Open Access
Add to BibSonomy
Abstract
We report a possible way to extend the emission wavelength of InyGa1-yN/InxGa1-xN quantum-well (QW) light-emitting diodes (LEDs) to the yellow-red spectral range with little degradation of the radiative efficiency. The InyGa1-yN well with high indium (In) content (HI-InyGa1-yN) was realized by periodic Ga-flow interruption (Ga-FI). The In contents of the HI-InyGa1-yN well and the InxGa1-xN barrier were changed to manipulate the emission wavelength of the LEDs. An In0.34Ga0.66N/In0.1Ga0.9N-QW LED, grown by continuous growth mode (C-LED), was prepared as a reference. The photoluminescence (PL) wavelengths of the HI-InyGa1-yN/InxGa1-xN QW LEDs were changed from 556 to 597 nm. The PL intensity of the HI-InyGa1-yN/InxGa1-xN LED with a peak wavelength of 563 nm was 2.7 times stronger than that of the C-LED (λ = 565 nm). The luminescence intensity for the HI-InyGa1-yN/InxGa1-xN QW LED emitting at 597 nm was stronger than those of the other LED samples with shorter wavelengths. Considering the previous works on degradation in crystal quality and increase in the quantum-confined Stark effect with increasing In content in InGaN, the approach in this work is very promising for yellow-red InGaN LEDs.
© 2017 Optical Society of America
1. Introduction
Recently, InGaN/GaN light-emitting diodes (LEDs) have attracted considerable attention for the visible spectral region applicable to display, lighting, and phosphor-free white light sources, because InGaN fully covers the visible-wavelength range [1–3]. However, the realization of wavelengths above ~550 nm has been limited, mainly because of difficulties in growing high-quality InGaN/GaN quantum wells (QWs) due to the large difference in lattice constants, thermal expansion coefficients, and optimal growth temperatures between InGaN and GaN [4–6]. That is, to extend the emission wavelength of InGaN/GaN QW LEDs, InGaN with high indium (In) content (HI-InGaN) should be grown with high crystal quality. However, it is intrinsically difficult to grow HI-InGaN. Generally, the growth temperature for InN (~600 °C) is lower than that for GaN (~1000 °C) [6,7]. In order to increase the In content of InGaN, the growth temperature should be as low as possible, which leads to low mobility of gallium (Ga) atoms hindering growth of a high-quality InGaN epilayer. Moreover, the structural properties of HI-InGaN grown at relatively low growth temperature may be degraded due to impurity incorporation, point defect formation, and phase separation [8,9]. More seriously, with increasing In content of InGaN, the well-known quantum-confined Stark effect (QCSE) is increased due to the increase in lattice mismatch between an InGaN well and a GaN barrier. The enhancement in QCSE in the InGaN/GaN QW region leads to a reduced overlap integral between electron and hole wave-functions, resulting in significant degradation of luminescent efficiency of LEDs [10,11]. Recently, III-nitride nano-structures such as quantum wires, quantum dashes, and nano-columns have been actively investigated as direct luminescent media or supporting surrounding structures for light emission with wavelengths over ~550 nm [12,13]. If we consider that commercial blue or green LEDs are based on InGaN/GaN QW structures and low-dimensional nano-structures have low density of states of carriers, it is meaningful to get visible wavelengths over ~550 nm by using QWs. Therefore, more investigation on possible ways to extend the emission wavelength of InGaN-based QWs without severe degradation in the structural and optical properties is necessary.
In this paper, a possible way to extend the emission wavelength of InGaN-based LEDs to the yellow-red spectral range with only little degradation of radiative efficiency is demonstrated by using HI-InyGa1-yN/InxGa1-xN QWs as an active layer, where the HI-InyGa1-yN well was formed by so called Ga-flow interruption (Ga-FI) technique. The structural, optical, and electrical properties of LEDs were investigated by transmission electron microscopy (TEM), high-resolution x-ray diffraction (HRXRD), photoluminescence (PL), and electroluminescence (EL). In addition, the emission wavelength of the HI-InyGa1-yN/InxGa1-xN was systematically manipulated by changing the In/Ga flow ratio and growth temperature for the InxGa1-xN barrier.
