Since they were first reported by Tang and VanSlyke in 1987, multilayer organic light-emitting diodes (OLEDs) have been widely investigated because of their remarkable range of applications in display and lighting technologies. OLEDs offer advantages that include flexibility, large device area, light weight, a solid-state nature, wide viewing angles, fast switching speeds, and low cost.^[1–3] Following the significant progress made in both academic and industrial researches over the past three decades, OLED technology is now entering the market. To ensure their widespread commercial application in the future, new material systems and manufacturing process technologies must be developed continuously on an ongoing basis.^[4]

Device efficiency and device lifetime, which are the two main concerns for practical use, are directly related to the energy consumption and long-term stability/durability and are largely governed by the charge balance in OLEDs.^[5–8] Therefore, to obtain an efficient OLED, it is essential to optimize the charge carrier injection, transport, and recombination processes in the device.^[5,6,9] Indium tin oxide (ITO) is the preferred anode material for use in electroluminescent devices because of its high conductivity and high transparency in the visible spectral regime.^[9] However, the relatively low work function of ITO together with the interface dipoles that occur at the conventional ITO/organic interfaces typically induce high barriers for hole injection.^[4,9] Therefore, insertion of a hole injection layer (HIL) between the ITO anode and the organic hole transport layer (HTL) has become a popular strategy for provision of favorable band alignment in OLEDs.^[10–13]

The metal oxides such as WO₃,^[14] V₂O₅,^[15] MoO₃,^{[3,13,16–18]} ReO₃,^[19] and NiO_x^[20] are important HIL materials and have been used widely to enhance the hole injection because they offer appropriate energy levels, good thermal stability, and excellent charge carrier mobility.^[8] However, these inorganic HILs are usually prepared by annealing a precursor layer at relatively high temperatures, which thus limits the application of these materials in flexible electronic products.^[8]

Conducting polymers represent another class of HIL/HTL materials. Their unique optical and electrical properties have allowed them to demonstrate promising potentials for use in flexible optoelectronic devices.^[8] Among these materials, poly-(3, 4-ethylenedioxythiophene): poly(styrene sulfonate) (PEDOT: PSS) is one of the most popular HIL/HTL materials because of its appropriate energy levels and ability to form smooth films on ITO substrates.^[21,22] Unfortunately, challenges including ITO etching, indium diffusion, water residue, exciton quenching, and weak adhesion still exist.^[2,9,23,24] In addition, water residue induced by aqueous PEDOT: PSS dispersion can introduce damage to the devices that typically leads to poor long-term durability.^[8,9]

Since the first demonstration of the use of copper phthalocyanine (CuPc) to modify the ITO anode of an OLED by Tang and VanSlyke in 1996,^[5] the use of metal phthalocyanines (MPcs) as anode buffer layers has been subject to intensive investigation with the aim of achieving higher OLED efficiencies and stabilities.^{[5,9,10,25–30]} In most cases, these MPc layers are fabricated via the vacuum sublimation method because of the poor solubility of typical MPcs in organic solvents.^[5,25,30] While vacuum processing eases the fabrication of complex OLED configurations with the desired performance and high reliability on a laboratory scale, this approach is very costly because of the expense of maintaining an ultrahigh vacuum combined with its high energy consumption, high operating temperatures, and huge materials losses.^[31,32] Unlike vacuum processing, solution-based processing offers the advantages of low-cost fabrication, compatibility with flexible substrates, and the potential for use in large-scale electronic device production,^[31,33] and is thus beneficial for industrial manufacturing applications. Numerous studies have thus been conducted to develop new soluble MPcs, in particular, MPcs that can be functionalized using solubilizing alkyl or alkoxy groups on peripheral or nonperipheral positions in the MPc rings are of considerable interest.^[31]

In our previous work, we presented two soluble tetraalkyl-substituted copper phthalocyanines (CuEtPc and CuBuPc) that were used as anode buffer layers in OLEDs and when compared with PEDOT: PSS-based OLEDs, enhanced device efficiency and stability were achieved.^[9] In this work, we change the central metal in the phthalocyanines from Cu to the less costly Zn and present two soluble tetraalkyl-substituted ZnPcs (ZnEtPc and ZnBuPc, as shown in Fig. 1(a)) for use as anode buffer layers in OLEDs. These ZnPcs demonstrate good solubility in common organic solvents and thus allow the OLED anode buffer layers to be prepared via spin-coating. Typical tris(8-hydroxyquinoline)aluminum (Alq₃)-based OLEDs are fabricated using these layers and characterized to examine the quality of the thin films. The performance and durability of OLEDs containing these ZnEtPc and ZnBuPc buffer layers are also assessed.

