Patterned Organic and Inorganic Composites for Electronic Applications

Mar 16, 2009 - Citation data is made available by participants in Crossref's Cited-by Linking service. For a more comprehensive list of citations to t...
1 downloads 0 Views 358KB Size
J. Phys. Chem. C 2009, 113, 5777–5783

5777

Patterned Organic and Inorganic Composites for Electronic Applications Paolo Vacca,* Giuseppe Nenna, Riccardo Miscioscia, Domenico Palumbo, Carla Minarini, and Dario Della Sala Enea Centro Ricerche Portici, Via Vecchio Macello, 80055 Portici (NA), Italy ReceiVed: September 2, 2008; ReVised Manuscript ReceiVed: January 23, 2009

Carbon nanotube (CNT) and polymer composite materials were obtained using two manufacturing processes. The first method is dispersion of CNT in a solvent-doped poly(ethylenedioxythiophene) polystyrene sulfonate (PEDOT:PSS) solution. The composite retains high optical properties because of the polymeric system and shows improved electrical properties. The second method is in situ polymerization of ethylenedioxythiophene (EDOT) in the presence of CNT. This procedure assures a uniform CNT distribution in a highly conductive p-EDOT layer with reduced optical transmittance. The composites were analyzed for optical transmittance, surface energy, polarity, distribution, and resistivity, and then they were used as an anodic layer in organic light-emitting diode (OLED) manufacturing. The device performances were characterized and compared to that of conventional devices with an indium tin oxide anode. p-EDOT composite layers have shown conductivity and optical transmittance suitable to produce an OLED with a 10 cd/A efficiency. Introduction Indium tin oxide (ITO) is most widely used as a transparent anode for organic light-emitting diodes (OLEDs) or polymer light-emitting diodes (PLEDs) due to its high optical transparency and electrical conductivity. The performance of electroluminescent diodes is highly affected by the hole injection properties of the ITO anode, and for this reason ITO surface treatments were adopted to control hole injection ability and improve the performance of OLEDs.1-10 However, ITO induces significant limitations for current and future generation of OLEDs because of diffusion of oxygen into proximate organic charge transporting and emissive layers,10-12 a relatively low work function (∼4.8 eV), and corrosion susceptibility.13 The fact that indium (ITO constituent) is in relatively short supply and, therefore, expensive also presents significant challenges for large-scale introduction of next generation display and photovoltaic technologies.14,15 Therefore, much effort has been made in improving the electrical properties of host semiconductor polymers by introducing guest conductive nanoparticles with low concentration and few side effects on optical properties. Carbon nanotubes (CNTs) were a good fit for polymer-based nanocomposites, which work well in a large variety of applications such as devices in nanoelectronics, field emitters, and reinforcing materials, because they exhibit high mechanical properties, have metal and p-type semiconductor properties, and have a relatively high work function. In addition, they also have a high aspect ratio, high conductivity, a low percolation threshold, and low absorption in the visible region.16-20 The development of a new class of nanostructured materials, e.g., polymer-carbon nanotubes (CNT), nanocomposites (NC), and polymer-TCO nanocomposites, was introduced.21-26 In 2005, a method was published in order to transform aggregated ITO nanoparticles into colloidal suspension by high-speed stirring and to stabilize the suspension by addition of poly(vinyl pyrrolidone).27 The obtained colloidal system cast on an optical grade substrate (PET) has produced ITO-PVP nanocomposite * Corresponding author. E-mail: [email protected]. Telephone: +39 081 7723386. Fax: +39 081 7723344.

films with high transmittance in the visible spectral region but with electrical resistivity that is four orders of magnitude higher than in respect to that of ITO layers. Recently, Chatterjee has reported46 this fabrication by electrospinning of polymer-ITO (PI) nanocomposite templates by mixing polyethylene oxide (PEO) with various ITO nanopowder concentrations. In these systems, current transport is accomplished by the tunnelling of carriers between neighboring ITO dots in the PEO matrix. To obtain a conductivity increment, we need to decrease the distance between neighboring ITO particles by increasing the ITO concentration until values of 30-40 wt % are obtained with subsequent reduction of the composite flexibility. Another approach developed recently has included increasing the conductivity of semiconductor polymers by modifying their level of conjugation. Poly(ethylenedioxythiophene) (PEDOT) is widely used for this purpose.14,28-39 PEDOT in its oxidized form has a relatively high conductivity, a band gap of 1.5-1.6 eV,15,40,41 and good optical transparency in the visible region, but it is insoluble. Combination with a polyelectrolyte, polystyrene sulfonate (PSS), resolved this problem, assuring a processable waterborne dispersion but strongly decreasing its conductivity.42,43 In this form, PEDOT is not suitable to compete with ITO as an anodic layer, but its use is limited to a hole injection layer application. Some authors have developed a solvent doping of PEDOT:PSS dispersions, obtaining considerable improvement in conductivity. In detail, polar solvents with high dielectric constants such as dimethyl sulfoxide (DMSO) induce a stronger screening effect between counterions and charge carriers, which reduces the Coulomb interaction between positively charged PEDOT and negatively charged PSS dopants. Therefore, the screening effect enhances the hopping rate and conductivity in the PEDOT:PSS systems.43,45 Previous X-ray diffraction (XRD) and optical absorbance experiments have demonstrated that increases in PEDOT:PSS conductivity by organic solvents are not due to a change in conformation.43 In order to preserve the conductive properties of PEDOT, layers are obtained by means of an in situ ethylenedioxythiophene (EDOT) polymerization performed after solution deposition. In this way, it is possible to manage an easily processable solution

