Microsoft discovers material that can replace about 70% of lithium in batteries in just a few days, run with simulation and AI model using Azure Quantum Elements
Discoveries in weeks, not years: How AI and high-performance computing are speeding up scientific discovery - Source
https://news.microsoft.com/source/features/sustainability/how-ai-and-hpc-are-speeding-up-scientific-discovery/
Unlocking a new era for scientific discovery with AI: How Microsoft's AI screened over 32 million candidates to find a better battery - Microsoft Azure Quantum Blog
https://cloudblogs.microsoft.com/quantum/2024/01/09/unlocking-a-new-era-for-scientific-discovery-with-ai-how-microsofts-ai-screened-over-32-million- candidates-to-find-a-better-battery/
Microsoft Azure Quantum and PNNL join forces to accelerate scientific discovery - YouTube
How Microsoft found a potential new battery material using AI - The Verge
https://www.theverge.com/24027031/microsoft-new-solid-state-battery-material-ai
Microsoft finds new material to replace lithium in batteries using AI | Windows Central
https://www.windowscentral.com/microsoft/microsoft-leverages-ai-to-identify-a-new-material-that-can-potentially-reduce-the-use-of-lithium-in-batteries-by- 70
Lithium-ion batteries, which are widely used at the time of writing, are liquid electrolytes, so while they are excellent at conducting energy, they can burst and cause a fire. In addition, with the increasing demand for lithium-ion batteries, various manufacturers are building battery factories, but it has been pointed out that the production of lithium-ion batteries requires large amounts of water and energy, which has a negative impact on the environment. .
Furthermore, according to the US Department of Energy, demand for lithium-ion batteries could increase up to 10 times by 2030. For this reason, scientists are working to develop next-generation batteries that use less lithium.
In 2023, Microsoft provided the PNNL research team with Azure Quantum Elements (AQE), a platform that integrates high-performance computing (HPC), quantum computing technology, and AI.
Microsoft declares to build a quantum supercomputer within 10 years - GIGAZINE
The PNNL research team asked AQE, 'Which battery material uses less lithium?' Then, AQE immediately proposed 32 million different candidates. The research team then selected approximately 500,000 materials from among the proposed candidates that had the potential to be stably used.
The researchers also estimated how well each material conducts energy and simulated how atoms and molecules move within each material. We also estimated how practical each candidate would be, including cost and availability.
As a result of selection by the research team, they were finally able to narrow down the battery material candidates to 23. It took only 80 hours to narrow down the 32 million candidates to 23. 'It would have probably taken a human over 20 years to manually examine 32 million pieces of material,' said Vijay Murugesan, staff scientist and materials science group leader at PNNL. 'We couldn't have done it without AQE.'
From the final 23 candidates, the research team synthesized one candidate with a high possibility of being used as a battery and conducted a test. Using the batteries manufactured at that time, we succeeded in powering light bulbs and clocks.
The newly manufactured battery uses sodium in addition to conventional lithium. According to Microsoft, the newly discovered material can reduce the lithium content by up to 70% compared to conventional lithium-ion batteries. Additionally, unlike lithium-ion batteries, which use a liquid electrolyte, they are said to be safer because they use a solid electrolyte.
On the other hand, solid electrolyte batteries have inferior energy transfer performance compared to liquid electrolyte batteries, and the newly discovered material has lower energy transfer performance than the research team originally predicted, so research will continue. It will be tested and fine-tuned by the team. Therefore, it is still some time before batteries using the newly discovered material will appear on the market.
Still, Krista Svoir, leader of the Microsoft Quantum - Redmond (QuArC) group at Microsoft Research, says, ``By using AI and AQE, we can compress chemical materials research that previously required 250 years into just 20 years.'' It is possible.”
Related Posts:
