Design of Transparent Multicolor LED Signage with an Oxide-Metal-Oxide Interconnect Electrode
JeeYeon Park, Hyunjee Jeon, Nayeon Park and Geonwook Yoo∗ School of Electronic Engineering, Soongsil University, Seoul 07027, Korea
Chul Jong Han and Min Suk Oh†
Display Research Center, Korea Electronic Technology Institute, Seongnam 13509, Korea
Byeong-Kwon Ju
Display and Nanosystem Laboratory, College of Engineering, Korea University, Seoul 136-713, Korea
Yoon-su Kim
Display Research Center, Korea Electronic Technology Institute, Seongnam 13509, Korea and Display and Nanosystem Laboratory, College of Engineering, Korea University, Seoul 136-713, Korea
(Received 3 April 2020; revised 27 April 2020; accepted 8 May 2020)
A transparent light emitting diode (LED) panel with a 16× 16 multicolor LED array for a signage application is proposed and studied based on experimental and simulation results. As a transparent electrode, oxide/metal/oxide (OMO) trilayers with various Al interlayer thicknesses were fabricated and characterized using an established a simulation model. Using commercial ray tracing optical and SPICE simulation tools, we designed the LED panel and investigated in its optical and electrical properties. Moreover, in order to resolve the waveguide effect and voltage drop issue across the panel, we integrated additional optical structures and various OMO width into the design. The results show that the OMO electrode interconnects and proposed design considerations are pivotal aspects toward achieving transparent LED digital signage applications.
PACS numbers: 81.05.−t, 42.15.Eq
Keywords: LED signage, Transparent, Oxide-metal-oxide, Interconnect DOI: 10.3938/jkps.77.82
I. INTRODUCTION
Transparent conductive oxides (TCOs) have been in great demand for displays based on organic light- emitting diodes and quantum-dot light-emitting diodes, solar cells, and so on [1–4]. In order to be success- fully adopted, TCOs have to possess high conductivity to deliver signals, high optical transparency and elec- trochemical stability. Various TCOs have been investi- gated, and indium-tin oxide (ITO) has been one of the most widely used electrodes. ITO exhibits high trans- mittance, low resistivity, a large bandgap, good etching characteristic, and chemical stability at room tempera- ture [5,6]. However, the limited supply and the high-cost of indium, lack of flexibility because of its brittle nature, and relatively high temperature annealing process for transparency have put constraints on its application as a
∗E-mail: [email protected]
†E-mail: [email protected]
transparent electrode [7]. Therefore, TCO materials such as silver nanowires, graphene, carbon nanotubes, con- ductive polymers and oxide/metal/oxide (OMO) multi- layer electrodes are being investigated [8–16].
On the other hand, OMO trilayer electrodes have been widely studied due to their excellent electrical con- ductivity and optical transparency. Furthermore, when the oxide thickness is decreased, OMO electrodes have better flexibility and thus suffer less damage when bent.
The metal interlayer in the OMO structure suppresses light reflection with the upper and the lower oxide layer, so OMO can get higher optical transmittance than that of a single film [17, 18]. Various combination of OMO structures have been reported, including ITO/Al/ITO, ITO/Ag/ITO, zinc-tin oxide (ZTO)/Ag/ZTO, indium- zinc oxide (IZO)/Ag/IZO, indium-zinc-tin oxide (IZTO)/Ag/IZTO, zinc oxide (ZnO)/Ag/ZnO, and aluminum-zinc oxide (AZO)/Ag/AZO [19–30].
In this work, a transparent light emitting diode (LED) panel with OMO electrode interconnects is proposed
pISSN:0374-4884/eISSN:1976-8524 -82- 2020 The Korean Physical Societyc
Fig. 1. (a) Cross-sectional schematic of the deposited OMO electrode (b) images of the OMOs for Al thicknesses of 3, 5, 7, 10 nm, respectively and (c) transmittance of the OMO electrodes.
for digital signage applications based on experimental and simulation studies. ITO/Al/ITO stacking electrodes with various Al thicknesses are deposited and character- ized optically and electrically; then, its simulation model is established using commercial tools such as LightTools and SPICE by Synopsys. The designed transparent LED panel consists of 16× 16 multicolor LEDs on a glass sub- strate with a size of 480× 480 mm2. Because the waveg- uide effect is anticipated when the LED light sources are encapsulated by a glass layer with relative low refractive index, reflective layers around the perimeter and anti- reflection coating were proposed and investigated based on simulation. Moreover, widths of OMO electrode in- terconnect are adjusted in order to circumvent voltage drop issue in the panel.
