论文部分内容阅读
Two strain-state samples of GaN,labelled the strain-relief sample and the quality-improved sample,were grown by hydride vapour phase epitaxy (HVPE),and then characterized by high-resolution X-ray diffraction,photoluminescence and optical microscopy.Two strain states of GaN in HVPE,like 3D and 2D growth modes in metal-organic chemical vapour deposition (MOCVD),provide an effective way to solve the heteroepitaxial problems of both strain relief and quality improvement.The gradual variation method (GVM),developed based on the two strain states,is characterized by growth parameters’ gradual variation alternating between the strain-relief growth conditions and the quality-improved growth conditions.In GVM,the introduction of the strain-relief amplitude,which is defined by the range from the quality-improved growth conditions to the strain-relief growth conditions,makes the strain-relief control concise and effective.The 300-μm thick bright and crack-free GaN film grown on a two-inch sapphire proves the effectiveness of GVM.
Two strain-state samples of GaN, labeled the strain-relief sample and the quality-improved sample, were grown by hydride vapor phase epitaxy (HVPE), and then characterized by high-resolution X-ray diffraction, photoluminescence and optical microscopy. Two strain states of GaN in HVPE, like 3D and 2D growth modes in metal-organic chemical vapor deposition (MOCVD), provide an effective way to solve the heteroepitaxial problems of both strain relief and quality improvement.The gradual variation method (GVM), developed based on the two strain states is is characterized by growth parameters’ gradual variation alternating between the strain-relief growth conditions and the quality-improved growth conditions. In GVM, the introduction of the strain-relief amplitude, which is defined by the range from the quality-improved growth conditions to the strain-relief growth conditions, makes the strain-relief control concise and effective. The 300-μm thick and crack-free GaN film grown on a two-inch sapphi re proves the effectiveness of GVM.