APPLIED PHYSICS LETTERS 88, 113501 共2006兲
Organic oxide/Al composite cathode in efficient polymer light-emitting diodes Tzung-Fang Guo,a兲 Fuh-Shun Yang, and Zen-Jay Tsai Institute of Electro-Optical Science and Engineering, Advanced Optoelectronic Technology Center, National Cheng Kung University, Tainan, Taiwan 701, Republic of China
Ten-Chin Wen and Sung-Nien Hsieh Department of Chemical Engineering, National Cheng Kung University, Tainan, Taiwan 701, Republic of China
Yaw-Shyan Fu Department of Environment and Energy, National University of Tainan, Tainan, Taiwan 700, Republic of China
Chia-Tin Chung Chi Mei Optoelectronics Corporation, Tainan Science-Based Industrial Park, Taiwan, 741, Republic of China
共Received 21 June 2005; accepted 13 February 2006; published online 13 March 2006兲 This work presents the fabrication of efficient polymer light-emitting diodes 共PLEDs兲 by thermally evaporating a salt-free neutral organic-oxide buffer layer onto the surface of the electroluminescent film in a vacuum before the device cathode, made of Al—rather than the low work function metals, such as Ca or Ba—is deposited. The electroluminescence 共EL兲 efficiency of phenyl-substituted poly共para-phenylene vinylene兲 copolymer-based PLEDs with an organic oxide/Al composite cathode, reaches 8.86 cd/ A, which is markedly higher than those, 5.26 cd/ A and 0.11 cd/ A, of devices with Ca/ Al and Al cathodes, respectively. The device performance is improved by the formation of a specific organic oxide/Al complex at the cathode interface during the deposition of Al, facilitating the injection of electrons and eliminating the metal-induced quenching sites of luminescence in the EL layer near the recombination region. © 2006 American Institute of Physics. 关DOI: 10.1063/1.2183808兴 In 1990, Burroughes et al.1 reported to use aluminum 共Al兲 as a device cathode for fabricating polymer lightemitting diodes 共PLEDs兲. The work applied a thin film of conjugated polymer, poly共phenylene-vinylene兲 共PPV兲, sandwiched between the indium-tin-oxide 共ITO兲/glass substrate and the Al electrode. However, the high barrier between the work function of Al and the lowest unoccupied molecular orbital 共LUMO兲 level of PPV causes the injection of electrons through the Al cathode into the PPV layer to be inefficient and dissimilar to that of the holes from the ITO anode side. The imbalance between holes and electrons is responsible for the poor electroluminescence 共EL兲 efficiencies of such devices. Accordingly, low work function metals, such as calcium 共Ca兲,2 or the placement of a buffer layer of lithium fluoride 共LiF兲 共⬍1 nm兲 at the interface with the Al electrode,3–5 are now the commonly used cathodes, supporting the balanced injection of holes and electrons and greatly improving the performance of PLEDs. Many research groups that have blended ionic surfactants into EL layers, or placed a thin layer of ionomers or organic salts that contain lithium 共Li兲 or Ca ions at the interface of cathode, have shown that the EL efficiencies of PLEDs with the high work function Al cathode can reach or exceed those of the level of the devices in which a conventional Ca/ Al or LiF / Al cathode is used.6–9 The improvement in the performance of the device follows from the interaction between additive metal ions with the EL layer near the cathode interface, even though the role of salt ions remains cona兲
Author to whom all correspondence should be addressed; electronic mail:
[email protected]
troversial. Nevertheless, Park et al.10 recently found that a nanolayer of insulated polymers placed at the interface between the EL film and the Al electrode can be used to manipulate the emission of electrons from the Al cathode. Deng et al.11 and Niu et al.12 showed that efficient PLEDs can be fabricated by blending poly共ethylene glycol兲 共PEG兲 into the EL layer or by casting a buffer layer of nonionic PEG-based surfactants on the EL layer with the Al cathode. Additionally, our earlier study reported the considerable enhancement in device performance achieved by simply spincasting an insulated, salt-free, and ultrathin interfacial layer of poly共ethylene oxide兲 polymer on the EL layer with the Al electrode.13 The specific interaction of Al with ethylene oxide groups, 共−CH2CH2O − 兲n, during the deposition of the Al cathode is presumed to be essential to improve the device performance. Salt ions need not to be added to the cathode buffer layer. Although spincoating is a convenient method for preparing the poly共ethylene oxide兲 buffer layer, the issue of the compatibility between the solvent and the organic layers cannot be ignored in the construction of PLEDs with multilayer structures. Besides, the spin-coating process cannot be used for organic light-emitting diodes 共LEDs兲 that are based on small organic molecules. This work develops a process for thermally evaporating an organic-oxide polymer layer with a low molecular weight, onto the surface of the EL film in a high vacuum, before the Al electrode is deposited. This process also enables the thickness and the functional groups of the buffer layer to be carefully controlled. An organic oxide/Al composite electrode, rather than an electrode formed from metals with low work functions, can be used as
0003-6951/2006/88共11兲/113501/3/$23.00 88, 113501-1 © 2006 American Institute of Physics Downloaded 22 Oct 2009 to 140.116.208.44. Redistribution subject to AIP license or copyright; see http://apl.aip.org/apl/copyright.jsp
113501-2
Guo et al.
