New AI approach accelerates targeted material discovery and creates foundation for self-driving experiments

Scientists have developed an AI-based method that allows them to collect data more efficiently in the search for new materials, helping them tackle complex design challenges more accurately and quickly.

This research emerged from a collaboration between computer science and materials science researchers at the Department of Energy’s SLAC National Accelerator Laboratory and Stanford University. The collaboration combines expertise in algorithm development, machine learning, and materials science.

Their work, published today in npj Computational Materialslays the foundation for “self-driving experiments,” in which an intelligent algorithm defines the parameters for the next set of measurements at facilities such as SLAC’s Linac Coherent Light Source (LCLS). The new method also enables the rapid discovery of new materials, which could hold promise in areas such as climate change, quantum computing, and drug design.

Traditional materials discovery has historically been a time-consuming and expensive process due to the expense of both creating and measuring the properties of new materials. The space of possible materials is also extremely large, with over 10 billion possibilities for materials with just four elements. For pharmaceutical applications, the challenge is even greater, with around 10⁶⁰ potential drug-like molecules containing only the most basic building blocks (C, H, O, N, and S atoms).

The task is further complicated by the need to meet complex design goals, such as discovering conditions to synthesize nanoparticles of different sizes, shapes, and compositions. Traditional methods, which typically maximize or minimize a simple property, are often too slow to sift through large search spaces to discover new materials that fit a researcher’s goals.

This paper proposes a new approach that automatically transforms complex goals into intelligent data acquisition strategies. A key feature is the ability to learn and improve from each experiment, using AI to suggest next steps based on the data collected so far. The innovation is based on the concept of Bayesian algorithm execution (BAX), recently developed by co-author Willie Neiswanger, a postdoctoral researcher in computer science at Stanford at the time of the study. In this method, a complex goal can be written as a simple shopping list or recipe, which excels in situations where multiple physical properties must be taken into account.

Another important aspect is that this method is user-friendly and open source, allowing scientists all over the world to use and adapt the method for their research. This promotes collaboration and innovation worldwide.

The researchers tested their approach on a variety of customized targets for nanomaterial synthesis and magnetic material characterization. The results showed that their methods were significantly more efficient than current techniques, especially in complex scenarios.

“Our method allows you to specify complex objectives, which enables automatic optimization over a large design space, increasing the likelihood of finding new, amazing materials,” said Sathya Chitturi, a Ph.D. student at SLAC and Stanford who led the research. “The Bayesian algorithm execution framework allows you to capture the complexity of materials design tasks in an elegant and simple way.”

The ability to design materials with specific catalytic properties, for example, could improve chemical processes that lead to more efficient and sustainable ways of producing goods and materials, reducing energy consumption and waste. In manufacturing, new materials could improve processes such as 3D printing, allowing for more precise and sustainable production. In healthcare, tailored drug delivery systems could improve the targeting and delivery of therapeutic agents, increasing efficacy and reducing side effects.

The researchers are already implementing ways to integrate this framework into experimental and simulation-based research projects to demonstrate its broad applicability and effectiveness.

“The project is a great example of multidisciplinary collaboration between SLAC and Stanford,” said collaborator Daniel Ratner, head of SLAC’s Machine Learning Initiative (MLI). “Sathya was able to adapt Willie’s core research in algorithmic computer science to tackle real scientific problems in materials science.”

MLI researchers are currently investigating applications for large-scale materials simulations, and Neiswanger, Ratner and collaborators recently published a related application of BAX to optimize the performance of particle accelerators.

“By combining advanced algorithms with targeted experimental strategies, our method makes the process of discovering new materials easier and faster,” said collaborator Chris Tassone, director of the Materials Science Division for the Stanford Synchrotron Radiation Lightsource (SSRL) at SLAC. “This could lead to new innovations and applications in many industries.”