Efficient non-doped monochrome and white phosphorescent organic light-emitting diodes based on ultrathin emissive layers

Organic Electronics 26 (2015) 451–457

Contents lists available at ScienceDirect

Organic Electronics journal homepage: www.elsevier.com/locate/orgel

Efficient non-doped monochrome and white phosphorescent organic light-emitting diodes based on ultrathin emissive layers Kaiwen Xue, Ren Sheng, Yu Duan, Ping Chen ⇑, Bingye Chen, Xiao Wang, Yahui Duan, Yi Zhao State Key Laboratory on Integrated Optoelectronics, College of Electronic Science and Engineering, Jilin University, Changchun 130012, People’s Republic of China

a r t i c l e

i n f o

Article history: Received 4 June 2015 Received in revised form 13 August 2015 Accepted 18 August 2015 Available online 24 August 2015 Keywords: White organic light-emitting diodes Non-doped Phosphorescent High efficiency Ultrathin layer

a b s t r a c t Efficient red, orange, green and blue monochrome phosphorescent organic light-emitting diodes (OLEDs) with simplified structure were fabricated based on ultrathin emissive layers. The maximum efficiencies of red, orange, green and blue OLEDs are 19.3 cd/A (17.3 lm/W), 45.7 cd/A (43.2 lm/W), 46.3 cd/A (41.6 lm/W) and 11.9 cd/A (9.2 lm/W). Moreover, efficient and color stable white OLEDs based on two complementary colors of orange/blue, three colors of red/orange/blue, and four colors of red/orange/ green/blue were demonstrated. The two colors, three colors and four colors white OLEDs have maximum efficiencies of 30.9 cd/A (27.7 lm/W), 30.3 cd/A (27.2 lm/W) and 28.9 cd/A (26.0 lm/W), respectively. And we also discussed the emission mechanism of the designed monochrome and white devices. Ó 2015 Elsevier B.V. All rights reserved.

1. Introduction White organic light-emitting diodes (OLEDs) exhibit great potential application in the solid-state lighting and full-color display due to their light weight, flexibility, high resolution, wide viewing angle [1–6]. Lots of device concepts have been proposed to improve the performance of white OLEDs, such as doping multiple dyes into one single emission layer (EML) [7–10], using multiple EMLs in which each layer emits different color light [11–14], using excimer or exciplex emission [15,16], using tandem or stacked structure [17]. So far, nearly all state of the art monochrome and white phosphorescent OLEDs were fabricated by the method of coevaporation [18–20]. As we all know, the doping method is beneficial to decreasing the severe concentration quenching effect, leading to high efficiency and low efficiency roll-off. But it also puts a few obstacles in the commercialization, such as accurately controlling the doping concentration by adjusting relative deposition rates of the host and guest, searching for the appropriate bipolar host materials with high energy to efficiently confine the excitons [21,22], which aggravate the complication in the manufacture process and increase the cost of the devices fabrication [23]. Since the non-doped white OLEDs were firstly demonstrated by Tsuji et al. [24], they have attracted increasing interest [25–31]. There are considerable advantages to be gained by using non⇑ Corresponding author. E-mail address: [email protected] (P. Chen). http://dx.doi.org/10.1016/j.orgel.2015.08.017 1566-1199/Ó 2015 Elsevier B.V. All rights reserved.

doped method to fabricate white OLEDs, such as the save of the materials, the simplified fabrication process, the better repeatability and the lower production cost, which make non-doped method more suitable for the industrial application for solid-state lighting and display. Liu et al. fabricated non-doped phosphorescent OLEDs based on double quantum-well structures with maximum power efficiency of 30.5 lm/W [28]. Chen et al. reported non-doped white OLEDs using ultrathin bluish-green and red dyes with maximum current efficiency of 7.4 cd/A [29]. Recently, Zhao et al. demonstrated efficient ultrathin nondoped blue/orange, blue/green/red, and blue/green/orange/red white OLEDs with maximum current efficiencies of 41.3 cd/A, 34.6 cd/A and 21.1 cd/A, respectively [31]. Moreover, the emission mechanism of the OLEDs using ultrathin nondoped emissive layers was still not clear. Generally speaking, to obtain efficient white non-doped OLEDs, the carrier injection layer, the transport layer and the spacer or interlayer between different color dyes are always necessary. But the more organic function materials used, the higher the cost. So it’s very meaningful and necessary to develop white OLEDs with simplified device structure. In this paper, we have performed a systematic study on how to realize efficient and stable two colors, three colors and four colors non-doped white OLEDs based on ultrathin emission layers. And all designed orange/blue, red/orange/blue and red/orange/green/blue white OLEDs exhibit very stable spectra and high efficiencies. Moreover, the working mechanism of the OLEDs was also discussed and we concluded that the position of the recombination region located at the interface of hole transport layer (HTL) and

