In this study, indium-tin oxide (ITO)/Al-doped zinc oxide (AZO) composite films were fabricated by pulsed laser deposition and used as transparent contact layers (TCLs) in GaN-based blue light emitting diodes (LEDs). The ITO/AZO TCLs were composed of the thin ITO (50 nm) films and AZO films with various thicknesses from 200 to 1000 nm. Conventional LED with ITO (200 nm) TCL prepared by E-beam evaporation was fabricated and characterized for comparison. From the transmittance spectra, the ITO/AZO films exhibited high transparency above 90% at wavelength of 465 nm. The sheet resistance of ITO/AZO TCL decreased as the AZO thickness increased, which could be attributed to the increase in a carrier concentration, leading to a decrease in the forward bias of LED. The LEDs with ITO/AZO composite TCLs showed better light extraction as compared to LED with ITO TCL in compliance with simulation. When an injection current of 20 mA was applied, the output power for LEDs fabricated with ITO/AZO TCLs had 45%, 63%, and 71% enhancement as compared with those fabricated using ITO (200 nm) TCL for the AZO thicknesses of 200, 460, and 1000 nm, respectively.
© 2011 OSA
High efficiency GaN-based light-emitting diodes (LEDs) are used in a large range of potential applications such as full-color displays, traffic signals, automobiles, solid-state lighting, backlights of liquid-crystal displays, and so on [1–3]. However, the external efficiency of LEDs is mainly limited by the difficulty of light extraction due to the trap of light emitting inside the LED by the total internal reflection at the semiconductor-air interface, which resulted from the refractive index difference of GaN-based materials and air . In order to solve the problem, several methods including surface roughening , photonic crystals , patterned sapphire substrates , and changing the chip shape  have been proposed. Additionally, the replacement of poorly transparent (60~70% in the visible region) Ni/Au alloy contact layer by a transparent contact layer (TCL) on p-GaN can also improve the light extraction. Recently, indium tin oxide (ITO) has emerged as one of most promising materials for the TCL in GaN-based LEDs due to its low resistivity and high transparency in the range of visible wavelength [9,10]. Unfortunately, indium is a rare metal with high price, and the low thermal stability of ITO leads to a poor device reliability. Therefore, many indium-free ZnO-based materials have increasingly studied as the alternative TCL in high efficiency LEDs due to their electrical and optical properties similar to ITO, better thermal stability, and smaller lattice mismatch to GaN [11–13]. ZnO-based thin films have been grown by several techniques such as sputtering , molecular beam epitaxy , metal–organic chemical vapor deposition , atomic layer deposition , E-beam evaporation , and pulsed laser deposition (PLD) . Among these techniques, PLD is the ideal production technology for the growth of epitaxial ZnO-based thin films because it can keep the consistent composition of the original material which is preserved in the interaction and the oxygen plasma created by the pulsed laser is very energetic. Our previous research has reported that the 200-nm-thick Al-doped ZnO (AZO) films with highly (002)-oriented structure grown by PLD at the growth temperature of 100 °C exhibits low resistivity of 233 μΩ-cm and high transmittance above 90% in both visible and UV regions . It indicates that the film has great potential in the TCL on LEDs.
In this work, the ITO (50 nm)/AZO (200-1000 nm) films, which were grown by PLD, were characterized and used as the TCLs on GaN blue LEDs to improve light extraction efficiency. In the ITO/AZO TCLs, the 50-nm-thick ITO film is served as an ohmic contact layer with p-GaN layer. On the other hand, the AZO films with various thicknesses are applied for uniform current spreading, high transparent and thick window layer to achieve the highly efficient LEDs. Moreover, the conventional LED with ITO (200 nm) TCL grown by E-beam evaporation was fabricated as contrasted sample.
In this study, two types of TCLs were fabricated. First, the ITO (200 nm) TCL was deposited by E-beam evaporation. The target material applied for this experiment was an ITO pellet which consisted of 95 wt % In2O3 and 5 wt % SnO2. The working pressure and substrate temperature were controlled to be 5 × 10−2 Pa and 250 °C, respectively. Second, the ITO (50 nm)/AZO (200-1000 nm) composite TCLs were grown by PLD (PLD/MBE-2000, PVD products) from the stoichiometric ITO and AZO ceramic targets (99.999% purity) of 3 in. diameter. The ITO target has the same composition used in E-beam evaporation, and the composition of AZO target is 98 wt % ZnO and 2 wt % Al2O3. The growth temperature of ITO and AZO films prepared by PLD was fixed at 100 °C. In the PLD system, the substrates were kept at 80 mm away from the targets, and the substrates were heated using a resistive heater during films growth. A KrF excimer laser (λ = 248 nm) was employed as an ablation source with 1 Hz repetition rate and the energy ñuence was 600 mJ/pulse. As the base pressure was smaller than 5 × 10−8 Torr, the Ar was introduced into the chamber, and the working pressure of films growth was fixed at 1 × 10−2 Torr. The ITO (50 nm)/AZO (200-1000 nm) TCLs were designated as ITO/AZO (200-1000 nm) TCLs.