FIG. 1. I-L-V curves of the devices with Al共쎲兲, PEGDE共12 Å兲 / Al共䉭兲, PEGDE共25 Å兲 / Al共䊏兲, PEGDE共50 Å兲 / Al共⽧兲, and PEGDE共75 Å兲 / Al共䉮兲 as the device cathode, respectively. The inset shows the results of the photovoltaic measurement for devices with Al and PEGDE/Al cathodes.
a device cathode in the fabrication of highly efficient PLEDs. The device configuration herein is comprised of ITO/ glass substrate as the anode, poly共3,4-ethylenedioxythiophene兲:polystyrenesulfonate 共PEDOT:PSS Bayer Corp. 4083兲 as the hole transport layer, “high-yellow” phenylsubstituted poly共para-phenylene vinylene兲 copolymer 共HY-PPV兲 共EL emission centered at 560 nm兲 共Ref. 14兲 as the light-emissive layer, organic-oxide polymer as the interface buffer layer, and the metal cathode electrode. The organicoxide film was prepared by thermally evaporating a thin polymer layer 共12, 25, 50, and 75 Å兲 of poly共ethylene glycol兲 dimethyl ether 共PEGDE兲 共Aldrich, Mn ⬃2000兲 onto the surface of the HY-PPV film inside a vacuum chamber 共10−6 Torr兲. The metal electrode was then evaporated on the substrates without breaking the vacuum. The active pixel area of the device was 0.06 cm2. The details of the fabrication procedure and the current-brightness-voltage 共I-L-V兲 measurement can be found elsewhere.13,15 The interfacial reaction between the cathode and the EL layer of the device was studied by x-ray photoelectron spectroscopy 共XPS兲. XPS measurements were made using a VG ESCALAB 210 spectrometer equipped with a magnesium K␣ source, at a photon energy of 1253.6 eV. The photovoltaic measurement was performed under the illumination supplied by a Thermo Oriel 150W solar simulator 共AM 1.5G兲. Figure 1 plots the I-L-V curves for devices with Al and Al electrodes with organic-oxide buffer layers of various thicknesses. The EL intensity and luminous efficiency of an ITO/PEDOT:PSS/HY-PPV/Al device biased at 8.50 V were 374 cd/ m2 and 0.11 cd/ A, respectively. The low EL is attributable to the inefficient injection of electrons because the injection barrier between the work function of Al and the LUMO level of HY-PPV was high. However, when an additional organic-oxide layer was placed at the cathode interface, the EL intensity of the ITO/PEDOT:PSS/HY-PPV/ PEGDE 共25 Å兲 / Al device exceeded 42 000 cd/ m2 at a bias of 8.50 V. The light turn-on voltage was reduced to 2.50 V; the corresponding value of the device with a cathode of only Al was about 3.80 V. The maximum luminous efficiency was 8.86 cd/ A, at 4493.1 cd/ m2 共50.73 mA/ cm2, 5.81 V兲. The organic-oxide buffer layer markedly improved the device performance. The inset in Fig. 1 presents the results of the photovoltaic measurements for devices with Al and PEGDE/Al cathodes. The open-circuit voltages 共Voc兲 of devices with Al and PEGDE/Al cathodes were 0.96 V and 1.65 V, respectively,
Appl. Phys. Lett. 88, 113501 共2006兲
FIG. 2. I-L-V curves of the devices with Ca/ Al共䊐兲, PEGDE共25 Å兲/Al /Al共䊏兲, PEGDE共25 Å兲/Ag共⫻兲, Al共쎲兲, and PEGDE共25 Å兲Å兲 / Al共䊊兲 as the device cathode.
