Conducting Polymer Composite

There has been continuous interest in the use of conjugated polymers for the  fabrication of numerous light and/or foldable electronic devices, for example, electrochromic displays, microelectronic devices, protective coatings, and rechargeable batteries. The motivation behind this intense interest is because of their unique electronic and optical properties, ability to be chemically tuned, most importantly their lightweight/foldable mechanical properties, processability, and low cost. They are also extremely a promising candidate for sensor applications because of their conductivity and electrochemical activity that are extremely sensitive to molecular interactions, which provide excellent signal transduction for molecular detection (Gardner et al. 1992; Janata and Josowicz 2003; Chaurasia et al. 2012). The uprising popularity of conducting polymer-based sensors lies in the fact that only specific chemicals can trigger a drastic conductance change. By functionalizing polymer molecules, it can be made more specific (Gao et al. 2003; Ramanathan et al. 2004). However, the limitations such as relative low conductivity and low mechanical and chemical stabilities restrict its use for some practical applications.

Potential Electronic Properties and Application of Various Carbon Nanotube (CNT)/Polymer Composites

Nowadays, tremendous efforts have been made to prepare polymer and carbon nanotube composites (because of remarkable electrical as well as thermal conductivities and the superior mechanical properties of carbon nanotubes (CNTs)) with the aim of synergistically combining the merits of each individual (Tasis et al. 2006; Iijima 1991; Ajayan et al. 1994; Dai and Mau 2001; Zengin et al. 2002; Cochet et al. 2001; Sainz et al. 2005; Moniruzzaman and Winey 2006) component. To form a perfect polymer/CNT composite with much more enhanced functionality, in situ polymerization of the desired monomers in the presence of carbon nanotubes would be expected to show best results compared to the post-mixing approaches (Cochet et al. 2001; Sainz et al. 2005; Li et al. 2003). But, well dispersion of the carbon nanotubes into solution is a must for in situ polymerization. There are reports of different dispersal approaches which impart different surface chemistries and electronic structures to the carbon nanotubes such as polymer wrapping, (Zheng et al. 2003; Dalton et al. 2000; Star et al. 2001) noncovalent adhesion of small molecules, (Dai and Mau 2001; Chen et al. 2002) and acidic oxidation (Wang et al. 2005). The key to the expected improvement in the nanocomposites depends critically on the monomer–nanotube interfacial chemical and electronic interactions during polymerization and polymer–nanotube interfacial interactions after polymerization. There are reports on the impacts of the surface chemistry and electronic structure of carbon nanotubes on the kinetics of polymerization and the electronic performances of the obtained composites (Cheung et al. 2009). Because of their well-documented unique surface chemistry and electronic structures, they have used single-stranded DNA dispersed and functionalized single-walled carbon nanotubes (ss-DNA-SWNTs) as a representative example. They have also discussed the multiple roles of ss-DNA-SWNTs during and after the in situ polymerization in the fabrication of highly conductive self-doped polyaniline/SWNT composites. The applications of these nanocomposites cover the wide area of biosensing and flexible electronics which will be discussed below.
Nowadays, there is increasing efforts for the use of single-walled carbon nanotube (SWNT) networks as sensing materials and conductive flexible electrodes due to its specific advantages. Fabrication of SWNT films can be done quite easily by various room temperature solution-based processes, such as spray coating, (Kaempgen et al. 2005; Artukovic et al. 2005) inkjet printing, (Kordás et al. 2006; Simmons et al. 2007) deposition by a layer-by-layer approach, (Shim et al. 2007; Kovtyukhova and Mallouk 2005) and deposition through a filter (Wu et al. 2004; Zhang et al. 2006a). Due to statistical averaging effects, the obtained network electrodes are highly reproducible and exhibit percolation-like electrical conductivity. A number of applications including electrodes for solar cells, (Rowell et al. 2006) organic light-emitting diodes, (Li et al. 2006) smart windows, (Gruner 2006) sensors (Ferrer-Anglada et al. 2006), and transparent transistors (Artukovic et al. 2005; Chaurasia et al. 2012) where SWNT networks can be very useful. But, all the experimentally measured conductivities of the SWNT networks are considerably lower than the conductivity of a SWNT rope (axial conductivity around 10,000–30,000 S/cm) (Thess et al. 1996). It has also been noted that the conductivity of the SWNT networks decreases as the temperature drops (Bekyarova et al. 2005). The existence of high junction resistance and tunneling barriers between nanotubes (which dominate the overall film conductivity in the network) is the result of its low conductivity and the strong temperature dependence conductivity. Therefore, it can be highly expected that decreasing the inter-tube resistance and lowering the number of these high-resistance junctions could increase the conductivity of the network. Actually, Lee and co-workers (Geng et al. 2007) reported that contact junctions can be improved by treating SWNT networks with a 12 M HNO3 which helps to remove the insulating surfactant in SWNT network, and indeed, it improves the conductivity of the SWNT network by 2.5 times. They have also observed the dramatic decrease in the percolation threshold to the greatly reduced contact resistance between the tubes in the SWNT network.

