Optimization of AlN Crystal during Iterative PVT Growth

Crystal quality optimization is crucial in physical vapor transport crystal growth, particularly when using poor-quality AlN seeds derived from heterogeneous growth on SiC substrates. Herein, the quality optimization of AlN crystals was explored in homogeneous iterative growth. Based on the dislocation reduction mechanism associated with crystal growth morphology and the low thermal stress condition, a convex-shaped crystal growth method has been developed by using a synergistic control strategy that optimizes the thermal and flow fields. Additionally, a wafer-surface etching method was designed to enable wafers to serve as qualified seeds for the next generations, effectively suppressing polycrystalline formation. The iterative growth process employed surface-etched seeds to ensure quality inheritance and convex growth for further improvement. Larger convexity in the earlier growth generations enabled rapid dislocation reduction, while reduced convexity in subsequent generations facilitated gradual quality optimization while preserving the crystal size and growth efficiency. Systematic analysis of crystal appearance, impurity content, structural quality, and transmittance demonstrated consistent improvements both within individual runs and across generations. Compared to the initial-generation wafer, the third-generation wafer showed a 67.2% reduction in the fwhm of the X-ray diffraction (XRD) (0002) rocking curve and a 52.7% reduction in the fwhm of the Raman E2(high) mode, along with a 45.2%, 56.8%, and 52.6% decrease in absorption coefficient at 240, 300, and 450 nm, respectively. The effective iterative growth process enables the realization of large-sized, high-quality AlN crystals via a heteroto-homogeneous growth route.