Development of deep neural network potential for studying water/diamond interfaces

• Main Author: Melvin, Daniel
• Other Authors: Li Shuzhou
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2023
• Subjects: Engineering::Materials
• Online Access: https://hdl.handle.net/10356/172366
• Institution: Nanyang Technological University

Diamond and Diamond-like carbon (DLC) are promising coating materials with high strength and outstanding tribological properties. The ultra-low friction and wear rate make diamonds attractive for real-world applications, especially micro- and nano- electromechanical systems (MEMS/NEMS). However, the termination species’ presence on the carbon surface and the interaction with the environment could significantly influence the tribological properties of the material. Plenty of studies have been dedicated to unveiling the main source of friction and improving the tribological properties of the carbon material surface. Nevertheless, our understanding of the interaction between terminated diamond surfaces and water molecules in the environment and its influence on the tribological properties of the material is still limited. In this study, we developed a deep neural network potential (DNNP) to accurately simulate the interaction between water and different diamond surface terminations. By utilizing a trained DNNP, the system size limitation can be overcome, allowing us to explore the influence of the water layer's structure on the tribological properties of terminated diamond surfaces. We found that the terminated specific terminational molecule species and sliding velocities could highly affect the tribological behavior of the carbon surface. These findings and studies help the advancement and understanding of the tribological properties of diamond-based material development.