A project of such scale that in 2023 & 2024 Researchers from Pacific Northwest National Laboratory (PNNL) undertook a massive endeavor at this scale. With this historic collaboration, we intend to radically transform how battery materials are discovered. They are screening over 32 million possible battery materials. Their ambition is to identify alternatives that are safer, cheaper and more environmentally friendly than lithium for future energy storage applications. They utilize generative design and advanced artificial intelligence (AI) models that can accelerate and optimize the material selection process. This process often involves a long series of laboratory assays.
The combined endeavor is distinguished not just by its size, but by the cutting edge technology powering it. Over the course of just one week, we were able to filter that initial pool of 32 million candidates into 500,000 stable materials. This example illustrates just how rapidly AI can sift through enormous datasets. This remarkable achievement prompted additional tinkering and editing. Ultimately, we produced a shortlist of 800 of the most promising candidates for more in-depth analysis.
The Role of Advanced AI Models
It is here that advanced AI models are critical to the evaluation process. Researchers use machine learning algorithms to rapidly predict key findings for systems that are similar. This novel approach can be done for orders of magnitude less than traditional computing approaches. This new capacity provides them the ability to review millions of records. They target “first-time right” candidates, making sure that only the best risk/reward molecules get progressed to the lab for synthesis and testing.
The incorporation of AI does more than speed up that discovery process, it increases precision as well. Traditional approaches have a tough time assessing such an extensive suite of new potential materials. Thanks to AI’s incredible predictive powers, researchers can focus their efforts on only the most promising options.
“AI can enable rapid predictions for similar systems at a fraction of the cost of classical computing,” stated a PNNL researcher familiar with the project.
Quantum Computing’s Impact on Material Simulations
Although AI radically speeds up the simulation-based evaluation process, those simulations are largely at the mercy of quantum computing state-of-the-art capabilities. Today, researchers can only afford to simulate molecules a few hundred atoms at most. This challenge is due to the immense computational resources required for precise simulations. The broader project seeks to enhance this important capability by building in more fault tolerance. It will do this by encoding quantum information redundantly in logical qubits.
And each logical qubit consists of hundreds of physical qubits. In order to do meaningful chemistry simulations, we’re going to require on the order of a million physical qubits. To achieve any reliability in predictions, qubits have to be of high quality. Research has shown that error rates must be near 10^-15.
“Meaningful chemistry simulations require hundreds to thousands of high-quality qubits with error rates of around 10^-15,” noted a quantum physicist involved in the initiative.
A Vision for Future Challenges
The possibilities for quantum-enhanced AI open up much further than just battery materials alone. In their view, these technological advancements hold the key to addressing many of the world’s greatest challenges. They could provide climate change and disease alleviation decades in advance of projections. Creatively and rapidly integrating novel materials can lead to transformative advances. Together these advancements have the potential to radically increase energy efficiency and reduce carbon footprints across the globe.
With climate change an ever-pressing threat, finding new battery materials is essential. Through this collaboration, we hope to not just improve technology, but to make a meaningful impact on environmental sustainability efforts.
“Set up on the earth, and the top of it reached to heaven. And behold the angels of God ascending and descending on it.” – Book of Genesis