The LEDs were prepared on c-plane sapphire using low-pressure metal-organic chemical vapor deposition. The LED structures consisted of a 3-μm-thick layer of u-GaN, a 2-μm-thick layer of n-type GaN:Si (n-GaN), 6 pairs of InGaN/GaN multiple quantum well (MQW) active layers, a 100-nm-thick p-type AlGaN layer and a 0.2-μm-thick p-type GaN:Mg layer. For the device process, a mesa pattern of LED sample was defined with the size of 12 × 24 mm2. Next, the ITO film was fabricated by PLD or E-beam evaporation on the LED. The ITO layer was partially etched in a diluted HCl solution, and the LED structure was etched by means of the inductively coupled plasma with Cl2/Ar gas to expose the n-GaN layer. Then, the AZO film was grown by PLD at the growth temperature of 100 °C. Finally, the Cr/Au (20 nm/200 nm) was deposited by E-beam evaporation as the n- and p-pad electrodes. The sheet resistances of the samples were measured using a conventional four-point-probe method. The transmittances of TCLs were determined from the N&K analyzer (model: 1280, N and K Tech.). The contact resistances between TCLs and p-GaN layer were evaluated by the transmission-line model (TLM) method . The current-voltage (I-V) characteristics of LED samples were determined using a semiconductor parameter analyzer (Keithley, 2400 sourcemeter), and the output powers of LEDs were measured with a calibrated integrating sphere. Trace-Pro software (commercial light-simulation software based on geometrical optics) is used to simulate the photon trajectories of LEDs with various TCLs to quantitatively analyze light-extraction efficiency. The films thickness and surface roughness were measured by the atomic force microscope (AFM).
3. Result and Discussion
Figure 1 exhibits the AFM images of ITO (200 nm) deposited by e-beam evaporation and ITO (50 nm) deposited by PLD. The root mean square (rms) roughness of the ITO (200 nm) and ITO (50 nm) were 2.6 and 3.8 nm, respectively. It can be expected that the optical and electrical properties of the LED samples were influenced by the surface roughness.
The transmission spectra with wavelength from 200 to 1000 nm of ITO (200 and 1000 nm) and ITO/AZO (200, 460, and 1000 nm) films were measured as shown in Fig. 2 . The ITO (1000 nm) film shows the transmittance about 70-75% at the wavelength between 400 and 700 nm. It indicates that the thicker ITO layer is not suitable for TCL in LED. Except for 1000-nm-thick ITO film, all films exhibit the transmittance higher than 85% at wavelength range of 400-700 nm. Transparent thick films with a few microns in thickness, such as 1000-nm-thick ITO in this work, usually have lower optical transmittance because of light scattering from longer optical path. However, we can found that the thicker AZO films still have high transmittance, even for the thickness increased to 1000 nm. This can be attributed to the strong absorption in UV-blue range and a smooth decrease in the extinction coefficient with wavelength for the AZO films, resulting from the band gap of 3.37 eV and its polycrystalline/partially amorphous nature. We also found that the absorption of ITO (50 nm)/AZO (200-1000 nm) almost does not change with the AZO thickness. It can be explained by the previous study by B. Z. Dong et al . In the previous study, the carrier concentration of thinnest AZO film (15 nm) is the smallest one, thus the occupied states in valence band are less and cause the small optical band gap. As the film thickness increases to larger than 56 nm, the carrier concentration becomes stable, resulting in the stable absorption and optical band gap. Therefore, we speculate that the carrier concentration of AZO film has become stable as the thickness is above 200 nm, leading to the stable absorption in our work. The transmittance at wavelength of 465 nm and sheet resistance of ITO/AZO (200-1000 nm), and the contact resistance between ITO/AZO (200-1000 nm) TCL and p-GaN layer evaluated from TLM were summarized in Table 1 . It is found that the transmittances of ITO/AZO films with various AZO thicknesses of 200, 460, and 1000 nm are 94, 96, and 90%, respectively. The results of sheet resistance show that the resistance of ITO/AZO films decreased with an increase of AZO film thickness from 200 to 1000 nm. The TLM results exhibit that the contact resistances between ITO/AZO TCL and p-GaN are 1.33, 3.00 × 10−1, and 1.32 × 10−3 Ω-cm2 as the AZO thicknesses are 200, 460, and 1000 nm, respectively. It indicates that the contact resistance also reduced as the thickness of AZO layer increased, which results from the decrease in the sheet resistance due to the increment in carrier concentration.
