Open Access
Add to BibSonomy
Abstract
A new process is presented for fabricating enhanced-efficiency micro-pixelated vertical-structured light-emitting diode (µVLED) arrays based on ion-implantation technology. High-resistivity selective regions are locally introduced in the n-GaN layer by ion implantation and then used as effective and non-destructive electrical isolation for realizing µVLED arrays with ultra-small pixel diameters. The implantation energy-dependent and size-dependent opto-electrical characteristics of fluorine (F-) implanted µVLED arrays are investigated systematically. The results show that the optimally designed F- ion implantation not only can achieve smaller reverse leakage current but also can realize ion-induced thermal relaxation effectively and is more suited for fabricating high-resolution µVLED arrays with higher optical output power. For the F--implanted µVLED array with pixel diameters of 10 µm, a measured output power density reaches a value of 82.1 W cm−2 at a high injection current density of 220 A cm−2, before power saturation. Further, the output power densities and external quantum efficiencies of F--implanted µVLED arrays with pixel diameters less than 10µm show strong dependences on pixel size due to the presence of defects-related SRH process. So, the high-efficiency µVLED arrays with ultra-small pixel sizes could be fabricated by an appropriately designed ion implantation combined with control of defect densities to meet the industrial requirement of microdisplay applications.
© 2021 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Recently, the nitride-based vertical-structured micropixellated light-emitting diode (µVLED) arrays have gained increasing research interests due to versatile applications in high-speed visible-light communication [1], self-emissive microdisplays [2–5], fluorescence-imaging-based measurement [6], and optogenetic neuromodulation [7]. For the state-of-the-art µVLED arrays, the mesa micro-fabrication processes are widely used as primary technologies to scale down the light-emitting pixel dimensions to around 10 microns or smaller, and then combining with mass-transfer technology to realize red-green-blue full-color microdisplays [8–10]. Nevertheless, due to disadvantages of high fabrication cost, low transfer velocity, and difficulty of inspection, the mass-transfer technology is still not available for mass production and commercialization of µVLED arrays [11]. Meanwhile, the plasma-assisted dry etching existed in traditional mesa isolation process, such as BCl3/Cl2-based inductively coupled plasma (ICP) etching, can show high possibility to form plasma bombardment on mesa sidewalls and causes plasma-induced sidewall damages, which may greatly increase the non-radiative recombination in multiple quantum wells (MQWs) active region [12–16]. Moreover, the sidewall damages forms lattice defects and N vacancies, which would lead µVLED arrays to suffer severely from surface leakage [17]. What is worse, sidewall damages significantly decrease the overall yield across LED wafer, and even more serious for smaller pixel dimensions. Taking a 20×20 µm2 µVLED for example, the available mesa area only accounts for 36% of the total mesa size [18].
To resolve the problems mentioned above, it is essential to develop new methods to fabricate µVLED arrays. In particular, the planar process such as ion implantation has unparalleled advantages in improving process reliability and yield. Generally, the mesa isolation based on ion implantation is realized by ion-induced lattice disorder, which can introduce deep defect levels causing an electrical conductivity decrease and achieving highly-resistive regions in nitride-based semiconductors. Researchers have achieved GaN-based HEMTs with enhanced performances based on ion-implantation processes by which there are almost no studies on fabricating µVLED arrays [19–21]. Usually, the heavy mass ions can introduce large displacement during ion implantation, which make the lattice crystalline hard to recover quality completely even through high-temperature thermal annealing and thus achieve higher thermal stability. For instance, the fluorine (F-) implant-isolation even can maintain thermal stability up to 550°C due to the formation of Ga-F bonds [22]. By contrast, the light ions species (including H+, He+) are known for having poor thermal stability, which may limit their practical applications for long-term moderate temperature operation [23,24].
In this work, a new process is presented to realize effective electrical isolation for fabricating µVLED arrays by F- ion implantation, via which the light emitted from MQWs underneath the F--implanted highly-resistive regions is intentionally limited, thereby leading to the formation of µVLED array. And then the electrical characteristics, luminescence performances and efficiency droops of F- ion-implanted µVLED arrays with different light-emitting pixel sizes are systematically analyzed under multiple ion-implantation energies. Finally, for realizing monolithic microdisplay, our work also proposes a new scheme that uses ion-implanted µVLED arrays with complementary metal oxide semiconductor (CMOS) active matrix driver according to the industrial requirement of microdisplay applications.
