Open Access
Add to BibSonomy
Abstract
In this study, deep ultraviolet light-emitting diodes (DUV-LEDs) with a reflective passivation layer (RPL) were investigated. The RPL consists of HfO2/SiO2 stacks as distributed Bragg reflectors, which are deposited on two DUV-LEDs with different p-GaN thicknesses. The RPL structure improved the external quantum efficiency droops of the DUV-LEDs with thick and thin p-GaN, thereby increasing their light output power by 18.4% and 39.4% under injection current of 500 mA and by 17.9% and 37.9% under injection current of 1000 mA, respectively. The efficiency droops of the DUV-LEDs with and without the RPL with thick p-GaN were 20.1% and 19.1% and with thin p-GaN were 18.0% and 15.6%, respectively. The DUV-LEDs with the RPL presented improved performance. The above results demonstrate the potential for development of the RPLs for DUV-LED applications.
© 2021 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
AlGaN-based deep ultraviolet light-emitting diodes (DUV-LEDs) have generated significant interest for applications in daily lives and industries [1,2]. LEDs can have lateral and flip–chip (FC) structures for blue LED or DUV-LED applications. The problem of using a lateral LED structure for DUV-LEDs in sapphire-based applications is the current accumulation due to its multi-finger geometry [3]. Using a lateral LED geometry, the distance between the n- and p-contacts was modified by design, resulting in a more uniform current injection throughout the active region of LED compare with the FC-LED geometry. With a lateral-design structure, most sapphire-based DUV-LEDs suffer from self-heating due to high operating voltages, low lighting efficiency, and low thermal conductivity of the sapphire substrate (0.35 W·cm−1K−1) [4]. The self-heating problem reduces the reliability of DUV-LEDs, shifts the emission wavelength, and decreases the emission efficiency on account of the shading rate of the electrode. When an FC-LED structure is used for DUV-LEDs, the n- or p-electrode covers most of the active region of the LED device and the electrode structure can consist of a high reflective metal in the DUV wavelength region. The emission light can be reflected by the n- or p-electrode and also be propagated through the sapphire substrate side, thereby enhancing the light efficiency. An FC-structured DUV-LED was mounted on an FC configuration with a high-thermal conductivity insulating AlN carrier (175 W·cm−1K−1) by thermocompression gold bonding.
Because of their compact size and environmental friendliness, DUV-LEDs can substitute the traditional mercury lamps. However, DUV-LEDs have serious disadvantages, such as a much lower output power than blue LEDs [5,6] resulting from serval factors: (1) low internal quantum efficiency (IQE) of DUV-LEDs caused by the high dislocation densities in the AlxGa1−xN-based active layers [7]. These high dislocation densities are, in turn, caused by the large lattice and thermal expansion coefficient mismatches between the AlGaN layers and the sapphire substrate [8], which lower the external quantum efficiency (EQE) and reduce the reliability of the DUV-LEDs. To reduce the dislocation densities in the AlxGa1-xN-based active layers, the use of a high-crystalline quality aluminum nitride (AlN) template/sapphire was investigated [9]. A thick AlN epilayer was grown on microscale patterned c-axis-oriented sapphire substrates by sputtering AlN thin films via hydride vapor-phase epitaxy (HVPE) [10]. HVPE-AlN epilayer deposition by a two-step approach was conducted to promote the lateral growth and coalescence of the AlN epilayer. A relatively smooth AlN epilayer surface, without discontinuities and holes, showed full width at half maximum (FWHM) values of the symmetric (0002) and asymmetric (10-12) measurements of 110 and 393 arcsec, respectively. The thickness