Abstract
A modified two-level model is proposed to study the spatially resolved current density distribution of GaN-based green miniaturized light-emitting diodes (mini-LEDs), combining with microscopic hyperspectral imaging. We found that the spatially resolved current density distribution reveals both the radiative and non-radiative recombination mappings, which can also be provided separately by this model. In addition, higher current density is not necessarily correlated with higher photon emission, especially for the regions around the electrode edges, where the high current density suggests current crowding and defect-related non-radiative recombination. The current density distribution of mini-LEDs is further verified by the laser-beam-induced current (LBIC) and the spatially resolved mappings of peak wavelength and FWHM. The modified two-level model also offers radiative/non-radiative mappings and is proved to be beneficial to determine the micro-zone current density distribution and to reveal the intrinsic radiative/non-radiative recombination mechanism of mini-LEDs.
© 2024 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
As a solid-state light source with tiny pitches, ultrahigh contrast, long lifetime, and high dynamic range (HDR), GaN-based miniaturized light-emitting diodes (mini-LEDs) have emerged as the potential next-generation display technology, such as high resolution display, digital signage, augmented reality (AR), and virtual reality (VR) [1–3]. Since the operating current density of mini-LEDs (about 1 A/cm2) is much smaller than that of conventional LEDs, one of the most troublesome challenges for the high-efficiency GaN-based mini-LEDs is the inhomogeneous current distribution. The non-uniform current spreading, or local current crowding can result in a high localized current density, anomalous heat generation, and non-uniform photon emission [4–6]. Latest researches demonstrate that by employing methods such as adopting a multi-layered structure, introducing a current diffusion layer, optimizing contact design, increasing contact area, and selecting specific materials, the issues related to local current crowding effects in devices can be effectively mitigated, thereby enhancing device performance [7–10].
Shim et al. proposed a method for current spreading analysis to demonstrate the local current-related output degradation [11]. Concurrently, the impacts of electrode patterns on the current spreading in LEDs have been investigated by a hybrid modeling method [12]. Given that the mini-LEDs usually operate at a relatively small current density, where Shockley–Read–Hall (SRH) recombination often saturates with increasing excitation density, the conventional ABC model is insufficient for understanding the radiative/nonradiative recombination behaviors in mini-LEDs [13]. The interplay of the extended defects (EDs) and point defects (PDs) have been introduced into the non-uniform spatial distribution of the external quantum efficiency (EQE) in GaN-based LEDs [14]. Lin and Zhang et al. proposed a two-level model, i.e. one band edge states related emission, the other one PD states-related radiative or nonradiative recombination, to investigate the recombination process and inner mechanism induced by EDs and PDs under different carrier generation levels in conventional LEDs [15]. The two-level model is further expanded to the EQE study of photo-voltaic materials [16]. A laser-beam-induced current (LBIC) retrieves information by detecting the generated photocurrent when a light beam scans across the surface of a semiconductor device [17]. Nejand et al. employed Electroluminescence (EL) and photoluminescence (PL) imaging, as well as LBIC mapping to measure homogeneous current collection and a low defect density over the entire module area. This in-depth assessment aimed to evaluate the performance and potential issues of tandem perovskite solar cells with both ends in series [18].
Empirically, the inhomogeneous distribution of current density is generally judged from the inhomogeneous photon emission distribution. However, the real current density distribution is not completely correlated to the spatial photon emission distribution because of the considerable defect related non-radiative recombination [19,20]. Given the large specific surface/sidewall area ratio and high defect density, the efficiency of mini-LEDs is sensitive to the local current density, which has been regarded as the critical challenges of high-performance GaN-based mini-LEDs [21,22]. Therefore, the investigation on micro-zone current density distribution is of importance to determine the luminescence and the intrinsic recombination mechanism of mini-LEDs.
In this paper, combining microscopic hyperspectral imaging technique, a modified two-level model is proposed to determine the spatially resolved micro-zone current density distribution of GaN-based mini-LEDs. The model also reveals the impacts of radiative and nonradiative recombination across the mesa and electrodes. Additionally, the inhomogeneous electroluminescence (EL) distribution of mini-LEDs do not completely agree with the actual current distribution. To compare with the simulation results of the proposed model, the LBIC is employed to obtain the spatially resolved photo-generated current mappings of mini-LEDs. LBIC detection provides positional information on the photoresponse of different regions within the device. The induced current is typically associated with micro-defects, thereby reflecting the local conductivity to some extent, which can be compared with the micro-region current density distribution [23]. This synergistic approach unveils the recombination behaviors of electrons and holes within the device, facilitating a comprehensive understanding of the electroluminescent mechanisms in the device.