2. Experimental
The investigated LED emitting with a specified wavelength of 460 nm was grown by metal-organic chemical vapor deposition on c-plane sapphire substrate, consisting of a GaN nucleation layer, a unintentionally-doped GaN buffer layer, a 3-µm-thick n-GaN layer, a five-period InGaN/GaN MQWs layer, a 60-nm-thick p-AlGaN electron-blocking layer, and a 70-nm-thick p-GaN layer. For the fabrication of VLED structure, a mirror system consisted of an indium-tin oxide (ITO) contact layer, an Al2O3 low index layer, a Pt metal layer, and a Ti diffusion barrier layer was successively deposited on p-GaN surface, followed by depositing an Au-Sn metallic bonding layer. Further, the VLED epi-wafer was bonded to silicon receptor substrate using thermo-compression bonding process under a comprehensive load of 10 kg/cm2. After the sapphire substrate was removed by laser lift-off process, a chemical mechanical polishing process was used to remove the GaN buffer layer and reduce the thickness of n-GaN layer to about 0.6 µm. Prior to the fabrication of µVLED arrays, ICP etching were used to form 800×800 µm2 mesas for ion-implantation processes which were carried out with a 7° tilt-angle using a NV-GSD-HE high-energy ion implanter at room temperature. More particularly, SiO2 layer rather than traditional photoresist layer was chosen as hard mask to protect the unimplanted region during the ion-implantation process. The transport-of-ions-in-matter software was used to simulate the implantation depth profile in VLED structure, and then a single-energy (90 KeV) and a dual-energy (90/180 KeV) for F- ions with a dose of 1×1015 ions/cm2, respectively, were selected for realizing the high-selective F--implanted isolation in the n-GaN layer. For systematic comparison, the F--implanted µVLED arrays having circular pixels with same center-to-center spacing of 20 µm were designed with diameters of 6 µm, 8 µm and 10 µm, respectively. Finally, a 200-nm-thick ITO conduction layer and a Ti/Au (50/150 nm) metal electrode layers were successively deposited onto the n-GaN layer and subsequently annealed at 500°C with nitrogen gas protection. Figure 1 shows the structural illustrations of ion-implanted VLED and µVLED array, which consists of 25×25 pixels with a corresponding display resolution of 1270 PPI.
The diffusion characteristics of ions in VLED structure under different implantation energies are identified by secondary ion mass spectrometry (SIMS). The electrical characteristics of F--implanted regions are characterized by circular transmission line model (CTLM) measurement. The electrical characteristics of F--implanted µVLED arrays with different pixel sizes under multiple implantation energies are clarified by current-voltage (I-V) measurements using an Agilent B1505A analyzer. The optical properties of F--implanted µVLED arrays are investigated by electroluminescence (EL) spectroscopy and light output power-current (L-I) measurement using integrating sphere system.
3. Results and discussion
Figure 2 shows the diffusions of ions in the F--implanted VLED structure. Owing to the ion implantation depth profiles are determined by implantation energy, the single-energy implantation with a lower energy (90 KeV) was employed to isolate the near-surface regions of n-GaN layer only, while the dual-energy implantation utilizes an additional higher energy (180 KeV) to isolate the n-GaN up to a deeper depth. As shown in Fig. 2, the F- SIMS depth profile of the dual-energy implantation fitted by Gaussian lineshapes can be resolved into two peaks located at about 0.16 µm and 0.33µm, respectively, which correspond to two sequentially performed ion implantations. On the other hand, as indicated by the drastic drop of distribution intensity in Si+ SIMS depth profile, the thickness of n-GaN layer for VLED structure is determined to be about 0.6 µm. It is noted that the depth profile distribution of F- ions for both the single-energy and dual-energy implantation exist almost entirely in the n-GaN layer, and thus could greatly avoid introducing ion-induced lattice damages and related new non-radiative recombination centers into MQWs region, which is critical to improving emissive capabilities for the F--implanted µVLED arrays.
The effect of implantation energies on electrical isolation of F--implanted n-GaN regions were further tested by CTLM method, which employed Ti/ITO bilayered contacts with inner radius (r0) of 70 µm and outer radii, ri (i=1 through 6), increasing from 80 µm to 120 µm. For reference, a same Ohmic contact fabricated on unimplanted n-GaN was also measured with identical test structure. The total resistances (Rtot) of both Ti/ITO contacts on F--implanted n-GaN were measured and the related electrical parameters can be calculated by linear fitting the CTLM data to the following equation [25], and summarized in Table 1.
