Open Access
Add to BibSonomy
Abstract
AlGaN heterostructure solar-blind avalanche photodiodes (APDs) were fabricated on a double-polished AlN/sapphire template based on a separate absorption and multiplication (SAM) back-illuminated configuration. By employing AlGaN heterostructures with different Al compositions across the entire device, the SAM APD achieved an avalanche gain of over 1×105 at an operated reverse bias of 92 V and a low dark current of 0.5 nA at the onset point of breakdown. These excellent performances were attributed to the acceleration of holes by the polarization electric field with the same direction as the reverse bias and higher impact ionization coefficient of the low-Al-content Al0.2Ga0.8N in the multiplication region. However, the Al0.2Ga0.8N layer produced a photocurrent response in the out of the solar-blind band. To retain the solar-blind detecting characteristic, a periodic Si3N4/SiO2 photonic crystal was deposited on the back of the AlN/sapphire template as an optical filter. This significantly improved the solar-blind characteristic of the device.
© 2020 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Solar-blind ultraviolet (UV) detection is attracting extensive attention owing to its wide range of applications for military and civilian purposes, such as biological and chemical threat alarms, missile warning and guidance, shipboard communication, corona monitoring, environmental detection, and deep space exploration [1–4]. AlGaN alloys are suitable materials for the fabrication of solar-blind avalanche photodiodes (APDs) because of their intrinsic solar-blind properties (Al > 40%) and tunable wide band-gap (∼3.4−6.2 eV) [5]. In contrast to photomultiplier tubes, which are large and fragile, AlGaN alloys exhibit high hardness, light weight, and low energy consumption [6,7]. When the applied voltage reaches the critical level of APDs, the AlGaN detector operates in Geiger mode and achieves a significant avalanche gain, enabling the detection of weak solar-blind UV signals. However, there are still several challenges in the preparation of high-performance AlGaN photodetectors, such as the slow migration of Al atoms on the epitaxial surface, the difficulties in p-type doping caused by the high Mg activation energy and low ionization coefficients of carriers in high-Al-content AlGaN alloys [8–10]. Therefore, the multiplication gain of AlGaN APDs is limited to the order of ∼102−103. Recently, AlGaN APDs with multiplication gains of 5500 and 15000 were reported [10,11]. Nevertheless, the performance of solar-blind AlGaN APDs is inferior compared to that of GaN UV devices, which have higher impact ionization coefficients (IIC) of carriers and a more established epitaxial technique [4,12,13].
In addition, the Al content in AlGaN APDs needs to be modulated above 0.4 to meet the band-gap requirements for solar-blind UV detection, because high-Al-content AlGaN experiences the epitaxial problem. Zoran et al. reported a silicon UV p-n detector combined with a NaF/Y2O3 photonic crystal (PC) that suppresses visible and infrared light absorption [14]. Photonic crystals are artificial microstructures formed by the periodic arrangement of materials with different refractive indices. When the electromagnetic wave propagates in the photonic bandgap material, it is modulated by Bragg scattering, and the electromagnetic wave forms an energy band structure. A photon whose energy is within the photonic band gap cannot enter the crystal. Accordingly, the combination of the device with a photonic crystal UV filter should be considered to achieve solar-blind detection. In our previous work, we also theoretically discussed the rationale of PC-integrated APDs [15].
In this study, we considered the IIC and polarization comprehensively and designed and fabricated a high-performance AlGaN APD. AlGaN heterostructures were employed based on a separate absorption and multiplication (SAM) structure to achieve higher hole IIC and introduce a polarization electric field that can accelerate hole transport. The fabricated AlGaN solar-blind APD devices exhibit a high avalanche gain of up to 105 and a low dark current density in the order of 10−6 A/cm2 at the onset point of breakdown. The corresponding physical mechanisms are investigated by analyzing the electric field distribution and energy band of the device. In addition, a periodic Si3N4/SiO2 PC was deposited on the back of the device as an optical filter in order to suppress the photocurrent response from bands outside the solar-blind UV range due to the low-Al-content (LAC) Al0.2Ga0.8N multiplication layer. This significantly improved the solar-blind characteristic of the device.
