Transparent and Flexible Carbon Nanotube Electrodes for Organic Light-Emitting Diodes |
CONTENTS
4.1 Organic Semiconductor Technology
4.2 Organic Light-Emitting Diodes
4.3.1 Conductive Polymer Electrodes
4.4.1.1 Composite Emission Layer
4.4.1.2 Composite Buffer Layer
4.4.1.3 CNT Composite Electrodes
4.4.2.1 CNT Thin-Film Fabrication
4.4.2.2 Transfer of CNT Thin Film onto a Substrate
4.4.2.3 SWCNT Electrode OLED Devices
4.4.2.4 Top Emission SWCNT Electrode OLED
4.4.2.5 MWCNT Electrode OLED Devices
4.4.2.6 Summary of CNT Electrode OLED Performance Criteria
In recent years, carbon nanotubes (CNTs) have gained popularity in various fields of applications. This is due to the fact that CNTs possess remarkable properties. CNTs excel in practically all qualities that a material could have, mechanical, electronic, thermal, chemical, etc. CNTs are simply rolled-up sheets of graphene that form cylindrical tubes referred to as single-wall carbon nanotubes (SWCNTs), or multiwall carbon nanotubes (MWCNTs) when more than one tube is concentrically nested together. Their chirality (roll-up vector) and diameter will determine the electronic properties of the CNTs, which could be of either metallic or semiconductor type, and with small or large band gaps [1]. Unlike other conducting materials, the transport of charges within a metallic SWCNT occurs ballistically, with no scattering, thus allowing for very high-current capacity coupled with no heat generation. CNTs are also known to have extraordinary mechanical properties. Comparing CNTs to high-strength steel, they can have one order of magnitude higher in tensile strength (~63 GPa) and Young’s modulus (~1 TPa). CNTs are also less dense than Al by about half and are twice as thermally conductive as natural diamond. Thus one can only begin to imagine why CNTs are so attractive in a very large number of applications. In this chapter, we pay particular attention to what CNTs can bring to the field of organic semiconductor devices and flexible devices. More specifically, we see that CNTs can actually be processed into thin films to form large-area electrodes. Thus when doing so, CNTs form a network that is highly transparent and flexible within the visible spectrum, and yet remains highly conductive, which is ideal for organic semiconductor device applications. In the following sections we start with a short introduction to organic semiconductor technology, followed by a review of different types of electrodes used for organic light-emitting diodes (OLEDs) on flexible substrates, and finally, a discussion of CNT transparent electrode fabrication and the process of OLEDs using those CNT electrodes by different research groups.
4.1 ORGANIC SEMICONDUCTOR TECHNOLOGY
Organic semiconductor technology has opened the doors to a whole new spectrum of novel electronics applications that were simply inconceivable with traditional inorganic semiconductors, such as devices on flexible substrates and disposable electronics. The ability to easily build devices that conform to any shape of surface is no longer just a dream. Furthermore, organic semiconductor processing techniques enable the possibility to fabricate on very large areas at very low cost. Two of the main applications of organic semiconductors are organic light-emitting diodes (OLEDs) and photovoltaic (PV) cells. Both types of applications require essentially the same fabrication process and similar device structures; only the operation mechanisms and materials of choice are different. Therefore here we concentrate our discussion on OLEDs only. Nevertheless, the technical aspects developed here could also be adapted to PV cells.
4.2 ORGANIC LIGHT-EMITTING DIODES
OLEDs have been the focus of many research centers across the world in the past two decades, ever since their invention [2]. The result of such intensive global research efforts led to the first commercial OLED-based electronic monochrome displays for car audio equipment before the beginning of the new millennium. More recently, higher-resolution full-color OLED displays have also emerged for cellular phone and television set applications. Furthermore, there has been a lot of work directed toward implementing OLEDs on flexible substrates; however, we have yet to see a Transparent and Flexible Carbon Nanotube Electrodes commercial product available. The main reason behind this comes from the fact that OLED technology heavily relies on transparent conductive oxides (TCOs), more specifically indium tin oxide (ITO) electrodes. Although ITO possesses a combination of desirable qualities such as low resistivity (~2 × 10–4 Ω-cm), high visible transparency (~85%) [3], and good commercial accessibility, it also holds properties that are detrimental to organic semiconductor devices. It has often been proven that the nonstoichiometric nature of the ITO surface results in a strong dependence on surface treatments in order to obtain the preferred surface properties (work function, roughness, surface energy, and morphology) needed to optimize device performance [4,5]. ITO has also been plagued with oxygen and indium diffusion into the organic active layer, thus destroying the device as the usage progresses [6,7,8].
FIGURE 4.1 Annual average price of indium 99.97%. (From Tolcin, A.C., Mineral Commodity Summaries, U.S. Geological Survey, January 2010. With permission.)
Moreover, indium, one of the main elements in the composition of ITO, is also a rare element that can be expensive to come by. Indium prices fluctuate heavily, governed by the supply and demand. Figure 4.1 shows the annual average price of indium 99.97% over the span of the last 19 years in the United States [9]. When demand rises and supply cannot keep up, prices rise tremendously, as has been witnessed recently with worldwide increased production of flat-panel displays.
Despite the above shortcomings, ITO has been quite successful in a wide range of applications, mainly due to its attractive electronic and optical properties. However, in order to fully take advantage of organic semiconductor technology, the mechanical properties of the electrodes used are also important. The electrode should be able to be as flexible and stress resistant as organic materials. In the case of ITO, it is obvious that it is not suitable for flexible substrate applications, as it is a brittle material that cracks and delaminates when subjected to bending [10]. Furthermore, in order to achieve low-resistivity values, ITO requires relatively high-temperature processing. This means that ITO is not suitable in certain applications, such as in the fabrication of organic top emission devices where one needs to deposit a transparent electrode on top of the organic materials.
In order to fully take advantage of organic material properties and be able to build flexible displays, transparent anode alternatives to ITO with properties that are more compatible with organic semiconductor technology and potentially more abundant and lower cost are being sought by researchers. CNT electrodes present themselves as one of the most promising materials for replacing ITO in organic semiconductor devices, which are especially compatible with the mechanical requirements of flexible and robust applications However, as will be discussed, first introductions of CNTs into OLEDs were accomplished by making composites with conductive polymers as either transparent electrodes or hole injector layers, and in some cases as electron injectors. A short review of conductive polymers that are also candidates for transparent flexible electrodes will also be presented.
4.3.1 CONDUCTIVE POLYMER ELECTRODES
First conducting polymers were discovered by Shirakawa et al. [11] in the form of doped polyacetylene in 1977. Since then, many other conducting polymers have emerged. Only a handful of them have been reportedly used as anodes in OLEDs.
