Open Access
Add to BibSonomyAbstract
The authors report the use of InN/GaN alternative structure to replace the thick InGaN well layers in the InGaN/GaN multiquantum well (MQW) and the fabrication of GaN-based green light-emitting diodes (LEDs). Using this method, it was found that we could achieve InGaN “well layers” with high crystal quality due to the enhanced migration of adatoms during the growth. It was also found that indium composition in the InGaN “well layers” and the thickness of the InGaN “well layers” both depend strongly on the growth time of InN and GaN. It was also found that we could achieve stronger electroluminescence (EL) intensities with narrower full-width-half-maxima (FWHMs) from the LEDs with InN/GaN alternative growth InGaN “well layers”. Furthermore, it was found that we could achieve better ideality factors and smaller reverse leakage currents from the proposed devices.
© 2013 Optical Society of America
1. Introduction
In the past decades, considerable progress has been made on GaN-based light-emitting diodes (LEDs) with InGaN/GaN multiple quantum wells (MQWs) active region [1–3]. Great effort has been exerted to improve the material quality [4,5], light-extraction efficiency (LEE) [6,7], and metal-semiconductor ohmic contacts [8] to drastically improve the luminous efficiency of GaN-based LEDs. Recent works based on photonic crystals [9] and microsphere array [10,11] have been employed to enhance LEE in GaN-based LEDs. New method to grow epitaxial layers on nano-patterned sapphire substrate has also been used to reduce dislocation density in these devices [12–14]. Although output power of current LEDs is high already, output efficiency of these LEDs requires further improvement to reach feasible solid-state lighting. On the other hand, it is possible to achieve GaN-based green LEDs with long emission wavelength by introducing high-indium mole fraction into the InGaN well layers. However, this could result in an enlarged lattice-mismatch-induced strain in the InGaN well layers and the formation of V-shaped defects. It is known that these defects are originated from the strain relaxation associated with stacking faults or indium segregation. It is also known that these defects are normally terminated on sample surface and could degrade the efficiency of GaN-based LEDs [15–20]. It should be noted that such a problem will become even more severe for the InGaN/GaN MQW structure with thick InGaN layers. It has been reported previously that one can improve the InGaN/GaN interfacial properties and thus reduce the number of V-shaped defects in GaN-based LEDs by pre-flowing trimethlyindium (TMIn) prior to the growth of InGaN quantum wells (QW) [21–25]. Therefore, one can enhance the emission efficiency of InGaN/GaN MQW green LEDs. It has also been shown that one can use pulse TMIn flow growth method to improve crystal quality of InN epitaxial layers by enhancing indium migration and suppressing the formation of indium droplets [26–28]. Using these methods, one should be able to achieve alternative growth of InN and GaN which could be used to replace thick InGaN well layers. In the present study, we report the use of InN/GaN alternative structure to replace the thick InGaN well layers in the InGaN/GaN MQW and the fabrication of GaN-based green LEDs. The effects of the growth sequences and source flow rates of InN and GaN on the material composition of the thick InGaN “well layers” will be discussed. The electro-optical properties of the fabricated InGaN/GaN MQW green LEDs will also be discussed.