2. Experiment
The LED samples used in this work were grown on patterned sapphire substrates by using a Thomas-Swan metal-organic chemical-vapor deposition system. Trimethygallium (TMGa), trimethylindum (TMIn), and ammonia (NH3) were used for Ga, In, and nitrogen sources, respectively. Disilane (Si2H6) and bis-cyclopentadienyl magnesium (CP2Mg) were used as n-type and p-type doping sources, respectively. Figure 1(a) shows the schematic structure of a conventional In0.34Ga0.66N/In0.1Ga0.9N LED (C-LED) formed by continuously supplying TMGa, TMIn, and NH3 at the same time for both well and barrier regions. The C-LED structure consists of a 2 μm-thick GaN buffer layer, a 3 μm-thick n-GaN layer, and five-stacked In0.34Ga0.66N/In0.1Ga0.9N QWs. For the multiple QWs (MQWs) of the C-LED, the growth time and growth temperature for an In0.34Ga0.66N well were 45 seconds and 780 °C, respectively. The growth temperature for the In0.1Ga0.9N barrier was 915 °C. Finally, a 150 nm-thick p-type GaN layer was grown on top of the In0.34Ga0.66N/In0.1Ga0.9N MQWs at a growth temperature of 1010 °C. Figure 1(b) shows the schematic diagram of a HI-InyGa1-yN/InxGa1-xN-MQW LED. The HI-InyGa1-yN well was formed by the Ga-FI technique schematically illustrated in the inset, where the TMGa supply was periodically interrupted during the deposition of In0.3Ga0.7N. The 2 second Ga-FI was carried out every 7 seconds during the deposition of In0.3Ga0.7N. Five periods of the In0.3Ga0.7N growth and Ga-FI were used for a single HI-InyGa1-yN layer. The HI-InyGa1-yN well was deposited at a growth temperature of 800 °C, which is a little higher compared to that 780 °C used for In0.34Ga0.66N for the C-LED. After forming HI-InyGa1-yN by the Ga-FI technique, an InxGa1-xN barrier was deposited. The active region of the LEDs was composed of five stacks of the HI-InyGa1-yN/InxGa1-xN QW. The periodic Ga-FI in this work is quite different from the growth interruption techniques for the InGaN-based QWs reported in previous works [14,15]. For an example, Cheong et al used the growth interruption between the InGaN well and the GaN barrier [14]. In addition, there is no report on the realization for a red spectral window from InGaN-based QWs by using a growth interruption technique. That is, the previous growth interruptions were used to improve the structural and optical properties of InGaN-based QWs for blue and green LEDs.
Fig. 1 Schematic diagrams for (a) the C-LED and (b) the HI-InyGa1-yN/InxGa1-xN LED with a conceptual illustration of the Ga-FI technique for the growth of a HI-InyGa1-yN well.
Table 1 shows the summary of the growth conditions for the LED samples. The In/Ga flow ratios, which were calculated by simply using the amount of flow for In and Ga sources, were 0 (S1-LED), 0.217 (S2-LED), and 0.434 (S3-LED) for the InxGa1-xN barrier of the LED samples. The growth temperatures of the InxGa1-xN barrier were 905 °C (S3-LED), 915 °C (S4-LED), and 895 °C (S5-LED). By using energy dispersive spectroscopy measurements (not shown here), the In content of the HI-InyGa1-yN well and the InxGa1-xN barrier for the LED samples were measured, as described in Table 2. The structural properties of the LED samples were measured by using cross-sectional TEM (JEM-2100F of JEOL Ltd.) and HRXRD (PANalytical X’Pert PRO). Figure 2 shows the cross-sectional TEM images for the LED samples. The thicknesses of the HI-InyGa1-yN well and the InxGa1-xN barrier for the LED samples are summarized in Table 2. For PL measurements, a He–Cd laser with a wavelength of 375 nm was used as excitation source. Luminescence spectra were detected by using a CCD detector. The electrical properties of the LED samples were measured by a LED chip tester.
Table 1. Summary on the growth conditions for the LED samples.
Table 2. Structural properties of the HI-InyGa1-yN well and the InxGa1-xN barrier for the LED samples
Fig. 2 Cross-sectional TEM images for (a) the C-LED, (b) the S1-LED, (c) the S2-LED, (d) the S3-LED, (e) the S4-LED, and (f) the S5-LED.
