Open Access
Add to BibSonomyAbstract
The epitaxial layers of n-type GaN were grown on both planar and patterned sapphire substrate (PSS) by metal organic chemical vapor deposition. By comparing the epitaxial layers grown on planar substrate, GaN grown on PSS exhibited many improvements both on surface morphology and crystalline quality according to the characterization of atoms force microscopy, and high resolution X-ray diffraction. Spatially resolved micro-Raman scattering results were performed for mapping the spatial variations in crystalline quality of the n-type GaN grown on PSS. According to the variations on the intensity and the full width at half maximum of GaN E2 (high) peaks, the crystalline quality improvement occurred in the lateral growth regions which correspond to center region of the pyramid patterns. We proposed that the bending of dislocations during the lateral growth plays an important role in the spatial variations of GaN crystalline quality. Cross sectional transmission electron microscope and spatial cathodoluminescence mapping results further supported the explanation of the dislocation inhibition during the growth process of GaN grown on PSS.
© 2016 Optical Society of America
1. Introduction
Gallium nitride (GaN) and related compound materials have been widely studied due to their outstanding properties for applications in electronic and optoelectronic devices, such as light-emitting diodes (LEDs), laser diodes (LDs), high-power, and high-frequency devices [1–3]. So far, the GaN based LED is mainly grown on foreign substrates owing to the high prices and small size of bulk GaN substrates. During the heteroepitaxial growth, the large lattice constant and thermal expansion coefficient mismatches between the foreign substrate and GaN could generate a large amount of dislocations (108–1010 cm−2) [4]. In general, the high density of dislocation which acts as the nonradiative recombination center could dramatically decrease the performance of GaN based LED. Thus, many researches have attempted to eliminate or inhibit the dislocations in the GaN epilayers during the past decades [5–7]. One of promising methods is the epitaxial lateral overgrowth (ELOG) which uses the strip-mask patterns to interrupt the propagation of threading dislocations (TDs) [8]. Although the ELOG technique can effectively inhibit the TDs, the process still suffers the disadvantages of both growth interruption due to the mask deposition and high TDs density in the window regions of the overgrowth layer. To overcome these problems, the mask-free patterned sapphire substrate (PSS) technology becomes very popular for high-efficiency InGaN/GaN-based LEDs of commercial productions [9]. A tremendous amount of researches related to GaN epilayers grown on PSS have focused on understanding the mechanism of TDs reduction, and evaluating the device’s capability [10, 11]. Also, some studies have compared the improvement of crystalline quality using different shapes of PSS [12, 13]. Raman spectroscopy is a sensitive method in study of stress distributions, crystalline quality and lattice properties of semiconductor materials. As a non-invasive, non-contact optical testing method, it is more convenient than transmission electronic microscopy (TEM) measurement considering its complex sample preparation process. However, to our knowledge, there are few studies reporting the evaluation of the spatial variations of crystalline quality at different regions of GaN epilayers grown on the PSS using Raman spectroscopy.
In this study, we investigate the improvements in the surface morphology and crystalline quality of n-type GaN grown on PSS comparing with that grown on planar substrate. In addition, we further demonstrate spatially resolved micro-Raman scattering results for mapping the spatial variations in crystalline quality of the n-type GaN grown on PSS.
2. Experimental
Figure 1 shows the top view and cross sectional scanning electronics microscopy (SEM) images of the patterned sapphire substrate we employed in this study. As shown in Fig. 1(a), the pattern is formed smoothly and arranged uniformly in a triangular pyramid shape. The pattern of the PSS was fabricated by standard photolithography and subsequent inductively coupled plasma dry etching to form an array of pyramids on the surface of c-plane sapphire substrate [14]. From the cross sectional SEM image shown in Fig. 1(b), it is noticed that the upper part of the pyramid has a higher inclined angle than the base part. Also, the typical height of the pyramid is about 1.63 μm. In order to identify the improvement in crystalline quality of GaN epitaxial layer grown on PSS, planar sapphire substrate was supplied as a comparison in the same growth run with same growth process.
