Expand this Topic clickable element to expand a topic
Skip to content
Optica Publishing Group

Toward high-bandwidth yellow-green micro-LEDs utilizing nanoporous distributed Bragg reflectors for visible light communication

Open Access Open Access

Abstract

In this study, high 3dB bandwidth yellow-green InGaN/GaN micro-LEDs grown on polar c-plane GaN substrates are realized by using nanoporous distributed Bragg reflectors, which can increase light extraction efficiency and serve as strain-relaxed buffers to mitigate the quantum-confined Stark effect, resulting in improved external quantum efficiency. Moreover, atomic layer deposition technology is introduced for surface defect passivation, thereby reducing the leakage current. As a result, the device exhibits the highest 3dB bandwidth up to 442 MHz and a data transmission rate of 800 Mbit/s at a current density of 2.5kA/cm2 with on–off keying modulation, and holds great promise for future high-speed visible light communication applications.

© 2022 Chinese Laser Press

1. INTRODUCTION

Nowadays, optical wireless communication (OWC) is regarded as a potential paradigm in communication areas including light fidelity (Li-Fi) [1,2], underwater optical communication (UWOC) [3], smart displaying signboards [4], and indoor positioning [5] due to its low power consumption, high spectrum bandwidth, and being license free [1,2,6]. Light-emitting diodes (LEDs) are a viable platform for visible light communication (VLC) since LEDs are widely utilized in our daily lives. Red-green-blue LEDs (RGB-LEDs) or phosphor-converted LEDs (pc-LEDs) are commonly used to generate white light, which can be further employed for VLC applications. However, because of the long carrier lifetime of phosphors, the bandwidth of pc-LEDs is limited up to a few MHz [6]. In addition, many researchers have employed quantum dots (QDs) to develop white light systems, which can be used in VLC applications [79]. Although the bandwidth of QD-micro-LED (μLED) systems has had a breakthrough reaching 1.3 GHz at 566 nm recently [9], there are still some challenges with using phosphor or QDs. On the other hand, introducing a yellow-green μLED subpixel into an RGB-μLED seems to be a promising way for improving the VLC system [10].

The 3dB bandwidth of the μLED can be described by Eq. (1) [11]:

f3dB=32π(1τr+1τnr+1τRC),
where τr, τnr, and τRC are the carrier lifetimes of radiative recombination and non-radiative recombination and resonant cavity (RC) time constant, respectively. In general, the modulation bandwidths of LEDs depend on the recombination lifetimes and RC time constant. However, for LEDs with sizes of 100μm×100μm and less, the radiative lifetime dominates the modulation bandwidth because the geometric capacitance is small enough to prevent the RC time constant from becoming the dominant factor [11]. Therefore, in this paper, the bandwidth is proportional to the carrier density. Nonetheless, the current density is limited due to the self-heating effect [12]. A smaller μLED can sustain higher current density due to lower junction temperature and less current crowding effect, resulting in a faster radiative lifetime. Hence, the high radiative recombination rate in μLEDs is very crucial to implement a high-bandwidth VLC system. The ABC rate equation model is usually used to clarify the correlation of radiative and non-radiative recombination lifetimes. The ABC rate equation model is described as
I=e·a·d(An+Bn2+Cn3),
where n is the carrier density in the active region, A, B, and C are the Shockley–Read–Hall (SRH), radiative, and Auger recombination coefficients, respectively, and e, a, and d represent the electronic charge, active region area, and thickness, respectively. The carrier lifetime can be determined by the recombination rate to the carrier density given as follows:
τ1=A+2Bn+3Cn2.

Thus, when the injection current increases, the number of injected carriers (i.e., electrons and holes) in the active region will increase, resulting in higher carrier density and short differential carrier lifetime. The bandwidth will increase with higher current density [11] and gradually saturate due to the self-heating effect [12]. The self-heating effect will increase the junction temperature and hinder the radiative recombination, resulting in efficiency roll-off phenomena [13]. Fortunately, a small sized μLED can effectively reduce self-heating at high current density. It has been proven that the operating range of current density can be greatly improved. According to Eq. (3), when a μLED operates at a high current density, the radiative coefficient and Auger coefficient will dominate the bandwidth.

