PF, PTAA homopolymers and PF-b-PTAA block copolymers were synthesized as reported [9,10]. Chemical structures of these polymers are shown in Figure 1. Characteristics of utilized polymers are listed on Table 1.

The performance of the EL devices based on blend systems with the device structure of ITO/PEDOT:PSS (30 nm)/polymer (30 nm)/BCP (50 nm)/LiF (0.5 nm)/Al (100 nm) was evaluated. As a hole blocking layer, 2,9-dimethyl-4,7-diphenyl-1,10-phenanthroline(bathocuproine, BCP) was used between the emission layer and cathode. EL devices were fabricated as described in a previous report [9]. Atomic force microscopic (AFM) measurements were performed on JEOL JSPM 4200 system in trapping mode (phase and topographic modes) with an MPP-11100-10 silicon probe (resonant frequency: 300 kHz, force constant: 40 N/m). All thin film of polymers were spin-cast onto glass slide with a MIKASA 1H-D7 spin coater from THF solutions at 1500 rpm for 30 s.

A series of EL devices based on PF1/PF-PTAA1 blend system were characterized to investigate the effect of the block copolymer. Figure 2 showed current efficiency vs. current density characteristics. The EL characteristics were summarized in Table 2.

With the increase of the content of  the block copolymer, the luminance (at 30 mA/m²) and the maximum current efficiency increased, and were almost equal to those of the block only device (100 wt%). As discussed in a previous report [10], it can be speculated that PTAA segment with hydrophilic moiety segregates at the interface between PEDOT:PSS and the active layer. Therefore, it is considered that the bipolar block copolymer effectively enhances the hole injection and electron blocking ability. It is noteworthy that blend system (77 wt%) showed higher current efficiency in the region of high current density than block only device. In the blend film, the size of PF domain is considered to be larger compared with the block copolymer film [17].

The large domain seemed to enhance the recombinetion probability and/or improve the carrier balance especially in the high current density region. As shown in Table 3, the device with PF2/PTAA showed the worst performance due to the severe phase separation (see Figure 4). It is revealed that the utilization of the block copolymer is much favorable in order to enhance the hole

Table 1. Characteristics of utilized polymers.

^a Determined by ¹H-NMR, ^b determined by GPC, ^c determined by DSC.

injection ability in blend systems.

In order to investigate the effect of the morphology or domain size of phase separated systems on EL performance, we examined the ternary blend systems consisting of both homopolymers and the block copolymer keeping the chemical composition constant (fluorene:triarylamine = 60:40). The addition of corresponding block copolymer is one of the effective strategies for more stable, and finer scale morphologies [18,19]. Figure 3 showed current efficiency vs. current density characteristics for the

Table 2. EL characteristics for the devices based on PF1/PF-b-PTAA1 blend systems.

Table 3. EL characteristics for the devices based on PF2/PTAA/PF-b-PTAA2 blend systems.

PF2/PTAA/PF-b-PTAA2 blend systems. The EL characteristics were summarized in Table 3. Although it is found that the addition of PTAA component to the active layer improved the performance as shown in Figure 2 and Table 2, the simple blend of both homopolymers afforded the worst performance we examined as above mentioned.

This result strongly suggests that the morphology in the active layer also plays an important role as well as its chemical characters. In fact, the maximum current efficiency and luminescence at 30 mA/cm² increased with the increase of the content of PF-b-PTAA2. The similar effect of a block copolymer as an additive was observed for electro-phosphorescent devices based on the blend systems consisting of two types of vinyl homopolymers (hole and electron transporting), and bipolar block copolymer with both segments [16]. Figure 4 represents the AFM images of blend films with the thickness of ca. 30 nm on PEDOT/PSS in a tapping mode. In the simple blend (Figure 4(f)), ca. 1 μm of domains were observed, and the domain size decreased as the content of PF-bPTAA2 increased. These results indicate that the finer domain size was achieved by the addition of the block copolymer as expected, and was favorable for better EL performance. Higher current efficiency was observed at the higher current density region for the blend system (80 wt% of block copolymer) than block only device. As discussed above, this is probably due to the appropriate domain size for the effective recombination and/or carrier balance in the blend system.

The relationship between the block copolymer content and EL performance for both blend systems are shown in Figure 5. For both cases, the maximum current efficien cies increased with the content of the block copolymer, and reached almost constant values at 80 wt%. The addition of block copolymers altered mainly the chemical nature of the active layer in the first blend system, and changed the morphological nature in the second system.

As shown in Figure 5, it is found that EL performance is governed by the both factors which can be controlled by the utilization of block copolymers as the additive to PF or PF/PTAA blend systems. Figure 6 shows EL spectra for typical two types of devices (PF1/PF-bPTAA1 77 wt%, and PF2/PTAA/PF-b-PTAA 80 wt%). These profiles are almost the same as that observed for the block copolymer based device [6]. Other devices also exhibited a similar blue EL profiles without unnecessary green-emitting bands.