Open Access
Add to BibSonomyAbstract
This study employed a UV-assisted roller imprinting technique using PDMS soft molds to imprint periodically inverted-pyramid structures on the surface of GaN-based LEDs. The refractive indices of the structures were 1.5, 1.7, and 1.9, which enhanced light output power by 26%, 43%, and 50%, respectively, compared to conventional LEDs. Materials with a greater refractive index indicate a larger critical angle of total internal reflection from the semi-conductor to the imprinted material. Once photons are extracted into the imprinted structure, they are prone to be extracted into the air through the structured surface. The numerical simulation performed using the ray tracing method proved useful for identifying the microstructure with optimal light extraction efficiency. Simulation results showed that LEDs with imprinted structures of varying refractive indices enhance optical efficiency in a manner similar to that demonstrated in these experiments.
©2011 Optical Society of America
1. Introduction
The poor light extraction efficiency (LEE) of GaN-based LEDs is due to the large difference in refractive index and smooth interface between air (n = 1) and GaN (n = 2.5), causing considerable Fresnel loss and total internal reflection (TIR) [1]. The light-emitting efficiency of LEDs can be improved by fabricating micro- or nanostructures inside or on the surface of the chip/substrate. The typical scale of structures applied in LEDs ranges from a few hundred nanometers to a few micrometers, which is roughly on par with the resolution limit of traditional optical lithographic technology. Imprint lithography is a micro/nanoscale technique for forming structures with highly reproducible patterns, making it apply adequately for the LEDs. Some studies have used imprint techniques to build structures of uniform size and periodicity on p-GaN and transparent conductive oxides of LEDs to enhance the LEE and provide homogeneity in the emitted light [2–5]. Methods to modulate the far-field light patterns and improve the light-extraction efficiency of LEDs using sub-microstructures enable most of the photons confined within the LED to be coupled outside the LED, using photonic crystals. Numerous studies have investigated the application of nanoimprinted photonic crystals in LEDs [6–8]. Instead of building micro/nanostructures on the surface of LEDs, patterns fabricated on a sapphire substrate through imprint lithography have also proven effective in enhancing optical efficiency [9, 10].
This study performed UV-assisted roller imprinting to build periodical microstructures on the ITO surface of LEDs. The refractive indices for the microstructures were 1.5, 1.7 and 1.9 respectively. The results show that a imprinting material with a higher refractive index contributes to superior optical efficiency, with enhancement of up to 50% over that of conventional LEDs at 20 mA current injection.
2. Experiments and simulations
2.1 Preparation of conventional LED chips
The GaN-base LED used in this study were grown on c-face (0001), two-inch sapphire substrates using a metal-organic chemical vapor deposition (MOCVD) system. The LED layer structure consisted of a low temperature GaN nucleation layer, a 2.5 μm thick unintentionally doped GaN layer, a 3 μm thick n-type GaN layer, an active region with 10 periods of InGaN/GaN multiple quantum wells (MQWs), a 30 nm thick Mg-doped p-Al0.15Ga0.85N cladding layer (p = 5 × 1017 cm−3), and a 200 nm thick Mg-doped first p+-GaN contact layer (n = 7 × 1017 cm−3). The active layer comprised a 2.3 nm thick InGaN-well layer and a 13 nm thick GaN-barrier layer for the InGaN/GaN MQW LED structures. The fabricated LED sample had indium tin oxide (ITO) evaporated onto it, as a transparent conductive layer. Inductively coupled plasma (ICP) was used to partially etch the LED sample to expose the n-GaN. Optical lithography defined the ITO pattern and wet etching was used to expose the p-GaN layer. Thermal evaporation with rapid thermal annealing was used to create the p- and n-electrodes deposited Cr/Au on the surfaces of p-GaN and n-GaN.
