Open Access
Add to BibSonomy
Abstract
Due to the increased surface-to-volume ratio, the surface recombination caused by sidewall defects is a key obstacle that limits the external quantum efficiency (EQE) for GaN-based micro-light-emitting diodes (µLEDs). In this work, we propose selectively removing the periphery p+-GaN layer so that the an artificially formed resistive ITO/p-GaN junction can be formed at the mesa edge. Three types of LEDs with different device dimensions of 30 × 30 µm2, 60 × 60 µm2 and 100 × 100 µm2 are investigated, respectively. We find that such resistive ITO/p-GaN junction can effectively prevent the holes from reaching the sidewalls for µLEDs with smaller size. Furthermore, such confinement of injection current also facilitates the hole injection into the active region for µLEDs. Therefore, the surface-defect-caused nonradiative recombination in the edge of mesa can be suppressed. Meantime, a reduction of current leakage caused by the sidewall defects can also be obtained. As a result, the measured and calculated external quantum efficiency (EQE) and optical output power for the proposed LED with small sizes are increased.
© 2021 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Thanks to the advantages in high resolution, low power consumption, high brightness, flexibility, fast response and high reliability, micro-size light-emitting diode (µLED)-based display technology is regarded as a promising candidate to replace conventional OLED and LCD display systems [1,2]. Therefore, µLED-based displays have great potential applications in e.g., smart watches, mobile phones, TVs, laptops, micro-projection displays, augmented reality and virtual reality [3,4]. Furthermore, because of the high modulation bandwidth and data safety, µLED also holds tremendous potential for visible light communication (VLC) [5,6]. However, at the current stage, although the marketing prospect for µLEDs is promising, there are still quite a few obstacles that require to be solved, such as mass transfer [7–9] and full-color conversion technology [10,11], insufficient emission efficiency [12]. According to our point of view, the design and fabrication of high-efficiency µLED display array shall be of the highest priority before making high-brightness µLED-based display system. Although the size reduction for LED devices will lead to a better light extraction efficiency, an improved current spreading and a reduced thermal heating effect [13,14], the decreased chip size increases the ratio of the perimeter-to-area in the meanwhile. Hence, unlike large-size LED devices, the surface nonradiative recombination caused by the sidewall defects is a non-negligible impact factor for µLEDs [15]. Therefore, the reported peak EQE for InGaN/GaN blue µLEDs is generally lower than 20% [16–18]. Aiming at suppressing the nonradiative recombination at surface imperfections, tremendous efforts have been made to repair the damaged surface. One of the methods that help reduce the defect density for sidewall defects is thermally annealing the devices, Tian et al. have attempted to increase the annealing time from 2 min to 3 min, which helps to suppress the surface nonradiative recombination rate and improve µLED performance [18]. However, the plasma-caused damages cannot be completely suppressed by the thermal annealing process. Hence, additional technique to annihilate the surface defects by sidewall passivation has been proposed, which is achieved by using the plasma-enhanced chemical vapor deposition (PECVD) or atomic layer deposition (ALD) systems [19,20]. Furthermore, to further eliminate the surface damages, sidewall chemical treatment is combined with the sidewall passivation during the manufacturing process [21]. Measured results indicate that the µLEDs after using ALD passivation and chemical treatment have a homogenized light emission across the mesa. Otherwise, no light emission can be observed from the 10 × 10 µm2 µLED. This is the manifestation of the strong suppression for sidewall defects by combining the processes of the ALD passivation and chemical treatment. Although all the mentioned techniques are utilized when fabricating µLEDs, it is less possible that sidewall defects can be completely suppressed. Hence, it is still necessary to take other measures to further suppress the surface nonradiative recombination. It is well known that µLEDs have better current spreading effect when compared with large-size LED device [15], which means that the current can easily reach the defected sidewall regions. One effective approach is adopting a buried tunneling junction (TJ) to keep the injection current apart from the mesa edges [22], and hence, a better current confinement can be achieved. Besides, a structure with an insulated aperture that is similar to GaN-based vertical-cavity-surface-emission laser diodes (VCSELs) also present a remarkable current confinement capability [23]. However, the buried tunnel junction and the aperture design are both technically challenging from the point of experimental view. Then, a simple and straightforward way is tuning the current spreading length by manipulating the vertical resistivity for LEDs [24]. Chang et al. have reported that by properly decreasing the quantum barrier thickness, a reduced vertical resistivity can be obtained, so that the current does not tend to spread to the mesa edges. As a result, the surface nonradiative recombination can be significantly suppressed according to the experimental measurement and numerical simulations. Furthermore, µLEDs have extremely poor hole injection capability because of the hole capturing effect by the surface defects [15]. Then, thinner quantum barriers also help to promote the hole injection for µLEDs [24].
