Argonne scientists create chatty X-ray microscope that zooms in where you tell it to
Researchers at Argonne National Laboratory have tested an X-ray microscope that responds to natural-language requests, directing its beam to examine areas of interest. The system matters because it could make nanoscale X-ray research easier to use and speed up analysis, while helping researchers identify promising areas for further study.
The microscope uses Argonne’s 26-ID hard X-ray nanoprobe beamline and was demonstrated on an integrated circuit, locating the interface between two regions. Its AI can reconstruct images as data arrives, rather than taking hours or days, and can analyse results and inspect additional areas it flags. The work is part of the SYNAPS-I project with Lawrence Berkeley National Laboratory and is intended to support research across fields, not only microelectronics.
- Argonne tested a microscope researchers can direct with natural-language requests.
- AI reconstructs X-ray images in real time as data arrives.
- The system can flag and investigate other areas of interest.
New here? Start with this
X-ray microscopes are scientific instruments that allow researchers to see extremely tiny structures far too small to view with ordinary light. They are important tools for studying materials across many fields, from computer chips to biological samples, but they have traditionally been difficult to operate.
Operating these microscopes has been a specialised task requiring researchers to manually direct the X-ray beam to areas of interest, examine the results, and decide where to look next. The process can be time-consuming and demands considerable technical expertise.
Scientists at Argonne National Laboratory have now built a version of this microscope that responds to everyday language instructions. Instead of requiring researchers to manually direct the beam and wait days to process data, this system automatically steers the microscope where asked, analyses results in real time, and can suggest new areas to examine. This advance could make X-ray microscopy more widely usable and accelerate research across many different fields.