of an AlN template epilayer grown by metal organic chemical vapor deposition is a critical factor affecting the crystalline quality of the AlxGa1-xN epilayer [11]. An ammonia (NH3) pulsed flow was controlled by enhancing the precursor migration to obtain high-quality AlN using an AlN template with a thick crack-free and atomically flat surface [12]. The growth ratios of NH3 pulsed flow and continuous flow modes were 0.6 and 6 µm/h, respectively. The FWHMs of the X-ray diffraction (0002) and (10-12) measurements of a 4 µm AlN buffer layer were 180 and 370 arcsec, respectively. (2) The low EQE of AlxGa1−xN-based DUV-LEDs is a major bottleneck due to the strong UV light absorption by the p-GaN Ohmic contact layer and the indium tin oxide (ITO) p-contact layer. The p-GaN contact layer not only provides a good Ohmic contact but also reduces the series resistance for achieving reliable high-power LED operation. To reduce the optical loss in the p-side of DUV-LEDs, the p-GaN contact layer is made thin [13], partially removed [14], or replaced by a transparent AlxGa1−xN epitaxial layer [15]. The EQE of a LED is defined as the ratio of the number of photons emitted by the LED per second to the number of electrons injected into the device per second [11] as follows:
In this study, hafnium dioxide (HfO2)/SiO2 stacks as reflectors were deposited by an e-gun evaporator and covered an entire LED chip to improve the light extraction efficiency of DUV-LEDs. This study not only examines in detail the light extraction effected by the reflector but also the effects of ITO and p-GaN layers with different thicknesses on the DUV-LEDs.
2. Experiments
The DUV-LED structure was grown on a (0001)-oriented c-plane sapphire substrate by metal organic chemical vapor deposition. Disilane and bis (cyclopentadienyl) magnesium were used for n- and p-type doping, respectively. Trimethylgallium and trimethylaluminum were used as the group-III element sources, and NH3 served as the group-V element source. The DUV-LED structure consisted of a 25 nm AlN template, 1.5 µm-thick AlN layer, 2 µm-thick Si-doped Al0.45Ga0.55N layer, Si-doped Al0.4Ga0.6N/Al0.6Ga0.4N multiple quantum well, 30 nm-thick p-Al0.75Ga0.25N as the electron blocking layer, p-GaN:Mg layer, and Mg-doped p+-GaN layer of 5 nm as the Ohmic contact layer. The device process flow is illustrated in Figs. 1(a)–1(j). The epitaxial samples were cleaned using H2SO2:H2O2:H2O = 5:1:1 for 10 min to remove the organic residues, which is shown in Fig. 1(a). A mesa pattern was formed by the photolithography process, which was followed by an inductively coupled plasma (ICP) dry-etching process to produce n-Al0.45Ga0.55N as the n-electrode, which is shown in Fig. 1(b). Subsequently, an isolation pattern was formed for the electric isolation of the LED device, which is displayed in Fig. 1(c). Ti/Al/Ni/Au (15/100/50/250 nm) layers were deposited by e-beam evaporation and subsequently annealed at 950 °C for 30 s to serve as the n-pad (Fig. 1(d)). ITO as the p-contact electrode with p-GaN was deposited by sputtering and subsequently annealed at 600 °C for 300 s (Fig. 1(e)). A metal stack dot pattern (Ni/Au: 50/250 nm) was deposited on the ITO thin film for p-spreading layer (PSL) deposition to expose partial ITO layer for the subsequent RPL deposition, as shown in Fig. 1(f). For resolving the reliability problem, a 400 nm-thick SiO2 layer was deposited on the entire device by plasma-enhanced chemical vapor deposition (PECVD) at 240 °C. It is worth mentioning that high-quality, dense, and thick SiO2 can be obtained by PECVD. The RPL consisted of HfO2 and SiO2 stack layers, which were deposited on the 400 nm passivation layer by electron beam evaporation. Note that the 400 nm SiO2 can reduce the stress and avoid cracking of the RPL. The chamber was evacuated to a base pressure of 1×10−6 mbar, and the HfO2 