Open Access
Add to BibSonomy
Abstract
While traditional tunnel junction (TJ) light-emitting diodes (LEDs) can enhance current diffusion and increase hole injection efficiency, their reliance on highly doped AlGaN layers to improve hole tunneling efficiency results in a higher conduction voltage, adversely impacting LED device performance. This paper proposes a non-heavy doped pnp-AlGaN TJ deep ultraviolet (DUV) LED with a low conduction voltage. By inserting the TJ near the active region, between the electron blocking layer and the hole supply layer, the need for heavily doped AlGaN is circumvented. Furthermore, the LED leverages the polarization charge in the pnp-AlGaN TJ layer to decrease the electric field strength, enhancing hole tunneling effects and reducing conduction voltage. The non-heavy doped pnp-AlGaN TJ LED effectively enhances carrier concentration in the quantum well, achieving a more uniform distribution of electrons and holes, thus improving radiative recombination efficiency. Consequently, at an injection current of 120 A/cm2, compared to the traditional structure LED (without TJ), the proposed LED exhibits a 190.7% increase in optical power, a 142.8% increase in maximum internal quantum efficiency (IQE) to 0.85, and a modest efficiency droop of only 5.8%, with a conduction voltage of just 4.1V. These findings offer valuable insights to address the challenges of high heavy doped TJ and elevated conduction voltage in high-performance TJ DUV LEDs.
© 2024 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
AlGaN-based deep ultraviolet (DUV) light-emitting diodes (LEDs) offer distinct advantages, including compact size, low power consumption, tunable wavelength, environmental friendliness, and extended lifespan. Their applications span various domains, encompassing disinfection, sterilization, biomedicine, and non-line-of-sight communication [1]. Presently, the maximum external quantum efficiency (EQE) for AlGaN-based DUV LEDs reaches 20% at an emission wavelength of 275 nm, while it achieves a maximum EQE of 10% at 265 nm, with EQE dropping below 1% for wavelengths under 250 nm [2,3]. The primary contributors to this lower EQE are the diminished internal quantum efficiency (IQE) and suboptimal light extraction efficiency (LEE). IQE is influenced by material quality, carrier injection efficiency, and radiation recombination, whereas LEE is determined by factors such as total internal reflection, light absorption, and optical polarization [4]. A significant challenge arises from the high activation energy of magnesium (Mg) in the p-type region, resulting in subpar hole injection efficiency, an asymmetric electron-hole flow, and subsequent non-radiative recombination outside the quantum well, leading to IQE droop [5]. Consequently, improving hole injection efficiency becomes crucial for enhancing the EQE of DUV LEDs. However, achieving efficient p-type doping in AlN-rich environments proves challenging, compounded by the significantly lower hole mobility in AlGaN compared to electron mobility, resulting in diminished hole injection efficiency and heightened electron leakage rates. Researchers have undertaken numerous strategies to address these challenges, including leveraging polarization effects to improve the electron blocking layer (EBL) and p-type hole supply layer (HSL) [6], utilizing three-dimensional hole gas [7], incorporating highly reflective photonic crystals on the p-AlGaN contact layer [8], optimizing the mesa area size to adjust current distribution [9], improving the conductivity of the contact layer with the p-electrode [10], and incorporating tunnel junctions (TJs) or polarized tunnel junctions (PTJs).
The development of the GaN TJ was initially reported in 2001. Despite its capability to improve current diffusion and enhance hole injection efficiency, the GaN TJ exhibits high series resistance and conduction voltage due to inefficient tunneling [11,12]. Subsequent studies by Grundmann et al. and Simon et al. explored a novel generation of nitride-PTJs, leveraging strong polarization dipoles generated by the GaN and AlN heterointerfaces, thereby improving the tunneling efficiency of GaN tunnel junctions [13,14]. A significant enhancement in tunneling efficiency was achieved by incorporating an ultra-thin AlN layer into the GaN-PN tunnel junction structure, reducing the depletion width and allowing for feasible interband tunneling. However, the tunneling probability remained constrained by the large band gap of AlN. Researchers addressed this limitation by replacing the AlN barrier with an ultra-thin AlGaN layer [15–17]. Both the TJ ($p^+-n^+$) structure and the PTJ ($p^+-i-n^+$) structure impose high doping requirements and exhibit elevated conduction voltage, presenting challenges for the deep ultraviolet region of the wide band gap.
