| Home | Log-in | Admin | Sitemap |
Home | Program | Detailed Program
Detailed Program
Paper Number : LD-I04
Time Frame : 13:55~14:20
Presentation Date : Thurseday, 27, November
Session Name : LED and Display Matrials
Session Chair 1# : Inhwan Lee
Session Chair 2# : Yasushi Nanishi
In situ XRD analysis of GaInN/GaN heterostructure grown on GaN underlying layer with different dislocation density
Motoaki Iwaya
Meijo University
A GaInN/GaN heterostructure system is widely used to high-brightness blue and green LEDs and high power violet laser diodes. The application of GaInN films in high-efficiency solar cells is also expected, because the bandgap of GaInN alloys ranges from 0.65 to 3.43 eV. Although there are many reports of the analysis of the relaxation process in GaInN/GaN heterostructure system characterized by TEM and XRD reciprocal space mapping (XRD-map), an understanding of the critical layer thickness at which misfit dislocations (MDs) are introduced in the GaInN/GaN
heterostructure is insufficient. For instance, a large fluctuation exists in the reported value of the critical thickness in GaInN/GaN heterostructure system. In this study, we investigated an in situ XRD analysis of GaInN/GaN heterostructure grown on GaN underlying layer with different dislocation density. The GaInN layers were grown on GaN template (threading dislocation density: ~3 ¡¿ 108 cm-2) on sapphire substrate covered with lowtemperature buffer layer and freestanding GaN substrates (threading dislocation density: <3 ¡¿ 106 cm- 2). We evaluated the GaInN films with symmetric (0002) Bragg diffraction using an in situ XRD system. The X-ray was focused on the sample surface using a Johansson curved crystal mirror. By using this method, the incidence angle of the X-rays can be changed without moving the X-ray source. Moreover, the diffracted X-ray was detected by a 1D X-ray CCD. By using this method, the scattered X-ray can be detected without moving the substrate and detector. By using this configuration, this system realized the equivalent of a (0002) 2¥è/¥ø scan without requiring the use of an analyzer crystal in 1 s during the rotation of the wafers. Although the resolution limit decreases slightly, this in situ XRD system is able to perform the equivalent of a (0002) 2¥è/¥ø scan at a resolution of 1 arcsec. In this setup, the full width at half maximum (FWHM) is both controlled by the dispersion in the lattice constant c and the mosaicity of the crystal. We observed the GaInN surface structure and defects close to the GaInN/GaN heterointerface by SEM, CL, AFM and TEM, respectively. Strain relaxation was also evaluated by a typical ex situ XRD-map. By comparing in situ XRD measurement results and ex situ characterization such as TEM, CL, AFM, SEM, and XRD-map, we found that it is possible to accurately determine the critical thickness of introduced the MDs in the GaInN are introduced by analyzing the FWHM of the in situ XRD spectrum from GaInN. In particular, we found that measurement accuracy of the critical thickness by this in situ XRD system is higher than that by XRD-maps. Moreover, critical thickness of introduced MDs in GaInN was significant depends on the dislocation density in the GaN underlying layer.
Acknowledgements :