Effect of acid treatment on carbon nanotube-based flexible transparent conducting films.

Despite recent efforts for fabricating flexible transparent conducting films (TCFs) with low resistance and high transmittance, several obstacles to meet the requirement of flexible displays still remain. Filtration method is a useful tool to control the film thickness but limited to the filter size, which is a drawback for large size applications.1-4 The spray coating method is a robust approach for large-size and uniform films, but the inherent sparse density increases the resistivity of film.5 In this Communication, a spray method was introduced to fabricate TCFs on polyethylene terephthalate (PET) films using sodium dodecyl sulfate (SDS)-dispersed single-walled carbon nanotubes (SWCNTs). These films were further immersed in various acids. Excellent conductivity was obtained with a negligible change in the transmittance in the visible range. This enhancement was attributed to the removal of remaining SDS and the subsequent densified film formation to improve the cross-junction resistance between SWCNT networks and enhanced metallicity of SWCNTs, whereas the chemical doping effect was negligible unlike that previously reported on bucky paper.8 The SWCNTs synthesized by the arc discharge method (Iljin Nanotechnology, Inc.) were used to fabricate TCFs by a spray method. The SWCNT powder was dispersed with SDS in deionized water and sonicated, followed by the centrifugation. The SWCNT supernatant solution was then sprayed on PET film followed by several rinsings in water to remove the remaining SDS (see experimental procedures in Supporting Infromation S1). Figure 1a shows a typical sheet resistance-transmittance curve (open-circle line) in a wide range of film thicknesses in comparison with the previously reported values.1-7 The film conductivity increased with increasing film thickness which was measured by atomic force microscopy (AFM) at step edges and was saturated to about 1500 S/cm at a thickness of about 50 nm (open-square line in Figure 1b). Although our results met the requirement of touch screen (TS: 500 Ω/sq, 85% T) for practical applications, the condition, for instance, for transparent conducting electrodes (50 Ω/sq, 85% T) was still not reached. The film performance strongly relies on the material qualities such as purity, diameter, defects, metallicity, and the degree of dispersion.9 Nevertheless, this approach provides a systematic way of fabricating films with reasonable film performance compared to other related works. To improve the conductivity of film, the film was simply immersed in various acids. Several changes were observed. The sheet resistance with 12 M of HNO3 treatment for 60 min was significantly reduced by a factor of about 2.5 times (Figure S1), while the changes in the transmittance were negligible in the visible region (Figure S2). The sheet resistances of ∼ 40 and 70 Ω/sq at the corresponding transmittances of 70 and 80% were obtained in this case (solid-circle line in Figure 1a). The low sheet resistance can meet the criteria of TCFs that may replace the conventional ITO with high flexibility for flat panel displays (FPD). The film thicknesses were reduced by about onefourth of the pristine film owing to the removal of bulky SDS. Those yielded an enhancement in the conductivity by a factor of ∼4 times (solid-star line shown in Figure 1b, open-circle dashed line for comparison by presuming no change in the thickness). The conductivity at large thickness reached 5500 S/cm. This value is much larger than 960 S/cm of SOCl2 treated bucky paper10 and ∼2000 S/cm of SOCl2-treated TCFs2 and comparable to that prepared by filtration method.1 The surface roughness (∼9 nm from AFM) was also improved by wetting some protruded nanotubes, as shown from the field-emission scanning electron microscopy (FESEM) images in Figure 1c,d. Similar effects were observed in the case of sulfuric acid treatment. To understand the effect of acid treatment, we present X-ray photoelectron spectroscopy (XPS) spectra in Figure 2. The peak position in C1s was ∼0.1 eV downshift by H2SO4 and no obvious shift by HNO3 (inset), which is not so prominent compared to 0.5 and 0.2 eV of H2SO4and HNO3-treated and 0.40 eV downshift of SOCl2-treated bucky paper. The development of oxygen-related peaks (inset) was not appreciable for both nitric and sulfuric acidtreated samples. NO-related peaks were developed with nitric acid treatment. What is more intriguing is the removal of sodium content and the SO4 peak particularly in the nitric acid-treated sample. This ensures that the SDS was removed from the nanotube surface. Similar behavior was also observed in the sulfuric acid-treated sample except for an additional addition of SO4 group after SDS removal (Figure S3). Our XPS analysis suggests that the main origin of the enhancement in conductivity is not the doping effect but the removal of insulating SDS in SWCNT networks and the subsequent densified film density. This is in good contrast with the previous § Sungkyunkwan University. ‡ Samsung Cheil Industries. Figure 1. (a) Sheet resistance versus transmittance at 550 nm before (opencircle) and after (solid-circle) nitric acid treatment with previously reported values for comparison and (b) the corresponding conductivity as a function of film thickness before and after acid treatment. SEM images of (c) before and (d) after acid treatment, taken at a viewing angle of 60°. Published on Web 05/31/2007