II. METHODS
For the fabrication of a transparent OMO electrode on a glass substrate, the ITO (150 nm)-coated glass was cleaned with acetone and isopropyl alcohol (IPA), rinsed with deionized (DI) water and blown dry with N2 gas.
After substrate had been cleaned, Al interlayers with thicknesses of 3, 5, 7 and 10 nm were deposited using thermal evaporation. Then, as an upper transparent conductive oxide, ITO layers with a thicknesses of 150 nm were deposited by using radio frequency (RF, 13.56 MHz) magnetron sputtering (150 W) of the ITO target (In2O3: SnO2= 90 : 10 wt%) in an Ar ambient at room temperature.
The sheet resistance of the OMO electrodes on a glass substrate was measured by using a resistivity meter (MCP-T610, Mitsubishi Chemical Analytech, Japan), and their transmittance at wavelengths from 390 to 740 nm was characterized by using a spectrophotometer (CM-3600d, Konica Minolta, Japan). The optical and the electrical simulations of the OMO electrode and its application for LED pixels and transparent panels were conducted using a ray-tracing optical design tool, Light-
Table 1. Characterized sheet resistance (Rsh) and resis- tivity (ρ) of the deposited OMO electrodes for various Al thicknesses (Thk).
Al (3 m) Al (5 nm) Al (7 nm) Al (10 nm)
Rsh [Ω/] 7.69 7.24 7.72 7.43
ρ [Ω · μm] 1.93 1.81 2.03 1.94
Tools and HSPICE from Synopsys.
III. RESULTS AND DISCUSSION
Figure 1(a) shows a cross-sectional schematic of the deposited ITO/Al/ITO structure with variable Al thick- ness. The electrical and the optical properties of the OMO structure composed of an oxide/metal/oxide tri- layer depend on the thickness of the inserted metal layer and the dielectric constant of the oxide layer. The over- coat oxide layer improves the transmittance of the OMO structure by acting as an anti-reflection film and by re- ducing destructive interference with the metal layer be- neath. Furthermore, it prevents metal oxidation. The deposited OMO films with corresponding Al thicknesses of 3, 5, 7, 10 nm are shown in Fig. 1(b), and their mea- sured optical transmittances are presented in Fig. 1(c), which shows relatively less transparency with increasing Al thickness. The transmittance value at 550 nm was adopted for the optical simulation. The characterized sheet resistance (Rsh) and resistivity (ρ) are summarized in Table 1, and no clear Al thickness dependence primar- ily due to thickness variation and surface roughness was observed.
In order to design a pixel and its array for a trans- parent LED digital signage panel, first we simulated the optical transmittance; then, we matched the calculated Rsh with the experimental results. Figure 2(a) show the
Fig. 2. (a) Schematic illustration of the simulation set-up for using a LightTools by Synopsys Inc. to measure the transmit- tance. Comparisons of the (b) simulated transmittances and the (c) calculated sheet resistances to the experimental results.
Table 2. Measured and simulated transmittances and sheet resistances (Rsh) of the deposited OMO electrodes for various Al thicknesses (Thk).
Al Al Al Al
(3 m) (5 nm) (7 nm) (10 nm) Transmittance Experimental 82 62 62 45
[%] Simulation 86.92 61.14 48.60 33.68 Rsh Experimental 7.69 7.24 7.72 7.43 [Ω/] Calculation 7.64 7.10 7.90 7.49
simulation setup to extract the optical transmittance in the LightTools; the thicknesses of the ITO films and the Al interlayer were the same as the nominal values in the experiment. A light source and receiver were placed on the other side of the OMO film structure, and the trans- mittance of the OMO film structure was calculated by measuring the amount of incident light on the receiver.
In Fig. 2(b), the simulated results are compared with the values measured for various Al thicknesses; similar thick- ness dependences with less than 10% discrepancy were observed. The calculated Rsh with ρ/t is also compared in Fig. 2(c). Instead of monotonically decreasing with in- creasing Al interlayer thickness, some fluctuations were observed due to the thickness variation, interface rough- ness, textured/grain size of the Al film as a result of the thermal deposition process [17,22]. The extracted resis- tivity corresponding to the fabricated OMO electrodes was used to calculate Rsh. As summarized in Table 2, the simulated optical and electrical properties of the OMO film structure were in accordance with the experimental results, and the OMO electrode simulated in LightTools was used to design a single pixel and 16× 16 array panel for LED signage.
