INTRODUCTION
Nanotechnology is an emerging field that studies fundamental nanoscale processes and the exploitation of those processes in the development and function of nanode-vices.[1] To date, development has been hampered by the need for materials and processes that can perform reproducibly at the nanometer scale. Although at present nanodevice development is dominated by materials made from SiO2 or metals, these materials can suffer from a lack of processability.[2] As an alternative that has met with some success, organic-based materials can be easily processed by a variety of methods including spin coating, evaporation, and printing.[3] Organic and organic/inorganic hybrid materials have been developed for organic light-emitting diodes (OLEDs), field-effect transistors, and other devices.1-4-6-1 However, organic devices usually suffer from a short lifetime as a consequence of poor mechanical and thermal stability associated with small organic molecules.[5] Although organic/inorganic hybrid materials have better mechanical and thermal stability,1-7-9-1 there still exists an inherent lack of component compatibility, leading to difficulties in processing, and intercomponent communication.
A promising class of materials for nanodevice application is conducting polymers (Fig. 1). Their thermal and environmental stability facilitates use in devices for photochemical and electrochemical applications,1-10-12-1 and their physical and chemical properties can be easily tailored for specific functions.[13,14] When conducting polymers were first introduced, their poor solubility in common processing solvents limited their versatility.1-11-1 However, this issue has been addressed through modifying the backbone of the polymer with side chains, such as alkyl and alkoxy groups. With this approach, processing of these materials in either aqueous or organic media has expanded the utility of conducting polymers in OLEDs, electrochromic displays, and coatings for more sensitive materials.1-6,15-17-1 Numerous articles are found in the literature reporting devices in the micrometer regime made with conducting polymers.[ , , ]
Although Moore’s law predicts a doubling in the number of transistors per chip approximately every 18 months,[22- the limits of traditional patterning techniques are threatening to hinder this advancement. The feature size required to continue this trend is in the <100 nm regime, where control over feature size is critical and difficult to achieve. New procedures and variations on current methods need to be developed to break this regime barrier. In this article, we will review current approaches to creating conducting polymer nanopatterns, from ''template synthesis'' to lithography. We discuss the advantages and disadvantages of each method and highlight the unique nanostructures formed with these techniques. These materials are likely to enhance the emerging field of nanoelectronics, while complementing current technology.
TEMPLATE SYNTHESIS
The first reported patterning of conducting polymer nanostructures used template synthesis, a method developed by Penner and Martin[23] in the mid-1980s. They reported the electrochemical polymerization of pyrrole in the pores of a polycarbonate membrane. Template synthesis, as the name implies, utilizes the microporous or nanoporous structure of materials as a template for the electrochemical or oxidative synthesis of conducting polymer nanotubules and nanofibrils.[24- This technique can also be used to form nanotubes and nanofibrils of metals, carbon, and semiconductors.1-24,25-1
Template Materials
The most common template materials used are ''track-etch'' membranes and alumina membranes.1-26-1 Track-etch membranes are formed by bombarding a nonporous membrane with nuclear fission fragments.1-27-1 These membranes are typically polycarbonate or polyester. The tracks that are formed are subsequently chemically etched to produce pores in the material. The size of the uniform pores can be as small as 10 nm in diameter with pore densities as high as 109 pores/cm2.1-26-1 The bombardment occurs randomly, which, at small pore diameters and high pore densities, can cause intersections of the pores.[26- If these intersections are not desired, alumina or other membranes can be used because the processes used to form these membranes are more controlled.
Fig. 1 Examples of conducting polymers.
Porous alumina membranes are formed from the anodization of Al metal in acidic solution. The uniform pores are cylindrical, with pore densities as high as 1011 pores/cm2 and a size range from >100 down to <5 nm.[26] Because of the nature of pore formation in the alumina membranes, interconnects do not develop between pores,[25] affording a template material that can form isolated nanostructures. Membranes, including SiO2, zeolites, and other nanoporous solids, have also been used as template materials, although to a lesser extent.[25]
Mechanism for Template Synthesis
The polymerization mechanism for conducting polymers using template synthesis is either electrochemical or oxi-dative. The most common electrochemical method is to plate one surface of the membrane with a conducting metal thin film and use this film as the anode in an electroplating process.[28] The polymer first nucleates on the walls of the pore, and subsequent deposition yields a well-ordered polymer outer layer that becomes more disordered as the tubule wall thickens.[29] In the oxidative method, the template is immersed in a solution containing monomer and a polymerization reagent, such as an aqueous iron(III) salt.[30'31] The deposition occurs in a manner similar to the electrochemical method, resulting in a similar polymer-order gradient for the nanostructures (Fig. 2).
Physical Properties of Conducting Polymer Nanotubules/Nanofibrils
Early studies of polypyrrole chain order in nanofi-brils (closed tubules) used polarized infrared absorption spectroscopy (PIRAS) to examine a range of fibril diameters from 30 to 600 nm. In addition, X-ray diffraction was used to observe 30 and 400 nm fibrils.[32] In the X -ray diffraction studies, 400-nm fibrils showed diffractograms similar to amorphous conducting polymer, while the 30-nm fibrils exhibited diffractograms similar to stretched polymer, suggesting an alignment of the polymer chains in these small-diameter fibrils.[32]
Fig. 2 Schematic of template synthesis.