Table 1. Electrical characteristics of Ti/ITO contacts on F--implanted and unimplanted n-GaN obtained by linear fitting CTLM data.
Next, the electrical characteristics of the F--implanted µVLED arrays with pixel diameters of 10 µm were investigated by I-V measurement. As shown in Fig. 3(a), at a reverse bias voltage of −5 V, the measured reverse leakage currents of µVLED arrays based on the single and dual-energy is 3.7×10−8 A and 3.9×10−9 A, respectively, which are far less than that of a traditional broad-area VLED (9.7×10−6 A) fabricated by ICP etching based on a same epi-wafer in our previous report [17]. Actually, the lattice defects, gallium dangling bonds and N vacancies introduced during ICP etching can lead to the formation of leakage pathways through which carriers tend to flow, and thus increase leakage current for a conventional VLED. By contrast, the F- implant-isolated regions existed in n-GaN can insulate carriers from leakage pathways located at etched sidewall, which in turn enhance the uniformity of carrier injection and result in a smaller reverse leakage current. Consequently, the substantial decreases of reverse leakage for both µVLED arrays can be attributed to the F- ion-induced variations of carrier transport pathways. In addition, the leakage current related to the dual-energy implantation is lower by one order of magnitude than that to the single-energy implantation at −5 V, indicating that deeper ion implantation is more suitable for improving reverse leakage for Micro-VLED arrays related to the more effective electrical isolation. As shown in Fig. 3(b), the typical forward voltage at 20 mA and the series resistance extracted from differential resistance for the single-energy F--implanted µVLED array are 3.09 V and 12.5 Ω, and for the dual-energy F--implanted µVLED array, 2.83 V and 8.33 Ω, respectively. Thus, the dual-energy F- implantations can not only optimize the leakage behavior but also can improve the forward electrical characteristic, which is favorable to fabricating the µVLED arrays with high wall-plug efficiency.
Fig. 3. Opto-electronic performances of F--implanted µVLED arrays under the single and dual-energy ion implantations: (a) reverse and (b) forward I-V characteristics, (c) light-output characteristics.
Subsequently, the light-output characteristics for F--implanted µVLED arrays with pixel diameter of 10 µm were investigated by integrating sphere measurement. As shown in Fig. 3(c), the optical output powers of both F--implanted µVLED arrays increase monotonously with increasing current, and can sustain high injection level before power saturation. As mentioned above, the carriers are forced to be redistributed due to the presence of F--implanted high-resistivity regions under which the light emission is limited, and thus create ion-induced cold zones. Therefore, the heat generated from the µVLED arrays under high injection conditions can be dissipated into the cold zones and realizes effective thermal relaxation, which can reduce the average junction temperature and significantly enhances the light-emission capabilities for the F--implanted µVLED arrays [26–28]. Moreover, the µVLED array based on dual-energy implantation can achieve substantially higher optical output under same current injection conditions. For example, an output power density of 82.1 W/cm2 can be achieved for the dual-energy implantation under an injection current density of 220 A/cm2, 1.36 times as large as that for the one with single-energy (60.2 W/cm2), indicating that the deeper ion implantation can actualize a more effective ion-induced thermal relaxation. Taken together, combining the studies on the CTLM, IV and LI measurements, it confirmed that the dual-energy F- implantation not only can achieve smaller reverse leakage current but also can realize ion-induced thermal relaxation more effectively, which is better suited for fabricating µVLED arrays with higher efficiency.
Once the ion-implantation energies were optimized, a special focus was putted on the size-reduction effect on the opto-electrical properties of the F--implanted µVLED arrays based on the dual-energy implantation schedule. The light emission uniformity of F--implanted µVLED arrays with pixel diameters ranging from 6 µm to 10 µm were illustrated by EL images operating at a driving current of 20 mA. It can be seen clearly in Fig. 4 that all µVLED arrays realize high-resolution display characteristics. However, the µVLED array with 6 µm pixels exhibits emission inhomogeneity with the appearing of lots of dim pixels for which lights are emitted with relatively low intensities, while the µVLED arrays with larger pixels show significantly more homogeneous light emission. Usami et al. experimentally showed that the closed-core pure screw dislocations with Burgers vector buvtw = [0001] formed in medium pits (diameters ranging from 1 nm to 2.3 nm) act as current leakage channel which will degrade the performances of vertical devices [29]. Therefore, the inhomogeneous light emission elucidate that the smaller pixels whose size close to the medium pits are more vulnerable to screw dislocations, and thus controlling dislocation density will be even more crucial for future microdisplay applications using F--implanted µVLED array with ultra-small pixel diameters less than 10 µm.