2. Methods
As shown in Fig. 1(a), the AlGaN SAM APD was fabricated by metal organic chemical vapor deposition on a double-polished AlN/Sapphire template using TMGa and TMAl as metal precursors and ammonia as the nitrogen source. On the top of the device, 50 nm of p-type GaN was used to obtain an ohmic contact with a hole concentration of 2×1018, as confirmed by Hall-effect measurements at room temperature. The SAM heterostructure consisted of a 120 nm-thick p-type Al0.1Ga0.9N layer, a 250 nm-thick unintentionally doped Al0.2Ga0.8N multiplication layer, a 60 nm-thick n-type Al0.35Ga0.65N interlayer for modulating the electric field distribution, a 200 nm-thick unintentionally doped Al0.4Ga0.6N absorption layer, and a 600 nm-thick n-type Al0.5Ga0.5N layer. A 1.2 µm-thick i-Al0.5Ga0.5N was employed to improve the crystal quality during the epitaxial process. The hole concentration of Al0.1Ga0.9N was measured to be 1×1018 cm−3 and the electron concentrations of the Al0.35Ga0.65N, Al0.5Ga0.5N, and unintentionally doped AlGaN layers were measured to be 1×1018, 2×1018, and 1×1016 cm−3, respectively.
Fig. 1. (a) Schematic of AlGaN heterostructure SAM APD with PC. (b) Bird view of the fabricated device.
Table 1. Recent progress in development of AlGaN/GaN avalanche photodetectors.
During the fabrication process, the wafer was cleaned with acetone, ethanol and deionized water consecutively. Hydrochloric acid was employed to remove surface oxides. Subsequently, the wafer was etched to n-Al0.5Ga0.5N layer by ICP etching combined with photolithography process. Next, the metal was evaporated on the wafer by electron beam evaporation under 10−7 Torr pressure condition, and lift-off process was carried out to form electrode morphology. Ti/Al/Ni/Au (30/130/50/100 nm) and Ni/Au (20/20 nm) were deposited on the n-type Al0.5Ga0.5N and p-type GaN, respectively. The metal on the n-type layer was annealed at 850 °C for 30 s in N2 ambient and that on the p-type layer was annealed at 500 °C for 12 min in air ambient to ultimately engender Ohmic contact. Inductively coupled plasma etching was used together with lithography to fabricate the double-mesa structure to suppress leakage current and adjust the electric field distribution. As shown in Fig. 1(b), the first mesa with diameter of 30 µm was etched to i-Al0.2Ga0.8N multiplication layer. The second mesa with diameter of 80 µm was etched to n-Al0.5Ga0.5N layer. Then the APD was dipped in KOH solution with 0.3 mol/L at 100 °C to reduce etching damage. Subsequently, SiO2 was deposited on the APDs by plasma enhanced chemical vapor deposition (PECVD) to form passivation layers, followed by opening test windows [22,23]. A periodic one-dimensional (1-D) PC consisting in an (L/2 H L/2)14 sandwiched structure, where L (51.5 nm) and H (38 nm) represent SiO2 and SiNx, respectively, was deposited on the back of the AlN/sapphire template by PECVD to ensure solar-blind detecting characteristics. The specific thicknesses of SiO2 and Si3N4 are chosen to keep the same center wavelength of 315 nm and ensure a 280 nm reflection rising edge. Finally, SiH4, N2O, and N2 were used as gas sources for growing SiO2 and Si3N4.
In this work, the Silvaco Atlas TCAD was employed to analyze the characteristics of the avalanche photodiodes. Continuity equation and Poisson’s equation were applied for numerical procedures. Newton iteration method was employed to solve nonlinear algebraic system. In addition, physical models consists of concentration dependent mobility, carrier statistical, impact ionization (BBT and Selb) and recombination (surface, trap, SRH and Auger). As for mesh size, the x-axis-mesh is uniformly distributed with spacing of 0.1 µm. In the y-axis-mesh, the mesh spacing is 0.01 µm in the bulk region and 1 nm around the heterojunction interface for accuracy of calculation.
3. Results and discussions
A Keithley 4200 semiconductor analyzer and a cascade probe station were employed to measure the I-V characteristics of the fabricated APD, as shown in Fig. 2. The UV back-illumination was performed by a Xe arc lamp source through an optical fiber. Both dark and light I-V curves present a stable trend before the reverse bias of 40 V.