and SiO2 layers were deposited under oxygen atmosphere. The purity of the HfO2 and SiO2 source materials was 99.9%. The working pressure and temperature were 2.4×10−4 mbar and 160 °C, respectively, which is shown in Fig. 1(g). The optical characteristics of the HfO2 and SiO2 stack were etched by an ICP dry-etching process, which is displayed in Fig. 1(h). The optical characteristics of the HfO2 and SiO2 stack reflectors were determined by transmission measurements using a Cary 5000 spectrophotometer (Varian, Australia). The PSL pattern was formed by photolithography process, which was followed by ICP dry-etching process to form a Ni/Au stack dot pattern contact, which was exposed for further PSL contact layer (Ni/Au: 50/250 nm); this is shown in Fig. 1(i). Subsequently, a 1 µm SiO2 layer was deposited on the entire device for passivation to ensure reliability by PECVD at 240 °C, which is depicted in Fig. 1(j). Cr/Pt/Au/AuSn (5/50/250/6000 nm) layers were deposited for the p- and n-contact bonding pads. After the conventional LED processes, the DUV-LED wafer was subjected to laser scribing and broken into chips of 40 mil × 40 mil dimensions, which is shown in Fig. 1(k). Finally, DUV-LED chips with the RPL were mounted on an AlN ceramics submount in an FC structure by soldering flux under 240 °C for 5 min for further electrical and optical measurements, which is displayed in Fig. 1(l). The thicknesses of the HfO2, SiO2, and metal electrode layers were measured by KLA-Tencor profilometer P-10. The reflectance of the RPL was measured using an n&k 1280 analyzer (n&k Technology, Inc.). The light output power of the packaged LEDs and the thermal behaviors (junction temperature) were measured using a calibrated integrating sphere (Instrument Systems) and a TeraLED and T3Ster combined system, respectively. The thin film carrier concentration and the mobility were measured using the standard Hall effect with a four-probe configuration (ACCENT HL5500 PC). The current–voltage (I–V) characteristics of the LEDs were determined at 25°C using an Agilent 4155B semiconductor parameter analyzer. To evaluate the effect of the RPL on the LED performance, four types of LED samples were fabricated using the same epitaxial structure with different p-GaN thicknesses: Ref-LED with thick p-GaN of 100 nm thickness and ITO thickness of 110 nm, Ref-RPL-LED with thick p-GaN of 100 nm thickness and ITO thickness of 12 nm and the RPL, Tp-GaN-LED with thin p-GaN of 40 nm thickness and ITO thickness of 110 nm, and Tp-GaN-RPL-LED with thin p-GaN of 40 nm thickness and ITO thickness of 12 nm and the RPL.
Fig. 1. Schematics of fabrication process of DUV-LED with RPL, (a) epitaxial structure of DUV-LED, (b) mesa process, (c) isolation pattern formed for electric isolation of LED device, (d) n-pad deposition and annealing, (e) ITO deposition for 110 nm and 12 nm, (f) Ni/Au metal, (g) RPL deposition, (h) RPL deposition after dry etching (i) p-spreading layer of Ni/Au(50/250nm), (j) passivation layer of 1 µm SiO2, (k) bonding pad formation, (l) mounting DUV-LED chip on AlN ceramics submount using flip-chip structure electrical connection.
3. Results and discussion
A major concern regarding DUV-LEDs with an RPL is their I–V characteristics. The forward I–V characteristics of the Ref-LED, Ref-RPL-LED, Tp-GaN-LED, and Tp-GaN-RPL-LED are shown in Fig. 2(a). All LED structures exhibited normal p–n diode behaviors at the forward dc bias. At an injection current of 500 mA, the forward voltages of the Ref-LED, Ref-RPL-LED, Tp-GaN-LED, and Tp-GaN-RPL-LED were 6.007, 6.351, 6.741, and 7.277 V under room temperature, respectively. It was found that the forward voltages of the Ref-LED and Ref-RPL-LED were lower than those of the Tp-GaN-LED and the Tp-GaN-RPL-LED. Therefore, the forward voltages of the DUV-LEDs were significantly affected by the p-GaN thickness.
