Abstract
Atomic layer deposited Al2O3 films are incorporated into miniature light emitting diodes (mini-LEDs) as an internal moisture barrier layer. The experimental results show that the water vapor transmission rate reaches ≤10−4 g/m2/day when the Al2O3 thickness is ≥40 nm. The mini-LED with a 40 nm-thick Al2O3 layer shows negligible degradation after 1000 h of 85°C/85% relative humidity testing, whereas the device without an Al2O3 layer fails after only 500 h due to delamination occurring at the GaN surface. Current-voltage characteristics of the device without an Al2O3 moisture barrier layer indicate an increase in series resistance and ideality factor. This study provides a simple, light-weighting method to have a satisfactory encapsulation function for miniature LEDs.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
GaN-based light emitting diode (LED) chips have been widely used in light-emitting fields such as in liquid-crystal display backlight, traffic signal lights, and indoor and outdoor lighting [1–4]. Recently, miniature LEDs have been developed and are considered as an advanced version of traditional LEDs owing to their much smaller backlight size, ability for local dimming, low power consumption, better contrast and improved brightness [5]. In the long run, miniature LEDs will be widely used in various display fields, from small screens to large screens.
For evaluating the performance of mini-LEDs, it is necessary to consider not only their brightness and photoelectric properties, but also stable reliability, which is crucial for the mini-LED chips to continue emitting light under different working environments. In high-temperature and high-humidity environments, mini-LED chips are most likely to be damaged. The failure mechanism of mini-LED chips at high temperatures is mainly driven by the deformation of the internal structure of the chip caused by the junction temperature, which causes the chip to lose its light-emitting function [6,7]. In high-humidity environments, one cause of failure is the oxidation of the electrodes of the LED chip by water vapor, resulting in the non-conductivity of the current [8]. Luo et al reported that the diffusion of moisture into LEDs not only decreases light output but also rises the risk of LED failure [9]. Therefore, special materials and sealing techniques must be incorporated into the design of LED chips in order to enhance their reliability. At the present, the most commonly used LED encapsulation materials are epoxy resin, silicone materials, poly(methyl methacrylate), polycarbonate, and UV curable silicone-epoxy hybrid materials [10,11]. However, these methods are mostly based on adding a thick encapsulation material to the outside of the devices and may no longer be suitable for devices requiring a high degree of flexibility or for places where bulky encapsulation is not favored. With the increasing trend of light-weighting or even removing of the bulky encapsulation layer, it is necessary to develop internal thin moisture barriers for miniature LEDs using a simple process.
In this study, a thin Al2O3 layer is incorporated into miniature LEDs as an internal moisture barrier layer. The thickness is varied from 40 to 80 nm to investigate its effect on device reliability in a high-temperature and high-humidity environment. A significant improvement in device lifetime evaluated from the reliability tests is demonstrated and discussed. The internal moisture barrier layer that enhances light output is also presented.
2. Experimental method
The Al2O3 films were prepared using ALD (R-200, Picosun, Finland) by alternately exposing trimethylaluminum (TMA, purity: 99.999%, Ai Mou Yuang Co. Ltd., Nanjing, China) and water. The optimized deposition parameters are summarized in Table 1. For the miniature LED fabrication, the n-GaN/multi-quantum well (MQW)/p-GaN layers were prepared on a patterned sapphire substrate. The p-GaN layer was doped with Mg, while the n-GaN layer was doped with Si. The thicknesses of n-GaN and p-GaN were 3 and 0.1 µm, respectively. The MQW layer had a total thickness of 160 nm and was composed of 10 pairs of 13 nm InGaN and 3 nm AlGaN layers. Next, the mesa patterning was performed using photolithography, followed by mesa etching using inductively coupled plasma reactive ion etching to expose n-GaN. A plasma enhanced chemical vapor deposition (PECVD) system was employed to deposit a SiO2 layer with a thickness of 230 nm as a blocking layer, which was then patterned and dry-etched. An indium tin oxide (ITO) layer with a thickness of 110 nm was sputtered on the SiO2 blocking layer. Afterwards, a first-electrode layer (FEL) containing Al/Pt/Ti multilayers was evaporated on the p and n-GaN regions. An ALD Al2O3 thin layer with a thickness varying from 20 to 80 nm was deposited as an internal moisture barrier layer, covered by a 650 nm-thick PECVD SiO2 layer and a Bragg reflector (DBR) containing SiO2 and Ti3O5 stacks. Subsequently, the DBR, SiO2, and Al2O3 layers were patterned and dry-etched until the underlying FEL was revealed. Finally, a Ti/Al/Ti/Au second electrode layer (SEL) as the p and n-metal pads was deposited. FEL and SEL both had a thickness of around 2 µm. The DBR included 25.5 pairs of 95 nm SiO2 and 73 nm Ti3O5 layers. For comparison, traditional miniature LEDs were also fabricated with the identical structure and processes, except for the incorporation of the Al2O3 layer. The length and width of the miniature LED die were around 600 and 200 µm, respectively.
