Open Access
Add to BibSonomy
Abstract
InGaN/GaN multiple quantum wells (MQWs) were fabricated under the same growth conditions on the planar and patterned sapphire substrates (PSS) with 10 nm and 20 nm sputtering AlN layers, respectively. Photoluminescence and electroluminescence results both showed that MQWs samples have significant differences in emission wavelengths. The wavelength of the samples on planar substrate is about 20 nm longer than that on the PSS. For samples with the same substrate, but different AlN layer thickness, also exhibit a small wavelength shift. High-resolution X-ray diffraction also revealed that the periodic thickness of MQWs on a planar substrate is thicker than that on PSS. Thermodynamic simulation was carried out to verify the effect of PSS on the heat conduction of GaN film. The PSS embedded in GaN film will affect its heat dissipation ability, and thus influence the wavelength of the MQWs samples by affecting the growth temperature.
© 2020 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Recently, GaN and other nitride semiconductors have attracted considerable attention and developed rapidly, especially in solid-state lighting [1–6]. The first GaN-based high-power blue light-emitting diode (LED) was fabricated on a sapphire substrate [7]. Currently, bulk GaN substrates are small and expensive, which are not suitable for large-scale commercial LED production. Therefore, hetero-substrates such as sapphire and silicon have become mainstream, in which sapphire substrates are most widely adopted. However, there are large lattice and thermal expansion mismatches between GaN and substrate in heteroepitaxy, which results in strong stress and high defect density of GaN film. Defects like dislocations act as non-radiative recombination centers and charge scattering centers, deteriorating the performance of LEDs. In order to improve the crystal quality of GaN films and the light extraction efficiency of LEDs, different patterned sapphire substrates (PSS) have been proposed, such as striped [8], conical [9] and hemispherical [10] patterns. The mechanism of reducing dislocation density in GaN films by PSS has been deeply studied [11–13]. However, the luminous efficiency of LEDs with longer wavelength such as green or yellow is still low [14–19]. Long wavelength means higher In composition in multiple quantum wells (MQWs). Higher In composition makes strain-induced piezoelectric polarization effect stronger in MQWs, and the quantum-confined Stark effect (QCSE) [20,21] is more serious, which limits the performance of green LEDs. In addition, the combination of In atoms is very sensitive to the growth temperature. The use of PSS results in parts of sapphire with low thermal conductivity embedded in GaN layer, which could change the thermal distribution near MQWs. Moreover, there is no discussion on the effect of PSS on the In combination in MQWs and the emission wavelength of LEDs.
In this paper, InGaN/GaN MQWs on planar sapphire substrate and PSS are prepared with the sputtered AlN layer thickness of 10 nm and 20 nm, respectively. By characterizing the crystal quality and luminous performance of the samples, the influence of the PSS on the growth of MQWs and the emission wavelength were systematically analyzed and discussed.
2. Experiment
Green InGaN/GaN MQWs structures were grown on the planar sapphire substrate and PSS by metal organic chemical vapor deposition (MOCVD) using Veeco K465i. The growth conditions of all samples were exactly the same except the substrate type and sputtered AlN layer thickness. Trimethylaluminum, trimethylgallium, trimethylindium and ammonia were used as precursors for Al, Ga, In and N, respectively. Biscyclopentadienylmagnesium and silane were used as p- and n-type dopant sources, respectively. The structure consisted of a low-temperature GaN buffer layer, a 1.5 µm unintentionally doped GaN layer, a 2.2 µm n-GaN layer, a InGaN/GaN superlattice structure to adjust stress, 8 periods of green InGaN/GaN MQWs and the thickness of 16 nm in one period, a low-temperature p-GaN to protect MQWs, a 4-cycle p-GaN/AlGaN superlattice electron blocking layer (EBL), and a 150 nm p-GaN layer. 4 samples are prepared, the samples with 10 nm and 20 nm sputtered AlN on planar substrate were recorded as sample A and B, and those on PSS were recorded as sample C and D, respectively.
