Single-crystal gallium nitride (GaN) is a semiconductor material known for its hardness and brittleness. This research aims to reveal the differences in the micro-mechanisms of material removal during traditional grinding and ultrasonic vibration-assisted grinding and to provide guidance for the high-efficiency, high-quality planarization processing of single-crystal GaN. To achieve this purpose, molecular dynamics (MD) simulation methods were used to establish a model (30nm × 40nm × 15nm) of single-crystal GaN being scratched by a single abrasive grain with and without ultrasonic vibration assistance. The differences in the surface morphology and subsurface damage formation mechanisms of single-crystal GaN under conditions with and without ultrasonic assistance were compared. The results indicate that, compared to traditional grinding, the periodic ultrasonic vibrations reduce the normal force and result in a more uniform distribution of stress and temperature, thereby mitigating local stress concentration and thermal accumulation effects. Ultrasonic vibration alters the motion of the abrasive grain, expanding the material removal range from 7.384 to 10.315nm, decreasing the number of residual atoms in the machining area, and lowering the chip pileup height at the leading edge of the abrasive grain from 2.063 to 1.528nm. Additionally, the micro-shear deformation induced by ultrasonic vibrations helps to suppress brittle fracturing caused by excessive local stress, thus reducing the thickness of the damaged subsurface. At a scratch depth of 2nm, the total length of dislocation lines in the ultrasonic-assisted scratch is reduced from 185.256 to 33.315 nm compared to traditional scratching. These findings offer new insights into the microscopic mechanisms of material removal in high-efficiency, high-quality grinding processes of single-crystal GaN. This study adopts LAMMPS software for MD simulations to investigate the physical behavior of GaN during the grinding process. The simulation results are then visualized and analyzed using OVITO software to elucidate microstructural changes in the material. During the simulation, the temperature of the Newtonian layer is calculated based on the statistical analysis of atomic velocities under the NVE ensemble. The grinding force applied to the GaN workpiece is determined by differentiating the atomic potential energy. Von Mises stress analysis is adopted to ascertain the stress distribution during the scratching process. Finally, the changes in the crystal structure during scratching are identified and classified using analysis of identified defect structures.
Read full abstract