After verifying the simulation results, we designed 16 × 16 transparent LED signage using the simulated OMO electrode; the panel size was set to 480 × 480 mm2 with a pixel pitch of 30 × 30 mm2. A multi- color LED (STF0A36C, Seoul Semiconductor, Inc., Ko-
Fig. 3. (a) Cross-sectional schematic and (b) top-view il- lustration of the proposed pixel.
Fig. 4. Simulation results: (a) coordinate system of the designed panel (b) light-rays emitted from the panel (c) sim- ulated forward illuminance in the unit of Lux and (d) simu- lated forward illuminance in color.
rea) was used, and its optical and electrical characteris- tics were adopted for the simulation. Figure 3(a) shows a cross-sectional schematic of the designed pixel. Instead of transparent plastic films such PET and PI, a glass substrate with a refractive index of 1.5 and a transmit-
Table 3. Summarized optical and electrical characteris- tics of the multicolor LED (STF0A36C, Seoul Semiconductor, Inc.).
Parameter Color Value Parameter Color Value
Dominant R 625 Forward R 20
λ G 525 Current G 20
[nm] B 457 [mA] B 20
Luminous R 19 Forward R 2.0
Flux G 48 Voltage G 2.7
[lm] B 37 [V] B 3.0
Viewing Angle [◦] R,G,B 120
tance of 91.9% was used as the substrate and cover layer.
Figure 3(b) illustrates a top view of the pixel, and the reader should note that the widths of the OMO electrode are not constant. This was intended to compensate for the voltage drop occurring through OMO interconnects with fixed widths, which will be discussed next.
The optical characteristics of the designed panel were simulated first. Figure 4(a) shows the designed 16 × 16 multicolor LED array panel drawn on the coordinate plane of LightTools, and the center of the panel was set as the origin of the coordinates. The verified OMO elec- trode, silicon resin, Cu as a bonding metal, and Pb as a solder were used. Optical parameters, including the refractive indices and the transmittances of the materi- als, the emission spectrum, and the optical intensity of the LED were measured. A small light-source surface of red (R), green (G) and blue (B) were integrated onto the multicolor LED module. Therefore, 256 surface light sources of R, G, B were simulated in total. The emis- sion spectrum, light intensities, and viewing angles for each R, G and B light sources, as summarized in Ta- ble 3, were considered. Then, the optical characteristics of the designed LED panel were simulated by measuring the amount of light incident on the receiver in front of the panel. Figure 4(b) shows light rays emitted from one of the LEDs and their propagation within the designed panel. Figures 4(c) and (d) show simulated results for the illuminance in the unit of Lux and the color, respec- tively, for which the data were collected in front of the panel; the light rays coming from each pixel of the 16× 16 array and R, G and B color are generated from the sub-surfaces of a single LED.
As anticipated from Fig. 4(b), we observed the waveg- uide effect because of the total internal reflection when the LED is encapsulated with a glass layer with a rela- tive low refractive index. Therefore, a significant amount of light emitted from the LED module is confined in the panel, resultig in edge-emission of the panel, not top- emission. Therefore, two additional structures were pro- posed and simulated to suppress the waveguide effect and, thus, to enhance top-emission: i) reflective layers around the perimeter and ii) an anti-reflection coating on the glass. As a reflective layer, a 0.5-mm-thick alu-
Fig. 5. Luminous flux improvement of the designed panel with reflective Al layers around the perimeter and an anti- reflection coating layer.
Fig. 6. (a) I-V curves of the R, G, B diodes generated by using the HSPICE model and their changes with the OMO interconnect. (b) Current flowing on the R, G, B diodes at different pixel locations.
minum (Al) layer with a reflectivity of 83% was attached around the panel, and the amount of top-emission was increased by about 53%. Furthermore, when an AR coat- ing was applied on both sides of the upper glass, the light reflected in the opposite direction of the receiver was re- duced and the luminous flux was enhanced by about 5%.
Figure 5 shows luminous flux vs. Al interlayer thick-
ness of the OMO electrode, and the flux of top-emission is seen to increase slightly increased with increasing Al thickness.
In order to predict the electrical characteristics and solve the aforementioned voltage-drop issue in the LED signage panel, we conducted a HSPICE simulation on the designed panel [31]. The current-voltage (I-V) curves in the specification sheet of the LED (STF0A36C) were generated using a simple diode model, and the results are presented in Fig. 6(a). The U model was used to simulate the OMO interconnect by taking its geometry and resistivity into account. The shift of I-V curves at the first pixel in a row is also shown in Fig. 6(a) due to resistance of the OMO with a 10-nm-thick Al inter- layer. Then, a pixel at a different location in a row was compared. The first, 8th, and 15th pixels were simu- lated by changing the length and the width of the OMO interconnect and the distance between the OMO inter- connects. Figure 6(b) compares the current flowing on R, G, B diodes at a fixed forward voltage for different lo- cations. Of note is that a constant current can be applied throughout the row, which is directly related to the lu- minance of the LEDs, and the proposed width variation across the panel is quite effective.
