Research Article Cite This: ACS Appl. Mater. Interfaces 2018, 10, 8148−8154
www.acsami.org
Extremely Low Roll-Off and High Efficiency Achieved by Strategic Exciton Management in Organic Light-Emitting Diodes with Simple Ultrathin Emitting Layer Structure Tianmu Zhang,† Changsheng Shi,† Chenyang Zhao,† Zhongbin Wu,† Jiangshan Chen,‡ Zhiyuan Xie,*,† and Dongge Ma*,†,‡ †
ACS Appl. Mater. Interfaces 2018.10:8148-8154. Downloaded from pubs.acs.org by UNIV OF KANSAS on 01/20/19. For personal use only.
State Key Laboratory of Polymer Physics and Chemistry, Changchun Institute of Applied Chemistry, University of Chinese Academy of Sciences, Changchun 130022, P. R. China ‡ Institute of Polymer Optoelectronic Materials and Devices, State Key Laboratory of Luminescent Materials and Devices, South China University of Technology, Guangzhou 510640, P. R. China S Supporting Information *
ABSTRACT: Phosphorescent organic light-emitting diodes (OLEDs) possess the property of high efficiency but have serious efficiency roll-off at high luminance. Herein, we manufactured high-efficiency phosphorescent OLEDs with extremely low roll-off by effectively locating the ultrathin emitting layer (UEML) away from the high-concentration exciton formation region. The strategic exciton management in this simple UEML architecture greatly suppressed the exciton annihilation due to the expansion of the exciton diffusion region; thus, this efficiency roll-off at high luminance was significantly improved. The resulting green phosphorescent OLEDs exhibited the maximum external quantum efficiency of 25.5%, current efficiency of 98.0 cd A−1, and power efficiency of 85.4 lm W−1 and still had 25.1%, 94.9 cd A−1, and 55.5 lm W−1 at 5000 cd m−2 luminance, and retained 24.3%, 92.7 cd A−1, and 49.3 lm W−1 at 10 000 cd m−2 luminance, respectively. Compared with the usual structures, the improvement demonstrated in this work displays potential value in applications. KEYWORDS: ultrathin, efficiency roll-off, exciton management, exciton distribution, organic light-emitting diodes
1. INTRODUCTION Organic light-emitting diodes (OLEDs) have generated attention for next generation display and solid-state lighting because of their attractive features, including homogeneous large area, color tunability, and flexibility.1−5 In past years, phosphorescent OLEDs have been demonstrated to achieve almost 100% internal quantum efficiency because they can harvest all generated singlet and triplet excitons but there still exists serious roll-off at high luminance when using a single doping emitting layer (EML) structure.6−10 It is known that the structural design of multiple doping EMLs could reduce the efficiency roll-off at high luminance to a certain extent while achieving high efficiency in phosphorescent OLEDs.11−15 However, the multiple doping procedure also greatly increases the complexity of device processing, the repeat ability of device performance, and importantly, the efficiency roll-off at high luminance still requires amelioration. To address these shortcomings, a simplified structure of the ultrathin emitting layer (UEML) without doping, which generally introduced an ultrathin layer (thickness < 1 nm) at the interface of the carrier transport layers, has been experimentally proved to be an effectual way not only for obtaining high efficiency but also greatly simplifying the device process.16−23 © 2018 American Chemical Society
However, the general UEML structure easily causes the interface charge accumulation, thus quenching the excitons, and the serious efficiency roll-off at high luminance still exists in the devices.22−26 Herein, we demonstrate a simple and efficient architecture that greatly improved the efficiency roll-off at high luminance in UEML-based OLEDs. Completely different from general UEML structures (Figure 1a), an ultrathin phosphor EML was inserted to the proper position in the electron-transporting layer (ETL), away from the exciton recombination interface, as shown in Figure 1b. This avoided serious exciton quenching and significantly minimized the efficiency roll-off of the fabricated OLEDs. By optimizing the UEML position, we fabricated a highefficiency green phosphor OLED, exhibiting not only higher efficiency but also extremely low-efficiency roll-off compared with general UEML and doping double EML OLEDs.22,23 The corresponding maximum current efficiency (CE), power efficiency (PE), and external quantum efficiency (EQE) Received: January 10, 2018 Accepted: February 13, 2018 Published: February 13, 2018 8148
DOI: 10.1021/acsami.8b00513 ACS Appl. Mater. Interfaces 2018, 10, 8148−8154
Research Article
ACS Applied Materials & Interfaces
Figure 1. Models of exciton emission process in two kinds of UEML architectures. (a) OLEDs with general UEML structure. (b) OLEDs with improved UEML structure in this work.
