Open Access
Add to BibSonomy
Abstract
AlGaN-based germicidal UV LEDs show promise in fighting the COVID-19 pandemic through disinfection of air, water, and surfaces. We report UV LEDs grown by MOCVD on SiC substrates, fabricated into thin-film flip chip devices. Replacing the uniform p-AlxGa1-xN layer (x = 0.2) with a short-period-superlattice of alternating (x = 0.1 and 0.8) Al-composition improved EQE from 1.3% to 2.7% (3.2% with encapsulation) at 20 A/cm2. Peak EQE and WPE values of 4.8% and 2.8% (287 nm) were measured at current densities below 2 A/cm2, and maximum output power of 7.4 mW (76 mW/mm2) was achieved at 284 nm. Further WPE improvements are expected with both superlattice and uniform layer optimization, improved p-contact metallization, and active region optimization.
© 2020 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Germicidal UV (GUV) radiation with dominant wavelengths around 250 nm - 290 nm is known to have an efficient inactivation effect on coronaviruses such as MERS [1] and, as was recently reported, SARS-CoV-2 [2]. GUV can also be used against all other bacterial, viral, and fungal pathogens, with widespread success in air and water purification applications [3]. A consumer market for water and air disinfection systems has emerged, and applications are expected to expand as GUV light sources with greater efficiency and power output are developed.
The universal efficacy of GUV disinfection has been attributed to pyramidine dimerization within DNA/RNA [3] and disulphide bridge breakdown in proteins [4,5]. According to most reports, the maximum GUV disinfection efficacy occurs with 265 nm illumination, but significant efficacy has been reported at wavelengths as long as 295 nm [6]. Despite the well-known maximum efficacy above 260 nm, most disinfection systems use 254 nm: the dominant wavelength of low-pressure mercury vapor lamps. It is expected that GUV disinfection could be much more widely applied if mercury lamps were replaced by a solid-state UV light source (UV LED) with controllable emission wavelength; solid-state technologies would enable instant on/off functionality and higher directionality, efficiency, power density, and lifetime, all while lowering cost and shrinking the form factor.
UV LEDs based on the AlxGa1-xN (AlGaN) alloy system can emit radiation spanning from 210 nm to 360 nm, but most research efforts have been focused on improving light output in the GUV region. As a result, GUV optoelectronics technologies have improved significantly in the last decade [7–10] and devices with output powers of some tens of milliwatts and efficiencies of up to 3-5% are now commercially available. For LEDs to become unambiguously superior to mercury lamps and other conventional GUV sources, wall-plug efficiencies must exceed 25% [11], and power output must reach hundreds of milliwatts or several watts.
Wall plug efficiency (WPE, the ratio of output optical power to input electrical power) of AlGaN-based LEDs is limited by several factors, including structural, optical, and electrical quality. Acceptor doping [12,13], light extraction [14], and ohmic contact formation [15] all become more challenging with increasing AlN mole fraction, x, in AlxGa1-xN alloys. Some of these effects can be ameliorated using superlattices, stacks of thin alternating layers of differing Al composition. The interfaces between these layers create valence band discontinuities and electrostatic fields due to polarization discontinuities, improving p-type conductivity despite the deep acceptor energy level of magnesium [16]. If the superlattice period is short enough (say, below 30 Å, referred to as a short-period superlattice, SPSL), quantum confinement and tunneling lead to the formation of superlattice energy bands with band gaps larger than the bulk gap of the low composition layer, so that low Al composition layers can be used without detrimental optical absorption [17].
The semiconductor layer stacks needed for UV LEDs are typically grown using metalorganic chemical vapor deposition (MOCVD) either on sapphire, which is optically transparent and inexpensive, or on AlN, which greatly improves crystal quality. In this work, we present AlGaN-based LEDs grown on SiC substrates, which enable high structural quality at a lower cost and larger maximum wafer size than AlN. The SiC substrate can be readily removed, leaving behind a thin-film flip chip device with vastly superior topside light extraction efficiency (LEE) compared to flat or patterned sapphire substrates [18]. We have previously demonstrated up to 2% external quantum efficiency (EQE) UV-C LEDs [19] on SiC via substrate removal and backside roughening [20]. More recently, improvements have been made in n-type conductivity in AlxGa1-xN with composition, x, greater than 0.6 [21] and reduced AlN/SiC dislocation density of 2×108 cm−2 [22].
