Effects of substrate treatment on the electroluminescence
performance of flexible OLEDs
q
Shih-Nan Lee, Shih Feng Hsu, Shiao Wen Hwang, Chin H. Chen
*Microelectronics and Information Systems Research Center, Department of Applied Chemistry, National Chiao Tung University, Hsinchu 30050, Taiwan, ROC
Received 30 August 2003; accepted 7 November 2003 Available online 27 April 2004
Abstract
Electroluminescence (EL) performance of flexible organic light-emitting device (FOLED) has been found to be highly dependent upon the morphology of the surface of the indium thin oxide (ITO)/plastic substrate as well as the patterning and processing conditions of the substrate. We will present evidence showing that luminance efficiency of FOLED can be greatly improved by pretreatment. Surface analysis of the ITO/PET by means of atomic force microscope (AFM) and optical microscope will be compared with that of the ITO/glass and the influence of flexible OLEDs substrate treatment by various methods on EL perfor-mance will also be discussed.
2004 Elsevier B.V. All rights reserved.
PACS: 06.60.)C; 06.60.Mr; 07.60.)j; 07.60.Pb; 07.60.Rd; 42.79.)e; 85.60.)q; 85.6.)w; 78.20.)e; 42.70.JK; 42.79.Kr; 61.41.+e; 81.05.Lg; 82.65; 68.37.PS; 85.60.Jb
Keywords: FOLED; PET; Surface analysis; LiF; ITO
1. Introduction
Since organic light-emitting diodes (OLED) have been found useful in flat panel display (FPD) applica-tions, several novel OLED technologies such as trans-parent OLED (TOLED) [1,2], Stack OLED (SOLED) [3] and flexible OLED (FOLED) [4–6] have been dis-covered. Fabrication of a display that is conformable, wearable and still fully functional even when it is bend or flexed is the key advantages of FOLED technology as compared to the others. This is because OLED encom-passes thin, amorphous solid-state films, which can be easily fabricated on any substrate, such as glass or flexible plastics and steel foil. Flexible OLEDs using plastic substrates are lighter and thinner than OLEDs
based on glass. They can be bent and twisted and are literally unbreakable.
Although flexible OLED possesses several advanta-ges, there are still many challenges to be overcome. Two of these challenges are to enhance the efficiency and stability of devices. Through substrate pretreatment, the efficiency and stability of the devices can be enhanced. By optimizing the cleaning conditions [7], performing oxidation of the indium thin oxide (ITO) surface by O2
plasma [8], and inserting a buffer layer or hole-injection layer [9], such as copper phthalocyanine (CuPc), be-tween the ITO and holo transfer layer (HTL) interfaces, the energy barrier of hole-injection can be reduced and the performance of the device can be improved. In previous reports, the insertion of LiF thin film between Alq3 and Al can effectively reduce the energy barrier of
electron injection [10]. Furthermore, Zhu [11] and Zhao [12] have published thin LiF layer on the ITO anode would improve device performance, but no results about device lifetime have been reported. Due to the fact that the plastic substrates can be only operated at tempera-ture below their Tg (glass transition temperature), the
deposited ITO film is often amorphous [13] and cannot
q
Original version presented at the 4th International Conference on Electroluminescence of Molecular Materials and Related Phenomena (ICEL4), 27–30 August 2003, Cheju Island, Korea.
*
Corresponding author. Tel.: 5712121x59200; fax: +886-3-5737681.
E-mail address:fredchen@eic.nctu.edu.tw(C.H. Chen).
1567-1739/$ - see front matter 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.cap.2003.11.064
Current Applied Physics 4 (2004) 651–654
be postannealed. This results in high resistance [13], which would cause the increase of turn on voltages. In this paper we will demonstrate the application of LiF thin film on ITO/PET substrates to improve the device performance and show the influence on device lifetime. 2. Experimental
The flexible ITO substrate used is the touch panel grade PET/ITO plastic substrate, supplied by U-shine in Hsinchu of Taiwan, with resistivity of 1· 103 Xcm.
The ITO glass substrate with resistivity of 1.8· 104
Xcm is purchased from Merck. The sequential steps for patterning were carried out. At first step, the cleaning of ITO substrates was subsequently done with isopropane alcohol (IPA), detergent, deionized water, and methanol in ultra-sonic bath to remove the organic contaminants that act as fatal impurities on the ITO substrates during the photolithographic patterning process.