2. Experiments
Samples used in this study were all grown on 2-inch (0001) patterned sapphire substrate (PSS) through a Thomas Swan close-coupled showerhead 19 × 2 metalorganic chemical vapor deposition (MOCVD) system. The periodic convex pattern on the PSS was formed by inductively coupling plasma (ICP) etcher. The pattern diameter, spacing, and height of the PSS were 3.5, 2, and 1.3 µm, respectively. During the MOCVD growth, TMIn, trimethylgallium (TMGa), trimethylaluminium (TMAl), and ammonia (NH3) were used as the source materials of In, Ga, Al, and N, respectively. Bicyclopentadienyl magnesium (CP2Mg) and silane (SiH4) were used as the p-type and n-type doping sources, respectively. The reactor temperature was first raised to 900°C to grow a 10-nm-thick in situ AlN nucleation layer on the PSS. The reactor temperature was then raised to 1050°C to grow a 2-µm-thick undoped GaN (u-GaN) epitaxial layer. A 2-µm-thick n-GaN layer was subsequently grown at 1050°C. A three-pair Si-doped InGaN (2 nm)/GaN(14 nm) strain releasing multiquantum well (SRMQW) structure was then grown at 800°C [29]. The reactor temperature was then ramped down to growth a 12-pair InGaN (3 nm)/GaN (16 nm) light-emitting MQW structure using the high-low temperature scheme. In other words, the growth temperatures of the InGaN “well layers” and the GaN barrier layers was kept at 730°C and 880°C, respectively. It should be noted that the InGaN “well layers” used in this study were prepared by the InN/GaN alternative growth method. To grow the InN/GaN alternative structure, the flow rate of TMGa was kept at 4.2 μmol/min during the GaN growth while TMIn flow rate was kept at 70 μmol/min during the growth of InN while the NH3 flow rate was kept at 26 slm throughout the growth. In other words, the V/III ratios during the growth of InN and GaN were kept at around 16000 and 28000, respectively. On the other hand, the growth pressure was kept at 225 torr. In this study, we varied the growth time of the InN and GaN of the InN/GaN alternative growth to 2 s/2 s for LED II, 4 s/4 s for LED III, and 8 s/8 s for LED IV, respectively. To achieve the 3-nm-thick InGaN “well layers”, we designed the pair numbers of the alternative InN/GaN structure at 40, 20, and 10 for LED II, LED III, and LED IV, respectively. For comparison, green LEDs with conventional growth method were also prepared (i.e., LED I). For LED I, we simultaneously introduced TMIn, TMGa and NH3 into the growth chamber for 80 s during the growth of the 3-nm-thick InGaN well layers. Figures 1(a) and 1(b) shows the source switching sequences and growth temperature profile during the growth of light-emitting MQWs with conventional InGaN well layers (i.e., LED I) and with InN/GaN alternative growth InGaN “well layers” (i.e., LED II, LED III and LED IV), respectively. The growth time for the GaN barrier was notably the same. After the growth of the light-emitting MQW structure, a 20 nm-thick Mg-doped Al0.1Ga0.9N electron blocking layer and a 150 nm-thick Mg-doped p-GaN contact layer were grown for all these LEDs. An X-ray diffraction (XRD) system and a transmission electron microscope were then used to evaluate crystal quality of these as grown samples. Subsequently, standard processing steps were used to fabricate 430 µm × 860 µm LED chips with indium tin oxide (ITO) upper contact [30]. Current–voltage (I–V) characteristics of the fabricated LEDs were then measured using an HP-4156C semiconductor parameter analyzer with a 100 mA current limit. Under high current injection, a Keithley 2400 source meter was used to measure I–V characteristics of these LEDs. The output powers and emission wavelength of the LEDs were measured using a calibrated integrating sphere and a spectrometer (Ocean Optics USB2000) at room temperature.
Fig. 1 Schematic of the source switching sequences of MQW growth conditions of (a) LED I and (b) LED II, LED III, and LED IV. Axis scales are not in proportion.
3. Results and discussion
Figure 2 shows the XRD spectra measured from the fabricated LEDs. It can be seen that distinct satellite peaks could be clearly observed from all these XRD spectra. These observations suggest good crystal quality with abrupt interfaces. The XRD results indicate that the periodicity was around 20 nm for all these samples. And the zero order InGaN peak of XRD spectra indicated the average In compositions of 5.6%, 5.5%, and 5.2% for LED II, LED III and LED IV, respectively. In other words, average indium percentages of InGaN/GaN MQWs decreased with the increased growth time for the LEDs with InN/GaN alternative growth structure. On the other hand, average indium percentage was 5.1% for LED I with the conventional structure.
Fig. 2 X-ray θ–2θ diffraction (XRD) spectra of all LEDs. X-ray diffraction peaks of the SRMQW and the Mg-doped Al0.1Ga0.9N layer are marked as square (■) and triangle (▲) respectively.