3. Results and discussion
Figure 2(a) and 2(c) show the cross-sectional TEM images of the C-LED with In0.34Ga0.66N/In0.1Ga0.9N MQWs and the S2-LED with HI-InyGa1-yN/InxGa1-xN MQWs, respectively. The thicknesses of the HI-InyGa1-yN well and the InxGa1-xN barrier for the S2-LED (C-LED) were measured to be 2.72 (3) nm and 17.12 (16.87) nm, respectively. The well thickness of the C-LED is thicker than that of the S2-LED, which is related to the difference in growth technique and growth rate affected by the growth temperature. In previous works, the growth rate of InGaN was relatively high when the growth temperature was low [16,17]. Similarly, the growth rate of the well for the C-LED (growth temperature of 780 °C) in this work was relatively higher than that of the S2-LED (growth temperature of 800 °C). The TEM image of the C-LED in Fig. 2(a) shows many dark spots corresponding to In-rich clusters. Formation of In-rich clusters can be explained by phase-separation, which is related to the miscibility gap between InN and GaN [18]. The interface between the In0.34Ga0.66N well and the In0.1Ga0.9N barrier of the C-LED cannot be clearly distinguished due to the presence of the In-rich clusters. On the other hand, the dark regions corresponding to HI-InyGa1-yN of the S2-LED sample in Fig. 2(c) are more clearly observed as compared to the C-LED, indicating that In distribution in the HI-InyGa1-yN well is more uniform than that in the C-LED. The interface quality between the HI-InyGa1-yN well and the InxGa1-xN barrier for the S2-LED was also improved compared to that of the C-LED. Generally, the crystal quality of InGaN with decreasing growth temperature is degraded mainly due to the difference in optimal growth temperatures for InN and GaN [6]. A Lin et al. reported an improvement of the surface morphology of an InGaN QW by the redistribution of surface adatoms through the InN treatment after the growth of the QW and attributed this to In being a surfactant [19]. The Ga-FI in this work was periodically performed during the growth of the HI-InyGa1-yN well region. As a result, the surface morphology at each step could be improved by In acting as a surfactant. In addition, the probability for the In atoms to remain or to exist in the HI-InyGa1-yN well formed by the Ga-FI technique was surely increased compared to the C-LED despite of the relatively higher growth temperature, resulting in the red-shift in the emission wavelength.
Figure 3(a) shows the HRXRD rocking curves of the (0002) diffractions for the C-LED and the S2-LED. The intensity (line-width) of the peak corresponding to HI-InyGa1-yN for the S2-LED was relatively stronger (narrower) than that of the C-LED. The growth temperature for the HI-InyGa1-yN well for the S2-LED sample was higher than that of the C-LED, resulting in the improvement in surface mobility of the group-III elements to form more uniform InGaN. Figure 3(b) shows the PL spectra for the C-LED and the S2-LED, measured at room temperature (RT). For clarity, the yellow-band emission (around λ = 557 nm) of an undoped GaN sample is inserted. The PL peak wavelengths for the C-LED and the S2-LED were measured to be 565 nm and 563 nm, respectively. The slight red-shift in emission wavelength and the higher PL intensity for the C-LED and the S2-LED compared the yellow-band emission indicate that the PL emission originates from the QWs. The PL intensity of the S2-LED at RT was 2.7 times stronger than that of the C-LED, indicating that the Ga-FI technique can provide an effective way to get HI-InyGa1-yN with high crystal quality.