Trace-Pro Simulation results are exhibited in Figs. 3(a) –3(d). Here, a 60 mW power (10000 light rays) was assumed to emit randomly from the MQW active layer. The simulated results reveal that the fluxes and rays are 17.16 mW and 1360 rays, 23.11 mW and 1373 rays, 25.57 mW and 1504 rays, and 33.51 mW and 1899 rays for the LEDs with various TCLs of ITO (200 nm), ITO/AZO (200 nm), ITO/AZO (460 nm) and ITO/AZO (1000 nm), respectively. In our simulation, the epilayer absorption and the effect from the silver cup in the epoxy lamp are not considered; however, the results also can demonstrate the influence of optical scattering at LEDs with various TCLs. The simulation results indicate that the LEDs with ITO/AZO (200-1000 nm) TCLs lead to more light extraction than that with ITO (200 nm) TCL. Moreover, the light extraction of LED with ITO/AZO TCL is increased with an increase of AZO thickness. It can be explained that the thicker AZO window layer can improve the light extraction by allowing additional light to escape through the side facets. Additionally, the larger surface roughness could decrease the refractive index, which improves light extraction of LED. From the results of surface roughness in Fig. 1, it reveals that the more light extraction in the LEDs fabricated with ITO (50 nm)/AZO (200-1000 nm) TCLs than that fabricated with ITO (200 nm) is partially attributed to the larger surface roughness in ITO (50 nm) film.
Figure 4 shows the current–voltage characteristics of the InGaN/GaN MQW LEDs with the ITO (200 nm) and ITO/AZO (200-1000 nm) TCLs measured at room temperature. It exhibits that the turn-on voltages of LEDs with ITO/AZO (200-1000 nm) TCLs are just a little higher than that with ITO (200 nm) TCL, indicating that the thin ITO film with thickness of 50 nm plays a role of ohmic contact layer successfully. According to the past research, the GaN-based LEDs with ZnO-based TCLs grown on the p-GaN directly leaded to the poor electrical properties [23,24]. This was attributed to the higher resistivity of ZnO-based films, which caused a schottky contact to p-GaN. However, as a thin ITO layer with lower resistivity was inserted between ZnO-based film and p-GaN, the turn-on voltage was reduced due to the formation of ohmic contact . On the other hand, the larger surface roughness in ITO (50 nm) as shown in Fig. 1 would cause a reduction in the contact area and an increment in contact resistance. This also leads to the higher turn-on voltage in the LEDs fabricated with ITO (50 nm)/AZO (200-1000 nm) TCLs than that with ITO (200 nm) TCL. We also found that the turn-on voltage of LEDs fabricated with ITO/AZO films decreased with increasing the AZO thickness. This could be attributed to the decrease in sheet resistance of the ITO/AZO TCL from 28.1 to 3.5 Ω/□ as the AZO thickness increased from 200 to 1000 nm, as shown in Table 1. The thicker AZO layer leads to a lower series resistance along the path of current spreading from the p-pad electrode, therefore, the lower turn-on voltage can be obtained.
Figures 5(a) –5(d) show the optical microscope images of these blue LEDs with various TCLs of ITO (200 nm) and ITO/AZO (200-1000 nm) operated at a low drive current of 0.5 mA. It can be found that the brightness of LED fabricated with ITO/AZO (200 nm) TCL is almost the same as that with ITO (200 nm) TCL. For the LED with ITO/AZO (460-1000 nm) TCL, the higher brightness can be observed, especially as the AZO thickness is 1000 nm. It reveals that the thick AZO window layer can used to provide current spreading and light extraction of LED sample efficiently, resulting in a great enhancement in brightness.
