**4. Results and Discussion**

The surface roughness of the fabric substrate was measured by AFM (Park System, Seoul, Korea) at each step of the planarization process. The surface roughness was plotted and also compared with the conventional plastic and glass substrate in Figure 3. The surface roughness of the PET substrate was reduced from 10 μm to 0.3 μm (as shown in Figure 3) after the planarization process, as described in the previous section. The PU layer decreased the roughness at the macro scale from 10 μm to 2 μm, and the PA decreased the roughness at the micro scale from 2 μm of the PET/PU to 0.3 m and also enhanced the process compatibility to the subsequent films by changing the hydrophobicity of the PET/PU as well. Although the surface roughness of the PET/PU/PA substrate (0.3 μm) was larger than that of poly carbonate (PC) plastic (0.025 μm) and glass (0.003 μm) substrates, it was comparable to the thickness of OTFTs (0.43 μm) and OLEDs (0.3 μm). Therefore, the devices could be fabricated on the planarized PET/PU/PA substrate.

**Figure 3.** The surface roughness variation of the polyethylene terephthalate fabric substrate according to deposition of the polyurethane (PU) and photo-acryl (PA) layers, including the AFM images of bare PET, PET/PU, and PET/PU/PA fabric substrates.

The hybrid electrode of CNT/Au for the S/D contacts together with PA for the gate dielectric in the OTFTs was employed for the first time for an AMOLED panel to improve performance as well as to reduce the process steps in this work. To reduce the interface states density and to obtain a smoother gate surface, the PA solution was diluted from the as-purchased state by adding the solvent of propylene glycol monomethyl ether acetate (PGMEA). The electrical characteristics were varied with the mixing ratio of PA to PGMEA, as shown in Figure 4. Representative transfer curves are depicted according to the various mixing ratios of the PA solutions. The electrical parameters are summarized in Table 1, where the values were averaged out of sixteen OTFTs for each ratio.

**Figure 4.** The transfer curves of the pentacene-OTFTs using photo-acryl (PA) for the gate dielectric with the various mixing ratios of PA to PGMEA, in which the CNT/Au (5 nm) electrodes were used for the S/D contacts.

**Table 1.** The average values of the electrical parameters of the pentacene OTFT using photo-acryl (PA) as the gate dielectric and the CNT/Au (5 nm) electrodes for the S/D contacts according to various mixing ratios of PA to the solvent PGMEA.


μFET: field effect mobility; Ion: the on-state current; Ioff: the off-state current: SS: sub-threshold slop: ε: dielectric constant.

The OTFTs with a ratio of 1:3 produced the largest mobility of 0.98 cm<sup>2</sup>/V·s, which was 49 times larger than the 0.02 cm<sup>2</sup>/V·<sup>s</sup> of the as-purchased PA, as shown in Table 1, and comparable to the 1.0 cm<sup>2</sup>/V·<sup>s</sup> of the OTFTs using the polyvinylphenol (PVP) gate dielectric and the CNT/Au (5 nm) for the S/D contacts, as reported in Reference [24]. The required performance enhancement for the OTFTs, as well as the reduction in process steps in the stacked AMOLED panel, were successfully achieved using PA for the gate dielectric and the CNT/Au (5 nm) electrodes for the S/D contacts.

The high performance of the OTFTs was degraded after the PL was deposited on them. In Figure 5, the transfer curves of the DR and SW OTFTs without the PL are compared to those with the PL; the transfer curves were measured from separated test pixels. As shown in Table 2, the mobility was reduced by about 40%, from 0.87 cm<sup>2</sup>/V·<sup>s</sup> and 0.75 cm<sup>2</sup>/V·<sup>s</sup> to 0.54 cm<sup>2</sup>/V·<sup>s</sup> and 0.49 cm<sup>2</sup>/V·<sup>s</sup> for DR and SW OTFT, respectively. Although the water in the w-PVA of the PL was expected to protect the hydrophobic pentacene of the OTFTs from the subsequent processes, the developing process of the d-PVA layer and the depositing process of the PA layer seemed to damage the pentacene. However, the on-state currents at 25 V were sufficiently large with 228 μA and 7.97 μA for the DR and SW OTFT, respectively. As a result, the DR OTFT could drive the large OLED and the SW OTFT could quickly charge up the storage capacitor. In addition, the off-state current of the SW OTFT, 1.69 pA/μm, was small enough to keep the charge on the storage capacitor during the time frame. Even though the performance of the DR and SW OTFTs deteriorated after depositing the PL, the performance was still able to operate the AMOLED panel.

**Figure 5.** The transfer curves of the DR and SW OTFTs with and without the protection layer in the test pixel; the performance was degraded by the protection layer; however, the on-state current was still large enough to drive the enlarged OLED of the stacked pixel.