The principle underlying PTJs involves introducing a robust spontaneous polarization and piezoelectric polarization electric field ($E_p$), influencing the coupling electric field ($E_c$) through contributions from both the built-in electric field ($E_b$) and $E_p$. Building upon prior research, this work extends the concept of PTJs by introducing a thin, non-heavily doped pnp-AlGaN TJ between the EBL and HSL to mitigate the challenges associated with heavy doping in AlN-rich AlGaN. The polarization-induced electric field $E_c$ at the p-AlGaN/n-AlGaN/p-AlGaN interface enhances hole injection efficiency. The results demonstrate substantial improvements in IQE and optical power, coupled with a reduction in conduction voltage.
2. Device architectures and physical parameters
In Fig. 1(a), Sample A represents the traditional structure (without a tunnel junction), as developed by Zhang et al. [18]. The LED is grown on a sapphire substrate, followed by an n-$Al_{0.6}Ga_{0.4}N$ electron supply layer (ESL) with a thickness of 2 ${\mathrm{\mu} }$m. The Si doping concentration of the n-AlGaN ESL is set to $1{\times }10^{18} cm^{-3}$. Subsequently, a five-period multi-quantum wells (MQWs) structure is formed, comprising five $Al_{0.45}Ga_{0.55}N$ quantum wells (2 nm thick) and six $Al_{0.56}Ga_{0.44}N$ quantum barriers (12 nm thick), ensuring a peak emission wavelength of 275 nm. Following this, a p-$Al_{0.9}Ga_{0.4}N$ electron blocking layer (EBL) with a thickness of 10 nm, a p-$Al_{0.6}Ga_{0.4}N$ hole supply layer (HSL) with a thickness of 50 nm, and a p-GaN layer with a thickness of 40 nm are fabricated. The Mg doping concentration is $2{\times }10^{18} cm^{-3}$ for EBL and $1.2{\times }10^{18} cm^{-3}$ for HSL and p-GaN. In Fig. 1(b), Sample B introduces an n-$Al_{0.5}Ga{0.5}N$ TJ layer (3 nm thick) between EBL and HSL. Similarly, Sample C inserts an n-$Al_{0.5}Ga_{0.5}N$ TJ layer (3 nm thick) between HSL and p-GaN (Fig. 1(c)). For Sample D, an n-$Al_{0.5}Ga_{0.5}N$ TJ layer (3 nm thick) is inserted between both EBL and HSL and between HSL and p-GaN (Fig. 1(d)). Sample E incorporates a 3nm p-$Al_{0.5}Ga_{0.5}N$/3nm n-$Al_{0.5}Ga_{0.5}N$/3nm p-$Al_{0.5}Ga_{0.5}N$ pnp-TJ layer between EBL and HSL (Fig. 1(e)). Sample F includes an n-$Al_{0.5}Ga_{0.5}N$ TJ layer (3 nm thick) on the p-GaN layer as a contact layer (Fig. 1(f)). The Si doping concentration of n-$Al_{0.5}Ga_{0.5}N$ is $3{\times }10^{17} cm^{-3}$, and the Mg doping concentration of p-$Al_{0.5}Ga_{0.5}N$ is $3{\times }10^{17} cm^{-3}$ to minimize resistance to hole transport.
Fig. 1. Schematic illustration of AlGaN DUV LEDs for (a) Sample A without TJ, (b) Sample B with n-AlGaN TJ between EBL and HSL, (c) Sample C with n-AlGaN TJ between HSL and p-GaN, (d) Sample D with n-AlGaN TJ between EBL and HSL and between HSL and p-GaN, (e) Sample E with pnp-AlGaN TJ between EBL and HSL, and (f) Sample F with n-AlGaN TJ on the p-GaN.
