Harnjoo Kim was working on his doctoral research at Georgia Tech when he and his professor stumbled onto something unexpected. They found a way to make tiny machines and structures that are thousands of times smaller than the width of a human hair — and they did it 100 times faster than before, without losing accuracy.
The breakthrough happened in the lab of Sourabh Saha, an associate professor at the George W. Woodruff School of Mechanical Engineering in Atlanta. For years, engineers have dreamed of using nanoscale 3D printing to build everything from advanced computer chips to medical devices. But there was a stubborn problem: methods fast enough for real manufacturing produced sloppy results, while precise methods were simply too slow to be practical.
"Many nanoscale 3D printing methods can produce very precise structures, but they are too slow for practical production," Kim said. "On the other hand, high-speed approaches often lose accuracy because it becomes difficult to control the printing process."
Saha and Kim solved this by turning a limitation into an advantage. Light, when you try to focus it very tightly, naturally spreads out in a phenomenon called diffraction. Engineers usually fight against diffraction. But the Georgia Tech team decided to work with it instead.
"Often, the diffraction limit is considered a nuisance," Saha said. "For us, it became an enabler."
Their technique uses a device with pixels that can only be switched on or off — like pixels on a basic screen. Most engineers assumed such a device could only produce two levels of brightness. But Saha and Kim found that arranging those simple on-off pixels into carefully designed patterns could create many different levels of light intensity. They built complex patterns from many small, predictable pieces, almost like assembling a painting from deliberate brushstrokes rather than throwing one wild splash of color.
"Our approach keeps things simple by assembling the final pattern from many small, predictable pieces," Saha said. "That are more like deliberate brushstrokes than a single wild splash."
The results, published in the journal Nature Communications, showed the method increased printing speed by roughly 100 times compared to the team's earlier work while also producing finer, more accurate details. Kim described the moment the experiments confirmed what the simulations had predicted.
"Seeing those simulations translate into real experiments and producing more accurate printed structures was extremely exciting," he said. "It felt immensely rewarding to discover a new way to overcome a long-standing limitation."
The implications stretch far beyond the lab. The technology could help manufacture micro-optical components for cameras and communication systems, next-generation sensors, biomedical implants, and mechanical materials with properties not found in nature. Saha's lab is also working toward a goal of producing fuel capsules for fusion energy — tiny containers that could help power the world with clean, abundant energy.
For Kim, who now works as a postdoctoral fellow in Saha's lab, the experience proved that perseverance pays off. "It felt immensely rewarding to discover a new way to overcome a long-standing limitation in projection-based nanoscale 3D printing."