Figures 3(a), 3(b), 3(c) and 3(d) show TEM images taken from LED I, LED II, LED III and LED IV, respectively. It can be seen that well thickness of LED I with conventional growth method was approximately 3.3 nm. In contrast, thicknesses of the InGaN “well layers” for LED II, LED III and LED IV were 7.2, 6.8, and 6.3 nm for LED II, LED III and LED IV, respectively. These numbers indicate that thicknesses of the InGaN “well layers” prepared by the InN/GaN alternative growth were at least two times thicker than that of the InGaN well layers prepared by the conventional growth method. From the XRD and TEM results, it was found that indium composition in the InGaN well layers was 33.3%. In contrast, indium compositions in LED II, LED III and LED IV were 17.6%, 17.9% and 18.6%, respectively. These numbers indicate that indium compositions in the InGaN “well layers” prepared by the InN/GaN alternative growth were significantly smaller than that in the InGaN well layers prepared by the conventional growth method, as shown in Fig. 5. Although the InN/GaN alternative growth method could be used to replace the growth of thick InGaN well layers, it was found that no clear InN and GaN interfaces could be observed in the TEM images. It is possible that the lack of clear indium modulation within the thick InGaN “well layers” was due to the poor TEM resolution. It is also possible that InN monolayers could not be grown during the switching sequences. It has been reported by Jamil et al [27] that the improvement of In migration by the pulse TMIn flow mode. In adatoms will easily migrate on the surface of the GaN or InGaN without Ga adatoms by using the InN/GaN alternative growth. Such migration could lead to the spreading of indium duing the subsequent GaN growth sequence. Thus, no clear InN/GaN interfaces could be observed in the TEM images.
Figure 4 shows electroluminescence (EL) spectra measured from the fabricated LEDs with 60 mA current injection. As shown in Fig. 4, it was found that EL peak occurred at 511 nm for both LEDs I and LED II while EL peak positions clearly blue-shifted for LED III and LED IV. It was also found that EL intensities observed from LED II, LED III and LED IV were all significantly larger than that observed from LED I. Furthermore, it was found that full-width-half-maximum (FWHM) of the EL peak observed from LED I was around 34 nm. In contrast, EL peak FWHMs obaservd from LED II, LED III and LED IV were all around 32 nm. Compared with LED I with conventional growth method, the stronger EL intensities and the narrower FWHMs observed from the LEDs with InN/GaN alternative growth InGaN “well layers” (i.e., LED II, LED III and LED IV) should be attributed to the improved crystal quality. Figure 5 summaries 60 mA EL peak positions and the average indium compositin observed from the fabricated LEDs. For the LEDs InN/GaN alternative growth InGaN “well layers”, it can be seen that the 60 mA emission wavelength became shorter as we increased the growth time for each InN and GaN layer. Although a longer growth time for each InN and GaN layer could result in a larger indium composition in the InGaN “well layers”, the thickness of the InGaN “well layers” also became smaller. Thus, emission wavelength depends strongly on all these parameters. For example, it has been shown from XRD and TEM results that average indium compositions were 33.3% for LED I and 17.6% for LED II. Thus, EL peak position observed from LED II should be notably shorter, as compared to that of LED I. It can be seen from Figs. 4 and 5 that EL peak positions were about the same for these two LEDs. Such a discrepancy should be attributed to the fact that thickness of the InGaN “well layers” for LED II was 7.2 nm while thicknesses of the InGaN well layers for LED I was only 3.3 nm. We could thus achieve similar EL peak wavelength from these two LEDs.
Fig. 5 Emission wavelength at a forward current of 60 mA, and the In composition in the InGaN/GaN MQWs of fabricated LEDs.
Figure 6(a) shows forward I–V characteristics measured from the fabricated LEDs. It can be seen that the 60 mA forward voltages, Vf, of LED I, LED II, LED III and LED IV were 3.21, 3.15, 3.16 and 3.07 V, respectively. Compared with LED I, the differences in 60 mA Vf observed LED II, LED III and LED IV were all within 5%. It was also found that the ideal factors of LED I, LED II, LED III and LED IV were 2.86, 2.64, 2.20, and 1.65, respectively. In other words, we could achieve smaller ideal factors from the LEDs with InN/GaN alternative growth InGaN “well layers” (i.e., LED II, LED III and LED IV), as compared to LED I with conventional structure. Figure 6(b) shows reverse I–V characteristics measured from the fabricated LEDs. With a reverse bias of −15 V, it was found that the reverse leakage current measured from LED I, LED II, LED III and LEDs IV were 11.9, 9.4, 7.3, and 5.5 µA, respectively. The better ideality factors and the smaller reverse leakage currents observed from LED II, LED III and LED IV should again be attributed to the improved crystal quality due to the use of InN/GaN alternative growth InGaN “well layers”.