The first report of such devices was in 1992 by Gustafsson et al., where a thin film of polyaniline (PANI) was spin-coated on a poly(ethylene terephthalate) (PET) flexible substrate to form the transparent anode [12]. The active emitting material was poly(2-methoxy, 5-(2′-ethyl-hexoxy)-1,4-phenylene-vinylene) (MEH-PPV). The device was completed by evaporating a layer of Ca and Al as cathode. The PANI film was reported to have very low sheet resistance (~100 Ω/□ at 70% transmittance). The turn-on voltage was measured to be around 1.8 V, which was reported to be very similar to ITO devices, and its work function was 4.3 eV compared to 4.1 eV for ITO at the time. No actual luminescence data value was actually supplied. However, PANI devices seemed to perform similar to the ITO devices. The major drawback of PANI films is that wavelengths under 475 nm were cut off; therefore blue emission devices could not be built with this type of electrode. Figure 4.2 shows the structure and photograph of the PANI OLED devices that were fabricated.
FIGURE 4.2 Structure and photograph of PANI anode MEH-PPV device on PET substrate. (From Gustafsson, G., et al., Nature, 1992; 357(6378): 477–479. With permission.)
Another report published in 1994 by Yang et al. also regarding the use of PANI as a hole injection electrode in MEH-PPV devices is of particular interest [13]. It was the first time where a conducting polymer was used as a hole injection layer between ITO and the active organic layer. It was proven that by inserting that conducting polymer in between, it effectively reduced the hole injection barrier height at the ITO/MEH-PPV interface. In this publication it was found that the barrier height at the PANI/MEH-PPV interface was 0.8–0.12 eV, compared to 0.2–0.24 eV for ITO/MEH-PPV. Given that device performance also depends on the sheet resistance at higher current densities, and that in order to obtain very low sheet resistance of PANI films, one needs a thicker layer of the material, which also reduces dramatically the transmittance, Yang et al. have opted to use a very thin layer of PANI (~600 Å) with 90% transmittance in the visible wavelength backed with an ITO anode, thus resulting in a best of both worlds situation. The combination ITO/PANI anode provided a low sheet resistance and a lower barrier height to the MEH-PPV. It was found that ITO/PANI anodes resulted in a reduced operating voltage of ~30–50%. Furthermore, another group [14] that investigated ITO/PANI anodes concluded that the introduction of a conducting polymer layer not only improved hole injection efficiency, but also had a smoother interface with the organic active layer, resulting in fewer shorts, improved reliability, and lifetime. However, this approach still needs the use of ITO and is not fully compatible with flexible displays.
Another conductive polymer, which is the most commonly used in OLEDs and PV cells today, is poly(3,4-ethylenedioxythiophene):poly(4-styrenesulfonate) (PEDOT:PSS). Commercial availability of PEDOT:PSS in aqueous solutions can be currently found at H.C. Starck and Agfa under the trade names Clevios and Orgacon, respectively. PEDOT:PSS has been mainly used as hole injection layer or buffer layer between ITO and the organic active layer [15]. Both commercial trades have also higher conductivity formulations of their PEDOT:PSS for conductive coating applications that are achieved by altering the ratio of PEDOT and PSS. For typical 1 µm thick higher-conductivity PEDOT:PSS films, the sheet resistance (Rs) is about 3 kΩ/◾, which translates to a conductivity of ~30 S/cm. Compared to typical ITO for OLED applications with a conductivity of ~20,000 S/cm (Rs = 15 Ω/◾, thickness = 130 nm, 90% transmittance), there seems to be a huge gap in conductivity in order to make this polymer conductive enough to replace ITO. Moreover, for a typical OLED application, PEDOT:PSS thickness is usually less than 1 µm in order to keep transparency high; thus the resulting conductivity at practical levels of transparency is even lower. Fortunately, there have been reports suggesting methods for increasing PEDOT:PSS conductivity without increasing the thickness, thus preserving the optical transparency, by the addition of controlled amounts of a polyalcohol such as glycerol [16,17] and sorbitol [18], or by mixing with various solvents such as dimethyl sulfoxide (DMSO), N,N-dimethyl formamide (DMF), and tetrahydrofuran (THF) [19]. Improvements of as much as two orders of magnitude were observed. Adding a polyalcohol has been said to decrease PEDOT:PSS resistivity by around five to six times, whereas mixing with DMSO could reduce this further by another 10-fold. Because of these reports, it is now possible to obtain much higher-conductivity films without sacrificing transparency.
Many research groups have published results of their OLEDs based on different combinations of PEDOT:PSS preparations. In 2002, Kim et al. [20] compared devices with high-conductivity PEDOT:PSS anodes with and without added glycerol. It was reported that these conducting polymer anodes compared well with ITO devices, but no actual comparison data were presented. It was also shown that devices with PEDOT:PSS with glycerol had a maximum luminescence of 1490 cd/m2, which was one order of magnitude higher than regular PEDOT:PSS electrode devices. Kim et al. [21] also published another set of results based on a highly conductive Orgacon PEDOT:PSS, which is capable of delivering 150 Ω/□ films at more than 70% visible transparency. It was reported that these devices showed lower turn-on voltages and better performance than the ITO anode devices at operating voltages lower than 8 V. This behavior is explained by a limited current injection at higher voltages due to the higher sheet resistance of the PEDOT:PSS. However, at low voltages, the better work function match of the conducting polymer to the emission layer wins over the ITO.
Further, similar results have been obtained by other groups with anodes based on high-conductivity PEDOT:PSS (Baytron P HC V4 and PH 500 from H.C. Starck) with added ethylene glycol and DMSO [22,23]. One report showed the results of a comparison between PH 500 + DMSO anodes and ITO transparent anodes in green, red, and blue OLEDs. For all the devices, the polymeric anodes showed better performance than ITO counterparts at low-drive voltages and luminescence. This performance was also attributed to a better work function match and lower hole injection barrier for PEDOT:PSS anodes, and also to its superior optical properties (lower refractive index). In another study, PH 500 and P HC V4 electrodes with conductivities of 300 and 450 S/cm, respectively, were compared with ITO. It was found that the surface morphology differences between P HC V4 and PH 500 gave the upper hand to PH 500 devices with higher luminescence. Further reducing the surface roughness with the addition of the methoxy-substituted 1,3,5-tris[4-(diphenylamino)phenyl]-benzene (TDAPB) hole transport (TDAPB HT) layer on the polymeric electrode not only reduced the surface roughness (0.3 nm) but also allowed for even higher luminescence (10,000 cd/m2) and better current efficiency up to 3.5 cd/A. Although these performance figures are quite respectable, they are still far from matching those of ITO-based electrode devices.
In this section, our discussion regarding CNTs for OLED applications will begin. CNTs were first introduced into OLEDs as CNT/polymer composites. Thus we will discuss these composites briefly as they were used in emission layers, hole injection layers, and conducting polymer electrodes. We will then give a more extensive insight into the current status of CNT technology as transparent conducting electrodes.