In this work, we present UV-B LEDs combining these previous improvements in n-AlGaN and AlN growth with a new optimization of the magnesium doped AlGaN (AlGaN:Mg) layers which supply holes to the LED active region. In particular, the effect of AlGaN composition is explored in layers of uniform composition, and in superlattice structures. Our results confirm the beneficial effects of superlattices with large compositional contrast and demonstrate that improved hole injection into the active region may lead to significant WPE improvements, even if at a slight cost to LEE.
2. Methods
Nitride samples were grown by MOCVD on 4H-SiC substrates (CREE) diced into 10 mm square pieces. Precursors included trimethylaluminum (TMAl), trimethylgallium (TMGa), triethylgallium, and trimethylindium for group-III elements, ammonia for nitrogen, disilane and bis(cyclopentadienyl)magnesium for n- and p-type dopants, and hydrogen carrier gas. Device layers were grown at a reactor pressure of 20 kPa.
The UV LED layer structure is represented schematically in Fig. 1. All AlN and AlGaN layer thicknesses were confirmed by a combination of in-situ monitoring, reflectometry, and transmission electron microscopy (TEM). AlN templates were grown on SiC substrates using reduced- and high-temperature layers as previously reported [22,23]. Al0.65Ga0.35N:Si n-type layers were grown under a reduced V/III ratio of ∼10, to maintain a smooth planar morphology under high Si incorporation conditions. Details of n-AlGaN growth will be presented in a future communication [21]. Active region growth conditions matched n-AlGaN conditions, without any disilane injection. Following a 27 nm unintentionally doped Al0.65Ga0.35N spacer layer, quantum wells and barriers were grown with nominal thicknesses (compositions) of 2.5 nm (40%) and 8.0 nm (65%), respectively.
Fig. 1. Epitaxial layer stacks with nominal thicknesses and compositions. For all devices, the buffer, unintentionally doped (UID), n-type, and active region layers were nominally identical. Mg-doped AlGaN layers were varied between devices to comprise either uniform, superlattice, or a combination of both layer types.
The magnesium doped AlGaN layers were grown with elevated V/III (i.e. NH3/[TMAl + TMGa]) ratios between approximately 2,000 for 20% composition, and approximately 6,000 for 80%. All were grown with a constant V/[TMAl]—differences in V/III were due to differing TMGa flow, used to control composition. All AlGaN:Mg multilayer structures had nominal total thickness of 35 nm, while composition and superlattice structure were varied as shown in Fig. 1. Short-period superlattice structures were grown with total period 17 Å (confirmed by cross-section TEM, not shown), wherein the thicknesses of the high- and low-Al composition layers were roughly the same. The list of various combinations of uniform and superlattice AlGaN:Mg structures is given in the second column of Table 1. Above the AlGaN:Mg layers, a p-GaN contact layer was grown under identical conditions for all samples. Based on calibrations performed on GaN/sapphire samples, the nominal thickness was around 10 nm, but various thicknesses and morphologies were observed for p-GaN layers grown above AlGaN, as shown in Fig. 2. The observed p-GaN morphology differences may be related to the combination of high Mg doping [24] and strain due to varied underlying AlGaN average composition, i.e. lattice constant.
Fig. 2. AFM images of p-GaN planar and 3D island layers on LED surfaces. Morphology of p-GaN grown on a 20% AlGaN uniform layer (left) representative of samples A and D is consistent with rough planar growth, whereas p-GaN grown on 10%/80% superlattice AlGaN (right) on sample C appears to form large Volmer-Weber islands. Line-scans along arrows show quantitative surface height data, indicating smooth surface (left) and island heights of 30 nm to 60 nm (right).
