INTRODUCTION
Supercapacitors, also referred to as electrochemical capacitors or ultracapacitors, are devices used to store and deliver electrical energy in high-power pulses.[1,2] With the advent of electric vehicles, digital communication, and other electronic devices that require significant bursts of electrical energy, the need for supercapacitors has expanded rapidly. At present, the most promising materials on which supercapacitors are based can be divided into two categories—those that make use of a double-layer charge storage mechanism (e.g., carbon nanotubes, carbon aerogels, and activated carbon black) and those employing a redox pseudo-capacitive charge storage mechanism (e.g., conducting polymers and transition metal oxi-des).[3,4] Already, the electrical charge that can be stored in each of these materials is typically several orders of magnitude larger than that of most commercially available conventional capacitors. However, it has been shown in recent times that even greater charge storage capacitances can be achieved in composites made by combining carbon nanotubes (a double-layer capacitive material) with a conducting polymer (a redox pseudo-capacitive material). The superior charge storage performance of carbon nanotube-conducting polymer composite supercapacitors arises from their ability to merge the properties that separately make carbon nanotubes and conducting polymers so suited to their respective charge storage mechanisms. That is to say, the composites are able to combine the high surface area and electrical conductivity of carbon nano-tubes with the redox electrochemistry of conducting polymers.[5]
CARBON NANOTUBE-CONDUCTING POLYMER COMPOSITES
Structure and Fabrication
Most carbon nanotube-conducting polymer composites used in supercapacitor devices display a nanoporous network structure in which each nanotube is coated by a thin layer of conducting polymer. This structure maximizes the exposed surface area of the conducting polymer while retaining the conductive network of the carbon nanotubes.
The nanoporous structures possible for carbon nano-tube-conducting polymer composites are generally not observed when making conducting polymer composites with other high surface area forms of carbon such as activated carbon black and carbon aerogels.[6,7] In these cases the conducting polymer generally blocks the electrolyte pores. The other carbon structures are also unable to achieve the exceptionally low percolation thresholds possible with carbon nanotubes.[8]
Electron microscopy, elemental analysis, X-ray diffraction, nuclear magnetic resonance, and Raman spec-troscopy studies indicate that there is good physical interaction between the carbon nanotubes and the conducting polymer (Fig. 1). In general, these techniques give no indication that there is a chemical reaction between the two components.[9-11] However, it has been reported that Raman spectroscopy on composites of arc-grown multi-walled carbon nanotubes and polyaniline indicates a site-selective interaction between the quinoid ring of the doped polymer and the carbon nanotubes.[12]
Recent reports have shown that composites of carbon nanotubes and conducting polymers such as polypyr-role, polyaniline, poly(3-methylthiophene), poly(3-hex-ylthiophene), poly(3-octylthiophene), and poly(p-phenyl-ene vinylene) can be prepared using a variety of chemical[9,11,12] and electrochemical[13-17] polymerization techniques or by simply mixing the polymer and nano-tubes.[18-20] To date, most carbon nanotube-conducting polymer composites have been made using multiwalled carbon nanotubes (MWNTs) as they generally offer uniform conductivity combined with a greater susceptibility to chemical oxidation and a lower cost relative to single-walled carbon nanotubes (SWNTs). The uniformity and thickness of the conducting polymer coating on each nanotube, in addition to the degree of porosity, are crucial in determining the composite’s properties and are largely dependent on the fabrication technique employed.
Routes for producing carbon nanotube-conducting polymer composites based on chemical polymerization generally require a solution containing the desired monomer (typically pyrrole or phenylacetylene) and carbon nanotubes in suspension.[9,11,21] Polymerization is then initiated by the addition of a dopant such as (NH4)2S2O8. While the conducting polymer is observed to coat each individual nanotube using this technique, there have been some reports that the deposit tends to be in-homogeneous with some aggregation of the deposited polymer.[22] This in situ method of polymerization offers excellent processing flexibility in that the coated nano-tubes can be deposited as a uniform composite film during polymerization or maintained as distinct coated nano-tubes in suspension.1-9-1 If the coated carbon nanotubes are maintained in suspension, they can subsequently be formed into a film or pellet using techniques such as spin-coating or vacuum drying.