2.2 Roller imprinting process
Instead of imprinting the entire surface at once, the roller and substrate were imprinted progressively, rendering the roller imprint suitable to the fabrication of structures on a large surface [11–14]. In this study, we built a two dimensional hexagonal-close-packed inverted-pyramid array, 3 μm in width, with 5 μm periodicity, and depth of 1.8 μm on the surface of the LED using UV-assisted roller imprinting, rather than etching with ITO or p-GaN. The refractive indices of the imprinted materials were ~1.5 (950 PMMA (polymethyl methacrylate) A4, MicroChem Corp.), ~1.8 (OptiNDEXTM D1, Brewer Science Inc.), and ~1.9 (OptiNDEXTMA54, Brewer Science Inc.). A soft polymer such as PDMS (Polydimethylsiloxane) was used as a mold to duplicate the pattern of the hard mold. PTFE (Polytetrafluoroethylene) was first sprayed onto the PDMS soft mold at 2000 rpm for 30 s before being cured at 200 °C for 10 min. to obtain a lower surface energy, and the imprinting materials were then sprayed onto PDMS mold at 3000 rpm for 40 s. The imprinting materials PMMA, D1 and A54 were cured at 120 °C for 2 min., 200 °C for 2 min., and 220 °C for 2 min., respectively. HMDS (Hexamethyldisilazane) was sprayed onto the LED substrate at 6000 rpm for 30 s before being cured at 110 °C for 1 min. to increase adhesion. Thus the imprinting materials have a better adhesion to the LED substrate and enabling detached from the PDMS mold. The LED sample was fixed at the translation stage using a vacuum holder, and sent for roller imprinting as a translation speed 5 mm/sec. We used a UV lamp as a curing source, the optical energy of which was capable of passing through a quartz cylinder and PDMS mold to focus on the imprinting contact area during the imprinting process. UV-assisted roller imprinting is a continuous room-temperature imprinting process which also can avoid the phenomenon of trapped-air. Figure 1 shows the roller imprinting process and the imprinted results.
2.3 Light extraction analysis using the ray tracing method
To calculate extraction efficiency, this study used Monte Carlo ray tracing to analyze the light propagation of LED chips with surface structure [15]. Ray tracing simulates emission, absorption, refraction, reflection, scattering, and the transmission of light rays. This study required the fabrication of a simulation model for an LED chip with a sapphire of 430 μm, an n-GaN of 4 μm, an MQW of 460 nm, a p-GaN of 200 nm, and an ITO of 240 nm. A two dimensional hexagonal-close-packed inverted-pyramid array was built on the entire LED surface as in the experiment. The chip size was assumed to be 100 × 100 μm2 to save calculation time. The total emission power from MQW was assumed 0.1 W. The real chip with mesa etching and electrodes on p- and n- GaN were neglected. Extraction light was calculated according to the fraction of light collected by a spherical detector outside 5 cm of the LED.
3. Results and discussions
3.1 Influence of microstructures with various refractive indices on LEDs
Figure 2 shows the light output power versus injection current of the LED chips with and without an imprinted structure on the surface. The figure indicates that imprinted structures are conducive to LEE enhancement, with higher light output power corresponding to a greater refractive indices. Table 1 lists the forward voltage and light output power in situations at 20 mA current injection. LEDs with inverted-pyramid structures and refractive indices of 1.5, 1.7, and 1.9 showed light-emitting efficiency exceeding that of conventional LEDs by 26%, 43%, and 50%, respectively. The electrical properties were not compromised, and the measured values of forward voltage were all within an acceptable range of error.
Fig. 2 Light output power of LEDs with and without imprinted structures, as a function of the driving current.
Table 1. Foward voltages and light ouput of conventional LED and imprinted LEDs with various refractive indices at 20 mA injection
Table 2 presents the critical angles of TIR among various materials. In terms of light traveling from the ITO to the imprinted material, it was discovered that a greater refractive index in the imprinted material resulted in a larger critical angle for TIR, increasing the chance of light entering the imprinted material from the ITO. However, for light traveling from the imprinted material to the air, a greater refractive index in the imprinted material resulted in a larger contrast between the refractive indices of the imprinted material and the air, which reduced the chance of light entering the air. With the cancelation of the two phenomena, the difference in refractive index had no apparent impact on the LEE of LEDs when using only one layer of imprinted material.
