Enhancing High-Dimensional Neural Network Potentials Accuracy in OLED Systems via Element-Relabeling

Key Points

• The aim is to improve the accuracy of atomistic simulations for OLED materials using high-dimensional neural network potentials.
• Utilized neural networks to model atomistic interactions
• Applied element-relabelling techniques for enhanced accuracy
• Compared results with traditional classical force fields
• Achieved higher accuracy in atomistic simulations compared to classical force fields
• Demonstrated significant improvement in the modeling of complex OLED environments