Fig. 1 Scanning electron microscopy image of the fractured cross section of multiwalled carbon nanotubes coated with polypyrrole. The tubes cleaved during fracture give an indication of the close interaction between the nanotubes and conducting polymer.
Electrochemical composite growth techniques can be further divided into those that polymerize the conducting polymer onto a carbon nanotube preform[14-17,23] and those in which the nanotubes and conducting polymer are simultaneously deposited from the electrolyte[5,13] during polymerization. When using an aligned MWNT preform, the conducting polymer (usually polypyrrole[14,16] or polyaniline[24]) tends to form a continuous, uniform coating over each individual carbon nanotube (Fig. 2). In this way, the basic structure of the carbon nanotube preform is retained, making it possible to produce composites with a very high surface area. While conducting polymers can also be deposited electrochemically on unaligned carbon nanotube preforms, there is a tendency in this case for the deposited conducting polymer to block electrolyte channels into the preform, preventing uniform coating of each carbon nanotube.[17]
For electrochemical synthesis routes in which the carbon nanotubes and conducting polymer are simultaneously deposited, the electrolyte used contains suspended carbon nanotubes in addition to the desired monomer (such as pyrrole or 3-methylthiophene) and any supporting electrolyte required.[5,25] Polymerization is then initiated via the application of an applied potential, resulting in the deposition of a nanoporous composite film in which each MWNT is uniformly coated by the conducting polymer (Fig. 3). If the suspended carbon nanotubes used in this technique are functionalized via an appropriate acid treatment step before they are added to the polymerization electrolyte, it is possible for the nanotubes to dope the deposited conducting polymer.[27,28] This eliminates the need for a supporting electrolyte during film growth and establishes several other advantages for supercapacitor applications that will be discussed later.
Composites formed by simply mixing the carbon nanotubes and conducting polymer generally make use of a suspension containing carbon nanotubes and a pre-synthesized conducting polymer such as poly(p-phenylene vinylene), poly(3-hexylthiophene), or poly(3-octylthio-phene).[18,20,29] The suspension is ultrasonicated to mix the two constituents and is then spin-cast to form a carbon nanotube-conducting polymer composite film. The resulting composite structure is quite dense with conducting polymer enveloping the carbon nanotubes and the spaces between them.[19] It is worth noting that carbon nanotube-conducting polymer composites have also been produced by simply spin coating the conducting polymer directly onto a MWNT preform.[30,31]
Fig. 2 Scanning electron microscopy image of the fractured cross section of an aligned array of MWNTs in which each MWNT has been coated with polypyrrole.
Fig. 3 Scanning electron microscopy images of nanoporous composite films formed by the simultaneous electrochemical deposition of MWNTs with (a) polypyrrole and (b) poly-(3-methylthiophene).
Properties and Applications
The responsiveness of conducting polymer films to their external environment makes them very attractive materials for a variety of applications including photo-voltaics,[32] sensors,[33'34] light emitting devices,[35'36] actuators,[37'38] and supercapacitors.[4] However, the response times of these devices are generally restricted by limited electrical conductivity and rates of ionic diffusion in the conducting polymer. The general improvement in electrical conductivity (up to 10 orders of magnitude) observed when carbon nanotubes are introduced into conducting polymers offers some alleviation of this problem.[18'19'29'39] Perhaps even more important for electrochemical applications is the increase in the amount of conducting polymer surface exposed to the electrolyte and the reduced diffusion distances in nanoporous carbon nanotube-conducting polymer composites.[13]
Carbon nanotube additions have been observed to increase the strength,[40] thermal conductivity,[41] stability in air,[19] and photodegradation resistance[11] of polymers. These properties combine well with the excellent photovoltaic efficiency[30] and strong photosensitivity[20] exhibited by composites of carbon nanotubes with conducting polymers such as poly(3-hexylthiophene), poly-(3-octylthiophene), and poly(p-phenylene vinylene), making them suitable for photovoltaic[31] and light-sensing[42] applications. For light-emitting applications, increasing the electrical conductivity of conducting polymers via the addition of untreated nanotubes is particularly advantageous as this does not introduce states within the polymer bandgap that quench luminescence, as is generally the case when conductivity is increased by doping the polymer.[19,43]
Electrochemical devices, such as actuators and super-capacitors, made from carbon nanotube-conducting polymer composites have been shown to display the electrochemical redox behavior of conducting polymers while retaining the conductive network of the carbon nanotubes.1-13’25’44-1 This combination of desirable properties is associated with an increase in the storage capacity of the composite supercapacitors relative to the component materials and will be discussed in detail in the following section.[5] Unfortunately, reports to date on the actuation of carbon nanotube-conducting polymer films appear to indicate that strain is limited by the nanotubes,[44] which are much stiffer than the conducting polymer.
