Surface treatment on polyethylenimine interlayer to improve inverted OLED performance
Wei Chang-Ting1, 2, Zhuang Jin-Yong2, Chen Ya-Li1, †, , Zhang Dong-Yu2, ‡, , Su Wen-Ming2, Cui Zheng2
Department of Chemistry, Shanghai University, Shanghai 200444, China
Printable Electronics Research Center, Suzhou Institute of Nano-Tech and Nano-Bionics, Chinese Academy of Sciences, Suzhou 215123, China

 

† Corresponding author. E-mail: ylchen@staff.shu.edu.cn

‡ Corresponding author. E-mail: dyzhang2010@sinano.ac.cn

Project supported by the National Key Basic Research Project of China (Grant No. 2015CB351901), the Strategic Priority Research Program of the Chinese Academy of Sciences (Grant No. XDA09020201), the Youth Innovation Promotion Association of Chinese Academy of Sciences (Grant No. 2013206), the National Natural Science Foundation of China (Grant No. 21402233), and the Natural Science Foundation of Jiangsu Province, China (Grant Nos. BK2012631 and BK20140387).

Abstract
Abstract

Polyethylenimine (PEI) interlayer rinsing with different solvents for inverted organic light emitting diodes (OLEDs) is systematically studied in this paper. In comparison with the pristine one, the maximum current efficiency (CEmax) and power efficiency (PEmax) are enhanced by 21% and 22% for the device rinsing by ethylene glycol monomethyl ether (EEA). Little effect is found on the work function of the PEI interlayer rinsed by deionized water (DI), ethanol (EtOH), and EEA. On the other hand, the surface morphologies of PEI through different solvent treatments are quite different. Our results indicates that the surface morphology is the key to improving the device performance for IOLED as the work function of PEI keeps stable.

1. Introduction

Over the past decades, organic light emitting diodes (OLEDs) have attracted a great deal of attention due to their intriguing applications in full-color display and solid-state lighting since the pioneering work of Tang and VanSlyke.[18] Recently, active matrix organic light emitting diode (AMOLED) display as a new-generation all-solid technology has been widely employed in smart phones, tablet personal computers (PCs), digital cameras, large-size televisions (TVs), and flexible wearable devices due to its high performance and low cost.[912] It is well known that the performance (particularly the stability and uniformity) of the thin-film transistor (TFT) in each AMOLED pixel is vital to achieving full color displays.[9,13] Generally, the oxide TFT backplane is adapted to the commercial AMOLED backplane due to its better uniformity, lower processing temperature, higher carrier mobility, higher transparency, and potentially better electrical stability than the low-temperature polycrystalline silicon TFT and amorphous silicon TFT backplane. The inverted OLED (IOLED) structure with bottom cathode is superior to the conventional OLED structure on account of the oxide TFT backplane possessing the n-type transistor characteristics. Simultaneously, the bottom cathode and n-type drain can be connected directly in IOLED structure, which can increase the stability and reduce the driving voltage.[14] In addition, the conventional structure of bottom-emitting OLED (IBOLED) device generally use indium tin oxide (ITO) as a transparent anode, in which the aperture ratio is only in a range from 30% to 50%. In contrast, the inverted (IOLED) device has a larger aperture ratio, higher stability, longer lifetime, and fewer image sticking phenomena than the conventional OLED.[9,1315]

Solution-processing for OLED fabrication has been developed rapidly for its low cost, large scale manufacturing and large-size OLED displays.[1619] As is well known, ITO is widely used as the anode for its high work function of about 4.8 eV, when used as the cathode, a 1.6–2.0 eV injection barrier exists between the lowest unoccupied molecular orbital (LUMO) level of the electron transport material (ETM) and that of the cathode. Therefore, the development of suitable solution method for processing electron injection materials is essential. It has been reported that by using the zinc oxide (ZnO) nanoparticles (NPs), the work function of ITO/ZnO can be tuned from 4.8 eV to 4.1 eV. However, there still exists an injection barrier between the organic layer and ZnO.[20] In order to solve this problem, Zhou et al. reported that introducing polyethyleneimine (PEI) between the metal oxides (such as ITO, ZnO), the organic layer can significantly improve the device performance.[21] The thin film layer of PEI will form interface dipoles and lower the work function. Since then, PEI has been used as an effective electron-injecting layer in inverted devices both for OLEDs and quantum dots light emitting diodes(QDLEDs).[2227] In 2014, Höfle et al. reported the deposition of PEI on the ZnO as a combined interlayer in an inverted polymer light-emitting diode, which can enhance the electron injection from the cathode into the emitting layer (EML). It has been reported that the PEI film was rinsed with water or ethanol to remove PEI surplus, though there is not any relevant mechanism described yet.[28,29]

