AI discovered a family of solid-state materials that conduct lithium ions. These electrolytes will lead to more efficient and safer batteries for electric vehicles and other devices.
British scientists have developed an artificial intelligence tool that has helped uncover four previously unknown materials.
They published their research in the journal Nature Communications.
Researchers at the University of Liverpool have combined machine learning technology with already accumulated human knowledge to accelerate the study of new combinations of chemical elements. Thus, it was possible to open a new family of solid-state materials that conduct lithium ions. Such electrolytes will lead to more efficient and safer batteries for electric vehicles and other devices.
The new tool explores the relationships between already known materials at a rate and in volumes that humans cannot reach. Artificial intelligence estimates the likelihood of mixing elements successfully and makes ratings of the most promising combinations. Using the findings of a computer, scientists conduct targeted research trying to obtain new materials.
Professor Mat Rosseinsky drew attention to the fact that previously it was considered the most effective to create new materials by analogy with existing ones. As a result, scientists received substances that are very similar to each other in many parameters. Therefore, all modern electronics run on batteries from materials developed back in the 1980s.
“We need new tools that reduce the time and effort required to discover truly new materials. We have tried to create such a tool that combines artificial intelligence skills and human knowledge,” said Rosseinsky.
Other materials are already in development. They can bring great benefits to humanity, stimulating new technologies to solve global problems, discovering new scientific phenomena and concepts. Researchers will have to study the proposals of AI and start developing substances with specific functions, for example, more efficient absorption of solar energy.