3. Results and discussions
3.1 Morphology and crystal quality tests
The surface morphologies of samples were obtained by atomic force microscope (AFM) measured by Bruke Fastscan Icon in tapping mode. The root-mean-square roughness of samples A, B, C and D are 0.649, 0.773, 0.837 and 0.538 nm, respectively. Those results indicate that all samples exhibit smooth surface, especially for sample D. As shown in Fig. 1, there are protrusions on the sample surface, and the density of raised spots in planar substrate samples (Figs. 1(a)–1(b)) is higher than that in PSS samples (Figs. 1(c)–1(d)). The protrusions are associated with the accumulation of In atoms, which means that planar substrate samples have more In clusters in MQWs, and the In component distribution in PSS samples are more uniform [22]. Figure 2 shows the cathodoluminescence (CL) mapping images of samples. CL measurement was conducted by MonoCL 3+ operating at 20 kV. The texts of samples A and B were carried out under 555 nm fixed grating, while samples C and D with 550 nm grating. According to the CL images, the luminescence of PSS samples is more uniform than that of planar substrate samples. In particular, sample D has the most uniform luminescence, indicating the best quality of MQWs.
The crystalline quality of green MQWs structures is characterized by the high-resolution X-ray diffraction (HRXRD). Figures 3(a) and 3(b) show the HRXRD rocking curve (RC) for the samples of (002) and (102) reflection planes, respectively. The full-width at half-maximum (FWHM) values of the samples are listed in Table 1. The largest FWHM of (002) and (102) planes appears in sample C on PSS, indicating the worst crystal quality among all the samples. This is attributed to the inappropriate thickness of 10 nm sputtered AlN layer for GaN film growth, which deteriorates the crystal quality of sample C, even worse than the samples on planar substrate [23]. In addition, HRXRD 2θ-ω scan was performed to characterize the quality of MQWs structure. Figures 3(c)–3(f) show the 2θ-ω result of samples A, B, C, and D. The dominant peak in the 2θ-ω image corresponds to the GaN layer, while the satellite peaks correspond to the InGaN/GaN MQWs. As can be clearly seen in Fig. 3, each sample has negative 4 and positive 2 orders of satellite peaks, which indicates that the entire MQWs have a good layer periodicity as well as interfaces between wells and barriers. In addition, the satellite peak intensities of samples C and D are stronger than those of samples A and B, revealing that the periodicity of MQWs grown on PSS is better than that on the planar substrates. This is due to the higher In content and the more severe In aggregation in the MQWs on the planar substrates, leading to worse interface and layer periodicity. Although sample C has the worst crystal quality, its satellite peak intensity is still higher than that of the planar substrates, because of the better periodicity of MQWs with low In content.
Fig. 3. Normalized HRXRD RC for samples of (a) (002) reflection planes, (b) (102) reflection planes; HRXRD 2θ-ω scan patterns of (c) sample A, (d) sample B, (e) sample C, and (f) sample D.
Table 1. FWHM data of rocking curve for samples A, B, C and D
Using the angular spacing of the satellite peaks, the calculated thickness of MQWs in one cycle (a well plus a barrier) is as followed. The computational formula of thickness [24]:
where, D is the thickness, λ is the wavelength of x-ray in XRD test, Δθ is the angular spacing of the satellite peaks, and the θB is the Bragg diffraction angle of GaN layer. The thickness of single quantum well and barrier (QWB) is 16.5 nm for planar substrate samples (samples A and B), and 15.9 nm for PSS samples (samples C and D), which reflects the different In component in MQWs. Because the In atom is larger than Ga, the MQWs of the planar samples has a higher composition of In.The cross-section transmission electron microscope (TEM) images taken under 300 kV by Themis-Z are explored in Figs. 4(a) and 4(b) for samples B and D, respectively. The 8-cycle MQWs and 4-cycle superlattice EBL structures can be clearly seen. According to the result of TEM, the thickness of quantum well period of samples B and D are 16.5 and 16.0 nm, respectively, which is consistent with the results calculated from the HRXRD 2θ-ω scan data. Therefore, the thickness difference of quantum wells is very small.