Electrical Conductivity CNTs Base Nonconducting Polymer Composites

As CNTs exhibit high aspect ratio and high electrical conductivity, they are excellent candidates for fabrication of electrically conducting nanocomposites. While the electrical conductivity of individual carbon nanotubes has been measured to be in the order of 106 S/m, (Baughman et al. 2002) the maximum electrical conductivity of SWCNT films has been reported to be in the range of 104–105 S/m (Ericson et al. 2004; Sreekumar et al. 2002) due to the contact resistance between the individual carbon nanotubes in the films. Therefore, the range of electrical conductivity of CNT/polymer composites is reported to be tremendously wide. On the other hand, this wide range advises that it is possible to control the electrical conductivity of CNT/polymer composites by varying the amount and degree of dispersion of CNTs in the composites. The CNT/polymer composites can be used for a variety of applications including electrostatic dissipation (<10-4 S/m), electrostatic painting (10-4 ~101 S/m), electromagnetic interference (EMI) shielding (>101 S/m), printable circuit wiring, and transparent conductive coatings.
Again, the electrical conductivity of CNT/polymer composites is widely defined by the percolation theory. The percolation theory predicts that there is a critical volume fraction at which nanocomposites containing conducting fillers in insulating polymers become electrically conductive. According to this theory, σc = A (V -  Vc)β, where σc is the conductivity of a composite, V is the CNT volume fraction, Vc is the CNT volume fraction at the percolation threshold, and A and β are constant. So far, there are several publications documented on the progress of electrical conductivity of different CNT/polymer composites (Shaffer and Windle 1999; Sandler et al. 1999, 2003). The percolation threshold has been reported to range from 0.0025 vol% (Sandler et al. 2003) to several vol%. Therefore, it is difficult to draw definite conclusions about the mechanism of electrical conductivity of CNT/polymer composites from the literature. This is because the reported levels of CNT loading to achieve a percolation threshold vary widely. The electrical conductivity and percolation threshold of different CNT/epoxy composite systems are shown in Table 3. It seems that different systems give a wide range of percolation values. However, even for the same system, for example, SWCNT/epoxy composites (Vc = 0.0025 ~  0.1 %), (Sandler et al. 2003) a wide variation in percolation value was observed.

