Real-time experiment-theory closed-loop interaction for autonomous materials science.

Area of Science:

• Materials Science
• Computational Materials Science
• Chemical Engineering

Background:

• The scientific method relies on iterative cycles of theoretical prediction and experimental validation.
• Closing the loop between experiments and theory is often challenging due to computational or time constraints.
• Autonomous systems are needed to accelerate materials discovery and characterization.

Purpose of the Study:

• To develop and demonstrate an autonomous materials search engine (AMASE) for self-driving, cyclical materials exploration.
• To apply AMASE for rapid mapping of temperature-composition phase diagrams.
• To showcase real-time, autonomous, and iterative interactions between experiments and computational predictions without human intervention.

Main Methods:

• Developed AMASE, an autonomous system integrating experimental and computational workflows.
• Employed CALPHAD (Calculation of Phase Diagrams) for real-time phase diagram prediction.
• Utilized AMASE to guide experimental determination of phase boundaries in thin films.

Main Results:

• AMASE successfully and autonomously mapped the eutectic phase diagram of the Sn-Bi thin-film system.
• Achieved a sixfold reduction in the number of experiments required for phase diagram determination.
• Demonstrated accurate phase boundary determination through a self-guided campaign covering a small fraction of the phase space.

Conclusions:

• AMASE enables efficient and autonomous materials exploration through continuous experiment-theory cycles.
• This approach significantly accelerates the process of phase diagram mapping.
• Autonomous systems represent a paradigm shift in materials discovery and characterization.