Research Article www.acsami.org
Optical Enhancement in Optoelectronic Devices Using Refractive Index Grading Layers Illhwan Lee,† Jae Yong Park,† Seungo Gim,† Kisoo Kim,†,‡ Sang-Hwan Cho,§ Chung Sock Choi,§ Seung-Yong Song,§ and Jong-Lam Lee*,† †
Department of Materials Science and Engineering, Division of Advanced Materials Science, Pohang University of Science and Technology (POSTECH), Pohang, Gyeongbuk 790-784, Korea § Samsung Display Company, Limited, Giheung-Gu, Yongin, Gyeonggi 446-711, Korea S Supporting Information *
ABSTRACT: We enhanced the optical transmittance of a multilayer barrier film by inserting a refractive index grading layer (RIGL). The result indicates that the Fresnel reflection, induced by the difference of refractive indices between SixNy and SiO2, is reduced by the RIGL. To eliminate the Fresnel reflection while maintaining high transmittance, the optimized design of grading structures with the RIGL was conducted using an optical simulator. With the RIGL, we achieved averaged transmittance in the visible wavelength region by 89.6%. It is found that the optimized grading structure inserting the multilayer barrier film has a higher optical transmittance (89.6%) in the visible region than that of a no grading sample (82.6%). Furthermore, luminance is enhanced by 14.5% (from 10 190 to 11 670 cd m−2 at 30 mA cm−2) when the grading structure is applied to organic light-emitting diodes. Finally, the results offer new opportunities in development of multilayer barrier films, which assist industrialization of very cost-effective flexible organic electronic devices. KEYWORDS: Fresnel reflection, thin-film passivation, OLED, refractive index grading, multilayer barrier film
■
structure with a multilayer barrier film composed of SixNy [high refractive index (n)] and SiO2 (low n). Two kinds of SixNy/ SiO2 stacking structures, no grading (NGR) (Figure 1a) and grading (GR) (Figure 1b) structures, on the OLED were considered in the paper. The NGR structures usually exhibit high Fresnel reflection loss as a result of differences in the refractive index between SixNy and SiO2 layers (Figure 1a). A number of attempts have been conducted to suppress the Fresnel reflection for anti-reflection, but it led to increasing processing cost and time.19−22 In this work, we demonstrate a highly transparent multilayer barrier film for OLEDs by inserting refractive index grading layers (RIGLs), which has an intermediate value between n values of SixNy and SiO2 layers. Optical simulation was conducted to determine optimized n and thickness of RIGLs for a highly transparent multilayer barrier film. The n of the SixNy layer could change by controlling the fraction of reaction gases, such as the ratio of SiH4/NH3,23,24 resulting in the control of the SixNy composition. From this, the n of the SixNy film changed from 1.74 to 2.03. Nitrogen-rich SixNy (n = 1.74) showed n values between SiO2 (n < 1.5) and SixNy (n > 1.9),
INTRODUCTION Flexible organic light-emitting diodes (FOLEDs) have attracted considerable attention as a result of their merits, such as being lightweight, inexpensive, and flexible and having the possibly of roll-to-roll processing; therefore, they have potential applications as next-generation display and lighting devices.1−5 Typically, FOLEDs are based on flexible plastic substrates, such as polyethylene terephthalate (PET), polydimethylsiloane (PDMS), polyimide (PI), and poly(methyl methacrylate) (PMMA).6−8 Unfortunately, most plastic substrates have a high water vapor transmission rate (WVTR), which produces degradation, such as dark spots and edge shrinkage over active areas, when they were exposed to the atmosphere.9,10 Especially, to obtain stable FOLEDs, a WVTR on the order of 10−6 g/m2/day is mandatory.11,12 To overcome these problems, multilayer barrier films are considered to be indispensable for the OLEDs.13,14 The multilayer barrier film is composed of barrier and sandwiching layers. Water permeation through a multilayer barrier film is controlled by defects of the barrier layer and diffusion path induced by sandwiching layers.15 Silicon nitride (SixNy) and silicon oxide (SiO2 ), deposited by plasma-enhanced chemical vapor deposition (PECVD), are representative barrier16,17 and sandwiching18 materials. Although the combination of SixNy and SiO2 has good barrier properties, use of them encounters some drawbacks. Figure 1 shows that the schematic OLED © XXXX American Chemical Society
Received: November 19, 2015 Accepted: January 22, 2016
A
DOI: 10.1021/acsami.5b11219 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Research Article
ACS Applied Materials & Interfaces
Figure 1. Schematic OLED structure with a multilayer barrier film composed of high and low n layers. Transmittance of (a) NGR and (b) GR structures. Insets show variation of n as a function of the distance from the substrate. dimensional (2D) finite-difference time-domain (FDTD) simulations (Fullwave, Synopsys, Inc.) to explain the effect of the RIGL. Boundary conditions for the FDTD simulation is set to be a perfect matched layer (PML) to avoid the reflected electromagnetic (EM) wave at the edge of the structure. The light is emitted by a dipole source with a wavelength of 510 nm. The simulations were performed during the light propagation over a distance of 12 μm to reach the steady state. When it reached the steady state, the electric field distribution does not changed over 0.1% in calculation domain as a function of the simulation time. The spatial cross-sectional electric field was obtained at the steady state using a discrete Fourier transform (DFT) monitor.
