Welcome Your New Lab Mate: Artificial Intelligence

How ATOMIC Works

Wang’s group studies two‑dimensional (2D) materials, crystals only one or a few atoms thick that are promising candidates for next-generation semiconductors, sensors and quantum devices. Their exceptional electrical properties and flexibility make them ideal for electronics, but fabrication defects can compromise these advantages. Determining how the layers stack and whether they contain microscopic defects requires laborious work and years of training.

“To characterize these materials, you usually need someone who understands every nuance of the microscope images,” Wang said. “It takes graduate students months to years of high-level science classes and experience to get to that point.”

To speed up the process, Wang’s team linked an off‑the‑shelf optical microscope to ChatGPT, allowing the model to handle basic operations like moving the sample, focusing the image and adjusting light levels. Layered on top was SAM, an open‑source vision model designed to identify discrete objects, which in the case of materials samples would include regions containing defects and pure areas.

Together, the two AIs formed a powerful tool in the lab, a kind of virtual lab mate that could see, analyze and act on its own.

Turning general-purpose AI into a reliable scientific partner, however, required significant customization from the Wang lab. SAM could recognize regions within the microscopic images, yet it struggled with overlapping layers, a common issue in materials research. To overcome that, they added a topological correction algorithm to refine those regions, isolating single-layer areas from multilayer stacks.

Finally, the team asked the system to sort the isolated regions by their optical characteristics, which ChatGPT could do autonomously.

What It Means for Researchers

For Wang, his excitement about the innovation isn’t just about speed. It’s also about teaching his students to use the technologies at their disposal to be modern-day researchers.

“In the last year, AI has advanced a lot and Dr. Wang said if we do not embrace this era and make use of these AI tools, they may replace us,” Yang said. “We tested the ATOMIC system on many samples and different conditions, and it’s quite robust.”

Wang sees potential applications across disciplines, from chemistry to biology, where tedious optical analysis often slows progress. Simplifying those workflows could open advanced research to students, industry engineers or anyone with curiosity and a microscope.

At the same time, Wang stresses the importance of keeping humans in the loop. Foundation models can behave unpredictably, sometimes generating different results for identical prompts. His group tested thousands of repetitions to assess robustness and found that while minor variations occur, overall accuracy remains high.

“The goal isn’t to replace expertise; it’s to amplify it,” Wang said. “We still need humans to interpret what the AI finds and decide what it means. But once you have a partner that can complete weeks of analysis in mere seconds, the possibilities for new discoveries are exponential.”

The work is led by Duke in collaboration with Massachusetts Institute of Technology, National University of Singapore, University of California, Berkeley, Australian National University and Google DeepMind. Wang and Zhichu Ren are corresponding authors, and Yang and Ruoyan Avery Yin are the co-first authors. Wang and Yang acknowledge the GCP Credit Award from the Google PaliGemma Academic Program.

CITATION: Yang, J., Yin, R. A., Jiang, C., Hu, Y., Zhu, X., Hu, X., Kumar, S., Holmes, S. K., Wang, X., Zhai, X., Rong, K., Zhu, Y., Zhang, T., Yin, Z., Cao, Y., Tang, H., Franklin, A. D., Kong, J., Gong, N. Z., Ren, Z., & Wang, H. (2025). Zero‑shot autonomous microscopy for scalable and intelligent characterization of 2D materials. ACS Nano, 19(40), 35493–35502. https://doi.org/10.1021/acsnano.5c09057