Table 1. Summary of UV-B LED data with uniform (U) and superlattice (SL) hole injection layers
UV LED devices were fabricated using conventional III-nitride process technologies, including a mesa etch using reactive ion etching with SiCl4/Cl2 chemistry [25] and metallization using electron-beam evaporation. The V (15 nm)/Al (80 nm)/Ni (20 nm)/Au (300 nm) n-contact was annealed under flowing N2 for 30 s at 900 °C [26]; the Ni (1 nm)/Al (100 nm)/Ni (100 nm)/Au (800 nm) p-contact was unannealed. This thin nickel p-side mirror contact was optimized with respect to reflectivity (data not shown), with some expected voltage penalty. On certain samples, diode test structures were fabricated without flip-chip, with p-contacts comprising Ni (2 nm)/Pt (3 nm)/Ti (10 nm)/Au (300 nm) for improved voltage performance.
Following topside fabrication, LEDs were flip-chip bonded to passivated SiC substrates using Au-Au thermocompression, and the growth substrate was removed using mechanical- and (dry) chemical methods [19,27]. The exposed nitrogen-face of the AlN thin-film LEDs was then roughened in aqueous KOH (0.25 M) at room temperature for 15 s [20]. The most efficient LED was encapsulated using a 1 mm thick layer of high-transparency fluororesin (CYTOP [28], optimized for spin-coating rather than molding, so no hemispherical encapsulation structure was fabricated).
Testing of UV LEDs was performed in a 75 mm diameter integrating sphere, which was recently calibrated by the manufacturer (Instrument Systems) using visible LEDs and a known spectral responsivity. The sphere calibration accuracy was confirmed with commercially available UV-B and UV-C LEDs. Pulsed measurements were performed using a signal generator and oscilloscope, providing a 10 kHz signal with an adjustable duty cycle. For low-power measurements, integrating time was allowed to increase up to 10 s (continuous) or 30 s (pulsed) as needed to maintain acceptable signal/background ratio.
3. Results
Five UV-B LED devices are summarized in Fig. 3 and Table 1, all with nominally identical substrate, AlN, n-AlGaN, and active-region structures, and differing only in the AlGaN:Mg layer(s) as shown in Fig. 1 (and in the morphology of the p-GaN contact layer as shown in Fig. 2). The target emission wavelength was 290 nm, although wavelengths between 287 nm and 299 nm were observed. The first device (A) contained an Al0.2Ga0.8N:Mg (20% composition) layer of uniform composition, and provided emission power and voltage metrics to be used as a reference for analysis of subsequent devices. Peak efficiencies of: EQE = 2.0%, WPE = 1.4%, and operational (measured at 20 A/cm2) efficiencies of: EQE = 1.3%, WPE = 0.7% were measured. Thin film flip-chip device operation voltage (at 20 A/cm2) was 8.3 V, with most of this excess voltage (above the photon energy < 4.5 eV) attributed to the resistive 1 nm thick Ni p-mirror contact. For comparison, the operating voltage of a diode test structure with low-voltage p-ohmic contact structure (Ni/Pt) was 5.7 V, corresponding to a reduction of excess voltage by nearly 70%.
Fig. 3. (a) EQE (solid) and WPE (dashed) for devices A-D. Devices with high-contrast p-superlattices (C, D) have EQE characteristics which peak at higher values, and peak at lower current densities, than the reference samples (A, B). Device C data for bare die (C) and encapsulated (C, enc.) structures. Current-voltage data (inset) shows increased voltage for all samples without a uniform 20% contact layer (B, C), and comparatively low voltage performance for both samples with a contact layer (A, D). (b) Schematic of thin film flip chip device architecture: emission through nitrogen-polar side after substrate removal, backside roughening.
To improve the LEE and reduce electron overflow (i.e. improve carrier injection efficiency, IE), the composition of the AlGaN:Mg layer was increased to 45%, expected to be more transparent at 290 nm than 20% AlGaN. However, this device (E) exhibited negligible light output power and high voltage. The 45% AlGaN layer was not sufficiently p-type to enable hole injection into the active region.
The poor electrical performance of p-type AlGaN layers with uniformly high Al composition was not surprising, and was expected to be improved by the implementation of superlattice AlGaN:Mg structures [9,29]. Thus, device B was grown with a superlattice of alternating 10% and 30% AlGaN:Mg layers (see Fig. 1) to investigate the effect of superlattice structure without changing the average Al composition (i.e. LEE) in comparison with the reference device (A). This device exhibited lower efficiency and higher voltage than device A, suggesting that the 20% net difference in composition was insufficient to produce the desired superlattice effects.
