Probabilistic Isolation of Crystalline Inorganic Phases.

Area of Science:

• Materials Science
• Crystallography
• Computational Chemistry

Background:

• Identifying unknown crystalline phases is crucial for materials discovery.
• Current methods for isolating unknown phases can be time-consuming and labor-intensive.
• Automating this process can significantly accelerate the exploration of new materials.

Purpose of the Study:

• To present Probabilistic Isolation of Crystalline Inorganic Phases (PICIP), a novel computational tool.
• To automate the isolation of unknown crystalline inorganic phases detected experimentally.
• To accelerate the overall process of materials discovery.

Main Methods:

• PICIP infers unknown phase composition using sample and known phase compositions.
• A novel algorithm estimates probability density over a linear compositional phase space.
• Iterative sampling strategies refine target compositions for increased accuracy.
• Chemical constraints like charge neutrality reduce phase space dimensionality.

Main Results:

• Simulations show >90% median purity of the unknown phase after four sequential samples.
• PICIP is robust to experimental errors in phase quantification (up to 13 wt %).
• The tool can identify scenarios with significant experimental error.

Conclusions:

• PICIP offers an automated, efficient approach to isolating unknown crystalline phases.
• The probabilistic method enhances accuracy and robustness in materials discovery workflows.
• This tool supports both traditional and high-throughput experimental approaches.