can regularly observe the specific oxidization levels of a painting in unparalleled detail and determine the ideal conditions for preserving it.”
Dave Burbank
Dave Burbank

Nyugen_edit.jpg

Nguyen won a Lemelson-MIT Student Prize, awarded only to four graduate students each year, for her data interpretation software for the EMPAD.
Beatrice Jin; Dave Burbank
Beatrice Jin; Dave Burbank

Software for an EMPAD—Making It Marketable

by Aditya Bhardwaj ’20

Growing up in Anaheim, California, Kayla Nguyen’s passion for physics stemmed from an enduring love for surfing and the technical aspects of skateboarding. “While working at a skate shop during my high school years, I would often disassemble and modify skateboards. I eventually developed a knack for tweaking all kinds of machines around me, trying to understand and improve them. I like to think that my love for science and invention today is a consequence of the hours I spent plugging away at skateboards in Anaheim.”

EMPAD, X-Ray Detector + Electron Microscope, Invented at Cornell

At Cornell, Nguyen’s award-winning research has revolved around the electron microscope pixel array detector (EMPAD), a ground-breaking device created by retrofitting an x-ray detector onto an electron microscope. As Nguyen explains, an x-ray detector previously optimized for x-rays of 10 keV energy is also compatible with the higher energy electron sources because they possess similar absorption profiles. The scattered electron signal can be collected by the EMPAD as a series of electron diffraction patterns, from which an image can be reconstructed.

Although the latest, most successful iteration of the device was engineered in 2014, its development began in 2006, when the David A. Muller and Sol M. Gruner research groups first attempted to integrate the electron microscope with a pixelated x-ray detector. The initial roadblock occurred, according to Nguyen, because the x-ray detector failed to adapt to the increased number of electrons that can be observed through an electron microscope. The increased electron counts would easily saturate the detector.

In the years that followed, the Gruner and Muller groups designed a new detector that could handle the large number of scattered electrons without saturating. “The dynamic range on the new detector,” Nguyen explains, “is 30,000 times larger than any other detector on the market, which meant it could observe a significantly higher number of electrons without saturating the detector, along with possessing the level of sensitivity to detect even a single electron.”

An Algorithm for EMPAD’s Imaging

Once the electron microscope and the detector were successfully combined, a new problem arose. The researchers had no algorithm through which to interpret the resulting diffraction pattern, and without doing so, it wasn’t possible to reconstruct the patterns into an image. “The interface between the electron beam and the detector created large sets of data, but unfortunately, the innovative nature of our work meant there was no way to understand this data yet,” Nguyen says. She was thus tasked with creating algorithms that could make sense of this information. She began analyzing the data in order to structure the diffraction patterns that emerged.

When an electron beam is projected onto a sample, it scatters through the sample and diffraction patterns are created onto the platform behind. The process is similar to light bouncing off an object. At each point where the light touches the object, it scatters into different directions, creating myriad unique diffraction patterns. Nguyen’s algorithms translate these patterns, condensing and reconstructing them into a single-image representation of all the patterns.

EMPAD, Helping to Preserve Art

The implications of the EMPAD project and Nguyen’s data interpretation software stretch across various disciplines. Art conservationists, for instance, can utilize the technology to better preserve artworks in their original state. Prominent art curators recently visited Cornell with fragments of the Norwegian artist Edvard Munch’s “Scream,” an iconic expressionist work. Nguyen and Barnaby Levin, a fellow Muller group researcher, observed the fragments through the EMPAD, attempting to determine the ideal conditions for preserving the work without having to restore it.

“Conservation has replaced restoration as the preferred method of maintaining art in the modern world; curators would like to exhibit paintings in their genuine state, without altering their characteristics. Unfortunately, both natural and artificial light oxidize a painting, and consequently, the artwork gradually loses color. With the EMPAD, art conservationists can regularly observe the specific oxidization levels of a painting in unparalleled detail and determine the ideal conditions for preserving it.”