Figure 4(a) shows the HRXRD rocking curves of the (0002) diffractions for the LED samples, where the strongest peak originates from GaN. The peak positions for HI-InyGa1-yN were measured to be 34.40°, 34.40°, and 34.36° for the S1-LED, the S2-LED, and the S3-LED, respectively, where the In content of the HI-InyGa1-yN well was increased by increasing the In/Ga flow ratio for the InxGa1-xN barrier. The full-width at half maximum (FWHM) of the peaks corresponding to HI-InyGa1-yN for the S1-LED, S2-LED and S3-LED were measured to be 0.093°, 0.043°, and 0.079°, respectively. For the S1-LED with an In/Ga flow ratio of 0 for the barrier, the FWHM of the peak corresponding to HI-InyGa1-yN was higher compared to those of the S2-LED and the S3-LED. The relatively narrow FWHMs for the S2-LED and the S3-LED compared to that of the S1-LED can be explained by the improvement in crystal quality of the HI-InyGa1-yN. Lin et al. reported that the reduction in compressive strain between an InGaN well and an InxGa1-xN barrier resulted in the improvement in crystal quality [20,21]. However, the FWHM of the HI-InyGa1-yN peak in the HRXRD rocking curves for the S3-LED is larger than that of the S2-LED sample, which can be interpreted as an increase in In phase separation and compositional fluctuation due to higher In/Ga flow ratio for the InxGa1-xN barrier and thus a higher overall In content [22,23]. Figure 4(b) shows the PL spectra for the LED samples, measured at RT. The peak wavelengths of the S1-LED, the S2-LED, and the S3-LED were measured as 556 nm, 563 nm, and 583 nm, respectively, which were mainly related to the amount of In content in the HI-InyGa1-yN well. The emission wavelength of the LED sample was red-shifted with increasing the In/Ga flow ratio for the InxGa1-xN barrier. This is attributed to the increase in the In content of the HI-InyGa1-yN well, mainly caused by the reduction in the diffusion probability of In atoms from the well to the barrier [24]. The relatively low potential of the InxGa1-xN barrier additionally contributed to the red-shift with increasing the In/Ga flow ratio. In addition, the PL spectrum for the S3-LED sample showed the double-peak feature, as fitted by two solid lines in the inset. The double peaks are originating from areas with different In content within the HI-InyGa1-yN.
Fig. 4 (a) HRXRD rocking curves and (b) PL spectra for the S1-LED (0), S2-LED (0.217), and S3-LED (0.434) with different In/Ga flow ratio for the InxGa1-xN barrier. (c) Normalized EL spectra for the LED samples at injection current of 20 mA, where the insets are the emission images for the LED samples taken during the EL measurements.
Figure 4(c) shows the normalized EL spectra for the LED samples, measured at an injection current of 20 mA. The peak wavelengths for the S1-LED, the S2-LED, and the S3-LED samples are measured to be 559 nm, 565 nm, and 581 nm, respectively. The inset of the Fig. 4(c) shows the emission images for the LED samples taken during the EL measurements by using In-ball bonding.
Figure 5(a) shows the PL spectra for the HI-InyGa1-yN/InxGa1-xN LED samples with different growth temperature for the InxGa1-xN barrier. The peak wavelengths for the S3-LED, the S4-LED, and the S5-LED were measured as 583 nm, 575 nm, and 597 nm, respectively. The emission peaks in the PL spectra for the LED samples are well coincident with the In content of the HI-InyGa1-yN well region. For the S4-LED with the growth temperature of 915 °C for the InxGa1-xN barrier, the In desorption during the growth of the barrier is higher than those of the S5-LED (895 °C) and the S3-LED (905 °C). As a result, less In coverage of the InxGa1-xN barrier at high growth temperature reduces In uptake in the QW. Also, there is more possibility for In atoms consisting of the HI-InyGa1-yN well to diffuse into the barrier at relatively high growth temperature for the S4-LED compared to those of the S3-LED and S5-LED. At the same time, Ga atoms diffuse more from the barrier into the well. As a result, the emission wavelength of the S4-LED sample was relatively short compared to other samples. In the same context, the peak position of the S5-LED was relatively longer than that of the S3-LED. If we only consider the amount of In in the HI-InyGa1-yN well, the QCSE should be increased for the S5-LED. As a result, the PL intensity might be reduced due to the degradation in the overlap integral between electron and hole wave-functions. However, the