**Table 2.** The average values of the electrical parameters of the DR and SW OTFTs in the stacked pixel with and without the protection layer.


In Figure 6a, the structure of the phosphorescent OLED used in this paper is presented. It consists of multiple organic layers of 1,4,5,8,9,11-hexaazatriphenylene-hexacarbonitrile (HAT-CN) (10 nm)/NPB (30 nm)/4,4'-Cyclohexylidenebis[N,N-bis(4-methylphenyl)benzenamine] (TAPC)(10 nm)/4,4'-Bis(N-carbazolyl)-1,1'-biphenyl (CBP):Ir (ppy) (20 nm)/2,2',"-(1,3,5-Benzinetriyl)-tris(1- phenyl-1-H-benzimidazole) (TPBi)(40 nm)/LiF (0.2 nm) between the cathode and anode electrodes. It also used a reflective anode of Ag (80 nm) and a transparent cathode of Al (1 nm)/Ag (20 nm)/NPB (40 nm) to implement the top emission on the opaque fabric substrate. The various types of OLED lights were also fabricated on the planarized PET/PU fabric substrate with an area of 20 cm × 20 cm, and a representative OLED light with an area of 7 cm × 7 cm is shown in Figure 6b. They successfully operated without an electrical short, reflecting that the surface roughness of the PET/PU substrate (0.3 μm) was smooth enough for OLEDs, even when the area was large. The OLEDs produced a phosphorescent green light with a wavelength of 534 nm and a luminance of 23,673 cd/m<sup>2</sup> at 7 V, as shown in Figure 6c.

**Figure 6.** (**a**) The structure of the phosphorescent OLED, (**b**) an OLED light fabricated on a PET fabric substrate with an area of 7 cm × 7 cm, (**c**) the luminance characteristics of the top emitting phosphorescent OLED fabricated on the PET fabric substrate, and (**d**) the degradation of the OLED luminance with and without the PVA/PA encapsulation.

In order to protect the AMOLED panel from damage due to the air exposure during measurements, a temporary encapsulation consisting of PVA and PA double layers was spin-coated with a thickness of 1 μm for each layer on the final AMOLED panel. It was expected that the water in the PVA solution would not affect the hydrophobic organic layers of the OLEDs, and the thick polymer layers would protect the OLEDs from air exposure. As shown in Figure 6d, as a layer was added, the luminance decreased from the initial luminance of 19,895 cd/m<sup>2</sup> at 7 V with a PVA single layer encapsulation to 16,636 cd/m<sup>2</sup> with the PVA/PA double-layer encapsulation. However, the encapsulation retarded the degradation by keeping air from permeating into the OLEDs. The luminance of the OLEDs gradually decreased to 64% of the initial luminance with the PVA/PA encapsulation and to 58% with the PVA encapsulation after 10 days in air. Meanwhile, the bare OLED decreased to 54%. The AMOLED panel with the encapsulation maintained visible brightness for 10 days, although the brightness became dim with time, as shown in the inset of Figure 6d.

In Figure 7, two types of AMOLED panels are compared. The side-by-side pixels produced an aperture ratio of 19%, and the space between the OLEDs (1130 μm), where the DR and SW OTFTs were contained, was discernable even with bare eyes.

**Figure 7.** Comparison of the stacked pixels with the side-by-side pixels in the AMOLED fabricated on the PET fabric substrate; the aperture ratio of 48% with the stacked pixels (pixel pitch: 1.0 mm × 0.77 mm, OLED area: 0.76 mm × 0.49 mm) is clearly identified with 19% of the side-by-side case (pixel pitch: 1.6 mm × 1.6 mm, OLED area: 1.04 mm × 0.47 mm).

Meanwhile, in the stacked pixels, the pixel space could not be distinguished because the light coming out of a pixel overlapped those of the neighboring pixels. The overlapping light occurred due to the smaller pixel space (280 μm) and the brighter luminance of the OLEDs caused by the larger on-state current (40 μA@VGS = −10 V) of the DR OTFT. The aperture ratio was 48%, approximately 2.5 times larger than that of the side-by-side pixels.

The high-aperture ratio in the AMOLED panel was successfully achieved using the stacked pixel structure, which was possible due to the PL between the OTFTs' pixel circuit and the OLED, as well as the improved performance of the OTFTs by using CNT/Au (5 nm) hybrid electrodes for the S/D contacts and the self-patterning PA gate dielectric.

The washing problem is a large obstacle to overcome for practical applications of electronic textiles including AMOLED textile displays. Recently, articles concerning water resistant encapsulation layers for e-textiles have been reported [25,26]. They have successfully protected the underlaid OLEDs on a fabric without performance deterioration, even after being washed 10 times. Therefore, the washing issue can be resolved in the near future.
