Indium-rich InGaN epitaxial layers with a p-i-n structure were grown pseudomorphically on a strain-relaxed InGaN template to reduce structural strain induced by lattice mismatch. We applied a nano-sculpting process to improve the crystal quality of the strain-relaxed InGaN template. The results show that the nano-sculpting process can suppress effectively the threading dislocation generation and improves significantly the I-V characteristic of the InGaN p-i-n structure. This InGaN template technique with nano-sculpting process shows great potential for future applications in indium-rich InGaN optic-electron devices.
© 2012 OSA
InGaN alloys have attracted considerable attention since the band gap value of InN was revised to 0.7 eV by Wu et al. . This alloy system has the variable direct band gap ranging from 0.7 to 3.4 eV, which covers full spectrum of visible solar and thus makes it a promising semiconductor material for applications in optoelectronic devices such as the full ultra violet A to near visible detectors , solar cells [3,4], light emitting diodes and lasers [5–7]. However, several barriers still exist in obtaining indium-rich InGaN alloys in order to realize green-emitting semiconductor lasers and other indium-rich InGaN optoelectronic devices. The main bottleneck is the lack of a suitable lattice-matched substrate for growing InGaN films. It is a common practice to grow InGaN films on thick GaN layers. However, the use of these conventional templates is inherently problematic because of the large lattice mismatch between GaN and the target InGaN films, which leads to high strain and unacceptable levels of performance-degrading material defects. Recently, some novel substrates for growing InGaN alloys have been proposed. For example, InN templates and ZnO substrates have been applied to grow indium-rich InGaN layers and lattice-matched In0.25Ga0.75N epilayer, respectively [8,9]. In spite of considerable effort in the past, there are a number of remaining barrier for applications of indium-rich InGaN materials and devices.
In fact, InGaN templates are also excellent candidates as lattice-matched substrates for indium-rich InGaN epilayers. In this study, we fabricated an indium-rich InGaN p-i-n structure on a fully strain-relaxed InGaN template by molecular beam epitaxy (MBE) in conjunction with a nano-sculpting process using glancing angle ion flux. It is important for the template to be relaxed because its key function is to provide a new lattice parameter for growing indium-rich InGaN alloys. The structural characteristics of the InGaN alloys including the template and epilayers were investigated using high-resolution X-ray diffraction (HR-XRD) methods and transmission electron microscopy (TEM) techniques.
The InGaN-based structures, as shown schematically in Fig. 1 , were grown by plasma assisted MBE on c-plane sapphire. The group III fluxes including high-purity elemental In and Ga, as well as fluxes of Si and Mg (n-type and p-type dopants) were supplied by effusion cells. Active nitrogen was produced by an RF plasma source from SVT Associates, using a nitrogen flow rate of about 3 SCCM that resulted in an N-limited growth rate of about 0.8 micron/hour. The growth was started by nitridation of the sapphire substrate, followed by growth of a thin AlN nucleation layer at about 800 °C. The growth temperature was reduced to about 700 °C for GaN and to about 500 °C for an 800-nm-thick InGaN template layer and a p-i-n structure. During the initial growth of InGaN template, a glancing angle ion flux treatment was applied concurrently in order to form nanoscale surface corrugation. When these corrugated structures grew to a bigger size, a smoothing process was then applied by slowly rotating the substrate without growth and smoothing the surface corrugated structures to a relative flat surface with a second ion flux treatment. Thus, the nanoscale structures were kept to a small size during the initial growth of the template, and finally, a nano-sculpted transition region was formed by repeating the nano-sculpting and smoothing processes. This transition region was grown to a size of about 50 nm. Further details about the nano-sculpting technique are available in the patent documents . Growth process was monitored in situ by reflection high energy electron diffraction and by a combination of emissivity-corrected pyrometry and two-color reflectometry from SVT Associates. A similar indium-rich InGaN p-i-n structure grown on an InGaN template without the nano-sculpted transition region was also prepared to confirm the effect of the nano-sculpting process. For I-V measurements, the n-InGaN ohmic contact was formed by depositing Ti/Al/Ni/Au (20/120/30/80 nm) layers using e-beam evaporation after etching down to the n-InGaN layer, and p-InGaN ohmic contact was fabricated by Ni/Au (30/80 nm) with a follow-up annealing treatment at 550 °C for 3 min in a flowing O2 atmosphere. Structural properties of the sample were characterized by XRD including XRD ω/2θ scan and reciprocal space mapping (RSM). The lattice constant, strain components and composition of each layer of InGaN alloy films were determined by RSM. Crystalline defects such as dislocations were studied by means of TEM.
