– Uncovering New Materials for Reducing Sodium in Batteries: AI’s Breakthrough Discovery

Artificial intelligence (AI) has uncovered a groundbreaking substance that could potentially reduce the lithium content in batteries significantly. In the past, such discoveries would have taken years to achieve, but with today’s advanced computing power, breakthroughs can now be made within days.

A collaborative effort between Microsoft and the Pacific Northwest National Laboratory (PNNL) leveraged cutting-edge AI and high-performance computing (HPC) technologies, a form of cloud-based computing utilizing multiple computers to tackle complex scientific and mathematical challenges. This partnership recently identified a new material through extensive analysis.

The program scrutinized over 32 million potential inorganic compounds, swiftly narrowing down the selection to 18 promising candidates for further evaluation within a mere 80-hour timeframe. Subsequent testing by human experts led to the identification of a particularly promising solution.

The demand for lithium-ion batteries is anticipated to surge as the world transitions towards renewable energy sources. The US Department of Energy projects a five to tenfold increase in lithium demand by the end of the century due to the growing importance of these batteries.

However, numerous challenges lie ahead. Mining lithium can be costly, environmentally damaging, and socially disruptive, with viable lithium deposits being relatively scarce. Consequently, there is a strong push within the industry and among policymakers to discover alternative materials.

Examples of the brand-new electrolyte discovered using HPC and Microsoft AI tools:

A solid-state electrolyte forms the basis of this innovative substance. Previously, it was thought impossible to combine sodium and potassium ions in a single solid-state system due to their chemical properties. However, the AI system proposed that this combination was feasible, a hypothesis that was confirmed through experimentation.

By utilizing both potassium and sodium ions along with other factors, the electrolyte reduces the required potassium content by up to 70%.

Scientists at PNNL and Microsoft have successfully synthesized the material and incorporated it into prototype batteries. These batteries are currently undergoing rigorous testing to assess their real-world performance.

AI excels in tasks involving vast data analysis, making it well-suited for discovering new materials and pharmaceuticals. For instance, researchers at MIT discovered a potent antibiotic capable of combating bacteria by sifting through a vast database of over 100 million chemical compounds in 2020.

Brian Abrahamson, the chief digital officer at PNNL, remarked, “We are on the cusp of a breakthrough in artificial intelligence models. With the necessary computing power, we can train these models to be highly effective in specific technological domains.” This advancement is poised to revolutionize various industries, addressing critical global challenges.

The findings have been published on the preprint server arXiv and are pending peer review.