Quantum Data Revolutionizes AI in Battery Material Research

Researchers are about to change the game in materials science. Now, they are teaming up to realize the full potential of quantum computing paired with artificial intelligence (AI). That’s what happened in the summer of 2001 when John P. Perdew created a highly innovative program. Eventually, their goal is to be able to screen 32…

Tina Reynolds Avatar

By

Quantum Data Revolutionizes AI in Battery Material Research

Researchers are about to change the game in materials science. Now, they are teaming up to realize the full potential of quantum computing paired with artificial intelligence (AI). That’s what happened in the summer of 2001 when John P. Perdew created a highly innovative program. Eventually, their goal is to be able to screen 32 million+ candidate battery materials, leading to safer, cheaper, greener energy.

The Pacific Northwest National Laboratory (PNNL) plays a crucial role in this endeavor by partnering with various researchers to harness advanced AI models. The project confronts timely problems in energy storage and climate change. First, it zeroes in on rapidly discovering new materials that current techniques would need to spend years—or even decades—researching. Today’s announcement further illustrates the ways in which AI, powered by quantum-accurate data, can accelerate predictions and help identify revolutionary new materials.

The Power of AI in Material Evaluation

In a world rapidly approaching the breaking point of concrete climate change, the demand for low-cost energy storage technologies is greater than ever before. This partnership between PNNL, along with other researchers, hopes to maximize the full potential of AI that will allow scientists to evaluate millions of potential battery materials. This sophisticated technology has the capability to sort through millions of candidates in order to find the most promising ones.

Using traditional methodologies, it would realistically take 20 years to evaluate this massive influx of materials. However, AI dramatically accelerates this process. We began with over 32 million different candidate battery materials. In less than a week, we honed it down to 500,000 stable candidates and from there whittled that list to 800 highly promising options.

Enabling this quick evaluation process helps researchers to target their efforts on materials that are best suited to excel through real life applications. If successful, these breakthroughs have the potential to upend the energy industry entirely. They do have the potential to improve battery efficiency and environmentally sustainable.

Quantum Computing’s Role in Chemistry

Quantum computing is poised to be a gamechanger in the field of material science and chemistry. It allows for the first time meaningful simulations to be performed that frankly are impossible today using classical computational techniques. To get there, we need hundreds — or better yet, thousands — of high-quality qubits with low error rates, preferably on the order of 10^-15.

The emerging field of fault-tolerant quantum computing will be essential to this endeavor. In order to achieve reliability, we must encode quantum information redundantly within logical qubits. This more fault tolerant approach is going to take something like a million physical qubits when all is said and done. This historic investment in quantum infrastructure is a big deal, and an important step towards making possible some of the more ambitious chemical simulations.

Electron correlation is vital to understanding systems with strong electron interactions. Gaining a more complete and fundamental understanding of these complex interactions is key to identifying and designing new materials with exotic electronic properties including high- T c superconductors. Further, electron correlation is critical for catalytic action by transition metal atoms, which are central to many industrially-relevant heterogeneous catalysis processes.

A New Era of Material Discovery

Together, the integration of AI and quantum computing represent a revolutionary moment in the quest for new materials. By enabling rapid predictions at a fraction of the cost of classical computing, researchers can identify “first-time right” candidates—those most likely to succeed upon synthesis and testing. In addition to speed, this efficiency directly translates into greatly reduced expenditures of people’s time and public resources.

The collaboration’s early results are promising. Now, scientists have frozen their selection process from millions of likely materials down to a list of hundreds of candidates that are prepared to undergo laboratory testing. Even with multiple shows, this lean approach can be highly effective. It further addresses the increasing demand for sustainable energy solutions as climate change continues to take center stage worldwide.

“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