Uncovering Hidden Atomic Patterns in Semiconductors
Advanced microscopy reveals motifs of trace atoms in semiconductors, paving the way for new microelectronics designed atom by atom.
The Science
Semiconductors are the foundation of modern electronics. Many semiconductors are made primarily of one element with a few other atoms added to the mix. These trace atoms may appear randomly mixed. However, at very small scales, they can form subtle patterns known as short-range order (SRO). These tiny atomic neighborhoods can alter how electrons move. This change affects how the material conducts electricity or light. Detecting SRO has been nearly impossible. Until now, no technique could zoom in close enough, and with enough clarity, to examine these atomic regions. Scientists used energy-filtered, advanced electron microscopy and machine-learning-assisted analysis to visualize and measure these hidden atomic motifs for the first time.
The Impact
Understanding and controlling SRO could allow researchers to control a semiconductor’s band structure without changing its overall composition. Band structure is the property that governs how semiconductors carry charge and interact with light. This discovery opens a path toward designing materials atom-by-atom. This technique could enable faster and more efficient devices. It could also potentially lead to scientists being able to create new microelectronic and quantum materials by design. With this approach, local atomic arrangements dictate the material’s quantum behavior and performance.
Summary
Theory predicts that trace atoms in semiconductors can form subtle, repeating arrangements known as SRO. SRO produces only faint, diffuse scattering that is drowned out by the bright diffraction from the matrix. This characteristic has made detecting them nearly impossible until now. Using an advanced energy-filtered four-dimensional scanning transmission electron microscopy (4D-STEM) technique, scientists have made these hidden atomic motifs visible for the first time in a silicon–germanium–tin alloy. Energy filtering was key. By isolating only the elastically scattered electrons, the team removed background noise from inelastic scattering. That change revealed the weak signal caused by SRO. To interpret this data, the researchers combined the experimental results with machine-learning–enabled simulations trained on quantum mechanical (DFT) calculations. Machine learning made it possible to model realistic samples containing millions of atoms—far beyond what traditional DFT can handle. This analysis uncovered SRO in the form of preferred atomic triplets. The results are exciting because these local atomic patterns influence the most important property for microelectronics, the band gap. The band gap controls how electrons move and interact, which shapes a material’s electrical and optical behavior.
Contact
Andrew Minor
National Center for Electron Microscopy, Molecular Foundry, Lawrence Berkeley National Laboratory
aminor@lbl.gov
Funding
Work was supported as part of the μ-ATOMS Energy Frontier Research Center funded by the US Department of Energy (DOE), Office of Science, Basic Energy Sciences. The research used the Molecular Foundry (a DOE Office of Science User Facility), which is supported by the Office of Basic Energy Sciences. This research also used resources of the National Energy Research Scientific Computing Center, a DOE Office of Science User Facility.
Publications
Vogl, L.M., et al., Identification of short-range ordering motifs in semiconductors. Science 389, 1342 (2025). [DOI: 10.1126/science.adu0719]
Related Links
Atomic Neighborhoods in Semiconductors Provide New Avenue for Designing Microelectronics, News from Berkeley Lab
Highlight Categories
Program: BES
Performer: DOE Laboratory , Foundry , NERSC