3. Results and discussion
The XRD ω/2θ scan of symmetric (0002) reflection for the InGaN p-i-n structure with nano-sculpting process is shown in Fig. 2 , from which the out-of-plane structural information can been obtained. The scan shows the GaN peak at a diffraction angle of 17.28°and a distinct shoulder on the low angle side of the GaN peak, from the p-GaN top layer. The broad diffraction peak around 16.92° is from the p-InGaN according to the sample structure. Three other diffraction peaks in the direction of low diffraction angle were recorded from the different InGaN layers with high In contents.
In order to determine the in-plane and out-of-plane lattice constants (a and c, respectively), we recorded the asymmetric (105) RSM of the InGaN sample as shown in Fig. 3 . In this map, there are four reciprocal lattice points (RLP). The top RLP corresponds to the contribution of the GaN layers at both top and bottom of the InGaN sample. The other three RLPs originate from the InGaN layers. The specific lattice constants (a, c) of InGaN with wurtzite structure can be given by 12]. In our calculation, we used the lattice parameters a0 = 0.31892 nm and c0 = 0.51850 nm for GaN . and a0 = 0.35378 nm and c0 = 0.57033 nm for InN , and the elastic constants c13 = 103 GPa and c33 = 405 GPa for GaN, and c13 = 92 GPa and c33 = 224 GPa for InN .
According to the calculation results, the InGaN (3) on the bottom of RSM patterns with the highest In composition of x = 0.69 is from the contribution of the i-InGaN layer as expected. The InGaN (1) and (2) exhibit the same in-plane lattice parameter but different In compositions. It is difficult to assign directly RLPs of InGaN (1) and (2) to the template and the n-InGaN layer, respectively, according to the results from the RSM. The selected area diffraction (SAD) was performed on the template and n-InGaN layers in order to distinguish their diffraction contribution in the map of RSM. The SAD patterns from the template and n-InGaN layer are shown in Fig. 4 . The in-plane lattice constant calculated from SAD spots is 0.3392 and 0.3346 nm for the n-InGaN layer and the template, respectively. Although a small discrepancy exists on lattice parameter taken from SAD and XRD, where the calculated in-plane lattice constant is 0.3407 and 0.3397 nm for the n-InGaN layer and the template respectively, it does not affect our judgment about the assignment of the InGaN layers. According to the consistency between XRD and SAD, the InGaN (1) with a smaller lattice constant in the RSM pattern is from the template, and the InGaN (2) corresponds to the n-InGaN layer.
The calculated degree of relaxation reveals that the InGaN template with a calculated In content of 59% is fully relaxed with respect to the underlying GaN layer. It can be seen from Fig. 2 and Fig. 3 that the template layer exhibits a good crystal quality that is comparable to the underlying GaN film in terms of the shape of (105) RSM, the full width at half maximum and peak intensity of the (0002) reflection. The nano-sculpted transition region formed by the glancing angle ion flux processing plays a key role in improving the crystal quality of the InGaN template. This growth method is surprisingly found to be able to effectively suppress dislocation growing, as shown in the TEM image of Fig. 5(a) . Threading dislocations in the
InGaN template mainly originate from the vertical penetration of dislocations through the interface of InGaN/GaN, and are partially annihilated in the InGaN template. We observe few new threading dislocations induced by the interface mismatch, implying that the nano-sculpted structure can serve as end points for the threading dislocations by providing a number of nano-sculpted interfaces. Colby et al.  gave a reasonable explanation about the mechanism of dislocation filtering in GaN nanostructures, where they thought that threading dislocations are likely to be excluded by the strong image forces of the nearby free surfaces with nanostructure, and a greater than 2 orders of magnitude reduction in threading dislocation density in GaN nanostructures was verified by TEM analysis. In contrast, we also provided the cross-sectional TEM image of another indium-rich InGaN template without using this process, as shown in Fig. 5(b). Obviously, a number of new threading dislocations generated at the interface of GaN/InGaN owing to the large lattice mismatch can be observed. Meanwhile, the symmetric (0002) reflection of the indium-rich InGaN p-i-n structure grown on this template is also provided in Fig. 6 , where the i- and n-InGaN layers are shown to grow coherently on the InGaN template and their indium content is determined to be about 50%. The InGaN layers exhibit lower relative diffraction intensity to the underlying GaN layer, when compared with the nano-sculpted template, as shown in Fig. 2. This indicates that the nano-sculpted InGaN template has better crystal quality, which is in good agreement with the result of TEM.