Fig. 1.

	Figure Option ViewDownloadNew Window
	Fig. 1. (a) Molecular structures of the materials used. (b) Device configurations and energy level diagrams of the OLEDs in this study.

The hole transport properties of the ZnEtPc and ZnBuPc were examined via top contact field-effect transistor (FET) technology, as shown in Fig. 2. Both the devices using the spin-coated thin films of ZnEtPc and ZnBuPc exhibit a p-type FET behavior, and the charge carrier mobilities of the ZnEtPc and ZnBuPc were determined to be 6.5 × 10⁻⁴ cm²/V⋅s and 9.5 × 10⁻⁴ cm²/V⋅s, respectively, which are similar to those of CuEtPc and CuBuPc in our previous work, suggesting that these materials are suitable for the application as anode buffer layers in OLEDs.^[9]

Fig. 2.

	Figure Option ViewDownloadNew Window
	Fig. 2. Output and transfer characteristics of top-contact FETs based on the spin-coating thin films of (a), (b) ZnEtPc and (c), (d) ZnBuPc.

The morphologies of the spin-coated ZnEtPc and ZnBuPc thin films on ITO were studied via atomic force microscopy (AFM) and compared with those of the pristine ITO and the PEDOT: PSS thin film. As shown in Fig. 3, the pristine ITO has a rough surface with a surface root-mean-square (RMS) roughness of 2.36 nm (Fig. 3(a)). The ZnEtPc and ZnBuPc thin films exhibit a smooth and uniform surface characteristic with average surface RMS roughness values of 1.38 nm and 0.92 nm, respectively (Figs. 3(c) and 3(d)), which are similar or lower than that of the PEDOT: PSS thin film (1.29 nm) (Fig. 3(b)). The flat and uniform property of the ZnEtPc and ZnBuPc anode buffer layers between ITO and NPB may help to create a conducting path for charge transport and reduce the pin-hole effect that will affect the long-term durability of OLEDs.

Fig. 3.

	Figure Option ViewDownloadNew Window
	Fig. 3. AFM images of (a) the pristine ITO glass, (b) the PEDOT: PSS film, and the spin-coated thin films of (c) ZnEtPc and (d) ZnBuPc on ITO substrates.

The electroluminescent (EL) spectra of the fabricated OLEDs are shown in Figs. 4(a) and 4(b). Both devices show emission maxima at approximately 530 nm, which is consistent with the EL characteristics of the Alq₃-based OLEDs and indicates that the buffer layers do not cause obvious changes in the emission spectrum of Alq₃.^[8,9,25] The transparency spectra of the ZnEtPc and ZnBuPc buffesr are illustrated in Fig. 5, the transparencies of > 99% and > 97% at the EL maximum (∼ 530 nm) are found for the ZnEtPc and ZnBuPc buffer layers, respectively, indicating that the absorption behavior of these buffer layers has no obvious effect on the EL of the devices.

Fig. 4.

	Figure Option ViewDownloadNew Window
	Fig. 4. Electroluminescent spectra of the OLEDs with (a) the ZnEtPc layer and (b) the ZnBuPc layer.

Fig. 5.

	Figure Option ViewDownloadNew Window
	Fig. 5. Transparency spectra of the ZnEtPc and ZnBuPc buffer layers.

The luminance–voltage (L–V) and luminance–current density (J–V) curves for all the OLEDs are shown in Figs. 6(a) and 6(b), respectively. When compared with the OLED without the anode buffer layer,^[9] the insertion of the ZnEtPc and ZnBuPc interlayers between the ITO and the HTL causes the L–V characteristics to shift toward a higher voltage. Turn-on voltages of 3.7 V and 3.8 V are observed for the OLEDs that contain the ZnEtPc and ZnBuPc buffer layers, respectively, while that of the diode with the single HTL is only 2.5 V, thus indicating that the anode buffer layers impede hole injection into the HTL.^[6,7,9] This hole-blocking characteristic is possibly caused by the alkyl groups (i.e., ethyl and butyl) on the ZnPc structures, which form thin “insulating layers” at both the ITO/ZnEtPc (ZnBuPc) and ZnEtPc (ZnBuPc)/NPB interfaces.^[9,35] As shown in Table 1, a significant OLED luminance enhancement when compared with that of the single HTL diode is achieved by inserting the ZnEtPc (ZnBuPc) buffer layer at the ITO/NPB interface. Both ZnPc-based OLEDs exhibit a luminance maximum similar to that of the PEDOT: PSS-based OLED, while the highest luminescence among all the diodes of 10560 cd/m² is found for the ZnEtPc-based OLED (Table 1). It was previously reported that hole migration into the Alq₃ layer in Alq₃-based OLEDs led to a reduced fluorescence quantum yield caused by the degradation products of the unstable cationic Alq₃ species ( ), which acted as fluorescence quenchers.^[6,7,36] The higher brightness of these devices is therefore attributed to the effect of the ZnEtPc and ZnBuPc buffer layers in achieving more balanced hole and electron injection processes by impeding hole injection into the NPB layer.^{[6,7,9,35,37,38]}