Table 2. Critical angles of TIR for ITO/imprinted material and imprinted material/air
Figure 3 depicts the various paths of light emitted from an LED through the transparent conductive layer into the imprinted material. In terms of material PMMA, light rays with incident angles exceeding the critical angle of internal reflection (5, 6, 7) are reflected back into the semi-conductor. For material A54, however, a greater critical angle is beneficial to the escape of photons into the imprinted material, whereupon these photons incident onto the surface of the inverted-pyramid at an angle highly likely to be smaller than the critical angle for TIR from imprinted material to air. More rays reach free space through the structured surface, and for this reason, greater refractive indices in the imprinting materials lead to better LEE in LEDs.
3.2 Ray tracing analysis
To confirm that inverted-pyramid structures contribute to LEE, two sets of LEDs were constructed for ray tracing analysis: one with an inverted-pyramid structure and imprinted PMMA material and the other without either. The number of light rays in the activation area was set at 20,000 in the simulations. Figure 4 shows the number of light rays escaping from the imprinted structure into free space far exceeds that of the LED without structures. These results confirm that the inverted-pyramid structure built by imprinting can significantly increase the LEE of LEDs.
Fig. 4 Results of ray tracing analysis: (a) LED without surface structures; (b) LED with inverted-pyramid structures.
Figure 5 compares the simulation results using imprinting structures with different refractive indices. Because the absorption coefficients of the imprinted materials were unknown, several optical simulations with various transmissions were conducted. To shorten simulation time, chip sizes were reduced to 100 × 100 μm2, and the light output power of the activation area was set at 0.1 W with even distribution. The material structures in the simulations were simplified versions of the actual LED structures. Even so, the simulation results indicate increasingly enhanced efficiency in LEDs with greater refractive indices, as shown in Fig. 5. The trend of enhancement was highly similar to that observed in the experiment. Therefore, optical simulations are reliable as analytical instruments prior to experimentation.
Fig. 5 Comparison of experiment and simulation results of enhanced efficiency using various imprinted materials, compared to conventional LEDs.
4. Conclusions
This study investigated the influence of imprinting microstructures on LED surfaces on the enhancement of optical efficiency in LEDs through experimentation and simulation. In our experiment, a UV-assisted roller imprinting technique was employed to distribute the imprinted material on a PDMS soft mold. A roller was used to apply gentle pressure in coordination with UN lighting to build the structures on the surface of the LED over the entire wafer. In the optical simulation, this study used Monte Carlo ray tracing to analyze the light propagation of LED chips with surface structure. In both our experiment and simulation, imprinted structures with greater refractive indices were proven to be more beneficial to the escape of photons from the LED to the imprinting structure prior to entering free space through the structured surface. Using the experiment on imprinting material with a refractive index of 1.9 as an example, light output power was increased to a level 50% greater than that of conventional LEDs.
Acknowledgments
The authors would like to acknowledge funding support from the National Science Council of Taiwan under grant number NSC 100-2628-E-033-002-MY3.