EMPAD, Enabling Fuel-Cell Cars

In addition to art conservation, the EMPAD finds significant potential use in the automobile industry. Fuel-cell cars are being increasingly viewed as an alternative to conventional gasoline cars, but the manufacturing process for such cars remains far too expensive for large-scale commercialization. The primary reason for high manufacturing costs is the need for platinum catalysts, which transform water into hydrogen and energy for the fuel-cell motor. In addition to being an expensive substance, platinum is mostly sourced from unstable regions of the world. The supply line remains not only expensive but also unreliable.

Nguyen argues that the EMPAD can be utilized to study the atomic structure of substances in far greater detail than other methods and can thus help scientists create a less expensive alternative to platinum, which can catalyze water with the same, or possibly even greater, efficiency.

Seeing the EMPAD’s Magic

The clearest evidence of the EMPAD’s exceptional capability, according to Nguyen, can be demonstrated with a comparison of a regular picture with an EMPAD-enhanced picture, using the signal captured by light. Under normal circumstances, a person can take a selfie with the sun behind them; but due to the highly-saturated brightness of the sun, the camera is only capable of focusing on either the person or the sun. Thus, the lack of dynamic range of the detector means a person can never take a clear selfie with the sun and all its sunspots in detail in the background. An EMPAD-enhanced image, on the other hand, can capture the person as well as the sun with hitherto unmatched specificity.

Exploring Commercialization Potential, Winning the Lemelson-MIT Prize

Nguyen recently presented her work at the Rev: Ithaca Startup Works event, held by Cornell’s Engineering Commercialization Fellows Program. As part of the program, Nguyen collaborated with Cornell MBA students to simulate a commercialization package for her data interpretation software, interacting with potential buyers to explain the implications of her research. The experience, she says, was especially beneficial in terms of helping her understand how best to present her technology to businesses in the non-science marketplace.

“A large part of commercializing scientific research involves communicating the benefits of a technology to people who aren’t scientists. The Commercialization Fellows program gave me exactly what I was looking for—the confidence to translate my work into non-scientific terms and pique investors’ interests.”

“As for my colleagues, they are always willing to be an outlet for each other’s ideas, and they never hesitate to use their expertise to guide each other in the right direction.”

Nguyen’s plans to commercialize her data software package have been buoyed by recognition from the Massachusetts Institute of Technology, in the form of the prestigious Lemelson-MIT Student Prize. The prize, awarded only to four graduate students each year and accompanied by a $15,000-dollar cash award, emphasizes the groundbreaking and relevant nature of Nguyen’s project. The Cornell Center for Technology Licensing (CTL), meanwhile, has licensed the EMPAD to Thermo-Fisher Scientific, a prominent scientific conglomerate, which has sold the technology to Australia’s Monash University and Israel’s Weizmann Institute of Science, among other scientific research centers.

Muller’s appreciation for her work, according to Nguyen, has provided her with immense support throughout her time at Cornell, as has the collaborative attitude of her colleagues. “Professor Muller has constantly reiterated his desire to encourage women in science, and he has validated this belief with the high number of female researchers at the Muller lab. As for my colleagues, they are always willing to be an outlet for each other’s ideas, and they never hesitate to use their expertise to guide each other in the right direction.”

The Track That Led to Success

Nguyen completed her undergraduate degree in the College of Creative Studies at the University of California, Santa Barbara (UCSB), where she pursued a physics major, albeit in an unusual curriculum. She was not graded in her physics courses during her freshman and sophomore years, instead being assessed on a fundamental understanding of important physical concepts for her physics courses with oral final exams.

“The focus was on small class sizes, and the professors encouraged us to collaborate and experiment without the added pressure of achieving a certain grade,” she says. Following the completion of her degree at UCSB, Nguyen joined Muller’s group at Cornell’s School of Applied and Engineering Physics to pursue a PhD degree.

Following her Cornell PhD in August 2018, Nguyen will continue her research as a postdoctoral researcher at the University of Illinois, Urbana-Champagne (UIUC). She plans to pursue a research-oriented career in education after her time at UIUC.