When Rose Cersonsky, the Conway Assistant Professor in chemical and biological engineering at the University of Wisconsin-Madison, was a graduate student at the University of Michigan, she had a “Eureka!” moment that would define the next decade of her research.
She was studying the design of 3D photonic nanocrystals, tiny structures that manipulate visible light. These crystals selectively let some wavelengths through while blocking or limiting others in a phenomenon known as a photonic band gap. The structures are what give chameleons their otherworldly color-shifting abilities and morpho butterflies their vibrant, iridescent blue hues.
They are also important in industry and are used, for example, to sort different petrochemicals, and play a role in optical communications and computing, lasers and LEDs, and biological sensing.
If researchers could better understand and then design various types of 3D nanocrystals on demand, their knowledge could advance all these applications and improve quantum computing and sensing, heat management and solar power and enable new types of energy harvesting and non-degrading color.
That’s why, during her PhD, Cersonsky focused on self-assembly of colloidal diamond. Researchers long targeted colloidal diamond as the most promising road for photonic crystals. However, in one of her simulations, Cersonsky noted that a phase transition in the diamond structure didn’t destroy its band gap as predicted in the literature.
While it was a small technical discrepancy, it was a huge eye-opener for Cersonsky. “I found a system that broke that diamond structure, but kept its optical properties, and that was really weird,” she says. “That discovery made it clear how much we still didn’t understand about which 3D photonic crystals are worth targeting.” This work was ultimately published in Physical Review Materials in 2018.
That revelation set Cersonsky on a fascinating, albeit frustrating, quest to understand the origins of these photonic properties and to develop rational design principles beyond what is known about the diamond structure.
Now, after almost nine years of work, Cersonsky and graduate students Sasawat Nayak and Seungmin Lee have collected what she’s learned so far in a paper in the March 2026 issue of the journal ACS Omega.
As previously published in Nature Communications in 2021, Cersonsky cataloged as many optical nanocrystal structures as she could find, calculating their photonic band gaps. She screened thousands of 3D structures found in various databases, running more than 150,000 calculations to determine their band gaps and eventually identifying 351 unique nanostructures that fit the parameters.
The survey, she hoped, would provide enough data to help her tease out some universal design paradigms, like why certain structures produced certain band gaps. That, however, proved more difficult than anticipated. As the work advanced, Cersonsky realized she would need more powerful tools to analyze the structures.
“I was like, ‘I don’t know what I’m doing. I need to learn,’” she remembers. “It’s the reason I got a postdoctoral position in machine learning. It’s what caused me to be steeped in that literature and community for such a long time.”
In fact, those machine learning tools are now at the core of Cersonsky’s research, which involves molecular modeling and simulation of colloids, soft matter and nanomaterials.
While that foray into machine learning opened up Cersonsky’s future research pathway, it did not help as much with her primary goal: cracking the photonic crystal code. As a faculty member at UW-Madison, she and her students tried again and again to find tools with the right physics and mathematics to analyze these structures. “The tools that I’ve gotten used to using, like crystal graphs, correlation functions, even just Cartesian coordinates, were really not suited for this,” she says. “Eventually, we all got very frustrated and annoyed; we just didn’t know what to do.”
Instead of abandoning the project, Cersonksy decided to turn it on its head. Up to that point, her approach was looking at the structures and trying to generalize how they might affect light. Instead, she decided to look at the end results and see what they could reveal about the structures. She and her students began analyzing the photonic densities of state, a histogram of frequencies these structures are able to transmit, to see what they could reveal about the structures.
The team compared histograms that were very similar, even though the structures were very different, trying to find commonalities between them. The researchers also looked at crystals that were structurally similar but produced very different histograms, trying to identify what created the differences among them.
Laid out in the ACS Omega paper, their results begin to sketch some of the design principles behind the crystals, including what crystalline features produce the largest band gaps, and how the composition, symmetry and geometry of the structures influence the size of their band gaps.
While this work gives shape to her research thus far, Cersonsky says she is not close to being done. “It’s a story where I’m not fully satisfied with the answer yet. It’s also one of the bigger questions in this area and one that has haunted a lot of people for a while now. And I’m thinking about it a lot differently than many others.”
The key to moving forward, she believes, is finding the right computational tools to understand the structures. “I’ve got suspicions about what may work,” she says. “But they’re tools from communities that know them far better than me. So I’m talking to people about that. I have an inbox of topological mathematicians I’m trying to reach.”
Continuing the work, says Cersonsky, isn’t just about scratching an intellectual itch; rather, understanding how to design these structures and move beyond the current focus on colloidal diamond could have a material impact on the world.
“There’s an incredibly potent space of applications that has been done a huge disservice by people being way too narrowly focused in what they’re trying to synthesize,” she says. “I think that there can be a much bigger impact if we stop thinking about what we’re told to make and start thinking about what’s best for us to make.”
Featured image of Rose Cersonsky by Joel Hallberg