Besides self-heating, the strong quantum-confined Stark effect (QCSE) in polar c-plane GaN-based μLEDs leads to reduced efficiency, wavelength shift, and low bandwidth [14,15]. The strain-induced polarization originates from the lattice mismatch within an InGaN/GaN quantum well (QW), resulting in a low carrier radiative recombination rate and low internal quantum efficiency (IQE) [16]. The polarization-related electric field leads to less overlap of the electron–hole wave function, which in turn reduces the carrier recombination, and hence the bandwidth of the μLED decreases. Furthermore, as the indium content in the InGaN QW increases, the stronger the QCSE becomes; this phenomenon is known as the “green gap,” limiting the development of green LEDs [17]. Since the QCSE in a polar c-plane GaN is more serious than in a semipolar one, semipolar μLEDs can reach higher bandwidths at lower current densities than μLEDs grown on a polar c-plane GaN as in a previous study [11]. As our previous study showed, semipolar μLEDs can improve electron–hole wave function overlap due to lower QCSE, leading to faster carrier recombination and hence increased μLED bandwidth. However, as the current density increases, due to the Coulomb screening effect screening the QCSE, the change of band bending in a c-plane LED is much more significant than that in a semipolar LED, leading to increased bandwidth at a faster rate with current density. As a result, at high current density, the bandwidth of the polar LED may eventually approach that of the semipolar LED [18]. In previous work, we utilized semipolar (202¯1) GaN to mitigate the polarization field, achieving 3dB bandwidth up to 756 MHz, and a 1.5 Gbit/s data rate under a current density of 2.0kA/cm2 [19]. We have also reported a semipolar blue μLED with a maximum bandwidth of 817 MHz and data transmission rate of 1.5 Gbit/s [8]. For InGaN-based yellow LEDs, a wall-plug efficiency of 565 nm yellow LEDs to 24.3% at 20A/cm2 and 33.7% at 3A/cm2 was achieved by improved material quality and reduced compressive strain of active layers [20].

Recent studies have shown that a high-bandwidth system can be realized by growing the μLED structure on a nanoporous distributed Bragg reflector (NP-DBR). The NP-DBR can improve light extraction efficiency (LEE) and serve as a strain-released layer for the InGaN active layer to enhance external quantum efficiency (EQE), leading to improved modulation bandwidth, light output power of the system of about 26%, and EQE of the system of about 43% [2127]. Compared with RC-LEDs and NP-GaN vertical-cavity surface-emitting lasers (VCSELs), the μLEDs fabricated in this paper employ only the bottom NP-DBR. Due to a lack of a top DBR, spontaneous emission light cannot be effectively confined within the active region [23,28,29].

Moreover, wavelengths around 530 nm to 570 nm for yellow-green LEDs have been vastly investigated [30,31]. In 2012, Zhang et al. fabricated an InGaN-based structure μ-pixel LED array capable of transmitting at a bit rate up to 250 Mbit/s with non-return-to-zero on–off keying (NRZ-OOK), achieving 3dB bandwidth up to 103.1 MHz under a current density of 92A/cm2 [32]. In 2018, Zhu et al. demonstrated a yellow-cyan RGBYC-LED with a yellow-green LED capable of transmitting at a bit rate up to 2.175 Gbit/s with 64-ary quadrature amplitude modulation (QAM) discrete multitone modulation (DMT), achieving 3dB bandwidth up to 300 MHz under a current density of 33.3A/cm2 [33]. Furthermore, in 2020, Haggar et al. demonstrated semipolar yellow-green GaN LEDs capable of achieving 3dB bandwidth up to 350 MHz under a current density of 92A/cm2 [34]. Nevertheless, compared to blue and red LEDs, the reported bandwidths for existing yellow-green LEDs are still low, and this needs to be improved for future VLC applications. In this study, we propose a high-bandwidth InGaN/GaN-based long-wavelength (547 nm to 574 nm) μLED grown on an NP n-GaN/GaN DBR structure. This layer structure is beneficial to improve the LEE and strain-relaxation behavior, compared to unporosified DBR, which showed poor optical performance from crack formation caused by strain accumulation [35,36], showing that fewer pairs are needed to have a high index contrast due to a high refractive index difference, low strain accumulation, better optical properties, and high crystalline quality [37,38]. Furthermore, the broad stop band, high reflectivity, superb lattice-matching, and high electric conductivity make electrical pumping possible [37], resulting in high-bandwidth yellow-green InGaN/GaN μLEDs, which hold great promise for future high-speed VLC applications.

2. EXPERIMENTS

Figure 1(a) is an image of the 30μm×8μLED array, and Fig. 1(b) shows the device structure of the yellow-green InGaN/GaN μLED fabricated on a 2 inch (5.08 cm) polar c-plane (0001) GaN epitaxial wafer. The (204¯3) patterned sapphire substrates (PSSs) with linear trenches, having groove depths and widths of 1 μm and 3 μm, respectively, with 6 μm stripe widths, were prepared by a photolithography process and an inductively coupled plasma reactive-ion etching (ICP-RIE) process. After the preparation of PSSs, a silicon oxide layer was deposited on the entire sapphire surfaces except for the selected c-plane sapphire sidewall via a self-aligned technique of angled evaporation. To produce a stacking-fault (SF)-free GaN, a 2 μm thick germanium-doped (Ge-doped) GaN layer was grown in the early stage of epitaxy. The coalescence of adjacent Ge-doped GaN stripes that can avoid the formation of SFs has been proven in previous work [39]. It helps to grow an 8 μm undoped GaN layer to form a continuous GaN film with a high-quality surface. After completing the SF-free GaN on a c-plane GaN epitaxy, the epitaxial wafer was planarized by chemical mechanical planarization (CMP) for subsequent μLED epitaxy structures.

 figure: Fig. 1.