In this work, an alternative approach to confine the current injection has been experimentally and numerically demonstrated. We selectively remove the p+-GaN layer in sidewall region, and then the ITO layer is directly contacted on the exposed p-GaN layer. By doing so, an artificially formed resistive ITO/p-GaN junction can be formed at the sidewall region, which is helpful to prevent the holes from arriving at the mesa sidewalls. It means that the current can be effectively confined in the non-defected mesa area, which suppresses the Shockley-Read-Hall (SRH) nonradiative recombination in the mesa edge. Furthermore, the confinement of current also enhances hole injection in the proposed LED structure. We have designed and fabricated three types of LED with different device dimensions of 100 × 100 µm2, 60 × 60 µm2 and 30 × 30 µm2, respectively. When compared with their reference counterparts, the electroluminescent (EL) spectra, optical power and the EQE are both improved for the LED with small sizes. However, the contrary conclusion is obtained such that the 100 × 100 µm2 LED with the ITO/p-GaN resistive junction shows the decreased EQE and the reduced optical output power when compared with the reference LED with the same device size. The poorer performance is attributed to the worse current spreading and additional junction heat when the sidewall ITO/p-GaN reverse junction is employed.
2. Numerical calculations and experimental measurements
To experimentally prove our point, the InGaN/GaN LED samples are grown on c-sapphire substrates by metal organic chemical-vapor deposition (MOCVD) system. Firstly, a 20 nm thick GaN buffer layer is grown on the C-plane sapphire substrates. Then, a 4 µm thick unintentional n-doped GaN layer is grown, which is followed by a 2 µm thick n-GaN layer with the electron concentration of 5 × 1018 cm−3. Next, five-period In0.15Ga0.85N/GaN multiple quantum wells (MQWs) are grown. The thicknesses for the quantum barrier and the quantum well are 12 nm and 3 nm, respectively. On the top of the active region, a 25 nm thick p-type Al0.15Ga0.85N layer is designed and serves as the p-type electron blocking layer (p-EBL). The hole concentration in the p-EBL is estimated to be 2 × 1017 cm−3. Subsequently, an 80 nm thick p-GaN layer with the hole concentration of 3 × 1017 cm−3 is employed as the hole injection layer. Next a 20-nm-thick heavily doped p+-GaN layer with the hole concentration of 3 × 1019 cm−3 is utilized for forming the p-type ohmic contact.
After growing the epi-layers in MOCVD system, the LED epi-wafer is processed into LED chips with different device sizes of 30 × 30 µm2, 60 × 60 µm2 and 100 × 100 µm2, respectively. For the reference LEDs, the LED mesas are dry-etched by using inductively coupled plasma (ICP) system. Then an indium tin oxide (ITO) current spreading layer of 30 nm is deposited by utilizing the atomic layer deposition (ALD) system. Next, the rapid thermal annealing (RTA) process is performed in the N2 ambient at the temperature of 550 °C for 3 minutes to achieve better p-type ohmic contact between the ITO layer and the p+-GaN layer [25]. Ti/Al/Ti/Au (20 nm/30 nm/60 nm/100 nm) and Ni/Au (10 nm/200 nm) are deposited by e-beam evaporation serving as the n-contact and p-contact, respectively. For the proposed devices, the same fabrication processes are adopted except that a 20-nm-thick and 4-µm-wide p+- GaN region around the device mesa is removed by using ICP system before depositing the ITO layer. For better illustration, the schematic diagrams for the reference devices and the proposed devices are shown in Figs. 1(a) and 1(b), respectively. Table 1 also shows titles for the investigated LEDs with different sizes and structures.
Fig. 1. Schematic diagrams for the reference device(a) and proposed device(b). (I), (II) and (III) show the emitting LED devices with different chips sizes at the applied voltage of 8V.