A device with a superlattice comprising 10% and 80% layers (C) was fabricated next, increasing both compositional contrast and optical transparency. In stark contrast to device E which had the same average composition of 45%, device C exhibited enhanced light output power and efficiency. A peak EQE of 4.8% (WPE = 2.8%) was measured at ∼1 A/cm2 after encapsulation, which enhanced the LEE by roughly 20%. At 20 A/cm2, the EQE (WPE) drooped to 3.2% (1.3%), while the voltage reached 10.5 V. Detailed data including pulsed measurements of this device are shown in Fig. 4. A maximum output power of 7.4 mW (76 mW/mm2) was measured at 13 V. Data for a commercially obtained UV-B LED are plotted in Fig. 4, and agree well with the manufacturer’s data sheet, confirming the calibration of the integrating sphere. Finally, the effect of integrating sphere collection duration was investigated, ensuring that the signal significantly outweighed the background. Neglecting to measure spectral power with sufficient integration time and signal to background ratio can result in erroneous power values and large errors in calculation of EQE at low currents.
Fig. 4. (a) Pulsed LIV data for the optimized UV LED device (C). Up to 7.4 mW of maximum output power (pulsed) was collected at around 13 V. (b) EQE (filled) and WPE (open) for device C (287 nm) compared with state-of-the-art commercially available (281 nm) UV-B LED. (c) Signal to background ratio (S.B.R.) for spectra collected at various injection currents and integration times. No parasitic emission was observed at any wavelength within the measured range.
Sample D was grown with a combination of a superlattice layer near the active region, and a 20% uniform AlGaN:Mg layer adjacent to the p-contact. Results indicate a promising route to future device improvements. With a similar low voltage to sample A, and improved EQE attributed to the superlattice, device D had nearly double the peak WPE (1.5%) of the reference samples A (0.8%), and B (0.6%).
4. Discussion
The aim of this experiment was primarily to study the difference between uniform AlGaN:Mg layers and superlattice layers. As expected, uniform layers show a very strong negative correlation between electrical performance and Al composition, making transparent GUV LEDs with uniform AlGaN:Mg layers impractical. Superlattice layers, in contrast, showed a much weaker negative correlation—voltage performance diminished only slightly with increased average Al composition in the AlGaN superlattice, while optoelectronic performance was markedly enhanced.
The relatively low voltage of device A is attributed to the low composition of the AlGaN:Mg layer. This 20% composition constitutes a low energy barrier for hole injection from the ohmic contact, through the p-GaN contact, into the AlGaN:Mg layer, and finally into the active region. It is also probable that this layer had increased equilibrium hole concentration, due to the reduced Mg activation energy at lower compositions [29]. On the other hand, the reduced conduction band barrier present in the 20% layer likely enabled electron overshoot beyond the active region. The planar p-GaN contact layer may have contributed to voltage reduction, while also reducing the LEE [30]. Nonetheless, the unexpectedly high EQE of this non-transparent device suggests that superior carrier injection efficiency may have outweighed losses in LEE (conversely, the 45% uniform device (E) had nominally higher LEE, but much lower IE).
There was no benefit in replacing a uniform 20% AlGaN:Mg layer (A) with a superlattice of the same average composition (B). Because there are numerous supposed benefits of superlattices on AlGaN:Mg performance—including polarization charges, valence band engineering, electron blocking characteristics, optical transparency, and Mg dopant activation [16,29,31–34]—we cannot conclusively explain the poor performance of sample B in comparison to sample A based simply on LIV experiments. However, given that all of the superlattice benefits listed above become more dominant with increasing Al content or compositional contrast, it is not surprising that the superlattice effects were much more beneficial in device C.
The UV LED with the highest efficiency in this study was sample C, with an AlGaN:Mg superlattice with 20 periods and with nominal compositions of 10% and 80%. The large composition contrast and high average composition of this layer may explain its optimal performance. As shown in Fig. 5, device C exhibited efficiency droop behavior, which has not been observed in our previous work. This may be because the devices in this work have lower dislocation density, improving IQE at very low carrier concentrations (where IQE increases rapidly with current density due to charged defect saturation [35]).