PL intensity of the S5-LED was 1.53 times stronger than that of the S4-LED. In addition, while double peaks were observed from the PL spectrum of the S4-LED, the shoulder-peak feature disappeared for the S5-LED showing high emission intensity. The increase in PL yield with a narrower line-width for the S5-LED is related to the improvement in In distribution at the HI-InyGa1-yN well. Figure 5(b) shows the EL spectra for the LED samples with a die size of 1x1 mm2, measured at injection current of 20 mA. The EL characteristics of the LED samples are well agreed with PL results. The EL intensity of the S5-LED was 1.83 times stronger than that of the S4-LED, which can be explained again by the improvement in In distribution at the HI-InyGa1-yN well. The inset of Fig. 5(b) shows the luminescent images for the LED samples measured at the injection current of 20 mA. Figure 5(c) shows the output powers of the LED samples depending on the injection currents ranging from 4 mA to 120 mA. The output power for the S3-LED, the S4-LED, and the S5-LED at an injection current of 120 mA were measured as 2.68 mW, 2.28 mW, and 3.79 mW, respectively. The output power of the S5-LED is 1.66 times stronger than that of the S4-LED. In addition, the increasing rate of the output power for the S5-LED with increasing the injection current is higher than those of the S3-LED and the S4-LED. The current dependences of external quantum efficiencies (EQEs) for the LED samples are summarized in the inset of the Fig. 5(c). The EQEs were calculated to be 1.1, 0.9, and 1.6% for the S3-LED, the S4-LED, and the S5-LED at an injection current of 60 mA, respectively. The EQE for the S5-LED sample operating at red spectral range is higher than those for the S3-LED and the S4-LED with shorter wavelengths. The increasing rate of the EQE for the S5-LED was also higher than those for the S3-LED and the S4-LED. The significant efficiency droop with increasing injection current was not observed for the LED samples. Figure 5(d) shows the summary on the emission wavelength for the LED samples with increasing injection current ranging from 4 to 120 mA. With increasing current, the emission wavelengths of the LED samples were blue-shifted mainly due to the band-filling effects of the localized energy states and the screening effect of the QCSE [25,26]. The amount of blue-shift in the emission wavelength for the S5-LED was measured to be 60.3 nm, which is larger than that (52.8 nm) for the S4-LED. This is attributed to the increase in the influence of the QCSE in the QW structure [27,28].
Fig. 5 (a) PL spectra of the S3-LED (905 °C), the S4-LED (915 °C), and the S5-LED (895 °C) with different growth temperature for InxGa1-xN barrier measured at RT. (b) EL spectra for the LED samples measured at injection current of 20 mA, where the insets are the luminescence images of the LED chips. Summaries on the (c) output powers with the EQEs (inset) and (d) emission wavelengths for the LED samples depending on the injection current ranging from 4 to 120 mA.
4. Conclusion
We shifted the emission wavelength of HI-InyGa1-yN/InxGa1-xN LEDs from 556 to 597 nm by changing the In/Ga flux ratio and growth temperature for the InxGa1-xN barrier, where the HI-InyGa1-yN well was formed by Ga-FI technique. The emission wavelength of the HI-InyGa1-yN/InxGa1-xN LED samples was red-shifted with increasing the In/Ga flow ratio for the InxGa1-xN barrier. This can be explained by the increase in the In content of the HI-InyGa1-yN well additionally introduced from the barrier via the inter-diffusion and low potential of the InxGa1-xN barrier. In addition, the emission wavelength of the HI-InyGa1-yN/InxGa1-xN LEDs was red-shifted with decreasing growth temperature, which was due to the less possibility of inter-diffusion for In and Ga atoms at the interface between the well and the barrier. The luminescent intensity of the HI-InyGa1-yN/InxGa1-xN LED emitting 597 nm (S5-LED) was relatively strong compared to other LED samples with shorter emission wavelengths. As a conclusion, the formation of HI-InyGa1-yN using the Ga-FI technique can be an effective way to extend emission wavelength to yellow-red spectral region with less degradation of radiative efficiency.
Acknowledgments
This work was supported by a National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT and Future Planning (No. 2015042417), and by the Ministry of Education (No. 2015R1D1A1A01060681).