4.4.1.1 Composite Emission Layer
First use of CNTs in OLEDs was reported by Curran et al. in 1998 [24] in the form of “doping” agents for the emitting conjugated polymer material, poly(m-phenylenevinylene-co-2,5-dioctoxy-p-phenylenevinylene) (PmPV). General consensus indicates Transparent and Flexible Carbon Nanotube Electrodes that emitting polymers in OLEDs need to remain undoped in order to maintain efficient luminescence. Traditional dopants act as trapping sites, quenching the radiative decay of excitons, thus reducing the overall luminescence. However, polymers have low conductivity, which in turn requires the device to operate at very high voltages, and therefore induce large thermal strains on the device. To minimize thermal effects, polymers would require dopants to increase charge carrier mobility and electrical conductivity. Curran et al. have taken advantage of the properties of carbon nanotubes by dispersing small quantities of MWCNTs in PmPV to act as nanometric heat sinks and improve conductivity at the same time. It was claimed that MWCNTs do not chemically dope the polymer; thus the electronic processes of radiative exciton decay are not altered in the emissive polymer. It was found that incorporating nanotubes into PmPV led to an increase in conductivity up to eight orders of magnitude, which reduced the operating voltage considerably and, at the same time, had very low impact on the electroluminescent characteristics. It was also found that MWCNT/PmPV composite devices lasted up to five times longer in air, possibly an indication of the effective nanometric head sink function of the nanotubes.
4.4.1.2 Composite Buffer Layer
Further studies on CNT/polymer composites employing CNTs as composites in hole-blocking layers, electron transport layers, hole transport layers, and hole injection layers in OLED structures have been carried out. We will summarize some of those results here.
Early experiments [25,26] have found that SWCNTs in the hole transport layer (HTL) PmPV/SWCNT composites placed between the hole injection layer (HIL) and the emissive layer (EL) acted has a hole trapping material that block holes from crossing the HT layer. The same phenomenon was also found when SWCNTs were incorporated into PEDOT:PSS [27]. Devices showed poor performance with SWCNT/PEDOT:PSS composites as HIL. Such use of CNTs did not yield improved OLED performance.
However, later studies by Wang et al. [28,29] had different results with MWCNT/composite HILs, which showed lower turn-on voltages and higher luminance. Another recent study has also showed improved performance, in a different manner, for HIL PEDOT:PSS/MWCNT composite devices. Different MWCNT loadings in the composite [30] were studied. The loading of MWCNT was varied as follows: 0, 0.7, 1.5, and 2.5 wt%. It was seen that as the MWCNT ratio was increased in the HIL composite, the turn-on voltage also increased, which suggests that MWCNTs worsen the hole injection and transport efficiencies of the PEDOT:PSS layer. On the other hand, the current efficiency is increased as the MWCNT concentration is increased. This could be explained by charge carriers being trapped by the MWCNT, causing better balance of both carrier types, and thus an increased luminescence current efficiency.
Given that the different studies did not yield consistent results with improved performance in CNT/polymer composite HILs or HTLs, it is safe to conclude that such use of CNTs is not the best.
4.4.1.3 CNT Composite Electrodes
So far, all the devices mentioned containing CNT/polymer composites have been used in conjunction with an ITO electrode. However, there have also been studies that tried to use CNT composites as ITO replacement, the same way as it was for conducting polymers. An example of composite electrode devices will be shown here.
FIGURE 4.3 I-V-L curve of the OLEDs based on PEDOT:PSS/SWCNT and ITO anodes. (From Wang, G.-F., X.-M. Tao, and R.-X. Wang. Flexible organic light-emitting diodes with a polymeric nanocomposite anode. Nanotechnology, 2008; 19(14): 145201−1. With permission.)
Wang et al. [31] have fabricated SWCNT/PEDOT:PSS transparent electrode OLEDs and compared them to ITO devices. The devices were fabricated on PET substrates and bending tests were performed. With no surprise, bending test results showed minimal increase in resistance for SWCNT/PEDOT:PSS devices (10% after 1600 cycles). For ITO devices, however, the resistance increased by two orders of magnitude only after 20 cycles. Figure 4.3 shows the luminance curves for two anodes. The performance seems to be similar with lower turn-on voltage for SWCNT/PEDOT:PSS devices, which is not surprising given PEDOT:PSS has a higher work function than ITO leading to a better charge injection and low voltage drive. However, at higher voltage drive, ITO devices will have greater luminance due to a lower resistivity of the anode compared to composite anode devices.
In order to minimize complexity and keep device fabrication simple, transparent and conductive electrodes can also be fabricated solely using CNTs, and many research laboratories have been geared toward its achievement. In the upcoming sections, the fabrication of CNT thin films and OLED devices based on such electrodes will be discussed.
4.4.2.1 CNT Thin-Film Fabrication
There are several methods that can be used for fabricating CNT thin films such as drop-drying from solvent, spin coating, airbrushing, and Langmuir-Blodgett deposition [32,33,34,35]. Regardless of the method used for forming the actual CNT thin film, as a first step, carbon nanotubes, in their initial powder-like form, need to be dispersed in some kind of solution, either aqueous or in a solvent. Dispersing CNTs in solution often involves putting the solution with carbon nanotubes through a sonication process in order to unbundle the CNTs as much as possible. The sonication process is usually carried out using an ultrasonic bath or an ultrasonic probe. Although very effective at dispersing CNTs in a solution, sonication is also known to break CNTs to shorter lengths [36]; therefore the power and duration of the sonication need to be controlled carefully so that CNT lengths are preserved as much as possible in order to obtain better conductivities on the resulting film.
FIGURE 4.4 SWCNTs in aqueous solution after centrifugation.
After the sonication process, the solution of CNTs is then transferred to an ultra-centrifuge to separate larger bundles of CNT from smaller bundles or individual CNTs. Through ultracentrifugation, large undispersed bundles of nanotubes will then be trapped at the bottom of the solution, leaving perfectly homogeneously dispersed CNTs at the top portion. Thus only the top portion of the solution is collected and used in the subsequent steps of the CNT thin-film fabrication. Figure 4.4 shows solutions of SWCNTs dispersed in deionized (DI) water after ultracentrifugation. The different concentrations of SWCNTs in water were achieved by varying the sonication time. A lighter color of solution represents a lower concentration of CNTs with longer nano-tubes, whereas a darker color indicates a high concentration but with shorter nanotubes.