452

K. Xue et al. / Organic Electronics 26 (2015) 451–457

the electron transport layer (ETL). The stable energy transfer was the determinate factor for achieving such excellent spectra and high efficiency. 2. Experimental Fig. 1 shows the chemical structures of red, orange, green and blue phosphorescent dyes, the detailed energy level diagram of the materials, and the structures of the two colors, three colors, and four colors white OLEDs. All the organic materials used in our work were purchased from Luminescence Technology Corporation. Indium tin oxide (ITO) with sheet resistance of 20 X per square, covered by 2 nm MoO3 was used as anode. ITO substrates were cleaned in an ultrasonic bath with acetone, ethanol and deionized water in sequence, and then treated in O2 plasma. 1,3-Bis (carbazol-9-yl)benzene (MCP) and 1,3-Bis[3,5-di(pyridin-3-yl)phe nyl]benzene (BmPyPhB) served as the HTL and ETL, respectively. In order to further simplify the device structure, MCP was also used as the spacer between EMLs to adjust the relative emission intensity. Bis(2-methyldibenzo[f,h]quinoxaline)(acetylacetonate)-iri dium(III) [Ir(MDQ)2(acac)], Bis(4-phenylthieno[3,2-c]pyridinato-N ,C20 ) acetylacetonate iridium(III) (PO-01), Tris(2-phenylpyridine)iri dium(III) [Ir(ppy)3], and Bis(3,5-difluoro-2-(2-pyridyl) phenyl-(2-c arboxypyridyl)iridium(III) (FIrPic) were used as red, orange, green and blue phosphorescent dyes, respectively. All OLEDs were fabricated in a single chamber tool with the substrate temperature of 30 °C by high vacuum thermal evaporation under the pressure of 3  104 Pa. In this work, we used in situ quartz-crystal monitor to monitor the thickness of the vacuum depositions. Typical deposition rates range from 0.001 to 10 Å/s. The deposition rate

could be controlled by temperature variation of the source material oven. The HTL and ETL were grown at the deposition rate of 1 Å/s, while the phosphorescent ultrathin EMLs were deposited at the rate of 0.01 Å/s. 100 nm Al cathode was defined through a shadow mask with an active area of 3 mm  3 mm. The electroluminescence (EL) characteristics were measured using a programmable Keithley 2400 source measure unit and PR655 spectroscan spectrometer. All the devices were measured without encapsulation under ambient conditions.

3. Results and discussion To demonstrate efficient OLEDs with ultrathin layer, it is very important to place the ultrathin EML at the main exciton recombination region to efficiently utilize the electrically generated excitons. We firstly fabricated four devices as shown in Fig. 2(a) to investigate the position of the recombination region. Fig. 2(b)–(d) shows the normalized EL spectra at the voltage of 5 V, the current efficiency–voltage, and luminance–voltage characteristics of devices A–D. As can be seen that devices A–C all show a primary orange emission from PO-01. The peak current efficiencies of devices A–C are 5.3 cd/A, 18.3 cd/A and 37.4 cd/A, indicating that the majority of excitons are generated at the narrow interface of MCP/BmPyPhB. The efficient orange emission of PO-01 in device B may be attributed to two possible channels: (i) the generated excitons at the narrow region diffuse through BmPyPhB or MCP and migrate into the ultrathin EML where they transfer to PO-01. (ii) From the energy level diagram in Fig. 1(a), we can see that PO-01 acts as deep traps for both holes and electrons. So the charge carriers

Fig. 1. The chemical structures of phosphorescent dyes, (a) the detailed energy level diagram of the materials and the structures of (b) two colors, (c) three colors, and (d) four colors white OLEDs. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

K. Xue et al. / Organic Electronics 26 (2015) 451–457

453

Fig. 2. (a) The device structures, (b) normalized EL spectra at 5 V, (c) the current efficiency–voltage and (d) the luminance–voltage characteristics of devices A–D.