Fig. 4. Electroluminescence images of F--implanted µVLED arrays with pixel diameters of (a) 6, (b) 8 and (c) 10 µm operated at 20 mA.
Integrating sphere measurements are further performed on all F--implanted µVLED arrays for which the light-output power dependences on current are normalized by light-emission zone. As shown in Fig. 5(a), the output power densities of all µVLED arrays increase monotonously with increasing injection current densities and can maintain high levels of current density before breakdown. Meanwhile, the µVLED array with bigger pixel size achieves larger output power under same injection conditions. On the other hand, the EQEs of all µVLED arrays also show size dependencies. That is, the bigger the pixel size is, the higher the EQE is, and the peak EQEs for µVLED arrays with pixel diameters of 10 µm, 8 µm and 6 µm are 20.4%, 15.3% and 10.1%, respectively. Interestingly, the µVLED array with bigger pixel has high peak efficiency at lower current density but also exhibits a stronger efficiency droop, while the array with smaller pixel shows little change in efficiency droop. For example, the droops at 220 A/cm2 from each peak EQEs are 31.7%, 17.9% and 8.4% for pixel diameters from 10 µm to 6 µm, respectively, which implies that the F--implanted µVLED arrays with different pixel sizes behave in distinctly different ways as current densities increases.
Fig. 5. (a) Optical output densities and EQEs as a function of current density for F--implanted µVLED array with pixel diameters ranging from 6 µm to 10 µm. (b) EQEs of F--implanted µVLED arrays versus carrier density, the dash lines represent fitted curves by carrier rate equation model.
Given that the device structures and fabricating processes are identical among these µVLED arrays, the reduced output performances and mitigation of efficiency droop for the array with smaller pixel size could be ascribed to the relative high dislocation density, which also means that the local density of non-radiative recombination centers in these arrays may be quite different. To get better understanding on efficiency-droop mechanism for the F--implanted µVLED arrays, the variation tendencies of efficiency droops related to different pixel diameters are analyzed using a carrier rate equation (ABC) model by which EQEs can be defined as
where ηext is the light-extraction efficiency, n is the carrier density. The coefficients A, B and C represent the SRH recombination, bimolecular radiative recombination, and Auger recombination, respectively [30]. As shown in Fig. 5(b), the values of A, B, and C coefficients can be iteratively achieved by fitting EQEs as a function of generation rate of carriers and the related fitted results are summarized in Table 2. Obviously, those three values of squared correlation coefficients (R-square) are all close to 1, indicating high accuracy of fitting processes based on the ABC model.Table 2. The values of A, B, and C coefficients obtained by fit to carrier rate equation (ABC) model.
The efficiency droop in GaN-based LEDs originates from non-radiative carrier loss mechanisms that have intrinsic relevance to defect density, especially as injection current increase. Generally, high defect densities increase the dominance of SRH non-radiative recombination and decrease non-radiative lifetime at low currents, which suppress the peak EQE and increase the value of the A coefficient, leading to the fact that the radiative recombination never turns into a dominant process even beyond the region dominated by some droop-causing mechanisms. In Table 2, the larger A coefficients indicate that defects-related SRH process become stronger as the decrease of pixel diameters. In other words, the relative high defect densities existed in µVLED array with smaller pixel sizes play a dominant role in suppressing peak efficiency, which is further confirmed by the corresponding smaller B coefficients. In addition to the competition with radiative recombination, the strong SRH process simultaneously relates to a reduced efficiency droop [30–33]. Accordingly, the larger A coefficients also means that carriers can be recombined more quickly via a strong SRH process, leading to a lower carrier density in MQWs, which in turn suppress the high-order droop-causing mechanisms (such as defect-assisted Auger non-radiative recombination) at a given high current, and hence mitigate the droop effect of µVLED array with smaller pixel sizes.