4.4.2.1.1 Vacuum Filtration
The most commonly used method for creating thin films of CNT from a dispersed solution is vacuum filtration. This method was originally developed by Wu et al. [37]. A membrane filter is used to filter a small quantity of a well-dispersed CNT in solution. Figure 4.5 shows a typical vacuum filtration system setup that can be used for making CNT thin films. The final thin film will be deposited on the membrane filter, followed by the transfer of the CNT onto the substrate or actual device. The type of filter used will be determined by the type of CNT solution used and the desired method of transfer of CNT thin film onto substrate, as will be explained later in this chapter. This method of film formation, vacuum filtration, has its advantages. First, the homogeneity of the CNT film is guaranteed by the process itself. As the CNTs are deposited on the filter, if thicker areas are developed, the filtration rate will also decrease in these localized areas, thus favoring the deposition of CNT in thinner areas where the permeation rate is higher, and so on. Therefore the resulting CNT thin film will be homogenous across the entire filtration area. The second advantage of this method comes from the fact that CNTs are extremely rigid for their size. Therefore they will tend to lie onto the filter straight, thus maximizing overlap and interpenetration with other CNTs within the film. This is the morphology required to maximize the electrical conductivity and mechanical integrity within the CNT thin film. Lastly, film thickness can be precisely controlled by the amount and concentration of CNT solution filtered through the membrane filter. Figure 4.6 shows resulting CNT thin films on cellulose ester membrane filters of different thicknesses. The amount of CNT filtered is proportional to the resulting conductivity of the film. However, the transparency of the film will suffer as the conductivity is increased. Therefore one needs to understand the trade-offs involved between transparency and conductivity.
FIGURE 4.5 Vacuum filtration system for CNT films.
4.4.2.2 Transfer of CNT Thin Film onto a Substrate
The final step in the processes of CNT film fabrication involves the transfer of the CNT film from the membrane filter onto the desired substrate. The original method developed by Wu et al. [37] uses a cellulose ester membrane filter that can be dissolved in organic solvents such as acetone. An illustration of the transfer and patterning process can be seen in Figure 4.7.
FIGURE 4.6 Different thicknesses of CNT thin films on mixed-cellulose ester membrane filters.
FIGURE 4.7 Transfer of CNTs from cellulose membrane filter to substrate by dissolving filter.
First, the membrane filter with the CNT film is placed face down onto the desired substrate. The cellulose ester membrane can then be dissolved by immersing the substrate with the filter in an organic solvent or just placed in a vapor bath of the organic solvent for a couple of hours. A drawback of this method is that the substrate needs to be able to resist organic solvents, and the device structure is limited by having CNT electrodes as the first structure on the device so that subsequent process steps are not affected; furthermore, in order to obtain the desired pattern of CNTs, one needs to use a traditional photolithographic process or a precisely cut shadow mask. A mask is required to protect wanted portions of the CNT film when the substrate is subjected to an oxygen plasma reactive ion etching to etch away CNTs from unprotected parts. If a photolithographic technique is used to create the desired electrode pattern, given the nature of this process, there is a very high probability of contamination of the substrate throughout the process. Furthermore, the porosity of CNTs makes the removal of residual photoresist a laborious task.
FIGURE 4.8 Transfer of CNTs from alumina membrane filter to substrate using a PDMS stamp.
A modified method for transferring CNT films from vacuum filtration was demonstrated by Zhou et al. [38] with the use of alumina membrane filters instead of cellulose ester membrane filters. The CNT film on alumina membrane filters can be transferred by picking up CNTs from the filter with a poly(dimethyl siloxane) (PDMS) stamp and then stamping the CNTs directly onto the desired surface. The latter method has the advantages of being a direct patterning and a “cleaner” method, and also of being able to deposit CNTs as intermediate or final layers of the device. This method is versatile and fast, as the deposition surfaces or substrates do not need to be solvent resistant.
A commonly used PDMS is provided by Dow Corning, Sylgard 184. The fabrication of a PDMS stamp is straightforward. A master mould needs to be first fabricated using traditional photolithography on a substrate, e.g., a silicon wafer. The PDMS base and curing agent are then mixed and poured onto the mold and cured in an oven. Once cured, the resulting PDMS stamp is peeled off the master mold and is ready to be used. Figure 4.8 illustrates the transfer process using a PDMS stamp. The transfer of the CNTs on the alumina membrane filter begins by pressing the PDMS stamp against the CNT film on top of the alumina filter. CNTs will adhere only to the protrusions of the PDMS stamp. Finally, the stamp is pressed against the substrate where CNT electrodes should be deposited, and CNTs are transferred from the stamp to the substrate, completing the electrode fabrication process. The curing time and temperature of the PDMS stamp are crucial for achieving a successful transfer of CNTs onto the substrate. Changing the curing time and curing temperature will change the adhesion of the PDMS surface. The right conditions must be found in order for the stamp to pick up the CNTs from the membrane filter and be able to release them onto the desired substrate. For example, curing for too long or raising the temperature too much will render the stamp too hard and it will lose its flexibility, and therefore will be unable to transfer the CNT films properly [39]. Figure 4.9 shows a sample of SWCNT electrodes transferred onto a poly(ethylene terephthalate) (PET) substrate using the above-mentioned transfer method.
FIGURE 4.9 Photograph of SWCNT electrodes deposited on a PET substrate by PDMS stamping.
4.4.2.3 SWCNT Electrode OLED Devices
OLEDs with SWCNT transparent electrodes were first reported by Aguirre et al. [40] in 2006. This group fabricated SWCNT electrodes on glass substrates using the vacuum filtration method and transfer from mixed-cellulose ester membrane filters.
The SWCNTs used in this experiment were produced in-lab by a pulsed laser vaporization technique and purified following a standard procedure. The purification process aims at eliminating amorphous carbon and metal catalyst impurities, and results in p-type charge transfer doping of the CNTs. The purified CNT powder was then dispersed in a 2% sodium cholate solution and centrifuged at 5000 g for 2 h.
The centrifuged solution was then used to filter several CNT sheets of various thicknesses and transferred onto clean glass slides by dissolving the cellulose filter in acetone. Figure 4.10 shows the resulting sheet resistance of the CNT thin film measured by a four-point probe versus the thickness and transmittance of the film for a wavelength of 520 nm. Electrical contacts to the CNT electrodes were made by evaporating 50 nm of Ti to one end of the CNTs.
The device structure of the SWCNT OLED is shown in Figure 4.11a. An ultra-thin (~1 nm) buffer layer of parylene was also deposited by CVD between the CNTs and the stacked OLED layers, which was absent in the control ITO OLEDs present. The addition of this layer claims to improve the wetting and adhesion of the evaporated organic semiconductor layer to the CNTs. The organic semiconductor layers that formed the OLED consisted of a 10 nm copper phthalocyanine (CuPc) hole injection layer (HIL), a 50 nm N,N-bis-1-naphthyl-N,N-diphenyl-1,1-biphenyl-4,4-diamine (NPB) hole transport layer (HTL), and a 50 nm tris-8-hydroxyquinoline aluminum (Alq3) electron transport layer (ETL) and emissive layer (EL) deposited in a thermal evaporator. The surface roughness of the CNT film was measured using atomic force microscopy (AFM) at 12 nm rms. The considerable roughness of the SWCNT films imposes a lower limit to the thickness of the organic layers. Therefore 100 nm layers of NPB and Alq3 were used instead for SWCNT OLEDs. The cathode in both ITO OLED and SWCNT OLED was made by evaporating 1 nm of lithium fluoride and 50 nm of aluminum. Worth noting, Figure 4.11b shows a scanning electron microscope (SEM) image of the cross section of the SWCNT OLED. The sample for the SEM image was prepared by cleaving the glass substrate at the center Transparent and Flexible Carbon Nanotube Electrodes of the emissive area. It can be seen that the SWCNT film network is overhanging across the glass substrate where the break was made, thus illustrating the flexibility and fabric quality of the SWCNT thin film.