Fig. 1. (a) Image of the 30μm×8μLED array, (b) schematic diagram of μLED structures, (c) scanning electron microscope image of the NP-DBR, and (d) simulated and experimental reflectance spectra of the DBR.

Download Full Size | PDF

The μLED epitaxial structure consisted of an undoped GaN layer, NP n-type GaN/GaN DBR, n-type GaN layer, followed by three pairs of InGaN/GaN multiple QWs (MQWs) as the active region and a 100 nm thick p-type GaN layer. The NP n-GaN layer shown in Fig. 1(c) was introduced into the μLED structure prepared by chemical etching through the following steps: (1) making indium contacts point on n-GaN, (2) immersing the n-GaN and platinum targets into the nitride acid solution with a concentration of 0.5 mol/L, (3) applying an 8 V forward voltage for about 20 min, and (4) rinsing the finished NP gallium nitride in deionized water for about 5 min, followed by blowing with N2 to dry. The simulated reflectance spectrum (performed with Synopsys RSoft) and measured reflectance spectrum of the NP-DBR are shown in Fig. 1(d); the refractive indices of the GaN and NP-GaN are approximately 2.39 and 1.7, respectively. It reveals that NP-DBR can greatly reflect light bouncing back at the wavelength region designed for improving the light extraction of yellow-green InGaN/GaN μLEDs.

In this study, we fabricated μLEDs with two mesa diameters of 30 μm and 50 μm. After the growth of the n-type GaN and three pairs of InGaN/GaN MQWs, a 100 nm thick indium tin oxide (ITO) layer was deposited by electron beam physical vapor deposition (EB-PVD) at 300°C on the p-GaN surface as a p-type contact layer. The different diameter of the emitter mesa was defined by a photolithography process. The mesa depth was 1 μm, which was etched by HCl solution for ITO and ICP-RIE for GaN sequentially. The ohmic contact for the p-type metal was formed by a rapid thermal process at 400°C in atmospheric environment after the mesa process. Thereafter, Ti/Al/Ni/Au with a thickness of 20/100/45/55 nm was deposited on the ITO layer as the electrodes. To reduce the sidewall defects generated by the mesa etching, a 5 nm thick aluminum oxide (Al2O3) passivation layer was grown on the full wafer by atomic layer deposition (ALD) technology. The growth condition of the thin film was a cycle of H2O and trimethylaluminum (TMA) in a 300°C argon (Ar) environment with Ar purging. A 200 nm thick SiO2 layer covered the thick aluminum-oxide layer through plasma-enhanced chemical vapor deposition (PECVD). Finally, the contact metal connected with the p-metal and n-type contact layer was deposited on the μLED device. The metal composition was Al/Ni/Au as a pad metal and a sidewall reflector, thus completing the fabrication of the μLED device.

3. RESULTS AND DISCUSSION

Reducing the diameters of LED devices is one of the most intuitive methods for high-speed VLC applications. The capacitance for a low-RC time constant can be achieved by minimizing the electrode through μLED device architecture. With the proper design of a ring contact and round mesa, the current spreading effect and the electrical performance can be improved. Furthermore, the influence of the sidewall defects arising from the etching process was mitigated via ALD technology. The passivation mechanism of ALD technology for μLEDs can be found in our previous research [19]. Moreover, the improved LEE and strain relaxation of the active region within the μLED device were realized with the assistance of NP-DBR.

Figure 2 summarizes the electrical and optical performances of the 30μm×8μLED array. The voltage and output power as a function of current density are shown in Fig. 2(a). Figure 1(a) shows an optical image of the 30μm×8μLED array illumination. The current density is normalized by the active region area. The 2.6 V turn-on voltage is at a current density of 18.56A/cm2, and the leakage current is also low (approximately 108A/cm2) at 4.8V due to the ALD process, which is regarded as a reasonable performance of a yellow-green μLED. Figure 2(b) shows that the maximum EQE of the μLED array is 8.7% before the efficiency droop at a low injection current of 12.73A/cm2. However, the efficiency droop phenomenon occurs while the current density is over 12.73A/cm2 because of the QCSE in polar c-plane GaN. Nevertheless, this device still suffers less efficiency droop compared to commercial c-plane wafers due to the presence of NP-DBR, which improves conversion efficiency at higher injection currents [40]. Figure 2(c) illustrates the electroluminescence (EL) emission spectra for increasing the injected current from 13A/cm2 to 255A/cm2 with an emission peak wavelength from 547 nm to 574 nm, belonging to the yellow-green spectral region. When the current density is increased from 13A/cm2 to 2500A/cm2, the peak wavelength shift is 27 nm as shown in Fig. 2(d). The peak wavelength tends to stabilize after the current density is over 1000A/cm2, which is regarded as excellent performance compared to conventional yellow-green μLEDs, where wavelength stability is a significant factor for device applications to prevent channel cross talk, since yellow-green μLEDs are commonly used in an RGBY wavelength-division multiplexing (WDM) system. Conforming to indoor illumination standards with OOK modulation, the minimum channel spacing is 33 nm [41]. Compared to conventional c-plane yellow-green LEDs in which the wavelength shift is 15 nm at a current density of 84.88A/cm2 [31], the wavelength shift of our device is only 27 nm at a current density of 2500A/cm2. The blueshift of the emission wavelength is attributed to the polarization-field-induced carrier screening effect and band filling effect, which becomes dominant at higher injection current densities. In addition to this, Fig. 2(d) shows that the full width at half maximum (FWHM) is shifted by 18 nm when the current density is varied from 13A/cm2 to 2500A/cm2. Table 1 summarizes the maximum current and maximum output power of the μLED array.

 figure: Fig. 2.