Fig. 5. EQE droop and peak wavelength shift versus current density for the optimized device (C), tested as bare die (dashed), with encapsulation (solid), and with encapsulation under pulsed operation (squares). Pulsed measurements appear to eliminate heating effects (redshift is eliminated), but do not significantly reduce droop, suggesting that efficiency droop is not primarily due to device heating.
Simultaneously, emission wavelength under continuous current injection redshifted at increased current densities, indicating device heating [36,37]. To isolate the effect of heating, continuous measurements were compared with pulsed measurements (duty cycles of 5%, 1%, and 0.1%). Pulsed current injection eliminated the redshift in the EL, without significantly affecting droop. This observation, and the observation that pulsed data are independent of duty cycle (data not shown), indicate that ohmic heating is not the primary mechanism behind efficiency droop in device C. Rather, we suggest that droop may result from Auger recombination, current crowding, or a combination of both effects. Both devices with enhanced hole injection via the high-contrast superlattice (C and D) show peak efficiency at much lower current densities than the reference samples, which is consistent with an Auger-related droop mechanism because improved hole injection increases hole concentration within the active region at a given current density. An increased hole concentration would lead to enhanced EL emission at low current densities, but also to enhanced Auger recombination at high current densities, in agreement with the observed trends.
In this series of samples, voltage seemed to be almost entirely determined by the composition of the topmost AlGaN in contact with the p-GaN layer, indicating that the superlattice layers adjacent to the active region may be further optimized without incurring a voltage penalty. Device D comprised a combination of the low-voltage-enabling 20% uniform layer of sample A with the high-EQE-enabling 10%/80% superlattice of device C. While the EQE was lower than that of device C (likely due to diminished LEE), the large voltage reduction resulted in an improved WPE compared to reference samples A and B.
Whereas achieving transparent and efficient GUV LEDs was impractical using AlGaN:Mg layers of uniform composition, it appears to be more practical with superlattice layers. High-Al composition superlattices provide benefits in carrier injection and transparency, without the detrimental effects of uniform high-composition layers [33]. LEE may have also been enhanced by the incomplete coverage of p-GaN grown above high-composition layers (Fig. 2), however this also may have led to current crowding and increased droop. Future studies will include optimization of the abruptness and smoothness of superlattice interfaces, as well as optimization of the low- and high-composition layer thicknesses independently.
5. Conclusion
AlGaN-based UV-B LEDs grown on SiC substrates have been investigated, and the effect of the AlGaN:Mg layers has been explored. Peak EQE and WPE of 4.8% and 2.8%, respectively, have been recorded at 287 nm; significant efficiency droop resulted in operational efficiencies of EQE = 3.2% and WPE = 1.3% at 20 A/cm2. Droop is unlikely to be due to heating and is rather more likely to be due to Auger processes and current crowding. The highest EQE values were achieved using an AlGaN:Mg superlattice of alternating 10% and 80% layers, which resulted in a forward voltage of 10.5 V at 20 A/cm2. Capping the 10%/80% superlattice with a 17 nm thick 20% uniform AlGaN:Mg layer reduced the forward voltage to 7.8 V at 20 A/cm2, while maintaining improved WPE over reference devices.
In summary, low Al composition (20%) uniform layers provide sufficient p-type conductivity for UV light emission; however, increasing Al content to 45% to achieve transparency results in negligible light output. Replacing the uniform layer with a superlattice layer of average composition 45% resulted in greatly increased EQE, with a moderate increase in voltage. This voltage increase was mitigated with the addition of a 20% AlGaN contact layer, which maintained high WPE because the enhanced voltage efficiency balanced the diminished LEE.
Funding
National Science Foundation (GRFP 1650114, MRSEC DMR 1121053); Solid State Lighting and Energy Electronics Center, University of California Santa Barbara.
Acknowledgements
Authors gratefully acknowledge the support of the staff at the CNSI X-ray and Microscopy Facilities, and at the UCSB Nanofab, especially A. Hopkins, D. Freeborn, and T. Bosch, without whom this work would not have been possible. Special thanks to B. K. Saif Addin for training in UV LED fabrication and encapsulation processes. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.