Fig. 2. I-V characteristics of the AlGaN SAM APD in the dark and under UV illumination at operated reverse bias. The inset shows a schematic of the measurement set up.
As the operating voltage exceeds 40 V, the dark current gradually increases with increasing applied bias. At the applied voltage of 88 V, the dark current starts to increase steeply, which indicates the avalanche breakdown of the APD. Notably, the device exhibits a low dark current 0.5 nA at the onset point of breakdown (86 V). The abrupt current rising criteria is defined as the current difference exceeds an order of magnitude within 0.5 V range of reverse bias, and the breakdown voltage is the initial voltage at which the current meets the rising criteria. The photo-generated carriers play a dominant role of impact ionization in photocurrent, which shows a smooth trend until avalanche breakdown occurs. Measured by a calibrated silicon photodetector, the input power to the APD is 1.3×10−5 W/cm2 with incident light set at 275 nm for I-V test. Notably, as the applied voltage increases further, the current increase tends to be flat and the avalanche gain presents a slight reduction after the peak. The impact ionization coefficient of hole increases with the increase of electric field intensity before avalanche. When the reverse bias increases further, the impact ionization and impurity scattering lead to the saturation of the hole mean free path. Therefore, the carrier will not be accelerated infinitely under the effect of electric field, thus resulting in the holes reaching the threshold energy to trigger impact ionization.
The avalanche gain is calculated as the difference between the primary multiplied current and multiplied dark current normalized by the difference between the primary unmultiplied current and the unmultiplied dark current [3]. It can be seen that the gain presents a significant upward trend at breakdown voltage and reaches a maximum of 1.12×105 at 92 V applied voltage. The expression of multiplication gain is M=(Ilight-Idark)/Iu, where M is multiplication gain, Ilight is light current and Idark is dark current, Iu is unmultipied mean current difference. In this study, to avoid the fluctuation of the current at small voltages, we take the average current value in the range of 10–40 V as the unity gain is kept with steady current trend. Hence, the multiplication gain is proportional to the light-dark current difference while Iu remains constant. In Fig. 2, photo-generated carriers play a dominant role of impact ionization in photocurrent. However, when impact ionization reaches the threshold (89 V), the contribution of photo-generated carriers to the current is no longer enhanced, and the photocurrent tends to be saturated. Whereas, dark current is primarily related to trap assisted tunneling, band-to-band tunneling, generation-recombination current and surface leakage current. Relatively, the electric field and applied voltage will have impacts on tunneling currents [24]. Therefore, as the reverse bias increases, the rising electric field intensity leads the dark current to increase slightly after reaching saturation. Consequently, the light-dark current difference does not increase all the time, and it exhibits a slight downward trend after reaching saturation, thus resulting in the reduction of gain.
In addition, there is a noteworthy gain fluctuation from 75 to 82 V. This is ascribed to the fluctuation of dark current before avalanche breakdown. As mentioned above, the generation and recombination (g-r) current will have impacts on the dark current. When the APD is operated with reverse bias, the device is kept in a non-equilibrium state. As the applied voltage increases, the impact ionization becomes more intensely and produces more excited carriers. Because the number of electrons and holes increases compared to the thermal equilibrium state, the probability of electron-hole recombination will increase. There will be a short period of net recombination leading to the fluctuation of dark current. Repeated measurements verify that the dark current fluctuation is more likely to occur near the onset of avalanche breakdown. The fluctuation of gain before breakdown has been also observed in other reports [19].
To further elucidate the properties of the AlGaN APD, we numerically simulated the electric field distribution and energy band of the device by Silvaco Atlas TCAD [25], as illustrated in Fig. 3. AlGaN materials with wurtzite structure have central asymmetry, so the positive and negative charge centers in the unit cells do not coincide and will form electric moments. It is the reason for the existence of spontaneous polarization effect. In addition, due to the different lattice constants, the AlxGa1-xN material heterojunction with different Al compositions has a strong piezoelectric polarization effect, and the epitaxial thin film will be influenced by the strain and stress. In this work, the photodetector is composed of various AlxGa1-xN/AlyGa1-yN heterostructures, so there will be strong piezoelectric polarization effects throughout the device.