SUPERCAPACITORS
Capacitors are devices used to accumulate and deliver electrical charge. Capacitors can be divided into conventional capacitors, generally based on a ceramic[45] or polymer[46] film dielectric, and supercapacitors, which make use of an electrolyte. The most promising super-capacitors at present make use of materials such as high surface area carbons,[47] conducting polymers,[4] and transition metal oxides[48] that are combined with an aqueous or organic electrolyte.
Traditional ceramic and polymer capacitors consist of two parallel conductive plates separated by a ceramic or polymeric dielectric, respectively. When a voltage is applied across the plates one of them adopts a positive charge and the other a negative charge. Conventional ceramic and polymer capacitors are frequently used in lighting, televisions, stereos, induction heaters, computers, and other electronic devices.[49] Like their more conventional counterparts, supercapacitors consist of two opposing electrodes; however, the electrodes are separated by an electrolyte instead of a polymer or ceramic dielectric. While supercapacitors are limited to operating at a few volts, their capacitance far exceeds that of conventional capacitors resulting in significantly higher energy densities.[50]
Table 1 Relative performance of modern supercapacitor materials
|
Maximum |
Maximum |
|
|
|
|
|
|
capacitance |
capacitance |
Power |
Energy |
Max voltage |
|
Material |
(Fg-x) |
(mF cm-2) |
(kW kg- x) |
(W hr kg- x) |
(V) |
|
High area carbons’55-60-1 |
180 |
173 |
20 |
7 |
3 |
|
Conducting polymers’17‘61-68- |
268 |
600 |
11 |
42 |
3 |
|
Ruthenium oxides’48‘69-73- |
840 |
500 |
0.5 |
27 |
1 |
The large capacitance exhibited by supercapacitor systems is related to two main charge storage mechanisms, either double-layer capacitance or redox pseudo-capacitance. Double-layer capacitance involves storing charge in the electrical double-layer formed at the electrode-electrolyte interface. Ions within the electrolyte accumulate at the electrode surfaces, compensating for the electronic charge injected into the electrodes. Depending on the electrolyte used, an electrical double-layer can typically store about 10-40 p.F cm"2 of interface; hence the surface area of the electrode material is critical.[51-53] Supercapacitors based on high surface area carbon electrodes, such as carbon nanotube mats, are of this type.
Redox pseudo-capacitance, on the other hand, involves the storage of charge via the reversible oxidation and reduction of an electrode material. Redox switching is accompanied by the transfer of charge-balancing coun-terions between the electrolyte and the solid electrode. In this case, charge is stored within the bulk of each redox-active electrode, not just at the interface with the electrolyte.[54] Conducting polymer and transition metal oxide-based supercapacitors employ this charge storage mechanism. The comparative performance of super-capacitors based on high surface area carbons, conducting polymers, and transition metal oxides is shown in Table 1.