Disclosures
The authors declare no conflict of interest.
References
1. K. Bedell, A. H. Buchaklian, and S. Perlman, “Efficacy of an automated multiple emitter whole-room Ultraviolet-C disinfection system against coronaviruses MHV and MERS-CoV,” Infect. Control Hosp. Epidemiol. 37(5), 598–599 (2016). [CrossRef]
2. R. J. Fischer, D. H. Morris, S. Sarchette, J. Matson, T. Bushmaker, C. K. Yinda, S. N. Seifert, A. Gamble, B. N. Williamson, S. D. Judson, J. O. Lloyd-smith, and V. J. Munster, “Assessment of N95 respirator decontamination and re-use for SARS-CoV-2,” medRxiv Prepr. (2020).
3. K. Song, M. Mohseni, and F. Taghipour, “Application of ultraviolet light-emitting diodes (UV-LEDs) for water disinfection: A review,” Water Res. 94, 341–349 (2016). [CrossRef]
4. M. T. Neves-Petersen, Z. Gryczynski, J. Lakowicz, P. Fojan, S. Pedersen, E. Petersen, and S. Bjørn Petersen, “High probability of disrupting a disulphide bridge mediated by an endogenous excited tryptophan residue,” Protein Sci. 11(3), 588–600 (2009). [CrossRef]
5. M. Correia, M. T. Neves-Petersen, P. B. Jeppesen, S. Gregersen, and S. B. Petersen, “UV-Light Exposure of Insulin: Pharmaceutical Implications upon Covalent Insulin Dityrosine Dimerization and Disulphide Bond Photolysis,” PLoS One 7(12), e50733 (2012). [CrossRef]
6. S. E. Beck, H. B. Wright, T. M. Hargy, T. C. Larason, and K. G. Linden, “Action spectra for validation of pathogen disinfection in medium-pressure ultraviolet (UV) systems,” Water Res. 70, 27–37 (2015). [CrossRef]
7. M. Shatalov, W. Sun, A. Lunev, X. Hu, A. Dobrinsky, Y. Bilenko, J. Yang, M. Shur, R. Gaska, C. Moe, G. Garrett, and M. Wraback, “AlGaN Deep Ultraviolet Light-Emitting Diodes with External Quantum Efficiency above 10%,” Appl. Phys. Express 5(8), 082101 (2012). [CrossRef]
8. S. I. Inoue, N. Tamari, and M. Taniguchi, “150 mW deep-ultraviolet light-emitting diodes with large-area AlN nanophotonic light-extraction structure emitting at 265 nm,” Appl. Phys. Lett. 110(14), 141106 (2017). [CrossRef]
9. J. Zhang, Y. Gao, L. Zhou, Y.-U. Gil, and K.-M. Kim, “Transparent deep ultraviolet light-emitting diodes with a p-type AlN ohmic contact layer,” in Light-Emitting Devices, Materials, and Applications, M. Strassburg, J. K. Kim, and M. R. Krames, eds. (SPIE, 2019), 10940, p. 1.