FIGURE 4.10 Sheet resistance versus thickness and transmittance of SWCNT thin film. (From Aguirre, C.M., et al., Applied Physics Letters, 2006; 88: 183104-1. With permission.)
FIGURE 4.11 (a) Device structure of SWCNT OLED. (b) Corresponding cross-sectional scanning electron microscopy image at a broken edge taken at a 20° angle from the surface normal. (From Aguirre, C.M., et al., Applied Physics Letters, 2006; 88: 183104-1. With permission.)
FIGURE 4.12 Current density (squares) and luminance (circles) as a function of applied voltage for OLEDs fabricated (a) on carbon nanotube anodes (SWNT OLED) and (b) on oxygen-plasma-treated ITO anodes (ITO OLED). (From Aguirre, C.M., et al., Applied Physics Letters, 2006; 88: 183104-1. With permission.)
The maximum luminescence reported for these devices was 2800 cd/m2 with 1.2 cd/A current efficiency at 20 V. These results are said to be comparable to ITO devices with the same structure with maximum luminescence of 6000 cd/m2 and 1.9 cd/A current efficiency. Detailed I-V curve characteristics are shown in Figure 4.12. The measured turn-on voltage of the SWCNT OLED was 6.6 V, only slightly higher than the ITO OLED, 6.2 V, despite the much thicker organic semiconductor layers. Furthermore, taking into consideration that the SWCNT film used in the maximum luminescence reported has a transmittance of only 44%, compared to ITO’s 90%, it is possible to conclude that SWCNT injection properties are very similar to those of ITO.
Two other publications involving SWCNT OLEDs were presented during the same year. These two research groups used commercially available SWCNTs instead to carry out their experiments. Li et al. [41] fabricated their CNT electrodes using the vacuum filtration method and PDMS stamping on PET substrates. Multiple combinations of spin-coated HILs, HTLs, and ELs were explored. Devices based on ITO electrodes on PET substrate were also fabricated as reference. The maximum luminescence and current efficiency for CNT devices were 3500 cd/m2 and 1.6 cd/A, respectively, compared to 20,000 cd/m2 and 6 cd/A for ITO on PET samples. However, the lifetimes of those devices were measured to be similar at ~50 h with an initial brightness of 1400 cd/m2.
FIGURE 4.13 SWCNT electrode OLED structure on PET. (From Li, J., et al., Nano Letters, 2006; 6(11): 2472–2477. With permission.)
SWCNT devices that were fabricated included PEDOT:PSS as HIL, PEDOT:SS with MeOH as HIL, and no HIL (see Figure 4.13). Also, devices with poly(9,9-dioctylfluorene-co-N-(4-butylphenyl)diphenylamine) (TFB) + poly(9,9-dioctylfluorene-co-benzothiadiazole (BT) and BT as EL were compared. TFB is a hole transport polymer, and BT is an electron-dominated emission material. Additionally, all the devices included a layer of TFB + 4,4′-bis[(p-trichlorosilylpropylphenyl)phenylamino] (TPD-Si2) as HTL. The following observations were found. First, devices with HIL and HTL performed better, which could indicate two things: PEDOT:PSS is effective in planarizing the CNTs and lowering the barrier for hole injection. Furthermore, devices with PEDOT:PSS with MeOH performed better than with PEDOT:PSS alone, which suggests that MeOH help reduce the surface roughness, and thus minimize leakage current. Lastly, devices with BT only as EL showed the best performances, indicating that SWNTs are more efficient hole-injecting electrodes than ITO since ITO devices show the opposite with better devices using TFB + BT as EL.
In the experiment from Zhang et al. [42], different SWCNT sources obtained from different growth methods are compared. It was found that P3 nanotubes (arc discharge from Carbon Solutions, Inc.) exhibit much greater surface smoothness and lower resistivity than HiPCO nanotubes (Carbon Nanotechnology, Inc.) when the CNT film is formed using the PDMS stamping method. The OLED devices were fabricated on glass substrates with thermally evaporated HTL (NPD) and EL (Alq3) layers and a spin-coated HIL layer (PEDOT:PSS). However, the performances obtained are below those of other reports using a similar structure. The maximum luminescence was only 17 cd/m2 at 20 V with a turn-on voltage of 5 V for P3 electrode devices. The authors attribute this poor performance to the surface roughness, high resistivity, and low work function of SWCNTs compared to ITO electrodes.
4.4.2.4 Top Emission SWCNT Electrode OLED
So far, only bottom emission devices were discussed. However, in 2010, Chien et al. [39] successfully demonstrated the first top emission SWCNT electrode OLED based on solution-processed HIL and EL.
In this report, raw SWCNTs were purchased commercially from Carbon Solutions, Inc. (P2-SWNT). P2-SWNTs are purified CNTs (>90%) with low functionality and used as received without further treatment. The SWCNTs in powder form were prepared into a colloidal aqueous solution of 0.1 wt% of CNT in a 1 wt% surfactant solution in deionized water. The surfactant used was sodium dodecyl sulfate (SDS). The CNT/SDS solution was left in an ultrasonic bath for 24 h and was followed by a centrifugation at 30,000 rpm (154,000 g) for 1 h. Only the top portion of the CNT solution was extracted for use in the following steps, as the centrifugation process removes the larger bundles of CNTs to the bottom of the tube that the ultrasonic bath was not able to disperse. The result is a highly homogenous solution of CNT dispersed in deionized water.
In order to build a top-emitting device over organic semiconductor layers, the process needs to be carried out at low temperature to avoid damaging the organic layers underneath. The top transparent electrode was therefore deposited using PDMS stamping of the SWCNT method on top of the PEDOT:PSS HIL, which was previously spin-coated over the EL. Such a process offers the possibility of implementing CNT OLED devices onto any type of the substrates, including nontransparent surfaces. Vacuum filtration through 0.1 µm porosity alumina filters (Whatman, Inc.) was carried out to obtain a uniform CNT film, and the transfer process was carried out as described earlier for PDMS stamping.