3.2 Optical and electrical tests
By means of the above morphology and crystal quality tests, it can be determined that high-quality InGaN/GaN MQWs structures are realized. Then, optical and electrical tests are performed as shown in Fig. 5. It is obvious that the luminescence intensity of MQWs on PSS is stronger than that of samples on planar substrates. Sample D (with 20 nm AlN layer on PSS) has particular advantages in photoluminescence (PL) intensity, which demonstrates a better quality than sample C and is consistent with the results of HRXRD. In addition, the PL peaks of PSS samples are 532 nm, while that of planar substrate samples are 559 nm. There exhibits a large wavelength shift between PSS samples and planar substrate samples. Moreover, wavelength shift also appears in the electroluminescence (EL) test. At 100 mA, the wavelength of sample A, B, C, and D is 548, 552, 526, and 530 nm, respectively. Obviously, the wavelength of the planar substrate samples is longer than that of the PSS samples at 100 mA.
Fig. 5. (a) PL spectrum, (b) voltage vs current density (c) LOP vs current density, and (d) normalized luminous efficiency vs current density for 4 samples.
Generally, the luminous wavelength of LEDs is mainly determined by the In component in MQWs, and the growth temperature has a great influence on the combination and decomposition of In atoms [25–27]. The thermal conductivity of GaN and sapphire are 190 and 30 W/(m·K), respectively [28]. It could be believed that the low thermal conductivity of sapphire affects the temperature control during epitaxial growth of MQWs structures.
Samples C and D have a cone-shaped PSS, whose bottom diameter, interval spacing, and height of the cones were 2.7, 0.2 and 1.5 µm, respectively. The total thickness of GaN between the sapphire substrate plane and MQWs was 4.5 µm, so the sapphire cones embedded in GaN had a large relative volume. We believe that a large amount of sapphire in GaN layer will reduce its thermal conductivity [29]. The growth temperature of GaN QB is 870°C, while the growth temperature of QW is 730°C. Because the boundary thermal resistance between GaN and gas is very large, the heat in the GaN film is mainly dissipated downward through sapphire substrate [28]. Therefore, the change of thermal conductivity of GaN layer will affect the actual growth temperature of MQWs. Since a large amount of sapphire with low thermal conductivity embedded in the GaN layer, as well as the lower thermal conductivity and the worse heat dissipation capability of the GaN layer in PSS samples compared with planar substrate samples, the actual growth temperature of QWs of the PSS samples is higher than that of the planar substrate samples. Furthermore, the lower temperature is easier for combination of In atoms, thus the QWs of planar substrate samples will have more In component than PSS samples. Therefore, the planar substrates samples show a longer emission wavelength in PL and EL tests than the PSS samples.
As shown in Fig. 5(b), due to the lower effective barrier between QWs of PSS samples, PSS samples exhibit lower working voltage than planar substrate samples under the same injection state. Low effective barrier improves carrier transport in MQWs, which results in low series resistance and forward voltage [30]. Fig. 5(c) shows that the light output power (LOP) of PSS samples is almost 30 times higher than that of planar substrate samples at 100 mA. Due to the advantages of working voltage and luminous intensity, the better luminous efficiency in PSS samples could be achieved compared with planar substrate samples. The efficiency of MQWs on PSS is an order of magnitude higher than that on planar substrates, which can be obtained from the normalized efficiency curves in Fig. 5(d). The longer wavelength means a higher In component in MQWs, which causes more serious polarization effect and polarization-induced QCSE. This sharply deteriorates the luminous performance of the LED when the wavelength becomes longer [20–21]. Therefore, the luminous performance of the planar substrate samples is so much worse than that of the PSS samples. In addition, PSS can improve light extraction efficiency, which leads to the increase in luminescence of PSS samples. For both PSS samples, although the EL wavelength of sample D is 4 nm longer than that of sample C, there is no significant difference in LOP and efficiency between them. This is because sample D has higher crystal quality compared with sample C, which improves the luminous performance of sample D.