10. Z. Zhang, M. Kushimoto, T. Sakai, N. Sugiyama, L. J. Schowalter, C. Sasaoka, and H. Amano, “A 271.8 nm deep-ultraviolet laser diode for room temperature operation,” Appl. Phys. Express 12(12), 124003 (2019). [CrossRef]
11. S. E. Beck, H. Ryu, L. A. Boczek, J. L. Cashdollar, K. M. Jeanis, J. S. Rosenblum, O. R. Lawal, and K. G. Linden, “Evaluating UV-C LED disinfection performance and investigating potential dual-wavelength synergy,” Water Res. 109, 207–216 (2017). [CrossRef]
12. M. L. Nakarmi, K. H. Kim, K. Zhu, J. Y. Lin, and H. X. Jiang, “Transport properties of highly conductive n-type Al-rich AlxGa1−xN(x≥0.7),” Appl. Phys. Lett. 85(17), 3769–3771 (2004). [CrossRef]
13. X. T. Trinh, D. Nilsson, I. G. Ivanov, E. Janzén, A. Kakanakova-Georgieva, and N. T. Son, “Stable and metastable Si negative-U centers in AlGaN and AlN,” Appl. Phys. Lett. 105(16), 162106 (2014). [CrossRef]
14. D. Y. Kim, J. H. Park, J. W. Lee, S. Hwang, S. J. Oh, J. Kim, C. Sone, E. F. Schubert, and J. K. Kim, “Overcoming the fundamental light-extraction efficiency limitations of deep ultraviolet light-emitting diodes by utilizing transverse-magnetic-dominant emission,” Light: Sci. Appl. 4(4), e263 (2015). [CrossRef]
15. K. Mori, K. Takeda, T. Kusafuka, M. Iwaya, T. Takeuchi, S. Kamiyama, I. Akasaki, and H. Amano, “Low-ohmic-contact-resistance V-based electrode for n-type AlGaN with high AlN molar fraction,” Jpn. J. Appl. Phys. 55(5S), 05FL03 (2016). [CrossRef]
16. P. Kozodoy, Y. P. Smorchkova, M. Hansen, H. Xing, S. P. DenBaars, U. K. Mishra, A. W. Saxler, R. Perrin, and W. C. Mitchel, “Polarization-enhanced Mg doping of AlGaN/GaN superlattices,” Appl. Phys. Lett. 75(16), 2444–2446 (1999). [CrossRef]
17. C. Weisbuch, “Fundamental Properties of III-V Semiconductor Two-Dimensional Quantized Structures: The Basis for Optical and Electronic Applications,” in Semiconductors and Semimetals (Bell Telephone Laboratories, Incorporated, 1987), 24.
18. H. Wan, S. Zhou, S. Lan, and C. Gui, “Light Extraction Efficiency Optimization of AlGaN-Based Deep-Ultraviolet Light-Emitting Diodes,” ECS J. Solid State Sci. Technol. 9(4), 046002 (2020). [CrossRef]
19. B. K. Saifaddin, A. S. Almogbel, C. J. Zollner, F. Wu, B. Bonef, M. Iza, S. Nakamura, S. P. Denbaars, and J. S. Speck, “AlGaN Deep-Ultraviolet Light-Emitting Diodes Grown on SiC Substrates,” ACS Photonics 7(3), 554–561 (2020). [CrossRef]
20. B. K. Saifaddin, M. Iza, H. Foronda, A. Almogbel, C. J. Zollner, F. Wu, A. Alyamani, A. Albadri, S. Nakamura, S. P. DenBaars, and J. S. Speck, “Impact of roughening density on the light extraction efficiency of thin-film flip-chip ultraviolet LEDs grown on SiC,” Opt. Express 27(16), A1074 (2019). [CrossRef]
21. A. S. Almogbel, C. J. Zollner, B. K. SaifAddin, M. Iza, J. Wang, Y. Yao, M. Wang, I. Prozheev, F. Tuomisto, A. Albadri, S. Nakamura, S. P. DenBaars, and J. S. Speck, “Growth of highly-conductive Al-rich AlGaN:Si with low group III vacancy concentration,” In Prep. (2020).