The top emission SWCNT OLEDs were fabricated on glass substrates in the initial investigation to determine the feasibility of such a procedure. Figure 4.14 shows the cross-sectional structure of the top emission OLED. The bottom cathode consists of a layer of Al (100 nm) and LiF (1 nm) evaporated through a shadow mask in a thermal evaporator at around 1 × 10–6 Torr. The organic emissive material was then spin-coated from a solution, and heat-treated in ambient air at 100°C for 5 min. The chosen organic material was prepared according to Park et al. [43] and consisted of a blend of poly(vinylcarbazole) (PVK), 2-(4-biphenylyl)-5-(4-tert-butylphenyl)-1,3,4 oxadiazole (PBD), tris(2-phenyl-pyridinato) iridium (Ir(ppy)3), and N,N′-diphenyl-N,N′-bis(3-methylphenyl)-1,1′-biphenyl-4,4′-diamine (TPD) into mixed solvent of 1,2-dichloroethane and chloroform. PEDOT:PSS (CLEVIOS P VP AI 4083 from H.C. Starck) was then spin-coated on top of the emission layer. The PEDOT:PSS was diluted with 10 wt% of 2-propanol to promote the wetting on the emission layer in order to obtain a more uniform buffer layer. The PDMS stamp was then used to transfer the CNTs from the alumina membrane filters onto the PEDOT:PSS layer. This process takes place on a hot plate. The appropriate temperature to transfer CNTs on top of a PEDOT substrate was determined to be 100°C. The PDMS “wetted” with CNTs is put in conformal contact with the PEDOT:PSS layer with light pressure from a fingertip for 30 s. The PDMS stamp is then carefully peeled off, leaving the CNT film on the PEDOT:PSS layer. Heating the substrate while stamping the CNTs serves two purposes. It helps the release of the CNTs from the PDMS and also helps the adhesion of the CNTs to the PEDOT:PSS layer. The top emission OLED is then completed by thermally evaporating through another shadow mask a layer of Al on the CNT film outside the active region in order to make the electrical contact of the CNTs. Some research groups [37,42,44] have deposited additional layers of Ti and Pd on the CNTs to obtain true ohmic contacts with the CNTs; however, Chien et al. did not find this to be necessary as the devices performed adequately to be able to take reliable measurements.
FIGURE 4.14 Top emission SWCNT OLED device structure.
CNT films of various thicknesses were also fabricated and characterized. Resistivity (Rs) and transmittance (%T) of the CNT films were measured with a four-point probe and a UV-visible spectrometer, respectively. For these measurements, the CNTs were transferred on a plain glass substrate spin-coated with PEDOT:PSS to simulate the surface conditions of a real device. As mentioned previously, resistivity and transmittance are affected by the thickness of the CNT film. As the thickness of the CNTs is decreased, the film becomes less conductive and more transparent. Figure 4.15 shows the resistivity versus the transmittance of the film at 510 nm corresponding to the peak emission of a green emission OLED. For an Rs: 100 Ω/sq CNT film, the transmittance is about 60%. At 90% transmittance, the film resistivity reaches 500 Ω/sq. These values are on par with numbers reported by other groups with films transferred from cellulose or alumina filters and on glass or plastic substrates. The devices in this report were fabricated using the 90% transmittance films. Figure 4.16 shows an AFM image of the CNT on a complete top emission device. The surface RMS roughness is about 7 nm, which is equivalent to the roughness of CNT films alone, showing that the transfer on top of the OLED devices does not affect the characteristics of the CNT films.
FIGURE 4.15 Sheet resistance as a function of transmittance of CNT films transferred onto the device measured for a 510 nm emission.
FIGURE 4.16 AFM image of CNT film already transferred onto an actual device (Rs: 500 Ω/sq, %T: 90%). Film RMS roughness is about 7 nm.
FIGURE 4.17 I-V curve characteristics of top emission SWCNT OLED.
OLED devices were encapsulated in a nitrogen glove box immediately after fabrication and were characterized with Keithley 2400 and Keithley 2610 DC source monitors. Luminance measures were taken with a photodiode with an infrared (IR) cutoff filter, which was calibrated with a Delta OHM HD 2102.1 photometer for their green OLEDs. About 30 top emissions with CNT anode OLEDs were fabricated in total. However, only half the devices had complete CNT film transfers. For these devices, the performance characteristics were very similar. Figure 4.17 shows the current density, luminance, and current efficiency as a function of voltage applied of a typical device. Maximum luminance and current efficiency obtained were 3588 cd/m2 and 1.24 cd/A, respectively. These numbers are comparable to devices reported by other groups for bottom emission and evaporated organic material devices and CNT electrodes. Figure 4.18 shows a photograph of an actual solution-processed top emission OLED with a CNT transparent anode. As can be seen, the light emission of the device is very uniform and comparable to typical OLEDs with ITO electrodes. However, even though these top emission OLEDs with CNT electrodes have excellent results compared to other CNT electrode devices reported in the literature, bottom emission devices with ITO electrodes fabricated in our lab, with the same spin-coated organic material, still have better performance than the top emission CNT devices. Optimized ITO electrode devices with luminance around 30,000 cd/m2 were achieved. It is believed that if further optimization is conducted, better performance should be expected from CNT top emission devices. For example, by reducing the resistivity of CNT films, better performances should be expected. This can be achieved by optimizing the CNT solution dispersion process and through chemical treatment of the prepared films. Additionally, charge injection properties from CNTs into the organic materials is of great importance in order to obtain better hole injection and PEDOT, which is appropriate for ITO electrodes, should be replaced by a more appropriate material for use with CNT electrodes. Finally, better control of the transferring process of CNTs with the help of a mechanical instrument is desired to obtain more reliable and consistent results.
FIGURE 4.18 Photograph of a working top emission SWCNT-OLED.
4.4.2.5 MWCNT Electrode OLED Devices
The first reported OLED devices using CNTs only as transparent hole-injecting electrodes were fabricated using MWCNTs on glass substrate [45]. This publication reported devices with spin-coated emission layers and hole injection layers, MEHPPV and PEDOT:PSS, respectively. The achieved performances were 500 cd/m2 for maximum luminescence, and a low turn-on voltage of 2.4 V. There was no mention of the efficiencies obtained. Additionally, this report uses a unique method in fabricating thin MWCNT films. The authors call it a dry solid-state process where transparent CNT sheets are drawn from a sidewall of MWCNT forests. Figure 4.19 illustrates such a process. Figure 4.19A shows a photograph of a self-supporting 3.4 cm wide and 1 m long MWCNT sheet that was hand drawn from a CNT forest. Figure 4.19B and C shows SEM images of the process at 35° and 90° angles with respect to the forest plane, respectively. Figure 4.19D shows a two-dimensionally reinforced structure of four overlaying MWCNT sheets with 45° shift in orientation between them. In order to make useful transparent conducting electrodes for organic devices, such sheets can also be directly placed onto a substrate and immersed momentarily in ethanol and pulled out vertically, which increases the density and transparency of the MWCNT sheets. One of the major advantages of such a process is the absence of sonication, which is known to shorten CNTs, thus decreasing electrical and thermal conductivities and mechanical properties.
FIGURE 4.19 MWCNT forest conversion into sheets (dry solid-state process). (From Zhang, M., et al., Science, 2005; 309(5738): 1215–1219. With permission.)
Another paper on MWCNT OLEDs was presented by Williams et al. [46] as a continuation to the previous publication focuses on improving the surface roughness of the MWCNT sheet by spin coating up to nine layers of PEDOT:PSS. Maximum luminance with 4500 cd/m2 with maximum current efficiency of 2.3 cd/A was achieved with this approach. This time, however, vacuum-evaporated active layers of α-NPB and Alq3 were deposited as active OLED layers.