The Advanced Physical Model of Semiconductor Devices (APSYS) tool is employed to numerically solve the Poisson equation, drift-diffusion model, and photon rate equation [19]. The 6$\times$6 k.p model is utilized to calculate energy band alignment [20]. AlN and GaN bandgaps are 6.2 eV and 3.4 eV, respectively. The activation energy of Mg-doped p-AlGaN ranges between 170 meV and 670 meV [21]. Band offset ratio and bowing parameter of AlGaN MQWs are assumed to be 0.3/0.7 and 0.94 eV, respectively [22]. Recombination parameters such as Schockley-Read-Hall (SRH) lifetime, Augur recombination coefficient, and radiative recombination coefficient are set at 100 ns, $1.0{\times }10^{-46} m^3/s$, and $0.5{\times }10^{-16}m^3/s$, respectively [23,24]. For LED chips measuring 300 ${\mathrm{\mu} }$m $\times$ 300 ${\mathrm{\mu} }$m, back loss is considered as material absorption loss at 2400 $m^{-1}$, representing the ideal loss for GaN devices. The built-in polarization charge density due to piezoelectric polarization and spontaneous polarization is assumed to be 50% of the theoretical value, calculated following the method proposed by Fiorentini et al. [25]. Trap-assisted tunneling is also considered in the model. All simulations are conducted at room temperature, and the LEE for the EQE calculation section is assumed to be 9%, as per the standard value reported in the literature [26].
3. Results and discussions
Figure 2 presents the simulation data for the DUV LED of Sample A alongside the experimental data cultivated by Zhang et al. Utilizing the material parameters outlined in Section 2, the optical power (OP) and EQE of Sample A align almost with the experimental findings of Zhang et al. under a current density of 120 $A/cm^2$. To investigate the impact of the position of the non-heavily doped n-AlGaN TJ layer on the photoelectric characteristics of the DUV LED, the OP, IQE, and current-voltage (I-V) characteristic curves for six Samples are computed, as depicted in Fig. 3. Furthermore, a comprehensive comparison of OP, IQE, droop efficiency, conduction voltage, and forward operating voltage for Samples A, B, C, D, E, and F is presented at 120 $A/cm^2$ in Fig. 4. While the OP and IQE of Samples C, D, and F slightly surpass those of Sample A, the conduction voltage and operating voltage experience a substantial increase, as evident in Figs. 3 and 4. This elevation can be attributed to the insertion of a non-heavily doped n-AlGaN TJ layer near the p-GaN, introducing additional series resistance. In thermionic emission, the insertion of the TJ layer results in a heightened Schottky barrier, leading to increased resistance [27]. Consistent with prior research, the introduction of an n-AlGaN TJ layer near the p-electrode mitigates the current crowding effect, enhances lateral current spreading, and promotes the uniformity of LED top luminescence due to its relatively high resistance contribution [10,11,15,28]. When the LED operates under forward bias voltage, the top TJ functions at reverse bias voltage, resembling a resistance that determines the forward voltage of the LED. Consequently, a higher bias voltage is required to achieve optimal tunneling efficiency. Researchers have historically addressed this by heavily doping the $n^+$ AlGaN TJ layer to limit the operating voltage [10,11]. However, when employing the non-heavily doped n-AlGaN TJ layer (Samples C, D, and F), the operating voltage experiences a significant increase, accompanied by low tunneling efficiency. Hence, inserting the non-heavily doped n-AlGaN TJ layer near the p-GaN is deemed unsuitable. As illustrated in Figs. 3 and 4, Sample E exhibits the most promising photoelectric properties among the six Samples. The OP of Sample E surpasses that of Sample A by 190.7% and is 44.3% higher than Sample B. Similarly, the maximum IQE of Sample E exceeds that of Sample A by 142.8% and outperforms Sample B by 45.17%. At an injection current of 120 $A/cm^2$, Sample E achieves an OP of 31.6 $W/cm^2$ and an IQE of 0.85, with only a marginal 5.8% efficiency droop. Notably, the conduction voltage of Sample E is merely 4.1 V, signifying that the TJ layer inserted between the EBL and HSL does not augment series resistance, presenting a considerable advantage for enhancing DUV LED performance. The computational outcomes affirm that utilizing the non-heavily doped pnp-AlGaN TJ layer significantly enhances tunneling efficiency without concomitant increases in conduction voltage.