Figure 2. Characteristics of Devices D1−D3: (a) current−voltage−luminance, (b) CE−luminance, (c) PE−luminance, and (d) EQE−luminance.
reached 98.0 cd A−1, 80.59 lm W−1, and 25.6%, and still kept at a high value of 92.7 cd A−1, 49.3 lm W−1, and 24.3% at 10 000 cd m−2 luminance, respectively.
3. RESULTS AND DISCUSSION 3.1. Device Structure and Performance. Green phosphorescent OLEDs were designed and fabricated to compare the advantage of UEML structure in this study. In these devices, ITO and Al were chosen as the anode and cathode, respectively. MoO3 and Liq were used as the hole and electron injection layers, respectively. N,N′-Di(naphthalene-1yl)-N,N′-diphenyl-benzidine (NPB) was selected for the holetransporting layer and bis(2-[2-hydroxyphenyl]pyridinato)beryllium (Bepp2) for the electron-transporting layer (ETL).33,34Green phosphorescence bis(2-phenylpyridine)iridium acetylacetonate (Ir(ppy)2(acac)) was used as the UEML. 4,4′,4″-Tri(N-carbazolyl)triphenylamine (TCTA) was used as the host. 1,1′-Bis[4-(di-p-tolylamino)phenyl]cyclohexane (TAPC) is used as the electron blocking layer.35 The fabrication of the three devices was as follows. 3.1.1. Device D1 (Doping Double EML). ITO (180 nm)/ MoO3 (10 nm)/MoO3: NPB (25%, 35 nm)/NPB (10 nm)/ TAPC (5 nm)/TCTA: Ir(ppy)2(acac) (8%, 5 nm)/Bepp2: Ir(ppy)2(acac) (8%, 5 nm)/Bepp2 (15 nm)/Bepp2: Liq (3%, 25 nm)/Liq (1 nm)/Al (150 nm). 3.1.2. Device D2 (UEML-1). ITO (180 nm)/MoO3 (10 nm)/ MoO3: NPB (25%, 35 nm)/NPB (10 nm)/TAPC (5 nm)/
2. EXPERIMENTAL SECTION The OLEDs studied here were fabricated on indium tin oxide (ITO) that has been fabricated on glass substrates, and their resistance was 10 Ω/□. Before use, the precoated glass substrates were bathed in waterbased ultrasonic cleanout fluid. Then, they were processed in a drying cabinet over 105 °C for 10 min, and oxygen plasma treatment for 2 min. Thermal evaporation method was used to fabricate the OLEDs in this work. The vacuum pressure of the evaporation equipment was around 6.0 × 10−4 Pa during processing. A Quartz crystal oscillator and Dektak 6M profiler was used to monitor and adjust the rate of material evaporation. The material evaporation was optimized to a proper rate to attain films of high quality. Owing to the ultrathin thickness, the UEML should not be continuous and the film is an average thickness measured by the film controlling system. The area of the emission zone was 4 mm × 4 mm overlapped between ITO and aluminum (Al). By using Keithley 2400 and Keithley 2000, the properties including current, luminance, and voltage could be measured. A SpectraScan spectrophotometer was introduced to realize measurement of the EL spectra. The characteristics of the devices were measured in ambient atmospheric conditions, and all of the OLEDs were fabricated without additional out-coupling techniques. 8149
DOI: 10.1021/acsami.8b00513 ACS Appl. Mater. Interfaces 2018, 10, 8148−8154
Research Article
ACS Applied Materials & Interfaces Table 1. Summary of EL Performance of Devices D1−D3 CEc (cd A−1)
EQEb (%) device
Von [V]
max
@5000 (cd m )
@10 000 (cd m )
max
@5000 (cd m )
@10 000 (cd m )
max
@5000 (cd m−2)
@10 000 (cd m−2)
D1 D2 D3
2.8 3 3
24.1 24.2 25.5
23.2 21.5 25.1
22.2 19.8 24.3
92.4 92.7 98
87.8 80.5 94.9
85.6 75.2 92.7
83.3 85.3 85.4
51.9 48.3 55.5
43.5 38.1 49.3
a
−2
−2
−2
PEd (lm W−1)
a
−2
Turn-on voltage. bExternal quantum efficiency (EQE). cCurrent efficiency (CE). dPower efficiency (PE).