FIGURE 4.20 (a) Photograph of working OLED with CNT electrode. (b) Structure of CNT OLED. (From Ou, E.C.W., et al., ACS Nano, 2009; 3(Compendex): 2258–2264. With permission.)
4.4.2.6 Summary of CNT Electrode OLED Performance Criteria
A recent report by Ou et al. [47] provided a good insight into the key elements that are important to the fabrication of a CNT-only transparent anode for OLEDs that performs well.
In this experiment, CNT thin films were deposited on PEN and PET substrates using a high-speed roll-to-roll slot die method. This report shows the highest-performing OLEDs with CNT electrodes to date with a luminescence of 9000 cd/m2 and a current efficiency of 10 cd/A. The active emission layer consisted of N,N-diphenyl-1,1-bihyl-4,4-diamine (NPB) and tris-(8-ydroxquinoline) aluminum (Alq3) coevaporated with coumarin 545. Figure 4.20 shows a photograph of a working CNT OLED and its corresponding structure. The experiments were centered around the use of a proprietary HIL that contains high-conductivity PEDOT:PSS and poly(ethylene glycol) (PEG), referred to as PSC in the report. Key elements to obtaining high-performance OLEDs with CNT electrodes discussed were surface roughness, sheet resistance, and work function of the injecting electrodes.
First, the surface roughness of the anode is important because protruding CNTs (Figure 4.21) generate very high local electrical fields, which can in turn cause local device failure and lead to a shorted device. Ou et al. reduced the surface roughness of their electrodes by more than half in a two-step process consisting of the deposition of a 5 nm polyvinylpyrrolidone (PVP) polymer with the Meyer rod method and ozone plasma treating the samples, followed by the spin coating of PSC (HIL). The resulting surface roughness can be seen in Table 4.1, comparing with the initial CNT film roughness.
FIGURE 4.21 SEM image of a SWCNT surface with occasional protruding CNT. (From Hu, L., et al., Nanotechnology, 2010; 21(Compendex). With permission.)
TABLE 4.1
Surface Roughness Change after PSC Coating on CNTs
Sample |
Description |
Initial Roughness (nm) |
Final Roughness after Spin-Coated with PsC (nm) |
2 |
PEN/CNT/PSC |
14.55 |
6.0 |
3 |
PEN/CNT/5 nm polymer/PSC |
14.65 |
4.6 |
4 |
PET/CNT/PSC |
9.3 |
5.9 |
5 |
PET/CNT/5 nm polymer/PSC |
9.72 |
5.2 |
6 |
PET/Doped CNT/PSC |
10.85 |
6.0 |
Source: Ou, E.C.W., et al., ACS Nano, 2009; 3(Compendex): 2258–2264. With permission.
Second, the sheet resistance and work function of the hole-injecting electrode were also improved by doping the CNT with HNO3, which resulted in a 50% sheet resistance drop and an increase of 0.13 eV for the work function. CNTs doped with HNO3 showed higher luminescence and lower turn-on voltage than undoped ones (see Figure 4.22).
Finally, to further illustrate the importance of the sheet resistance of the electrode in an OLED, PEN/PSC anode (310.5 ohms/sq) and anode (102.9 ohms/sq) devices were compared. As shown in Figure 4.23, the PEN/CNT/PSC anode OLED was a huge improvement over the PEN/PSC anode OLED, with as much as three times the luminescence for the same driving voltage.
The development of a true high-performance transparent flexible electrode to be used for OLED fabrication is still a challenge to be resolved. Among various alternatives the use of CNT electrodes is one of the more promising approaches. CNT electrodes with sheet resistivities around 100 ohms/sq and transparencies around 60% have been produced. OLEDs made with those electrodes have shown high luminescence and current efficiency. Even though those performances are still inferior to the ones obtained with standard ITO electrodes, various approaches for reducing resistivity of the CNT films without losing transparency are presently under study and should lead to the development of real flexible displays in the near future.
FIGURE 4.22 Luminescence versus voltage of OLEDs for PET/CNT/PEDOT:PSS versus PET/doped CNT/PEDOT:PSS. (From Ou, E.C.W., et al., ACS Nano, 2009; 3(Compendex): 2258–2264. With permission.)
FIGURE 4.23 Luminescence characteristics of PEN/PSC and PEN/CNT/PSC devices. (From Ou, E.C.W., et al., ACS Nano, 2009; 3(Compendex): 2258–2264. With permission.)
1. Yilun, L., et al. Engineering the band gap of carbon nanotubes. In 8th IEEE Conference on Nanotechnology 2008 (NANO ’08). 2008.
2. Tang, C.W., and S.A. VanSlyke. Organic electroluminescent diodes. Applied Physics Letters, 1987; 51(12): 913–915.
3. Kim, H., et al. Transparent conducting electrode materials grown by pulsed laser deposition for organic light-emitting devices. In Organic Light-Emitting Materials and Devices III, July 19–21, 1999.
4. Vaufrey, D., et al. Sol gel deposited Sb doped SnO2 as transparent anode for OLED: Process, patterning and hole injection characteristics. In Organic Light-Emitting Materials and Devices V, July 30–August 1, 2001. 2002.
5. Besbes, S., et al. Effect of surface treatment and functionalization on the ITO properties for OLEDs. Materials Science and Engineering C, 2006; 26(2–3): 505–510.
6. Gardonio, S., et al. Degradation of organic light-emitting diodes under different environment at high drive conditions. Organic Electronics, 2007; 8(1): 37–43.
7. Melpignano, P., et al. Mechanism of dark-spot degradation of organic light-emitting devices. Applied Physics Letters, 2005; 86(4): 041105-3.
8. de Jong, M.P., L.J. van Ijzendoorn, and M.J.A. de Voigt. Stability of the interface between indium-tin-oxide and poly(3,4-ethylenedioxythiophene)/poly(styrenesulfonate) in polymer light-emitting diodes. Applied Physics Letters, 2000; 77(14): 2255–2257.
9. Tolcin, A.C. Mineral commodity summaries. U.S. Geological Survey, January 2010.
10. Lin, Y.C., W.Q. Shi, and Z.Z. Chen. Effect of deflection on the mechanical and optoelectronic properties of indium tin oxide films deposited on polyethylene terephthalate substrates by pulse magnetron sputtering. Thin Solid Films, 2009; 517(5): 1701–1705.
11. Shirakawa, H., et al. Synthesis of electrically conducting organic polymers: Halogen derivatives of polyacetylene, (CH)x. Journal of the Chemical Society, Chemical Communications, 1977; 1977(16): 578–580.
12. Gustafsson, G., et al. Flexible light-emitting diodes made from soluble conducting polymers. Nature, 1992; 357(6378): 477–479.