The film thickness was evaluated by using an alpha-step profilometer (D500, KLA Tencor, USA) and an ellipsometer (M2000, J.A. Woollam Co., Inc., USA). The water vapor transmission rate (WVTR) was determined using the Ca corrosion test method [12]. The samples used in the Ca test were prepared as follows. The Al film with a thickness of 300 nm was sputtered as the electrodes on the both sides of a glass substrate, and then the Ca film with a thickness of 50 nm and an area of 1 cm2 was deposited between the electrodes. Afterwards, the samples were encapsulated by a 650 nm-thick SiO2 or a Al2O3 layer (with varying thickness of 20-60 nm). The samples were placed in a programmable constant humidity and temperature test chamber. The real-time change in Ca film resistance was monitored with a four probe resistance tester. The cross-sectional images of the miniature LEDs were observed using a scanning electron microscope (SEM, Sigma 500, Zeiss, Germany). The light distribution was measured using a light intensity instrument with an integrating sphere. The optical power was measured using an optical power meter coupled into an integrated sphere. The device reliability tests were performed at a high relative humidity (RH) of 85% and a high temperature of 85 °C for 1000 h. The electroluminescence (EL) spectra were obtained using a spectrophotometer (UV2600, Shimadzu, Japan). The current-voltage (I-V) curves were obtained using a Keithley 2400 source meter (Keithley Instruments, US).
3. Results and discussion
The deposition parameters of the ALD Al2O3 layer are optimized to ensure ALD growth with self-limiting surface reactions. The optimized parameters are substrate temperature of 220 °C, TMA pulse time of 0.08 s, H2O pulse time of 0.1 s, TMA purge time of 3 s, and H2O purge time of 6 s. The eventually saturated growth-per-cycle (GPC) is 1.07 Å/cycle, which is similar to reported values [13]. A linear relationship between GPC and ALD cycle number, which is also an important feature for confirming the ALD growth, has also been obtained.
The optimized Al2O3 layer is applied to the miniature LED die fabrication. The schematic diagram of the device structure is shown in Fig. 1(a), consisting of patterned sapphire, n-GaN, MQW, p-GaN, SiO2, ITO, FEL, Al2O3 moisture barrier layer, SiO2, DBR layers, and SEL. The device without the Al2O3 barrier layer refers to a traditional miniature LED die. The cross-sectional SEM images of miniature LED dies without and with an Al2O3 layer (80 nm-thick as an example) are shown in Figs. 1(b) and 1(c). The Al2O3 layer provides excellent coverages on different materials including FELs (side-walls), ITO, p-GaN, and n-GaN. The thickness on each region is nearly the same thanks to the high conformality of the ALD technique [14]. It is believed that no pinholes in the Al2O3 layer and no uncovered areas at the interface between Al2O3 and underlying layers.