Carbon Nanotube Supercapacitors
High surface area carbon electrochemical double-layer capacitors generally exhibit an excellent response time, cycle life, specific power, and stability. Because the capacitance of these devices is related to the amount of available electrode surface area, there has been extensive research aimed at increasing the surface area of carbon electrodes for double-layer capacitors. From this work, several techniques have been reported on the use of additives such as metals and metal oxides,[50] the pyrolysis of carbon-based polymers,[58] and the production of carbon foam, paste, and nanotube electrodes.[74,75] The molecular dimensions and high aspect ratio of carbon nanotubes in particular enable them to form exceptionally high surface area structures.[76]
It has been widely observed that carbon micropores less than 2 nm in size have detrimental effects on the kinetics and amount of charge stored by carbon-based capacitors.[77] This limitation arises because the pore dimensions approach that of the double-layer thereby inhibiting double-layer formation.[74] In addition to their extremely high surface areas, carbon nanotube electrodes typically exhibit an open structure with pore sizes in excess of 2 nm.[77] This is particularly advantageous for the electrolyte accessibility of carbon nanotubes relative to other high surface area carbon electrode materials, such as activated carbon black, which generally possess a significant portion of pores smaller than 2 nm. As a result, carbon nanotube supercapacitors are frequently capable of a greater capacitance, specific energy, specific power, and response time relative to other carbon structures used in double-layer supercapacitors.[56]
Conducting Polymer Supercapacitors
Conducting polymer supercapacitors utilize their reversible redox states to store and deliver charge. While other redox pseudo-capacitors, such as those based on transition metal oxides such as ruthenium oxide, may have higher specific capacitances by mass, conducting polymer supercapacitors generally have comparatively superior rates of response, specific power capabilities, and operating voltages.[ , ] Conducting polymers are also more amenable to use in complex shapes and lower in cost relative to ruthenium oxide-based supercapacitors.[83]
The majority of the promising conducting polymer supercapacitors currently being investigated are based on polypyrrole,[84] polyaniline,[85] poly(3-phenylthio-phene),[86-88] and poly(3-methylthiophene).[61] While these conducting polymers are generally used in liquid electrolytes, it is possible to employ polymer and gel electrolytes to create solid-state capacitors. It should be noted, however, that the response time and specific capacitance of solid-state supercapacitors are generally inferior to those obtained when using a liquid electrolyte.[89]
CARBON NANOTUBE-CONDUCTING POLYMER COMPOSITE SUPERCAPACITORS
Fundamental Findings
While the capacitance and specific energy of super-capacitors far exceed that of conventional capacitors, the charge storage mechanisms of supercapacitors generally result in a rate of charge and discharge that is several orders of magnitude slower than many commercially available conventional capacitors. For example, it can take redox supercapacitors tens of seconds to accumulate and deliver charge, whereas many commercial ceramic and polymer dielectric capacitors can charge and discharge in a matter of nanoseconds. To improve super-capacitor kinetics and specific power, it is essential that the combined resistivity of the matrix and electrolyte be minimized. In the case of redox supercapacitors, it is also important to select a redox reaction that has a large exchange current density.[1]
Double-layer capacitors have a natural advantage in terms of capacitance kinetics as their charge storage mechanism does not involve the slower ion intercalation and deintercalation processes observed in redox pseudo-capacitors. In fact, carbon-based double-layer capacitors have been produced that can accept and deliver charge in a period of microseconds.[57] In addition, modern double-layer capacitors generally exhibit cycle lives in excess of 105 cycles and good stability,[74] typically surpassing that of redox pseudo-capacitive systems. On the other hand, the three-dimensional adsorption process found in redox pseudo-capacitors enables much higher values of specific energy storage and greater voltage consistency during delivery than is found in processes that are limited to the electrode surface such as double-layer capacitance.[1] Supercapacitors based on composites of high surface area carbon nanotubes (a double-layer capacitive material) and a conducting polymer (a redox pseudo-capacitive material) can, to a large extent, bridge the gap between fast capacitor kinetics and substantial energy storage.