The epitaxial layers of N-type GaN in this work were performed in a commercial metal organic chemical vapor deposition (MOCVD) system on both planar c-plane substrate and PSS. Trimethylgallium (TMGa), Trimethylaluminium (TMAl) were used as the Ga and Al sources respectively. Ammonia (NH3) was used as the N source. Disilane (Si2H6) were used for n-type doping source. H2 was carrier gas during the growth of n-type GaN epilayers. The growth was initiated with a low temperature GaN nucleation layer grown at 520°C. Then, the growth temperature was risen up to 1030°C immediately. After the growth of 2.7 μm undoped high temperature GaN buffer layer, 100nm AlGaN/GaN superlattices were grown to adjust the strain of the epilayers. Finally, 2.5 μm Si-doped n-GaN layer was grown at 1080°C with n = 1 × 1018cm−3. Here we name the n-type GaN epitaxial layers grown on planar c-plane substrate as sample A while the one grown on PSS as sample B. After the growth process, the surface roughness and crystalline quality of two samples were characterized by atomic force microscopy (AFM), high-resolution X-ray diffraction (HRXRD), and cathodoluminescence (CL) spectroscopy. The distribution of crystalline quality was investigated according to the mapping results of Micro-Raman spectroscopy. All Raman spectra were measured in Z(_,_)-Z backscattering geometries at room temperature. Here we defined z‖ [0001], x‖ [110], and y‖ [100]. The Raman spectroscopy used in this study was a confocal Jobin Yvon LavRam HR800 micro-Raman spectrometer with a charge-coupled device (CCD) detector and an optical microscopy system. The wavelength of the excitation laser was 514 nm and the laser power was about 2 mW. The laser beam spot was focused to a diameter of about 1 μ m on the sample surface. An automatic XY stage allowed the samples to be moved in a minimum step of 0.1 μm.
3. Results and discussion
The surface roughness of the n-type GaN was examined by the AFM measurement. In Figs. 2(a) and 2(b), we compare the AFM images (10 × 10 μm2) of the n-type GaN grown on both the planar substrate (a) and the PSS (b). From the AFM images, we can calibrate the surface roughness of 1.282 nm in sample A and 0.777 nm in sample B. The decrease of the surface roughness in sample B is clear evidence showing that using PSS could effectively improve the surface morphology of the n-type GaN epilayers. It should also be noted that several black pits are observed in Fig. 2(a), which is absent in Fig. 2(b). Different pit intensities between two samples indicate that there exists a variation of dislocation density between sample A and sample B.
XRD analysis is a typical measurement for revealing material quality, which can evidently prove the effectiveness of patterns in crystal growth. In order to identify the influence of PSS on the inhibition of dislocation and the improvement in crystal quality, XRD rocking curves were employed to compare the TD densities of n-GaN grown on both the PSS and planar substrate. It was reported that the ω-scan rocking curves on the symmetric (002) planes are influenced by screw-type dislocations, whereas the asymmetric (102) planes rocking curves are sensitive to edge-type dislocations [15, 16]. Figures 3(a) and 3(b) show HRXRD ω-scan rocking curves of the symmetric (002) and asymmetric (102) reflections for n-GaN on PSS and planar substrate respectively. The full width at half maximum (FWHM) of the symmetry (002) ω-scan rocking curve for GaN on PSS (279 arc sec) was found to be much smaller than that of on the planar substrate (436 arc sec). Furthermore, the FWHM of the asymmetry (102) rocking curve on planar substrate (784 arc sec) was much large than that of on the PSS (306 arc sec). Therefore, it is clearly seen that there is a significant reduction of TDs in the n-GaN grown on PSS, comparing with that grown on the planar substrate, especially the edge-type dislocations.
Fig. 3 (a) Symmetric (002) and (b) asymmetric (102) reflection XRD x-scan rocking curves measured for both samples.
The optical properties of semiconductors could also reflect the crystalline quality of materials. Temperature dependent PL are very effective methods to evaluate the improvement of crystalline quality of GaN film grown on PSS. Macroscopic optical properties of the GaN film grown on both the planar substrate and PSS were studied by PL spectra in the temperature ranged from 80k to 280 K, using a 325nm line of a He-Cd laser. Figure 4 illustrates typical PL spectra obtained from the GaN surface of two samples as a function of temperature. Many features are shown in Fig. 4(a) and4(b). Firstly, with the decrease of temperature, the intensity of the near band edge (NBE) peak becomes larger, while the peak position is blue shifted. This trend happens in both samples. Meanwhile, the yellow luminescence (YL) is observed in the PL spectra of both samples. The intensity of YL is decreased with increasing the temperature. Several variations are very noticeable in the PL spectra. The intensity of both NBE peak and YL in sample B is larger than in sample A. Also, the FWHM of the NBE peak in sample B is smaller than sample A at the same temperature. This indicates that the crystalline quality of GaN films in sample B is better than in sample A.