2. Experiment
Two flip-chip green mini-LEDs with rectangular and T-type electrode geometries have been used in this work, marked as Sample A and Sample B. Both samples, featuring a patterned sapphire substrate (PSS), are 76.2 × 127 µm2 in size with a 45 × 80 µm2 mesa. The schematic diagrams of the samples are illustrated in Fig. 1(a), the epi-layer of green mini-LED is composed of n-GaN layer, InGaN/GaN multiple quantum wells active layer, p-GaN layer, Indium tin oxide (ITO) layer, and distributed Bragg reflection (DBR) layer. The sample was mounted on a temperature-controlling stage (TEC 264M-BB-DB9, Arroyo Instruments Inc.) with heatsink temperature maintained at 300 K, and driven by a source/measure unit (KeysightB2912A, Keysight Inc.). The spatially resolved photon emission distribution was captured by a microscopic hyperspectral imaging system (GaiaFieldF-V10, Dualix Inc.) with a 50X object lens (NA = 0.55).The spatial resolution and the spectral resolution is 500 nm and 4 nm, respectively. A scanning spectrometer (Spectro-320e, Instrument Systems Inc.) was utilized to calibrate the absolute spectral power distribution of mini-LEDs. During the LBIC measurement, the mini-LEDs were excited by a 405-nm-laser with an approximate excitation density of 50 W/cm2, the spatial resolution is about 450 nm with a 50X microscope lens (NA = 0.55). The induced current was measured by a current measure unit (Keithley 2611A, Keithley Inc.).
3. Results and discussion
The normalized EL spectra of mini-LEDs at different current densities are shown in Fig. 2(a) and (c), as the current density increases from 1 to 100 A/cm2, the peak wavelengths of the EL spectra blueshift from 543 nm to 520 nm for sample A, from 541 nm to 522 nm for sample B, respectively. The blueshift is attributed to the screening of Quantum Confined Stark Effect (QCSE) [24]. In Fig. 2(b) and (d), the non-uniform spatially resolved EL emission mappings are due to the non-uniform current density distribution (it may bring deviations of mini-LEDs’ performance from chip to chip), while the band edge recombination and defect states are recognized as the critical impact factors on the current density distribution.
For the InGaN/ GaN green mini-LEDs with considerable EDs and PDs, the lateral and vertical current flows can lead to the non-uniform current density distribution across the mesa. Herein, to investigate spatially resolved efficiency of LED, the micro-zone EQE can be expressed as
After acquiring hyperspectra and absolute spectral power distribution of mini-LED under a series of injection currents, the spectral power distribution at different injection currents during the EQE rising stage was fitted for each pixel based on Eq. (4). To increase the signal to noise ratio, we usually combine 2 × 2 pixels as a new one. Eventually, the mappings of α′, β′ and ξm were obtained. According to Eq. (4), the spatially resolved distribution of current density can be represented by the mappings of ξm. As illustrated in Fig. 3(a), the radiative recombination (α′) mainly occurs in the central bright regions of the mesa, marked as Region 1. Whereas, discrete dark spots are also exhibited across the mesa, suggesting the relatively lower radiative recombination rate in these regions. This phenomenon is consistent to the photon emission distribution of the mesa in Fig. 2(b). Meanwhile, the region inside the p-electrode possesses a relatively high recombination rate and high current density, marked as Region 2, which, however, is unrevealed in the EL mappings due to the light absorption of anode.
Note worthily, as illustrated in Fig. 3(b), contrarily to Fig. 3(a), the bright spots of coefficient β′ are mainly distributed in the complementary regions of the mesa and the edges of p-electrode, suggesting that the local PD density is higher and the SRH recombination is dominant in these edge regions. The current density distribution (ξm) is illustrated in Fig. 3(c). In the center area of mesa, the distribution of ξm is similar to that of α′ (Region 1), suggesting that the brighter photon emission distribution is mainly attributed to the higher current density in this region. However, this is not necessarily a vice-versa case, especially around the edge of p-electrode (Region 3 in white curve), where the current density is remarkably high, leading to severe current crowding but dominant non-radiative recombination, consistent with the counterpart mapping of β′ in Fig. 3(b).