3.3 Theoretical simulation
In order to further clarify the effect of PSS on heat transfer in GaN film, theoretical simulation was carried out. The finite element method was used to simulate the transient thermal conduction of GaN film. The simulation is based on the thermal conductivity, density and specific heat capacity of GaN and sapphire, as shown in Table 2. Because the layer structure obtained by epitaxy is isotropic in the horizontal direction and the heat transfer mainly occurs in the vertical direction, a 2-dimensional model for simulation was used. The geometric model in the simulation is shown in Figs. 6(a) and 6(b), the size of cone-shaped PSS and the thickness of GaN layer are the same as the actual samples, but only 6 patterns are taken in horizontal direction to simplify the calculation. As shown in Figs. 6(a) and 6(b), the heat dissipation modes of GaN film have surface thermal radiation and downward thermal conduction. Figure 6(c) shows the temperature-time curves of the surfaces of the two models. When the temperature of the planar substrate model drops from 870°C to 730°C, the temperature of PSS model is 738°C, 8°C higher than the former. The simulation results show that sapphire embedded in GaN does affect the thermal conductivity of GaN film, and the actual temperature of PSS samples after cooling under the same conditions is higher than that of planar substrate samples, which is consistent with previous speculation. It is well known that InGaN is very sensitive to the growth temperature. Different growth temperatures lead to different In compositions in samples on different substrates, thus having different emission wavelengths.
Fig. 6. GaN film heat dissipation model of (a) planar substrate and (b) PSS, (c) temperature curves of GaN surface over time.
Table 2. Heat transfer parameters of GaN and sapphire
4. Conclusion
Green InGaN/GaN MQWs samples with sputtering AlN layers of different thicknesses were fabricated on planar substrate and PSS. AFM and CL tests show that the surface and luminescence of PSS samples are more uniform than that of planar substrates. Larger wavelength difference occurs between different substrate samples. PL and EL emission wavelengths of planar substrate samples are about 20 nm longer than those of PSS samples. The large wavelength change could be considered as the result of the embedded sapphire of PSS in GaN layer, which weakens the thermal conductivity of GaN film. This results in the higher actual growth temperature of QWs on PSS than that on planar substrate. Therefore, the In component in the PSS samples is lower than that in the planar ones, causing the wavelength change. The effect of PSS on the thermal conductivity and temperature of GaN film is also verified by simulations.
Funding
Shanxi Provincial Key Research and Development Project (2018ZDCXL-GY-01-02-02); National Key Research and Development Program of China (2016YFB0400801, 2018YFB0406601); National Natural Science Foundation of China (61574108).
Disclosures
The authors declare no conflicts of interest.
References
1. A. Laubsch, M. Sabathil, J. Baur, M. Peter, and B. Hahn, “High-power and high-efficiency ingan-based light emitters,” IEEE Trans. Electron Devices 57(1), 79–87 (2010). [CrossRef]
2. H. Morkoc and S. N. Mohammad, “High-luminosity blue and bluegreen gallium nitride light-emitting diodes,” Science 267(5194), 51–55 (1995). [CrossRef]