Fig. 2. (a) Forward I–V and (b) dynamic resistance characteristics of Ref-LED, Ref-RPL-LED, Tp-GaN-LED, and Tp-GaN-RPL-LED.
The dynamic resistances of the Ref-LED, Ref-RPL-LED, Tp-GaN-LED, and Tp-GaN-RPL-LED as functions of the applied voltage are shown in Fig. 2(b). When the applied voltage was higher than 4.0 V, the Ref-LED and the Ref-RPL-LED presented lower dynamic resistances, which were lower than those of the Tp-GaN-LED and the Tp-GaN-RPL-LED. The dynamic resistances of the Ref-LED, Ref-RPL-LED, Tp-GaN-LED, and Tp-GaN-RPL-LED were 3.098, 4.161, 11.656, and 15.322 Ω, respectively, at an applied voltage of 5.5 V. The dynamic resistances of the Ref-LED and the Ref-RPL-LED were lower than those of the Tp-GaN-LED and the Tp-GaN-RPL-LED. Because the injection current is injected from the top contact metal through the ITO and p-GaN to the entire active layer of the LED, the effect of the p-GaN thickness on the electrical properties of the LEDs was more remarkable than on those of the LEDs with the ITO layer.
The electrical properties of ITO thin films with 110 nm and 12 nm thicknesses were obtained by Hall measurement, and are listed in Table 1. Noticeably, the carrier concentration and mobility of the 110 nm ITO thin film were higher than those of the 12 nm ITO thin film. This will result in a lower resistivity for the former than that of the latter. The other concerning factor is the contact resistance. The specific contact resistance (ρc) values of different p-GaN thickness contacts and 110 nm and 12 nm ITO thin films are also listed in Table 1. Notably, the ρc values of the thick p-GaN (100 nm) contact with the 110 nm and 12 nm ITO thin films were 5.22 × 10−3 and 9.22 × 10−3 Ω-cm, respectively. Thus, the thickness of p-GaN has the dominant effect on the electrical properties of the DUV-LEDs compared with the ITO thickness. The obtained results agree well with the results shown in Fig. 2.
Table 1. Electrical properties of ITO thin films and specific contact resistance (ρc) of p-GaN/ITO after annealing.
Figure 3(a) shows the measured reflectance spectra of the RPL for different angles of incidence. The reflection band width changes with the incidence angle. From the theoretical optical analysis of the incident light angle-dependent reflectance of the RPL, the simulation results show a similar trend to the measured data in the same ranges of the wavelength and incident angle, as shown in Fig. 3(b). The reflection bandwidth is defined as that in which the wavelength range of reflectivity > 90% when the incident angle is 10°–50° and reflectivity > 85% when the incident angle is 60°–80°. As summarized in Table 2, the bandwidth decreases as the incident angle increases. The reflection bandwidth becomes narrow and the spectrum shifts toward a shorter wavelength when the incident angle increases, which is shown in Fig. 3(c). The transmittance spectra of the thick p-GaN of 100 nm thickness, thin p-GaN of 40 nm thickness, ITO of 110 nm thickness, and ITO of 12 nm thickness in the 200-500 nm wavelength range are shown in Fig. 3(d). The transmittance of the p-GaN and ITO thin films increases with their thickness decrease. However, the transmittance levels of the thick p-GaN, thin p-GaN, 110 nm ITO, and 12 nm ITO are 15.7%, 45.7%, 13.0%, and 61.8% at the wavelength of 280 nm, respectively. More details of the transmittance of the p-GaN and ITO thin films are shown in the insert table in Fig. 3(d). This can be attributed to the strong absorption in the DUV and the steady decrease in the extinction coefficient caused by the physical properties of materials. The different thicknesses of the p-GaN and ITO thin films will result in different electrical and optical properties and further affect the LED device performance.
