Stress-Driven Grain Boundary Structural Transition in Diamond by Machine Learning Potential

Area of Science:

• Materials Science
• Computational Materials Science
• Nanotechnology

Background:

• Understanding carbon grain boundary dynamics, especially in diamond, is vital for advanced device applications.
• Experimental and computational limitations hinder the analysis of these complex structures.
• Carbon's diverse allotropes offer significant potential, but their boundary behavior remains challenging to study.

Purpose of the Study:

• To develop a machine learning-based molecular dynamics potential for predicting structural transitions in diamond grain boundaries.
• To elucidate the atomic-scale mechanisms governing these transitions.
• To quantify the impact of these transitions on thermal properties.

Main Methods:

• Developed a machine learning potential trained on ab initio data.
• Utilized molecular dynamics simulations to study incoherent twin boundaries in diamond.
• Analyzed atomic-scale mechanisms and interfacial thermal conductance.

Main Results:

• The machine learning potential accurately predicts structural transitions in diamond grain boundaries.
• Atomic-scale mechanisms driving these transitions were identified.
• An 80% reduction in interfacial thermal conductance was observed during grain boundary transition.

Conclusions:

• The study provides significant insights into the behavior of diamond grain boundaries.
• A novel mechanism regulating thermal properties at grain boundaries was uncovered.
• Findings pave the way for improved thermal management in diamond-based technologies.