2, and 3 nm in devices D1 and D2 in sequence, and −1, 0, 1, 2, 3, 5, and 6 nm in device D3 from the interface TCTA/Bepp2 (the interface is defined as the position “0 nm”, ranging in the TCTA direction as negative distance and in the Bepp2 direction as positive distance). The PO-01 sensing layer is extremely thin, and its introduction should not significantly affect the charge transport in devices (Supporting Information, Figure S3). On comparing the intensity of PO-01 (peak = 552 nm) to that of Ir(ppy)2(acac) (peak = 520 nm), the relative intensity of PO-01 spectra could represent the tendency of the exciton distribution in each device well (Supporting Information, Figures S1 and S2). Figure 3 shows the tendency of the exciton distribution in the emission zone within devices D1− D3.
TCTA (5 nm)/Ir(ppy)2(acac) (0.2 nm)/Bepp2 (20 nm)/ Bepp2: Liq (3%, 25 nm)/Liq (1 nm)/Al (150 nm). 3.1.3. Device D3 (UEML-2). ITO (180 nm)/MoO3 (10 nm)/ MoO3: NPB (25%, 35 nm)/NPB (10 nm)/TAPC (5 nm)/ TCTA (5 nm)/Bepp2 (4 nm)/Ir(ppy)2(acac) (0.2 nm)/Bepp2 (16 nm)/Bepp2: Liq (3%, 25 nm)/Liq (1 nm)/Al (150 nm). The structures of the compared devices are similar except for the EMLs. Device D1 is a green phosphor OLED with double EMLs by doping Ir(ppy)2(acac) in TCTA and Bepp2 by a ratio of 8%, respectively. Device D2 is a green phosphor OLED with inserting an Ir(ppy)2(acac) UEML of 0.2 nm into the interface of TCTA/Bepp2. Device D3 is a green phosphor OLED with locating the same UEML in Bepp2 4 nm away from the interface of TCTA/Bepp2. The three devices displayed similar turn-on voltage and exhibited high efficiency, indicating good exciton utilization. However, as shown in Figure 2, it remarkably exhibits higher efficiency at high luminance compared with devices D1 and D2, indicating the validity of the device structure designed here. The efficiency of device D3, respectively, attained the EQE of 25.5%, CE of 98.0 cd A−1, and PE of 85.4 lm W−1 at max. They, respectively, retained values of 24.3%, 92.7 cd A−1, and 49.3 lm W−1 at 10 000 cd m−2 high luminance, higher than those of devices D1 and D2. This indicates the exciton quenching in device D3 was significantly improved. Comparatively, device D2 showed more serious efficiency roll-off. Efficiency was reduced to 19.8%, 75.2 cd A−1, and 38.1 lm W−1 at 10 000 cd m−2 luminance from the maximum efficiency of 24.2%, 92.7 cd A−1, and 85.3 lm W−1, respectively. Although the architecture of doping double EML in device D1 improved the efficiency rolloff compared to that of device D2, reaching 22.2%, 85.6 cd A−1, and 43.5 lm W−1 at 10 000 cd m−2 luminance, it can be seen that the exciton quenching is still serious than that in device D3. Exciton emission from the green phosphor was within the exciton recombination zone in devices D1 and D2, easily quenching the radiative excitons, whereas the exciton emission and exciton recombination in device D3 were completely separated. The radiative excitons were far away from the highconcentration exciton recombination zone. Therefore, the exciton quenching was greatly suppressed. This should be an important feature of the UEML architecture we designed. Table 1 summarizes the EL performance of devices D1−D3. 3.2. Dependence of Exciton Distribution on the Emissive Layer Structure. To further elucidate the mechanism by which efficiency roll-off at high luminance was improved in device D3, we explored the tendency of the exciton distribution in recombination zones of devices D1−D3. A general method is inserting an ultrathin red or yellow phosphor as a sensing layer in the destination area, and by comparing the relative emission intensity of the red or yellow light to the control emission, information about the spatial distribution of the triplet excitons in OLEDs can be provided well.13,27,28 In this article, a 0.02 nm PO-01 yellow phosphor layer was placed, respectively, at distances of −3, −2, −1, 0, 1,