Fig. 2. (a) Optical power (P-I) curves and (b) external quantum efficiency of simulation data (Sample A) and experimental data of LEDs grown by Zhang et al.
Fig. 3. Calculated (a) optical power, (b) internal quantum efficiency, and (c) current-voltage (I-V) characteristics for Samples A, B, C, D, E, and F at 120 $A/cm^2$.
Fig. 4. Calculated optical power (OP), internal quantum efficiency (IQE), droop efficiency, conduction voltage, and forward operating voltage for Samples A, B, C, D, E, and F at 120 $A/cm^2$.
The tunneling efficiency of Samples C and D exhibits low values, resulting in elevated conduction and operating voltages, significantly detrimental to LED performance. Consequently, we focus our in-depth physical mechanisms analysis solely on Samples A, B, and E. Among the four TJ Samples, only Samples B and E witness a marginal reduction in conduction voltages, attributed to the unique position of their TJ layers, situated in proximity to the quantum well. However, Sample E surpasses Sample B in both OP and maximum IQE by a substantial margin (Figs. 3 and 4). The key distinction lies in the insertion of a p-AlGaN layer in both upper and lower layers of the n-AlGaN TJ in Sample E (Fig. 1). To further explore the utility of the non-heavily doped pnp-AlGaN TJ layer (Sample E), enhancing tunneling efficiency, resulting in a remarkable increase in LED OP and IQE with a lower conduction voltage. As depicted in Fig. 5, the electric field strengths of Samples A, B, and E within the TJ layer are calculated. The non-heavy doped pnp-AlGaN TJ layer, proposed here, extends the concept of introducing a polarization electric field in PTJs. The pnp-AlGaN TJ layer, when inserted, generates a polarization electric field ($E_P$) between the upper and lower layers of the n-AlGaN TJ, influencing the coupling electric field ($E_C$). The coupling electric field in the TJ layer can be expressed as [15]:
Fig. 5. Calculated electric field of Sample A without TJ, Sample B with non-heavy doped n-AlGaN TJ, and Sample E with non-heavy doped pnp-AlGaN TJ at 120 $A/cm^2$
In the equation, $e$ represents the unit charge, $\varepsilon _{r}$ and $\varepsilon _{0}$ represent the average relative dielectric constant and absolute dielectric constant, respectively. $N_{dop}$ represents the ion doping concentration in the space charge region, $L_{d}$ represents the depletion layer width, and $\sigma _p$ represents the charge density of the polarization-induced sheet. The sign "$\pm$" indicates the alignment of the polarization-induced electric field with $E_{b}$, "$+$" when identical, and "$-$" when different. When the LED operates under forward bias voltage, the PTJs function at reverse bias voltage, decreasing both resistance and depletion layer width ($L_{d}$). Although a smaller $L_{d}$ will reduce the electric field $E_{c}$, potentially affecting threshold voltage, current, and tunneling efficiency, the pnp-AlGaN TJ layer proposed here introduces a polarized electric field ($E_{p}$) in the TJ layer to manipulate the coupling electric field ($E_{c}$). As shown in Fig. 5, the electric field strength ($E_{c}$) of Sample E significantly decreases that of Samples A and B. This result is attributable to the direct insertion of the pnp-AlGaN TJ layer between EBL and HSL, manipulating the charge density ($\sigma _p$) of the polarization-induced sheet in Eq. (1). The lowered $E_{c}$ in the TJ layer promotes hole tunneling and positively impacts light absorption avoidance, proving highly advantageous for improving IQE and OP of the LED.