More notably, the fitting values of C coefficient are 1.04×10−28, 6.93×10−29 and 5.62×10−29 cm6 s−1 for arrays with pixel diameters of 6, 8 and 10 µm, respectively. Owing to no reported Auger recombination effect can explain such a large C coefficient which is bigger than 10−28 cm6 s−1, it is reasonable to assume that there must exists some other carrier loss mechanisms besides the Auger recombination for the array with 6 µm pixel diameter. Furthermore, the values of C coefficient decrease as pixel sizes increase, suggesting that the defect-assisted efficiency droop mechanisms at high injection currents can be effectively suppressed by relative lower defect density. Therefore, controlling defect densities is crucial to improving light-output performances for F--implanted µVLED arrays with ultra-small pixel sizes.
It should be emphasized that, for monolithic microdisplay, the metal lines for ohmic contacts and signal transmission paths are normally integrated on a same wafer, which requires a separate driving circuit to operate. However, for high-density microdisplay, it is very difficult to connect enormous amount of control signals from each individual pixel to separate driving circuit within a limited space [30]. Moreover, such a monolithic microdisplay can only be driven by a passive mode for which high source voltage is required, thus leading to serious issues related to efficiency and thermal dissipation. Therefore, as shown in Fig. 6, for the practical microdisplay applications, the F--implanted µVLED array can be designed and driven with high-density CMOS active matrix driver whose function is to address each pixel independently [34], so that specific images can be displayed on µVLED array.
4. Conclusion
We fabricated high-efficiency InGaN-based µVLED arrays through non-destructive ion-implantation process and investigated implantation energy and size dependences of light-output performances for F--implanted µVLED arrays. The results show that composite energy F- ion implantation in n-GaN layer can realize more effective electrical isolation for µVLED arrays, which not only can avoid ion-induced damages but also can realize ion-induced heat relaxation effectively, and hence more suitable for fabricating the µVLED arrays with smaller reverse leakage current and higher optical output power. A measured output power density of 82.1 W cm−2 at an injection current density of 220 A cm−2 can be achieved for the dual-energy (90/180 KeV) F--implanted µVLED array with pixel diameters of 10 µm. Furthermore, the µVLED array with bigger pixel diameter shows a striking efficiency peak but followed by a stronger efficiency droop, while the array with smaller pixel exhibits low peak efficiency and little efficiency droop. Analysis revealed that the output power density and EQEs of µVLED arrays with pixel diameters less than 10 µm are severely limited due to the relative higher defect densities and associated defects-related SRH process. This work will greatly promote the understanding of device physics in ion-implantation-based µVLEDs and provide a new compatible strategy for high-resolution, high-efficiency µVLED array fabrication to meet the industrial requirement of microdisplay applications.
Funding
Key Industry Technology Innovation Program of Suzhou (SYG201928); China Postdoctoral Science Foundation (2019M661969); Jiangsu Planned Projects for Postdoctoral Research Funds (2018K008C).
Acknowledgments
Technical support was provided by Nano Fabrication Facility, Suzhou Institute of Nano-Tech and Nano-Bionics, Chinese Academy of Sciences.
Disclosures
The authors declare no conflicts of interest.