Figure 3. Exciton distribution of the emission zone in devices D1−D3.
Device D3 displayed a wider exciton distribution than that of devices D1 and D2. Device D1 illustrates a relative narrow exciton distribution with a peak at 1 nm in ETL, which should be due to the hindering effect of Ir(ppy)2(acac) on electron transport within the doping layer of Bepp2: Ir(ppy)2(acac) (Supporting Information, Figure S6), and slightly higher hole mobility of TCTA than the electron mobility of Bepp2.29,30 Additionally, the triplet energy level of the host TCTA (2.86 eV) was higher than that of the host Bepp2 (2.6 eV), thus the triplet excitons on the TCTA molecules could diffuse into the Bepp2. This may result in the exciton recombination shifting to the ETL. The energy of generated excitons within the recombination zone of TCTA and Bepp2 will ultimately be transferred to the Ir(ppy)2(acac), finally leading to green emission (Figure 4a). Because the doping of Ir(ppy)2(acac) covered the entire emission layer consisting of TCTA and Bepp2, the generated excitons could attain excellent utilization and effective improvement of the exciton quenching. Therefore, the device D1 represented high efficiency and reduced efficiency roll-off. For device D2, because of the energy level structure of TCTA (lowest unoccupied molecular orbital (LUMO) = −2.3 eV, highest occupied molecular orbital 8150
DOI: 10.1021/acsami.8b00513 ACS Appl. Mater. Interfaces 2018, 10, 8148−8154
Research Article
ACS Applied Materials & Interfaces
Figure 4. Working principles of devices. (a) Device D1, (b) device D2, and (c) device D3.
(HOMO) = −5.7 eV, and T1 = 2.86 eV)29 and Bepp2 (LUMO = −2.6 eV, HOMO = −5.7 eV, and T1 = 2.6 eV),30 the lower triplet energy level of Ir(ppy)2(acac) was confined the generated excitons in the interfacial region.22 Therefore, a narrow exciton distribution with a peak at the interface of TCTA/Bepp2 was observed. Similar to device D1, in D2, the generated exciton energy could also effectively transfer to Ir(ppy)2(acac) (LUMO = −2.6 eV, HOMO = −5.0 eV, and T1 = 2.3 eV),31 resulting in high efficiency (Figure 4b). However, high exciton concentration at the interface also leads to serious annihilation that manifests as a dramatic efficiency roll-off, as shown in Figure 2. In device D2, the UEML was located at the interface of high-exciton concentration, leading the highconcentration exciton accumulation to severely quench the radiative excitons with the increase of current density. This
problem was resolved by locating the UEML away from the high-concentration exciton formation region in device D3. As shown in Figure 3, the spatial distribution illustrated a sharp reduction of the exciton density at 0 to −1 nm, which implies effective confinement of the exciton transfer from the interface to TCTA due to its higher triplet energy level. Although the exciton distribution showed a gentle slope from 0 to 3 nm of a high concentration level, it gradually lowered from 3 nm, suggesting exciton transfer in a single direction, recombining and emitting on the UEML 4 nm from the TCTA/Bepp2 interface. This wide exciton distribution undoubtedly reduces the exciton quenching. As shown in Figure 4c, the thin Ir(ppy)2(acac) emission layer would hardly impact the charge carrier transport; therefore, the electrons and holes will first be injected to the interface between TCTA/Bepp2 and then 8151
DOI: 10.1021/acsami.8b00513 ACS Appl. Mater. Interfaces 2018, 10, 8148−8154
Research Article
ACS Applied Materials & Interfaces
Figure 5. Normalized EQE−current density characteristics of devices. (a) Device D1, (b) device D2, and (c) device D3. (d) The extracted J0 values are given. The red solid lines are the fitted results from the TTA model.
