It's one of the many promises of artificial intelligence: The technology should dramatically speed up the search for new materials and molecules, and thus help solve some of the most pressing problems of our time. Experts hope to find chemical blueprints for better catalysts, more powerful batteries and other innovative materials. Good A team from Microsoft in collaboration with the Pacific Northwest National Laboratory (PNNL), a research facility affiliated with the US Department of Energy, said it has reached an important stage towards achieving this vision. With the help of artificial intelligence, they filtered out a previously unknown substance from 32 million possible substances and then synthesized it in the laboratory. According to the researchers, the material has great potential as a resource-efficient energy storage device. The results have not yet been independently verified.
Research into new chemicals is usually a complex, expensive and long process. It usually takes years, if not decades, to find, produce and test new compounds. For example, the lithium-ion battery widely used today took about two decades to develop. The entire process – from searching for suitable candidate materials, through to selecting, testing and producing a prototype battery – has now been shortened to nine months. “We are at the beginning of a new era of scientific discovery,” says Jason Zander, vice president of strategic mission and technology at Microsoft. Quote in press release. “Our success in finding a new battery material using artificial intelligence is just one of many examples of how our innovative approach to materials research can improve our daily lives in the future.”
Microsoft's “Azure Quantum Elements” platform relies on various artificial intelligence systems, cloud computing, high-performance computing – and in the distant future – on a quantum computer and combines them with each other. First, the team trained the AI to identify viable combinations of different chemical elements for that specific application. The algorithm then suggested 32.6 million candidates. The scientists then used another artificial intelligence system to identify all the materials that form a stable composition under natural conditions. A third AI tool filtered out molecules that could be considered battery materials based on their reactivity and ionic conductivity. About 800 items remain. All AI models used in this selection process are based on a graphical neural network. Such networks can process data that can be represented in graphs.
“Alcohol buff. Troublemaker. Introvert. Student. Social media lover. Web ninja. Bacon fan. Reader.”
More Stories
Skin rash after eating asparagus? What could be behind it?
Warning signs of Alzheimer’s disease: Researchers find new evidence
Zoonoses: Do squirrels transmit leprosy?