How to accelerate the inorganic materials synthesis: from computational guidelines to data-driven method?

Area of Science:

• Materials Science
• Computational Chemistry
• Chemical Engineering

Background:

• Novel functional materials are crucial for global challenges.
• Experimental synthesis is a key bottleneck in materials development.
• Computational power and machine learning (ML) offer solutions for optimizing synthesis.

Purpose of the Study:

• To review the latest advancements in computation-guided and ML-assisted inorganic material synthesis.
• To provide a comprehensive overview of physical models, data-driven methods, and ML techniques.
• To highlight challenges and opportunities in the field.

Main Methods:

• Introduction to common inorganic material synthesis methods.
• Discussion of thermodynamic and kinetic physical models for synthesis feasibility.
• Exploration of data acquisition, material descriptors, and ML techniques.
• Classification of ML applications based on material data sources.

Main Results:

• Computational and ML approaches significantly accelerate and optimize material synthesis.
• Physical models aid in understanding synthesis feasibility.
• Various ML techniques are applied across different material data sources.
• The review consolidates current progress in ML-assisted inorganic material synthesis.

Conclusions:

• ML-assisted synthesis is a powerful tool for developing novel functional materials.
• Further research is needed to address challenges and unlock opportunities in ML-guided synthesis.
• This review offers scientific guidance for future advancements in the field.