Fig. 2. (a) Voltage and output power as a function of current density of the 30μm×8μLED array; (b) EQE-current density curve of μLEDs; (c) electroluminescence spectra at different currents; (d) wavelength shift and FWHM as a function of current.

Download Full Size | PDF

Tables Icon

Table 1. Maximum Current Density and Maximum Output Power of the μLED Array

A schematic of the measurement for optical response is shown in Fig. 3(a). To measure the optical response of the μLED array, a vector network analyzer (VNA, Keysight, HP 8720ES) was used in this experiment. The μLEDs were probed by a high-speed microprobe (Cascade, ACP40-GS-250). A high-speed bias tee was utilized to couple the RF signal and DC current into μLEDs. A plastic fiber collected an optical signal and directly sent it into a photodetector (Graviton, SPA-3). The probe system and photodetector were de-embedded before the measurement. It ensured that the normalized response did not include the system response. According to the experimental setup, the optical response of μLEDs could be obtained. Figure 4 shows the measurement results of optical responses through different diameters and numbers. The 3dB bandwidth of each curve is summarized in Table 2 and Fig. 4(h). The 3dB bandwidth did not decrease with the number of emitters at the same emitter diameter. Therefore, the RC time constant did not dominate the modulation bandwidth as described in a previous paper [11]. Additionally, the 3dB bandwidth is proportional to the current density, which corresponds to Eq. (3). If the current crowding effect and self-heating do not occur, the 3dB bandwidth can be improved by increasing the current density regardless of the emitter diameter. However, the current crowding effect and self-heating cannot be ignored. The roll-off point of light versus current density will happen much earlier at a larger emitter diameter, which implies that the self-heating effect of the μLED array is more serious. To demonstrate the self-heating effect, the roll-off point of light versus current density and a thermal infrared image are further described in Fig. 5. Measurement results show that the 3dB bandwidth is proportional to the current density and inversely proportional to the emitter diameter, but it is almost independent of the emitter number. The maximum bandwidth could achieve 442 MHz at a current density of 2500A/cm2 for a 30μm×8μLED array. Furthermore, at a current density of 2500A/cm2, the 3dB bandwidth of the μLED array with a diameter of 30 μm was higher than that of the μLED array with a diameter of 50 μm. It implies that τRC is not the main factor in this experimental design. The dominant factor should be current density and emitter diameter. Due to less self-heating, a smaller μLED can endure higher current densities, resulting in shorter carrier lifetimes. As shown in Fig. 5(a), the 50μm×6μLED array fails when the current density is higher than 2500A/cm2. By comparison, the 30μm×8μLED shows a smaller rollover effect when the current density is up to 6500A/cm2. Thus, the thermal infrared image of the devices operated at 2500A/cm2 shown in Figs. 5(b)–5(d) implies that the 30 μm device can sustain a higher current density due to less self-heating. This phenomenon can be observed from the different current density levels in Fig. 4. The bandwidth is improved with increased injection current density and a shorter carrier lifetime. In addition to the high injection current, a small emitter diameter is also promoted for a uniform current spreading effect. Hence, a small diameter μLED can reach higher optical bandwidth even if the emitter number is increased. Apart from diameter, the NP-DBR structure can further improve the optical bandwidth. A high bandwidth μLED array also demonstrates that the NP-DBR serves as a strain-relaxed layer to mitigate the QCSE influence. We summarize the bandwidth record of the yellow-green μLED array mentioned in Section 1 and the results of this work together in Table 3. To the best of our knowledge, this work reaches the highest 3dB bandwidth for yellow-green μLEDs by far.

 figure: Fig. 3.

Fig. 3. Schematic of the measurement for (a) optical response and (b) transmission performance.

Download Full Size | PDF

 figure: Fig. 4.

Fig. 4. Normalized frequency response for (a) 30μm×1; (b) 30μm×4; (c) 30μm×6; (d) 30μm×8; (e) 50μm×1; (f) 50μm×4; (g) 50μm×6μLEDs arrays. (h) Summarized 3dB bandwidth for μLED arrays.

Download Full Size | PDF

Tables Icon

Table 2. 3dB Bandwidth for μLED Arrays

 figure: Fig. 5.

Fig. 5. (a) Current density versus optical power of 30μm×8 and 50μm×6μLED arrays, (b) shooting angle of the thermal infrared camera, and thermal infrared images of (c) 30μm×8 and (d) 50μm×6μLED arrays operated at 2500A/cm2.

