Phosphorescent Material Search Using a Combination of High-Throughput Evaluation and Machine Learning.

Area of Science:

• Materials Science
• Thin Film Deposition
• Photoluminescence

Background:

• High-throughput experiments and machine learning (ML) accelerate material discovery.
• Combinatorial chemistry generates large datasets rapidly.
• Improving phosphorescent material afterglow lifetime is crucial for applications.

Purpose of the Study:

• To combine high-throughput experiments and ML to search for semiconducting thin films.
• To enhance the afterglow lifetime of Eu and Dy codoped SrAl2O4-based phosphorescent materials.
• To optimize thin film deposition and calcination conditions.

Main Methods:

• Fabricated a combinatorial library of thin films using various oxide targets (MgO, GeO2, Ga2O3, ZnO, Bi2O3, Ta2O5, TiO2, Y2O3) on SrAl2O4 substrates.
• Utilized high-throughput evaluation to generate a dataset of 800 examples with systematic changes in material ratios and calcination conditions.
• Employed machine learning for interpolation of afterglow lifetime using film thickness and calcination conditions as explanatory variables.

Main Results:

• Identified magnesium oxide (MgO) thin films as effective in improving afterglow lifetime.
• Determined optimal conditions for MgO thin films: approximately 100 nm thickness and calcination at 400-600 °C in air.
• Achieved accurate interpolation of afterglow lifetime using ML, validated by correlation coefficient and root mean squared error.

Conclusions:

• The combined approach of high-throughput experimentation and machine learning is effective for accelerating material development.
• MgO thin films offer a promising route to enhance the afterglow lifetime of SrAl2O4-based phosphors.
• Optimization of film thickness and calcination conditions is key to maximizing performance.