The next step is the fabrication of patterned anode plate. First of all, the positive commercial photoresist were spin-coated on the substrate to a thickness of about 1 lm. After baked, it was placed on a mask aligner and exposed to ultra-violet(UV) light for 90 s. Then it was developed in a developer for about 60 s, followed by baking process for 2 min at 75 C. Subsequently the samples were etched by 3% H2C2O4ðaqÞfor 2 min at bath
temperature of 35–40 C. Surface analyses of the ITO substrates were performed by means of atomic force microscope (AFM) and optical microscope after wet etching. The step for pretreatment is to deposit LiF layer in different thickness, 5, 10, and 15 A at a deposition rate of 0.1 A/s using ULVAC coater, or treat it with 60 sccm of O2 plasma without LiF, and following by
sequential deposition of CuPc, N,N0
-bis(1-naphthyl)-N,N0-diphenyl-1,10-biphenyl-4,40-diamine (NPB),
tris(8-hydroxyquinolinato)aluminum (Alq3) in the 105 Pa
and cathode in the 104 Pa. The device structure is
[ITO/LiF (x A)/CuPC (150 A)/NPB (600 A)/Alq3 (750
A)/LiF (10 A)/Al (2000 A)]. The electroluminescence characteristics of OLEDs were measured with a Photo Research PR650 spectrophotometer and a programma-ble current source. The surface roughness and inspection measurements were performed by AFM and Olympus optical microscope, respectively. The device lifetime was measured by photodiode connected to a data acquisition instruments and computer.
3. Results and discussion 3.1. Dark spots
Compared to the traditional OLEDs, which were fabricated on glass substrate, we found the FOLEDs reveal more dark spots under the same test conditions.
This phenomenon is observed immediately after device fabrication, so the factor of O2 and H2O permeation is
excluded. From the microscopic and macroscopic observation, we could seek the clue about why FOLEDs show more dark spots than traditional OLED. Com-pared with ITO/glass substrate, PET substrates show the smaller root-mean-square (Rms¼ 3.4 nm) roughness than that of glass substrates (Rms¼ 3.8 nm). The big Rms value may be due to the different crystal of ITO in glass substrate. The crystal of ITO could lead to smaller resistivity than that of ITO/PET substrate. Further-more, there was observed a spike on ITO/PET surface. Some papers have pointed out the influence of spikes on device degradation [14,15]. From the viewpoint of OLED, the spikes could be more determined than Rms value on device reliability. Fig. 1(a) and (b) shows the optical microscope (OM) graphs of two ITO substrates after aluminum layer was deposited immediately in the same batch. The aluminum layer deposited on glass and PET substrates enhance the contrast for OM detection. Although the roughness of ITO on PET (Rms¼ 3.4 nm) is smaller than that on glass (Rms¼ 3.8 nm), a lot of pinholes have been found on the optical graph of minum on ITO/PET. By contrast, the surface of
alu-Fig. 1. (a) Microscope image (·50) of ITO/glass with Al layer. (b) Microscope image (·50) of ITO/PET with Al layer.
minum on ITO/glass is uniform. The pinholes may result from the spike in PET substrate. In summary, the spikes of substrates and the non-uniformity of deposited films may cause the shorting of two adjacent electrodes, which subsequently result in the dark spots formation. 3.2. Pretreatment
Fig. 2(a) and (b) shows the current density and luminance vs. voltages of the FOLED devices with bare ITO or different pretreatments of the ITO. It has been reported that oxygen plasma can effectively increase the work function of ITO and improve the performance of OLED. Compare to the traditional OLED, the O2
plasma treatment on ITO/PET does not reduce operat-ing voltage obviously. But the operatoperat-ing voltages of devices with proper LiF thickness on anode are clearly reduced. Under current density of 20 mA/cm2, the
operating voltages of devices are 9.3, 7.9, 7.9, and 9.5 V for devices with 0.5, 1, 1.5 nm LiF, and bare ITO, respectively. The operating voltages of devices with 1 or 1.5 nm of LiF on anode are close to 7.5 V of traditional
OLED device. Similarly, the voltages required to reach a luminance of 1000 cd/m2 are 9.4, 7.8, 7.8, and 9.4 V for
devices with 0.5, 1, 1.5 nm, and bare ITO, respectively. As shown in Fig. 2(b), the plasma treatment does not enhance the luminance at the same operating voltages.
This enhancement of carrier-injection is due to the tunneling junction [16,17] caused by inserting a thin insulating layer between ITO and HTL. The energy level of LiF is at around 12 eV, which is an insulator [11]. By adding an ultra-thin layer of LiF between ITO and CuPc, the hole-injection can be enhanced by means of tunneling junction. Hence, the operating voltages can be reduced. By increasing the LiF thickness, the effect of carrier tunneling will decrease, therefore, leading to a weaker carrier-injection process and higher operating voltages.
Fig. 3 shows the power efficiency vs. current density of devices with bare ITO and with different thickness of LiF on the ITO. The graph clearly shows that the power efficiency of the device is improved by adding at least 1 nm of LiF. Furthermore, the power efficiency of the device with plasma treatment is worse than that of bare ITO. This is due to that the plasma treatment only re-duces the hole-injection barrier, but does not increase the hole-electron balance. Otherwise, as 1 nm LiF layer was inserted between ITO and CuPC, not only the voltage was reduced, but also the brightness could be enhanced. Due to the lower hole-injection barrier and better hole-electron balance than those of devices with-out LiF layer, the power efficiency was improved. Therefore, the performances of FOLEDs using ITO/ PET substrate with unfavorable electronic properties could be dramatically improved by LiF pretreatment.