In this paper, the properties of PEI layer treated with deionized water (DI), ethanol (EtOH) and ethylene glycol monomethyl ether (EEA) are systematically investigated. The work function and surface morphology of PEI are measured. Our work indicates that the solvent rinsing has little effect on the work function. On the other hand, the surface morphology of PEI is quite different from various solvent treatments. The electroluminescent properties of both fluorescent and phosphorescent OLEDs based on PEI are characterized. A maximum current efficiency and power efficiency (PEmax) of EEA treated phosphorescent devices (CEmax) reach 30.54 cd·A−1 and 13.90 lm·W−1, respectively, which are increased by 21% and 22% compared with that of the untreated one. Based on the investigations of the PEI layer, the significant device improvement can be attributed to better interlayer contact with the neighboring organic layer.

2. Experiment
2.1. Materials

All chemicals and reagents were used as received from commercial sources without further purification. PEI (Mw = 25000 g/mol) was purchased from Sigma-Aldrich. EtOH was of analytically pure grade from Sinopharm Chemical Reagent Co., Ltd. (China), EEA was of analytically pure grade from Sigma-Aldrich (USA). DI with resistance of 18 MΩ· cm@25 °C used all through the experiment was purified by Direct-Q5uv manufactured by Merck Millipore. Alq3, Ir(ppy)3, 4,4-bis(9-carba-zolyl)-2,2-biphenyl(CBP), [N-(1-naphthy1)-npheny1-amino] biphenyl (NPB) and 1-bis [(di-4-tolylamino) phenyl]cyclohexane (TAPC) were purchased from Nichem Fine Technology Co. Ltd. MoO3 was purchased from Alfa Aesar. ZnO NPs were synthesized according to the previously reported procedures, and were dispersed in acetone with a concentration of 10 mg·mL−1 prior to use.[30]

2.2. General information

The film surface morphology was measured with Veeco Dimension 3100 (USA) at ambient temperature in a tapping mode. The work function was measured on the Veeco Dimension 3100 Kelvin probe force microscope (USA) at ambient temperature. Highly ordered pyrolytic graphite, whose work function in air is 4.6 eV, was taken as the reference. The thickness values of solution-processed films were measured using an AlphaStep profilometer (Veeco, Dektak150). All electrical testing measurements were performed under ambient conditions without further encapsulation. The electroluminescent spectra were measured with a Spectra Scan PR655. The current–voltage (IV) and luminance–voltage (LV) characteristics were measured with a computer controlled Keithley 2400 Sourcemeter. The active area of the device was 2 mm×2 mm, and only the luminance in the forward direction was measured.

2.3. Device fabrication of IOLED devices

The chemical structures, device configurations, and the energy level diagrams of IOLEDs with the PEI rinsing with EEA are shown in Fig. 1 and Fig. 2. OLEDs were fabricated on an ITO substrate with a sheet resistance of ∼20 Ω/sq. Prior to use, the ITO substrate was degreased with solvents and treated with O2-plasma for 3 min. First, a 30-nm ZnO layer was deposited by spin-coating (2300 rpm for 60 s) on a pre-cleaned ITO substrate and then baked at 120 °C for 10 min in the glove box. The PEI was dissolved in EEA with a concentration of 0.4 wt% and spin-coated at 5000 rpm for 50 s onto the ZnO layer.[28] After the coating of PEI, the substrate was baked at 100 °C for 10 min to remove residual solvent. The resulting solvents, including DI (in air), EtOH (in N2) and the EEA (in N2), were spin-coated on PEI film after annealing, at 2000 rpm for 60 s and then dried at 100 °C for 10 min in nitrogen atmosphere. Then the samples were directly loaded into the evaporation system. After the deposition of PEI, the EML of Alq3 (70 nm), a hole-transporting layer (HTL) of NPB (30 nm), a hole-injection layer (HIL) of MoO3 (10 nm) and Al were deposited layer by layer in the vacuum of 5×10−4 Pa. As for the green phosphorescent device, the EML consists of CBP and Ir(ppy)3(8%, 40 nm), and TAPC (30 nm) as the HTL.