Fig. 3. (a) Simulated electric field distribution at a reverse bias of 90 V. The right-hand axis indicates the calculated polarization charge concentration. (b) Energy band profiles for ionization-enhanced APD (blue) and conventional high-Al-content APD (black) at zero bias, respectively.
The electric field distribution demonstrates the ability of the SAM structure to regulate electric field intensity. Ppz1 and Ppz2 are the piezoelectric polarization intensity in i-Al0.2Ga0.8N multiplication region introduced by n-Al0.35Ga0.65N interlayer and p-Al0.2Ga0.8N layer, respectively. The calculated pure piezoelectric polarization intensity in the Al0.2Ga0.8N multiplication layer was 1.57 ×10−3 C/cm2. This polarization electric field in the same direction as the applied voltage will promote the impact ionization of holes in the avalanche region. Furthermore, it can be observed that there is a significant polarization charge concentration of 3.38 C/cm−3 at the interface of Al0.2Ga0.8N/Al0.35Ga0.65N in Fig. 3(a). Such a high density of negative charges caused by the polarization effect will induce a built-in electric field in the interlayer.
As depicted in Fig. 3(b), compared to the conventional “approximate-homostructure”, the ionization-enhanced APD exhibits a steep drop in the valence band potential due to the built-in electric field. Herein, the “approximate-homostrucutre” is a referenced APD with Al0.4Ga0.6N alloy throughout the multiplication layer to absorption layer, while the other parts are the same as the ionization-enhanced structure. The drop in potential further accelerates the transfer of holes from the absorption region to the multiplication region, which contributes to the impact ionization of holes. In addition, the numerous negative charges form barriers in the conduction band [Fig. 3(b)], which hinder the transport of electrons, ensuring the mechanism of hole initiating multiplication. The IIC of holes will increase with decreasing Al composition in the AlGaN alloys [26,27]. Therefore, the polarization effect and the high IIC of the LAC Al0.2Ga0.8N in the multiplication region account for the high-gain performance of the SAM APD.
In addition, the p-layer consists of a GaN/Al0.1Ga0.9N heterostructure. The p-GaN cap layer is designed for preparing better Ohmic contact. Because of the existence of piezoelectric effects, charges will be accumulated at the interface of GaN/Al0.1Ga0.9N heterojunction, thus leading to the potential jump in Figs. 3(a) and 3(b). Furthermore, as shown in Fig. 3(b), a potential barrier generated in conduction band by the piezoelectric polarization will obstruct the electron transport from entering the multiplication region. It is beneficial to reduce the excessive noise and promote single-carrier (hole) to trigger avalanche.
Meanwhile, a comprehensive survey table is summarized to clearly demonstrate the development of AlGaN-based solar-blind APD (Table 1). The back-illuminated SAM APD in this work exhibits superior performances compared with other reported APDs. However, the gain of GaN UV avalanche detector maintains a high level due to its mature epitaxial technique and high crystal quality. Therefore, the multiplication gain of APD could be further improved through the improvement of epitaxial technology. Moreover, device design can also improve the APD performance. In addition to the SAM merits of triggering impact ionization by pure holes, the device structure could be further optimized. The multiplication region is one of the factors that could be optimized, because it can alter the probability of carrier impact ionization while influencing the critical electric field. The insert-layer is able to regulate electric field distribution. The thickness and doping concentration of interlayer will influence the electric field intensity in the functional regions. Reasonable interlayer design can prevent potential premature breakdown and control the carrier transport. Furthermore, a GaN/AlN periodically stacked-structure was also proposed and demonstrated to manipulate the carrier transport and realize a highly efficient and controllable carrier multiplication process [28]. Consequently, the parameters of device could be further optimized and the performance can be improved by introducing energy-band engineering with specific structure.
Whereas, the LAC AlGaN will absorb light outside the solar-blind light range, resulting in weaker out-of-band signal suppression. To suppress the photocurrent response from out of solar-blind bands, a 1-D periodic Si3N4/SiO2 PC was installed on the sapphire-side of the device.