suggesting the difference between the work functions of the cathodes and the built-in potential of the devices.9,11–13 In Fig. 2, the threshold voltages of both the charge injection and the light emission of the device with the PEGDE 共25 Å兲 / Al cathode are close to those of the device with the Ca/ Al cathode. The formation of a specific organic oxide/Al complex during the deposition of Al electrode facilitates the injection of electrons and improves the device performance. The considerable enhancement of EL efficiency is attributed to the balance of injected charge carriers, holes, and electrons. However, this phenomenon was observed only in devices in which the PEGDE buffer layer had an Al electrode. Silver 共Ag兲 with a PEGDE buffer layer did not improve the device performance. In Fig. 2, the device, ITO/PEDOT:PSS/HYPPV/PEGDE 共25 Å兲 / Ag, exhibits a lower injected current and a higher light turn-on voltage than those of the device with the PEGDE 共25 Å兲 / Al cathode. The performance of the device varies with the thickness of the organic-oxide buffer layers. Figure 1 shows the lower light turn-on voltages and the higher EL intensities of the device with the PEGDE 共25 Å兲 / Al cathode. The luminous efficiencies of ITO/PEDOT:PSS/HY-PPV/PEGDE/Al devices with PEGDE layers of thicknesses 12, 25, 50, and 75 Å, biased at 100 mA/ cm2, are 7.15, 8.74, 5.89, and 4.11 cd/ A, respectively, as shown in Fig. 4. The region in which Al atoms effectively react with the organic-oxide layer is limited to the contact interface at approximately several tens of Å. The performance is best for the device with a 25 Å thick PEGDE film and an Al electrode. The EL efficiency declines as the thickness of the organic-oxide layers increases, because the PEGDE film is basically an insulating polymer. A thin layer of polyethylene 共PE兲 共Aldrich, Mn ⬃1700兲, 共–CH2CH2 – 兲n, was used 共thermally evaporated兲 instead of PEGDE as the cathode buffer layer to elucidate the function of the PEGDE/Al cathodes. Figure 2 reveals that the threshold of charge injection of the device with the PE 共25 Å兲 / Al cathode equals that of the device with the Al cathode. The PE buffer layer did not interact with Al in the same way as the PEGDE film. The lower injected current indicates the higher series resistance of the device with the PE/Al cathode, since the carrier conductivity of the PE layer is relatively poor. However, the device with the PE/Al cathode has an increasingly higher EL intensity than the device with the Al cathode as the bias is increased, as shown in Fig. 4, probably because the PE or PEGDE insulated film somehow acts as a wide band-gap hole-blocking layer, which confines the holes in the
Downloaded 22 Oct 2009 to 140.116.208.44. Redistribution subject to AIP license or copyright; see http://apl.aip.org/apl/copyright.jsp
113501-3
Appl. Phys. Lett. 88, 113501 共2006兲
Guo et al.
FIG. 3. 共Color online兲 XPS C 1s spectra of pristine HY-PPV film共〫兲, HY-PPV/ Al共30 Å兲共쎲兲, and HY-PPV/ PEGDE共25 Å兲 / Al共30 Å兲共䊏兲. The inset shows the photograph of the device 共3.0⫻ 3.5 cm兲 fabricated with PEGDE共12 Å兲 / Al cathode.
EL layer, increasing the probability of recombination of oppositely charged carriers. The interaction at the organic/metal interface was investigated by XPS. In Fig. 3, the C 1s core-level spectrum of the pristine HY-PPV surface has a main peak at a binding energy 共BE兲 of 284.6 eV and a second peak at 286.2 eV. These peaks are associated with hydrocarbon atoms 共C–C and C–H兲 and carbon atoms attached to the oxygen 共C–O兲 of HY-PPV molecules, respectively.16,17 The HY-PPV film with the 30 Å thick layer of thermally evaporated Al, as presented in Fig. 3, yielded a reduced BE peak at 286.2 eV, which is attributable to the breaking of C–O bonds in HY-PPV molecules at the contact interface. This finding is of an interfacial reaction of Al with the organic EL layer. However, when an additional PEGDE buffer layer 共25 Å兲 was placed on the surface of the HY-PPV film with 30 Å of thermally evaporated Al, the feature at 286.2 eV did not obviously change; rather, the third C 1s peak at 288.6 eV grew. PEGDE is a polymer with the sequent carbon-oxide functional group, 共–CH2CH2O – 兲n, so a strong interaction between Al and the lone-pair electrons in the oxygen atoms of the PEGDE chains is expected. The third C 1s feature at 288.6 eV is associated with the formation of the organic oxide/Al complex. Therefore, a PEGDE buffer layer at the cathode interface effectively suppresses the diffusion of the metal atoms, and blocks the doping reaction in the EL layer. The diffusion of cathode metals into the EL layer dopes the conjugated molecules near the cathode interface,18,19 quenching luminescence at those reaction sites. The luminous efficiency is related to the doping level of the EL film. Figure 4 shows that the devices with PEGDE/Al cathodes have higher EL efficiencies than that of the device with Ca/ Al cathode 共5.26 cd/ A biased at 100 mA/ cm2兲. The EL efficiencies remain high in the high brightness regime when the devices are biased at high current. No variation in the EL spectra or the CIE coordinates was observed. In summary, the thickness, the functional groups in the organic buffer layer and the metal electrodes with various work functions, were changed to verify the multiple functions of the PEGDE layer, and thereby significantly improve the device performance. The inset in Fig. 3 presents a photograph of the device fabricated with a PEDGE 共25 Å兲 / Al cathode. The uniformity of the EL across the light-emissive region was excellent, as determined by microscopy. A study of the feasibility of using an organic oxide/Al composite
FIG. 4. The luminous efficiency versus current for the devices applying Al共쎲兲, PEGDE共12 Å兲 / Al共䉭兲, PEGDE共25 Å兲 / Al共䊏兲, PEGDE共50 Å兲 / Al 共⽧兲 PEGDE共75 Å兲 / Al共䉮兲, Ca/ Al共䊐兲, PEGDE共25 Å兲 / Ag共⫻兲, and PE共25 Å兲 / Al共䊊兲 as the cathode, respectively.
