Interfacial Optimization for AlN/Diamond Heterostructures via Machine Learning Potential Molecular Dynamics...

Area of Science:

• Materials Science
• Semiconductor Physics
• Computational Materials Science

Background:

• Aluminum nitride (AlN)/diamond heterostructures offer exceptional properties for high-power electronics.
• Poor interfacial adhesion between AlN and diamond leads to device failure.
• Developing robust interfaces is critical for next-generation electronic devices.

Purpose of the Study:

• Investigate the uniaxial tensile failure mechanisms of AlN/diamond heterogeneous interfaces.
• Identify strategies to improve interfacial mechanical properties and adhesion.
• Provide solutions for optimizing ultrawide band gap semiconductor heterostructures.

Main Methods:

• Molecular dynamics simulations utilizing a neuroevolutionary machine learning potential (NEP) model.
• Trained NEP accurately describes interatomic interactions and predicts cleavage planes.
• Analysis of the effects of annealing, strain engineering, surface roughness, and nanostructures.

Main Results:

• Annealing treatment enhances interfacial binding by improving C-N atom interactions, reducing potential energy.
• Strain engineering of AlN significantly impacts interfacial mechanical properties.
• Reduced surface roughness, AlN strain engineering, and annealing collectively improve interfacial stability.

Conclusions:

• A combination of surface roughness reduction, AlN strain engineering, and annealing offers a promising approach to enhance interfacial mechanical properties.
• Optimized interfacial properties are crucial for the reliable performance of AlN/diamond heterostructures in high-power devices.
• This study provides a pathway for developing superior ultrawide band gap semiconductor heterostructures.