Atomic musical chairs: how tiny nanocrystals are informing the future of energy-efficient electronics
While most people, when asked about energy innovation, think about some of the “large” technologies, such as wind turbines, long transmission lines, or massive power plants, some of the most important advances in how we use energy are happening at a scale so small that millions of the “machines” involved could fit on the head of a pin.
New research led by University of Colorado Boulder professor Gordana Dukovic, working in collaboration with Sadegh Yazdi and Dmitri Talapin from the University of Chicago, reveals new insights on a high-speed game of “atomic musical chairs.” This collaboration involved two large teams working together. Researchers from IMOD and the NSF Science and Technology Center on Real-Time Functional Imaging (STROBE) employed cutting-edge microscopy techniques to directly visualize, for the first time at this scale, how atoms swap places inside tiny semiconductor nanocrystals, which is a crucial step toward understanding the composition, and ultimately the properties, of these materials.
Almost all of our electronic devices are built from semiconductors. Whether it is the screen on your smartphone, the components in your car, or the microchips in your computer, these electronics rely on semiconductors. Traditionally, these materials are “grown” through rigid and often expensive processes. Tuning the properties of a semiconductor using this approach is not straightforward. If you want a specific color of light for a display, or a specific energy absorption profile for a solar panel, you often have to start from scratch with an entirely different material.
This is where semiconductor nanocrystals offer remarkable opportunities. The specific size, shape, and composition of these tiny nanocrystals determine the physical and electronic properties of the overall material. A particularly powerful process with such nanocrystals is called cation exchange. Instead of building a new crystal from scratch, you can take an existing one and swap out its internal atomic components to change its properties.
“This is a project that we have been working on for a long time,” explains Ben Hammel, a graduate student in the Dukovic Group, and lead author on this research. “We have been looking at these materials from the Talapin Group for a long time.”
This work, just published in ACS Nano, focuses on what are called III–V nanocrystals, which are tiny, four-sided pyramids, or tetrahedrons, named for the groups of the periodic table their constituent elements come from (Group III includes elements like indium, gallium, and aluminum; Group V includes phosphorus, arsenic, and antimony). In this research, the nanocrystals are made of a mixture of indium, phosphorus, and arsenic. To exert more control over the properties of these nanocrystals, the researchers introduced gallium. Adding gallium is like tuning a guitar string: it changes the energy of the crystal, influencing how it interacts with light.
“A lot of people have developed ways to make III-V bulk semiconductors, but the real challenge is making them into nanocrystals, where you have more control over their properties, and the Talapin Group have developed a really neat molten salt process to do this,” explains Hammel. The molten salt work was published in Science in 2024.
Imagine the inside of one of these tiny crystals as a perfectly ordered lattice of “seats.” There are two types of players: anions (the phosphorus and arsenic atoms) and cations (the indium atoms). A key observation from the team was that the “house” never moves. The anions are like the floor and the chairs: they stay perfectly still, maintaining the overall crystal framework. The cations, on the other hand, are the players sitting in those chairs.
In this work, the nanocrystals were placed into a “hot bath” of molten gallium salts, essentially starting the music on the game of atomic musical chairs. Previous work had shown that the atoms exchange, but there was not a lot of evidence for how this process worked. “Understanding how this works is very important, and finding out more about the local elemental composition and how the gallium atoms move can inform how we design these systems in the future,” explains Hammel.
These nanocrystals are only 5 to 10 nanometers wide. A typical human hair is between 80,000 and 100,000 nanometers wide. These crystals are called “nano” for a reason! To observe this game of atomic musical chairs in action, the team used scanning transmission electron microscopy (STEM), an instrument that uses a focused beam of electrons to probe and image matter at the atomic scale. “Early on, there were some signs that there was heterogeneity within the particles, but it was unclear, so a big technical challenge we had to overcome was how we can actually measure the gallium moving through the nanocrystal,” said Hammel.
A key challenge they had to figure out was the sensitivity of the nanocrystals to the very tool being used to study them. The electron beam of the STEM, if used at high intensity, can damage the nanocrystals before a useful image can even be collected. To solve this, the team developed an innovative “statistical” imaging approach. Rather than blasting a single crystal with a high dose of electrons to get a sharp image, the researchers instead took many low-dose, and individually blurry, snapshots of hundreds of different crystals at different stages of the molten salt reaction. “We essentially stacked the data on top of each other,” describes Hammel. “If I can add together 10 nanocrystals, I can get 10 times the signal.”
Adding these kinds of signals together hadn’t been done before with semiconductor nanocrystals. “A lot of this came together from teamwork—I got a lot of really great suggestions from collaborators on how to collect and analyze this information. I used a suite of open source Python tools, which I was a little lost with until I met the researcher who developed them at a conference—Josh Taillon from NIST—who gave me some great suggestions and ideas,” says Hammel. Using these advanced computer algorithms, they aligned and stacked hundreds of images on top of each other. Much like a long-exposure photograph of the night sky reveals stars the naked eye cannot see, this averaged stacked image revealed a detailed map of where the gallium atoms were moving inside the nanocrystals. To the team’s knowledge, this signal-averaging approach for elemental mapping has not previously been applied to semiconductor nanocrystals.
The gallium atoms rush in to claim “seats”, but not randomly. Gallium grabs the seats near the surface first. Because of the high surface-to-volume ratio of these tiny particles, this surface exchange causes a dramatic and rapid change in overall composition: within the first 15 minutes in the molten salt bath, the outside of the nanocrystals is substantially transformed. However, as the game goes on, it gets progressively harder. The indium atoms sitting in the seats at the center of the nanocrystal are crowded in, and for a gallium atom to reach the core, an indium atom must fight its way out through an increasingly gallium-rich lattice. This sets up a compositional gradient, essentially a smooth transition from a gallium-rich exterior to an indium-rich core, that persists even after 16 hours of reaction.
This new methodology, combining STEM with advanced computational image processing, is sensitive enough to detect and map the movement of atoms through individual nanocrystals. Applying it here directly revealed that the cation exchange process (indium being replaced by gallium) creates a graded composition rather than a simple sharp boundary between materials. The team also used computer simulations (finite element analysis in COMSOL) to model this exchange as a diffusion-limited process, finding that the rate of exchange slows dramatically as more gallium enters the lattice, likely because the smaller gallium atoms cause the lattice to contract, making it progressively harder for further exchange to occur.
Importantly, the methods developed in this work are broadly applicable and could be used to determine the elemental composition of many other types of nanocrystals that have previously been difficult to study due to their sensitivity to electron beams.
The ability to observe and better understand the cation exchange process in these semiconductor nanocrystals has significant implications for the development of next-generation materials. It has been suggested that graded compositions, like those observed here, could help suppress certain energy-loss processes in semiconductor devices, potentially enabling more efficient lighting and lower-power electronics. Whether these specific nanocrystals deliver on that promise remains an open and exciting research question, but this work provides the observational foundation needed to begin answering it. Additionally, the molten salt synthesis approach that underpins this research is an active area of development as a potentially more versatile route to III-V semiconductor nanocrystals—materials that have historically been among the most challenging to synthesize with fine compositional control.
By developing new tools to better observe the game of “atomic musical chairs,” the researchers are providing the field with insights into how to engineer materials at the atomic scale and revealing that the path from one material to another is more nuanced, and more interesting, than previously understood.