Figure 4(a) illustrates the reflection spectrum of the 1-D PC obtained with Ocean Optics DH-2000-BAL (Light source) and Nanocalc-XR in the back-illuminated condition. It is exhibited that the reflectivity increases sharply at a wavelength of 280 nm and exceeds 90% between 300 and 340 nm. This confirms the good optical filtering characteristics of the Si3N4/SiO2 PC. To further investigate the effects of the 1-D PC on the SAM APDs, the spectral responsivity of the devices with and without PC were also measured under back illumination by using a calibrated monochromator and a Xe arc lamp as the light source. The power of the monochromatic light was measured with calibrated silicon photodetector. In addition, to prevent device damage, the APD should not maintain the avalanche state for a long time. Accordingly, we obtained the responsivity spectra at lower biases to illustrate the solar-blind detection performance and verify the filtering effect of the 1-D PC.
Fig. 4. (a) Reflection spectrum of the PC. The inset shows a schematic of the 1-D PC structure. (b) Spectral responsivity of the SAM APDs with (red) and without (purple) PC at different reverse biases. (c) I-V curves under UV illumination of APDs with (grey) and without (red) photonic crystal (PC).
As depicted in Fig. 4(b), both samples display pronounced peaks at 280 nm, indicating the nature of solar-blind detection. In addition, the device without PC displays a weak response in the wavelength range of 320–340 nm, which is attributed to the LAC AlGaN absorption. In contrast, the device with the 1-D PC exhibits superior out- of-band cut-off edges and effectively reduces the absorption of light outside the solar-blind band by two orders of magnitude. In Fig. 4(a), the cut-off wavelength in the reflection spectrum is 348 nm and the reflectance at 340 nm is still at 90%. Therefore, it can be observed from Fig. 4(b), that the responsivity at 340 nm remains at a low order of magnitude, while the PC has a lower impact after 350 nm. Consequently, it is confirmed that the 1-D PC significantly improves the solar-blind detecting capability of the AlGaN SAM APD.
In addition, the slightly lower responsivity of the sample with PC is attributed to the 10% reflectivity at wavelengths below 280 nm, which is negligible with the rising reverse bias. When the wavelength reaches 305 nm, the reflectivity exceeds 90%, which corresponds to the onset of suppression in the response of the APD. As shown in the Fig. 4(c), the photocurrents of the APDs are illustrated with and without photonic crystal (PC), respectively. It can be observed that the sample with PC has a slightly lower photocurrent than that of the APD without PC at low reverse bias. Since the photonic crystal retains 10% reflectivity at wavelength below 280 nm, the back-illumination light intensity will be affected by the PC. However, as the applied voltage increases, the photocurrents of the devices gradually coincide and show the same trend. Because multiplication initiated by the photo-generated carriers plays a dominant role in photocurrent at high reverse bias, the effect of light intensity becomes weaker, which is consistent with the trend observed in the spectral responsivity.
4. Conclusion
In summary, we comprehensively considered the IIC and polarization and proposed an enhanced AlGaN heterostructure solar-blind APD. By employing polarization engineering and an entire LAC Al0.2Ga0.8N as the multiplication region to enhance the impact ionization, the APD achieved a high avalanche gain of over 1×105 at the reverse bias of 92 V and a low dark current of 0.5 nA at 86 V. Simulated electric field and energy band demonstrated that the polarization electric field introduced by the heterostructure facilitates the transport of holes. Although the LAC AlGaN had higher hole IICs, the long-wavelength UV absorption in the multiplication region was unfavorable for solar-blind detection. Hence, a 1-D Si3N4/SiO2 periodic crystal was employed as an optical filter and it effectively improved the solar-blind detection capability of the APD.
Funding
National Key Research and Development Program of China (2016YFB0400903); National Natural Science Foundation of China (61634002); NSAF (U1830109); Key R&D Project of Jiangsu Province, China (BE2016174); “333” Project of Jiangsu Province, China (BRA2018040); Zhejiang Lab’s International Talent Fund for Young Professionals.
Disclosures
The authors declare no conflicts of interest.