To date, investigations into the supercapacitive properties of carbon nanotube-conducting polymer composites have been based on samples made by chemically or elec-trochemically polymerizing either polypyrrole’[5'14'22'90] polyaniline,[17,91] or poly(3-methylthiophene)[26,92] in the presence of SWNTs orMWNTs. These first studies already indicate that carbon nanotube-conducting polymer composites are capable of specific capacitances per mass and geometric electrode area as high as 265 F g"1[90] and 2.6 F cm" 2,[14] respectively. These reported values are already significantly higher than that of the pure polymer or carbon nanotubes used in each case. It has also been shown that carbon nanotube-conducting polymer composite films are capable of charge and discharge rates that are approximately an order of magnitude faster than similarly prepared and tested samples of the relevant conducting polymer.[5]
As the available reports on carbon nanotube-conducting polymer composite supercapacitors were compiled by various researchers, using a range of sample conformations and an assortment of test conditions, it is difficult to compare the relative merits of each composite system. Indeed, even the process of measuring capacitance can be accomplished using a range of techniques including cyclic voltammetry, electrochemical impedance spectroscopy, and charge-discharge analysis—each of which has its own merits, which are too exhaustive to discuss here. However, the collective data available for carbon nano-tube-conducting polymer composite supercapacitors do indicate some fundamental deductions that can be applied generally across the range of composite systems studied.
The first of these deductions is the importance of using nanotubes that are able to form a connected three-dimensional network that promotes efficient charge transfer within the composite without inhibiting electrolyte access (i.e., avoids pore sizes <2 nm[77]). It is also advantageous to maximize the surface area of conducting polymer exposed to the electrolyte, thereby optimizing the redox-pseudocapacitive contribution of the conducting polymer. For this reason, it is desirable to obtain a thin, uniform, and continuous coating of conducing polymer over each of the carbon nanotubes. Composite growth techniques in which the carbon nanotubes and conducting polymer are deposited simultaneously have an advantage in this respect relative to techniques that make use of a carbon nanotube preform; polymer deposition onto carbon nanotube preforms, particularly those in which the nano-tubes are randomly oriented, can be limited by pore blocking and uneven distribution of the conducting polymer. It has also been reported that electrochemical growth techniques are able to produce a more homogeneous polymer coating with less aggregation than observed for composite synthesis routes that make use of chemical polymerization.[22]
Recent work has shown that if MWNTs are acid treated (resulting in the attachment of hydroxyl, carbonyl, and carboxylic functional groups to the tube surface) prior to electrochemical synthesis of MWNT-polypyrrole composite films, the MWNTs are able to partially dope the conducting polymer.[93] As mentioned earlier, this eliminates the need for a supporting electrolyte during film growth. However, the use of functionalized carbon nanotubes also benefits the supercapacitive behavior of the composites produced as will be discussed in the following section.
Carbon Nanotube Doping of the Conducting Polymer
Fig. 4 shows scanning electron microscopy images of MWNT-polypyrrole composite films made using different concentrations of functionalized MWNTs.[28] It can be seen that as the concentration of functionalized MWNTs in the polymerization electrolyte is increased from 0.025 to 0.4 wt.%, the polypyrrole coating thickness on each MWNT decreases, thereby increasing the loading fraction of MWNTs in the composite. For comparison, it is also possible to electrochemically grow MWNT-polypyrrole composite films using pristine (not functionalized) aligned arrays of MWNTs. In this case, the loading fraction of aligned MWNTs can be raised by simply using longer MWNTs with the same amount of deposited conducting polymer.[14] The length of the aligned MWNT array can be varied, while keeping the nanotube diameter and packing density approximately constant, via manipulation of the reaction parameters used to synthesize the MWNTs.[94]
Fig. 5 shows the specific capacitance vs. film-formation charge (a measure of the amount of polymer deposited) for MWNT-polypyrrole composite films made using functionalized MWNTs and aligned MWNT arrays that were not functionalized. In both cases, the capacitance of similarly prepared pure polypyrrole films (chloride ion doped) is shown for comparison. For the pure polypyrrole samples, it can be seen that as the amount of polymer deposited increased (as indicated by the film-formation charge), the proportion of total polymer that contributed to the measured capacitance decreased. This restriction arises from the limited electronic and ionic conductivity in the conducting polymer. Such limitations are alleviated in the composite films by the high surface area conductive MWNT network that gives rise to reduced diffusion distances, improved electrolyte access, and superior electronic conductivity. As a result, it is possible to deposit significantly more conducting polymer in the composite films without a marked deterioration in the proportion of polymer contributing to the measured capacitance.