10.1021/jp807803z CCC: $40.75  2009 American Chemical Society Published on Web 03/16/2009

5778 J. Phys. Chem. C, Vol. 113, No. 14, 2009 made of EDOT and oxidant agents and to obtain the insoluble PEDOT material on the substrate.45This work uses PEDOT and PEDOT:PSS in order obtain organic and inorganic composites suitable for conductivity, optical transmittance, and morphology for replacing transparent conductive oxides in electronic applications. The first procedure includes the dispersion of functionalized double-walled carbon nanotubes (DWCNTs) in water and then a mixture preparation of doped PEDOT:PSS-DWCNT composites. Another procedure was developed in order to obtain a uniform DWCNT distribution in the organic layer. CNTs were dispersed in an EDOT solution and spin coated onto a substrate, and then an in situ polymerization of EDOT was performed by an oxidative process. The composites were analyzed for physical and electrical properties, and then they were used as an anodic layer in OLED manufacturing. The devices’ performances were characterized and compared to that of conventional devices with an ITO anode. Experimental Section DMSO-Doped PEDOT:PSS. A 2.7 wt % dispersion in water with PEDOT:PSS (1:20 w/w, Aldrich) was filtered and then diluted with 5% v/v DMSO. The mixed solution of PEDOT: PSS and the organic solvent were stirred continuously for 24 h at room temperature, filtered, and used for composite preparation according to the subsequent procedures. Double-WalledCarbonNanotubeWaterDispersion. Doublewalled carbon nanotubes (DWCNTs) functionalized with a water soluble polar substituent COOH group, covalently bonded directly to the nanotube, were chosen in order to improve the nanoparticle dispersion and its subsequent deposition.48,49 The employed CNTs, 2 nm in diameter and 0.5-0.6 µm in length, were analyzed by SEM and appear highly entangled with macroscopic particles. For this reason, a significant effort was made to break this association and bring the material into solution. We have dispersed 5 mg of CNTs in 2 mL of water and sonicated the mixture for 30 min. Then an aliquot of 8 mL of water was added, and the resulting solution, 0.5 mg/mL in concentration, was sonicated for an additional 90 min, producing a completely dispersed material according to visual observation. After this procedure, we allowed the dispersion to stand overnight at room temperature, and then we verified the stability of the solution. DMSO-Doped PEDOT:PSS-CNT Composite. In order to prepare a CNT-doped PEDOT:PSS formulation, an aqueous dispersion of the conductive polymer with a weight ratio of PEDOT to PSS of 1:20 was used. The dispersion was filtered and then diluted with 5% v/v DMSO. The mixed solution of PEDOT:PSS and the organic solvent was stirred continuously for 24 h at room temperature and filtered. An aliquot of DWCNT-COOH dispersion was added to the doped PEDOT: PSS solution, and the resulting PEDOT:PSS-CNT (0.02% w/w) formulation was sonicated for 3 h at room temperature. p-EDOT in Situ Polymerization. An iron(III) tosylate solution, 40 wt %, was prepared by dissolving 16.80 mmol of iron(III) p-toluene sulfonate in 20 mL of n-butanol. Then 9.84 mmol of imidazole was added, and the solution was allowed to stir for 3 h at room temperature. Imidazole facilitates oxidation of organic compounds that normally lose hydrogen atoms. The solution was filtered by using a 0.45 µm filter, and 9.42 mmol of EDOT was added drop by drop. The reactive mixture was stirred for 2 min and then spin coated onto a glass substrate. The samples were heated to 100 °C for 2 min, causing EDOT to polymerize and imidazole to evaporate. Finally, the iron was removed by washing the sample with water. In this procedure,

Vacca et al.