References and links
1. A. David, T. Fujii, B. Moran, S. Nakamura, S. P. DenBaars, C. Weisbuch, and H. Benisty, “Photonic crystal laser lift-off GaN light-emitting diodes,” Appl. Phys. Lett. 88(13), 133514 (2006). [CrossRef]
2. S. J. Chang, C. F. Shen, W. S. Chen, C. T. Kuo, T. K. Ko, S. C. Shei, and J. K. Sheu, “Nitride-based light emitting diodes with indium tin oxide electrode patterned by imprint lithography,” Appl. Phys. Lett. 91(1), 013504 (2007). [CrossRef]
3. H. W. Huang, C. H. Lin, C. C. Yu, B. D. Lee, C. H. Chiu, C. F. Lai, H. C. Kuo, K. M. Leung, T. C. Lu, and S. C. Wang, “Enhanced light output from a nitride-based power chip of green light-emitting diodes with nano-rough surface using nanoimprint lithography,” Nanotechnology 19(18), 185301 (2008). [CrossRef] [PubMed]
4. W. Zhou, G. Min, Z. Song, J. Zhang, Y. Liu, and J. Zhang, “Enhanced efficiency of light emitting diodes with a nano-patterned gallium nitride surface realized by soft UV nanoimprint lithography,” Nanotechnology 21(20), 205304 (2010). [CrossRef] [PubMed]
5. Y. C. Lee, M. J. Ciou, and J. S. Huang, “Light output enhancement for nitride-based light emitting diodes via imprinting lithography using spin-on glass,” Microelectron. Eng. 87(11), 2211–2217 (2010). [CrossRef]
6. B. S. Cheng, C. H. Chiu, K. J. Huang, C. F. Lai, H. C. Kuo, C. H. Lin, T. C. Lu, S. C. Wang, and C. C. Yu, “Enhanced light extraction of InGaN-based green LEDs by nano-imprinted 2D photonic crystal pattern,” Semicond. Sci. Technol. 23(5), 055002 (2008). [CrossRef]
7. Y. Naoi, M. Matsumoto, T. Tan, M. Tohno, S. Sakai, A. Fukano, and S. Tanaka, “GaN-based light emitting diodes with periodic nano-structures on the surface fabricated by nanoimprint lithography technique,” Phys. Status Solidi C 7(7–8), 2154–2156 (2010). [CrossRef]
8. A. Z. Khokhar, K. Parsons, G. Hubbard, F. Rahman, D. S. Macintyre, C. Xiong, D. Massoubre, Z. Gong, N. P. Johnson, R. M. De La Rue, I. M. Watson, E. Gu, M. D. Dawson, S. J. Abbott, M. D. B. Charlton, and M. Tillin, “Nanofabrication of gallium nitride photonic crystal light-emitting diodes,” Microelectron. Eng. 87(11), 2200–2207 (2010). [CrossRef]
9. H. W. Huang, J. K. Huang, S. Y. Kuo, K. Y. Lee, and H. C. Kuo, “High extraction efficiency GaN-based light-emitting diodes on embedded SiO2 nanorod array and nanoscale patterned sapphire substrate,” Appl. Phys. Lett. 96(26), 263115 (2010). [CrossRef]
10. C. C. Kao, Y. K. Su, C. L. Lin, and J. J. Chen, “The aspect ratio effects on the performances of GaN-based light-emitting diodes with nano patterned sapphire substrates,” Appl. Phys. Lett. 97(2), 023111 (2010). [CrossRef]
11. J. J. Lee, S. Y. Park, K. B. Choi, and G. H. Kim, “Nano-scale patterning using the roll typed UV-nanoimprint,” Microelectron. Eng. 85(5–6), 861–865 (2008). [CrossRef]
12. H. Tan, A. Gilbertson, and S. Y. Chou, “Roller nanoimprint lithography,” J. Vac. Sci. Technol. B 16(6), 3926–3928 (1998). [CrossRef]
13. M. T. Gale, “Replication technique for diffractive optical elements,” Microelectron. Eng. 34(3–4), 321–339 (1997). [CrossRef]
14. T. Mäkelä, T. Haatainen, P. Majander, J. Ahopelto, and V. Lambertini, “Continuous double-sided roll-to-roll imprinting of polymer film,” Jpn. J. Appl. Phys. 47(6), 5142–5144 (2008). [CrossRef]
15. T. X. Lee, K. F. Gao, W. T. Chien, and C. C. Sun, “Light extraction analysis of GaN-based light-emitting diodes with surface texture and/or patterned substrate,” Opt. Express 15(11), 6670–6676 (2007). [CrossRef] [PubMed]