The n-InGaN layer with a calculated In content of 62% has mostly identical in-plane lattice constant with the nano-sculpted template, indicating that the n-InGaN layer was grown pseudomorphically on the template, which can be observed intuitively in the RSM patterns. The RLP of the n-InGaN layer broadens obviously along both the qx and qy axes in comparison with that of the template and shows degradation in the crystal quality due to a higher In content incorporation and Si-doping, but generally, the n-InGaN layer maintains the good crystal quality of the template in terms of their line width and peak intensity of the (0002) reflection.
The 180-nm-thick i-InGaN layer exhibits a partial relaxation of about 59% with respect to the template according to the calculation results of the RSM. This partial strain relaxation should be mainly attributed to the lattice mismatch between the i-InGaN layer and the underlying InGaN structure, and results in a further degradation in the crystal quality. However, phase separation in such an indium-rich and thick InGaN film is not observed, which occurs usually in indium-rich or strained InGaN alloys [17,18]. The p-InGaN layer with a designed In composition of 15% and Mg doping concentration of 5 × 1017 has poor crystal quality verified by the broad (0002) diffraction peak with a low intensity in Fig. 2 and an absence of the asymmetric (105) diffraction in Fig. 3. The poor crystal quality is attributed to the large lattice mismatch between the p-InGaN layer and the underlying indium-rich InGaN structure. Another important cause for poor crystal quality is associated with Mg doping. As Mg doping is known to degrade the crystal quality and facilitate phase separation [19,20]. Although the the p-InGaN layer has poor crystal quality, the InGaN structure with the nano-sculpting process still exhibits an obvious pn-junction behavior, as shown in Fig. 7 , and has a smaller reverse leakage current compared to the similar structure without the nano-sculpting process. Further investigations on device performances are limited owing to the poor crystal quality of the p-InGaN layer.
In summary, we grew an indium-rich InGaN epitaxial structure on a fully strain-relaxed InGaN template. To reduce the threading dislocations in the InGaN template induced by the large lattice mismatch between GaN and InGaN, we employed a nano-sculpting process during the growth of the InGaN template. The results show that the nano-sculpting process can filter effectively threading dislocations in the InGaN template and improve significantly the crystal quality and the I-V characteristic of the indium-rich InGaN p-i-n structure grown on the nano-sculpted template when compared to a similar InGaN p-i-n structure grown on an InGaN template without using this nano-sculpting process.
This work was supported by the National 973 project, China (2012CB619306, 2011CB301900, 2009CB320300), the Natural Science Foundation of Jiangsu Province (BK2010045), NSFC (60936004, 60990311), and Ph.D. Programs Foundation of Ministry of Education of China (20110091110032).
References and links
1. J. Wu, W. Walukiewicz, K. M. Yu, J. W. Ager, E. E. Haller, H. Lu, W. J. Schaff, Y. Saito, and Y. Nanishi, “Unusual properties of the fundamental band gap of InN,” Appl. Phys. Lett. 80(21), 3967–3969 (2002). [CrossRef]
2. E. Muñoz, “(Al,ln,Ga)N-based photodetectors. some materials issues,” Phys. Stat. Solidi B 244(8), 2859–2877 (2007). [CrossRef]
3. R. Dahal, B. Pantha, J. Li, J. Y. Lin, and H. X. Jiang, “InGaN/GaN multiple quantum well solar cells with long operating wavelengths,” Appl. Phys. Lett. 94(6), 063505 (2009). [CrossRef]