Fig. 4 Temperature-dependent PL spectra of (a) sample A and (b) sample B. The temperature ranges from 80k to 280k.
In order to investigate the spatial variations in crystal quality of n-type GaN grown on PSS, micro-Raman measurements have been carried out for both samples. According to group theory predicts [17], taking hexagonal GaN belonging to the space group, there exist six first order Raman-active optical modes at Γ point: A1 (LO) + A1 (TO) + E1 (LO) + E1 (TO) + E2 (low) + E2 (high) in various scattering geometry configurations. Under Z (_, _) -Z geometry, only E2 (low), E2 (high) and A1 (LO) mode peaks are allowed to appear in the spectra. The E2 (high) mode peak is considered to be the most sensitive to stress and always used to characterize the residual stress in GaN thin films [18]. The E2 (high) mode peak of strain-free GaN should be at 567.6 cm−1, while the peak would up shift when it is under compressive stress [19]. Figure 5 illustrates representative Raman spectra of both samples. As show in Fig. 5, three peaks located at about 142 cm−1, 570 cm−1 and 734 cm−1 appear in all spectra which represent to the E2 (low), E2 (high) and A1 (LO) mode peak respectively [20]. The peaks at about 418 cm−1and 748 cm−1 are corresponding to sapphire Raman peaks [21]. It can be noticed from the insets that all the E2 (high) mode peaks in the two samples are obviously larger than 567.6 cm−1. This means that both samples are under compressive stress, just as predicted for GaN epilayers grown on sapphire substrates [19]. In addition, there are two peaks appeared at 535cm−1 and 560 cm−1 respectively, which are corresponding to the A1 (TO) and E1 (TO) mode peaks of GaN. However, these two forbidden Raman mode peaks are absent in the spectrum of sample A. The observation of the A1 (TO) and E1 (TO) modes peaks may be attributed to the slight deviation of the backscattering geometry due to more prominent mosaic structure caused by the patterns in sample B.
Fig. 5 Representative Raman spectra obtained of both samples. (a) Sample A and (b) Sample B. Insets illustrate the magnification of the E2 (high) mode peaks of GaN.
Useful information on spatial distributions of crystalline quality in n-type GaN grown on PSS could be observed through a spatial mapping of the E2 (high) mode. Figures 6(a) and 6(b) show the intensity and the FWHM maps of the E2 (high) mode in a 10 × 10 μm2 zone measured in sample B. The mapping scan data is given in a false-color image, with hues ranging from black for the weakest to yellow for the highest intensity. To extract values for the intensities and FWHM, individual E2 (high) mode is fitted to a Lorentzian-Guassian mixed function. As shown in Fig. 6(a), the distinction in intensity between pattern and period regions is easily distinguished. The blue regions with the lowest intensity should be corresponding to the center parts of the patterns, meanwhile the yellow and red regions with highest intensity refer to the rest part of the patterns. At the same time, the variations in the of E2 (high) peaks exhibit similar trends. The regions corresponding to the center region of patterns has the minimum FWHM. The largest FWHM exists in regions represent for the rest part of the patterns which is relatively larger than that in the planar regions. In fact, the intensity and FWHM variations of E2 (high) mode peak could effectively reflect the crystalline quality of GaN epilayers. The smaller FWHM of E2 (high) indicates better crystalline quality of material with lower dislocation density. Low temperature Raman scattering mapping measurements on both samples were also carried out. Figures 7(a) and 7(b) show the intensity and the FWHM maps of the E2 (high) mode in a 10 × 10 μm2 zone measured at 80k in sample B. Comparing with the mapping results taken at room temperature, the variations of the intensity and the FWHM maps of the E2 (high) mode have the similar trends. The regions corresponding to the center region of patterns has the minimum intensity and FWHM. The difference between the results taken at low temperature and room temperature is that the intensity of E2 (high) mode is larger at low temperature while the FWHM becomes smaller. The variation is totally caused by the influence of temperature [22].
Fig. 6 (a) Raman mapping spectrum of E2 (high) mode intensity of sample B; (b) Raman mapping spectrum of E2 (high) mode FWHM of sample B.
Fig. 7 (a) Low temperature Raman mapping spectrum of E2 (high) mode intensity of sample B; (b) Low temperature Raman mapping spectrum of E2 (high) mode FWHM of sample B.