22. C. J. Zollner, A. Almogbel, Y. Yao, B. K. SaifAddin, F. Wu, M. Iza, S. P. DenBaars, J. S. Speck, and S. Nakamura, “Reduced dislocation density and residual tension in AlN grown on SiC by metalorganic chemical vapor deposition,” Appl. Phys. Lett. 115(16), 161101 (2019). [CrossRef]
23. H. M. Foronda, F. Wu, C. Zollner, M. E. Alif, B. Saifaddin, A. Almogbel, M. Iza, S. Nakamura, S. P. DenBaars, and J. S. Speck, “Low threading dislocation density aluminum nitride on silicon carbide through the use of reduced temperature interlayers,” J. Cryst. Growth 483, 134–139 (2018). [CrossRef]
24. M. Akiba, H. Hirayama, Y. Tomita, Y. Tsukada, N. Maeda, and N. Kamata, “Growth of flat p-GaN contact layer by pulse flow method for high light-extraction AlGaN deep-UV LEDs with Al-based electrode,” Phys. Status Solidi C 9(3–4), 806–809 (2012). [CrossRef]
25. D. Selvanathan, L. Zhou, V. Kumar, and I. Adesida, “Low resistance Ti/Al/Mo/Au Ohmic contacts for AlGaN/GaN heterostructure field effect transistors,” phys. stat. sol. (a) 194(2), 583–586 (2002). [CrossRef]
26. B. Saifaddin, C. J. Zollner, A. Almogbel, F. Wu, A. Albadri, A. Al Yamani, M. Iza, S. Nakamura, S. P. Denbaars, and J. S. Speck, “Developments in AlGaN and UV-C LEDs grown on SiC,” Proc. SPIE 10554, 105541E (2018). [CrossRef]
27. B. K. Saifaddin, A. Almogbel, C. J. Zollner, H. Foronda, A. Alyamani, A. Albadri, M. Iza, S. Nakamura, S. P. Denbaars, and J. S. Speck, “Fabrication technology for high light-extraction ultraviolet thin-film flip-chip (UV TFFC) LEDs grown on SiC,” Semicond. Sci. Technol. 34(3), 035007 (2019). [CrossRef]
28. S. Nagai, K. Yamada, A. Hirano, M. Ippommatsu, M. Ito, N. Morishima, K. Aosaki, Y. Honda, H. Amano, and I. Akasaki, “Development of highly durable deep-ultraviolet AlGaN-based LED multichip array with hemispherical encapsulated structures using a selected resin through a detailed feasibility study,” Jpn. J. Appl. Phys. 55(8), 082101 (2016). [CrossRef]
29. T. C. Zheng, W. Lin, R. Liu, D. J. Cai, J. C. Li, S. P. Li, and J. Y. Kang, “Improved p -type conductivity in Al- rich AlGaN using multidimensional Mg-doped superlattices,” Sci. Rep. 6(1), 21897 (2016). [CrossRef]
30. N. Susilo, J. Enslin, L. Sulmoni, M. Guttmann, U. Zeimer, T. Wernicke, M. Weyers, and M. Kneissl, “Effect of the GaN:Mg Contact Layer on the Light-Output and Current-Voltage Characteristic of UVB LEDs,” Phys. Status Solidi A 215(10), 1700643 (2018). [CrossRef]
31. J. Zhang, Y. Gao, L. Zhou, and Y. Gil, “Surface hole gas enabled transparent deep ultraviolet light-emitting diode,” Semicond. Sci. Technol. 33(7), 07LT01 (2018). [CrossRef]
32. K. Sato, S. Yasue, K. Yamada, S. Tanaka, T. Omori, S. Ishizuka, S. Teramura, Y. Ogino, S. Iwayama, H. Miyake, M. Iwaya, T. Takeuchi, S. Kamiyama, and I. Akasaki, “Room-temperature operation of AlGaN ultraviolet-B laser diode at 298 nm on lattice-relaxed Al0.6Ga0.4N/AlN/sapphire,” Appl. Phys. Express 13(3), 031004 (2020). [CrossRef]
33. H. Hirayama, Y. Tsukada, T. Maeda, and N. Kamata, “Marked enhancement in the efficiency of deep-ultraviolet AlGaN light-emitting diodes by using a multiquantum-barrier electron blocking layer,” Appl. Phys. Express 3(3), 031002 (2010). [CrossRef]
34. L. Li, Y. Zhang, S. Xu, W. Bi, Z.-H. Zhang, H.-C. Kuo, L. Li, Y. Zhang, S. Xu, W. Bi, Z.-H. Zhang, and H.-C. Kuo, “On the Hole Injection for III-Nitride Based Deep Ultraviolet Light-Emitting Diodes,” Materials 10(10), 1221 (2017). [CrossRef]
35. A. David, N. G. Young, C. Lund, and M. D. Craven, “Review—The Physics of Recombinations in III-Nitride Emitters,” ECS J. Solid State Sci. Technol. 9(1), 016021 (2020). [CrossRef]
36. Y. P. Varshni, “Temperature dependence of the energy gap in semiconductors,” Physica 34(1), 149–154 (1967). [CrossRef]
37. J. Cho, C. Sone, Y. Park, and E. Yoon, “Measuring the junction temperature of III-nitride light emitting diodes using electro-luminescence shift,” Phys. Status Solidi A 202(9), 1869–1873 (2005). [CrossRef]