Download Full Size | PDF

Tables Icon

Table 3. Benchmark of 3dB Bandwidth for Yellow-Green μLED Arrays

A schematic of the measurement for transmission performance is shown in Fig. 3(b). The transmission performance of the μLED array was demonstrated on an OOK system. The test bit sequence was an NRZ 271 pseudorandom binary sequence (PRBS7) generated by a bit pattern generator (Anritsu MP1800A). The eye diagrams were analyzed and recorded by the 3.5 GHz bandwidth Tektronix DPO 7354C oscilloscope. The measured eye diagrams for the μLED array are shown in Fig. 6; the vertical scale of the eye diagram is 50 mV/div. The eye diagram was measured with a bandwidth of 442 MHz at 2500A/cm2; the clear open eye can be observed at 200 Mbit/s and gradually becomes smaller when the bit rate is up to 800 Mbit/s, where the bit error rate (BER) is 2.6×103, satisfying the forward error correction (FEC) limit (3.8×103). This result responds to the bandwidth measurement and proves that VLC can be implemented via a yellow-green μLED. A benchmark for the yellow-green μLED array data transmission experiment is summarized in Table 4. In 2016, Luo et al. showed a 650 Mbit/s VLC system based on NRZ-OOK at a wavelength of 560 nm and a 3dB bandwidth of 238 MHz [42]. Furthermore, in 2019, Shi et al. demonstrated a 2.4344 Gbit/s underwater VLC system based on discrete Fourier transform-spread (DFT-S) orthogonal frequency-division multiplexing (OFDM) at a wavelength of 562 nm [43]. Moreover, in 2020, Jung et al. achieved a 16.2 kbit/s full-duplex LED-to-LED VLC system based on direct current-biased optical (DCO)-OFDM at a wavelength of 587 nm [44]. Also in 2020, Milovančev et al. showed a 1.25 Gbit/s VLC system based on OFDM at a wavelength of 588 nm and a 3dB bandwidth of 260 MHz [45]. Hence, despite the relatively low data rate of our work, it is possible to achieve a higher data rate through a novel transmission format due to the high 3dB bandwidth results of this work.

 figure: Fig. 6.

Fig. 6. Eye diagrams for the data rate measurement of the 30μm×8μLED array.

Download Full Size | PDF

Tables Icon

Table 4. Benchmark of the Data Rate for Yellow-Green μLED Arrays

4. CONCLUSION

In summary, we demonstrated a high 3dB bandwidth yellow-green InGaN/GaN μLED with the introduction of NP-DBR. This novel structure is beneficial to light extraction and strain release and can lessen the problem of QCSE, leading to improved EQE of yellow-green InGaN/GaN μLEDs. In addition, ALD technology was introduced for surface defect passivation, thereby reducing the leakage current. Consequently, the device exhibited the highest 3dB bandwidth of 442 MHz with a 547 nm peak wavelength at a current density of 2500A/cm2. Moreover, the data transmission experiment was also successfully shown, achieving a data rate of 800 Mbit/s with the NRZ-OOK modulation format. These results reveal the excellent optical and electrical properties of the yellow-green InGaN/GaN μLED device with an embedded NP-DBR, which has considerable potential for future high-speed VLC.

Funding

Ministry of Science and Technology, Taiwan (110-2124-M-A49-003-, 108-2221-E-009-113-MY3).

Acknowledgment

The authors thank Prof. Jun Han from Yale University and Prof. Chia-Feng Lin from National Chung Hsing University for fruitful discussions.

Disclosures

The authors declare no conflicts of interest.

Data Availability

The data that support the findings of this study are available from the authors upon reasonable request.

REFERENCES

1. H.-H. Chou and W.-T. Huang, “Asymmetrical bidirectional optical wireless communication system based on a transmissive 1D LC-SLM for NG-PON2,” Opt. Lett. 45, 4543–4546 (2020). [CrossRef]  

2. H. H. Chou and W. T. Huang, “Wavelength tunable asymmetric B-OWC system based on self-injection locking for TDM-PONs,” IEEE Photon. Technol. Lett. 33, 370–372 (2021). [CrossRef]  

3. C. Li, H. H. Lu, W. S. Tsai, Z. H. Wang, C. W. Hung, C. W. Su, and Y. F. Lu, “A 5 m/25 Gbps underwater wireless optical communication system,” IEEE Photon. J. 10, 7904909 (2018). [CrossRef]  

4. T. Wu, Y. Lin, Y. M. Huang, M. Liu, K. J. Singh, W. Lin, T. Lu, X. Zheng, J. Zhou, H. C. Kuo, and Z. Chen, “A highly stable full-color display device with VLC application potential using semipolar micro-LEDs and all-inorganic encapsulated perovskite nanocrystal,” Photon. Res. 9, 2132–2143 (2021). [CrossRef]  

5. T.-H. Do and M. Yoo, “An in-depth survey of visible light communication based positioning systems,” Sensors 16, 678 (2016). [CrossRef]  