Fig. 6 (a) Forward I–V characteristics and (b) Reverse I–V characteristics of all the fabricated LEDs.
Figure 7 shows output powers and external quantum efficiencies (EQEs) as functions of the injection current for these LEDs. As we increased the injection current, it was found that output power increased for all these LEDs. It was also found that output powers measured from the LEDs with InN/GaN alternative growth InGaN “well layers” (i.e., LED II, LED III and LED IV) were all larger than that measured from LED I with conventional structure. With 60 mA current injection, it was found that output powers were 38.7, 49.9, 54.6 and 54.9 mW, which correspond to 26.6%, 34.3%, 37.3% and 37.0% EQEs for LED I, LED II, LED III and LED IV, respectively. In other words, we could achieve significant enhancement in both output power and EQE by replacing the thick InGaN well layers with InN/GaN alternative growth InGaN “well layers” for GaN-based green LEDs. Compared with LED II, it was also found that output powers measured from LED III and LED IV were larger. This should be attributed to the reduce thickness of the InGaN “well layers”. Here, we define efficiency droop as the efficiency degradation from the peak of EQE to the EQE of 220 mA. With this definition, it was found that the efficiency droops were 38.2%, 41.4%, 40.4% and 45.0% for LED I, LED II, LED III and LED IV, respectively. In other words, the LEDs with InN/GaN alternative growth InGaN “well layers” suffered more severely from the efficiency droop, as compared to LED I with conventional structure. We also checked the EL peak position as a function of injection current for these LEDs. It was found that the amount of EL peak blue shifted for 16.5, 16.7 and 14.2 nm for LED II, LED III and LED IV, respectively, as we increased the injection current from 3 and 220 mA (not shown here). The blue shift of the LEDs decreased with the increase of growth time of InN and GaN should be attributed to the decrease in thickness for the thick InGaN “well layers” On the other hand, EL peak position only blue shifted by 11.5 nm for LED I. The larger blue shift and more significant efficiency droop should thus be attributed to the increased thickness of the InGaN “well layers”. It should be noted that output powers and EQEs observed from LED II, LED III and LED IV were all larger than those observed from LED I, despite the thicker InGaN “well layers” of LED II and LED IV.
Fig. 7 Curves of measured output power and EQE as a function of the injection current for all the LEDs.
4. Conclusions
In summary, we report the use of InN/GaN alternative structure to replace the thick InGaN well layers in the InGaN/GaN MQW and the fabrication of GaN-based green LEDs. Using this method, it was found that we could achieve InGaN “well layers” with high crystal quality due to the enhanced migration of adatoms during the growth. It was also found that indium composition in the InGaN “well layers” and the thickness of the InGaN “well layers” both depend strongly on the growth time of InN and GaN. It was also found that we could achieve stronger EL intensities with narrower FWHMs from the LEDs with InN/GaN alternative growth InGaN “well layers”. Furthermore, it was found that we could achieve better ideality factors and smaller reverse leakage currents from the proposed devices.
Acknowledgments
The authors thank the National Science Council of Taiwan for their financial support under Contract No. NSC101-2221-E-006-066-MY3 and 98-2221-E-006-013-MY3. This research was also made possible through the Advanced Optoelectronic Technology Center, NCKU as a project of the Ministry of Education, Taiwan, and through the financial support of the Bureau of Energy, Ministry of Economic Affairs of Taiwan, under Contract No. 102-E0603.
References and links
1. 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]
2. S. J. Chang, W. C. Lai, Y. K. Su, J. F. Chen, C. H. Liu, and U. H. Liaw, “InGaN-GaN multiquantum-well blue and green light-emitting diodes,” IEEE J. Sel. Top. Quantum Electron. 8(2), 278–283 (2002). [CrossRef]
3. S. J. Chang, C. H. Kuo, Y. K. Su, L. W. Wu, J. K. Sheu, T. C. Wen, W. C. Lai, J. F. Chen, and J. M. Tsai, “400-nm InGaN-GaN and InGaN-AlGaN multiquantum well light-emitting diodes,” IEEE J. Sel. Top. Quantum Electron. 8(4), 744–748 (2002). [CrossRef]
4. S. Nakamura, M. Senoh, N. Iwasa, S. Nagahama, T. Yamada, and T. Mukai, “Superbright green InGaN single-quantum-well-structure light-emitting diodes,” Jpn. J. Appl. Phys. 34(Part 2, No. 10B), L1332–L1335 (1995).