leading to a reduction of the Fresnel reflection at the SixNy/ SiO2 interface.
■
EXPERIMENTAL SECTION
Fabrication of Multilayer Barrier Films and OLEDs. A glass with a thickness of 500 μm was used as the substrate and was ultrasonically cleaned in acetone, isopropyl alcohol, and deionized (DI) water each for 3 min and then dried with blowing N2 gas. To form the GR structure, SiO2 and SixNy were deposited on the cleaned glass substrates using PECVD. The partial pressures of gases, such as SiH4, NH3, and He, were individually controlled to tune the n of SixNy. To quantify how the GR structure affects OLED properties, bottom emission OLEDs were fabricated. In the OLED, indium tin oxide (ITO)-coated glass was used as the substrate. The substrates were loaded into the chamber, and then 3 nm WO3, 190 nm WO3-doped 4,4′-N,N′-dicarbazole-biphenyl (CBP), 5 nm intrinsic CBP, 20 nm Tris[2-phenylpyridinato-C2,N]iridium(III) (Irppy3)-doped CBP, 70 nm 1,3,5-Tris(1-phenyl-1H-benzimidazol-2-yl)benzene (TPBi), 1 nm LiF, and 100 nm Al were deposited in sequence as a hole injection layer (HIL), hole transport layer (HTL), emissive layer (EML), electron-transporting layer (ETL), electron-injection layer (EIL), and reflective cathode, respectively.25 Characterization Methods. The total (intergrating sphere) transmittance of samples was recorded at wavelengths of 400−700 nm using an Agilent Technologies Cary 4000 ultraviolet−visible (UV−vis) spectrometer. The current density−voltage characteristics of the OLEDs were measured using a Keithley 2400 source meter in a N2 ambient condition. Luminance characteristics of OLEDs were measured by a calibrated Si photodiode in a N2 ambient condition. Electroluminescence (EL) spectra of OLEDs were measured using a spectrometer (Spectral Products, SM240) with an optical fiber. To investigate the angular dependencies of samples, we measured the luminance as a function of the viewing angle (from −70° to 70°) for the bottom emission OLED. The bare glass, GR, and NGR samples are attached on the back side of the substrate of the OLED using a sample holder, allowing for the samples to cling to the OLED without any adhesive. For reliable values, each experiment was repeated at least twice. Simulation. Optical transmittance calculations of NGR and GR structures were performed using commercial software (Essential Macleod, Thin Film Center, Inc.) based on the characteristic matrix method. Angular-dependent transmittance was calculated by rigorous coupled wave analysis (RCWA) software (DiffractMOD 3.1, Synopsys, Inc.). In angular-dependent transmittance, RCWA simulation was used for calculation under an unpolarized plane wave with an equal contribution from both polarization states [50% transverse magnetic (TM) and 50% transverse electric (TE)].26,27 The angle of incident was tuned from 0° to 85° in intervals of 5°. We employed two-
■
RESULTS AND DISCUSSION To study the effect of the GR structure on the anti-reflection property, optical simulation was performed. Commercial software (Essential Macleod, Thin Film Center, Inc.) based on the characteristic matrix method was employed for the optical analysis involving a multilayer structure. The optical constants of SiO2 and SixNy were measured by spectroscopic ellipsometry. To determine the optimum thicknesses of SixNy (N) and SiO2 (O), we calculated two kinds of NON structures. One is to use a SiO2 layer with a fixed thickness [270 nm (3λ/ 4n)], sandwiched with SixNy having a different thickness (Figure S1a of the Supporting Information). The other is to used two SixNy layers with a constant thickness [330 nm (5λ/ 4n)], but the SiO2 thickness is changed (Figure S1b of the Supporting Information). No distinct change in the average transmittance was found with the thickness of SixNy (Figure S1c of the Supporting Information) and SiO2 (Figure S1d of the Supporting Information). Therefore, thicknesses of SixNy and SiO2 layers are to be 330 and 270 nm, respectively, where λ is the incident wavelength of 510 nm (peak center of the emission spectrum of green OLED). Figure 2a shows the optical simulation results for NGR structures with an increase of a set of SiO2 and SixNy layers. The oscillation in transmittance was observed as a set of SiO2 and SixNy layers were deposited onto the SixNy layer. As the number of the layer set increases, the difference between the wavelengths of the peak becomes short but the amplitude of the oscillation increased, leading to the decrease in the average transmittance. This wave fluctuation could be explained in terms of the Fresnel reflection induced by the difference between refractive indices of SixNy and SiO2. Such oscillation induced by the Fresnel reflection could be suppressed by inserting the RIGL between SixNy and SiO2 layers. A schematic diagram of the GR B