The mechanism for percolation threshold for electrical conductivity of CNT/polymer composites is determined by numerous factors, and a number of publications have reported the factors affecting the percolation mechanism of CNT/polymer composites. The common factors affecting the percolation threshold of electrical conductivity are dispersion, (Sandler et al. 2003; Li et al. 2007b) alignment, (Choi et al. 2003; Du et al. 2003) aspect ratio, (Li et al. 2007b; Bai and Allaoui 2003; Bryning et al. 2005) degree of surface modification (Georgakilas et al. 2002) of CNTs, types and molecular weights of the matrix polymer, (Pan et al. 2010; Ramasubramaniam et al. 2003) and composite processing methods (Liu et al. 2008). The aligned CNTs in epoxy decrease the percolation threshold by one order of magnitude compared to entangled nanotubes (Sandler et al. 2003). The electrical conductivity of SWCNT/epoxy composites with SWCNTs aligned under a 25-T magnetic field was increased by 35 % compared to similar nanocomposites without magnetically aligned SWCNTs (Choi et al. 2003).
In contrast, Du et al. (2003) found that the electrical conductivity of CNT/PMMA composite with 2.0 vol% CNTs was 10-10 S/cm with the aligned CNTs in the matrix and 10-4 S/cm with unaligned CNTs. This indicates that the alignment of the CNTs in the composite decreased the electrical conductivity. The reason is that there are fewer contact points between the CNTs when they are highly aligned in the composites, so CNT-aligned composites require more nanotubes to reach the percolation threshold. The aspect ratio of CNTs has a tremendous influence on the percolation threshold of CNT/polymer composites without changing other important parameters, such as the polymer matrix or the dispersion and aggregation state of the CNTs. On the other hand, it is well known that chemical functionalization may disrupt the extended conjugation of nanotubes and hence reduce the electrical conductivity of functionalized CNTs. For example, silanefunctionalized CNT/epoxy composites showed a lower electrical conductivity than that of the untreated CNT composites at the same nanotube content (Ma et al. 2007).
Cho et al. (2005) reported that the electrical conductivity of the acid-treated MWCNT composites was lower than that of the untreated MWCNT composites at the same content of MWCNTs. This is attributed to the increased defects in the lattice structure of carbon–carbon bonds on the nanotube surface as a result of the acid treatment. In particular, the severe modification of carbon nanotubes may significantly lower their electrical conductivity. However, there are several publications reporting that the functionalization of CNTs can improve the electrical conductivity of the nanocomposites (Tamburri et al. 2005). Tamburri et al. (2005) found that the functionalization of SWCNTs with –COOH and –OH groups enhanced the nanocomposites’ electrical conductivity as compared to the use of untreated SWCNTs.

Electrical Conductivity CNTs Base Conducting Polymer Composites

Most of the reported conducting polymer/carbon nanotube composites show conductivity enhancement over polymeric materials but much lower electronic performance compared to CNT films alone (Bekyarova et al. 2005). Few years back, Sun et al. (Wang et al. 2008) demonstrated that bulk-separated metallic SWNTs show superior performance than the as-produced nanotube sample in conductive polymer composites which can be obtained by blending with poly (3-hexylthiophene) and also poly(3,4-ethylene dioxythiophene):poly(styrene sulfonate). They did not show the performance of films prepared from SWNT alone as a control experiment. It was evidenced by Blanchet et al. (2004) that the percolation threshold of a SWNT network was drastically downshifted by replacing the insulating dispersing reagents in the network with a conducting polymer. But, the conductivity of the SWNT network was not increased by the replacement after percolation. However, they have fabricated a water-soluble and highly conductive self-doped polyaniline/SWNT composite (Fig. 4; Ma et al. 2006a) by the in situ polymerization of a thin skin of PABA (poly 3-aminophenylboronic acid) along and around ss-DNA-SWNTs.

They have also measured the thickness of the polymer layer on the carbon nanotube using transmission electron microscopy (TEM) and it was found to be around 1–3 nm (Fig. 5c; Ma et al. 2006b). The thin conducting polymer layer remarkably improves the contacts between the tubes and hence acts as a “conductive glue” which effectively assembles the SWNTs into a conductive network (Fig. 5b).
SWNTs with a PABA layer (referred to composite–SWNT networks) can be fabricated by vacuum filtration and dip coating (same methods used for SWNT alone) (Hu et al. 2004). Also a post mixture can be produced by simple mixing the pre-formed PABA with the same amount of ss-DNA-SWNTs. However, this postmixing process is not as effective as the premixing one in terms of interlinking the tubes (Fig. 5d). Furthermore, it has also been observed that the morphology of the post-mixture composite (Fig. 5d) is akin to the ss-DNA-SWNT network alone (Fig. 5a), except some large particles or aggregates, which could be due to the presence of neat PABA which has not being uniformly mixed with the SWNTs. Surprisingly, they have found that the percolation threshold of the SWNT networks increased by threefold and the conductivity of the post-mixture network significantly decreased (Fig. 6a).