cathode to fabricate small molecular organic LEDs is currently underway.20 The authors would like to thank the National Science Council 共NSC兲 of Taiwan 共NSC94-2113-M-006-007兲 and the Center for Micro-NanoTechnology of National Cheng Kung University 共NCKU兲 共NSC93-212-M-006-006兲 for financially supporting this research. Dr. Ruei-Tang Chen from Eternal Chemical Co., Ltd, is appreciated for providing the HY-PPV polymer. The assistance of Reui-Chin Lee from NSC Instrument Development Center at NCKU is highly appreciated for performing XPS measurements. 1
J. H. Burroughes, D. D. C. Bradley, A. R. Brown, R. N. Marks, K. Mackay, R. H. Friend, P. L. Burn, and A. B. Holmes, Nature 共London兲 347, 539 共1990兲. 2 D. Braun and A. J. Heeger, Appl. Phys. Lett. 58, 1982 共1991兲. 3 L. S. Huang, C. W. Tang, and M. G. Mason, Appl. Phys. Lett. 70, 152 共1997兲. 4 G. E. Jabbour, Y. Kawabe, S. E. Shaheen, J. F. Wang, M. M. Morrell, B. Kippelen, and N. Peyghambarian, Appl. Phys. Lett. 71, 1762 共1997兲. 5 J. Yoon, J. J. Kim, T.-W. Lee, and O. O. Park, Appl. Phys. Lett. 76, 2152 共2000兲. 6 Y. Cao, G. Yu, and A. J. Heeger, Adv. Mater. 共Weinheim, Ger.兲 10, 917 共1998兲. 7 J. Kido and T. Matsumoto, Appl. Phys. Lett. 73, 2866 共1998兲. 8 T.-W. Lee, O. O. Park, L.-M. Do, T. Zyung, T. Ahn, and H.-K. Shim, J. Appl. Phys. 90, 2128 共2001兲. 9 Q. Xu, J. Ouyang, Y. Yang, T. Ito, and J. Kido, Appl. Phys. Lett. 83, 4695 共2003兲. 10 J. H. Park, O. O. Park, J.-W. Yu, J. K. Kim, and Y. C. Kim, Appl. Phys. Lett. 84, 1783 共2004兲. 11 X. Y. Deng, W. M. Lau, K. Y. Wong, K. H. Low, H. F. Chow, and Y. Cao, Appl. Phys. Lett. 84, 3522 共2004兲. 12 Y.-H. Niu, H. Ma, Q. Xu, and A. K.-Y. Jen, Appl. Phys. Lett. 86, 083504 共2005兲. 13 T.-F. Guo, F.-S. Yang, Z.-J. Tsai, T.-C. Wen, S.-N. Hsieh, and Y.-S. Fu, Appl. Phys. Lett. 87, 013504 共2005兲. 14 H. Becker, H. Spreitzer, W. Kreuder, E. Kluge, H. Schenk, I. Parker, and Y. Cao, Adv. Mater. 共Weinheim, Ger.兲 12, 42 共2000兲. 15 Y. Shi, L. Liu, and Y. Yang, J. Appl. Phys. 87, 4254 共2000兲. 16 S. Li, E. T. Kang, K. G. Neoh, Z. H. Ma, K. L. Tan, and W. Huang, Appl. Surf. Sci. 181, 201 共2001兲. 17 G. G. Andersson, W. J. H. van Gennip, J. W. Niemantsverdriet, and H. H. Brongersma, Chem. Phys. 278, 159 共2002兲. 18 J. Birgerson, M. Fahlman, P. Bröms, and W. R. Salaneck, Synth. Met. 80, 125 共1996兲. 19 F. J. J. Janssen, L. J. van IJzendoorn, A. W. Denier van der Gon, M. J. A. de Voigt, and H. H. Brongersma, Phys. Rev. B 70, 165425 共2004兲. 20 We have the promising result. This part of the research will be published. Downloaded 22 Oct 2009 to 140.116.208.44. Redistribution subject to AIP license or copyright; see http://apl.aip.org/apl/copyright.jsp