References
1. X. Y. Liu, R. Z. Lin, H. L. Chen, S. L. Zhang, Z. Y. Qian, G. F. Zhou, X. W. Chen, X. J. Zhou, L. R. Zheng, R. Liu, and P. F. Tian, “High-bandwidth InGaN self-powered detector arrays toward MIMO visible light communication based on micro-LED arrays,” ACS Photonics 6(12), 3186–3195 (2019). [CrossRef]
2. M. S. Wong, D. Hwang, A. I. Alhassan, C. Lee, R. Ley, S. Nakamura, and S. P. Denbarrs, “High efficiency of III-nitride micro-light emitting diodes by sidewall passivation using atomic layer deposition,” Opt. Express 26(16), 21324–21331 (2018). [CrossRef]
3. J. Day, J. Li, D. Y. C. Lie, C. Bradford, J. Y. Lin, and H. X. Jiang, “III-Nitride full-scale high-resolution microdisplays,” Appl. Phys. Lett. 99(3), 031116 (2011). [CrossRef]
4. 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]
5. X. Z. Bao, J. Q. Liang, Z. Z. Liang, W. B. Wang, C. Tian, Y. X. Qin, and J. G. Lü, “Design and fabrication of AlGaInP-based micro-light-emitting-diode array devices,” Opt. Laser Technol. 78, 34–41 (2016). [CrossRef]
6. J. A. Chediak, Z. Luo, J. Seo, N. Cheung, L. P. Lee, and T. D. Sands, “Heterogeneous integration of CdS filters with GaN LEDs for fluorescence detection Microsystems,” Sens. Actuators, A 111(1), 1–7 (2004). [CrossRef]
7. N. Grossman, V. Poher, M. S. Grubb, G. T. Kennedy, K. Nikolic, B. McGovern, P. R. Berlinguer, Z. Gong, E. M. Drakakis, M. A. Neil, M. D. Dawson, J. Burrone, and P. Degenaar, “Multi-site optical excitation using ChR2 and micro-LED array,” J. Neural Eng. 7(1), 016004 (2010). [CrossRef]
8. Y. G. Huang, G. J. Tan, F. W. 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]
9. Z. Gong, H. X. Zhang, E. Gu, C. Griffin, M. D. Dawson, V. Poher, G. Kennedy, P. M. W. French, and M. A. A. Neil, “Matrix-Addressable Micropixellated InGaN Light-Emitting Diodes With Uniform Emission and Increased Light Output,” IEEE Trans. Electron Devices 54(10), 2650–2658 (2007). [CrossRef]
10. W. Y. Tian and J. H. Li, “Size-dependent optical-electrical characteristic of blue GaN/InGaN micro-light emitting diodes,” Appl. Opt. 59(29), 9225–9232 (2020). [CrossRef]
11. S. W. H. Chen, C. C. Shen, T. Wu, Z. Y. Liao, L. F. Chen, J. R. Zhou, C. F. Lee, C. H. Lin, C. C. Lin, C. W. Sher, P. T. Lee, A. J. Tzou, Z. Chen, and H. C. Kuo, “Full-color monolithic hybrid quantum dot nanoring micro light-emitting diodes with improved efficiency using atomic layer deposition and nonradiative resonant energy transfer,” Photonics Res. 7(4), 416–422 (2019). [CrossRef]
12. 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]
13. D. Hwang, A. Mughal, C. D. Pynn, S. Nakamura, and S. P. DenBaars, “Sustained high external quantum efficiency in ultrasmall blue III-nitride micro-LEDs,” Appl. Phys. Express 10(3), 032101 (2017). [CrossRef]
14. S. X. Jin, J. Li, J. Y. Lin, and H. X. Jiang, “InGaN/GaN quantum well interconnected microdisk light emitting diodes,” Appl. Phys. Lett. 77(20), 3236–3238 (2000). [CrossRef]
15. L. Chang, Y. W. Yeh, S. Hang, K. K. Tian, J. Q. Kou, W. G. Bi, Y. H. Zhang, Z. H. Zhang, Z. J. 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), 160 (2020). [CrossRef]
16. S. Hang, C. M. Chuang, Y. H. Zhang, C. S. Chu, K. K. Tian, Q. Zheng, T. Z. Wu, Z. J. 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]
17. F. Xu, C. H. Gao, Y. M. Fan, P. Chen, and B. S. Zhang, “Enhanced performance of vertical-structured InGaN micro-pixelated light-emitting-diode array fabricated using an ion implantation process,” Opt. Lett. 44(18), 4562–4565 (2019). [CrossRef]
18. J. Q. Kou, C. C. Shen, H. Shao, J. M. Che, X. H. Chunshuang, S. Chu, K. K. Tian, Y. H. 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), A643–653 (2019). [CrossRef]
