Semiconductor crosshatch patterns form as a result of strain relaxation and dislocation pileups during growth of lattice mismatched materials. Due to their connection with the internal misfit dislocation network, these patterns are a complex fingerprint of internal strain relaxation and growth anisotropy. Therefore, these patterns not only describe the residual strain state at the surface, but could also provide an indicator of electron transport quality through the material. Here, we present a method utilizing computer thresholding and machine learning to analyze antiferromagnetic (AFM) crosshatch patterns that exhibits this correlation. Our analysis reveals optimized electron transport for moderate values of $\lambda$ (crosshatch wavelength) and $\epsilon$ (crosshatch height), roughly 1 $\mu \mathrm{m}$ and 4 nm, respectively, that define the average wave form of the pattern. Simulated two-dimensional AFM crosshatch patterns are used to train a machine learning model to correlate the crosshatch patterns to dislocation density. This model is used to evaluate the experimental AFM images and predict a dislocation density based on the crosshatch wave form. Predicted dislocation density, experimental AFM crosshatch data, and transport measurements are used to train a final model to predict electron mean free path. Our model shows electron scattering is dominated by elastic effects (e.g., dislocations) below 150 nm ${\overline{\lambda }}_{\mathrm{MFP}}$.

Close

PrevNext

View figure in article

• Quantum transport
• Roughness
• Atomic force microscopy
• Machine learning
• Molecular beam epitaxy