The tunneling efficiency of carriers is influenced by both the electric field strength within the TJ layer and the energy band. Expressing hole tunneling efficiency involves the following equation [29]:
where $m_{t}^{*}$ is the effective mass of the carrier, $E_{g}$ is the energy band with width $x$, and $h$ is the Planck constant. The depletion layer width $L_{d}$ in Eqs. (1) and (2) can be expressed as: where $N_A$ and $N_D$ represent the acceptor and donor concentrations, $\varepsilon$ is the dielectric constant in free space, and $q$ is the electron charge. Notably, $L_d$ decreases with a reduction in doping level, providing compelling evidence for the utilization of a non-heavily doped pnp-AlGaN TJ layer. Furthermore, equation (refequation) illustrates that $L_{d}$ is not solely influenced by the doping concentration but is also impacted by the energy band $E_{g}$ also affects the hole tunneling probability $P_{b}$. As shown in Fig. 6, the energy bands of Samples A, B, and E near the TJ layer are computed. The insertion of the pnp-TJ between the EBL and HSL in Sample E results in a rearrangement of its energy band. Compared to Samples A and B, the entire energy band of Sample E is elevated, contributing to an increased tunneling probability $P_{b}$ for holes. The computational results in Fig. 6 reveal that the effective barrier height ($\theta _{e}$) of the conduction band in Sample E is heightened to 430.26 meV, reinforcing the confinement of electrons. Simultaneously, the effective barrier height ($\theta _{c}$) of the valence band is reduced to 104.99 meV, facilitating the hole transport process. Although Sample B also experiences a partial increase in $\theta _{e}$ and a decrease in $\theta _{c}$, the magnitude of increase is comparatively smaller than that of Sample E. This aligns with the findings in Fig. 5, where Sample B, incorporating only n-AlGaN TJ, exhibits a lesser electric field strength in this layer compared to the pnp-AlGaN TJ layer in Sample E. Consequently, Sample E demonstrates substantial potential for improving hole injection efficiency and minimizing electron leakage.Fig. 6. Energy bands (conduction band, valence band) of (a) Sample A without TJ, (b) Sample B with non-heavy doped n-AlGaN TJ, and (c) Sample E with non-heavy doped pnp-AlGaN TJ. $E_c$, $E_fe$, $E_fh$, and $E_v$ denote the conduction band, electron quasi-Fermi level, hole quasi-Fermi level, and valence band, respectively.
From a physical perspective, Sample E, featuring a non-heavy doped pnp-AlGaN TJ layer, exhibits a smaller coupling electric field, fostering enhanced manipulation of the energy band for efficient electron and hole transport. It is crucial to note, however, that the total thickness of the pnp-AlGaN TJ layer in Sample E is 9 nm, while the thickness of the n-AlGaN TJ layer in Sample B is merely 3 nm. Although a thicker TJ layer can indeed contributes more to the electric field and energy band within the layer, elevating the effective barrier height for electrons, reducing the effective barrier height for the valence band, improving hole tunneling efficiency, and mitigating electron leakage, excessively thick tunnels, in theory, can severely impede in-band tunneling of holes, leading to substantial electron leakage [30]. As illustrated in Figs. 7 and 8, we calculate the electron and hole concentrations in the quantum well for the three Samples, respectively. The electron concentration in the five QWs of Sample E is 32.5% higher than that of Sample A and 29.8% higher than that of Sample B (Fig. 7(a)). To elucidate the reason for the increased electron concentration, we calculate the electron concentration near the TJ layer for the three Samples, as shown in Fig. 7(b). It is evident that the pnp-TJ layer in Sample E effectively hinders most electrons from leaking into the p-type region, a result of the distinctive position of the pnp-TJ layer. The insertion of pnp-TJ between EBL and HSL proves advantageous in preventing electron leakage, a distinct advantage over TJ design at the top of the p-type layer. The pnp-TJ layer’s capability to block electron leakage stems from the increased effective barrier height of the conduction band (Fig. 6(c)) and the decreased electric field intensity in the TJ layer (Fig. 5). The hole concentration in the five quantum wells of Sample E is 155.5% higher than that of Sample A and 123.5% higher than that of Sample B. To probe the reason for the augmented hole concentration, we calculate the hole current densities for the three Samples, as shown in Fig. 8(b). The hole current density of Sample E is the highest, indicative of superior hole injection efficiency. This enhancement is attributed to the increased hole tunneling efficiency $P_{b}$ and the reduced effective barrier height of the valence band (Fig. 6(c)), facilitating in-band tunneling of holes and improving hole injection efficiency. From the calculated electron and hole concentrations in Figs. 7 and 8, it is evident that, despite the pnp-AlGaN TJ layer’s total thickness being 9 nm, it effectively improves hole transport efficiency and reduces electron leakage. This improvement is attributed to the fact that the thickness of n-AlGaN in the pnp-AlGaN TJ is only 3 nm, equivalent to the TJ thickness in Sample B. The insertion of p-AlGaN layers above and below the n-AlGaN layer further enhances the tunneling efficiency of the TJ layer, thereby ameliorating the uneven distribution of carrier concentration in the quantum well.