could be directly trapped by PO-01, resulting in the excitons directly formed on it. However, when we introduced an ultrathin EML of Ir(MDQ)2(acac) in device D to efficiently harvest the generated excitons at the narrow interface of MCP/BmPyPhB, we can only observe one primary red emission originating from Ir (MDQ)2(acac) in the spectra. The absence of the orange emission implies that directly excitons generated on ultrathin PO-01 layer is negligible. So we can conclude that the emission of PO-01 in device B mainly due to the energy transfer from MCP excitons diffusion from the interface of MCP/BmPyPhB. The negligible chargetrapping effect in such the device structure is attributed to the following reason. Firstly, we can tell from the energy level diagram that most of holes and electrons accumulate at the interface of MCP/BmPyPhB, but few electrons could be injected into MCP because of the big offset of LUMO between MCP and BmPyPhB, consistent with the poor electron-transporting property of MCP. The lack of electron in MCP layer decreases the probability of directly exciton generation on PO-01. And furthermore, the ultrathin EML structure also reduces the effect of direct charge trapping. Based on the above results, we conclude the recombination region mainly locates at the interface of MCP/BmPyPhB and both chargecarriers and excitons could well confine in such device structure. To investigate the impact of ultrathin EML thickness on the devices performance, we fabricated efficient non-doped red, orange, green and blue phosphorescent OLEDs with the structures of ITO/MoO3 (2 nm)/MCP (60 nm)/ultrathin EML (X nm)/BmPyPhB (30 nm)/Liq (1 nm)/Al (100 nm), where X was varied, being 0.05 nm, 0.1 nm, 0.15 nm, 0.5 nm, and the ultrathin EML was chosen as Ir(MDQ)2(acac), PO-01, Ir(ppy)3 and FIrPic, corresponding to the red, orange, green and blue OLEDs. Fig. 3 shows the current efficiency–luminance characteristics of the monochrome OLEDs with varying thicknesses of ultrathin EML. The detailed EL perfor-

mances are summarized in Table 1. The maximum efficiencies of red, green and blue OLEDs are 19.3 cd/A (17.3 lm/W), 46.3 cd/A (41.6 lm/W), 11.9 cd/A (9.2 lm/W), all obtained for the devices with 0.1 nm ultrathin EML. While the orange device with the optimum PO-01 thickness of 0.15 nm has the maximum efficiencies of 45.7 cd/A (43.2 lm/W). Though the orange device has lower efficiency than the recently reported work by Li et al. [18] and Cheng et al. [19], the state-of-art performance makes our orange OLEDs among the best reported results. The lower efficiency of the OLEDs with thicker EML is due to the serious concentration quenching effect [32]. Moreover, the red, orange, green and blue monochrome OLEDs with optimum thickness EML all display low efficiency roll-off with an efficiency of 13.3 cd/A, 38.9 cd/A, 41.7 cd/A and 6.6 cd/A at 5000 cd/m2. It’s known that charge balance and exciton annihilation are the main mechanisms of the efficiency roll-off. The electron mobility of BmPyPhB is as high as 104 cm2/Vs, which agrees with the hole mobility of MCP in the same order [33]. The remarkable feature of the low efficiency roll-off indicates the good charge balance in such the device structure. The slightly decreasing of the efficiency is due to the triplet–polaron annihilation and/or triplet–triplet annihilation. The inset shows the normalized EL spectra at 3.5 V of the monochrome devices. White light emission can be easily produced by mixing two complementary colors because organic materials have broader emission spectra than inorganic materials [34–38]. So we firstly fabricated six non-doped devices based on two complementary colors dyes of PO-01 and FIrPic with the structures of ITO/MoO3 (2 nm)/MCP [(60–X) nm]/PO-01 (0.15 nm)/MCP (X nm)/FIrPic (0.1 nm)/BmPyPhB (30 nm)/Liq (1 nm)/Al (100 nm), where X was varied being 1, 1.5, 2, 3, 3.5, and 4, corresponding to devices D1, D2, D3, D4, D5, and D6 respectively.