Finally, the conductance of the post-mixture composite is five orders of magnitude lower than the network formed from the in situ polymerized PABA composite and is three orders of magnitude lower than the network prepared from SWNT alone (Fig. 6a). They have also found from Fourier transform infrared (FTIR) spectroscopic study that the structure of the PABA layer in the composite prepared by in situ polymerization is very different from that of the neat PABA (Fig. 7a) and also drastically different from the PABA in the composite formed by post-mixing with the pre-formed neat PABA (Fig. 7b, purple curve). The FTIR peak of PABA in the composite formed by in situ polymerization at 1,120 cm-1 has been assessed by MacDiarmid et al. (1994, Yan et al. 2007) as the “electronic-like band” and is considered to be a measure of the degree of delocalization of electrons.

Therefore, it can be considered as a characteristic peak of polyaniline conductivity. These results are the strong indication towards the conclusion that the PABA has much higher conductivity and existed in the more stable and conductive emeraldine state (Zengin et al. 2002; Sainz et al. 2005), compared to the neat PABA and the post-mixture PABA which were in the nonconductive pernigraniline state (Ma et al. 2006a; Wang et al. 2005). They have also prepared another PABA composite by in situ polymerization in the presence of pre-oxidized ss-DNA/SWNTs (“seed” method) (Zhang et al. 2004; Zhang and Manohar 2004). The intensity of the “electronic-like” FTIR peak is slightly lower than that of the PABA in the in situ polymerized composite with the intact ss-DNA-SWNTs, but much higher than the neat PABA and the post-mixture PABA composite (Fig. 7b). Though the percolation threshold of the composite formed by the seed approach is threefold higher than the in situ composite but similar to the post-mixture composite, the conductance after the threshold is four and six orders of magnitude lower than the SWNT network alone and the in situ composite with intact SWNTs, respectively (Fig. 6a). The morphology of the seed composite film has been studied using AFM and noticed that the PABA in the composite did not interlink the nanotubes into a conductive network. Instead, serious aggregation of the nanotubes into large particles (Fig. 5e, f) has been induced by PABA. The aggregation mechanism is still not well understood yet and currently under investigation. The author thinks that it might be related to defects along the tubes (formed by the pre-oxidation process) which remarkably weaken the mechanical strength of the carbon nanotube. Thus, it can be concluded that not only the polymer’s molecular structure but also the arrangement or distribution of the carbon nanotubes in the composites dictates the overall percolation behavior and macroscopic electronic property of the composites. It is also important to mention that the fabrication process significantly impacts the electronic and molecular structure of the PABA formed in the composites as well as the arrangement or lateral distribution of the carbon nanotubes in the composites. To effectively optimize the fabrication parameters and ensure the formation of SWNT networks in a controllable fashion for a variety of potential applications, it is crucial to understand these reaction characteristics. Furthermore, it is worth mentioning that ss-DNA-SWNTs played multiple roles during in situ fabrication of conducting polymer nanocomposites. First, it functioned as catalytic molecular templates during in situ polymerization of ABA as the polymerization process might be 4,500 times faster (Ma et al. 2008a). Additionally, the quality of the resulting PABA was also drastically improved, observed by the fact that the backbone of the self-doped polyaniline had longer conjugated length as fewer short oligomers were produced and they existed in the more stable and conductive emeraldine state, which in turn can be exploited to produce conducting polymer composite materials with a much more enhanced electronic performance. Secondly, the ss-DNA-SWNTs also worked as unique conductive polyanionic doping agents in the resulting polyaniline film with enhanced conductivity and redox activity both in low pH and neutral pH solutions. In addition, it also acted as active stabilizers after the polymerization. The final advantage will be that the large surface area of the carbon nanotubes greatly enhanced the density of the functional groups available for sensitive detection of the target analytes.