From the device structure, the thick SiO2 layer could already provide a certain level of moisture resistance, and the added Al2O3 layer enhances the function. To quantify and compare moisture resistivity, WVTR measurements using the Ca-corrosion method are respectively performed on 650 nm-thick SiO2 layer and 20 to 80 nm-thick Al2O3 layer. The scheme of sample preparation for the Ca corrosion tests is shown in Fig. 2(a), and the WVTR value can be determined by [12]:
The Al2O3 films are applied to miniature LED fabrication. Figures 3(a)-(c) shows the light distribution images for the device without, with 40 nm and 80 nm-thick Al2O3 layers. It is seen that the latter two devices exhibit higher light intensity as indicated by, for instance, the more white-colored regions. The angular light distribution shown in Fig. 3(d) agrees with the result of the enhancement brought by the incorporation of the Al2O3 layer. This is due to that when light is emitted from the MQW, a part of it will move toward the back (DBR) side at random angles. The Al2O3 layer can enhance reflectance at the interface between Al2O3 and the upper thick SiO2 layer. Light that might originally move to the sides of the miniature LED die in the case of no Al2O3 layer will be reflected in advance, and this helps avoid light from escaping from the sides of the device. Thus, the devices with an Al2O3 layer having lower light loss show improved light output at the front side.
The light output power-current-voltage (L-I-V) curves for the miniature LEDs before high temperature and high humidity testing are shown in Fig. 4(a). It is seen that the miniature LEDs with Al2O3 layers exhibit a slightly higher light output power than the device without an Al2O3 layer, agreeing with the observation in Fig. 3. In addition, the EL spectra in Fig. 4(b) indicate that all of the devices have an emission wavelength at around 450 nm.
Reliability tests are carried out for 1000 h in an 85 C/85% RH environment for devices with different Al2O3 barrier layer thicknesses. The traditional device (without the Al2O3 layer) is used for comparison. The optical power degradation of the devices is shown in Fig. 5(a). Before testing, the absolute output powers of the miniature LEDs without, with 20, 40, 60, and 80 nm Al2O3 moisture barrier layers are 101.9, 104.5, 105, 105.6, and 107.2 mW, respectively. The optical power greatly drops by more than half at 200 h and completely fails at around 500 h for the device without Al2O3 barrier layer. With the 20 nm-thick Al2O3 barrier layer, the total degradation reduces and is still around 57%. The devices with Al2O3 layer thicknesses of 40 nm or greater show almost no degradation after 1000 h. This result validates the Al2O3 films used as an internal moisture barrier layer. Figure 5(b) shows the cross-sectional SEM image of the failure device after 1000 h. Cracking or delamination occurs near the top surface of the p-GaN as a result of water vapor intrusion. It is noted that the devices are powered on throughout the reliability tests. If there is no sufficient moisture resistivity, moisture can diffuse into the devices along the SiO2/GaN interface and accumulate to form a high-conductivity layer. The current flow in this case mostly passes through the interfacial moisture-induced layer, and the heat generated at the interface causes moisture expansion, which leads to delamination in regions where the GaN surface possesses defects or where the adhesion between SiO2 and GaN is weak. The internal Al2O3 layer with a thickness of 40 nm or above protects the devices by providing adequate moisture resistance.
To further investigate the effect of the internal Al2O3 barrier layer on device performance, I-V measurements are performed on the devices without and with the 40 nm-thick Al2O3 barrier layer after 200 h of reliability testing. The I-V curves of the devices in Fig. 6(a) are fitted to extract the series resistance and ideality factor by using [21]:
Figure 7(a) shows the semi-logarithmic I-V curves for the devices without and with the 40 nm Al2O3 moisture barrier layer after 200 h of reliability testing. It can be seen that the device without the Al2O3 layer has a significantly higher leakage current. The flatter curve at higher voltages (around 2.5-3 V) also reflects the increased series resistance. In addition, the electrostatic discharge (ESD) characteristics of the devices after 200 h of reliability testing are shown in Fig. 7(b). The survival rate of the device without the Al2O3 layer starts to decrease at −1.5 kV, and reduces to 60% at −2.5 kV. The survival rate drops to 0 at −3.5 kV or higher negative voltages. In contrast, the device with the 40 nm Al2O3 layer has a survival rate of 62% at −5 kV, which drops to 0% at −5.5 kV.