The performance of light output was measured by calibrating an integrating sphere with a Si photodiode on the package device, which can collect the light emitted in any direction from the LED. Figure 6 plots the light output powers as a function of injection current for the LEDs fabricated with 200-nm-thick ITO and ITO/AZO (200-1000 nm) TCLs. The light output powers of the LEDs with ITO/AZO (200-1000 nm) TCLs are higher than that with ITO (200 nm) TCL in the whole measured currents from 0 to 100 mA. The output powers for the LEDs fabricated with ITO/AZO (200 nm), ITO/AZO (460 nm) and ITO/AZO (1000 nm) TCLs had 45%, 63%, and 71% enhancement compared to that fabricated with ITO (200 nm) TCL at a 20 mA operating current, respectively. The improvement of light output power could be due to the increase of light extraction efficiency by substituting ITO/AZO (200-1000 nm) TCLs for ITO (200 nm) TCL. Furthermore, the increased light output power of LEDs with ITO/AZO TCLs as the thickness of AZO film increased from 200 to 1000 nm can result from the increment of light extraction through sidewalls in thicker AZO layer. Moreover, the improvements in brightness and optical output power of the LEDs with ITO (50 nm)/AZO (200-1000 nm) TCLs were also affected by the larger surface roughness in the ITO films, as discussed in the Fig. 3.
In conclusion, the ITO/AZO films with various AZO thicknesses of 200-1000 nm were deposited by pulsed laser deposition at 100 °C and utilized as TCLs for the fabrication of bright InGaN/GaN MQWs blue LEDs. The ITO/AZO (1000 nm) film shows high transmittance of 90% at wavelength of 465 nm and low contact resistance of 1.32 × 10−3 Ω-cm2 to p-GaN. Trace-Pro simulation results reveal that the LEDs with ITO/AZO (200-1000 nm) TCLs bring out more light extraction compared to that with ITO (200 nm) TCL. Moreover, the light extraction of LED with ITO/AZO TCL is increased as the AZO thickness increased. This could be explained that the light extraction is improved through the side facets in thicker AZO film. Compared to the conventional LED with ITO (200 nm) TCL, it was found that the light output power of the LEDs fabricated with ITO/AZO TCLs can be improved, especially as the thickness of AZO layer is 1000 nm.
This work was supported by the National Science Council (Taiwan) under Grant Nos. 98-2623-E-005-002-IT and 100-2623-E-005-001-IT and the Ministry of Economic Affairs under grant No. 99-EC-17-A-07-S1-158.
References and links
1. S. Nakamura, T. Mukai, and M. Senoh, “Candela-class high-brightness InGaN/AlGaN double-heterostructure blue-light-emitting diodes,” Appl. Phys. Lett. 64(13), 1687–1689 (1994). [CrossRef]
2. H. X. Jiang, S. X. Jin, J. Li, J. Shakya, and J. Y. Lin, “III-nitride blue microdisplays,” Appl. Phys. Lett. 78(9), 1303–1305 (2001). [CrossRef]
3. K. M. Uang, S. J. Wang, S. L. Chen, C. K. Wu, S. C. Chang, T. M. Chen, and B. W. Liou, “High-power GaN-based light-emitting diodes with transparent indium zinc oxide films,” Jpn. J. Appl. Phys. 44(4BNo. 4B), 2516–2519 (2005). [CrossRef]
4. H. K. Cho, J. Jang, J.-H. Choi, J. Choi, J. Kim, J. S. Lee, B. Lee, Y. H. Choe, K.-D. Lee, S. H. Kim, K. Lee, S.-K. Kim, and Y.-H. Lee, “Light extraction enhancement from nano-imprinted photonic crystal GaN-based blue light-emitting diodes,” Opt. Express 14(19), 8654–8660 (2006). [CrossRef] [PubMed]
5. T. Fujii, Y. Gao, R. Sharma, E. L. Hu, S. P. DenBaars, and S. Nakamura, “Increase in the extraction efficiency of GaN-based light-emitting diodes via surface roughening,” Appl. Phys. Lett. 84(6), 855–857 (2004). [CrossRef]
6. T. N. Oder, K. H. Kim, J. Y. Lin, and H. X. Jiang, “III-nitride blue and ultraviolet photonic crystal light emitting diodes,” Appl. Phys. Lett. 84(4), 466–468 (2004). [CrossRef]
