Numerical Simulation and Experimental Research on Reducing Basal Plane Dislocation(BPD) in 8-inch SiC Single Crystal

Numerical Simulation and Experimental Research on Reducing Basal Plane Dislocation(BPD) in 8-inch SiC Single Crystal

PAM-XIAMEN is able to manufacture 8 inch SiC wafers, specific wafer parameters please refer to: https://www.powerwaywafer.com/200mm-wafers.html

The quality of SiC single crystals has a significant impact on device performance, and basal dislocations (BPD) are one of the important defects that affect the application of SiC chips. The Physical Vapor Transport (PVT) method is a mature method for growing SiC crystals, but a large number of dislocations are generated during the growth process. Therefore, controlling and reducing dislocations, especially BPD, is an important issue for improving the quality of SiC chips.

Because of such problems, researchers studied the effects of cooling process, seed bonding method, and graphite crucible material on BPD density in 8-inch N-type 4H-SiC single crystal growth through numerical simulation and corresponding experiments. The aim was to analyze and find effective methods to reduce BPD density and improve the quality of SiC single crystal.

1. Numerical Simulation for 8-inch SiC Single Crystal

The growth and cooling process of 8-inch SiC single crystals were simulated using STR-VR software, and the stress and BPD density of SiC crystals were analyzed. According to the simulation results, four groups of experiments were designed to verify the effectiveness of the simulation results. In the experiment, the cooling rate, the seed bonding method and the thermal expansion coefficient of the graphite crucible were adjusted, shown as Fig. 2.

Fig. 1 Schematic diagram and temperature distribution of SiC single crystal growth based on PVT

Fig. 1 Schematic diagram and temperature distribution of SiC single crystal growth based on physical vapor transport (PVT) method. This is the foundation for understanding the entire experimental setup and simulation conditions.

Fig. 2 Simulation conditions for SiC crystal growth

Fig. 2 Simulation conditions for SiC crystal growth: cooling rate, seed bonding method, and thermal expansion coefficient of graphite crucible

By comparing the maximum shear stress and dislocation density under different conditions, the effects of cooling rate, seed crystal bonding method, and graphite crucible material on BPD density were analyzed. Please refer to the following figure for details:

Fig. 3 Under simulation condition A, evolution of shear stress and dislocation density during SiC crystal growth process

Fig. 3 Under simulation condition A, evolution of shear stress and dislocation density during SiC crystal growth process. (a1)-(a4) show the changes in shear stress during the SiC crytsal growth process, while (a5)-(a8) show the changes in dislocation density. (b1)-(b4) and (b5)-(b8) respectively show the changes in shear stress and dislocation density during the cooling process.

Fig. 4 Changes in maximum shear stress and dislocation density during SiC single crystal (a, b) growth process and (c, d) cooling process

Fig. 4 Changes in maximum shear stress and dislocation density during SiC single crystal (a, b) growth process and (c, d) cooling process under four simulated conditions. By comparing different simulation conditions, the effects of cooling rate, stress boundary conditions, and thermal expansion coefficient of graphite crucibles on dislocation density can be analyzed.

Fig. 5 Compares the dislocation density of SiC crystals at (a1-a4) growth process and at (b1-b4) cooling process to 500 ° C

Fig. 5 Compares the dislocation density of SiC crystals at (a1-a4) growth process and at (b1-b4) cooling process to 500 ° C under four simulated conditions. The results showed that increasing cooling rate, optimizing seed crystal bonding method, and graphite crucible material can significantly reduce dislocation density.

Fig. 6 The distribution of etching pits on SiC wafers [(a) sample A, (b) sample B, (c) sample C, (d) sample D]

Fig. 6 The distribution of etching pits on SiC wafers [(a) sample A, (b) sample B, (c) sample C, (d) sample D] obtained from four experimental groups after processing and etching. The elliptical etching pits in the figure were identified as BPDs, and the number showed a decreasing trend, indicating that the measures taken in the experiment effectively reduced the BPD density.

Fig. 7 The comparison of BPD density under different conditions

Fig. 7 The comparison of BPD density under different conditions. The BPD density of SiC wafers: A to D were 4689cm-2, 2925 cm-2, 1560 cm-2, and 704 cm-2, respectively. This indicates that by increasing the cooling rate, optimizing the seed crystal bonding method, and using graphite crucibles with lower thermal expansion coefficients, the BPD density can be effectively controlled and reduced.

2. Conclusions for Numerical Simulation in 8 inch SiC Crystal Growth

Research has found that increasing cooling rate, optimizing seed crystal bonding methods, and using graphite crucible materials with thermal expansion coefficients similar to SiC single crystals can effectively reduce BPD density.

The experimental results show that high cooling rate reduces the BPD density of 8-inch SiC wafers from 4689 cm-2 to 2925 cm-2; After optimizing the seed crystal bonding method, The BPD density of SiC wafers decreased to 1560 cm-2; By using graphite crucibles with lower thermal expansion coefficients, The SiC wafer BPD density further decreased to 704 cm-2.

The conclusion of the article validates the numerical simulation results and points out that these optimization methods can significantly improve the quality of SiC single crystals.

powerwaywafer

For more information, please contact us email at [email protected] and [email protected].


has been added to your cart.
Checkout