Fig. 1. Chemical structures of PEI and the IOLEDs structures based on Alq3 and Ir(ppy)3 as EML.
Fig. 2. Energy level diagrams of (a) Alq3, (b) Ir(ppy)3 IOLEDs with the PEI rinsing with EEA.
3. Results and discussion

Since the PEI layer is rinsed by different solvents, the work function and film mophology are investigated. The thickness of untreated PEI film is about 10 nm, then decreases to 8 nm after being treated by solvent, which is in accordance with the reported result.[22] Firstly, the work functions of ZnO/PEI with different-solvent surface treatment were performed by Kelvin probe force microscopy (KFM). As shown in Table 1, the work function values of ZnO/PEI rinsed with DI, EtOH, and EEA are 3.40, 3.37, and 3.35 eV, respectively, in comparison with 3.50 eV of the untreated (W/O) one. As a result, the work function of ZnO/PEI is slightly affected by the solvent treatment.

Table 1.

Values of work function and surface roughness of ZnO/PEI with different-solvent surface treatment.

.

As is well known, the film morphology is very important for the device performance. Here, the film forming properties of PEI are investigated by AFM. Figure 3 shows the AFM images of PEI film surfaces treated with W/O (Fig. 3(a)) and with (Fig. 3(b)) DI, (Fig. 3(c)) EtOH, and (Fig. 3(d)) EEA. As shown in Table 1, the values of surface roughness (Ra) of PEI film W/O rinsed with DI, EtOH, and EEA are 4.12, 4.35, 2.76, 2.65 nm, respectively. These results imply that the different treatments have different influences on the film-forming properties of PEI. In Fig. 3, there are a lot of “island” bulges for the PEI layer treated by DI. As for the PEI treated with EtOH and EEA, the surface roughness and film uniformity are significantly improved compared with the W/O one. As a result, the physical contact between PEI and EML becomes more efficient, and better device performance can be expected. The physical properties of DI, EtOH, and EEA are compared further and find that the main difference among them consists in polarity value and boiling point. The polar values and the boiling points of DI, EtOH, and EEA are 10.2, 4.3, 5.0,[31] 100 °C, and 78 °C, 135 °C, respectively. From these data, it is possible that the differences in solvent polarity and boiling point have a direct influence on the film-forming state of PEI, leading to the different performances of devices.[26,3234]

Fig. 3. AFM topographic images (2 μm × 2 μm) of the PEI surface W/O (a) and with (b) DI, (c) EtOH, and (d) EEA treatment.

In other words, the PEI layer experiences the EtOH and EEA treatment, some agglomeration particles and interface defects can be removed, thereby the ability to inject the electrons from ILs to EML is improved. But in the case of the DI treated one, the interface defects are actually increased, then the ability to inject and tranport the electrons is reduced, eventually leading to the lowest device performance.[26,3234]

The current density–voltage–luminance (JVL) and the current efficiency–luminance–power efficiency (CELPE) characteristics of the fluorescent IOLEDs are shown in Figs. 4 and 5, respectively. It is obious that both the current density and luminance for the device treated with DI are significantly lower than those of the device without rinsing. As for devices treated with EtOH and EEA, both current density and brightness are significantly enhanced at the same voltage. As shown in Table 2, the turn-on voltage (Von), the maximum luminance (Lmax), CEmax, PEmax and the maximum external quantum efficiency (EQEmax) of the optimized EEA treated device are 5.2 V, 10090 cd·m−2, 3.16 cd·A−1, 0.98 lm·W−1, and 1.05%, in contrast to the untreated one, the turn-on voltage is reduced by 0.4 V, the Lmax, CEmax, PEmax, and EQEmax are increased by 28%, 17%, 21%, and 18%, respectively. In other words, EEA treatment improves the injection and the balance of the carrier in the device. From the electroluminescent (EL) spectra of the devices as shown in the inset of Fig. 6, all EL spectra are completely overlapped, indicating that the treatment only changes the interface contact beween the EML and ILs, so there is no side effect on the properties of emitting material. This is very valuable for fabricateing the solution preparation multilayer device.

