UCLA study sets new benchmarks for 3D, atom-by-atom maps of disordered materials

FINDINGS

Researchers at the California NanoSystems Institute at UCLA published a step-by-step framework for determining the three-dimensional positions and elemental identities of atoms in amorphous materials. These solids, such as glass, lack the repeating atomic patterns seen in a crystal. The team analyzed realistically simulated electron-microscope data and tested how each step affected accuracy.

The team used algorithms to analyze rigorously simulated imaging data of nanoparticles - so small they're measured in billionths of a meter. For amorphous silica, the primary component of glass, they demonstrated 100% accuracy in mapping the three-dimensional positions of the constituent silicon and oxygen atoms, with precision about seven trillionths of a meter under favorable imaging conditions.

BACKGROUND

While 3D atomic structure determination has a history of more than a century, its application has been limited to crystal structures. Such techniques depend on averaging a pattern that is repeated trillions of times.

In contrast, the precision and accuracy required to map individual atoms in a single, non-repeating structure have been out of reach until recently. Imaging amorphous materials in 3D at the atomic level is expected to have such a widespread impact on science and engineering that this UCLA study appears back-to-back with another paper on the same topic in the journal Nature.

METHOD

The study focuses on two imaging techniques developed over the last 25 years by corresponding author Jianwei "John" Miao, a professor of physics and astronomy in the UCLA College.

One is atomic electron tomography, or AET, which takes many images from different angles as an electron beam passes through a sample, then uses computation to reconstruct a three-dimensional map of the atoms. The other is ptychography, a computational microscopy technique that records patterns of scattered electrons as a tightly focused beam scans a sample and uses algorithms to reconstruct an image without the need for a physical lens.

To validate their approach, the researchers used rigorously simulated AET and ptychographic data that closely mimic real experiments. Their algorithms had to contend with sources of error such as image noise, small variations in focus and vibrations of atoms caused by ambient temperatures. All of these factors were incorporated into electron scattering simulations based in quantum mechanics, the counterintuitive rules that govern subatomic particles. The computation also took advantage of known constraints, such as the types of atoms present and the typical distances between neighboring atoms, to refine the final 3D atomic map.

IMPACT

As a rule, algorithms typically gain power over time from improvements in programming and hardware. So computational microscopy, including AET and ptychography, is poised to provide a long-lasting boost to 3D atomic imaging. What's more, unlocking the 3D structure of amorphous materials is expected to drive both technological innovation and new insights into fundamental aspects of nature.

For example, glass has an amorphous structure and happens to be one of the most ubiquitous and useful materials on the planet. Emerging technologies for ultrathin electronics, solar cells, rewritable memory, medical devices and quantum technologies often rely on materials that also lack long-range atomic order. In the future, biologists could gain access to 3D imaging once advances make it possible to identify individual carbon and nitrogen - atoms that are essential to life, and the close neighbors of oxygen in the periodic table, the element mapped alongside silicon in this study.

AUTHORS

The study's first author is UCLA postdoctoral researcher Yuxuan Liao. Other co-authors are Haozhi Sha, Colum O'Leary and Hanfeng Zhong, all of UCLA; and Yao Yang, a doctoral alumnus of UCLA and former UCLA postdoctoral researcher who is now an assistant professor at Westlake University in Hangzhou, China.

DISCLOSURES

There are no disclosures associated with this research.

JOURNAL

The study was published in Nature.

FUNDING

The study was supported by STROBE - a National Science Foundation Science and Technology Center for which Miao serves as deputy director - and the U.S. Air Force Office of Scientific Research.