DOI: 10.1021/acsami.5b11219 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Research Article
ACS Applied Materials & Interfaces
etry (Figure 3a). The refractive index linearly increases as a function of the ratio of SiH4/NH3. It revealed that the refractive
Figure 2. Calculated transmittance as a function of the wavelength from 400 to 800 nm. (a) NGR structure (N = 330 nm, and O = 270 nm) and (b) GR structure (N = 200 nm, O = 180 nm, and RIGL = 73 nm). We label each spectrum with the corresponding stack structure of SixNy (N)/SiO2 (O). (c) Calculated transmittance averaged from a wavelength of 400−700 nm as a function of the number of SixNy and SiO2 layers.
Figure 3. (a) Experimental value of the refractive index for SixNy versus the ratio of SiH4/NH3. Simulated and measured transmittances of two kinds of structures: (b) NONON (N = 330 nm, and O = 270 nm) and (c) NONON with RIGL (N = 200 nm, O = 180 nm, and RIGL = 73 nm) as a function of the wavelength from 400 to 800 nm.
structure, including the RIGL, is shown in the inset of Figure 2b. In the design of the RIGL, an incident wavelength of 510 nm was used and the n of the RIGL (nr) is chosen by (n1n2)0.5, where n1 and n2 are the refractive indices of dielectric layers. The thickness of the RIGL is typically λ/4nr. Thus, the thickness of the RIGL can be calculated to be 73 nm. Also, thicknesses of SixNy and SiO2 layers are 200 nm (3λ/4n) and 180 nm (2λ/4n) because of the maintenance of the total thickness of a multilayer barrier film. As a result, the total thicknesses of NON (inset of Figure 2a) and NON with RIGL (inset of Figure 2b) samples are 930 and 889 nm, respectively. In the case of GR structures, the transmittance is unchanged after using RIGLs. This is due to the suppression of the Fresnel reflection by RIGLs (Figure 2b). The averaged transmittance (Tavg) of the NGR structure using one layer was 82.16% in the visible wavelength region (400−700 nm), but it gradually decreased to ∼65.47% at the sample with seven layers. However, in GR structures with one layer, the Tavg was 89.93%, but no distinct changes (Tavg = 89.93−89.22%) were found with increasing the number of layers (Figure 2c). We determined the refractive indices of SixNy films with various ratios of SiH4/NH3 using the spectroscopic ellipsom-
index could be effectively controlled by the ratio of SiH4/ NH3.23,24 The simulated and measured transmittances of the NGR structure show low transmittance in a specific wavelength, such as 510 and 720 nm wavelength regions, as a result of the Fresnel reflection (Figure 3b), leading to the decrease of Tavg to 82.6%. This provides evidence that the NGR structure composed of high and low n layers induced the Fresnel reflection depending upon the incident wavelength. As the RIGL was sandwiched in the NGR layer, such oscillation disappeared and the magnitude of transmittance is independent of the wavelength, as shown in Figure 3c. Tavg was measured to be 89.6%, which is similar to that measured in bare glass (91.3%). To study the effect of the RIGL, we employed three kinds of substrates, such as the bare glass, NGR, and GR samples, on the back side of bottom green OLEDs (Figure S2a of the Supporting Information). Figure 4a shows the normalized EL spectra of the samples. No changes in the peak wavelength at 510 nm and the shape of spectrum were found in both the bare glass and GR samples. It means that the GR structure does not C
DOI: 10.1021/acsami.5b11219 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Research Article
ACS Applied Materials & Interfaces
Figure 4. (a) Normalized EL spectra of OLEDs on bare glass, NGR (NONON; N = 330 nm, and O = 270 nm), and GR (NONON with RIRL; N = 200 nm, O = 180 nm, and RIGL = 73 nm) samples at normal direction. (b) Luminance−current density characteristics of OLEDs on bare glass, NGR, and GR samples. (c) Measured angular emission pattern of OLEDs with bare glass, NGR, and GR samples (from −70° to 70°; injected J was set at a constant 30 mA/cm2).
