Skip to main content
Researchers Md Sariful Sheikh, Ryan Jacobs, Jun Meng and Dane Morgan
June 13, 2024

An oxide in a haystack: From a field of 34,000, advanced computational techniques identify exactly the right material

Written By: Jason Daley

Using advanced computational techniques and molecular-level simulation, a team of materials science engineers at the University of Wisconsin-Madison has identified, synthesized and tested a new class of materials that act as fast oxygen conductors. These types of materials could eventually improve a host of sustainable, energy-related technologies. The research appears in the June 13, 2024, issue of the journal Nature Materials.

Materials that rapidly conduct oxygen are critical to many emerging and developing technologies, including solid oxide and proton ceramic fuel cells, gas sensors, electrolyzers that produce sustainable hydrogen, oxide-based memristors for digital memory, and gas separation membranes.

However, most of the commercially available and emerging oxygen conductors only work at high temperatures, around 1,400 Fahrenheit, making them costly to run and reducing their longevity.

That’s why Dane Morgan, a professor of materials science and engineering, along with postdoctoral researchers Jun Meng and Md Sariful Sheikh, and staff scientists Ryan Jacobs, Will Nachlas and Xiangguo Li, decided to investigate whether there might be novel classes of oxide materials that could operate at lower temperatures.

In most oxygen transport materials, the oxygen ions move through holes or vacant sites in the crystalline lattice of the material. But researchers have identified another class, called interstitial transport materials, in which oxygen ions are more or less jammed into the crystalline lattice where, somewhat counterintuitively, they often move about more freely. “Observations from data in the literature show that the energy required to move oxygen by an interstitial mechanism is typically much lower than by a vacancy mechanism,” says Jacobs. “So we felt there was an opportunity to try and find new materials that could exploit this mechanism. But the fundamental challenge is that, usually, oxygen doesn’t want to be jammed into a compact oxide lattice.”

The team’s goal was to find a material that was fairly easy to produce and also allowed for interstitial transport. To start, Meng used computational techniques to assess all 34,000 materials in a database of most known oxides. First, she designed easy-to-calculate, physically-motivated intuitive criteria to examine the structure and chemical properties of the oxides, finding those with lattice structures that had large enough free space to allow for oxygen interstitial formation and transport. Using molecular-level simulations, the team whittled things down by examining thermodynamic properties, the energy cost of introducing and diffusing oxygen ions in the lattice, whether a material had been previously synthesized, and how difficult it is to produce. That resulted in three candidate families of materials.

The team synthesized one illustrative composition from one of these families, LMS, a member of the perrierite/chevkinite family, and tested its properties. “We measured the oxide conductivity and found that it is very high, comparable to the best known materials, which is quite exciting,” says Sheikh, who led the experiments.

The team, however, isn’t pinning its hopes on LMS, which they say was a test case. “We found an example that has impressive performance, which demonstrated the validity of the approach,” says Morgan. “Now we can continue to evolve the screening methods and target materials with really specific properties and exceptional performance that can improve some of these energy technologies.”

Jacobs says the computational process the team used to identify LMS was extremely powerful. “Not only did it give us this particular material, it actually gave us several distinct materials families that seemed to be promising,” he says. “That suggests that there are really a lot of interesting materials for interstitial oxygen transport that haven’t been discovered yet, even in the catalogs and databases of known materials.”

In fact, Morgan says it’s unlikely any materials scientist, including himself, would have intuitively believed that LMS has these properties. This, he says illustrates the increasing power of modern, computationally-driven materials design. “We screened thousands of compounds, approaching all the known oxides in the world—and the simulations gave us a material none of us would have guessed,” he says. “And it had exactly the properties it was supposed to.”

Featured image caption: From left to right, researchers Md Sariful Sheikh, Ryan Jacobs, Jun Meng and Dane Morgan. Credit: Joel Hallberg.

Dane Morgan is the Harvey D. Spangler Professor of Engineering. Other authors include Jian Liu of the DOE National Energy Technology Laboratory, Morgantown, West Virginia.

The authors acknowledge funding from the US Department of Energy (DOE), Office of Science, Basic Energy Sciences (BES), under Award # DE-SC0020419. This work used the Extreme Science and Engineering Discovery Environment (XSEDE), which is supported by National Science Foundation Grant Number ACI-1548562.