Fig. 6.

	Figure Option ViewDownloadNew Window
	Fig. 6. (a) Luminance–voltage and (b) luminance–current density characteristics of OLEDs. The data for the OLEDs without an anode buffer layer and for those with a PEDOT: PSS layer are taken from the literature.[9]

Table 1.

Table 1.

Table 1. OLED performance. . Anode buffer layer λEL, maxa/nm CIEb (x, y) Vturn-onc/V LEmaxd/cd⋅A−1 PEmaxe/lm⋅W−1 Lmaxf/cd⋅m−2 W/o anode buffer layer 530 0.313, 0.541 2.5 2.96 2.04 6966 PEDOT: PSS 530 0.314, 0.533 3.0 3.77 1.97 9477 ZnEtPc 530 0.285, 0.544 3.7 3.19 1.31 10560 ZnBuPc 530 0.289, 0.547 3.8 3.22 1.27 9152 a Maximum peak value of EL spectrum. b Measured at maximum luminance. c Turn-on voltage recorded at luminance of 1 cd/m2. d Maximum luminous efficiency. e Maximum power efficiency. f Maximum luminance. The data for the OLEDs without the anode buffer layer and with the PEDOT: PSS are taken from the literature.[9]

Table 1. OLED performance. .

The enhanced charge balance achieved when using the ZnEtPc and ZnBuPc anode buffer layers can also be observed in the luminance–current density (L–J) and current efficiency–current density characteristics (Figs. 7(a) and 7(b), respectively). The luminance increases almost linearly with increasing current density from a low initial current density (approximately 0–100 mA/cm²) and when the current density exceeds 100 mA/cm², a significant roll-off effect can be observed in both diodes, thus indicating the reduced charge balance in the device at high current densities (Fig. 7(a)). Possible causes include exciton–exciton interactions, exciton–charge carrier interactions, and exciton dissociation when the device is operating at a high current density.^{[8,9,25,39–41]} More serious roll-off effects were observed in the single HTL diode and the PEDOT: PSS-based diode, indicating that the ZnEtPc and ZnBuPc layers can significantly optimize both charge carrier injection and transport processes in the OLEDs.^[8,9,35] These results are consistent with the characterization results for the current efficiency–current density characteristics, in which the EL efficiencies of all diodes see a sharp rise, thus achieving their maximum values at low current densities, and these efficiencies then decay with increasing current density in the high current density range (Fig. 7(b)). While the PEDOT: PSS-based OLED produces the highest maximum efficiency of 3.77 cd/A at a low current density, the device shows a more obvious efficiency roll-off than that of the OLEDs containing the ZnEtPc and ZnBuPc buffer layers (Fig. 7(b)). The ZnEtPc- and ZnBuPc-based OLEDs generally demonstrate superior EL efficiencies when compared with those of the other two device structures within the main measurement range. It was previously reported that the improved hole impedance into NPB produced by the use of a hole-blocking layer can help to increase the electron density in the Alq₃ layer zone adjacent to the NPB/Alq₃ interface, thus reducing the population and improving direct hole injection from the NPB into the anionic Alq₃ species ( ) to generate more excited states.^[6,9] Similar to the results reported in our previous work,^[9] the hole-blocking characteristics of the ZnEtPc and ZnBuPc buffer layers produce better charge balance in the OLEDs, thus giving rise to the improved device performances.^{[6,7,9,35,38,42]}

Fig. 7.