recombine to form excitons, finally leading to green emission by the long-range energy transfer of excitons across Bepp2. It can be seen that this process is very effective. The formed exciton energy can be 100% utilized with much lower quenching due to the exciton radiation far from the highconcentration exciton formation zone. The realization of two discrete channels of exciton radiation and exciton formation provides a promising new avenue to achieve high-efficiency OLEDs with extremely low-efficiency roll-off. 3.3. Analysis of Quenching Mechanism Models Driven by Exciton Density. Extensive experiments have shown that triplet-polaron quenching and triplet−triplet annihilation (TTA) are considered as the prime quenching mechanisms for efficiency roll-off in phosphorescence OLEDs. To analyze the quenching mechanism among our OLEDs with UEML architecture, we fitted efficiency properties of devices D1−D3 by a TTA model. As shown in Figure 5, the efficiency degradation properties were fitted well using the TTA model, implying that it is the dominant annihilation process in devices D1−D3. The TTA model refers to the biexcitonic quenching mechanism proposed by Baldo and group.32−35 It can be expressed as follows ηEQE η0
⎞ J ⎛ 8J = 0 ⎜⎜ 1 + − 1⎟⎟ 4J ⎝ J0 ⎠
137.4, and 324.4, respectively. As D3 had the largest J0 value, this provides additional validity for suppressing exciton quenching in device D3.
4. CONCLUSIONS We successfully resolved the problem of efficiency roll-off at high luminance in phosphorescent OLEDs by locating the UEML away from the high-concentration exciton formation region. The realization of two discrete channels of exciton radiation and exciton formation greatly suppressed the exciton annihilation due to the expansion of the exciton recombination region, thus not only achieving high efficiency but also significantly suppressing the efficiency roll-off at high luminance. The resulting green phosphorescence OLEDs exhibited the maximum EQE of 25.5%, CE of 98.0 cd A−1, and PE of 85.4 lm W−1. They retained 25.1%, 94.9 cd A−1, and 55.5 lm W−1 at 5000 cd m−2 luminance, respectively. This novel design provides a promising avenue to fabricate high-efficiency phosphorescent OLEDs with extremely low efficiency roll-off.