The LBIC mapping across the mesa of sample A is depicted in Fig. 3(d), offering an alternative perspective for the validation of the electrical channels. It is observable that, despite the limitations imposed by resolution on LBIC detection, the distribution of photocurrent at the electrode edges, mesa center, and periphery remains consistent with the observations in Figs. 3(a)-(c). Since the incident light is partly absorbed by the p-electrode, the photogenerated current distribution inside the p-electrode is hardly resolvable, however, different photogenerated current distribution still exists inside the p-electrode after carefully resolved, as also revealed in Region 2 of Fig. 3(c). The photogenerated current approaches its maximum at the center of the mesa, indicating that the photogenerated carrier channels are dominant in this area, consistent with Region 1 in ξm mapping of Fig. 3(c). Meanwhile, the carrier localization effect in the center region can enhance the radiative recombination by trapping carriers into localization centers, leading to the non-uniform current density. Generally, current density distribution can reveal the spatial photon emission distribution. However, higher current density is not necessarily correlated with higher photon emission, especially for the region around the electrode edge, where the current crowding and the defect-related non-radiative recombination are dominant. Likewise, for the Sample B with T-type p-electrode, the complementary distribution of radiative recombination and non-radiative recombination via defect level are illustrated in Fig. 4(a) and (b). Comparing to the EL distribution in Fig. 2(d), the mappings of current density (coefficient ξm) obtained by the modified two-level model in Fig. 4(c) suggests a dominant distribution of bright regions across mesa (Region 1), further confirmed by LBIC distribution (Fig. 4(d)). Similar to the results of Sample A, Region 2 and Region 3 in Sample B reveal higher current density distributions in p-electrode.
The spatially resolved mapping of spectral properties including peak wavelength and Full Width at Half Maximum (FWHM) of sample A at small and large current density are illustrated in Fig. 5. The peak wavelength exhibits an average blueshift of ∼23 nm from 0.1 A/cm2 to 100 A/cm2, further confirming the screening of QCSE, as shown in Fig. 5(a) and (b). The non-uniform wavelength distribution across the mesa is possibly associated with the carrier localization effect by indium fluctuations or the non-uniform current density distribution. Considering the current crowding effect and the higher current distribution around the p-electrode, where a more significant blueshift in wavelength occurs (Fig. 5(b)), especially at 100A/cm2, the non-uniform wavelength distribution across the mesa is more likely attributed to the current density distribution. Moreover, as illustrated in Fig. 5(c) and (d), comparing the low and high injections, the mapping of FWHM around the p-electrode and the center of mesa display obvious expansions at high injection. The experimental results of peak wavelength and FWHM mappings further confirm the current density analysis in Fig. 3.
4. Conclusion
In conclusion, we performed a spatially resolved study of the current density distribution in GaN-based flip-chip green mini-LEDs by combining microscopic hyperspectral imaging and a modified two-level model simulation. Two samples with different electrodes are tested to verify the experimental and simulation results. The current density (ξm) distribution suggests a dominant distribution of bright regions across mesa and in the p-electrode. In addition, current crowding effect is clearly illustrated around the p-electrode edge, which is related with the non-radiative recombination as also revealed in the β′ mapping. The LBIC and the spatially resolved mappings of peak wavelength and FWHM further confirm the current density distribution. The modified model offers a helpful approach for determining the micro-zone current density distribution, facilitates a better understanding of the local intrinsic radiative/non-radiative recombination mechanism, and will be beneficial to improve the performance of mini-LEDs.
Funding
National Natural Science Foundation of China (62275227); Major Science and Technology Project of Fujian Province (2019H6004, 2020H6017).
Acknowledgments
The authors wish to thank anonymous reviewers for their valuable suggestions.
Disclosures
The authors declare no conflicts of interest.
Data availability
The data underlying the results presented in this paper are not publicly at this time but can be obtained from the authors upon reasonable request.
References