Table 1. Information of device size and device structure for LEDs A1, A2, B1, B2, C1 and C2
In order to reveal the device physics in the in-depth level, we also conduct numerical simulations by using APSYS, which is able to solve current continuity equations, Poisson equations and Schrödinger equations with proper boundary conditions. The band offset ratio between the conduction band and valence band for InGaN/InGaN quantum wells is set to 0.70/0.30 [26]. The Auger recombination coefficient is set to 1 × 10−30 cm6s−1 and the SRH nonradiative recombination lifetime is assumed to be 100 ns [27]. In addition, considering the existence of sidewall defects, the nonradiative recombination occurring at mesa surfaces are specially set in our models for µLEDs. In our sidewall defect model, the trap levels for electrons and holes are set at 0.24 eV below the conduction band (i.e., Ec - 0.24 eV) and 0.46 eV above the valence band (i.e., Ev + 0.46 eV), respectively. The capture cross section of 3.4 × 10−17 cm2 and the trap density of 1 × 1013 cm−3 are assumed for electron traps [28]. The capture cross section of 2.1 × 10−15 cm2 and the trap density of 1.6 × 1013 cm−3 are set for holes [29]. The width of defected region at the sidewall is empirically set to 4 µm [15]. Other parameters can be found elsewhere [30].
3. Results and discussion
The measured EL spectra at varying injection current density levels for different LED devices are shown in Fig. 2. It can be found from Figs. 2(a) and 2(b) that the EL intensity for Device A2 has been enhanced when compared with Device A1. This is due to the fact that the resistive ITO/p-GaN junction in sidewall region prevents the current from spreading to the mesa edge, which leads to the reduced SRH recombination in the sidewall region. As a result, the improved hole injection can be obtained in the device. Further investigations into the Figs.2(c) and 2(d) also infer that the EL intensity for Device B2 is slightly higher than the Device B1, i.e., such emission intensity enhancement becomes less obvious when the LED size increases. Nevertheless, when we compare Devices C1 and C2 with the large dimension of 100 × 100 µm2, it shows that Device C1 has a higher EL intensity when compared with its counterpart. This indicates that sidewall non-radiative SRH recombination plays a small role for large LEDs. Moreover, the poor current spreading and the accompanying Joule heating effect are believed to cause the degraded EL intensity for Device C2.
Fig. 2. Measured electroluminescence spectra for different LED devices in terms of various injection current density levels.
We then experimentally and numerically obtain the optical power and EQE as the function of the injection current density for different devices as shown in Fig. 3. Our calculated data have similar trend as the experimental ones, which indicates that our computational model is reliable. According to the results in Figs. 3(a) and 3(b), both the EQE and the optical power for Device A2 are larger than that for Device A1 experimentally and numerically. Further observations into Figs. 3(c) and 3(d) indicate that the enhancement of the EQE and optical power becomes less significant. The opposite trends are obtained for Devices C1 and C2, such that Device C2 presents the worse EQE and optical power than Device C1. The results demonstrated by Figs. 3(a)-3(f) are consistent with the EL profiles in Fig. 2. It is worth noting that the calculated EQE and the optical power show the deviation from the experimentally measured ones, and this is likely due to the less optimized physical parameters for III-nitrides, e.g., the trap energy levels at mesa surfaces might vary across the wafer especially for small chips; the light extraction efficiency is possibly to be affected by the chip size. Nevertheless, the calculated trending details are consistent with the experimental ones.
Fig. 3. (a) Measured and calculated EQEs and optical power in terms of various injection current density for difference devices. The chip sizes for (a)/(b), (c)/(d) and (e)/(f) are 30 × 30 µm2, 60 × 60 µm2 and 100 × 100 µm2, respectively. (g) Measured EQE droop for different devices and (h) measured and calculated power enhancement for the proposed LEDs when compared with the reference LEDs.