It can be seen in Fig. 5a that as the loading fraction of functionalized MWNTs is increased, via an increase in the concentration of MWNTs in the polymerization electrolyte, the specific capacitance also increases for a given film-formation charge. On the contrary, as the loading fraction of aligned MWNTs (not functionalized) is increased by using longer MWNTs, the specific capacitance is not significantly effected for a given film-formation charge (Fig. 5b). Although there are factors to be considered such as the difference in diameter, length, con-ductivity, and distribution of the functionalized MWNTs relative to the pristine aligned MWNTs, it would appear that such primarily kinetic influences are insufficient to fully account for this difference in behavior. Instead, the increased level of doping provided by the function-alized MWNTs with increasing loading fraction is thought to be linked with the increase in capacitance for a given film-formation charge. This is an important finding because it indicates that the role of carbon nanotubes can be extended beyond that of a high surface area conductive substrate in carbon nanotube-conducting polymer films. By adopting the additional role of macromolecular dopant, carbon nanotubes can facilitate even greater performance benefits in carbon nanotube-conducting polymer composite supercapacitors.
Fig. 4 Scanning electron microscopy images of MWNT-polypyrrole composite films made using polymerization electrolytes containing the indicated concentrations of functionalized MWNTs.
Fig. 5 The specific capacitance vs. film-formation charge (a measure of the amount of polymer deposited) for MWNT-polypyrrole composite films made using (a) functionalized MWNTs of various concentrations and (b) pristine (not functionalized) aligned MWNT arrays of various lengths.
The doping of conducing polymers with functionalized MWNTs is still under investigation, although there does appear to be some similarities with doping using anions that are relatively immobile within the conducting polymer, such as dodecyl sulphate ions.[93] However, large anions such as dodecyl sulphate do not contribute to the formation of a high surface area conductive network as functionalized MWNTs can. Accordingly, dodecyl sulphate-doped polypyrrole does not offer the large capacitances of MWNT-polypyrrole composite films. For the samples shown in Fig. 4, it is possible to calculate the theoretical maximum thickness of polypyrrole that could be doped by the negatively charged surface groups along the MWNT surface (2.4 meg g" 1).[95] Such calculation yields a polypyrrole coating thickness of approximately 8 nm (assuming there are four pyrrole units per negative surface charge[96,97]). This result implies that only a fraction of the deposited polypyrrole is actually in the oxidized state during film-growth when functionalized MWNTs are the only source of dopant anions added to the polymerization electrolyte. Therefore the functionalized MWNTs allow the continuous electrochemical growth of polypyrrole, as does a more conventional anionic dopant, which would otherwise cease due to the insulating nature of unoxidized polypyrrole. However, the mechanism by which this occurs seems to be the formation of a conductive network comprising the MWNTs and adjacent layer of doped polypyrrole, rather than the incorporation of sufficient anionic dopant to ensure comprehensive oxidation of the polypyrrole.
Fig. 6 Scanning electron microscopy images of (a) a MWNT-polypyrrole composite film made using a polymerization electrolyte containing 0.4 wt.% functionalized MWNTs, 0.01 M sodium dodecyl sulphate, and 0.5 M pyrrole; and (b) a pure polypyrrole film made using a polymerization electrolyte containing 0.5 M KCl and 0.5 M pyrrole.
It is interesting to note the effect of competing dopant anions when growing carbon nanotube-conducting polymer composites using functionalized MWNTs. Adding additional sources of dopant anions to the polymerization electrolyte, such as dodecyl sulphate ions, makes it possible for polypyrrole to grow in areas further displaced from the functionalized MWNT surfaces. Consequently, the porosity of the composite structures produced is significantly reduced.[28] In the case of dodecyl sulphate ions, the reduction in film porosity was sufficient to produce a surface morphology quite similar to that observed for pure polypyrrole (Fig. 6). The reduced porosity of composite films grown with dodecyl sulphate additions is likely to be associated with a decrease in the loading fraction of functionalized MWNTs. These structural changes were accompanied by a sizable decline in capacitance and rate of response relative to composite films grown without competing dopant anions. This finding provides further indication of the importance of the three-dimensional network of nanopores and loading fraction of functiona-lized MWNTs in carbon nanotube-conducting polymer composites for supercapacitors.