Fig. 3. Angle-dependent reflectance spectra of (a) measured, (b) calculated RPL, and (c) RPL structures for different wavelengths by varying detected angles from 10° to 90° and (d) transmittance of p-GaN and ITO thin films.
Table 2. Reflective bandwidths with different incident angles.
Figure 4(a) illustrates the output power characteristics of the Ref-LED, Ref-RPL-LED, Tp-GaN-LED, and Tp-GaN-RPL-LED as functions of the injection current measured at room temperature. From Fig. 4(a), all LEDs with different epitaxial structures and LED structure designs exhibit a linear relationship between the output power and the injection current. The output powers of the Ref-LED, Ref-RPL-LED, Tp-GaN-LED, and Tp-GaN-RPL-LED were 40.21, 47.62, 48.38, and 67.45 mW at an injection current of 500 mA, respectively, and were 74.22, 87.53, 90.8, and 125.24 mW at an injection current of 1000 mA, respectively. The output powers of the samples with the RPL increased by 18.4% and 39.4% at an injection current of 500 mA and by 17.9% and 37.9% at an injection current of 1000 mA, respectively, for thick and thin p-GaN epitaxial structures. It is worth mentioning that the LEDs with the RPL structure presented the best output power performance for both epitaxial structures even though they did not exhibit the best I–V characteristics. To evaluate the light extraction efficiencies (ηextrac) of the different LEDs, their EQEs as functions of the injection current are shown in Fig. 4(b). The maximum EQEs of the Ref-LED, Ref-RPL-LED, Tp-GaN-LED, and Tp-GaN-RPL-LED were 1.90%, 2.27%, 2.33%, and 3.09% at the injection currents of 20, 20, 20, and 300 mA, respectively. When the injection current was raised from 1 mA to 1600 mA, the EQE increased markedly, first reaching a maximum at a certain current, and subsequently slightly decreased. To further investigate the carrier transfer and the recombination mechanism of the LED structures, the droop efficiencies (definition: [(EQEmax-EQEmin)/EQEmax] × 100%) of the Ref-LED, Ref-RPL-LED, Tp-GaN-LED, and Tp-GaN-RPL-LED were calculated as 20.1%, 19.1%, 18.0%, and 15.6%, respectively. Noticeably, the droop efficiency was improved by introducing the thin p-GaN layer and the RPL owing to their reduced optical absorption. From temperature-dependent photoluminescence measurements, the IQE of a DUV-LED was estimated as 40.8% [20], and the results are shown in Fig. 4(c). The peak wavelengths of the UV epilayer were 273.55 and 274.64 nm, respectively. The LEEs of the Ref-LED, Ref-RPL-LED, Tp-GaN-LED, and Tp-GaN-RPL-LED were calculated using LEE = IQE/EQE as 4.68%, 5.57%, 5.75%, and 7.57%, respectively. The LEEs of the Ref-RPL-LED, Tp-GaN-LED, and Tp-GaN-RPL-LED were improved by 19.02%, 22.86%, and 61.75% compared to that of the Ref-LED. The maximum wall plug efficiency of the Ref-LED, Ref-RPL-LED, Tp-GaN-LED, and Tp-GaN-RPL-LED were 2.01%, 2.06%, 1.74%, and 2.59%, respectively. Thus, the thin GaN and the thin ITO can reduce the UV absorption. This can be expected to allow more light to be reflected by the RPL and escape from the sapphire substrate.
Fig. 4. (a) Output power characteristics and (b) EQE as functions of current for Ref-LED, Ref-RPL-LED, Tp-GaN-LED, and Tp-GaN-RPL-LED after encapsulation.
Figure 5 shows the emission peaks of the Ref-LED, Ref-RPL-LED, Tp-GaN-LED, and Tp-GaN-RPL-LED at an injection current of 400 mA. The FWHMs of the electroluminescence peaks of the Ref-LED, Ref-RPL-LED, Tp-GaN-LED, and Tp-GaN-RPL-LED were at 9.8, 10.1, 9.9, and 10.1 nm at an injection current of 400 mA, respectively, showing that the FWHMs are similar.