Furthermore, the measured efficiency droop at the current density level of ∼ 500 A/cm2 for different devices are also shown in Fig. 3(f). We can observe that the efficiency droop becomes less obvious as the chip size gradually decreases for all the investigated LEDs, which is the signature for the even more remarkable defect-induced nonradiative recombination rate when the chip size becomes small, such that most holes are captured by sidewall defects and the Auger recombination in the MQWs cannot be triggered [31]. Meanwhile, the low injection also gives rise to the stronger electron leakage level [32]. However, if we compare Devices A1 and A2, it can be found that Device A2 shows a reduced efficiency droop, which is attributed to the fact that the resistive ITO/p-GaN junction can concentrate the holes at the central undefective area and then enhance the hole injection into the active region therein. Such enhanced hole injection efficiency enables a reduced electron leakage level and the decreased efficiency droop. Devices B1 and B2 show negligible difference in the efficiency droop, and this illustrates that the impact of resistive ITO/p-GaN junction on the hole injection efficiency becomes small as the chip size increases. Further analysis for Devices C1 and C2 shows that Device C2 shows the even larger efficiency droop. This is ascribed to the current crowding effect when the resistive ITO/p-GaN junction is adopted, and the current crowding effect also generates more Joule heating effect and the thermal droop is yielded [32]. Furthermore, in order to show the difference of optical power between the reference devices and proposed devices more clearly, the power enhancement at the current density of 500 A/cm2 is shown in Fig. 3(h). The experimentally measured results show that adopting a resistive ITO/p-GaN junction can improve the optical power for Devices A2 and B2. We believe that such enhancement will become more significant as the chip size further decreases.
Then, the measured current-voltage (I-V) characteristics for different devices are shown in Fig. 4. When the devices are operated in the reverse bias condition, Device A2 possesses a smaller leakage current than Device A1 [see Fig. 4(a)]. the difference of the leakage current between Devices B1 and B2 becomes small when compared with that for Devices A1 and A2 [see Fig. 4(c)]. Such difference becomes even smaller for Devices C1 and C2 according to Fig. 4(e). This is due to the fact that the impact of sidewall defects on the leakage current becomes small with the chip size increases. Furthermore, it is worth noting that the leakage current intensity for Devices A1 and A2 with the chip size of 30 × 30 µm2 are ∼ 10−2 A/cm2, which is much higher than that for Devices B and C with the chip size of 60 × 60 µm2 and 100 × 100 µm2. This indicates that leakage current caused by sidewall defects has a great impact for LEDs with small sizes, which further confirms that the contribution of the resistive ITO/p-GaN junction for decreasing the sidewall non-radiative SRH recombination and can help to decrease the leakage current for µLEDs with even smaller size. We believe that such enhancement becomes more apparent if LED size further decreases. In addition, it is worth noting that the leakage current for Devices C1 and C2 is ∼ 10−3 A/cm2, which is higher than ∼ 10−4 A/cm2 for Devices B1 and B2, which is likely attributed to the more bulky defect-induced leakage current when the mesa size increase. When the applied voltage ranges from 0 V to 10 V, Device A2 shows the smaller forward current than Device A1 [see Fig. 4(b)], which is attributed to the resistive sidewall ITO/p-GaN junction. When chip size increase to 60 × 60 µm2, the impact of the resistive ITO/p-GaN junction on the forward conduction becomes reduced [see Fig. 4(d)]. Such impact is negligible for Device C2 with the size of 100 × 100 µm2 according to Fig. 4(f).
Fig. 4. Measured current-voltage (I-V) characteristics for Devices A in (a) and (b), Devices B in (c) and (d) and Devices C in (e) and (f), respectively. The current density profiles are in logarithmic scale and linear scale under the reverse bias and forward bias, respectively.
To even better reveal that the impact of the adopted resistive ITO/p-GaN junction on the confinement of injection current. Firstly, we show the two-dimensional (2-D) hole distribution profiles selectively for Devices A1 and A2 in Figs. 5(a) and 5(b) at the current density of 500 A/cm2, respectively. It can be found that the hole concentration in the sidewall region of Device A2 is significantly lower than that in Device A1, which proves that the resistive ITO/p-GaN junction can effectively prevent holes from spreading to the sidewall region. Then, we calculate the SRH nonradiative and radiative recombination rate at the current density of 500 A/cm2 in the first quantum well closest to the p-EBL for Devices A1 and A2 in Fig. 5(c). It indicates that Device A2 possesses a smaller SRH non-radiative recombination in the mesa edge and a higher radiative recombination rate in the center of device when compared with the Device A1. This numerically proves that the resistive ITO/p-GaN junction can help to make the injection current apart from the mesa edge and suppress the sidewall non-radiative SRH recombination rate for µLEDs. Moreover, according to Fig. 5(d), such current confinement also facilitates the hole injection into the non-defected central device region and enhances the radiative recombination therein. The simulation results here also confirm that the sidewall non-radiative SRH recombination must be taken into account for LED with small size. Besides, the sidewall non-radiative SRH recombination can be suppressed by utilizing the resistive ITO/p-GaN junction.