4. Conclusion
The ALD Al2O3 films are prepared as an internal moisture barrier layer in miniature LEDs. The Al2O3 thickness of 40 nm can lead to a water vapor transmission rate of 4.3 × 10−4 g/m2/day, which is two orders of magnitude lower than the 650 nm-thick SiO2 that is the only moisture resistive layer used in traditional devices. Moreover, the devices with Al2O3 thicknesses of 40 nm or greater show almost no degradation after 1000 h of 85°C/85%RH reliable tests, while the traditional device degrades more than half at 200 h and completely fails at 500 h. From the I-V curve analysis and cross-sectional SEM images, the device without an Al2O3 layer suffers from not only an increase in series resistance due to FEL oxidation, but also a rise in ideality factor due to the deteriorated SiO2/GaN interface. Overall, the 40 nm-thick Al2O3 is capable of providing a satisfactory moisture resistance. Compared to traditional devices requiring heavy or complex encapsulation, the method used in the study is helpful for miniature LEDs used in high-humidity environments.
Funding
National Natural Science Foundation of China (61704142); Natural Science Foundation of Fujian Province (2020H0025).
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. H. Masui, H. Yamada, K. Iso, J. S. Speck, S. Nakamura, and S. P. DenBaars, “Non-polar-oriented InGaN light-emitting diodes for liquid-crystal-display backlighting,” J. Soc. Inf. Disp. 16(4), 571–578 (2008). [CrossRef]
2. C. Lin, W. Gui-Gen, Z. Hua-Yu, Z. Fu-Qiang, and H. Jie-Cai, “Progress in Preparation of Patterned Sapphire Substrate for GaN-based Light Emitting Diodes,” J. Inorg. Mater. 27(9), 897–905 (2012). [CrossRef]
3. M. Zhuo, Q. Li, B. Lv, M. Huang, X. Xue, X. Zheng, Y. Lin, B. Liu, X. Liao, Z. Guo, and Z. Chen, “Enhance the Photosynthetic and Color Performances of Multi-Primary White Light-Emitting Diodes for the Indoor Farming,” IEEE Photonics J. 14(6), 1–6 (2022). [CrossRef]
4. Z. Wang, H. Zhu, M. He, Y. Zhou, X. Luo, and N. Zhang, “GAN and Multi-Agent DRL Based Decentralized Traffic Light Signal Control,” IEEE Trans. Veh. Technol. 71(2), 1333–1348 (2022). [CrossRef]
5. Z. Gao, H. Ning, R. Yao, W. Xu, W. Zou, C. Guo, D. Luo, H. Xu, and J. Xiao, “Mini-LED Backlight Technology Progress for Liquid Crystal Display,” Crystals 12(3), 313 (2022). [CrossRef]
6. L.-R. Trevisanello, M. Meneghini, G. Mura, C. Sanna, S. Buso, G. Spiazzi, M. Vanzi, G. Meneghesso, and E. Zanoni, “Thermal stability analysis of high brightness LED during high temperature and electrical aging,” in I. T. Ferguson, N. Narendran, T. Taguchi, and I. E. Ashdown, eds. (2007), p. 666913.
7. Y. C. Hsu, Y. K. Lin, C. C. Tsai, J. H. Kuang, S. B. Huang, H. L. Hu, Y. I. Su, and W. H. Cheng, “Failure Mechanisms Associated with Lens Shape of High-Power LED Modules in Aging Test,” in LEOS 2007 - IEEE Lasers and Electro-Optics Society Annual Meeting Conference Proceedings (2007), pp. 570–571.
8. P. Xu, H. Xiu, L. Yin, P. Wen, Y. Xue, and J. Yang, “The effect of humidity on the degradation mechanisms of GaN-based green laser diodes,” Opt. Laser Technol. 157, 108662 (2023). [CrossRef]
9. X. Luo, B. Wu, and S. Liu, “Effects of Moist Environments on LED Module Reliability,” IEEE Trans. Device Mater. Reliab. 10(2), 182–186 (2010). [CrossRef]
10. J. Gu, H. Xu, B. Peng, Z. Liu, and G. Zhu, “The Study on Thermal Aging Mechanism of Silicone Materials for LED Encapsulation,” in 2021 22nd International Conference on Electronic Packaging Technology (ICEPT) (2021), pp. 1–3.