■
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsami.8b00513. Original normalized spectra by inserting a sensing layer of PO-01 into the series position in devices D1−D3 for exciton distribution exploring; current density−voltage of a 0.02 nm PO-01 sensing layer inserted at different positions to the interface TCTA/Bepp2; properties of the devices within the optimizing process of the UEML
(1)
The maximum EQE in the absence of TTA is represented as η0. The current density is J, and J0 stands for the current density when EQE drops to half of its maximum value, called as the critical density. Efficiency roll-off and the J0 value are inversely related. The values J0 of devices D1, D2, and D3 were 256.1, 8152
DOI: 10.1021/acsami.8b00513 ACS Appl. Mater. Interfaces 2018, 10, 8148−8154
Research Article
ACS Applied Materials & Interfaces
■
Utilization of Excitons by an Efficient Sandwich-Type Emissive Layer Architecture for Reduced Efficiency Roll-Off and Improved Operational Stability in Organic Light-Emitting Diodes. ACS Appl. Mater. Interfaces 2016, 8, 3150−3159. (13) Kim, S. Y.; Jeon, W. S.; Park, T. J.; Pode, R.; Jin, J.; Kwon, J. H. Low Voltage Efficient Simple p-i-n Type Electrophosphorescent Green Organic Light-Emitting Devices. Appl. Phys. Lett. 2009, 94, No. 133303. (14) Lee, J.-H.; Lee, S.; Yoo, S.; Kim, K.; Kim, J. Langevin and TrapAssisted Recombination in Phosphorescent Organic Light Emitting Diodes. Adv. Funct. Mater. 2014, 24, 4681−4688. (15) Park, Y. S.; Lee, S.; Kim, K. H.; Kim, S. Y.; Lee, J. H.; Kim, J. J. Exciplex-Forming Co-Host for Organic Light-Emitting Diodes with Ultimate Efficiency. Adv. Funct. Mater. 2013, 23, 4914−4920. (16) Xia, D.; Wang, B.; Chen, B.; Wang, S.; Zhang, B.; Ding, J.; Wang, L. X.; Jing, X. B.; Wang, F. S. Self-Host Blue-Emitting Iridium Dendrimer with Carbazole Dendrons: Nondoped Phosphorescent Organic Light-Emitting Diodes. Angew. Chem., Int. Ed. 2014, 53, 1048−1052. (17) Zhao, Y.; Chen, J.; Ma, D. Realization of High Efficiency Orange and White Organic Light Emitting Diodes by Introducing an UltraThin Undoped Orange Emitting Layer. Appl. Phys. Lett. 2011, 99, No. 163303. (18) Tan, T.; Ouyang, S.; Xie, Y.; Wang, D.; Zhu, D.; Xu, X.; Fong, H. H. Balanced White Organic Light-Emitting Diode with Non-Doped Ultra-Thin Emissive Layers Based on Exciton Management. Org. Electron. 2015, 25, 232−236. (19) Zhao, F.; Zhu, L.; Liu, Y.; Wang, Y.; Ma, D. Doping-Free Hybrid White Organic Light-Emitting Diodes with Fluorescent Blue, Phosphorescent Green and Red Emission Layers. Org. Electron. 2015, 27, 207−211. (20) Li, S. H.; Wu, S. F.; Wang, Y. K.; Liang, J. J.; Sun, Q.; Huang, C. C.; Wu, J. C.; Liao, L. S.; Fung, M. K. Management of Excitons for Highly Efficient Organic Light-Emitting Diodes with Reduced Triplet Exciton Quenching: Synergistic Effects of Exciplex and Quantum Well Structure. J. Mater. Chem. C 2018, 6, 342−349. (21) Wang, Q.; Oswald, I. W.; Yang, X.; Zhou, G.; Jia, H.; Qiao, Q.; Chen, Y.; Hoshikawa-Halbert, J.; Gnade, B. E. A Non-Doped Phosphorescent Organic Light-Emitting Device with above 31% External Quantum Efficiency. Adv. Mater. 2014, 26, 8107−8113. (22) Zhao, Y.; Chen, J.; Ma, D. Ultrathin Nondoped Emissive Layers for Efficient and Simple Monochrome and White Organic LightEmitting Diodes. ACS Appl. Mater. Interfaces 2013, 5, 965−971. (23) Wu, S.; Li, S.; Qi, S.; Huang, C.; Fung, M. K. Highly Efficient White Organic Light-Emitting Diodes with Ultrathin Emissive Layers and a Spacer-Free Structure. Sci. Rep. 2016, 6, No. 25821. (24) Liu, B.; Tao, H.; Wang, L.; Gao, D.; Liu, W.; Zou, J.; Xu, M.; Ning, H. L.; Peng, J. B.; Cao, Y. High-Performance Doping-free Hybrid White Organic Light-Emitting Diodes: the Exploitation of Ultrathin Emitting Nanolayers (