References and links
1. M. H. Crawford, “LEDs for solid-state lighting: performance challenges and recent advances,” IEEE J. Sel. Top. Quantum Electron. 15(4), 1028–1040 (2009). [CrossRef]
2. B. Damilano and B. Gil, “Yellow-red emission from (Ga,In)N heterostructures,” J. Phys. D Appl. Phys. 48(40), 403111 (2015). [CrossRef]
3. C. F. Huang, C. F. Lu, T. Y. Tang, J. J. Huang, and C. C. Yang, “Phosphor-free white-light light-emitting diode of weakly carrier-density-dependent spectrum with prestrained growth of InGaN/GaN quantum wells,” Appl. Phys. Lett. 90(15), 151122 (2007). [CrossRef]
4. S. Y. Kwon, H. J. Kim, H. Na, Y. W. Kim, H. C. Seo, H. J. Kim, Y. Shin, E. Yoon, and Y. S. Park, “In-rich InGaN GaN quantum wells grown by metal-organic chemical vapor deposition,” J. Appl. Phys. 99(4), 044906 (2006). [CrossRef]
5. H. S. Chen, C. F. Lu, D. M. Yeh, C. F. Huang, J. J. Huang, and C. C. Yang, “Orange–red light-emitting diodes based on a prestrained InGaN–GaN quantum-well epitaxy structure,” IEEE Photonics Technol. Lett. 18(21), 2269–2271 (2006). [CrossRef]
6. A. Yamamoto, K. Sugita, and A. Hashimoto, “Elucidation of factors obstructing quality improvement of MOVPE-grown InN,” J. Cryst. Growth 311(22), 4636–4640 (2009). [CrossRef]
7. D. Iida, K. Nagata, T. Makino, M. Iwaya, S. Kamiyama, H. Amano, I. Akasaki, A. Bandoh, and T. Udagawa, “Growth of GaInN by raised-pressure metalorganic vapor phase epitaxy,” Appl. Phys. Express 3(7), 075601 (2010). [CrossRef]
8. T. Langer, A. Kruse, F. A. Ketzer, A. Schwiegel, L. Hoffmann, H. Jönen, H. Bremers, U. Rossow, and A. Hangleiter, “Origin of the “green gap”: Increasing nonradiative recombination in indium‐rich GaInN/GaN quantum well structures,” Phys. Status Solidi., C Curr. Top. Solid State Phys. 8(7–8), 2170–2172 (2011). [CrossRef]
9. K. G. Belyaev, M. V. Rakhlin, V. N. Jmerik, A. M. Mizerov, Y. V. Kuznetsova, M. V. Zamoryanskaya, S. V. Ivanov, and A. A. Toropov, “Phase separation in InxGa1‐xN (0.10< x< 0.40),” Phys. Status Solidi., C Curr. Top. Solid State Phys. 10(3), 527–531 (2013). [CrossRef]
10. H. K. Cho, J. Y. Lee, C. S. Kim, and G. M. Yang, “Influence of strain relaxation on structural and optical characteristics of InGaN/GaN multiple quantum wells with high indium composition,” J. Appl. Phys. 91(3), 1166–1170 (2002). [CrossRef]
11. T. H. Ngo, B. Gil, P. Valvin, B. Damilano, K. Lekhal, and P. D. Mierry, “Internal quantum efficiency in yellow-amber light emitting AlGaN-InGaN-GaN heterostructures,” Appl. Phys. Lett. 107(12), 122103 (2015). [CrossRef]
12. H. Sekiguchi, K. Kishino, and A. Kikuchi, “Emission color control from blue to red with nanocolumn diameter of InGaN/GaN nanocolumn arrays grown on same substrate,” Appl. Phys. Lett. 96(23), 231104 (2010). [CrossRef]
13. S. Jahangir, M. Mandl, M. Strassburg, and P. Bhattacharya, “Molecular beam epitaxial growth and optical properties of red-emitting (λ = 650 nm) InGaN/GaN disks-in-nanowires on silicon,” Appl. Phys. Lett. 102(7), 071101 (2013). [CrossRef]
14. M. G. Cheong, H. S. Yoon, R. J. Choi, C. S. Kim, S. W. Yu, C. H. Hong, E. K. Suh, and H. J. Lee, “Effects of growth interruption on the optical and the structural properties of InGaN/GaN quantum wells grown by metalorganic chemical vapor deposition,” J. Appl. Phys. 90(11), 5642–5646 (2001). [CrossRef]