Figure 1. EDOT polymerization reaction scheme.

iron(III) p-toluene sulfonate works as a polymerizing agent, while the imidazole facilitates oxidation of the EDOT compound through the loss of hydrogen atoms. In detail, the reaction of EDOT with iron(III) tosylate is summarized in Figure 1. The overall polymerization reaction can be separated into two principal steps: (1) oxidative polymerization of the monomer to the neutral polythiophene and (2) oxidative doping of the neutral polymer to the conductive polycation. p-EDOT-CNT Composite. An iron(III) tosylate solution, 40 wt %, was prepared by dissolving 16.80 mmol of iron(III) p-toluene sulfonate in 20 mL of n-butanol. Then 9.84 mmol of imidazole was added, and the solution was allowed to stir for 3 h at room temperature. The solution was filtered using a 0.45 µm filter. An aliquot of 0.5 mL of DWCNT-COOH dispersion (0.5 mg/mL) was added to the iron tosylate solution, and the resulting formulation was sonicated for 3 h at room temperature. In the subsequent step, 9.42 mmol of EDOT was added drop by drop, and the reactive mixture was stirred for 2 min and then spin coated onto a glass substrate. The samples were heated to 100 °C for 2 min, causing EDOT to polymerize and imidazole to evaporate. Finally, the iron was removed by washing the sample with water. Composite Layer Patterning. A limit for the use of conductive materials in solution in organic electronics is represented by the difficulty in obtaining a patterned area by means of conventional patterning processes. In order to avoid this limit, we have introduced a new photomechanical process for organic layers in order to define polymeric anode areas and prevent the formation of shorts during top electrode contact soldering.50 A photolithographic step was carried out in combination with the deposition and subsequent peeling of an elastomeric polymer, obtaining a dry patterning of organic layers. This process can be used with pattern polymeric (spinned) materials as well as low molecular (evaporated) materials, ink-jetted materials, sprayed materials, and so on. In detail, the performed dry patterning process of organic materials consists of seven steps (Figure 2). Initially, the process has included photoresist deposition onto a substrate and subsequent UV irradiation with a photolithographic mask and wet development (Figure 2, steps 1-3). In this way, thin film structures were produced with a large thickness (20 µm) and a layout defined by the employed mask. Successively, a silicon elastomer resin was deposited onto the obtained photoresist structures by spin coating (Figure 2, step 4). Poly(dimethylsiloxane) (PDMS) molecules have a unique combination of properties, resulting from the presence of an inorganic siloxane backbone and organic methyl groups attached to silicon.51 They have very low glass transition temperatures and, hence, are fluid at room temperature. The PDMS elastomer that we used is a two-part kit: (1) a liquid silicon rubber base (i.e., a vinyl-terminated PDMS) and (2) a catalyst or curing agent (copolymers of methylhydrosiloxane and dimethylsiloxane). After the liquids are mixed, poured over a master, and heated to elevated temperatures, the liquid mixture becomes a solid, cross-linked elastomer in a few hours via the hydrosilylation reaction between vinyl (SiCH)CH2) groups and hydrosilane (SiH) groups. The mixture was spin coated and thermally cured in order to obtain a poly(dimethylsiloxane)

Organic/Inorganic Composites for Electronics

Figure 2. Scheme of the patterning process.

Figure 3. Profilometric profile and picture (inset) for an anodic composite area.

(PDMS) layer with a thichness larger than that of the photoresist layer. The PDMS layer was obtained in all areas where the photoresist was removed. We chose PDMS because it ensures conformal adhesion to a very large variety of materials, and it shows thermodynamic stability within a large temperature range, high chemical stability, and a low surface energy that allows for an easy peeling process. In the subsequent step (Figure 2, step 5), the photoresist structures were stripped by dipping the sample in acetone for 5 min, and holes were obtained in the PDMS layers. The hole array is defined by the employed mask during the photolithography process. After the hole production, we developed the deposition of polymer anodic structures (Figure 2, step 6). The dispersions, obtained according to the above-reported procedures, were deposited by spin coating onto the structured substrate, and thin film elements were formed in the hole areas. The final step of the process included a peeling of the PDMS layer (Figure 2, step 7). The mechanical peeling is a dry process, and it does not induce any defects in the deposited thin structures. After PDMS peeling, only the polymer anodic structures remained on the substrate, with geometry induced by the layout of the employed photolithographic mask. An obtained anodic structure is observed in its profilometric profile and picture in the inset in Figure 3. Characterization of Materials. The prepared dispersions, (1) DMSO-doped PEDOT:PSS, (2) DMSO-doped PEDOT: PSS-CNT, (3) p-EDOT, and (4) p-EDOT-CNT, were analyzed

