基于三轴X射线衍射方法的n-GaN位错密度的测试条件分析

发布时间：2018-05-18 02:17

  本文选题：氮化镓 + 高分辨三轴X射线衍射　； 参考：《物理学报》2017年21期

【摘要】：三轴X射线衍射技术广泛应用于半导体材料参数的精确测试,然而应用于纤锌矿n-GaN位错密度的测试却可能隐藏极大的误差.本文采用三轴X射线衍射技术测试了两个氢化物气相外延方法生长的n-GaN样品,发现两样品对应衍射面的半高全宽都基本一致,按照镶嵌结构模型,采用Srikant方法或Williamson-Hall方法,两样品的位错密度也应基本一致.但van der Pauw变温霍尔效应测试表明,其中的非故意掺杂样品是莫特相变材料,而掺Si样品则是非莫特相变材料,位错密度有数量级的差别.实验表明,位错沿晶界生长导致的晶粒尺寸效应,表现为三轴X射线衍射技术检测不到晶界晶格畸变区域的位错,给测试带来极大误差,这对正确使用Srikant方法和Williamson-Hall方法提出了测试要求.分析表明,当扭转角与倾转角之比β_(twist)/β_(tiit)≥2.0时,Srikant方法是准确的,否则需进一步由Williamson-Hall方法确定晶粒大小(面内共格长度L//),当L//≥1.5μm时,Srikant方法是准确的.
[Abstract]:Triaxial X-ray diffraction technique is widely used in the accurate measurement of semiconductor material parameters, however, the measurement of dislocation density of wurtzite n-GaN may conceal great errors. In this paper, the n-GaN samples grown by two hydride vapor phase epitaxy methods have been measured by using triaxial X-ray diffraction technique. It is found that the half height full width of the corresponding diffraction surfaces of the two samples are basically the same. According to the mosaic structure model, the Srikant method or Williamson-Hall method is used. The dislocation density of the two samples should be basically the same. However, the van der Pauw variable-temperature Hall effect test shows that the unintentionally doped sample is a Mott phase change material, while the Si doped sample is a non-Mott phase change material, and the dislocation density varies by an order of magnitude. The experimental results show that the grain size effect caused by the growth of dislocations along grain boundaries is characterized by triaxial X-ray diffraction, which can not detect the dislocation in the lattice distortion region of grain boundaries, which brings great error to the measurement. This puts forward the test requirements for the correct use of the Srikant method and the Williamson-Hall method. The analysis shows that the Srikant method is accurate when the ratio of torsion angle to tilting angle 尾 is greater than 2.0, otherwise it is necessary to determine the grain size by the Williamson-Hall method (the in-plane coherent length L / R), and the Srikant method is accurate when L / L / 鈮，

本文编号：1903932