	Figure Option ViewDownloadNew Window
	Fig. 7. (a) Luminance–current density and (b) current efficiency–current density characteristics of the four types of OLEDs. The data for the OLEDs without an anode buffer layer and with the PEDOT: PSS buffer layer are taken from the literature.[9]

It can be found that the CIE coordinate value at the OLED maximum luminance undergoes a slight change due to the insertion of the PEDOT: PSS thin film, and a relatively significant alteration when the ZnEtPc and ZnBuPc thin films are introduced. And similar CIE coordinate values can be found for the ZnEtPc- and ZnBuPc-based OLEDs (Table 1). According to the L–V and J–V characteristics of the OLEDs (Fig. 6), all the anode buffer layers exhibit a hole-blocking property for hole injection from the ITO anode, resulting in the enhanced operating voltages compared with the OLED without anode buffer layer. And higher operating voltages of the ZnEtPc- and ZnBuPc-based OLEDs than that of the PEDOT: PSS-based device can be found. It was reported that the recombination zone shifts to the anode side with increased applied voltages, which results in the change of chrominance in the multilayer devices.^[43] Thus, the variation in the CIE coordinate is attributed to higher operating voltages due to the hole-blocking properties of these anode buffer layers, which lead to the movement of the recombination zone away from the cathode side and closer to the anode side.

To investigate the effects of the quality of the ZnEtPc and ZnBuPc layers and that of the enhanced charge balance on the long-term device durability, the OLED lifetimes were tested under a constant drive current density of 100 mA/cm². As shown in Fig. 8, the OLED lifetimes are significantly improved when using the ZnEtPc and ZnBuPc buffer layers. The luminance of the single HTL diode decays rapidly, showing a poor half-lifetime of 355 min.^[9] The OLED lifetime is improved by the insertion of the PEDOT: PSS layer and this OLED exhibits a half-lifetime of 619 min.^[9] Significant enhancements in device durability are observed for the OLEDs with the ZnEtPc and ZnBuPc layers, which have half-lifetimes as long as 2245 min and 2367 min, respectively.

Fig. 8.

	Figure Option ViewDownloadNew Window
	Fig. 8. (a) Normalized luminance vs. operating time characteristics of the OLEDs at a constant current density of 100 mA/cm2. The data for the OLEDs without an anode buffer layer and with the PEDOT: PSS buffer layer are taken from the literature.[9]

These results reveal the strong correlation between the hole injection and transport processes and the device lifetime.^[9,35,41] Indeed, the hole-blocking properties of the ZnEtPc and ZnBuPc anode buffer layers can help to drive the recombination zone away from the cathode, which leads to increased electron density at the NPB/Alq₃ interface and reduced production of the unstable species and thus results in longer OLED lifetimes.^[6,35,41]

Another factor that is responsible for the increased OLED durability under continuous operation conditions is the quality of the anode buffer layers. It was previously reported that the chemical reaction of PEDOT: PSS with the ITO layer induces instability in ITO/PEDOT: PSS system that typically accelerates the device degradation.^[35,44,45] This process can occur during both OLED fabrication and device operation^[35,44] and the degradation becomes more serious when the device is exposed to moisture because large amounts of indium can then diffuse further into the neighboring layers and accelerate the degradation of these layers.^[9,35,44] Furthermore, because the reaction between Alq₃ and any trace water is believed to be another major failure pathway for Alq₃-based OLEDs,^[46] any potential water residue in an OLED induced by aqueous PEDOT: PSS dispersion after fabrication would introduce this danger to the device.^[9] Unlike PEDOT: PSS, which is acidic, the ZnEtPc and ZnBuPc layers are highly chemically and thermally stable and thus offer effective protection to the ITO anode. In addition, the spin-coated ZnEtPc and ZnBuPc layers are fabricated from nonaqueous organic solvents (in this work, dichlorobenzene) that can suppress water damage to the device and thus play a role in retarding device degradation.^[8,9,35]

In summary, two soluble tetraalkyl-substituted ZnPcs (ZnEtPc and ZnBuPc) for use as anode buffer layers in OLEDs are presented. The hole-blocking properties of these ZnPc layers significantly impede hole injection into the Alq₃ emissive layer and thus reduce generation of the unstable species, which leads to enhanced OLED luminance and efficiency when compared with the corresponding properties of PEDOT: PSS-based OLEDs. The high thermal and chemical stabilities of these ZnPcs, the nonaqueous ZnPc film fabrication processes, and the enhanced charge balance in the resulting OLEDs lead to the enhancement of the OLEDs℉ durability. The results of this study indicate a promising direction to follow in future work.