Forum on the Future of Synthesis

Joseph Montoya, Toyota Research Institute

Successes and Vision for Practical Materials Discovery

Written by Jun Meng

The Forum on the Future of Synthesis at the MRS 2024 Fall Meeting highlighted how artificial intelligence (AI) and machine learning (ML) can reshape materials discovery—but with measured optimism. Joseph Montoya from the Toyota Research Institute explored the ongoing challenge of turning AI-predicted materials into real-world discoveries.

While advances in data science have accelerated structure predictions since 2019, Montoya emphasized that synthesizing even one novel material remains a challenging process. He noted that several AI-predicted materials have been attempted in the lab, but 5 out of 6 failed to be successfully synthesized. However, Montoya stressed that failure in synthesis doesn’t negate the potential of these materials—it often depends on the precise conditions under which they are made.

A notable example was Ca1.6​RuOx​, which degraded quickly during testing, revealing unexpected challenges. Montoya introduced a synthesis theory called PIRO, a strategy to bridge theory and practice by incorporating thermodynamic data and phase stability into experimental workflows.

This iterative approach, he explained, holds promise for accelerating materials’ synthesizability, allowing a vast number of candidate materials to be screened and refined in practice. AI in materials discovery is still evolving, but forums like this are paving the way toward actionable breakthroughs.