6. T.-C. Yu, W.-T. Huang, W.-B. Lee, C.-W. Chow, S.-W. Chang, and H.-C. Kuo, “Visible light communication system technology review: devices, architectures, and applications,” Crystals 11, 1098 (2021). [CrossRef]  

7. S. Mei, X. Liu, W. Zhang, R. Liu, L. Zheng, R. Guo, and P. Tian, “High-bandwidth white-light system combining a micro-LED with perovskite quantum dots for visible light communication,” ACS Appl. Mater. Interfaces 10, 5641–5648 (2018). [CrossRef]  

8. K. J. Singh, X. Fan, A. S. Sadhu, C. H. Lin, F. J. Liou, T. Wu, Y. J. Lu, J. H. He, Z. Chen, T. Wu, and H. C. Kuo, “CsPbBr3 perovskite quantum-dot paper exhibiting a highest 3 dB bandwidth and realizing a flexible white-light system for visible-light communication,” Photon. Res. 9, 2341–2350 (2021). [CrossRef]  

9. L. Wang, L. Wang, C. J. Chen, K. C. Chen, Z. Hao, Y. Luo, C. Sun, M. C. Wu, J. Yu, Y. Han, and B. Xiong, “Green InGaN quantum dots breaking through efficiency and bandwidth bottlenecks of micro-LEDs,” Laser Photon. Rev. 15, 2000406 (2021). [CrossRef]  

10. Y. Wang, X. Chen, X. Liu, Z. Wang, L. Yu, and X. Zhang, “Experimental testing of high-capacity bandwidth efficient visible light communication with silicon-based RGBY-LED,” in IEEE International Symposium on Broadband Multimedia Systems and Broadcasting (BMSB) (2021), pp. 1–3.

11. M. Haemmer, B. Roycroft, M. Akhter, D. V. Dinh, Z. Quan, J. Zhao, P. J. Parbrook, and B. Corbett, “Size-dependent bandwidth of semipolar (112¯2) light-emitting-diodes,” IEEE Photon. Technol. Lett. 30, 439–442 (2018). [CrossRef]  

12. Z. Gong, S. Jin, Y. Chen, J. McKendry, D. Massoubre, I. M. Watson, E. Gu, and M. D. Dawson, “Size-dependent light output, spectral shift, and self-heating of 400 nm InGaN light-emitting diodes,” J. Appl. Phys. 107, 013103 (2010). [CrossRef]  

13. A. Rashidi, M. Monavarian, A. Aragon, and D. Feezell, “Thermal and efficiency droop in InGaN/GaN light-emitting diodes: decoupling multiphysics effects using temperature-dependent RF measurements,” Sci. Rep. 9, 19921 (2019). [CrossRef]  

14. S.-W. H. Chen, Y. M. Huang, K. J. Singh, Y. C. Hsu, F. J. Liou, J. Song, J. Choi, P. T. Lee, C. C. Lin, Z. Chen, and J. Han, “Full-color micro-LED display with high color stability using semipolar (20-21) InGaN LEDs and quantum-dot photoresist,” Photon. Res. 8, 630–636 (2020). [CrossRef]  

15. J. Piprek, “Efficiency droop in nitride-based light-emitting diodes,” Phys. Status Solidi A 207, 2217–2225 (2010). [CrossRef]  

16. D. A. B. Miller, D. S. Chemla, T. C. Damen, A. C. Gossard, W. Wiegmann, T. H. Wood, and C. A. Burrus, “Band-edge electroabsorption in quantum well structures: the quantum-confined Stark effect,” Phys. Rev. Lett. 53, 2173–2176 (1984). [CrossRef]  

17. M. Usman, M. Munsif, U. Mushtaq, A.-R. Anwar, and N. Muhammad, “Green gap in GaN-based light-emitting diodes: in perspective,” Crit. Rev. Solid State Mater. Sci. 46, 450–467 (2021). [CrossRef]  

18. M. Monavarian, A. Rashidi, A. A. Aragon, S. H. Oh, A. K. Rishinaramangalam, S. P. DenBaars, and D. Feezell, “Impact of crystal orientation on the modulation bandwidth of InGaN/GaN light-emitting diodes,” Appl. Phys. Lett. 112, 041104 (2018). [CrossRef]  

19. S.-W. H. Chen, Y. M. Huang, Y. H. Chang, Y. Lin, F. J. Liou, Y. C. Hsu, J. Song, J. Choi, C. W. Chow, C. C. Lin, and R. H. Horng, “High-bandwidth green semipolar (20–21) InGaN/GaN micro light-emitting diodes for visible light communication,” ACS Photon. 7, 2228–2235 (2020). [CrossRef]  

20. F. Jiang, J. Zhang, L. Xu, J. Ding, G. Wang, X. Wu, X. Wang, C. Mo, Z. Quan, X. Guo, and C. Zheng, “Efficient InGaN-based yellow-light-emitting diodes,” Photon. Res. 7, 144–148 (2019). [CrossRef]  