A wide variety of conducting polymers such as polyaniline (PANI), poly(diphenylamine) (PDPA), polypyrrole, polythiophene, etc., are currently used in different applications including metallic interconnects in circuits, electromagnetic radiation shielding, and chemical sensors (Stutzmann et al. 2003). The conductivity of such polymers arises due to the existence of charge carriers and mobility of those charge carriers along the bonds of the polymer chains. These polymers also show chemical selectivity, which makes them act as ideal candidate for the immobilization of gas molecules, and exhibit highly reversible redox behavior with a distinguishable chemical memory. So, these conducting polymers can potentially act as a gas sensor. Let us define gas sensor first. It is a device which detects the presence of different gases in an area, especially those gases which might be harmful to human beings or animals. The fabrication of ammonia gas sensors via a scanned-tip electrospinning method (Fig. 8) using a single 10-camphorsulfonic acid (HCSA) doped PANI/poly(ethylene oxide) (PEO) nanofiber with a diameter of 100–500 nm on gold electrodes has been reported by Craighead et al. (Liu et al. 2004) in 2004. The characterization of the well-defined single fiber material and the sensor response has been thoroughly studied. They have demonstrated that the sensor showed a rapid and reversible resistance change upon exposure to NH3 gas at concentrations as low as 500 ppb via the protonation and deprotonation of PANI.  The performance of nanofiber sensors, for example, response time, can be estimated by considering the diffusion of ammonia into the fiber and the reaction of ammonia with doped PANI. Another aspect was the correlation between response times with fiber diameter. Indeed, the response times with various diameters refer to radiusdependent differences in the diffusion time of ammonia gas into the fibers. Manesh et al. have prepared another type of ammonia gas sensor with a detection limit of 1 ppm (Manesh et al. 2007) using electrospun PDPA/poly(methyl methacrylate) (PMMA) nanofibers as sensing materials. They have demonstrated that the changes in resistance of the nonwoven membrane showed linearity with the concentration of ammonia in the range of 10–300 ppm. Additionally, the detection target was expandable and can be expanded from ammonia to other amines according to Gong et al. (2008) (Fig. 9) using PANI nanotubes which can be easily made using electrospun PVA fiber mat membrane as the template. The small diameter, high surface areas and porous nature of the PANI nanotubes gave considerably better performance with regard to both time response and sensitivity. They have also observed that the responses follow the orders: (C2H5)3 N > NH3 > N2H4. However, the PANI nanotubes showed higher sensitivity and quicker response to (C2H5)3 N compared with PANI prepared without a template. In addition, a reasonable reproducibility has been observed in case of the reversible circulation response change of PANI nanotubes. However, Li et al. (Ji et al. 2008) described coaxial PANI/PMMA composite nanofibers using the electrospinning technique and an in situ polymerization method. The responses of the gas sensors based on these PANI/PMMA composite nanofibers towards triethylamine (TEA) vapor were investigated at RT, and it was found out that the sensors showed a sensing magnitude as high as 77 towards TEA vapor of 500 ppm. Furthermore, the responses were linear, reproducible, and reversible towards TEA vapor concentrations ranging from 20 to 500 ppm. Additionally, it was revealed that the concentration of doping acids only brought changes in resistance of the sensor without affecting its sensing characteristics. For example, the gas sensor with a doping acid (toluene sulfonic acid) exhibited the highest sensing magnitude, which can be explained by understanding its sensing mechanism and the interactions of TEA vapor with doping acids.

