New Methodology for Developing Innovative Materials
Author Information
Author(s): Yukari Katsura, Masakazu Akiyama, Haruhiko Morito, Masaya Fujioka, Tohru Sugahara
Primary Institution: National Institute for Materials Science (NIMS)
Hypothesis
Can materials informatics and machine learning accelerate the discovery of new inorganic materials?
Conclusion
The study successfully developed new methodologies and tools for discovering new inorganic materials, leading to the synthesis of numerous new phases.
Supporting Evidence
- Developed Element Reactivity Maps to predict compound formation probabilities.
- Created the Crystal Cluster Simulator for intuitive crystal structure design.
- Conducted large-scale synthesis experiments resulting in over 7,000 samples.
Takeaway
Researchers created new tools to help find and make new materials, discovering many new types in the process.
Methodology
The study utilized machine learning on crystal structure databases and developed tools like the Element Reactivity Maps and Crystal Cluster Simulator to explore new materials.
Potential Biases
Potential biases in the data used for machine learning could affect the predictions of material properties.
Limitations
The study's reliance on existing databases may limit the discovery of truly novel materials not represented in the data.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website