According to our earlier research [23], the pattern regions of the PSS play the role of SiNx in the traditional ELOG, so GaN grown on the pattern is the ‘wing’ and the GaN grown on the planar region is the ‘window’. It is well known that many dislocations in the window region tend to bending to the side of GaN grain instead of propagating toward the top surface due to the dislocation following a path of minimum elastic energy per unit of growth length of materials [24]. As shown in Fig. 8, during the lateral growth process, the majority of dislocations bend to the patterns in the dislocation annihilation regions. In contrast, very few dislocations bend in the lateral growth regions. There are only a very small amount of dislocations propagating towards the top surface.
Fig. 8 Schematic diagram of dislocation growth mechanism during the growth process of GaN grown on PSS.
It is well known that Transmission electron microscope (TEM) could make accurate estimation of the dislocations distribution in the material. In order to further identify dislocation growth mechanism, the sample grown on PSS was examined by TEM. Figure 9 shows the cross sectional TEM image (with g = 110) of sample B. From Fig. 9, the distribution of dislocations can be clearly observed. Firstly, it can be observed that many dislocations bending towards the pyramid sapphire patterns. This is in agreement of the TEM results illustrated in earlier reported studies [25–27]. The TEM image also shows that voids are generated at the top of the patterns, which is also an evidence of the ELOG process. Meanwhile, only a small amount of dislocations propagated upwards from the top region of the void, which may be caused by the emergence of two adjacent lateral growth regions. It can be also noticed that the regions corresponding to the center part of patterns have lowest dislocation density while the areas corresponding to the dislocation annihilation regions exist more dislocations. The distribution of dislocations is in good agree with our schematic diagram as shown in Fig. 8.
In order to further identify the different spatial distributions of TDs density in two samples, SEM-CL studies have been carried out. Here, TDs appear as dark spots on a light background since the dark spots may be due to the TDs acting as nonradiative recombination centers [28]. Meanwhile, it is expected that the regions with high TD densities, the dark spots will merge and lower the luminescence intensity of the regions [28]. Figure 10 illustrates the plan-view monochromatic 366 nm CL mapping images of the two samples with electron acceleration voltages of 5 kV. From the CL mapping images of 10 × 10 μm2 in dimension, a total of 467 dark spots are counted in Fig. 10(a), indicating a density of 4.7 × 108 cm−2 TDs in sample A. In contrast, there are only 92 dark spots calculated on the surface of n-type GaN grown on PSS from Fig. 10(b), referring to the TD density of 9.2 × 107 cm−2. The low TD density of sample B is comparable with the result of traditional ELOG [23]. Almost 5 times decreasing of the TDs density also confirms the PSS could dramatically improve the crystalline quality of the n-type GaN epilayers. It should also be noticed that the patterns of the PSS could be simulated in Fig. 10(b). The red dash lines are corresponding to the profiles of the patterns, meanwhile the black dash lines refers to the profiles of the lateral growth regions. There are very few dark points in the lateral growth regions while more dislocations are observed in the dislocation annihilation regions. The spatial distributions of dislocations in sample B obtained from the CL characterization could further support the dislocation growth mechanisms during the growth process of GaN grown on PSS we propose according to the Raman mapping results.
Fig. 10 Plan-view monochromatic 366 nm CL mapping images of the two samples with electron acceleration voltages of 5 kV: (a) sample A and (b) sample B.
4. Conclusion
In conclusion, n-type GaN epilayers were grown on both planar sapphire and PSS. AFM images indicate that the surface roughness is improved by using the PSS. Also, the reduction of symmetric (002) and asymmetric (102) FWHMs of HRXRD reveals the decrease of dislocation density in the GaN grown on PSS. The spatial variation in crystalline quality of the n-type GaN grown on PSS is measured by the spatially resolved mapping of micro-Raman scattering. The variations on the intensity and the FWHM of GaN E2 (high) peak indicate that the crystalline quality improvement occurs in lateral growth regions corresponding to the center region of patterns. Bending of dislocations during the lateral growth is proposed to play an important role in the spatial variations of GaN crystalline quality. Cross sectional TEM image and spatial CL mapping results further identify the inhibition of dislocations by using PSS. It also supports the interpretation of the dislocation growth mechanisms during the growth process of GaN grown on PSS.
Acknowledgement
The authors wish to acknowledge support by the National Natural Science Foundation of China (Grant No. 61204006, 61574108), the Fundamental Research Funds for the Central Universities (Grant No. 7214570101).