7. C. C. Wang, H. Ku, C. C. Liu, K. K. Chong, C. I. Hung, Y. H. Wang, and M. P. Houng, “Enhancement of the light output performance for GaN-based light-emitting diodes by bottom pillar structure,” Appl. Phys. Lett. 91(12), 121109 (2007). [CrossRef]
8. J. Y. Kim, M. K. Kwon, J. P. Kim, and S. J. Park, “Enhanced light extraction from triangular GaN-based light-emitting diodes,” IEEE Photon. Technol. Lett. 19(23), 1865–1867 (2007). [CrossRef]
9. T. Margalith, O. Buchinsky, D. A. Cohen, A. C. Abare, M. Hansen, S. P. DenBaars, and L. A. Coldren, “Indium tin oxide contacts to gallium nitride optoelectronic devices,” Appl. Phys. Lett. 74(26), 3930–3932 (1999). [CrossRef]
10. S. M. Pan, R. C. Tu, Y. M. Fan, R. C. Yeh, and J. T. Hsu, “Enhanced output power of InGaN–GaN light-emitting diodes with high-transparency nickel-oxide–indium-tin-oxide ohmic contacts,” IEEE Photon. Technol. Lett. 15(5), 646–648 (2003). [CrossRef]
11. T. Y. Park, Y. S. Choi, J. W. Kang, J. H. Jeong, S. J. Park, D. M. Jeon, J. W. Kim, and Y. C. Kim, “Enhanced optical power and low forward voltage of GaN-based light-emitting diodes with Ga-doped ZnO transparent conducting layer,” Appl. Phys. Lett. 96(5), 051124 (2010). [CrossRef]
12. B. Y. Oh, M. C. Jeong, W. Lee, and J. M. Myoung, “Properties of transparent conductive ZnO:Al films prepared by co-sputtering,” J. Cryst. Growth 274(3-4), 453–457 (2005). [CrossRef]
13. T. Minami, T. Miyata, and T. Yamamoto, “Stability of transparent conducting oxide films for use at high temperatures,” J. Vac. Sci. Technol. A 17(4), 1822–1826 (1999). [CrossRef]
14. C. L. Jia, K. M. Wang, X. L. Wang, X. J. Zhang, and F. Lu, “Formation of c-axis oriented ZnO optical waveguides by radio-frequency magnetron sputtering,” Opt. Express 13(13), 5093–5099 (2005). [CrossRef] [PubMed]
15. S. M. Yang, S. K. Han, J. W. Lee, J. H. Kim, J. G. Kim, S. K. Hong, J. Y. Lee, J. H. Song, S. I. Hong, J. S. Park, and T. Yao, “Microstructural investigation of ZnO films grown on (1 1 1) Si substrates by plasma-assisted molecular beam epitaxy,” J. Cryst. Growth 312(9), 1557–1562 (2010). [CrossRef]
16. N. Yoshii, A. Nakamura, S. Hosaka, and J. Temmyo, “Investigation of morphology and crystallinity of ZnO crystal formed by side-flow-type MOCVD,” J. Electrochem. Soc. 156(7), K117–K120 (2009). [CrossRef]
17. M. C. Lin, Y. J. Chang, M. J. Chen, and C. J. Chu, “Characteristics of Zr-doped ZnO thin films grown by atomic layer deposition,” J. Electrochem. Soc. 158(6), D395–D398 (2011). [CrossRef]
18. areA. Mahmood, N. Ahmed, Q. Raza, T. Muhammad Khan, M. Mehmood, M. M. Hassan, and N. Mahmood, “Effect of thermal annealing on the structural and optical properties of ZnO thin films deposited by the reactive e-beam evaporation technique,” Phys. Scr. 82(6), 065801 (2010). [CrossRef]
20. S. P. Liu, D. S. Wuu, S. L. Ou, Y. C. Fu, P. R. Lin, M. T. Hung, and R. H. Horng, “Highly ultraviolet-transparent ZnO:Al conducting layers by pulsed laser deposition,” J. Electrochem. Soc. 158(5), K127–K130 (2011). [CrossRef]
21. G. K. Reeves and H. B. Harrison, “Obtaining the specific contact resistance from transmission line model measurements,” IEEE Electron Device Lett. 3(5), 111–113 (1982). [CrossRef]
22. B. Z. Dong, G. J. Fang, J.-F. Wang, W.-J. Guan, and X.-Z. Zhao, “Effect of thickness on structural, electrical, and optical properties of ZnO:Al films deposited by pulsed laser deposition,” J. Appl. Phys. 101(3), 033713 (2007). [CrossRef]
23. J. K. Sheu, Y. S. Lu, M. L. Lee, W. C. Lai, C. H. Kuo, and C. J. Tun, “Enhanced efficiency of GaN-based light-emitting diodes with periodic textured Ga-doped ZnO transparent contact layer,” Appl. Phys. Lett. 90(26), 263511 (2007). [CrossRef]
24. J. K. Sheu, M. L. Lee, Y. S. Lu, and K. W. Shu, “Ga-doped ZnO transparent conductive oxide films applied to GaN-based light-emitting diodes for improving light extraction efficiency,” IEEE J. Quantum Electron. 44(12), 1211–1218 (2008). [CrossRef]