There are reports of use of polymer as brushes as a composite, intended to make many applications; a significant advantage of polymer brushes compared to other surface modification methods is their mechanical and chemical stability, accompanied by a high level of synthetic flexibility towards the introduction of functional groups. This is in contrast to the physisorbed, non-bound polymer films where chemical modification by using wet chemistry is difficult to conduct. Additionally, it is now possible to grow brushes on virtually every surface (flat surfaces, particles, or macromolecules), to any thickness, of every composition, incorporating a multitude of functional groups and containing series of blocks. More recent applications of polymer brushes include nano-patterned surfaces (Shah et al. 2000), photochemical devices (Whiting et al. 2006), new adhesive materials (Raphael and De Gennes 1992), protein-resistant biosurfaces, (Saha et al. 2012) chromatographic devices, (van Zanten 1994) lubricants, (Joanny 1992) polymer surfactants, (Milner 1991) polymer compatibilizers, eight and many more.
One of the most attractive applications of surface-initiated polymerizations is the formation of nano-patterned surfaces by soft lithography techniques that combine microcontact printing (μCP) and graft polymerization. An elegant example is that of Hawker et al. who combined photolithography with nitroxide-mediated “living” free radical polymerization to obtain patterned polymer brushes with well-defined hydrophobic and hydrophilic domains (Fig. 10). They extended this concept to synthesize patterned polymer layers by aqueous ATRP (Vidal et al. 1980).

Conducting Polymer Composites Made from Polymer Brushes

Recently, Huck and coworkers (Kong et al. 2007) have shown that charge-transporting polymer brushes (polytriphenyl amine acrylate) can be used for a variety of organic electronic device fabrications using composite methodology. These polymer brush films contain a greater level of ordering at the molecular level and display higher charge mobility than spin-coated films of the same polymer, which was attributed to the controlled polymer brush architecture and morphology. As, for example, when CdSe nanocrystals (with diameter in the range of 2.5–2.8 nm) subjected into the polymer brush layers form a polymer composite (Fig. 11), its photovoltaic quantum efficiencies of up to 50 % (Snaith et al. 2005). In another report, Advincula and coworkers (Fulghum et al. 2008) successfully grafted holetransporting PVK (poly(vinyl carbazole)) brushes on transparent ITO electrodes. Using cyclic voltammetry, the PVK brush was electrochemically cross-linked to form a conjugated polymer network film. Covalent linkage of PVK led to a direct electroluminescent PLED device, in which the electroluminescent polymer layer can be simply solution-cast onto the modified ITO.

A more ambitious challenge in surface science is the design of smart surfaces with dynamically controllable properties (Lahann et al. 2003). Such surfaces have characteristics that can be changed or tuned in an accurate and predictable manner by using an external stimulus. Recently, Huck and coworkers have shown that wetting properties of surfaces modified with cationic polyelectrolyte brushes strongly depend on the nature of the counter ion (Fig. 12). Coordination of polyelectrolyte brushes bearing quaternary ammonium groups (QA+) with sulfate anions resulted in highly hydrophilic surfaces, (Moya et al. 2005) whereas coordination of similar brushes with ClO4¯ rendered the surface hydrophobic (Azzaroni et al. 2005).

Recent research has focused on the Cu(I)-catalyzed, highly specific, and efficient formation of 1,2,3-triazoles via the 1,3-dipolar cycloaddition of azides and terminal alkynes (“click” chemistry) (Feldman et al. 2004). This methodology has been used to modify surfaces of solid metals and cells, because the reaction provides high yields and stereospecificity and proceeds under mild conditions (Tornøe et al. 2002; Lewis et al. 2002), Click chemistry also has been used for functionalizing polymers in solution (Sumerlin et al. 2005; Gao et al. 2005). Research in nanobiotechnology and biomedical sciences often involves the manipulation of interfaces between man-made surfaces and biomolecules (and cells), which generally requires the construction of surfaces that present chemically active functional groups from non-biofouling supporting materials. Choi and coworkers (Lee et al. 2007) used “click” chemistry to couple azide groups at the terminal of the non-biofouling polymeric film of poly(oligoethylene glycol methacrylate) with incoming molecules of interest containing terminal acetylenes (Fig. 13). As a model for bioconjugation, biotin was immobilized onto the poly(oligoethylene glycol methacrylate) film via click chemistry, and biospecific recognition of streptavidin was demonstrated.