Fig. 4 exhibits the lifetime curves of devices with and without LiF in glove box. Excluding the extrinsic factor, such as the attack of O2 and H2O, the graph indicates
that the device with LiF is more stable than the device
4 6 8 10 12 14 16 0 20 40 60 80 100
Current Density (mA/cm
2 ) Voltage (V) 0.5 nm LiF 1 nm LiF 1.5 nm LiF No LiF Plasma 4 6 8 10 12 14 16 0 1000 2000 3000 4000 5000 6000 Brightness (cd/m 2 ) Voltage (V) 0.5 nm LiF 1 nm LiF 1.5 nm LiF No LiF Plasma (a) (b)
Fig. 2. (a) Current vs. voltage graph of various LiF thickness. (b) Brightness vs. voltage graph of various LiF thickness.
0 20 40 60 80 100 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 Power Efficiency (lm/W)
Current Density (mA/cm2) 0.5 nm LiF 1 nm LiF 1.5 nm LiF No LiF Plasma
Fig. 3. Power efficiency vs. current density graph of various LiF thicknesses.
without LiF. Further investigation is needed to elucidate detailed mechanism.
4. Conclusion
From the AFM analysis, although the grains of ITO on glass substrate result in the big Rms value, the local spike on ITO/PET substrate can affect the uniformity of deposited films. The spike and the non-uniform of deposited films may cause more dark spots in FOLEDs than in traditional OLED. In addition, by inserting LiF layer between ITO and CuPc interface, the device operating voltage was reduced, and luminance effi-ciency was increased. The performances of FOLEDs using PET substrate with unfavorable electronic properties of ITO could be dramatically improved by LiF pretreatment. The lifetime test also indicates that the device with LiF is more stable than the device without.
Acknowledgements
This work was supported by the National Science Council of Taiwan, Republic of China under the grant of Joint Industrial and Academic Research Project (NSC91-2622-L-009-001).
References
[1] G. Gu, V. Bulovic, P.E. Burrows, S.R. Forrest, M.E. Thompson, Appl. Phys. Lett. 68 (1996) 2606.
[2] V. Bulovic, G. Gu, P.E. Burrows, M.E. Thompson, S.R. Forest, Nature 380 (1996) 29.
[3] P.E. Burrows, V. Khalfin, G. Gu, S.R. Forrest, Appl. Phys. Lett. 73 (1998) 435.
[4] A.N. Krasnova, Appl. Phys. Lett. 80 (2002) 3853.
[5] Z. Chen, B. Cotterell, W. Wang, Eng. Fract. Mech. 69 (2002) 597.
[6] A.B. Chwang, M.A. Rothman, S.Y. Mao, R.H. Hewitt, M.S. Weaver, J.A. Silvernail, K. Rajan, M. Hack, J.J. Brown, X. Chu, L. Moro, T. Krajewski, N. Rutherford, Appl. Phys. Lett. 83 (2003) 413.
[7] Th. Kugler, A. Johansson, I. Dalsseg, U. Gelius, W.R. Salaneck, Synthetic Met. 91 (1997) 143.
[8] D. Lu, Y. Wu, J. Guo, G. Lu, Y. Wang, J. Shen, Mater. Sci. Eng. B97 (2003) 141.
[9] S.A. VanSlyke, C.H. Chen, C.W. Tang, Appl. Phys. Lett. 69 (1996) 2160.
[10] L.S. Hung, C.W. Tang, M.G. Mason, Appl. Phys. Lett. 70 (1997) 152.
[11] F. Zhu, B. Low, K. Zhang, S. Chua, Appl. Phys. Lett. 79 (2001) 1205.
[12] Y. Zhao, S.Y. Liu, J.Y. Hou, Thin Solid Films 397 (2001) 208.
[13] D.C. Paine, T. Whitson, D. Janiac, R. Beresford, C.O. Yang, B. Lewis, J. Appl. Phys. 85 (1999) 8445.
[14] Y.H. Tak, K.B. Kim, H.G. Park, K.H. Lee, J.R. Lee, Thin Solid Films 411 (2002) 12.
[15] J. McElvain, H. Antoniadis, M.R. Hueschen, J.N. Miller, D.M. Roitman, J.R. Sheats, R.L. Moon, J. Appl. Phys. 80 (1996) 6002.
[16] X.M. Ding, L.H. Hung, L.F. Cheng, Z.B. Deng, X.Y. Hou, C.S. Lee, S.T. Lee, Appl. Phys. Lett. 76 (2000) 2704.
[17] Y.E. Kim, H. Park, J.J. Kim, Appl. Phys. Lett. 69 (1996) 599.
0 20 40 60 80 100 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 Time (hrs) No LiF with LiF L0=1141 cd/m2 at 20 mA/cm2 L0=1018 cd/m2 at 20 mA/cm2
Fig. 4. Device lifetime comparison between device with LiF and device without.