Fig. 5. Calculated 2-D hole distribution profiles for Device A1 (a) and Device A2 (b). SRH and radiative recombination rate in the last quantum well closest to p-EBL for Device A1 and A2 (c). Lateral hole distribution in the last quantum well for Devices A1 and A2 (d). The data for (a), (b), (c) and (d) are calculated at the injection current density of 500A/cm2.
To further reveal the impact of internal joule heating on the performance for the investigated devices. Our program can demonstrate the temperature range in the device structure. We thus calculate the minimum/maximum temperatures for different devices at the current density of 500 A/cm2 in Fig. 6. We find that the device temperature increases with the increased device size. The biggest junction temperature for Device C2 interprets the thermal droop in Figs. 3(e) and 3(f). Moreover, the proposed devices with the resistive ITO/p-GaN junctions have the increased device temperature when compared with their counterparts, which means that the resistive ITO/p-GaN layer may cause more Joule heating effect due to the less spreaded current in the mesa region. Hence, we believe that if the heat can be better dissipated, the µLEDs can even further promote the EQE and the optical power. Note, unlike LEDs with small device sizes, the sidewall defect-induced nonradiative recombination rate is less significant for Devices C1 and C2. Hence, the Joule heating and the current crowding effects shall need even more concern for conventional LEDs with big sizes.
Fig. 6. Simulated minimum and maximun device temperatures for each device at the current density of 500 A/cm2.
4. Conclusion
In this work, aiming to suppress the SRH nonradiative recombination caused by the sidewall defects, the µLED with the resistive ITO/p-GaN junction is investigated both experimentally and numerically. It is proven that the sidewall resistive ITO/p-GaN junction can effectively makes the injection current apart from the mesa edge. Thanks to the confinement for the injection current, the enhanced hole injection and a suppressed SRH nonradiative recombination can be obtained for LED device with small size. Accordingly, the EQE and the optical power can be both improved. However, for the LED with the mesa size of 100 × 100 µm2, the resistive ITO/p-GaN junction may cause the current crowding effect, which will then give rise to the increased Joule heating effect. Thus, a reduced EQE and the efficiency rollover can be observed. Therefore, we strongly believe that the proposed resistive ITO/p-GaN layers can be very useful for making high-efficiency µLEDs.
Funding
National Natural Science Foundation of China (62074050); Natural Science Foundation of Hebei Province (F2020202030); State Key Laboratory of Reliability and Intelligence of Electrical Equipment, Hebei University of Technology (EERI_PI2020008); Joint Research Project for Tunghsu Group and Hebei University of Technology (HI1909).