1. P. J. Parbrook, Brian Corbett, Jung Han, et al., “Micro-Light Emitting Diode: From Chips to Applications,” Laser Photon. Rev. 15, 1 (2021). [CrossRef]
2. .Y. Huang, Guanjun Tan, Fangwang Gou, et al., “Prospects and challenges of mini-LED and micro-LED displays,” J. Soc. Inf. Disp. 27(7), 387–401 (2019). [CrossRef]
3. .M.-Y. Deng, En-Lin Hsiang, Qian Yang, et al., “Reducing Power Consumption of Active-Matrix Mini-LED Backlit LCDs by Driving Circuit,” IEEE Trans. Electron Devices 68(5), 2347–2354 (2021). [CrossRef]
4. .J. Che, Hua Shao, Le Chang, et al., “Doping-Induced Energy Barriers to Improve the Current Spreading Effect for AlGaN-Based Ultraviolet-B Light-Emitting Diodes,” IEEE Electron Device Lett. 41, 1 (2020). [CrossRef]
5. .J. Iveland, Lucio Martinelli, Jacques Peretti, et al., “Direct Measurement of Auger Electrons Emitted from a Semiconductor Light-Emitting Diode under Electrical Injection: Identification of the Dominant Mechanism for Efficiency Droop,” Phys. Rev. Lett. 110(17), 177406 (2013). [CrossRef]
6. .H. Kim, Jaehee Cho, Jeong Wook Lee, et al., “Consideration of the actual current-spreading length of GaN-based light-emitting diodes for high-efficiency design,” IEEE J. Quantum Electron. 43(8), 625–632 (2007). [CrossRef]
7. .H. K. Su, Shengrui Xu, Hongchang Tao, et al., “Improving the Current Spreading by Fe Doping in n-GaN Layer for GaN-Based Ultraviolet Light-Emitting Diodes,” IEEE Electron Device Lett. 42(9), 1346–1349 (2021). [CrossRef]
8. .H. C. Tao, Shengrui Xu, Yanrong Cao, et al., “Enhanced Performance of N-Polar AlGaN-Based Ultraviolet Light-Emitting Diodes With Lattice- Matched AlInGaN Insertion in n-AlGaN Layer,” IEEE Photonics J. 15(3), 1–5 (2023). [CrossRef]
9. .Y. H. Hsu, Yi-Hsin Lin, Ming-Hsien Wu, et al., “Current Confinement Effect on the Performance of Blue Light Micro-LEDs with 10 µm Dimension,” ACS Omega 8(38), 35351–35358 (2023). [CrossRef]
10. .Y. L. Wang, Peixian Li, Xinyu Zhang, et al., “Using a Multi-Layer Stacked AlGaN/GaN Structure to Improve the Current Spreading Performance of Ultraviolet Light-Emitting Diodes,” Materials 13(2), 454 (2020). [CrossRef]
11. .S. Hwang and J. Shim, “A method for current spreading analysis and electrode pattern design in light-emitting diodes,” IEEE Trans. Electron Devices 55(5), 1123–1128 (2008). [CrossRef]
12. .P. Wang, Wei Wei, Bin Cao, et al., “Simulation of current spreading for GaN-based light-emitting diodes,” Opt. Laser Technol. 42(5), 737–740 (2010). [CrossRef]
13. .J. Piprek, “Efficiency droop in nitride-based light-emitting diodes,” Physica Status Solidi (a) 207(10), 2217–2225 (2010). [CrossRef]
14. .N. S. Averkiev, A.E. Chernyakov, M.E. Levinshtein, et al., “Two channels of non-radiative recombination in InGaN/GaN LEDs,” Phys. B 404(23-24), 4896–4898 (2009). [CrossRef]
15. .Y. Lin, Yong Zhang, Zhiqiang Liu, et al., “Spatially resolved study of quantum efficiency droop in InGaN light-emitting diodes,” Appl. Phys. Lett. 101(25), 1 (2012). [CrossRef]
16. .F. Zhang, Jose F. Castaneda, Shangshang Chen, et al., “Comparative studies of optoelectrical properties of prominent PV materials: Halide perovskite, CdTe, and GaAs,” Mater. Today 36, 18–29 (2020). [CrossRef]
17. W. C. Qiu and W. D. Hu, “Laser beam induced current microscopy and photocurrent mapping for junction characterization of infrared photodetectors,” Sci. China-Phys. Mech. Astron. 58(2), 1–13 (2015). [CrossRef]
18. .B. A. Nejand, David B. Ritzer, Hang Hu, et al., “Scalable two-terminal all-perovskite tandem solar modules with a 19.1% efficiency,” Nat. Energy 7(7), 620–630 (2022). [CrossRef]
19. .M. Sheikhi, Wei Guo, Yijun Dai, et al., “Mechanism of Improved Luminescence Intensity of Ultraviolet Light Emitting Diodes (UV-LEDs) Under Thermal and Chemical Treatments,” IEEE Photonics J. 11(6), 1–8 (2019). [CrossRef]
20. .W. Guo, Changwen Su, Hao Lu, et al., “Origins of Inhomogeneous Light Emission From GaN-Based Flip-Chip Green Micro-LEDs,” IEEE Electron Device Lett. 40(7), 1132–1135 (2019). [CrossRef]
21. .S. Lai, Wansheng Lin, Jinlan Chen, et al., “The impacts of sidewall passivation via atomic layer deposition on GaN-based flip-chip blue mini-LEDs,” J. Phys. D: Appl. Phys. 55(37), 374001 (2022). [CrossRef]
22. .Y. Huang, Ming-Yang Deng, Shin-Tson Wu, et al., “Mini-LED, Micro-LED and OLED displays: present status and future perspectives,” Light: Sci. Appl. 9(1), 105 (2020). [CrossRef]
23. .M. Burghard and A. Mews, “High-Resolution Photocurrent Mapping of Carbon Nanostructures,” ACS Nano 6(7), 5752–5756 (2012). [CrossRef]
24. .J.-H. Ryou, P. Douglas Yoder, Jianping Liu, et al., “Control of Quantum-Confined Stark Effect in InGaN-Based Quantum Wells,” IEEE J. Sel. Top. Quantum Electron. 15(4), 1080–1091 (2009). [CrossRef]