Fig. 7. Electron concentration (a) in the MQWs and (b) in the p-type region of Sample A without TJ, Sample B with non-heavy doped n-AlGaN TJ, and Sample E with non-heavy doped pnp-AlGaN TJ.
Fig. 8. (a) Hole concentration and (b) hole current density of Sample A without TJ, Sample B with non-heavy doped n-AlGaN TJ, and Sample E with non-heavy doped pnp-AlGaN TJ.
The elevation in electron and hole concentrations, particularly in hole concentration, contributes to a more even distribution of carriers within the quantum well, yielding substantial advantages in terms of improving radiative recombination efficiency. Illustrated in Fig. 9, the radiative recombination rate of Sample E surpasses that of Sample A by 147.7% and that of Sample B by 103.5%. This improvement is primarily attributed to the incorporation of a non-heavy doped pnp-AlGaN Tunnel Junction (TJ) in Sample E, leading to enhanced hole tunneling efficiency Pb and decreased the electric field strength Ec in the TJ layer (Fig. 5). The non-heavy doped pnp-AlGaN TJ layer plays a pivotal role in manipulating the entire energy band, increasing the effective barrier height of the conduction band, and decreasing the effective barrier height of the valence band (Fig. 6). Consequently, this alteration elevates the concentrations of electrons and holes within the quantum well (Figs. 7 and 8). The resulting more uniform carrier distribution within the quantum well enhances transverse current expansion, leading to an increased radiative recombination rate in the quantum well and a reduction in non-radiative recombination in the p-type region. This cascading effect amplifies the IQE and OP of Sample E (Figs. 3 and 4).
Fig. 9. Radiative recombination rate comparison among Sample A without TJ, Sample B with non-heavy doped n-AlGaN TJ, and Sample E with non-heavy doped pnp-AlGaN TJ. The horizontal positions of the radiative recombination rate for the three Samples have been appropriately adjusted for convenient comparison.
4. Conclusion
This paper proposes a novel approach for a deep ultraviolet light-emitting diode (DUV LED) with a non-heavy doped pnp-AlGaN Tunnel Junction (TJ) and an emission wavelength of 275 nm. The insertion of the pnp-AlGaN TJ layer between the Electron Blocking Layer (EBL) and Hole Supply Layer (HSL) addresses the challenge of heavy doping encountered in traditional TJ layers. In comparison to devices employing n-AlGaN TJ, the DUV LEDs with a non-heavy doped pnp-AlGaN TJ leverage the polarized electric field within the TJ layer to augment the energy band and decrease electric field intensity, thereby enhancing hole tunneling probability. Crucially, the non-heavy doped pnp-AlGaN TJ layer manipulates the entire energy band of the device, elevating the effective barrier height of the conduction band while diminishing the effective barrier height of the valence band. This dual effect not only improves carrier concentration within the quantum well but also ensures a more uniform distribution of carriers. Consequently, the radiative recombination efficiency of carriers in the quantum well experiences a remarkable increase of 147.7%. Notably, the distinctive placement of the non-heavy doped pnp-AlGaN TJ layer circumvents the challenge of high conduction voltage induced by elevated resistance in conventional TJ structures. The resulting conduction voltage is approximately 4.1V. As a result, the internal quantum efficiency (IQE) and optical power (OP) of the proposed non-heavy doped pnp-AlGaN TJ DUV LED are heightened to 0.85 and 31.6 $W/cm^2$, respectively, with a minimal efficiency droop of only 5.8%. The growth of sharp pnp junctions by using MOCVD is challenging due to the diffusion of the Mg dopants and the memory effect. Nevertheless, such issues can be solved by inserting the ultrathin AlN layer or growing at lower temperatures. This study contributes novel insights towards overcoming issues related to high doping and conduction voltage in TJ-based DUV LEDs.
