Siamese Graph Neural Networks for Melting Temperature Prediction of Molten Salt Eutectics

Key Points

• To develop a model that predicts melting temperatures of molten salt eutectics using Siamese graph neural networks.
• Utilized Siamese graph neural networks for structure-property predictions.
• Performed high-throughput screening to identify optimal models.
• Analyzed different molten salt eutectic mixtures for comprehensive predictions.
• Achieved significant improvements in melting temperature predictions compared to traditional methods.
• Demonstrated the capability of the model to generalize across different eutectics.
• Identified key structural features influencing melting temperature in diverse mixtures.