References and links
1. S. Nakamura, “The Roles of Structural Imperfections in InGaN-Based Blue Light-Emitting Diodes and Laser Diodes,” Science 281(5379), 956–961 (1998). [CrossRef] [PubMed]
2. S. Nakamura, M. Senoh, N. Iwasa, and S. I. Nagahama, “High-brightness InGaN blue, green and yellow light-emitting diodes with quantum well structures,” Jpn. J. Appl. Phys. 34(Part 2, No. 7A), L797–L799 (1995). [CrossRef]
3. U. K. Mishra, P. Parikh, and Yi-Feng Wu, “AlGaN/GaN HEMTs: An overview of device operation and applications,” Proc. IEEE 90(6), 1022–1031 (2002). [CrossRef]
4. T. Wei, Z. Huo, Y. Zhang, H. Zheng, Y. Chen, J. Yang, Q. Hu, R. Duan, J. Wang, Y. Zeng, and J. Li, “Efficiency enhancement of homoepitaxial InGaN/GaN light-emitting diodes on free-standing GaN substrate with double embedded SiO2 photonic crystals,” Opt. Express 22(S4Suppl 4), A1093–A1100 (2014). [CrossRef] [PubMed]
5. T. Jiang, S. R. Xu, J. C. Zhang, Y. Xie, and Y. Hao, “Spatially resolved and orientation dependent Raman mapping of epitaxial lateral overgrowth nonpolar a-plane GaN on r-plane sapphire,” Sci. Rep. 6, 19955 (2016). [CrossRef] [PubMed]
6. S. R. Xu, Y. Hao, L. A. Yang, J. C. Zhang, J. S. Xue, X. Y. Xue, Z. Y. Liu, Z. Y. Lin, J. C. Ma, P. X. Li, J. T. Li, and Q. He, “Improvement of (112) Semipolar GaN Crystal Quality with TiN Interlayer by Metal Organic Vapor Phase Epitaxy,” Jpn. J. Appl. Phys. 50, 115502 (2011).
7. Y. J. Liu, T. Y. Tsai, C. H. Yen, L. Y. Chen, T. H. Tsai, C. C. Huang, T. Y. Chen, C. H. Hsu, and W. C. Liu, “Performance investigation of GaN-based light-emitting diodes with tiny misorientation of sapphire substrates,” Opt. Express 18(3), 2729–2742 (2010). [CrossRef] [PubMed]
8. C.-Y. Cho, J.-B. Lee, S.-J. Lee, S.-H. Han, T.-Y. Park, J. W. Kim, Y. C. Kim, and S.-J. Park, “Improvement of light output power of InGaN/GaN light-emitting diode by lateral epitaxial overgrowth using pyramidal-shaped SiO2.,” Opt. Express 18(2), 1462–1468 (2010). [CrossRef] [PubMed]
9. X. H. Huang, J. P. Liu, J. J. Kong, H. Yang, and H. B. Wang, “High-efficiency InGaN-based LEDs grown on patterned sapphire substrates,” Opt. Express 19(S4Suppl 4), A949–A955 (2011). [CrossRef] [PubMed]
10. J. Y. Cho, J. S. Kim, Y. D. Kim, H. J. Cha, and H. Lee, “Fabrication of oxide-based nano-patterned sapphire substrate to improve the efficiency of GaN-based of LED,” Jpn. J. Appl. Phys. 54(2S), 02BA04 (2015). [CrossRef]
11. Y. B. Tao, T. J. Yu, Z. Y. Yang, D. Ling, Y. Wang, Z. Z. Chen, Z. J. Yang, and G. Y. Zhang, “Evolution and control of dislocations in GaN grown on cone-patterned sapphire substrate by Metal Organic Vapor Phase Epitaxy,” J. Cryst. Growth 315(1), 183–187 (2011). [CrossRef]
12. S. Zhou, H. Wang, Z. Lin, H. Yang, X. Hong, and G. Q. Li, “Study of defects in LED epitaxial layers grown on the optimized hemispherical patterned sapphire substrates,” Jpn. J. Appl. Phys. 53(2), 025503 (2014). [CrossRef]
13. V. C. Su, P. H. Chen, R. M. Lin, M. L. Lee, Y. H. You, C. I. Ho, Y. C. Chen, W. F. Chen, and C. H. Kuan, “Suppressed quantum-confined Stark effect in InGaN-based LEDs with nano-sized patterned sapphire substrates,” Opt. Express 21(24), 30065–30073 (2013). [CrossRef] [PubMed]