13. Yang, Y., and A.J. Heeger. Polyaniline as a transparent electrode for polymer light-emitting diodes: Lower operating voltage and higher efficiency. Applied Physics Letters, 1994; 64(10): 1245–1247.
14. Karg, S., et al. Increased brightness and lifetime of polymer light-emitting diodes with polyaniline anodes. Synthetic Metals, 1996; 80(2): 111–117.
15. Fung, M.K., et al. Anode modification of polyfluorene-based polymer light-emitting devices. Applied Physics Letters, 2002; 81(8): 1497–1499.
16. Ghosh, S., and O. Inganäs. Nano-structured conducting polymer network based on PEDOT-PSS. Synthetic Metals, 2001; 121(1–3): 1321–1322.
17. Makinen, A.J., et al. Hole injection barriers at polymer anode/small molecule interfaces. Applied Physics Letters, 2001; 79(5): 557–559.
18. Granlund, T., L.A.A. Pettersson, and O. Inganas. Determination of the emission zone in a single-layer polymer light-emitting diode through optical measurements. Journal of Applied Physics, 2001; 89(11): 5897–5902.
19. Kim, J.Y., et al. Enhancement of electrical conductivity of poly(3,4-ethylenedioxythiophene)/poly(4-styrene sulfonate) by a change of solvents. Synthetic Metals, 2002; 126(2–3): 311–316.
20. Kim, W.H., et al. Molecular organic light-emitting diodes using highly conducting polymers as anodes. Applied Physics Letters, 2002; 80(20): 3844–3846.
21. Kim, W., et al. Efficient silole-based organic light-emitting diodes using high conductivity polymer anodes. Chemistry of Materials, 2004; 16(Compendex): 4681–4686.
22. Fehse, K., et al. Highly conductive polymer anodes as replacements for inorganic materials in high-efficiency organic light-emitting diodes. Advanced Materials, 2007; 19(3): 441–444.
23. Kajii, H., et al. Organic light-emitting diodes with highly conductive polymer electrodes as anode and their stress tolerance. Japanese Journal of Applied Physics, 2008; 47(1 Part 2): 460–463.
24. Curran, S.A., et al. A composite from poly(m-phenylenevinylene-co-2,5-dioctoxy-p-phenylenevinylene) and carbon nanotubes: A novel material for molecular optoelectronics. Advanced Materials, 1998; 10(14): 1091–1093.
25. Woo, H.S., et al. Hole blocking in carbon nanotube-polymer composite organic light-emitting diodes based on poly (m-phenylene vinylene-co-2, 5-dioctoxy-p-phenylene vinylene). Applied Physics Letters, 2000; 77(9): 1393–1395.
26. Woo, H.-S., et al. Tailoring hole transport in organic light-emitting devices using carbon nanotube-polymer nanocomposites. Journal of the Korean Physical Society, 2004; 45: 507–511.
27. Woo, H.S., et al. Organic light emitting diodes fabricated with single wall carbon nano-tubes dispersed in a hole conducting buffer: The role of carbon nanotubes in a hole conducting polymer. In Fourth International Topical Conference on Optical Probes of pi-Conjugated Polymers and Photonic Crystals, February 2000. Switzerland: Elsevier, 2001.
28. Wang, G.-F., et al. Improvement in performance of organic light-emitting devices by inclusion of multi-wall carbon nanotubes. Journal of Luminescence, 2007; 126(Compendex): 602–606.
29. Wang, G.-F., X.-M. Tao, and R.-X. Wang. Fabrication and characterization of OLEDs using PEDOT:PSS and MWCNT nanocomposites. Composites Science and Technology, 2008; 68(Compendex): 2837–2841.
30. Su, S.-H., et al. Enhancing efficiency of organic light-emitting diodes using a carbon-nanotube-doped hole injection layer. Japanese Journal of Applied Physics, 2010; 49(Compendex).
31. Wang, G.-F., X.-M. Tao, and R.-X. Wang. Flexible organic light-emitting diodes with a polymeric nanocomposite anode. Nanotechnology, 2008; 19(14): 145201-1.
32. Duggal, R., F. Hussain, and M. Pasquali. Self-assembly of single-walled carbon nano-tubes into a sheet by drop drying. Advanced Materials, 2006; 18(Compendex): 29–34.
33. Southard, A., et al. Solution-processed single walled carbon nanotube electrodes for organic thin-film transistors. Organic Electronics: Physics, Materials, Applications, 2009; 10(Compendex): 1556–1561.
34. Li, J., and Y. Zhang. Langmuir-Blodgett films of single-walled carbon nanotubes. Carbon, 2007; 45(3): 493–498.
35. Jong Hyuk, Y., et al. Fabrication of transparent single wall carbon nanotube films with low sheet resistance. Journal of Vacuum Science and Technology B (Microelectronics and Nanometer Structures), 2008; 26: 851–855.
36. Vichchulada, P., et al. Sonication power for length control of single-walled carbon nano-tubes in aqueous suspensions used for 2-dimensional network formation. Journal of Physical Chemistry C, 2010; 114(29): 12490–12495.
37. Wu, Z., et al. Transparent, conductive carbon nanotube films. Science, 2004; 305(5688): 1273–1276.
38. Zhou, Y., L. Hu, and G. Gruner. A method of printing carbon nanotube thin films. Applied Physics Letters, 2006; 88(12): 123109-3.
39. Chien, Y.-M., et al. A solution processed top emission OLED with transparent carbon nanotube electrodes. Nanotechnology, 2010; 21(13): 134020.
40. Aguirre, C.M., et al. Carbon nanotube sheets as electrodes in organic light-emitting diodes. Applied Physics Letters, 2006; 88: 183104-1.
41. Li, J., et al. Organic light-emitting diodes having carbon nanotube anodes. Nano Letters, 2006; 6(11): 2472–2477.
42. Zhang, D., et al. Transparent, conductive, and flexible carbon nanotube films and their application in organic light-emitting diodes. Nano Letters, 2006; 6(9): 1880–1886.
43. Park, J.H., et al. Double interfacial layers for highly efficient organic light-emitting devices. Applied Physics Letters, 2007; 90(15): 153508-3.
44. Nosho, Y., et al. Relation between conduction property and work function of contact metal in carbon nanotube field-effect transistors. Nanotechnology, 2006; 17(14): 3412–3415.
45. Zhang, M., et al. Strong, transparent, multifunctional, carbon nanotube sheets. Science, 2005; 309(5738): 1215–1219.
46. Williams, C.D., et al. Multiwalled carbon nanotube sheets as transparent electrodes in high brightness organic light-emitting diodes. Applied Physics Letters, 2008; 93(Compendex).
47. Ou, E.C.W., et al. Surface-modified nanotube anodes for high performance organic light-emitting diode. ACS Nano, 2009; 3(Compendex): 2258–2264.
48. Hu, L., et al. Flexible organic light-emitting diodes with transparent carbon nanotube electrodes: Problems and solutions. Nanotechnology, 2010; 21(Compendex).