IV. CONCLUSION
In summary, a transparent 16 × 16 multicolor LED panel with a size of 480 × 480 mm2 was designed us- ing OMO electrode interconnects. The fabricated OMO electrodes with Al interlayer thicknesses of 3, 5, 7, and 10 nm were characterized, and their optical and electri- cal simulation models were established. By integrating a reflective layer of 0.5-mm-thick aluminum (Al) around the panel and by applying on AR coating on both sides of the upper glass, we were able to increase the amount of top emissions by about 53% and 5%, respectively. Fur- thermore, by modifying the widths of OMO electrode interconnect depending on the pixel locations, we were able to solve the issue of voltage drop across the panel.
Through the experimental and the simulation studies, the feasibility of using OMO electrodes for transparent LED signage was explored, and the proposed additional structure may be effective for resolving several design issues.
ACKNOWLEDGMENTS
This research was funded by the National Research Foundation of Korea (NRF-2019M3C1B909055), and
partly supported by the Industy technology R&D pro- gram of MOTIE/KEIT (10063316, Development of core technology of tiling active matrix panel, aiming 200-inch UHD class display). The simulation tools were supported by the IDEC (IC Design Education Center) Program and CYFEM, Korea.
REFERENCES
[1] S. Sutthana, N. Hongsith and S. Choopun, Curr. Appl.
Phys.10, 813 (2010).
[2] S. Song et al., Vacuum85, 39 (2010).
[3] R. D. Sahu, S. Lin and J. Huang, Appl. Surf. Sci.253, 4886 (2007).
[4] Y. Park, K. Choi, H. Kim and J. Kang, Electrochem.
Solid-State Lett.13, J39 (2010).
[5] M. Bender et al., Thin Solid Films326, 67 (1998).
[6] M. Shakiba, A. Kosarian and E. Farshidi, J. Mater. Sci.:
Mater. Electron.28, 787 (2017).
[7] R. D. Cairns et al., Appl. Phys. Lett.76, 1425 (2000).
[8] C. Li et al., Appl. Surf. Sci.285, 505 (2013).
[9] J. Kim et al., ACS Nano5, 3222 (2011).
[10] C. Liu and X. Yu, Nanoscale Res. Lett.6, 75 (2011).
[11] D. Leem et al., Adv. Mater.23, 4371 (2011).
[12] H. Park et al., Nano Lett.14, 5148 (2014).
[13] K. Kim et al., Nature457, 706 (2009).
[14] C. P. Chen et al., ACS Nano4, 4403 (2010).
[15] Y. Xia, L. Sun and J. Ouyang, Adv. Mater. 24, 2436 (2012).
[16] W. Cao, Y. Li, E. Wrzesniewski and T. W. Hammond, Org. Electron.13, 2221 (2012).
[17] C. Guill´en and J. Herrero, Thin Solid Films 520, 1 (2011).
[18] W. Wei et al., J. Mater. Sci. Technol.33, 1107 (2017) [19] J. Yu and S. Lee, Trans. Electr. Electron. Mater.19, 212
(2018).
[20] K. Choi et al., Appl. Phys. Lett.92, 223302 (2008).
[21] R. D. Sahu, S. Lin and J. Huang, Appl. Surf. Sci.252, 7509 (2006).
[22] M. Theuring, M. Vehse, K. von Maydell and C. Agert, Thin Solid Films558, 294 (2014).
[23] T. Winkler et al., Org. Electron.12, 1612 (2011).
[24] N. E. Cho, P. Moon, C. Kim and I. Yun, Expert Syst.
Appl.39, 8885 (2012).
[25] K. Choi, J. Kim, Y. Lee and H. Kim, Thin Solid Films 341, 152 (1999).
[26] H. Kim, K. Seo and H. Kim, Trans. Mater. Res. Soc.
Jpn.40, 401 (2015).
[27] M. Girtan, Sol. Energy Mater. Sol. Cells100, 153 (2012).
[28] S. Lee, Y. Park and T. Seong, J. Alloys Compd. 776, 960 (2019).
[29] S. Lee, E. Cho and S. Kwon, Appl. Surf. Sci. 487, 990 (2019).
[30] C. Wu, RSC Adv.8, 11862 (2018).
[31] Synopsys Inc., CA, USA, HSPICE and RF Command Reference (2007).