4. J. J. Xue, D. J. Chen, B. Liu, Z. L. Xie, R. L. Jiang, R. Zhang, and Y. D. Zheng, “Au/Pt/InGaN/GaN heterostructure Schottky prototype solar cell,” Chin. Phys. Lett. 26(9), 098102 (2009). [CrossRef]
5. S. Nakamura, M. Senoh, S. Nagahama, N. Iwasa, T. Yamada, T. Matsushita, H. Kiyoku, and Y. Sugimoto, “InGaN-based multi-quantum-well-structure laser diodes,” Jpn. J. Appl. Phys. 35(Part 2, No. 1B), L74–L76 (1996). [CrossRef]
6. S. Nakamura, T. Mukai, and M. Senoh, “Candela-class high-brightness InGaN/AlGaN double-heterostructure blue-light-emitting diodes,” Appl. Phys. Lett. 64(13), 1687–1689 (1994). [CrossRef]
7. Y. Narukawa, Y. Kawakami, M. Funato, S. Fujita, S. Fujita, and S. Nakamura, “Role of self-formed InGaN quantum dots for exciton localization in the purple laser diode emitting at 420 nm,” Appl. Phys. Lett. 70(8), 981–983 (1997). [CrossRef]
8. H. Naoi, M. Kurouchi, S. Takado, D. Muto, T. Araki, and Y. Nanishi, “Structural and luminescence properties of In-rich InGaN layers grown on InN templates by RF-MBE,” Phys. Status Solidi A 202(14), 2642–2647 (2005). [CrossRef]
9. A. Kobayashi, J. Ohta, and H. Fujioka, “Low temperature epitaxial growth of In0.25Ga0.75N on lattice-matched ZnO by pulsed laser deposition,” J. Appl. Phys. 99(12), 123513 (2006). [CrossRef]
10. P. I. Cohen and B. Cui, US patent 20100090311A1 (2010).
11. S. Pereira, M. R. Correia, E. Pereira, K. P. O’Donnell, E. Alves, A. D. Sequeira, N. Franco, I. M. Watson, and C. J. Deatcher, “Strain and composition distributions in Wurtzite InGaN/GaN layers extracted from x-ray reciprocal space mapping,” Appl. Phys. Lett. 80(21), 3913–3915 (2002). [CrossRef]
12. M. Schuster, P. O. Gervais, B. Jobst, W. Hosler, R. Averbeck, H. Riechert, A. Iberl, and R. Stommer, “Determination of the chemical composition of distorted InGaN GaN heterostructures from x-ray diffraction data,” J. Phys. D Appl. Phys. 32(10A), A56–A60 (1999). [CrossRef]
13. T. Detchprohm, K. Hiramatsu, K. Itoh, and I. Akasaki, “Relaxation process of the thermal strain in the GaN/alpha-Al2O3 heterostructure and determination of the intrinsic lattice constants of GaN free from the strain,” Jpn. J. Appl. Phys. 31(Part 2, No. 10B), L1454–L1456 (1992).
14. W. Paszkowicz, “X-ray powder diffraction data for indium nitride,” Powder Diffr. 14, 258–260 (1999).
15. A. F. Wright, “Elastic properties of zinc-blende and wurtzite AlN, GaN, and InN,” J. Appl. Phys. 82(6), 2833–2839 (1997). [CrossRef]
16. R. Colby, Z. W. Liang, I. H. Wildeson, D. A. Ewoldt, T. D. Sands, R. E. García, and E. A. Stach, “Dislocation filtering in GaN nanostructures,” Nano Lett. 10(5), 1568–1573 (2010). [CrossRef] [PubMed]
17. C. Tessarek, S. Figge, T. Aschenbrenner, S. Bley, A. Rosenauer, M. Seyfried, J. Kalden, K. Sebald, J. Gutowski, and D. Hommel, “Strong phase separation of strained In(x)Ga(1-x)N layers due to spinodal and binodal decomposition: formation of stable quantum dots,” Phys. Rev. B 83(11), 115316 (2011). [CrossRef]
18. B. N. Pantha, J. Li, J. Y. Lin, and H. X. Jiang, “Evolution of phase separation in In-rich InGaN alloys,” Appl. Phys. Lett. 96(23), 232105 (2010). [CrossRef]
19. Y. Huang, A. Melton, B. Jampana, M. Jamil, J. H. Ryou, R. D. Dupuis, and I. T. Ferguson, “Compositional instability in strained InGaN epitaxial layers induced by kinetic effects,” J. Appl. Phys. 110(6), 064908 (2011). [CrossRef]
20. L. Sang, M. Takeguchi, W. Lee, Y. Nakayama, M. Lozach, T. Sekiguchi, and M. Sumiya, “Phase separation resulting from Mg doping in p-InGaN film grown on GaN/Sapphire template,” Appl. Phys. Express 3(11), 111004 (2010). [CrossRef]