Disclosures
The authors declare no conflicts of interest related to this paper.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. X. Jiang and J. Y. Lin, “Nitride micro-LEDs and beyond–a decade progress review,” Opt. Express 21(S3), A475–484 (2013). [CrossRef]
2. T. Wu, C.-W. Sher, Y. Lin, C.-F. Lee, S. Liang, Y. Lu, S.-W. Huang Chen, W. Guo, H.-C. Kuo, and Z. Chen, “Mini-LED and Micro-LED: Promising Candidates for the Next Generation Display Technology,” Appl. Sci 8(9), 1557 (2018). [CrossRef]
3. N. Li, K. Han, W. Spratt, S. Bedell, J. Ott, M. Hopstaken, F. Libsch, Q. Li, and D. Sadana, “Ultra-low-power sub-photon-voltage high-efficiency light-emitting diodes,” Nat. Photonics 13(9), 588–592 (2019). [CrossRef]
4. G. Tan, Y. Huang, M. C. Li, S. L. Lee, and S. T. Wu, “High dynamic range liquid crystal displays with a mini-LED backlight,” Opt. Express 26(13), 16572–16584 (2018). [CrossRef]
5. N. Arvanitakis, R. Bian, J. J. D. McKendry, C. Cheng, E. Xie, X. He, G. Yang, M. S. Islim, A. A. Purwita, E. Gu, H. Haas, and M. D. Dawson, “Gb/s Underwater Wireless Optical Communications Using Series-Connected GaN Micro-LED Arrays,” IEEE Photonics J 12(2), 1–10 (2020). [CrossRef]
6. M. Monavarian, A. Rashidi, A. A. Aragon, S. H. Oh, A. K. Rishinaramangalam, S. P. DenBaars, and D. Feezell, “Impact of crystal orientation on the modulation bandwidth of InGaN/GaN light-emitting diodes,” Appl. Phys. Lett 112(4), 041104 (2018). [CrossRef]
7. B. Corbett, R. Loi, W. Zhou, D. Liu, and Z. Ma, “Transfer print techniques for heterogeneous integration of photonic components,” Prog. Quantum Electron 52, 1–17 (2017). [CrossRef]
8. J. Yoon, S. Jo, I. S. Chun, I. Jung, H. S. Kim, M. Meitl, E. Menard, X. Li, J. J. Coleman, U. Paik, and J. A. Rogers, “GaAs photovoltaics and optoelectronics using releasable multilayer epitaxial assemblies,” Nature 465(7296), 329–333 (2010). [CrossRef]
9. L. Zhang, F. Ou, W. C. Chong, Y. Chen, and Q. Li, “Wafer-scale monolithic hybrid integration of Si-based IC and III-V epi-layers-A mass manufacturable approach for active matrix micro-LED micro-displays,” J. Soc. Inf. Disp 26(3), 137–145 (2018). [CrossRef]
10. J. He, H. Chen, H. Chen, Y. Wang, S. T. Wu, and Y. Dong, “Hybrid downconverters with green perovskite-polymer composite films for wide color gamut displays,” Opt. Express 25(11), 12915–12925 (2017). [CrossRef]
11. M. Kim, M. Ryu, J. H. J. Cha, H. S. Kim, T. Jeong, J. Jang, and Ten micrometer pixel, “Quantum dots color conversion layer for high resolution and full color active matrix micro-LED display,” J. Soc. Inf. Disp 27(6), 347–353 (2019). [CrossRef]
12. S. Hang, C.-M. Chuang, Y. Zhang, C. Chu, K. Tian, Q. Zheng, T. Wu, Z. Liu, Z.-H. Zhang, Q. Li, and H.-C. Kuo, “A review on the low external quantum efficiency and the remedies for GaN-based micro-LEDs,” J. Phys. D: Appl. Phys 54(15), 153002 (2021). [CrossRef]
13. W. Choi, C. W. Jeon, M. D. Dawson, P. R. Edwards, R. W. Martin, and S. Tripathy, “Mechanism of enhanced light output efficiency in InGaN-based microlight emitting diodes,” J. Appl. Phys 93(10), 5978–5982 (2003). [CrossRef]
14. L. Ploch, H. Rodriguez, C. Stolmacker, M. Hoppe, M. Lapeyrade, J. Stellmach, F. Mehnke, T. Wernicke, A. Knauer, V. Kueller, M. Weyers, S. Einfeldt, and M. Kneissl, “Effective thermal management in ultraviolet Light-Emitting Diodes With Micro-LED Arrays,” IEEE Trans. Electron Devices 60(2), 782–786 (2013). [CrossRef]
15. J. Kou, C.-C. Shen, H. Shao, J. Che, X. Hou, C. Chu, K. Tian, Y. Zhang, Z.-H. Zhang, and H.-C. Kuo, “Impact of the surface recombination on InGaN/GaN-based blue micro-light emitting diodes,” Opt. Express 27(12), A643–A653 (2019). [CrossRef]