454

K. Xue et al. / Organic Electronics 26 (2015) 451–457

Fig. 3. The current efficiency–luminance characteristics of (a) red, (b) orange, (c) green and (d) blue OLEDs. The inset shows the normalized EL spectra at 3.5 V. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Table 1 Performances of the optimized red, orange, green, and blue OLEDs. Devices optimized

Turn-on voltage (V)

Current efficiency (cd/A)

Power efficiency (lm/W)

gmax/g100/g1000/g5000

gmax/g100/g1000/g5000

Red Orange Green Blue

3.00 2.97 2.95 3.23

19.3/18.2/17.9/13.3 45.7/43.9/44.0/38.9 46.3/40.6/45.4/41.7 11.9/10.1/11.9/6.6

17.3/17.2/14.6/9.0 43.2/42.0/37.3/28.6 41.6/39.3/38.8/30.9 9.2/8.7/8.5/3.6

gmax: maximum efficiency, g100, g1000, g5000: efficiencies at 100 cd/m2, 1000 cd/m2, 5000 cd/m2.

Fig. 4 shows the EL performances of devices D1–D6. The normalized EL spectra of the six devices are shown in Fig. 4(a). We can clearly see two primary peaks at 468 nm and 560 nm originating from FIrPic and PO-01, respectively. The color of the emission light changes from yellowish white to bluish white with the thickness of MCP spacer increasing from 1 nm to 4 nm. When X is 1, 1.5 and 2 nm, more energy is transferred from FIrPic to PO-01, leading to the strong orange emission. As can be seen, the blue emission initiates to dominate the spectra for device D6 which can be explained that the energy transfer is suppressed by thicker MCP spacer. Fig. 4 (c) and (d) show the normalized spectra of white devices D4 and D5 under different luminance. Both the white devices D4 and D5 show very stable white emission spectra with satisfied Commission International de L’Eclairage (CIE) coordinates. Fig. 4(b) denotes current efficiency–voltage–luminance characteristics of devices D1– D6. The current efficiency decreases markedly with X increasing. And white device D4 has maximum current efficiency and power efficiency of 30.9 cd/A and 27.7 lm/W. Generally speaking, for three colors (red/green/blue or red/ orange/blue) white OLEDs, the red-shift with increasing voltage is always observed due to the direct charge trapping effect of the

red dyes [2,39,40]. Considering the negligible charge-trapping effect of the ultrathin layer in the designed structure, we inserted Ir(MDQ)2(acac) ultrathin layer into device D4 to fabricate red/ orange/blue white OLEDs with the structures of ITO/MoO3 (2 nm)/MCP [(57–Y) nm]/Ir(MDQ)2(acac) (0.1 nm)/MCP (Y nm)/ PO-01 (0.15 nm)/MCP (3 nm)/FIrPic (0.1 nm)/BmPyPhB (30 nm)/ Liq (1 nm)/Al (100 nm). In order to realize balanced white emission, a thin MCP spacer with different thickness is placed between orange and red ultrathin EMLs. Here, Y is 0.5, 2, 3 and 4 corresponding to devices T1, T2, T3 and T4. Fig. 5 depicts the EL properties of red/orange/blue white OLEDs T1–T4. The EL spectra normalized to the blue emission peak are shown in Fig. 5(a). In device T1 with only 0.5 nm MCP spacer, the spectra exhibit two main emission peaks from the Ir(MDQ)2(acac) and FIrPic. The weak intensity of PO-01 is attributed to the efficient energy transfer from PO-01 layer to Ir(MDQ)2(acac). The orange emission initiates to dominate the EL spectra with the thickness of MCP spacer increasing. The observation can be explained by that the energy transfer from PO-01 to Ir(MDQ)2(acac) is suppressed by thicker MCP spacer. Fig. 5(b) shows the current efficiency–lumi nance–power efficiency characteristics of devices T1–T4. The efficiency of device T1 is the lowest due to the absence of the efficient PO-01 emission. It can be seen that the efficiency increases clearly with Y increasing. Table 2 summarizes the EL performances of the devices T1–T4. Device T3 has maximum current efficiency and power efficiency of 25.8 cd/A and 23.1 lm/W and device T4 has the highest efficiencies of 30.3 cd/A and 27.2 lm/W. The normalized EL spectra of white devices T3 and T4 at different driving voltages are shown in Fig. 5(c)–(d). We can see both the spectra show only a little change. The observation is mainly due to the stable sequential energy transfer from higher energy dyes to the lower ones. It’s so complicated to construct red/orange/green/blue four colors white OLEDs based on traditional multiple EMLs or one single

K. Xue et al. / Organic Electronics 26 (2015) 451–457

455

Fig. 4. (a) Normalized EL spectra at 7 V, (b) the current efficiency–voltage–luminance characteristics of devices D1–D6. Normalized EL spectra of (c) device D4 and (d) device D5 under different luminance.