The PIRAS experiment supported the X-ray diffraction evidence that polymer chain alignment in a fibril increases as the diameter decreases. In a typical PIRAS experiment, the absorbance is measured for light polarized parallel and perpendicular to the major axis of a material (Fig. 3). In the case of a polymer fibril, if the polymer chain alignment is completely random, then the absorbance of light parallel to the pore axis will be the same as the absorbance of light perpendicular to the pore axis and will give a dichroic ratio (ratio of absorbances parallel and perpendicular to the pore axis) of 1. Smaller fibrils had dichroic ratios significantly different from 1, with the fibrils preferentially absorbing one polarization over another.[32] Amorphous polymer and large-diameter fibrils do not absorb polarized light preferentially, giving dichroic values very close to 1.[32,33]
Polarized infrared absorption spectroscopy experiments were also performed on polycarbonate membranes; the polycarbonate was also found to be stretch-oriented.[32] Studies have shown that conducting polymers can be induced to align with the orientation of the substrate.1-34-36-1 This orientation is lost as the thickness of the deposited polymer film grows. Similarly, the outer walls of conducting polymer tubules and fibrils deposited in stretch-oriented polycarbonate membranes align with the axis of the membrane, but this alignment is slowly lost as polymerization continues and the walls thicken. Wall thickness is controlled by the polymerization time.[25-
The thickness of the polymer walls is also correlated with the conductivity of the tubule. While thick-walled nanotubules had conductivities similar to bulk material, the narrow-walled nanotubules had significantly higher values.1-35-1 One explanation for this phenomenon is that the alignment in the polymer chains increases the conjugation length of the polymer, enhancing the electronic properties of the material. As the walls thicken, the overall randomness of the structure increases, decreasing the conjugation length and hence the conductivity.[35- This method provides a means to control the conductivity of the material through polymerization time. It has been postulated that the walls in a polycarbonate membrane nucleate the ordered polymer in two ways: 1) as the polymerization commences, the polycationic oligomers forming are not as soluble as the monomer units and 2) there is a coulombic attraction between the anionic walls of the membrane and the forming cationic polymer.[26]
Recently, this ”greater ordering” claim has been disputed. Most of the above studies of polypyrrole nanotubules were performed with a polycarbonate membrane; Mativetsky and Datars[37] used an alumina membrane to form nanotubules of polypyrrole oxidatively and found the tubules to have a highly irregular morphology. Indeed, PIRAS studies with alumina membranes have shown no preferential polarized absorption (dichroic ratios near 1).[37] Because there are no charged sites in an alumina membrane, the nucleation occurs more randomly in the alumina pores. Polypyrrole then grows radially outward from each nucleation site until it encounters an obstacle: that of the pore wall or another ”polypyrrole mass,” disrupting order. Although Mativetsky and Datars[37- were not able to measure conductivity directly, their measurements suggest that the conductivity of these random fibrils is comparable to the fibrils of similar size formed from a polycarbonate membrane. They therefore concluded that the enhanced conductivity of these nanotubules is not a result of increased order imposed by the polycarbonate walls.
Fig. 3 Schematic of a typical PIRAS experiment.
Applications of Template Synthesis
Conducting polymer nanofibrils and nanotubules have a wide variety of potential applications. They are highly permeable, and studies have shown that they can be highly selective for certain gas mixtures; the high-density conducting polymer fibrils could be used as a gas permeation membrane.[38- Conducting polymer nanotubules are also well suited for encapsulated sensors because the polymerization process can be stopped at any time, thus controlling the size of the cavity. Researchers have encapsulated a variety of enzymes and chemical species in thin-walled tubules and have found that their activity is greater than that of enzymes trapped in a thin polymer film.[30,39,40- These reactions could be carried out in aqueous or organic solvents because the conducting polymer ”microreactor” is isolated from the environmental surroundings by the walls of the conducting polymer, utilizing the relative insolubility of the polymer in common solvents.[30- The surface area for enzyme/substrate interaction can be adjusted by changing the porosity of the membrane, producing a microreactor with a high loading capability.1-30-1
With the enhanced conductive properties of these materials, numerous groups have reported using nanotubules to form a variety of novel systems. For example, Granstrom et al.[41- polymerized 3,4-ethylene-dioxythio-phene (EDOT) in polycarbonate membranes with pore sizes of 10 nm and 100 nm for the preparation of a hole-injection contact in a polymeric light-emitting diode (PLED). An electroluminescent layer was spin-coated on the surface of the membrane, followed by an Al/Ca layer for the electron-injecting contact. Although the PLED tested had a very low efficiency, with optimization, the authors believed that increased efficiencies up to 1-10% could be obtained.1-41-1 The advantage of this technique is in the sheer quantity of nano-LEDs that can be formed at one time.
Fig. 4 Scanning electron microscope image of polyaniline nanofibrils formed from a porous alumina membrane using template synthesis.
In other applications, polyaniline tubules have been studied for use as microscale/nanoscale transistors and for field emission because of their excellent switching properties and mechanical stability (Fig. 4).[42,43- Polymer composites have been made to blend physical properties, and many have been found to enhance desirable characteristics, such as conductivity and ”rectification effect.”1-31,44-46-1
