Physics Ph.D. Student Develops AI Tools to Tackle Industry ‘Bottleneck’

Physics Ph.D. Student Develops AI Tools to Tackle Industry 'Bottleneck'

Nonuniformities can be a big deal on a manufacturing line.

But they can be tough to identify on materials that are measured on a nanometer scale, as Alex Bordovalos came to appreciate during an internship with a solar cell manufacturer that was integrated into his master's program at The University of Toledo.

"It's not like you can take a ruler and see it," Bordovalos said. "You need specialized tools."

Those specialized tools often align with a technique known as spectroscopic ellipsometry, which physicists use to measure thickness and other optical properties of almost impossibly thin materials by detecting changes to the polarization state of light. As Bordovalos now works toward a doctoral degree in physics at UToledo, he's advancing this technique by developing artificial intelligence technology to automate its most time-consuming and labor-intensive element - namely, interpreting raw data into measurements that actually mean something.

"Alex is doing great work that addresses a major bottleneck in industry and research," said Dr. Nik Podraza, who advises Bordovalos as a professor of physics and interim dean of the College of Natural Sciences and Mathematics. "The tool he's developing promises huge decreases in analysis time in measurements of materials that may be many times thinner than a human hair. This has significant potential for manufacturers in the semiconductor and energy sectors."

Bordovalos is working toward his third degree at his hometown University, crediting positive experiences at UToledo with his decision to remain a Rocket as he continues toward his Ph.D. He earned a bachelor's degree in physics in 2020, then continued into the Professional Science Master's in Photovoltaics Program, which is designed to prepare students for successful careers in the photovoltaics industry by integrating laboratory skills, business management practices and real-life work experience through an industry internship. He earned that graduate degree in 2022.

The internship proved especially impactful for Bordovalos, with his firsthand experiences in solar cell production impressing on him the potential for a more readily accessible technique to measure the ultra-thin components of a variety of devices.

These devices might include cadmium telluride solar cells, one of the thin-film photovoltaic technologies that are a specialty of researchers at UToledo's Wright Center for Photovoltaics Innovation and Commercialization. Thin-film photovoltaics research contributes to a broader strength in materials science at UToledo, where researchers routinely publish highly cited academic articles that support a global distinction in this academic discipline as recognized by U.S. News & World Report.

Bordovalos also tapped into a long-standing interest in computer science in his quest to develop a more readily accessible approach to spectroscopic ellipsometry, which he believes can support both the research and development of devices with ultra-thin components as well as quality control on the production lines that turn them out. He began to develop an AI tool that could perform the technique more efficiently than physicists who, in order to account for irregularities in the materials they measure, must continually adjust the mathematical models that they use to make sense of their raw data.

The process has required him to train the tool on a lot of simulated data.

"I'm working with half a million simulations," Bordovalos said. "The simulations cover all these different possibilities. For example, I'll start with the thickness that I expect, and I might vary it by about 50% in either direction."

Bordovalos has shared his preliminary research in conferences and journals, including a first-author article last year in the peer-reviewed Journal of Applied Physics and presentations at the IEEE Photovoltaics Specialists Conference and the AVS International Symposium and Exhibition. The Spectroscopic Ellipsometry Technical Group recognized him with the J.A. Woollam Outstanding Student Poster Award at the latter in 2024.

As he now nears the end of his doctoral program and begins to eye job opportunities in both academia and industry, he's ready to begin real-world testing of his model using cadmium telluride solar cells that are, conveniently, readily available at UToledo.

"That's the last hurdle," Bordovalos said. "I need to see if the model can make the jump handling the simulated stuff very well to handling the real stuff very well."