Fig. 5. (a) Normalized EL spectra at 7 V, (b) the current efficiency–luminance–power efficiency characteristics of devices T1–T4. Normalized EL spectra of (c) device T3 and (d) device T4 under different voltages.

456

K. Xue et al. / Organic Electronics 26 (2015) 451–457

Table 2 Performances of devices T1–T4.

a

Devices

a

T1 T2 T3 T4

(0.47, (0.44, (0.43, (0.41,

CIE (x, y)

0.38) 0.41) 0.43) 0.43)

Luminance (max)

Current efficiency (cd/A) gmax/g100/g1000

Power efficiency (lm/W) gmax/g100/g1000

16,580 18,810 20,200 20,860

17.9/16.4/13.6 23.7/22.3/18.4 25.8/25.1/20.7 30.3/27.3/23.3

16.1/12.8/8.6 21.3/17.6/12.0 23.1/19.8/13.6 27.2/21.8/15.4

At 1000 cd/m2.

EML doped with four dyes that there are only a few relevant results [41–44]. Based on the above result, we designed red/orange/green/ blue white device just by introducing a green ultrathin EML between FIrPic and PO-01 EMLs. And the structure of four colors non-doped white device is ITO/MoO3 (2 nm)/MCP (51 nm)/Ir (MDQ)2(acac) (0.1 nm)/MCP (3 nm)/PO-01 (0.15 nm)/MCP (3 nm)/Ir(ppy)3 (0.1 nm)/MCP (3 nm)/FIrPic (0.1 nm)/BmPyPhB

(30 nm)/Liq (1 nm)/Al (100 nm). The ultrathin EMLs are separated by 3 nm MCP spacer to obtain the balanced white emission by controlling the energy transfer. And we also fabricated one conventional doped white device with multiple EMLs as the control device. The structure is as follows: ITO/MoO3 (2 nm)/NPB (40 nm)/CBP (2 nm)/CBP: 6% Ir(MDQ)2(acac) (3 nm)/CBP: 10% Ir (ppy)3 (5 nm)/MCP: BmPyPhB (1: 1, 2 nm)/MCP: 10% FIrPic: 0.5% PO-01 (10 nm)/BmPyPhB (30 nm)/Liq (1 nm)/Al (100 nm). NPB and CBP stand for N,N0 -di(naphthalen-1-yl)-N,N0 -diphenylbenzidine and 4,40 -bis(N-carbazolyl)biphenyl, respectively. The normalized EL spectra of the two four colors white OLEDs under different voltages are depicted in Fig. 6(a)–(b). The nondoped white device emits warm yellow–white light with CIE coordinates of (0.342, 0.457) at 7 V. Clearly to see that the non-doped white device shows excellent spectra stability within a large voltage range due to stable sequential energy transfer, as shown in Fig. 7. While the spectra of the doped white device depend on the applied voltage. From energy level diagram, we can tell that Ir(MDQ)2(acac) acts as deep trap center for both holes and

Fig. 6. Normalized EL spectra of (a) non-doped and (b) doped four colors white devices at different voltages. (c) Current efficiency–luminance–power efficiency, (d) current density–voltage–luminance characteristics of non-doped and doped white devices.

Fig. 7. Operational principles of four colors red/orange/green/blue white OLEDs.