J. Phys. Chem. C, Vol. 113, No. 14, 2009 5779 in dynamic viscosity to evaluate the effect of the nanoparticles on the rheological behavior. The dynamic viscosity measurements were performed using a Brookfield viscometer (Visco Basic Plus) that measures the shear stress of the engine to keep the rotor in motion at a given shear rate chosen by the operator. In order to develop a composite characterization, the dispersions were processed by spin coating onto glass and quartz substrates. The deposition experimental parameters were chosen in order to produce layers with similar morphological properties by all dispersions. The film thickness and roughness were evaluated with a surface profiler (KLA Tencor P-10). Raman spectra of composite layers were recorded by means of a Raman microscope (Renishaw), using a 514.5 nm excitation line. Contact angle measurements were performed with Dataphysics OCA 20 equipment at 21 °C and 50% relative humidity in order to study the surface energy of the polymer anodic layers. The resistivity of the PEDOT:PSS coating was measured according to ASTM procedure.52 In detail, the electrodes were painted with silver on the PEDOT:PSS layers deposited onto glass. The obtained electrodes are 1.5 mm wide, 20 mm long, and 20 mm apart. For the optical measurements, the samples were prepared in a form of thin films spin coated onto glass substrates. The transmittance spectra were obtained with a lambda 900 spectrophotometer (PerkinElmer). p-EDOT layers were structurally characterized by X-ray diffraction (XRD) measurements carried out with an MDP-X PERT (Philips) diffractometer, using a Cu KR radiation source. OLED Manufacturing and Characterization. To manufacture multilayer structures, polymeric anode (pA)-N,N’diphenyl-N,N’-bis(3-methylphenyl)benzidine (TPD)-tris(8hydroxyquinoline)aluminum(Alq3)-AlOLEDsweremanufactured using the developed organic and inorganic composites. In order to aquire the anode layer, different composites were obtained by spin coating onto glass substrates. Deposition experimental parameters were chosen in order to produce a 200 nm pA layer. A reference sample of an ITO layer was used in a conventional OLED with a structure of ITO-TPD-Alq3-Al. The spincoating process was developed using a Brewer Science Model 100 spin coater. After anode layer baking, TPD was used as a hole transporting material and evaporated over the organic anode. Alq3 was used as an electron-transporting and emitter material and was deposited by thermal evaporation onto a TPD layer, without losing vacuum conditions. The process was performed in a high-vacuum chamber, and the base pressure was 10 µPa. The layer thickness is 40 nm for TPD and 60 nm for Alq3. An Al cathode, 200 nm thick, was evaporated as a final layer. The active device area is 12.57 mm2. Current-voltage (I-V) characteristics were measured with a semiconductor characterization system (Keithley 4200) in a constant voltage mode with logarithmic increment steps and a delay time of 300 ms before each measurement point. The spectral radiance was monitored using a calibrated spectroradiometer (Optronics OL770), and it has permitted the calculation of the chromaticity coordinates (CIE). All characterizations were performed in air at room temperature. Results and Discussion The prepared dispersions were analyzed in dynamic viscosity in order to study the rheological behavior of the composite systems in their liquid state. The measurements revealed a stronger increment of dynamic viscosity due to CNT doping of the PEDOT:PSS dispersion (Table 1). In this case, PEDOT and PSS chains are linked tightly by ionic interactions and form an ionic polymer complex that induces the association of carbon

5780 J. Phys. Chem. C, Vol. 113, No. 14, 2009

Vacca et al.

TABLE 1

sample ITO DMSO-doped PEDOT:PSS (1:20) DMSO-doped PEDOT:PSS-DWCNT p-EDOT p-EDOT-DWCNT

dynamic viscosity (mPa s)

surface energy (mJ/m2)

polarity (%)

resistivity (Ω cm)

– 13.0

32.6 70.6

49.2 39.2

2.0 × 10-4 3.4 × 101

13.8

71.2

40.3

3.2 × 101

25.2 25.2

67.4 68.6

35.8 38.4

2.3 × 10-3 1.2 × 10-3

nanotubes in macroscopic particles with subsequent increasing of the dynamic viscosity. In an EDOT solution, the low dipole moment of the dissolved materials avoids any orientation of the dispersed nanoparticles, and the result is an unchanged dynamic viscosity for a low doping concentration (