14. H. Gao, F. Yan, Y. Zhang, J. Li, Y. Zeng, and G. Wang, “Enhancement of the light output power of InGaN/GaN light-emitting diodes grown on pyramidal patterned sapphire substrates in the micro- and nanoscale,” J. Appl. Phys. 103(1), 014314 (2008). [CrossRef]
15. B. Heying, X. H. Wu, S. Keller, Y. Li, D. Kapolnek, B. P. Keller, S. P. DenBaars, and J. S. Speck, “Role of threading dislocation structure on the x-ray diffraction peak widths in epitaxial GaN films,” Appl. Phys. Lett. 68(5), 643 (1996). [CrossRef]
16. S. R. Xu, Y. Hao, J. C. Zhang, X. W. Zhou, L. A. Yang, J. F. Zhang, H. T. Duan, Z. M. Li, M. Wei, S. G. Hu, Y. R. Cao, Q. W. Zhu, Z. H. Xu, and W. P. Gu, “Improvements in a-plane GaN crystal quality by AlN/AlGaN superlattices layers,” J. Cryst. Growth 311(14), 3622–3625 (2009). [CrossRef]
17. H. Harima, “Properties of GaN and related compounds studied by means of Raman scattering,” J. Phys. Condens. Matter 14(38), R697 (2002). [CrossRef]
18. T. Jiang, S. R. Xu, J. C. Zhang, Z. Y. Lin, R. Y. Jiang, and Y. Hao, “Growth of a-Plane GaN Films on r-Plane Sapphire by Combining Metal Organic Vapor Phase Epitaxy with the Hydride Vapor Phase Epitaxy,” Chin. Phys. Lett. 32(8), 088103 (2015). [CrossRef]
19. K. Hiramatsu, T. Detchprohm, and I. Akasaki, “Relaxation Mechanism of Thermal Stresses in the Heterostructure of GaN Grown on Sapphire by Vapor Phase Epitaxy,” Jpn. J. Appl. Phys. 32(1), 1528–1533 (1993). [CrossRef]
20. J. S. Song, H. Rho, M. S. Jeong, J. W. Ju, and I. H. Lee, “Spatially resolved photoluminescence and Raman mapping of epitaxial GaN laterally overgrown on sapphire,” Phys. Rev. B 81(23), 233304 (2010). [CrossRef]
21. A. Hushur, M. H. Manghnani, and J. Narayan, “Raman studies of GaN/sapphire thin film heterostructures,” J. Appl. Phys. 106(5), 054317 (2009). [CrossRef]
22. A. Link, K. Bitzer, W. Limmer, R. Sauer, C. Kirchner, V. Schwegler, M. Kamp, D. G. Ebling, and K. W. Benz, “Temperature dependence of the E2 and A1(LO) phonons in GaN and AlN,” J. Appl. Phys. 86, 6256 (1999). [CrossRef]
23. S. R. Xu, P. X. Li, J. C. Zhang, T. Jiang, J. J. Ma, Z. Y. Lin, and Y. Hao, “Threading dislocation annihilation in the GaN layer on cone patterned sapphire substrate,” J. Alloys Compd. 614, 360–363 (2014). [CrossRef]
24. H. Klapper, “Röntgentopographische untersuchungen der defektstrukturen im thioharnstoff,” J. Cryst. Growth 15(4), 281–287 (1972). [CrossRef]
25. H. Y. Shin, S. K. Kwon, Y. I. Chang, M. J. Cho, and K. H. Park, “Reducing dislocation density in GaN films using a cone-shaped patterned sapphire substrate,” J. Cryst. Growth 311(17), 4167–4170 (2009). [CrossRef]
26. M. T. Wang, K. Y. Liao, and Y. L. Li, “Growth Mechanism and Strain Variation of GaN Material Grown on Patterned Sapphire Substrates With Various Pattern Designs,” IEEE Photonics Technol. Lett. 23(14), 962 (2011). [CrossRef]
27. J. H. Cheng, Y. S. Wu, W. C. Liao, and B. W. Lin, “Improved crystal quality and performance of GaN-based light-emitting diodes by decreasing the slanted angle of patterned sapphire,” Appl. Phys. Lett. 96(5), 051109 (2010). [CrossRef]
28. M. A. Moram, C. S. Ghedia, D. V. S. Rao, J. S. Barnard, Y. Zhang, M. J. Kappers, and C. J. Humphreys, “On the origin of threading dislocations in GaN films,” J. Appl. Phys. 106(7), 073513 (2009). [CrossRef]