modify the trasmitted light or distort the emitted color. However, as a result of the Fresnel reflection induced by the difference of n between SixNy and SiO2, the NGR sample exhibits a narrower EL spectrum than that of the bare glass sample. Figure 4b shows the current density (J)−luminance (L) characteristics of OLEDs with the different types of substrate. The L at J = 30 mA/cm2 decreased from 11 580 to 10 190 cd/ m2 when the bare glass sample was replaced with the NGR sample. It means that the decrease of L originates from the Fresnel reflection (Figure 3b), leading to a loss of optical efficiency. However, it was slightly increased to 11 670 cd/m2 when the GR sample was employed. This result indicates that the GR structure is effective in reducing the Fresnel reflection and, thereby, improving the light out-coupling efficiency of OLEDs. To accurately investigate the angular dependencies of the GR structure, we measured the luminance as a function of the viewing angle (from −70° to 70°) for the bottom OLED (J = 30 mA/cm2; Figure 4c). Figure S2a of the Supporting Information shows that the schematic explanation of the measurement for the angular distribution. We used the same OLED device that is not degraded during the measurement of angular distribution (Figure S2b of the Supporting Information). The angular distribution of the bare glass sample exhibits Lambertian emission, and no distinct changes were found with the GR sample. However, the angular dependence based on the NGR sample is quite different from the bare and GR samples. The luminance of the NGR sample is decreased with the viewing angles. No changes in both the peak wavelength of 510 nm and the shape of the spectrum with the viewing angle were found with bare glass and GR samples (panels a and c of Figure S3 of the Supporting Information). It means that the GR structure did not modify the trasmitted light or distort the emitted color. However, the EL spectrum of the NGR structure with the viewing angle is different from that of the bare glass sample (Figure S3b of the Supporting Information). These
differences in the EL spectrum of the NGR structure indicate that reflection occurred in a specific angle. Panels a−c of Figure 5 present the calculated angle-resolved transmittance of the bare glass, NGR, and GR structures under an unpolarized plane wave. The unpolarized light transmittance is calculated by taking average of transmittances in TEpolarized and TM-polarized light (Figure S4 of the Supporting Information). Figure 5d represents the difference of average transmittance compared to bare galss as a function of the incident angle. In bare glass and GR samples, a high angular transmittance was found at the visible wavelength region (panels a and c of Figure 5). The angular transmittance (0− 85°) averaged in the visible wavelength region (400−700 nm) was calculated to be 82.76% in the GR sample. This value is higher than the bare glass value (80.96%). However, in the NGR sample, a low transmittance appeared at a specific wavelength region near 510 nm (Figure 5b). The angular transmittance averaged in the visible region was calculated to be 72.17%. Such a low transmittance originates from the Fresnel reflection as a result of high/low refractive index stacks. By introduction of the RIGLs, the Fresnel reflection is suppressed, leading to a broad increase in the angular transmittance. To investigate the response of the structures for OLEDs, 2D FDTD simulations were empolyed under dipole sources. The cross-section square of the electric