Funding
National Natural Science Foundation of China (62174148); National Key Research and Development Program of China (2016YFE0118400, 2022YFE0112000); Key Program for International Joint Research of Henan Province (231111520300); Ningbo Major Project of ‘Science, Technology and Innovation 2025’ (2019B10129); Zhengzhou 1125 Innovation Project (ZZ2018-45).
Acknowledgments
We thank Muhammad Nawaz Sharif and Yongjie Zhou for helpful discussions.
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. L. Dabing, J. Ke J, and S. Xiaojuan, “AlGaN photonics: recent advances in materials and ultraviolet devices,” Adv. Opt. Photonics 10(1), 43 (2018). [CrossRef]
2. T. Takano, T. Mino T, and J. Sakai, “Deep-ultraviolet light-emitting diodes with external quantum efficiency higher than 20% at 275 nm achieved by improving light-extraction efficiency,” Appl. Phys. Express 10(3), 031002 (2017). [CrossRef]
3. S. 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]
4. K. Kojima, Y. Nagasawa, A. Hirano, et al., “Carrier localization structure combined with current micropaths in AlGaN quantum wells grown on an AlN template with macrosteps,” Appl. Phys. Lett. 114(1), 011102 (2019). [CrossRef]
5. S. Xiao, H. Yu, and M. Memon, “In-depth investigation of deep ultraviolet MicroLED geometry for enhanced performance,” IEEE Electron Device Lett. 44(9), 1520–1523 (2023). [CrossRef]
6. L. Khan, N. Maeda, J. Yun, et al., “Achieving 9.6% efficiency in 304 nm p-AlGaN UVB LED via increasing the holes injection and light reflectance,” Sci. Rep. 12(1), 2591 (2022). [CrossRef]
7. Z. Zhang, S. Tan, Z. Kyaw, et al., “p-doping-free InGaN/GaN light emitting diode driven by three-dimensional hole gas,” Appl. Phys. Lett. 103(26), 263501 (2013). [CrossRef]
8. Y. Kashima, N. Maeda, E. Matsuura, et al., “High external quantum efficiency (10%) AlGaN-based deep-ultraviolet light-emitting diodes achieved by using highly reflective photonic crystal on p-AlGaN contact layer,” Appl. Phys. Express 11(1), 012101 (2018). [CrossRef]
9. R. Horng, K. Chen, C. Tien, et al., “Effects of mesa size on current spreading and light extraction of GaN-based LED,” J. Disp. Technol. 11(12), 1010–1013 (2015). [CrossRef]
10. Z. Zhang, S. Tan, W. Liu, et al., “Improved InGaN/GaN light-emitting diodes with a p-GaN/n-GaN/p-GaN/n-GaN/p-GaN current spreading layer,” Opt. Express 21(4), 4958–4969 (2013). [CrossRef]
11. S. Jeon, Y. Song, H. Jang, et al., “Lateral current spreading in GaN-based light-emitting diodes utilizing tunnel contact junctions,” Appl. Phys. Lett. 78(21), 3265–3267 (2001). [CrossRef]
12. T. Takeuchi T, G. Hasnain G, and S. Corzine S, “GaN-based light emitting diodes with tunnel junctions,” Jpn. J. Appl. Phys. 40(8B), L861 (2001). [CrossRef]
13. M. Grundmann, “Polarization-induced tunnel junctions in III-nitrides for optoelectronic applications,” University of California (2007).
