Many breakthrough materials have been found by accident. The first superconductor, for example, was discovered while researchers were looking for something else. A superconductor is a class of material that can conduct electricity with zero resistance, meaning electricity can flow without losing energy as heat.
That kind of efficiency could transform everyday technology. Much of modern life depends on moving electrical energy from point A to point B, and a lot of that energy is wasted along the way. Superconductors could dramatically reduce that waste.
If that sounds too good to be true, that's because it is. Superconductors are rare and require very specific conditions to function, including extremely low temperature and in some cases, incredibly high pressure. Nonetheless, scientists continue to search for these elusive materials.
Rather than waiting to discover new superconductors by accident, scientists from the U.S. Department of Energy's (DOE) Argonne National Laboratory and Northwestern University are working on designing and predicting new materials with unique properties - including superconductivity - by understanding and controlling how their atomic structures form.
"Finding new materials that exhibit quantum phenomena, including superconductivity, magnetism and other exotic effects is a very hot area in physics," said Mercouri Kanatzidis, a professor at Northwestern University with a joint appointment as an Argonne materials scientist. "Superconductivity is probably the biggest prize but also the most difficult to achieve by design. Almost all superconductors have been discovered by accident. Our project asks: How do you go about finding these materials-"
Luck or Design
According to Kanatzidis, there are two main paths to discovering new materials: try many experiments and hope to stumble on something interesting or develop a design strategy that makes discovery more reliable and faster.
"To do that, you need to develop what we call 'the science of synthesis,'" Kanatzidis said. "This isn't just making new materials to see what happens. It's about studying how materials come together and figuring out what makes their structures form and stay together in a certain way."
In this work, the team focused on a family of inorganic materials made from barium (Ba), antimony (Sb) and a mix of sulfur (S) and tellurium (Te). Even though the overall recipe - the ratio of elements - stayed the same for each compound, the structures did not.
The materials all fit the same general formula BaSbQ3, where Q is sulfur or tellurium, and the ratio of Ba:Sb:Q is fixed at 1:1:3. What changed is the number of sulfur and tellurium atoms and how they were arranged inside the crystal.
Starting from a well-known compound of barium, antimony and tellurium, the team set out to see what happened when they removed some tellurium atoms and replaced them with sulfur. Because sulfur and tellurium are in the same group of the periodic table, they have the same number of outer electrons. In many materials, swapping one for the other simply produces a solid solution, where the two elements mix randomly. That's not what happened here.
"The surprise was that as we added more sulfur, almost every sample turned out to be a different compound," said Argonne postdoctoral researcher Xiuquan Zhou. "Instead of just mixing together, it was forming a new compound every time. Every structure was different, but when we looked closer, we realized they were related by a mathematical relationship that put them all into the same family, called a homologous series."
Kanatzidis added, "That was remarkable. I'd never seen anything like this."
Predictable Patterns
A homologous series is a family of related compounds built from the same basic building blocks, arranged in a repeatable pattern. Each new member of the family is like the next step in a sequence. It's made by adding or changing one block in the structure - in this case, swapping a tellurium atom for a sulfur atom. Because the change follows a consistent rule, if you understand one member of the family, you can often predict what the next one will look like.
For this family of compounds, changing the balance of sulfur and tellurium didn't just make a small adjustment to one material - it produced a whole sequence of distinct crystal structures, each related to the others in an orderly way.
To confirm what they had made, the researchers relied on several sophisticated tools. At the Advanced Photon Source (APS), they used small-angle X-ray scattering at the DuPont-Northwestern-Dow Collaborative Access Team beamline and high-resolution powder X-ray diffraction at beamline 11-BM to identify and determine new crystal structures.
At the Center for Nanoscale Materials (CNM), they used scanning electron microscopy with energy-dispersive X-ray spectroscopy to verify the materials' composition. They also used transmission electron microscopy at the Northwestern University Atomic and Nanoscale Characterization Experimental Center to image the materials at the atomic level and validate the structures. The APS and CNM are both DOE Office of Science user facilities at Argonne.
Using these tools, the team demonstrated 10 distinct compounds, all with the same ratio of elements (1:1:3) but each with different structures.
"Each compound is new, so each is worth investigating for the original goal of superconductivity, quantum phenomena and other exotic effects," said Argonne postdoctoral researcher Hengdi Zhao. "On the synthesis science side, we've made real progress because this approach can be used again and again. Our results help us come up with new ideas for making other families of materials, each with their own unique compounds and structures."
Beyond Algorithms
As artificial intelligence (AI) and machine learning become increasingly popular in materials science, many research teams are turning to computer algorithms to predict new compounds. However, this discovery highlights the continued importance of human intuition and chemical understanding.
"AI gives you what is already known; it's trained on existing data," explained Kanatzidis. "All the examples so far of new materials from AI are known structures and paradigms, just new analogs or members of known families. We want to find new families. We're trying to stay ahead of AI so that if we succeed, AI can be trained on our knowledge."
The larger goal is not just to add 10 more materials to a list but to develop a clearer playbook for discovering materials with the kinds of electronic behavior that could power future technologies.
The results of this research were published in Science.
This study was funded by DOE Office of Basic Energy Sciences and Argonne's Laboratory Directed Research and Development program.
Other contributors to this work include Duck-Young Chung, Stephan Rosenkranz and Saul Lapidus from Argonne and Ziliang Wang, Patricia Meza, Christopher Wolverton, Vinayak Dravid, Denis Keane and Steven Weigand from Northwestern University.