field distributions of the NGR and GR samples is shown in panels e and f of Figure 5. The horizontal and vertical dipoles for the NGR simulated are given in panels a and b of Figure S5 of the Supporting Information, and those for the GR are in panels c and d of Figure S5 of the Supporting Information. Both the horizontal and vertical dipoles were incoherently combined to obtain the electric field distributions in panels e and f of Figure 5. The field intensity at the center of the interface between SiO2 and RIGL in the GR structure (Figure 5f) was enhanced up to 25.7% compared to that of the interface between SiO2 and SixNy in the D
DOI: 10.1021/acsami.5b11219 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Research Article
ACS Applied Materials & Interfaces
Figure 5. Contour plot of the simulated angular-resolved transmittance of (a) bare glass, (b) NGR (NONON; N = 330 nm, and O = 270 nm), and (c) GR (NONON with RIGL; N = 200 nm, O = 180 nm, and RIGL = 73 nm) samples under an unpolarized plane wave. (d) Calculated angular Tavg (400−700 nm) difference (ΔT) between bare glass and NGR or GR samples as a function of the incident angle. Cross-sectional square of the electric field distributions of the (e) NGR and (f) GR structures. (g) Schematic diagram of the simulated OLED structure (inset) and calculated normalized extraction efficiency. Cross-sectional square of the electric field distributions of the (h) NGR and (i) GR structures.
NGR structure (Figure 5e). The enhanced field intensity originated from the RIGL, which played a role in reducing the Fresnel reflection at the interface between SixNy and SiO2. To study the effect of the GR structure on the OLED and the electric field distribution for OLEDs, 2D FDTD simulations were performed under dipole sources. Figure 5g shows the schematic diagram of the simulated OLED structure (inset). Boundary conditions for the FDTD simulation are set to a perfect matched layer (PML) to avoid the reflected electromagnetic wave at the edge of the structure, except the bottom layer, which has been set to a perfect electric conductor (PEC) to approximate the metallic mirror at the bottom of the OLED. Extraction efficiency is calculated from light output power normalized by the power of the excitation source. A crosssectional DFT monitor was used to obtain the spatial electric field distribution. We employed three kinds of structures, such as the no barrier film (bare), NGR (NONON), and GR (NONON with RIGL) structures, on the thin Ag cathode of top green OLEDs. The extraction efficiency was increased to 21.8% when the NGR (0.888) structure was replaced with the GR (1.082) structure (Figure 5g). This result indicates that the GR structure is effective in reducing the Fresnel reflection and, thereby, improving the light out-coupling efficiency of OLEDs.
Simulation results (21.8% increase) agreed well with experimental results (14.5% increase; Figure 4b). The cross-sectional square of the electric field distribution of the NGR and GR structures on the top green OLED is shown in panels h and i of Figure 5. The field intensity at the center of the interface between air and the GR structure was enhanced in comparison to that of the interface between air and the NGR structure. The enhanced field intensity originated from the RIGL, which played a role in reducing the Fresnel reflection. However, the amplitude of the electric field inside the NGR structure increases, because the electromagnetic wave inside the NGR structure is captured by the internal reflection.
