Scientists in England have developed a new computer algorithm to aid the search for new materials. The system helped researchers create two new promising materials in the lab.
Machine learning allowed the model to acquire an understanding of the structural properties of different materials. The algorithm uses its knowledge to predict new combinations of atoms to form materials with valuable physical and structural properties.
With the endless possible atomic combinations, materials scientists can be overwhelmed. Computer models can help researchers narrow their focus.
"Understanding which atoms will combine to form new materials from the vast space of possible candidates is one of the grand scientific challenges, and solving it will open up exciting scientific opportunities that could lead to important properties," Matt Rosseinsky, a professor of chemistry, said in a news release.
The model works by identifying only chemically stable combinations that can be reliably synthesized in the lab. Researchers hope the model can help material scientists develop new materials for energy generation and storage.
Rosseinsky and his colleagues detailed their work in the journal Nature.
"The key step in this research was the ability to generate large numbers of truly representative structures that could be used to assess which element combinations were stable," Rosseinsky said, "which greatly narrowed the space that had to be explored experimentally — like having a map with someone's address, rather than knowing they live in London somewhere."
New catalytic converter composite reduces rare earth element usage
Automobiles are facing increasingly strict emissions regulations in an effort to reduce the amount of harmful air pollutants that are released into the environment. In Japan, for example, the current emissions standards for NOx and nonmethane hydrocarbons are less than 0.05 g/km. Currently, one method of reducing harmful emissions is with a high-performance, three-way catalytic (TWC) converter. … read more