K. Xue et al. / Organic Electronics 26 (2015) 451–457

electrons. So the dominant red emission in the doped white device at low voltage is due to the direct charge trapping effect of Ir (MDQ)2(acac). The great shift of the spectra in the doped white device is mainly attributed to the fact that the position of the exictons recombination region changes with the increasing voltages [45]. Moreover, the trapping sites are filled at high voltage and less excitons are generated directly on Ir(MDQ)2(acac), which also causes the decreasing of the red emission. The EL characteristics of the two white devices are compared in Fig. 6(c)–(d). The maximum efficiencies of the non-doped white device are 26.0 lm/W and 28.9 cd/A, while those of the doped device are 15.8 lm/W and 25.6 cd/A. Note that the power efficiency of the non-doped white device is greatly improved by 39% compared to the doped one. The lower power efficiency of the white device with multiple doped EMLs is owing to the higher operational voltage. The energy barriers for holes and electrons injection at the interfaces in doped white device result in the carriers accumulated at the narrow interfaces, leading to higher turn-on voltage and lower power efficiency. 4. Conclusion In summary, we reported simplified non-doped monochrome and white OLEDs based on ultrathin EMLs. All the designed white OLEDs have very stable spectra due to the avoidance of the movement of charges recombination region and the elimination of the direct charge trapping effect on ultra-thin dyes. Moreover, the orange/blue, red/orange/blue and red/orange/green/blue white OLEDs have maximum efficiencies of 30.9 cd/A (27.7 lm/W), 30.3 cd/A (27.2 lm/W) and 28.9 cd/A (26.0 lm/W), respectively. Acknowledgments The work was funded by the National Natural Science Foundation of China (61377026, 61275024, 61275033, 61274002), Program of international science and technology cooperation (2014DFG12390), the Ministry of Science and Technology of China (2013CB834802), the National High Technology Research and Development Program of China (2011AA03A110) and the Project of Science and Technology Development Plan of Jilin Province (20140101204JC, 20130206020GX, 20150101045JC, 20140520071JH). References [1] J. Kido, K. Hongawa, K. Okuyama, K. Naga, Appl. Phys. Lett. 64 (1994) 815. [2] S. Chen, Q. Wu, M. Kong, X. Zhao, Z. Yu, P. Jia, W. Huang, J. Mater. Chem. C 1 (2013) 3508. [3] J. Kido, M. Kimura, K. Nagai, Science 267 (1995) 1332. [4] L. Ying, C.L. Ho, H. Wu, Y. Cao, W.Y. Wong, Adv. Mater. 26 (2014) 2459.