■
CONCLUSION In summary, enhancement of out-coupling efficiency of the multilayer barrier film on OLEDs can be achieved using RIGLs to suppress the Fresnel reflection. This improvement was achieved by controlling the n of the SixNy film for RIGLs. RCWA and 2D FDTD simulation results showed the elimination of the Fresnel reflection in the interface between SixNy and SiO2 and enhanced the transmittance. With the RIGLs, the following was successfully achieved: (1) effectively reduced Fresnel reflection loss at the SixNy−SiO2 interface, E
DOI: 10.1021/acsami.5b11219 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Research Article
ACS Applied Materials & Interfaces
(7) Lee, H.; Choi, S.; Kim, K.; Yun, J.; Jung, D.; Park, S.; Park, J. A Novel Solution-Stamping Process for Preparation of a Highly Conductive Aluminum Thin Film. Adv. Mater. 2011, 23, 5524−5528. (8) Han, T.; Lee, Y.; Choi, M.; Woo, S.; Bae, S.; Hong, B.; Ahn, J.; Lee, T. Extremely Efficient Flexible Organic Ligth-Emitting Diodes with Modified Graphene Anode. Nat. Photonics 2012, 6, 105−110. (9) Park, S.; Yun, W.; Kim, L.; Park, S.; Kim, S.; Park, C. Inorganic/ Organic Multilayer Passivation Incorporating Alternating Stacks of Organic/Inorganic Multilayers for Long-Term Air-Stable Organic Light-Emitting Diodes. Org. Electron. 2013, 14, 3385−3391. (10) Schaer, M.; Nüesch, F.; Berner, D.; Leo, W.; Zuppiroli, L. Water Vapor and Oxygen Degradation Mechanism in Organic Light Emitting Diodes. Adv. Funct. Mater. 2001, 11, 116−121. (11) Burrows, P.; Graff, G.; Gross, M.; Martin, P.; Hall, M.; Mast, E.; Bonham, C.; Bennett, W.; Michalski, L.; Weaver, M.; Brown, J.; Fogarty, D.; Sapochak, L. Gas Permeation and Lifetime Tests No Polymer-Based Barrier Coatings. Proc. SPIE 2000, 4105, 75−83. (12) Yun, W.; Jang, J.; Nam, S.; Kim, L.; Seo, S.; Park, C. Thermally Evaporated SiO Thin Films as a Versatile Interlayer for Plasma-Based OLED Passivation. ACS Appl. Mater. Interfaces 2012, 4, 3247−3253. (13) Yong-Qiang, Y.; Yu, D. Optimization of Al2O3 Films Deposited by ALD at Low Termperatures for OLED Encapsulation. J. Phys. Chem. C 2014, 118, 18783−18787. (14) Xiao, W.; Yu, D.; Bo, S.; Qiang, Y.; Dan, Y.; Ping, C.; Hui, D.; Yi, Z. The Improvement of Thin Film Barrier Performances of Organic-Inorganic Hybrid Nanolaminates Employing a Low-Temperature MLD/ALD Method. RSC Adv. 2014, 4, 43850−43856. (15) Graff, G.; Williford, R.; Burrows, P. Mechanisms of Vapor Permeation through Multilayer Barrier Films: Lag Time versue Euilibrium Permeation. J. Appl. Phys. 2004, 96, 1840−1849. (16) Kim, H.; Kim, M.; Kang, J.; Kim, J.; Yi, M. High-Quality ThinFilm Passivation by Catalyzer-Enhanced Chemical Vapor Deposition for Organic Light-Emitting Diodes. Appl. Phys. Lett. 2007, 90, 013502. (17) Wuu, D.; Lo, W.; Chiang, C.; Lin, H.; Chang, L.; Horng, R.; Huang, C.; Gao, Y. Water and Oxygen Permeation of Silicon Nitride Films Prepared by Plasma-Enhaced Chemical Vapor Deposition. Surf. Coat. Technol. 2005, 198, 114−117. (18) Lifka, H.; van Esch, H. A.; Rosink, J. J. W. M. 50.3: Thin Film Encapsulation of OLED Displays with a NONON Stack. Dig. Tech. Pap. - Soc. Inf. Disp. Int. Symp. 2004, 35, 1384−1387. (19) Fan, Y.; Chen, J.; Ma, D. Enhancement of Light Extraction of Green Top-Emitting Organic Light-Emitting Diodes with Refractive Index Gradually Changed Coupling Layers. Org. Electron. 2013, 14, 3234−3239. (20) Minot, M. Single-Layer, Gradient Refractive Index Antireflection Films Effective from 0.35 to 2.5 μ. J. Opt. Soc. Am. 1976, 66, 515−519. (21) Nagel, H.; Aberle, A.; Hezel, R. Optimised Antireflection Coatings for Planar Silicon Solar Cells using Remote PECVD Silicon Nitride and Porous Silicon Dioxide. Prog. Photovoltaics 1999, 7, 245− 260. (22) Hong, K.; Yu, H.; Lee, I.; Kim, K.; Kim, S.; Lee, J.