5. T. Mukai, S. Nagahama, M. Sano, T. Yanamoto, D. Morita, T. Mitani, Y. Narukawa, S. Yamamoto, I. Niki, M. Yamada, S. Sonobe, S. Shioji, K. Deguchi, T. Nau, H. Tamaki, Y. Murazaki, and M. Kameshima, “Recent progress of nitride-based light emitting devices,” Phys. Status Solidi A 200(1), 52–57 (2003). [CrossRef]
6. W. C. Lai, Y. Y. Yang, L. C. Peng, S. W. Yang, Y. R. Lin, and J. K. Sheu, “GaN-based light emitting diodes with embedded SiO2 pillars and air gap array structures,” Appl. Phys. Lett. 97(8), 081103 (2010). [CrossRef]
7. J. K. Kim, T. Gessmann, E. F. Schubert, J. Q. Xi, H. Luo, J. Cho, C. Sone, and Y. Park, “GaInN light-emitting diode with conductive omnidirectional reflector having a low-refractive-index indium-tin oxide layer,” Appl. Phys. Lett. 88(1), 013501 (2006). [CrossRef]
8. Y.-L. Li, E. F. Schubert, J. W. Graff, A. Osinsky, and W. F. Schaff, “Low-resistance ohmic contacts to p-type GaN,” Appl. Phys. Lett. 76(19), 2728–2730 (2000). [CrossRef]
9. J. Jewell, D. Simeonov, S. C. Huang, Y. L. Hu, S. Nakamura, J. Speck, and C. Weisbuch, “Double embedded photonic crystals for extraction of guided light in light-emitting diodes,” Appl. Phys. Lett. 100(17), 171105 (2012). [CrossRef]
10. P. Zhu, G. Liu, J. Zhang, and N. Tansu, “FDTD analysis on extraction efficiency of GaN light-emitting diodes with microsphere arrays,” J. Display Technol. 9(5), 317–323 (2013). [CrossRef]
11. X. Li, P. Zhu, G. Liu, J. Zhang, R. Song, Y. Ee, P. Kumnorkaew, J. F. Gilchrist, and N. Tansu, “Light Extraction Efficiency Enhancement of III-Nitride Light-Emitting Diodes by Using 2-D Close-Packed TiO2 Microsphere Arrays,” J. Display Technol. 9(5), 324–332 (2013). [CrossRef]
12. Y. K. Ee, J. M. Biser, W. Cao, H. M. Chan, R. P. Vinci, and N. Tansu, “Metalorganic vapor phase epitaxy of III-Nitride light-emitting diodes on nanopatterned AGOG sapphire substrate by abbreviated growth mode,” IEEE J. Sel. Top. Quantum Electron. 15(4), 1066–1072 (2009). [CrossRef]
13. Y. K. Ee, X. H. Li, J. Biser, W. Cao, H. M. Chan, R. P. Vinci, and N. Tansu, “Abbreviated MOVPE nucleation of III-nitride light-emitting diodes on nano-patterned sapphire,” J. Cryst. Growth 312(8), 1311–1315 (2010). [CrossRef]
14. Y. Li, S. You, M. Zhu, L. Zhao, W. Hou, T. Detchprohm, Y. Taniguchi, N. Tamura, S. Tanaka, and C. Wetzel, “Defect-reduced green GaInN/GaN light-emitting diode on nanopatterned sapphire,” Appl. Phys. Lett. 98(15), 151102 (2011). [CrossRef]
15. X. H. Wu, C. R. Elsass, A. Abare, M. Mack, S. Keller, P. M. Petroff, S. P. DenBaars, J. S. Speck, and S. J. Rosner, “Structural origin of V-defects and correlation with localized excitonic centers in InGaN/GaN multiple quantum wells,” Appl. Phys. Lett. 72(6), 692–694 (1998). [CrossRef]
16. H. K. Cho, J. Y. Lee, G. M. Yang, and C. S. Kim, “Formation mechanism of V defects in the InGaN/GaN multiple quantum wells grown on GaN layers with low threading dislocation density,” Appl. Phys. Lett. 79(2), 215–217 (2001). [CrossRef]