457

[5] M.C. Gather, A. Köhnen, K. Meerholz, Adv. Mater. 23 (2011) 233. [6] Y. Liu, L. Cui, M. Xu, X. Shi, D. Zhou, Z. Wang, Z. Jiang, L. Liao, J. Mater. Chem. C 2 (2014) 2488. [7] Z. Wang, Y. Lou, S. Naka, H. Okada, Appl. Phys. Lett. 97 (2010) 203302. [8] J.H. Jou, Y.S. Chiu, C.P. Wang, R.Y. Wang, H.C. Hu, Appl. Phys. Lett. 88 (2006) 193501. [9] G. Li, J. Shinar, Appl. Phys. Lett. 83 (2003) 5359. [10] P.I. Shih, C.F. Shu, Y.L. Tung, Y. Chi, Appl. Phys. Lett. 88 (2006) 251110. [11] J.H. Seo, J.H. Seo, J.H. Park, Y.K. Kim, J.H. Kim, G.W. Hyung, K.H. Lee, S.S. Yoon, Appl. Phys. Lett. 90 (2007) 203507. [12] Y.S. Wu, S.W. Hwang, H.H. Chen, M.T. Lee, W.J. Shen, C.H. Chen, Thin Solid Films 488 (2005) 265. [13] Y.R. Sun, N.C. Giebink, H. Kanno, B.W. Ma, M.E. Thompson, S.R. Forrest, Nature 440 (2006) 908. [14] S. Tokito, T. Iijima, T. Tsuzuki, F. Sato, Appl. Phys. Lett. 83 (2003) 2459. [15] J. Feng, F. Li, W. Gao, S.Y. Liu, Y. Liu, Y. Wang, Appl. Phys. Lett. 78 (2001) 3947. [16] J. Thompson, R.I.R. Blyth, M. Mazzeo, M. Anni, G. Gigli, R. Cingolani, Appl. Phys. Lett. 79 (2001) 560. [17] L. Ding, Y.Q. Sun, H. Chen, F. Zu, Z. Wang, L.A. Liao, J. Mater. Chem. C 2 (2014) 10403. [18] G. Li, D. Zhu, T. Peng, Y. Liu, Y. Wang, M.R. Bryce, Adv. Funct. Mater. 24 (2014) 7420. [19] G. Cheng, S.C.F. Kui, W.H. Ang, M.Y. Ko, P.K. Chow, C.L. Kwong, C.C. Kwok, C. Ma, X. Guan, K.H. Low, S.J. Su, C.M. Che, Chem. Sci. 5 (2014) 4819. [20] S.J. Su, C. Cai, J. Kido, Chem. Mater. 23 (2010) 274. [21] Z.S. Zhang, P.R. Yan, S.Z. Yue, Y. Chen, Q.Y. Wu, D.L. Qu, R.D. Guo, P. Wang, Y. Zhao, S.Y. Liu, Org. Electron. 14 (2013) 2172. [22] D. Hu, F. Shen, H. Liu, P. Lu, Y. Lv, D. Liu, Y. Ma, Chem. Commun. 48 (2012) 3015. [23] C.T. Chen, Chem. Mater. 16 (2004) 4389. [24] T. Tsuji, S. Naka, H. Okada, H. Onnagawa, Appl. Phys. Lett. 81 (2002) 3329. [25] M.T. Lee, M.T. Chu, J.S. Lin, M. Tseng, J. Phys. D Appl. Phys. 43 (2010) 442003. [26] Y. Divayana, X.W. Sun, Org. Electron. 10 (2009) 320. [27] S. Liu, B. Li, L. Zhang, H. Jiang, Appl. Phys. Lett. 97 (2010) 083304. [28] S. Liu, B. Li, L. Zhang, S. Yue, Appl. Phys. Lett. 98 (2011) 163301. [29] S.M. Chen, Z.J. Zhao, B.Z. Tang, H.S. Kwok, J. Phys. D Appl. Phys. 43 (2010) 095101. [30] Y.M. Yin, J. Yu, H.T. Cao, L.T. Zhang, H.Z. Sun, W.F. Xie, Sci. Rep. 4 (2014) 06754. [31] Y.B. Zhao, J.S. Chen, D.G. Ma, ACS Appl. Mater. Interfaces 5 (2013) 965. [32] C.L. Wu, C.H. Chang, Y.T. Chang, C.T. Chen, C.T. Chen, C. Su, J. Mater. Chem. C 2 (2014) 7188. [33] J.H. Jou, W.B. Wang, S.Z. Chen, J.J. Shyue, M.F. Hsu, C.W. Lin, S.W. Shen, C.J. Wang, C.P. Liu, C.T. Chen, M.F. Wu, S.W. Liu, J. Mater. Chem. 20 (2010) 8411. [34] G. Cheng, Y.F. Zhang, Y. Zhao, S.Y. Liu, Y.G. Ma, Appl. Phys. Lett. 88 (2006) 083512. [35] P. Chen, W.F. Xie, J. Li, T. Guan, Y. Duan, Y. Zhao, S.Y. Liu, Appl. Phys. Lett. 91 (2007) 023505. [36] U.S. Bhansali, H. Jia, I.W.H. Oswald, B.E. Gnade, W. Chen, M.A. Omary, Appl. Phys. Lett. 100 (2012) 183305. [37] H. Kanno, R.J. Holmes, Y. Sun, S.K. Cohen, S.R. Forrest, Adv. Mater. 18 (2006) 339. [38] B.W. D’Andrade, S.R. Forrest, Adv. Mater. 16 (2004) 1585. [39] C.H. Hsiao, Y.H. Lan, P.Y. Lee, T.L. Chiu, J.H. Lee, Org. Electron. 12 (2011) 547. [40] Q. Wang, J. Ding, D. Ma, Adv. Mater. 21 (2009) 2397. [41] Y.L. Chang, Y. Song, Z. Wang, M.G. Helander, J. Qiu, L. Chai, Z.W. Liu, G.D. Scholes, Z.H. Lu, Adv. Funct. Mater. 23 (2013) 705. [42] S.M. Zhang, S.Z. Yue, Q.Y. Wu, Z.S. Zhang, Y. Chen, X.H. Wang, Z.Y. Liu, G.H. Xie, Q. Xue, D.L. Qu, Y. Zhao, S.Y. Liu, Org. Electron. 14 (2013) 2014. [43] J. Yu, H. Lin, F. Wang, Y. Lin, J. Lin, J. Zhang, H. Zhang, Z. Wang, B. Wei, J. Mater. Chem. 22 (2012) 22097. [44] N. Sun, Y. Zhao, F. Zhao, Y. Chen, D. Yang, J. Chen, D. Ma, Appl. Phys. Lett. 105 (2014) 013303. [45] P. Chen, Q. Xue, W. Xie, Y. Duan, G. Xie, Y. Zhao, J. Hou, S. Liu, L. Zhang, B. Li, Appl. Phys. Lett. 93 (2008) 153508.