21. G.-Y. Shiu, K. T. Chen, F. H. Fan, K. P. Huang, W. J. Hsu, J. J. Dai, C. F. Lai, and C. F. Lin, “InGaN light-emitting diodes with an embedded nanoporous GaN distributed Bragg reflectors,” Sci. Rep. 6, 29138 (2016). [CrossRef]  

22. S. S. Pasayat, C. Gupta, M. S. Wong, Y. Wang, S. Nakamura, S. P. Denbaars, S. Keller, and U. K. Mishra, “Growth of strain-relaxed InGaN on micrometer-sized patterned compliant GaN pseudo-substrates,” Appl. Phys. Lett. 116, 111101 (2020). [CrossRef]  

23. C.-J. Wang, Y. Ke, G. Y. Shiu, Y. Y. Chen, Y. S. Lin, H. Chen, and C. F. Lin, “InGaN resonant-cavity light-emitting diodes with porous and dielectric reflectors,” Appl. Sci. 11, 8 (2021). [CrossRef]  

24. C. Zhang, S. H. Park, D. Chen, D. W. Lin, W. Xiong, H. C. Kuo, C. F. Lin, H. Cao, and J. Han, “Mesoporous GaN for photonic engineering—highly reflective GaN mirrors as an example,” ACS Photon. 2, 980–986 (2015). [CrossRef]  

25. C. C. Yang, C. F. Lin, C. M. Lin, C. C. Chang, K. T. Chen, J. F. Chien, and C. Y. Chang, “Improving light output power of InGaN-based light emitting diodes with pattern-nanoporous p-type GaN:Mg surfaces,” Appl. Phys. Lett. 93, 203103 (2008). [CrossRef]  

26. K. J. Lee, S.-J. Kim, J.-J. Kim, K. Hwang, S.-T. Kim, and S.-J. Park, “Enhanced performance of InGaN/GaN multiple-quantum-well light-emitting diodes grown on nanoporous GaN layers,” Opt. Express 22, A1164–A1173 (2014). [CrossRef]  

27. N. Nakada, M. Nakaji, H. Ishikawa, T. Egawa, M. Umeno, and T. Jimbo, “Improved characteristics of InGaN multiple-quantum-well light-emitting diode by GaN/AlGaN distributed Bragg reflector grown on sapphire,” Appl. Phys. Lett. 76, 1804–1806 (2000). [CrossRef]  

28. B.-C. Shieh, Y. C. Jhang, K. P. Huang, W. C. Huang, J. J. Dai, C. F. Lai, and C. F. Lin, “InGaN light-emitting diodes with embedded nanoporous GaN distributed Bragg reflectors,” Appl. Phys Express 8, 082101 (2015). [CrossRef]  

29. R. T. Elafandy, J.-H. Kang, C. Mi, T. K. Kim, J. S. Kwak, and J. Han, “Study and application of birefringent nanoporous GaN in the polarization control of blue vertical-cavity surface-emitting lasers,” ACS Photon. 8, 1041–1047 (2021). [CrossRef]  

30. S. Yamamoto, Y. Zhao, C. C. Pan, R. B. Chung, K. Fujito, J. Sonoda, S. P. DenBaars, and S. Nakamura, “High-efficiency single-quantum-well green and yellow-green light-emitting diodes on semipolar (2021) GaN substrates,” Appl. Phys. Express 3, 122102 (2010). [CrossRef]  

31. C. Du, Z. Ma, J. Zhou, T. Lu, Y. Jiang, P. Zuo, H. Jia, and H. Chen, “Enhancing the quantum efficiency of InGaN yellow-green light-emitting diodes by growth interruption,” Appl. Phys. Lett. 105, 071108 (2014). [CrossRef]  

32. S. Zhang, J. J. McKendry, Z. Gong, B. R. Rae, S. Watson, E. Xie, P. Tian, E. Richardson, E. Gu, Z. Chen, and G. Zhang, “Directly color-tunable smart display based on a CMOS-controlled micro-LED array,” in IEEE Photonics Conference (2012), pp. 435–436.

33. X. Zhu, F. Wang, M. Shi, N. Chi, J. Liu, and F. Jiang, “10.72 Gb/s visible light communication system based on single packaged RGBYC LED utilizing QAM-DMT modulation with hardware pre-equalization,” in Optical Fiber Communication Conference (2018), paper M3K.3.