References
1. S. Oh, C. K. Kim, and J. Kim, “High Responsivity beta-Ga2O3 Metal-Semiconductor-Metal Solar-Blind Photodetectors with Ultraviolet Transparent Graphene Electrodes,” ACS Photonics 5(3), 1123–1128 (2018). [CrossRef]
2. H. Chen, J. T. Zhang, Z. X. Chen, H. Q. Liu, X. Z. Ma, Q. G. Li, G. Chu, and S. Chu, “Avalanche solar blind photodetectors based on single crystalline Mg0.47Zn0.53O thin film on Ga:ZnO substrate,” J. Phys. D: Appl. Phys. 51(17), 175104 (2018). [CrossRef]
3. R. McClintock, A. Yasan, K. Minder, P. Kung, and M. Razeghi, “Avalanche multiplication in AlGaN based solar-blind photodetectors,” Appl. Phys. Lett. 87(24), 241123 (2005). [CrossRef]
4. Z. Y. Zheng, Y. X. Mai, and G. Wang, “Monte Carlo study of device characteristics of GaN-based avalanche photodiode devices,” J. Appl. Phys. 106(2), 023716 (2009). [CrossRef]
5. Z. Q. Huang, J. F. Li, W. L. Zhang, and H. Jiang, “AlGaN Solar-Blind Avalanche Photodiodes with Enhanced Multiplication Gain Using Back-Illuminated Structure,” Appl. Phys. Express 6(5), 054101 (2013). [CrossRef]
6. H. Wu, W. Wu, H. Zhang, Y. Chen, Z. Wu, G. Wang, and H. Jiang, “All AlGaN epitaxial structure solar-blind avalanche photodiodes with high efficiency and high gain,” Appl. Phys. Express 9(5), 052103 (2016). [CrossRef]
7. H. D. Sun, J. Yin, E. F. Pecora, L. Dal Negro, R. Paiella, and T. D. Moustakas, “Deep-Ultraviolet Emitting AlGaN Multiple Quantum Well Graded-Index Separate-Confinement Heterostructures Grown by MBE on SiC Substrates,” IEEE Photonics J. 9(4), 1–9 (2017). [CrossRef]
8. O. Ambacher, B. Foutz, J. Smart, J. R. Shealy, N. G. Weimann, K. Chu, M. Murphy, A. J. Sierakowski, W. J. Schaff, L. F. Eastman, R. Dimitrov, A. Mitchell, and M. Stutzmann, “Two dimensional electron gases induced by spontaneous and piezoelectric polarization in undoped and doped AlGaN/GaN heterostructures,” J. Appl. Phys. 87(1), 334–344 (2000). [CrossRef]
9. J. Kim, M. H. Ji, T. Detchprohm, R. D. Dupuis, J. H. Ryou, A. K. Sood, N. D. Dhar, and J. Lewis, “Comparison of AlGaN p-i-n ultraviolet avalanche photodiodes grown on free-standing GaN and sapphire substrates,” Appl. Phys. Express 8(12), 122202 (2015). [CrossRef]
10. T. Tut, M. Gokkavas, A. Inal, and E. Ozbay, “AlxGa1-xN-based avalanche photodiodes with high reproducible avalanche gain,” Appl. Phys. Lett. 90(16), 163506 (2007). [CrossRef]
11. Y. Huang, D. J. Chen, H. Lu, K. X. Dong, R. Zhang, Y. D. Zheng, L. Li, and Z. H. Li, “Back-illuminated separate absorption and multiplication AlGaN solar-blind avalanche photodiodes,” Appl. Phys. Lett. 101(25), 253516 (2012). [CrossRef]
12. M.-H. Ji, J. Kim, T. Detchprohm, Y. Zhu, S.-C. Shen, and R. D. Dupuis, “p-i-p-i-n Separate Absorption and Multiplication Ultraviolet Avalanche Photodiodes,” IEEE Photonics Technol. Lett. 30(2), 181–184 (2018). [CrossRef]
13. S. Choi, H. J. Kim, Y. Zhang, X. G. Bai, D. Yoo, J. Limb, J. H. Ryou, S. C. Shen, P. D. Yoder, and R. D. Dupuis, “Geiger-Mode Operation of GaN Avalanche Photodiodes Grown on GaN Substrates,” IEEE Photonics Technol. Lett. 21(20), 1526–1528 (2009). [CrossRef]