17. N. Sharma, P. Thomas, D. Tricker, and C. Humphreys, “Chemical mapping and formation of V-defects in InGaN multiple quantum wells,” Appl. Phys. Lett. 77(9), 1274–1277 (2000). [CrossRef]
18. C. J. Sun, M. Zubair Anwar, Q. Chen, J. W. Yang, M. Asif Khan, M. S. Shur, A. D. Bykhovski, Z. Liliental-Weber, C. Kisielowski, M. Smith, J. Y. Lin, and H. X. Xiang, “Quantum shift of band-edge stimulated emission in InGaN-GaN multiple quantum well light-emitting diodes,” Appl. Phys. Lett. 70(22), 2978–2980 (1997). [CrossRef]
19. I. H. Kim, H. S. Park, Y. J. Park, and T. Kim, “Formation of V-shaped pits in InGaN/GaN multiquantum wells and bulk InGaN films,” Appl. Phys. Lett. 73(12), 1634–1636 (1998). [CrossRef]
20. Y. Chen, T. Takeuchi, H. Amano, I. Akasaki, N. Yamada, Y. Kaneko, and S. Y. Wang, “Pit formation in GaInN quantum wells,” Appl. Phys. Lett. 72(6), 710–712 (1998). [CrossRef]
21. 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(Part 2, No. 4B), L371–L373 (2001).
22. M. S. Kumar, J. Y. Park, Y. S. Lee, S. J. Chung, C. H. Hong, and E. K. Suh, “Improved internal quantum efficiency of green emitting InGaN/GaN multiple quantum wells by In preflow for InGaN well growth,” Jpn. J. Appl. Phys. 47(2), 839–842 (2008). [CrossRef]
23. 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]
24. Y. J. Lee, Y. C. Chen, C. J. Lee, C. M. Cheng, S. W. Chen, and T. C. Lu, “Stable temperature characteristics and suppression of efficiency droop in InGaN green light-emitting diodes using pre-TMIn flow treatment,” IEEE Photon. Technol. Lett. 22(17), 1279–1281 (2010). [CrossRef]
25. S. W. Feng, C. Y. Tsai, H. C. Wang, H. C. Lin, and J. I. Chyi, “Optical properties of InGaN/GaN multiple quantum wells with trimethylindium treatment during growth interruption,” J. Cryst. Growth 325(1), 41–45 (2011). [CrossRef]
26. M. C. Johnsona, S. L. Konseka, A. Zettla, and E. D. Bourret-Courchesne, “Nucleation and growth of InN thin films using conventional and pulsed MOVPE,” J. Cryst. Growth 272(1–4), 400–406 (2004). [CrossRef]
27. M. Jamil, H. Zhao, J. B. Higgins, and N. Tansu, “Influence of growth temperature and V/III ratio on the optical characteristics of narrow band gap (0.77 eV) InN grown on GaN/sapphire using pulsed MOVPE,” J. Cryst. Growth 310(23), 4947–4953 (2008). [CrossRef]
28. M. Jamil, H. Zhao, J. B. Higgins, and N. Tansu, “MOVPE and photoluminescence of narrow band gap (0.77 eV) InN on GaN/sapphire by pulsed growth mode,” Phys. Status Solidi A 205(12), 2886–2891 (2008). [CrossRef]
29. C. Y. Chang, S. J. Chang, C. H. Liu, S. G. Li, and T. K. Lin, “GaN-based LEDs with double strain releasing MQWs and Si delta-doping layers,” IEEE Photon. Technol. Lett. 24(20), 1809–1811 (2012). [CrossRef]
30. J. M. Phillips, M. E. Coltrin, M. H. Crawford, A. J. Fischer, M. R. Krames, R. Mueller-Mach, G. O. Mueller, Y. Ohno, L. E. S. Rohwer, J. A. Simmons, and J. Y. Tsao, “Research challenges to ultra-efficient inorganic solid-state lighting,” Laser Photon. Rev. 1(4), 307–333 (2007). [CrossRef]