34. J. I. H. Haggar, Y. Cai, S. S. Ghataora, R. M. Smith, J. Bai, and T. Wang, “High modulation bandwidth of semipolar (11–22) InGaN/GaN LEDs with long wavelength emission,” ACS Appl. Electron. Mater. 2, 2363–2368 (2020). [CrossRef]  

35. G. M. Yang, M. H. MacDougal, and P. D. Dapkus, “Ultralow threshold current vertical-cavity surface-emitting lasers obtained with selective oxidation,” Electron. Lett. 31, 886–888 (1995). [CrossRef]  

36. G. S. Huang, T. C. Lu, H. H. Yao, H. C. Kuo, S. C. Wang, C. W. Lin, and L. Chang, “Crack-free GaN/AlN distributed Bragg reflectors incorporated with GaN/AlN superlattices grown by metalorganic chemical vapor deposition,” Appl. Phys. Lett. 88, 061904 (2006). [CrossRef]  

37. X. Yang, Z. Chen, D. Cao, C. Zhao, L. Shen, C. Luan, Z. Pang, J. Liu, J. Ma, and H. Xiao, “Large-area, liftoff nanoporous GaN distributed Bragg reflectors: fabrication and application,” Appl. Surf. Sci. 489, 849–855 (2019). [CrossRef]  

38. J. Park, J.-H. Kang, and S.-W. Ryu, “High diffuse reflectivity of nanoporous GaN distributed Bragg reflector formed by electrochemical etching,” Appl. Phys Express 6, 072201 (2013). [CrossRef]  

39. J. Song, J. Choi, C. Zhang, Z. Deng, Y. Xie, and J. Han, “Elimination of stacking faults in semipolar GaN and light-emitting diodes grown on sapphire,” ACS Appl. Mater. Interfaces 11, 33140–33146 (2019). [CrossRef]  

40. J. H. Son and J.-L. Lee, “Strain engineering for the solution of efficiency droop in InGaN/GaN light-emitting diodes,” Opt. Express 18, 5466–5471 (2010). [CrossRef]  

41. L. Cui, Y. Tang, H. Jia, J. Luo, and B. Gnade, “Analysis of the multichannel WDM-VLC communication system,” J. Lightwave Technol. 34, 5627–5634 (2016). [CrossRef]  

42. J. Luo, Y. Tang, H. Jia, Q. Zhu, and W. Xue, “750 Mb/s monochromatic LED-based real-time visible light communication system employing a low-complexity cascaded post-equalizer,” Chin. Opt. Lett. 14, 120604 (2016). [CrossRef]  

43. J. Shi, X. Zhu, F. Wang, P. Zou, Y. Zhou, J. Liu, F. Jiang, and N. Chi, “Net data rate of 14.6 Gbit/s underwater VLC utilizing silicon substrate common-anode five primary colors LED,” in Optical Fiber Communication Conference (OFC) (2019), paper M3I.5.

44. H. Jung and S.-M. Kim, “A full-duplex LED-to-LED visible light communication system,” Electronics 9, 1713 (2020). [CrossRef]  

45. D. Milovančev, N. Vokić, H. Hübel, and B. Schrenk, “Gb/s visible light communication with low-cost receiver based on single-color LED,” J. Lightwave Technol. 38, 3305–3314 (2020). [CrossRef]  

Data Availability

The data that support the findings of this study are available from the authors upon reasonable request.

Cited By

Optica participates in Crossref's Cited-By Linking service. Citing articles from Optica Publishing Group journals and other participating publishers are listed here.

Alert me when this article is cited.


Figures (6)

Fig. 1.
Fig. 1. (a) Image of the 30μm×8μLED array, (b) schematic diagram of μLED structures, (c) scanning electron microscope image of the NP-DBR, and (d) simulated and experimental reflectance spectra of the DBR.
Fig. 2.
Fig. 2. (a) Voltage and output power as a function of current density of the 30μm×8μLED array; (b) EQE-current density curve of μLEDs; (c) electroluminescence spectra at different currents; (d) wavelength shift and FWHM as a function of current.
Fig. 3.
Fig. 3. Schematic of the measurement for (a) optical response and (b) transmission performance.
Fig. 4.
Fig. 4. Normalized frequency response for (a) 30μm×1; (b) 30μm×4; (c) 30μm×6; (d) 30μm×8; (e) 50μm×1; (f) 50μm×4; (g) 50μm×6μLEDs arrays. (h) Summarized 3dB bandwidth for μLED arrays.
Fig. 5.
Fig. 5. (a) Current density versus optical power of 30μm×8 and 50μm×6μLED arrays, (b) shooting angle of the thermal infrared camera, and thermal infrared images of (c) 30μm×8 and (d) 50μm×6μLED arrays operated at 2500A/cm2.
Fig. 6.
Fig. 6. Eye diagrams for the data rate measurement of the 30μm×8μLED array.

Tables (4)

Tables Icon

Table 1. Maximum Current Density and Maximum Output Power of the μLED Array

Tables Icon

Table 2. 3dB Bandwidth for μLED Arrays

Tables Icon

Table 3. Benchmark of 3dB Bandwidth for Yellow-Green μLED Arrays

Tables Icon

Table 4. Benchmark of the Data Rate for Yellow-Green μLED Arrays

Equations (3)

Equations on this page are rendered with MathJax. Learn more.

f3dB=32π(1τr+1τnr+1τRC),
I=e·a·d(An+Bn2+Cn3),
τ1=A+2Bn+3Cn2.
Select as filters


Select Topics Cancel
© Copyright 2024 | Optica Publishing Group. All rights reserved, including rights for text and data mining and training of artificial technologies or similar technologies.