-L. Enhaced Light Out-Coupling of Organic Light-Emitting Diodes: Spontaneously Formed Nanofacet-Structured MgO as a Refractive Index Modulation Layer. Adv. Mater. 2010, 22, 4890−4894. (23) Nayar, P. Refractive Index Control of Silicon Nitride Films Prepared by Radio-Frequency Reactive Sputtering. J. Vac. Sci. Technol., A 2002, 20, 2137−2139. (24) Mattsson, K. Plasma-Enhanced Growth, Composition, and Refractive Index of Silicon Oxy-Nitride Films. J. Appl. Phys. 1995, 77, 6616−6623. (25) Lee, I.; Park, J.; Gim, S.; Ham, J.; Son, J.; Lee, J.-L. Spontaneously Formed Nanopatterns on Polymer Films for Flexible Organic Light-Emitting Diodes. Small 2015, 11, 4480−4484. (26) Crouse, D.; Keshavareddy, P. Polarization Independent Enhanced Optical Transmission in One-Dimensional Gratings and Device Applications. Opt. Express 2007, 15, 1415−1427. (27) An, C.; Cho, C.; Choi, J.; Park, J.; Jin, M.; Lee, J.; Jung, H. Highly Efficient Top-Illuminated Flexible Polymer Solar Cells with a
leading to high transparency in the visible wavelength region (400−700 nm) by 89.6% (NGR structure: 82.6%), and (2) simple and low-cost process without an additional deposition method. Consequently, by replacing NGR to GR structure, we can reduce the internal reflection, resulting in enhanced luminance of 14.5% (from 10 190 to 11 670 cd m−2 at 30 mA cm−2) without changes in both the emission peak center of 510 nm and the shape of the spectrum. Finally, the results offer new opportunities in development of the multilayer barrier films, which assist industrialization of very cost-effective FOLEDs.
■
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsami.5b11219. Calculated transmittance spectrum of NON structures (Figure S1), schematic explanation of the measurement of the angular distribution (Figure S2a), current density− voltage characteristics of OLED (Figure S2b), normalized EL spectrum (Figure S3), contour plot of the simulated angle-resolved transmittance under TE-polarized and TM-polarized light (Figure S4), and crosssectional square of the electric field distributions (Figure S5) (PDF)
■
AUTHOR INFORMATION
Corresponding Author
*Telephone: 82-54-279-2152. E-mail:
[email protected]. Present Address ‡
Kisoo Kim: POSCO, Pohang, Gyeongbuk 37859, Korea.
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work was supported by the National Research Foundation of Korea (NRF) Grants funded by the Korea Government (MSIP, NRF-2013R1A2A2A01069237 and NRF2014H1A2A1021655, Global Ph.D. Fellowship Program). The authors also acknowledge support from Samsung Display Co., Ltd.
■
REFERENCES
(1) Forrest, S. The Path to Ubiquitous and Low-Cost Organic Electronic Applications on Plastic. Nature 2004, 428, 911−918. (2) Sandstrom, A.; Dam, H.; Krebs, F.; Edman, L. Ambient Fabrication of Flexible and Large-Area Organic Light-Emitting Devices using Slot-Die Coating. Nat. Commun. 2012, 3, 1002. (3) Kang, H.; Jung, S.; Jeong, S.; Kim, G.; Lee, K. Polymer-Metal Hybrid Transparent Electrodes for Flexible Electronics. Nat. Commun. 2015, 6, 6503. (4) Ok, K.; Kim, J.; Park, S.; Kim, Y.; Lee, C.; Hong, S.; Kwak, M.; Kim, N.; Han, C.; Kim, J. Ultra-Thin and Smooth Transparent Electrode for Flexible and Leakage-Free Organic Light-Emitting Diodes. Sci. Rep. 2015, 5, 9464. (5) Yu, Z.; Zhang, Q.; Li, L.; Chen, Qi; Niu, X.; Liu, J.; Pei, Q. Highly Flexible Silver Nanowire Electrodes for Shape-Memory Polymer LightEmitting Diodes. Adv. Mater. 2011, 23, 664−668. (6) Kim, S.; Cho, H.; Hong, K.; Son, J.; Kim, K.; Koo, B.; Kim, S.; Lee, J. − L. Design of Red, Green, Blue Transparent Electrodes for Flexible Optical Devices. Opt. Express 2014, 22, A1257−A1269. F
DOI: 10.1021/acsami.5b11219 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Research Article
ACS Applied Materials & Interfaces Nanopatterned 3D Microresonant Cavity. Small 2014, 10, 1278− 1283.
G
DOI: 10.1021/acsami.5b11219 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