Prototype Devices
The work described above has primarily aimed at characterizing the charge storage properties of carbon nano-tube-conducting polymer composites. Of the carbon nanotube-conducting polymer composite systems investigated to date, those based on polypyrrole have received the widest attention because of the charge storage capabilities, ease of polymerization, relative stability, and processing flexibility of this particular conducting polymer system. While polypyrrole is capable of existing in an oxidized (p-doped) and neutral state, displaying excellent redox properties between these two states, it is unable to redox cycle beyond the neutral state into a truly reduced form (n-doped). While this limitation does not prevent the characterization of charge storage in polypyrrole-based composites, it does inhibit the use of such composites in real supercapacitor devices where it is desirable for one of the two electrodes in the capacitor to be oxidized while the other is reduced during charging. In this case, polypyrrole-based films are ideally only suited to use in the oxidized electrode of supercapacitors.
Several other conducting polymers, such as poly(3-methylthiophene), are capable of being oxidized (p-doped) or reduced (n-doped) from the neutral state and have also been shown to possess excellent charge storage properties.[6,7,83,98,99] Carbon nanotube-conducting polymer composites based on such polymers can be used effectively in both electrodes of a supercapacitor. For this reason, current work aimed at developing prototype carbon nanotube-conducting polymer composite supercapacitors is looking closely at the choice of conducting polymer for each of the electrodes within the supercapacitor device. For example, it is possible to produce symmetric supercapacitors based entirely on, say, carbon nanotube-poly(3-methylthiophene) composites, or unsymmetric supercapacitors with, say, a carbon nano-tube-poly(3-methylthiophene) composite as the reduced electrode and a carbon nanotube-polypyrrole composite as the oxidized electrode. It is also possible to produce hybrid supercapacitors that combine a carbon nanotube-conducting polymer composite electrode with a simple carbon nanotube electrode. However, recent work indicates that such hybrid supercapacitors are outperformed by those based on a suitable combination of carbon nanotube-conducting polymer composites.[92]
It is recognized that the properties of conducting polymers can deteriorate significantly with prolonged potential cycling, which is of concern for many super-capacitor applications. The cycle life limitations of conducting polymers are often attributed to progressive assimilation of free space by the polymer chains.[96,100] This free space is required by the counterions that diffuse in and out of the polymer during potential cycling, hence its assimilation by the polymer chains limits the degree of doping possible for each successive cycle. Cycle life testing on prototype MWNT-polypyrrole composite films indicates that, for as many as 10,000 cycles, the capacitance dropped by less than 36%.[25,101] This is a substantial improvement relative to similarly prepared pure polypyrrole films in which capacitance dropped by 95% over 10,000 cycles. It is thought that the rigid nanotubes in carbon nanotube-conducting polymer composites inhibit the expansion and contraction of the composite, thereby limiting progressive compaction with potential cycling. In this way, the deleterious effect of potential cycling on the incorporation of counterions is reduced. The network of nanopores maintained in carbon nanotube-conducting polymer composites also helps to ensure effective electrolyte access over the full potential cycle.
CONCLUSION
Nanoporous carbon nanotube-conducting polymer composites are able to effectively combine the high surface area electrically conductive network of carbon nanotubes with the three-dimensional charge storage capabilities of redox-active conducting polymers. This combination of properties is particularly advantageous for super-capacitors, which need to be able to store and deliver relatively large amounts of electrical energy in high power pulses. Recent work has shown carbon nanotube-con-ducting polymer composites to exhibit specific capacitances per mass and geometric electrode area as high as 265 F g"1[90] and 2.6 F cm"2,[14] respectively. These values are already significantly higher than that of the pure polymer or carbon nanotubes used in each case and there is still much room for further optimization. Manipulation of the various composite synthesis techniques available—particularly those producing films in which the nanotubes are able to dope the conducting polymer—is expected to provide many opportunities to build on the exceptional findings discussed in this article.