Fig. 5. Emission peaks of Ref-LED, Ref-RPL-LED, Tp-GaN-LED, and Tp-GaN-RPL-LED as functions of injection current.
Figure 6 shows the far-field radiation patterns of the Ref-LED, Ref-RPL-LED, Tp-GaN-LED, and Tp-GaN-RPL-LED at an injection current of 350 mA. The output intensities of the LEDs with the RPL structure were higher than that of the LED without the RPL structure for both epilayer structure regarding the full angles. This was caused by the reflection of the RPL structure being higher than that of the single SiO2 layer. In the Ref-LED and the Ref-RPL-LED, the p-GaN is thick, which will further affect the light extraction. Therefore, the light extractions of the Ref-LED and the Ref-RPL-LED presented similar radiation patterns. In contrast, as the p-GaN thickness is reduced to 40 nm in the Tp-GaN-LED and the Tp-GaN-RPL-LED, the light absorption by p-GaN was also reduced and more light was reflected by the RPL, and subsequently escaped from LED surfaces. The LEEs of the Tp-GaN-LED and the Tp-GaN-RPL-LED were improved. The results of the intensity in the normal direction (0° direction) were consistent with the optoelectronic performance of the LEDs with the RPL. The viewing angles of the Ref-LED and Ref-RPL-LED, Tp-GaN-LED, and Tp-GaN-RPL-LED were 136.5°, 134.6°, 135.2°, and 126.9°, respectively. More light was extracted from the Ref-RPL-LED and Tp-GaN-RPL-LED surfaces than from the Ref-LED and Tp-GaN-LED surfaces, thus reducing the light absorption and finally narrowing the viewing angle, particularly in the Tp-GaN-RPL-LED. The far-field radiation pattern of the Tp-GaN-RPL-LED between 0° and 30° was significantly enhanced by the RPL, and the improvement was caused by two factors: (1) the thin p-GaN layer reduces the light absorption and (2) the high reflection of the RPL, approximately 92%, between the measured angles of 10° and 40°, results in a narrower viewing angle. The far-field radiation pattern of the Tp-GaN-RPL-LED between 30° and 45° is far lower than that at 45°. It is probably caused by three factors: (1) The refractive index of sapphire is approximately 1.87 at 280 nm wavelength. Therefore, the critical angle of the FC LED structure between the sapphire and air for the Tp-GaN-RPL-LED is 33°. (2) The mesa edge angle after dry etching is 30°–35°. (3) The RPL is covered on the top of p-Al0.45Ga0.55N and n-Al0.45Ga0.55N. The refractive indices of Al0.45Ga0.55N and SiO2 are 2.480 [21] and 1.490 at the wavelength of 280 nm, respectively. The first layer of the RPL is SiO2, and the critical angle between Al0.45Ga0.55N and SiO2 is 37°. Based on the above three factors, the light escaping from the active layer between 0° and 30° will improve the light extraction by the RPL and the thin p-GaN. When the light escapes from 30° to 90°, the critical angles of sapphire/air and Al0.45Ga0.55N/SiO2 and the mesa edge angle will be limited. Therefore, the unescaped light will be trapped inside the LED as the guiding mode propagation and escape from the chip side wall, causing the radiation pattern of the Tp-GaN-RPL-LED between 45° and 90° to be larger than that of the Tp-GaN-LED.
Fig. 6. Far-field radiation patterns of (a) Ref-LED and Ref-RPL-LED (b) Tp-GaN-LED and Tp-GaN-RPL-LED at injection current of 350 mA.