15. S. J. Leem, M. H. Kim, J. Shin, Y. Choi, and J. Jeong, “The effects of In flow during growth interruption on the optical properties of InGaN multiple quantum wells grown by low pressure metalorganic chemical vapor deposition,” Jpn. J. Appl. Phys. 40(2), L371–L373 (2001). [CrossRef]
16. C. J. Deatcher, C. Liu, S. Pereira, M. Lada, A. G. Cullis, Y. J. Sun, O. Brandt, and I. M. Watson, “In situ optical reflectometry applied to growth of indium gallium nitride epilayers and multi quantum well structures,” Semicond. Sci. Technol. 18(4), 212–218 (2003). [CrossRef]
17. R. W. Martin, P. R. Edwards, R. Pecharroman-Gallego, C. Liu, C. J. Deatcher, I. M. Watson, and K. P. O’Donnell, “Light emission ranging from blue to red from a series of InGaN/GaN single quantum wells,” J. Phys. D Appl. Phys. 35(7), 604–608 (2002). [CrossRef]
18. P. Li, H. Li, Z. Li, J. Kang, X. Yi, J. Li, and G. Wang, “Strong carrier localization effect in carrier dynamics of 585nm InGaN amber light-emitting diodes,” J. Appl. Phys. 117(7), 073101 (2015). [CrossRef]
19. H. C. Lin, R. S. Lin, and J. I. Chyi, “Enhancing the quantum efficiency of InGaN green light-emitting diodes by trimethylindium treatment,” Appl. Phys. Lett. 92(16), 161113 (2008). [CrossRef]
20. Z. Lin, H. Wang, W. Wang, Y. Lin, M. Yang, S. Chen, and G. Li, “Employing low-temperature barriers to achieve strain-relaxed and high-performance GaN-based LEDs,” Opt. Express 24(11), 11885–11896 (2016). [CrossRef] [PubMed]
21. L. Zhang, K. Cheng, S. Degroote, M. Leys, M. Germain, and G. Borghs, “Strain effects in GaN epilayers grown on different substrates by metal organic vapor phase epitaxy,” J. Appl. Phys. 108(7), 073522 (2010). [CrossRef]
22. Y. S. Lin, K. J. Ma, C. Hsu, S. W. Feng, Y. C. Cheng, C. C. Liao, C. C. Yang, C. C. Chou, C. M. Lee, and J. I. Chyi, “Dependence of composition fluctuation on indium content in InGaN/GaN multiple quantum wells,” Appl. Phys. Lett. 77(19), 2988–2990 (2000). [CrossRef]
23. M. V. Klymenko, O. Shulika, and I. A. Sukhoivanov, “Theoretical study of optical transition matrix elements in InGaN/GaN SQW subject to indium surface segregation,” IEEE J. Sel. Top. Quantum Electron. 17(5), 1374–1380 (2011). [CrossRef]
24. H. Y. Liu, X. D. Wang, B. Xu, D. Ding, W. H. Jiang, J. Wu, and Z. G. Wang, “Effect of In-mole-fraction in InGaAs overgrowth layer on self-assembled InAs/GaAs quantum dots,” J. Cryst. Growth 213(1–2), 193–197 (2000). [CrossRef]
25. L. W. Wu, S. J. Chang, T. C. Wen, Y. K. Su, J. F. Chen, W. C. Lai, C. H. Kuo, C. H. Chen, and J. K. Sheu, “Influence of Si-doping on the characteristics of InGaN-Gan multiple quantum-well blue light emitting diodes,” IEEE J. Quantum Electron. 38(5), 446–450 (2002). [CrossRef]
26. E. Kuokstis, J. W. Yang, G. Simin, M. A. Khan, R. Gaska, and M. S. Shur, “Two mechanisms of blueshift of edge emission in InGaN-based epilayers and multiple quantum wells,” Appl. Phys. Lett. 80(6), 977–979 (2002). [CrossRef]
27. T. Mukai, M. Yamada, and S. Nakamura, “Characteristics of InGaN-based uv/blue/green/amber/red light-emitting diodes,” Jpn. J. Appl. Phys. 38(7), 3976–3981 (1999). [CrossRef]
28. T. H. Ngo, B. Gil, B. Damilano, K. Lekhal, and P. D. Mierry, “Internal quantum efficiency and auger recombination in green, yellow and red InGaN-based light emitters grown along the polar direction,” Superlattices Microstruct. 103, 245–251 (2017). [CrossRef]