14. J. Simon J, Z. Zhang, and K. Goodman, “Polarization-induced Zener tunnel junctions in wide-bandgap heterostructures,” Phys. Rev. Lett. 103(2), 026801 (2009). [CrossRef]
15. Y. Wang, Z. Zhang, and L. Guo, “Calculating the effect of AlGaN dielectric layers in a polarization tunnel junction on the performance of AlGaN-based deep-ultraviolet light-emitting diodes,” Nanomaterials 11(12), 3328 (2021). [CrossRef]
16. R. Wang, H. Yu, S. Xiao, et al., “Efficient AlGaN-based deep-ultraviolet LED with N-side located tunnel junction,” IEEE Photonics Technol. Lett. 35(24), 1391–1394 (2023). [CrossRef]
17. S. Xiao, H. Yu, H. Jia, et al., “Performance evaluation of tunnel junction-based N-polar AlGaN deep-ultraviolet light-emitting diodes,” Opt. Lett. 47(16), 4187–4190 (2022). [CrossRef]
18. Z. Zhang, J. Kou, and S. Chen, “Increasing the hole energy by grading the alloy composition of the p-type electron blocking layer for very high-performance deep ultraviolet light-emitting diodes,” Photonics Res. 7(4), B1–B6 (2019). [CrossRef]
19. http://www.crosslight.com/ (2018).
20. T. Chuang and C. S. Chang, “K p method for strained Wurtzite semiconductors,” Phys. Rev. B 54(4), 2491–2504 (1996). [CrossRef]
21. S. Lee, J. Son, and T. Sakong, “Investigation of optical and electrical properties of Mg-doped p-InxGa1-xN, p-GaN and p-AlyGa1-yN grown by MOCVD,” J. Cryst. Growth 272(1-4), 455–459 (2004). [CrossRef]
22. C. Coughlan, S. Schulz, M. A. Caro, et al., “Band gap bowing and optical polarization switching in AlGaN alloys,” Phys. Status Solidi B 252(5), 879–884 (2015). [CrossRef]
23. Z. Xing, Y. Wang, F. Wang, et al., “Improvement of the optoelectronic characteristics in deep-ultraviolet laser diodes with tapered p-cladding layer and triangular electron blocking layer,” Appl. Phys. B 128(11), 197 (2022). [CrossRef]
24. Z. Xing, Y. Wang, F. Wang, et al., “Enhanced performance in deep-ultraviolet laser diodes with undoped BGaN electron blocking layer,” Opt. Express 30(20), 36446 (2022). [CrossRef]
25. V. Fiorentini, F. Bernardini, and O. Ambacher, “Evidence for nonlinear macroscopic polarization in III–V nitride alloy heterostructures,” Appl. Phys. Lett. 80(7), 1204–1206 (2002). [CrossRef]
26. H. Ryu, I. Choi, H. Choi, et al., “Investigation of light extraction efficiency in AlGaN deep-ultraviolet light-emitting diodes,” Appl. Phys. Express 6(6), 062101 (2013). [CrossRef]
27. L. Sulmoni, F. Mehnke, and A. Mogilatenko, “Electrical properties and microstructure formation of V/Al-based n-contacts on high Al mole fraction n-AlGaN layers,” Photonics Res. 8(8), 1381 (2020). [CrossRef]
28. Z. Zhang, S. Tan, Z. Kyaw, et al., “Demir, H.V. InGaN/GaN light-emitting diode with a polarization tunnel junction,” Appl. Phys. Lett. 102(19), 193508 (2013). [CrossRef]
29. V. Arcara, “Tunnel junctions in nitride heterostructures for optoelectronic applications,” Doctoral dissertation, Université Côte d’Azur and CNRS France, (2020).
30. A. Pandey, W. Shin, J. Gim, et al., “High-efficiency AlGaN/GaN/AlGaN tunnel junction ultraviolet light-emitting diodes,” Photonics Res. 8(3), 331–337 (2020). [CrossRef]