4. Conclusion
This paper presents the effects of an RPL on the performance of DUV-LEDs with thick and thin p-GaN layers. The RPL can provide wide-angle and broadband highly reflective characteristics for DUV wavelength applications. Comparing DUV-LEDs with thick and thin p-GaN layers and with and without the RPL, not only is the LEE improved but the droop properties of the DUV-LEDs with different p-GaN thickness are also reduced. Because the reflectance of the RPL is higher than that of a single passivation layer. The RPL and thin p-GaN thickness epitaxial structures show potential for DUV-LED applications.
Funding
National Natural Science Foundation of China (11904302); Major Science and Technology Project of Xiamen (3502Z20191015).
Acknowledgments
The authors would like to thank Prof. Nakamura of the UCSB, Prof. Hedeto Mayate of Mie University for their helpful discussions, and the Advanced Semicoductor Processing and Devices Lab (http://rhhlab.wixsite.com/astdl) of National Yang Ming Chiao Tung University for the measurement support.
Disclosures
The authors declare no conflicts of interest.
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. A. Khan, K. Balakrishnan, and T. Katona, “Ultraviolet light-emitting diodes based on group three nitrides,” Nat. Photonics 2(2), 77–84 (2008). [CrossRef]
2. M. S. Shur and R. Gaska, “Deep-ultraviolet light-emitting diodes,” IEEE Trans. Electron Devices 57(1), 12–25 (2010). [CrossRef]
3. X. Guo and E. F. Schubert, “Current crowding in GaN/InGaN light emitting diodes on insulating substrates,” J. Appl. Phys. 90(8), 4191–4195 (2001). [CrossRef]
4. A. Chitnis, J. Sun, V. Mandavilli, R. Pachipulusu, S. Wu, M. Gaevski, V. Adivarahan, J. P. Zhang, M. A. Khan, A. Sarua, and M. Kuball, “Self-heating effects at high pump currents in deep ultraviolet light-emitting diodes at 324 nm,” Appl. Phys. Lett. 81(18), 3491–3493 (2002). [CrossRef]
5. T.-C. Hsu, Y.-T. Teng, Y.-W. Yeh, X. Fan, K.-H. Chu, S.-H. Lin, K.-K. Yeh, P.-T. Lee, Y. Lin, Z. Chen, T. Wu, and H.-C. Kuo, “Perspectives on UVC LED: Its progress and application,” Photonics 8(6), 196 (2021). [CrossRef]
6. H. Chen, Z. Zhao, Y. Lin, L. Zhu, E. Ma, W. Guo, T. Wu, F. Lin, Y. Lu, Z. Chen, T. Shih, Y. Lei, and A. Pan, “On the exciton-assisted radiative recombination via impurity trap levels in AlGaN deep ultraviolet light-emitting diodes,” Nanotechnology 32(37), 375204 (2021). [CrossRef]
7. T. Wu, Y. Lin, Z. Peng, H. Chen, Z. Shangguan, M. Liu, S.-W. H. Chen, C.-H. Lin, H.-C. Kuo, and Z. Chen, “Interplay of carriers and deep-level recombination centers of 275-nm light-emitting diodes - Analysis on the parasitic peaks over wide ranges of temperature and injection density,” Opt. Express 27(16), A1060–A1073 (2019). [CrossRef]
8. C. Chu, K. Tian, Y. Zhang, W. Bi, and Z.-H. Zhang, “Progress in external quantum efficiency for III-Nitride based deep ultraviolet light-emitting diodes,” Phys. Status Solidi A 216(4), 1800815 (2019). [CrossRef]
9. Y. Narukawa, M. Ichikawa, D. Sanga, M. Sano, and T. Mukai, “White light emitting diodes with super-high luminous efficacy,” J. Phys. D: Appl. Phys. 43(35), 354002 (2010). [CrossRef]
