ABSTRACT
XRD (0002) omega-2theta profiles of the InGaN epilayers, which were grown with different In/(In+Ga) flux ratio ranging from 0.93 to 0.16, are shown in Fig.2. For all the samples, diffraction peaks from the InGaN layers were clearly observed and the peak angle was increased with the decrease of In/(In+Ga) flux ratio. In the higher In/(In+Ga) flux ratio region 0.93-0.62, a single diffraction peak from the InGaN layer was observed(CASE 1). However, in the lower In/(In+Ga) flux ratio region below 0.53, two diffraction peaks were observed(CASE 2). For example, the diffraction peaks were observed at 16.1° and 16.4° for the InGaN sample grown with the In/(In+Ga) of 0.53. The plausible reason for this phenomenon may be the compositional phase separation or the mixing of cubic phase in the InGaN layers. It is difficult to distinguish the h-InGaN(0002) and c-InG aN (lll) diffraction peaks because each diffraction angle is the same position in the mixed configuration of h-InGaN (0001 )//c-InGaN(111). Therefore, the mixing of cubic phase was investigated by measuring the asymmetric c-InGaN(002) diffraction angle by XRD. Figure3 shows the XRD profiles around asymmetric c-InGaN(002) diffraction angle for the InN and two case (CASE 1 and 2) of the InGaN samples. Any peaks were not observed for the InN and CASE 1 InGaN samples. However, for the CASE 2 InGaN samples, a cubic (002) diffraction peak was surely observed between the c-GaN(002) diffraction angle at 20.0° and the c-InN(002) diffraction angle at 18.0°. This result indicates that the mixing of the cubic phase InGaN occurred in this sample. The cubic (111) diffraction angle of the cubic InGaN calculated from c-InGaN(002) diffraction angle was about 16.1°. This value agrees well with the lower angle peak of the two peaks observed for the CASE2 InGaN samples grown with the In/(In+Ga) flux ratio of 0.53. For the other InGaN samples grown at the In/(In+Ga) flux ratio below 0.53, the cubic (002) diffraction peaks were also observed. Therefore, lower angle peaks are attributed to the cubic phase mixing. These results indicate that hexagonal InGaN is unstable in this region because of poor Ga atom migration. It is thought that the optimum temperature for h-InGaN growth gets higher with decrease of InN molar fraction. Next, for 400nm thick Ino.93Gao.07N samples, XRD reciprocal space mapping around (20-25) diffraction was measured in order to examine lattice relaxation. This result indicates that the InGaN layer on InN was fully relaxed. Therefore, for all the InGaN samples in this study, the InN molar fraction was obtained by using Vegard’s law.