19. W. C. Xing, Z. H. Liu, H. D. Qiu, G. I. Ng, and T. Palacios, “Planar-Nanostrip-Channel InAlN/GaN HEMTs on Si With Improved gm and fT Linearity,” IEEE Electron Device Lett. 38(5), 619–622 (2017). [CrossRef]
20. C. F. Lo, T. S. Kang, L. Liu, C. Y. Chang, S. J. Pearton, I. I. Kravchenko, O. Laboutin, J. W. Johnson, and F. Ren, “Isolation blocking voltage of nitrogen ion-implanted AlGaN/GaN high electron mobility transistor structure,” Appl. Phys. Lett. 97(26), 262116 (2010). [CrossRef]
21. S. Arulkumaran, K. R. I. N. Kennedy, P. P. M. T. N. Bhat, and S. Tripathy, “Thermally stable device isolation by inert gas heavy ion implantation in AlGaN/GaN HEMTs on Si,” J. Vac. Sci. Technol., B 34(4), 042203 (2016). [CrossRef]
22. S. Huang, H. W. Chen, and K. J. Chena, “Effects of the fluorine plasma treatment on the surface potential and Schottky barrier height of AlxGa1-xN/GaN heterostructures,” Appl. Phys. Lett. 96(23), 233510 (2010). [CrossRef]
23. S. C. Binari, H. B. Dietrich, G. Kelner, and H. He, “and N implant isolation of n-type GaN,” J. Appl. Phys. 78(5), 3008–3011 (1995). [CrossRef]
24. C. Saguy, J. Salzman, R. Kalish, V. Richter, U. Tish, S. Zamir, and S. Prawer, “Electrical isolation of GaN by ion implantation damage: Experiment and model,” Appl. Phys. Lett. 74(17), 2441–2443 (1999). [CrossRef]
25. A. Weimar, A. Lell, G. Brüderl, S. Bader, and V. HÄrle, “Investigation of Low-Resistance Metal Contacts on p-Type GaN Using the Linear and Circular Transmission Line Method,” Phys. Status Solidi A 183(1), 169–175 (2001). [CrossRef]
26. J. H. Li, X. R. Zhang, C. Zhou, J. G. Zheng, D. S. Ge, and W. H. Zhu, “New applications of an automated system for high-power LEDs,” IEEE/ASME Trans. Mechatron. 21(2), 1035–1042 (2016). [CrossRef]
27. J. H. Li, B. K. Ma, R. S. Wang, and L. Han, “Study on a cooling system based on thermoelectric cooler for thermal management of high-power LEDs,” Microelectron. Reliab. 51(12), 2210–2215 (2011). [CrossRef]
28. Z. Liu, J. H. Li, and X. H. Liu, “Novel functionalized BN nanosheets/epoxy composites with advanced thermal conductivity and mechanical properties,” ACS Appl. Mater. Interfaces 12(5), 6503–6515 (2020). [CrossRef]
29. S. Usami, Y. Ando, A. Tanaka, K. Nagamatsu, M. Deki, M. Kushimoto, S. Nitta, Y. Honda, H. Amano, Y. Sugawara, Y. Z. Yao, and Y. Ishikawa, “Correlation between dislocations and leakage current of p-n diodes on a free-standing GaN substrate,” Appl. Phys. Lett. 112(18), 182106 (2018). [CrossRef]
30. Z. Q. Liu, T. B. Wei, E. Q. Guo, X. Y. Yi, L. C. Wang, J. X. Wang, G. H. Wang, Y. Shi, I. Ferguson, and J. M. Li, “Efficiency droop in InGaN/GaN multiple-quantum-well blue light-emitting diodes grown on free-standing GaN substrate,” Appl. Phys. Lett. 99(9), 091104 (2011). [CrossRef]
31. H. Y. Ryu, H. S. Kim, and J. I. Shim, “Rate equation analysis of efficiency droop in InGaN light-emitting diodes,” Appl. Phys. Lett. 95(8), 081114 (2009). [CrossRef]
32. W. Liu, D. G. Zhao, D. S. Jiang, P. Chen, Z. S. Liu, J. J. Zhu, X. Li, F. Liang, J. P. Liu, L. Q. Zhang, H. Yang, Y. T. Zhang, and G. T. Du, “Shockley-Read-Hall recombination and efficiency droop in InGaN/GaN multiple-quantum-well green light-emitting diodes,” J. Phys. D: Appl. Phys. 49(14), 145104 (2016). [CrossRef]
33. H. Y. Ryu, D. S. Shin, and J. I. Shim, “Analysis of efficiency droop in nitride light-emitting diodes by the reduced effective volume of InGaN active material,” Appl. Phys. Lett. 100(13), 131109 (2012). [CrossRef]
34. J. Seong, J. Jang, J. Lee, and M. Lee, “CMOS Backplane Pixel Circuit with Leakage and Voltage Compensation for an Micro-LED Display achieving 5000 PPI or higher,” IEEE Access 8, 49467–49476 (2020). [CrossRef]