Fig. 4. JVL characteristics of Alq3 IOLEDs with different surface treatment processes on the PEI layer.
Fig. 5. CELPE characteristics of Alq3 IOLEDs with different surface treatment processes on PEI layer.
Table 2.

Device performances of IOLEDs with different-solvent surface treatment.

.
Fig. 6. EQEV (Inset: EL spectra.) characteristics of Alq3 IOLEDs with different surface treatment processes on PEI layer.

For studing the universality of surface treatment method on PEI, we apply it to an Ir(ppy)3 green phosphorescent device. The JVL and the CELPE characteristics of the Ir(ppy)3 IOLEDs are shown in Figs. 7 and 8, respectively. In accordance with the results in the above-mentioned Alq3 IOLEDs, compared with those in the W/O Ir(ppy)3 IOLEDs, the performance of EtOH and EEA treated devices are significantly improved too. As shown in Table 3, the values of Von, Lmax, CEmax, PEmax, and EQEmax of the optimized EEA treated device are 5.1 V, 18830 cd·m−2, 30.54 cd·A−1, 13.90 lm·W−1, and 8.95%, respectively. The turn-on voltage is reduced by 0.6 V, the Lmax, CEmax, PEmax, and EQEmax are increased by 36%, 21%, 22%, and 20% compared with those of the W/O one, respectively. On the contrary, the turn-on voltage and the Lmax of the DI treated device are 7 V and 6216 cd·m−2, which are significantly inferior to those of the W/O one. From the EL spectra of the devices with and without being treated as shown in the inset of Fig. 9, we can find that all EL spectra are completely overlapped, indicating that the solvent surface treatment does not spoil the phosphorescent emitting material. It proves again that EEA treatment has potential applications in the solution process multilayer phosphorescent devices.

Fig. 7. JVL characteristics of Ir(ppy)3 IOLEDs with different surface treatment processes on PEI layer.
Fig. 8. CELPE characteristics of Ir(ppy)3 IOLEDs with different surface treatment processes on PEI layer.
Table 3.

Device performances of phosphorescent IOLEDs with different-solvent surface treatment.

.
Fig. 9. EQEV (Inset: EL spectra.) characteristics of Ir(ppy)3 OLEDs with different surface treatment processes on PEI layer.

We suggest that the improved device performance of the EEA treated phosphorescent device also results from the optimizing of morphology and the suppressed interface defect states on PEI surface as mentioned earlier.[26,3234] In addition, from Table 3, it is noteworthy that the EQE of the optimized device only drops about 26.34% when the brightness is increased from the 1000 cd·m−2 to 10000 cd·m−2, while that of the W/O one drops about 32.22%. The EQE roll–off, furthermore the stability of EEA treated phosphorescent device, are significantly improved. In contrast, the DI treated device, its carrier injection and transmission are deteriorated due to the increased interface defects and bad interface contact, which results in the inferior device performance.

4. Conclusions

In this work, we systematically investigate the effects of PEI injection layer rinsed with different polar solvents, including DI, EtOH, and EEA. Our work indicates that there are slight effects on the work function of PEI rinsed with these different solvents, but a big influence is exerted on the surface morphology. The EEA treatment can effectively lower the surface roughness, reduce the interfacial defects and improve the physical contact between ILs and the EML. Both the fluorescent and phosphorescent OLEDs based on the PEI rinsed with EEA achieve the best device performances. A maximum current efficiency (CEmax) and power efficiency (PEmax) are enhanced by 21% and 22% for the phosphorescent OLEDs. Our work indicates that the solution-processed OLED performance can be achieved by simple solvent treatment.