3. H. Kim, S. J. Park, and H. Hwang, “Design and fabrication of highly efficient GaN-based light-emitting diodes,” IEEE Trans. Electron Devices 49(10), 1715–1722 (2002). [CrossRef]
4. T. Mukai, M. Yamada, and S. Nakamura, “Characteristics of InGaN-Based UV/Blue/Green/Amber/Red Light-Emitting Diodes,” Jpn. J. Appl. Phys. 38(Part 1, No. 7A), 3976–3981 (1999). [CrossRef]
5. E. F. Schubert and J. K. Kim, “Solid-State Light Sources Getting Smart,” Science 308(5726), 1274–1278 (2005). [CrossRef]
6. M. R. Krames, O. B. Shchekin, R. Mueller-Mach, G. O. Mueller, L. Zhou, G. Harbers, and M. G. Craford, “Status and future of high-power light-emitting diodes for solid-state lighting,” J. Disp. Technol. 3(2), 160–175 (2007). [CrossRef]
7. S. Nakamura, T. Mukai, and M. Senoh, “High-Power GaN P-N Junction Blue-Light-Emitting Diodes,” Jpn. J. Appl. Phys. 30(Part 2, No. 12A), L1998–L2001 (1991). [CrossRef]
8. H. G. Chen, T. S. Ko, S. C. Ling, T. C. Lu, H. C. Kuo, S. C. Wang, Y. H. Wu, and L. Chang, “Dislocation reduction in GaN grown on stripe patterned r-plane sapphire substrates,” Appl. Phys. Lett. 91(2), 021914 (2007). [CrossRef]
9. 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]
10. C. T. Chang, S. K. Hsiao, E. Y. Chang, Y. L. Hsiao, J. C. Huang, C. Y. Lu, H. C. Chang, K. W. Cheng, and C. T. Lee, “460-nm InGaN-Based LEDs Grown on Fully Inclined Hemisphere-Shape-Patterned Sapphire Substrate With Submicrometer Spacing,” IEEE Photonics Technol. Lett. 21(19), 1366–1368 (2009). [CrossRef]
11. F. W. Lee, W. C. Ke, C. H. Cheng, B. W. Liao, and W. K. Chen, “Influence of different aspect ratios on the structural and electrical properties of GaN thin films grown on nanoscale-patterned sapphire substrates,” Appl. Surf. Sci. 375, 223–229 (2016). [CrossRef]
12. R. S. Peng, X. J. Meng, S. R. Xu, J. C. Zhang, P. X. Li, J. Huang, J. J. Du, Y. Zhao, X. M. Fan, and Y. Hao, “Study on Dislocation Annihilation Mechanism of the High-Quality GaN Grown on Sputtered AlN/PSS and Its Application in Green Light-Emitting Diodes,” IEEE Trans. Electron Devices 66(5), 2243–2248 (2019). [CrossRef]
13. Q. C. Nie, Z. M. Jiang, Z. Y. Gan, S. Liu, H. Yan, and H. S. Fang, “Defect analysis of the LED structure deposited on the sapphire substrate,” J. Cryst. Growth 488, 1–7 (2018). [CrossRef]
14. S. D. Lester, M. J. Ludowise, K. P. Killeen, B. H. Perez, J. N. Miller, and S. J. Rosner, “High-efficiency InGaN MQW blue and green LEDs,” J. Cryst. Growth 189-190(6), 786–789 (1998). [CrossRef]
15. S. Saito, R. Hashimoto, J. Hwang, and S. Nunoue, “InGaN Light-Emitting Diodes on c-Face Sapphire Substrates in Green Gap Spectral Range,” Appl. Phys. Express 6(11), 111004 (2013). [CrossRef]
16. P. Li, Y. Zhao, H. Li, Z. Li, Y. Zhang, J. Kang, M. Liang, Z. Liu, X. Yi, and G. Wang, “Highly efficient InGaN green mini-size flip-chip light-emitting diodes with AIGaN insertion layer,” Nanotechnology 30(9), 095203 (2019). [CrossRef]
17. H. P. Zhao, G. Y. Liu, J. Zhang, J. D. Poplawsky, V. Dierolf, and N. Tansul, “Approaches for high internal quantum efficiency green InGaN light-emitting diodes with large overlap quantum wells,” Opt. Express 19(S4), A991–A1007 (2011). [CrossRef]