16. Y. Huang, G. Tan, F. Gou, M.-C. Li, S.-L. Lee, and S.-T. Wu, “Prospects and challenges of mini-LED and micro-LED displays,” J. Soc. Inf. Disp 27(7), 387–401 (2019). [CrossRef]
17. F. Olivier, S. Tirano, L. Dupré, B. Aventurier, C. Largeron, and F. Templier, “Influence of size-reduction on the performances of GaN-based micro-LEDs for display application,” J. Lumin 191, 112–116 (2017). [CrossRef]
18. P. Tian, J. J. D. McKendry, Z. Gong, B. Guilhabert, I. M. Watson, E. Gu, Z. Chen, G. Zhang, and M. D. Dawson, “Size-dependent efficiency and efficiency droop of blue InGaN micro-light emitting diodes,” Appl. Phys. Lett 101(23), 231110 (2012). [CrossRef]
19. R. T. Ley, J. M. Smith, M. S. Wong, T. Margalith, S. Nakamura, S. P. DenBaars, and M. J. Gordon, “Revealing the importance of light extraction efficiency in InGaN/GaN microLEDs via chemical treatment and dielectric passivation,” Appl. Phys. Lett 116(25), 251104 (2020). [CrossRef]
20. C.-M. Yang, D.-S. Kim, S.-G. Lee, J.-H. Lee, Y. S. Lee, and J.-H. Lee, “Improvement in Electrical and Optical Performances of GaN-Based LED With SiO2/Al2O3 Double Dielectric Stack Layer,” IEEE Electron Device Lett 33(4), 564–566 (2012). [CrossRef]
21. M. S. Wong, D. Hwang, A. I. Alhassan, C. Lee, R. Ley, S. Nakamura, and S. P. DenBaars, “High efficiency of III-nitride micro-light-emitting diodes by sidewall passivation using atomic layer deposition,” Opt. Express 26(16), 21324–21331 (2018). [CrossRef]
22. M. Malinverni, D. Martin, and N. Grandjean, “InGaN based micro light emitting diodes featuring a buried GaN tunnel junction,” Appl. Phys. Lett 107(5), 051107 (2015). [CrossRef]
23. T.-C. Lu, C.-C. Kao, H.-C. Kuo, G.-S. Huang, and S.-C. Wang, “CW lasing of current injection blue GaN-based vertical cavity surface emitting laser,” Appl. Phys. Lett 92(14), 141102 (2008). [CrossRef]
24. L. Chang, Y.-W. Yeh, S. Hang, K. Tian, J. Kou, W. Bi, Y. Zhang, Z.-H. Zhang, Z. Liu, and H.-C. Kuo, “Alternative strategy to reduce surface recombination for InGaN/GaN Micro-light-emitting diodes—thinning the Quantum barriers to manage the current spreading,” Nanoscale Res. Lett 15(1), 1–9 (2020). [CrossRef]
25. H. L. Hao, L. K. Wu, W. J. Chung, Y. Zhang, and W. Z. Shen, “Process optimization of RTA on the characteristics of ITO-coated GaN-based LEDs,” Microelectron. Reliab. 55(11), 2263–2268 (2015). [CrossRef]
26. J. Piprek and S. Nakamura, in IEE Proceedings - Optoelectronics (2002), Vol. 149, pp. 145.
27. M. Meneghini, N. Trivellin, G. Meneghesso, E. Zanoni, U. Zehnder, and B. Hahn, “A combined electro-optical method for the determination of the recombination parameters in InGaN-based light-emitting diodes,” J. Appl. Phys 106(11), 114508 (2009). [CrossRef]
28. L. Rigutti, A. Castaldini, and A. Cavallini, “Anomalous deep-level transients related to quantum well piezoelectric fields in InyGa1-yN∕GaN-heterostructure light-emitting diodes,” Phys. Rev. B 77(4), 045312 (2008). [CrossRef]
29. T. Narita, Y. Tokuda, T. Kogiso, K. Tomita, and T. Kachi, “The trap states in lightly Mg-doped GaN grown by MOVPE on a freestanding GaN substrate,” J. Appl. Phys 123(16), 161405 (2018). [CrossRef]
30. I. Vurgaftman and J. R. Meyer, “Band parameters for nitrogen-containing semiconductors,” J. Appl. Phys 94(6), 3675–3696 (2003). [CrossRef]
31. F. Olivier, A. Daami, C. Licitra, and F. Templier, “Shockley-Read-Hall and Auger non-radiative recombination in GaN based LEDs: A size effect study,” Appl. Phys. Lett 111(2), 022104 (2017). [CrossRef]
32. J. Cho, E. F. Schubert, and J. K. Kim, “Efficiency droop in light-emitting diodes: Challenges and countermeasures,” Laser Photonics Rev 7(3), 408–421 (2013). [CrossRef]