Reference
1Tang C WVanSlyke S A 1987 Appl. Phys. Lett 51 913
2Burroughes JBradley DBrown AMarks RMackay KFriend RBurns PHolmes A 1990 Nature 347 539
3Baldo M AO'brien DYou YShoustikov ASibley SThompson MForrest S 1998 Nature 395 151
4Friend RGymer RHolmes ABurroughes JMarks RTaliani CBradley DDos Santos DBredas JLögdlund M 1999 Nature 397 121
5Müller C DFalcou AReckefuss NRojahn MWiederhirn VRudati PFrohne HNuyken OBecker HMeerholz K 2003 Nature 421 829
6Sun YGiebink N CKanno HMa BThompson M EForrest S R 2006 Nature 440 908
7Yu J NZhang M YLi CShang Y ZLv Y FWei BHuang W 2012 Chin. Phys. 21 083303
8Miao Y QGao Z XZhang A QLi Y HWang HJia H SLiu X GTsuboi T 2015 Chin. Phys. 24 057802
9Hsieh H HTsai T TChang C YWang H HHuang J YHsu S FWu Y CTsai T CChuang C SChang L H 2010 SID Int. Symp. Dig. Tech. Pap 41 140
10Fan YZhang HChen JMa D 2013 Org. Electron 14 1898
11Shih T HTing H CLin P LChen C LTsai LChen C YLin L FLiu C HChen C CLin H S 2014 SID Int. Symp. Dig. Tech. Pap 45 766
12Song ENam H 2014 Displays 35 118
13Lee J HWang P SPark H DWu C IKim J J 2011 Org. Electron 12 1763
14Lee HPark IKwak JYoon D YLee C 2010 Appl. Phys. Lett 96 153306
15Chen JShi CFu QZhao FHu YFeng YMa D 2012 J. Mater. Chem 22 5164
16Ding ZXing RFu QMa DHan Y 2011 Org. Electron 12 703
17Wilson PLekakou CWatts J F 2012 Org. Electron 13 409
18Perelaer JAbbel RWünscher SJani RVan Lammeren TSchubert U S 2012 Adv. Mater 24 2620
19Coenen M JSlaats T MEggenhuisen T MGroen P 2015 Thin Solid Films 583 194
20Kabra DLu L PSong M HSnaith H JFriend R H 2010 Adv. Mater 22 3194
21Zhou YFuentes-Hernandez CShim JMeyer JGiordano A JLi HWinget PPapadopoulos TCheun HKim J 2012 Science 336 327
22Kim Y HHan T HCho HMin S YLee C LLee T W 2014 Adv. Funct. Mater 24 3808
23Stolz SScherer MMankel ELovrincic RSchinke JKowalsky WJaegermann WLemmer UMechau NHernandez-Sosa G 2014 ACS Appl. Mater. Interfaces 6 6616
24Min XJiang FQin FLi ZTong JXiong SMeng WZhou Y 2014 ACS Appl. Mater. Interfaces 6 22628
25Fan CLei YLiu ZWang RLei YLi GXiong ZYang X 2015 ACS Appl. Mater. Interfaces 7 20769
26Yan LSong YZhou YSong BLi Y 2015 Org. Electron 17 94
27Yang WLiu ZChen JHuang LZhang LPan HWu BLin Y2015Sci. Rep5
28Höfle SSchienle ABruns MLemmer UColsmann A 2014 Adv. Mater 26 2750
29Höfle SSchienle ABernhard CBruns MLemmer UColsmann A 2014 Adv. Mater 26 5155
30Beek W JWienk M MKemerink MYang XJanssen R A 2005 J. Phys. Chem. 109 9505
31West S D 1987 J. Chromatogr. Sci 25 122
32Liu SHo SChen YSo F 2015 Chem. Mater 27 2532
33Dong W FLiu SWan LMao GKurth D GMöhwald H 2005 Chem. Mater 17 4992
34Staudigel JStößel MSteuber FSimmerer J 1999 J. Appl. Phys 86 3895