18. R. Sharma, P. M. Pattison, H. Masui, R. M. Farrell, T. J. Baker, B. A. Haskell, F. Wu, S. P. DenBaars, J. S. Speck, and S. Nakamura, “Demonstration of a semipolar (2005) InGaN∕GaN green light emitting diode,” Appl. Phys. Lett. 87(23), 231110 (2005). [CrossRef]
19. D. M. Yeh, C. F. Huang, C. Y. Chen, Y. C. Lu, and C. C. Yang, “Localized surface plasmon-induced emission enhancement of a green light-emitting diode,” Nanotechnology 19(34), 345201 (2008). [CrossRef]
20. T. Takeuchi, C. Wetzel, S. Yamaguchi, H. Sakai, H. Amano, and I. Akasaki, “Determination of piezoelectric fields in strained GaInN quantum wells using the quantum-confined Stark effect,” Appl. Phys. Lett. 73(12), 1691–1693 (1998). [CrossRef]
21. S. F. Chichibu, A. C. Abare, M. S. Minsky, S. Keller, S. B. Fleischer, J. E. Bowers, E. Hu, U. K. Mishra, L. A. Coldren, and S. P. DenBaarsb, “Effective band gap inhomogeneity and piezoelectric field in InGaN/GaN multiquantum well structures,” Appl. Phys. Lett. 73(14), 2006–2008 (1998). [CrossRef]
22. S. T. Liu, X. Q. Wang, G. Chen, Y. W. Zhang, L. Feng, C. C. Huang, F. J. Xu, N. Tang, L. W. Sang, M. Sumiya, and B. Shen, “Temperature-controlled epitaxy of InxGa1-xN alloys and their band gap bowing,” J. Appl. Phys. 110(11), 113514 (2011). [CrossRef]
23. H. P. Hu, S. J. Zhou, X. T. Liu, Y. L. Gao, C. Q. Gui, and S. Liu, “Effects of GaN/AlGaN/Sputtered AlN nucleation layers on performance of GaN-based ultraviolet light-emitting diodes,” Sci. Rep. 7(1), 44627 (2017). [CrossRef]
24. Z. Q. Guan, J. L. Tang, Z. P. Wei, S. Z. Niu, X. Fang, D. Fang, D. K. Wang, H. M. Jia, and X. H. Wang, “Application of X-Ray Dual Crystal Diffraction in Semiconductor Materials and Structure Analysis,” Material Sciences 08(01), 37–44 (2018). [CrossRef]
25. S. Hammersley, M. J. Kappers, F. C. P. Massabuau, S. L. Sahonta, P. Dawson, R. A. Oliver, and C. J. Humphreys, “Effects of quantum well growth temperature on the recombination efficiency of InGaN/GaN multiple quantum wells that emit in the green and blue spectral regions,” Appl. Phys. Lett. 107(13), 132106 (2015). [CrossRef]
26. Z. C. Li, J. P. Liu, M. X. Feng, K. Zhou, S. M. Zhang, H. Wang, D. Y. Li, D. Q. Li, L. Q. Zhang, D. G. Zhao, D. S. Jiang, H. B. Wang, and H. Yang, “Suppression of thermal degradation of InGaN/GaN quantum wells in green laser diode structures during the epitaxial growth,” Appl. Phys. Lett. 103(15), 152109 (2013). [CrossRef]
27. Y. T. Moon, D. J. Kim, K. M. Song, C. J. Choi, S. H. Han, T. Y. Seong, and S. J. Park, “Effects of thermal and hydrogen treatment on indium segregation in InGaN/GaN multiple quantum wells,” J. Appl. Phys. 89(11), 6514–6518 (2001). [CrossRef]
28. I. Choi, K. Lee, C. R. Lee, J. S. Lee, S. M. Kim, K. U. Jeong, and J. S. Kim, “Application of Hexagonal Boron Nitride to a Heat-Transfer Medium of an InGaN/GaN Quantum-Well Green LED,” ACS Appl. Mater. Interfaces 11(20), 18876–18884 (2019). [CrossRef]
29. D. I. Florescu, V. M. Asnin, F. H. Pollak, A. M. Jones, J. C. Ramer, M. J. Schurman, and I. Ferguson, “Thermal conductivity of fully and partially coalesced lateral epitaxial overgrown GaN/sapphire (0001) by scanning thermal microscopy,” Appl. Phys. Lett. 77(10), 1464–1466 (2000). [CrossRef]
30. A. I. Alhassan, N. G. Young, R. M. Farrell, C. Pynn, F. Wu, A. Y. Alyamani, S. Nakamura, S. P. Denbaars, and J. S. Speck, “Development of high-performance green c-plane III-nitride light-emitting diodes,” Opt. Express 26(5), 5591–5601 (2018). [CrossRef]