11. S. Yang, S.-Y. Kwak, J. Jin, J.-S. Kim, Y. Choi, K.-W. Paik, and B.-S. Bae, “Thermally resistant UV-curable epoxy–siloxane hybrid materials for light emitting diode (LED) encapsulation,” J. Mater. Chem. 22(18), 8874–8880 (2012). [CrossRef]
12. H.-I. Lu, D.-P. Tran, C.-K. Lin, and B.-D. To, “Effects of Long-Term Static Bending Deformation on a Barrier Thin Film for Flexible Organic Optoelectronic Devices,” Coatings 8(4), 127 (2018). [CrossRef]
13. S. Kim, S.-H. Lee, I. H. Jo, J. Seo, Y.-E. Yoo, and J. H. Kim, “Influence of growth temperature on dielectric strength of Al2O3 thin films prepared via atomic layer deposition at low temperature,” Sci. Rep. 12(1), 5124 (2022). [CrossRef]
14. H. H. Sønsteby, A. Yanguas-Gil, and J. W. Elam, “Consistency and reproducibility in atomic layer deposition,” J. Vac. Sci. Technol., A 38(2), 020804 (2020). [CrossRef]
15. C.-L. Fan, M.-C. Shang, B.-J. Li, Y.-Z. Lin, S.-J. Wang, W.-D. Lee, and B.-R. Hung, “Teflon/SiO2 Bilayer Passivation for Improving the Electrical Reliability of Oxide TFTs Fabricated Using a New Two-Photomask Self-Alignment Process,” Materials 8(4), 1704–1713 (2015). [CrossRef]
16. C.-H. Hsu, Y.-S. Lin, H.-Y. Wu, X.-Y. Zhang, W.-Y. Wu, S.-Y. Lien, D.-S. Wuu, and Y.-L. Jiang, “Deposition of Silicon-Based Stacked Layers for Flexible Encapsulation of Organic Light Emitting Diodes,” Nanomaterials 9(7), 1053 (2019). [CrossRef]
17. V. Hatty, H. Kahn, and A. H. Heuer, “Fracture Toughness, Fracture Strength, and Stress Corrosion Cracking of Silicon Dioxide Thin Films,” J. Microelectromech. Syst. 17(4), 943–947 (2008). [CrossRef]
18. W. N. Sharpe, J. Pulskamp, D. S. Gianola, C. Eberl, R. G. Polcawich, and R. J. Thompson, “Strain Measurements of Silicon Dioxide Microspecimens by Digital Imaging Processing,” Exp. Mech. 47(5), 649–658 (2007). [CrossRef]
19. H. Wang, Y. Liu, H. Liu, Z. Chen, P. Xiong, X. Xu, F. Chen, K. Li, and Y. Duan, “Effect of Various Oxidants on Reaction Mechanisms, Self-Limiting Natures and Structural Characteristics of Al2O3 Films Grown by Atomic Layer Deposition,” Adv. Mater. Interfaces 5(14), 1701248 (2018). [CrossRef]
20. S. Zhang, W. Xue, and Z. Yu, “Moisture barrier evaluation of SiOx/SiNx stacks on polyimide substrates using electrical calcium test,” Thin Solid Films 580, 101–105 (2015). [CrossRef]
21. Y.-C. Yao, J.-M. Hwang, Z.-P. Yang, J.-Y. Haung, C.-C. Lin, W.-C. Shen, C.-Y. Chou, M.-T. Wang, C.-Y. Huang, C.-Y. Chen, M.-T. Tsai, T.-N. Lin, J.-L. Shen, and Y.-J. Lee, “Enhanced external quantum efficiency in GaN-based vertical-type light-emitting diodes by localized surface plasmons,” Sci. Rep. 6(1), 22659 (2016). [CrossRef]