10. T. Kinoshita, T. Obata, T. Nagashima, H. Yanagi, B. Moody, S. Mita, S.-i. Inoue, Y. Kumagai, A. Koukitu, and Z. Sitar, “Performance and reliability of deep-ultraviolet light-emitting diodes fabricated on AlN substrates prepared by hydride vapor phase epitaxy,” Appl. Phys. Express 6(9), 092103 (2013). [CrossRef]
11. T. Kinoshita, K. Hironaka, T. Obata, T. Nagashima, R. Dalmau, R. Schlesser, B. Moody, J. Xie, S.-I. Inoue, Y. Kumagai, A. Koukitu, and Z. Sitar, “Deep-ultraviolet light-emitting diodes fabricated on AlN substrates prepared by hydride vapor phase epitaxy,” Appl. Phys. Express 5(12), 122101 (2012). [CrossRef]
12. S. Xiao, K. Shojiki, and H. Miyake, “Thick AlN layers grown on micro-scale patterned sapphire substrates with sputter-deposited annealed AlN films by hydride vapor-phase epitaxy,” J. Cryst. Growth 566-567, 126163 (2021). [CrossRef]
13. M. Shatalov, W. Sun, R. Jain, A. Lunev, X. Hu, A. Dobrinsky, Y. Bilenko, J. Yang, G. A. Garrett, L. E. Rodak, M. Wraback, M. Shur, and R. Gaska, “High power AlGaN ultraviolet light emitters,” Semicond. Sci. Technol. 29(8), 084007 (2014). [CrossRef]
14. H. Hirayama, N. Maeda, S. Fujikawa, S. Toyoda, and N. Kamata, “Recent progress and future prospects of AlGaN-based high-efficiency deep-ultraviolet light-emitting diodes,” Jpn. J. Appl. Phys. 53(10), 100209 (2014). [CrossRef]
15. H.-Y. Ryu, I.-G. Choi, H.-S. Choi, and J.-I. Shim, “Investigation of Light Extraction Efficiency in AlGaN Deep-Ultraviolet Light-Emitting Diodes,” Appl. Phys. Express 6(6), 062101 (2013). [CrossRef]
16. Y. J. Sung, M. S. Kim, H. k. Kim, S. Choi, Y. H. Kim, M. H. Jung, R. J. Choi, Y. T. Moon, J. T. Oh, H. H. Jeong, and G. Y. Yeom, “Light extraction enhancement of AlGaN-based vertical type deep-ultraviolet light-emitting-diodes by using highly reflective ITO/Al electrode and surface roughening,” Opt. Express 27(21), 29930 (2019). [CrossRef]
17. N. Maeda, M. Jo, and H. Hirayama, “Improving the efficiency of AlGaN deep-UV LEDs by using highly reflective Ni/Al p-type electrodes,” Phys. Status Solidi A 215(8), 1700435 (2018). [CrossRef]
18. Y. Zheng, J. Zhang, L. Chang, C. Chu, K. Tian, Q. Zheng, Q. Li, Y. Zhang, W. Bi, and Z. H. Zhang, “Understanding omni-directional reflectors and nominating more dielectric materials for deep ultraviolet light-emitting diodes with inclined sidewalls,” J. Appl. Phys. 128(9), 093106 (2020). [CrossRef]
19. J. W. Lee, D. Y. Kim, J. H. Park, E. F. Schubert, J. Kim, J. Lee, Y. I. Kim, Y. Park, and J. K. Kim, “An elegant route to overcome fundamentally-limited light extraction in AlGaN deep ultraviolet light-emitting diodes: Preferential outcoupling of strong in-plane emission,” Sci. Rep. 6(1), 22537 (2016). [CrossRef]
20. E. Matioli and C. Weisbuch, “Direct measurement of internal quantum efficiency in light emitting diodes under electrical injection,” J. Appl. Phys. 109(7), 073114 (2011). [CrossRef]
21. U. Tisch, B. Meyler, O. Katz, E. Finkman, and J. Salzman, “Dependence of the refractive index of AlxGa1-xN on temperature and composition at elevated temperatures,” J. Appl. Phys. 89(5), 2676–2685 (2001). [CrossRef]