14. Z. Djuric, T. Dankovic, Z. Jaksic, D. Randjelovic, R. Petrovic, W. Ehrfeld, A. Schmidt, and K. Hecker, “A silicon UV flame detector utilizing photonic crystal,” Proc. SPIE 3680, 601–610 (1999). [CrossRef]
15. Q. Cai, M. Ge, J. J. Xue, L. Q. Hu, D. J. Chen, H. Lu, R. Zhang, and Y. D. Zheng, “An Improved Design for Solar-Blind AlGaN Avalanche Photodiodes,” IEEE Photonics J. 9(4), 1–7 (2017). [CrossRef]
16. T. Tut, S. Butun, B. Butun, M. Gokkavas, H. B. Yu, and E. Ozbay, “Solar-blind AlxGa1-xN-based avalanche photodiodes,” Appl. Phys. Lett. 87(22), 223502 (2005). [CrossRef]
17. L. Sun, J. L. Chen, J. F. Li, and H. Jiang, “AlGaN solar-blind avalanche photodiodes with high multiplication gain,” Appl. Phys. Lett. 97(19), 191103 (2010). [CrossRef]
18. Z. G. Shao, D. J. Chen, H. Lu, R. Zhang, D. P. Cao, W. J. Luo, Y. D. Zheng, L. Li, and Z. H. Li, “High-Gain AlGaN Solar-Blind Avalanche Photodiodes,” IEEE Electron Device Lett. 35(3), 372–374 (2014). [CrossRef]
19. J. Kim, M. H. Ji, T. Detchprohm, J. H. Ryou, R. D. Dupuis, A. K. Sood, and N. K. Dhar, “AlxGa1-xN Ultraviolet Avalanche Photodiodes With Avalanche Gain Greater Than 10(5),” IEEE Photonics Technol. Lett. 27(6), 642–645 (2015). [CrossRef]
20. L. Hahn, F. Fuchs, L. Kirste, R. Driad, F. Rutz, T. Passow, K. Kohler, R. Rehm, and O. Ambacher, “Avalanche multiplication in AlGaN-based heterostructures for the ultraviolet spectral range,” Appl. Phys. Lett. 112(15), 151102 (2018). [CrossRef]
21. L. Watschke, T. Passow, F. Fuchs, L. Kirste, R. Driad, F. Rutz, S. Leone, R. Rehm, and O. Ambacher, “AlGaN avalanche Schottky diodes with high Al-content,” Jpn. J. Appl. Phys. 58(SC), SCCC11 (2019). [CrossRef]
22. K. K. Xu, L. W. Snyman, J. L. Polleux, K. A. Ogudo, C. Viana, Q. Yu, and G. P. Li, “Silicon LEDs toward high frequency on-chip link,” Optik 127(17), 7002–7020 (2016). [CrossRef]
23. H. Mackel and R. Ludemann, “Detailed study of the composition of hydrogenated SiNx layers for high-quality silicon surface passivation,” J. Appl. Phys. 92(5), 2602–2609 (2002). [CrossRef]
24. E. Bellotti, F. Bertazzi, S. Shishehchi, M. Matsubara, and M. Goano, “Theory of Carriers Transport in III-Nitride Materials: State of the Art and Future Outlook,” IEEE Trans. Electron Devices 60(10), 3204–3215 (2013). [CrossRef]
25. K. K. Xu, “Silicon MOS Optoelectronic Micro-Nano Structure Based on Reverse-Biased PN Junction,” Phys. Status Solidi A 216(7), 1800868 (2019). [CrossRef]
26. E. Bellotti, F. Bertazzi, and M. Goano, “Alloy scattering in AlGaN and InGaN: A numerical study,” J. Appl. Phys. 101(12), 123706 (2007). [CrossRef]
27. E. Bellotti and F. Bertazzi, “A numerical study of carrier impact ionization in AlxGa1-xN,” J. Appl. Phys. 111(10), 103711 (2012). [CrossRef]
28. J. Y. Zheng, L. Wang, X. Z. Wu, Z. B. Hao, C. Z. Sun, B. Xiong, Y. Luo, Y. J. Han, J. Wang, H. T. Li, J. Brault, S. Matta, M. Al Khalfioui, J. C. Yan, T. B. Wei, Y. Zhang, and J. X. Wang, “A PMT-like high gain avalanche photodiode based on GaN/AlN periodically stacked structure,” Appl. Phys. Lett. 109(24), 241105 (2016). [CrossRef]
