a review on the synthesis of carbon nanotube thin...

In: Carbon Nanotubes ISBN: 978-1-62081-914-2 Editor: Ajay Kumar Mishra © 2013 Nova Science Publishers, Inc.

Chapter 7

A REVIEW ON THE SYNTHESIS OF CARBON

NANOTUBE THIN FILMS

Rachel L. Muhlbauer and Rosario A. Gerhardt School of Materials Science and Engineering

Georgia Institute of Technology Atlanta, GA, US

ABSTRACT Thin film properties, including their electrical, optical, and mechanical properties, are

highly influenced by the way the film is fabricated; therefore, it is essential to understand what controls the formation of the thin films. Parameters that may affect the growth depend on the type of fabrication method used and the original size and purity of the carbon nanotubes and the types of precursors or substrates used. Carbon nanotube thin film fabrication can be divided into two broad categories: direct growth or transfer deposition. During direct synthesis techniques, the thin films are constructed directly on the substrate being used for film testing or the final device. Direct growth methods include, but are not limited to, chemical vapor deposition and a long list of wet deposition methods such as electrophoretic deposition, dropcasting, Langmuir-Blodgett film formation, and self-assembled monolayers to name a few. Each method results in films with very different characteristics even if one uses the same carbon nanotube source. Conversely, transfer synthesis techniques require the use of a dummy substrate and then the generated film is relocated to the final substrate of choice. Transfer films are generally created using a filter and then they are moved by either dissolving the filter or by utilizing one of the transfer methods (e.g. PDMS stamp transfer, surface energy exploitation or other methods). It will be shown that using different filters with the same nanotube source can also result in films with different properties, which can then be exploited to create films with either specific properties or structures. Understanding the science behind each technique through examples from our own work and the extensive scientific literature will help elucidate the connection between film deposition and film properties.

No part of this digital document may be reproduced, stored in a retrieval system or transmitted commercially in any form or by any means. The publisher has taken reasonable care in the preparation of this digital document, but makes no expressed or implied warranty of any kind and assumes no responsibility for any errors or omissions. No liability is assumed for incidental or consequential damages in connection with or arising out of information contained herein. This digital document is sold with the clear understanding that the publisher is not engaged in rendering legal, medical or any other professional services.

Rachel L. Muhlbauer and Rosario A. Gerhardt 108

INTRODUCTION Since the discovery of the mutliwalled carbon nanotube (MWNT) in 1991 [1] and the

single-walled carbon nanotube (SWNT) in 1993 [2], the carbon nanotube (CNT) has been researched heavily as a material for a variety of applications due to its remarkable electronic, chemical, mechanical, and thermal properties. One architecture of the CNT which takes advantage of these properties and allows for integration into modern device applications is the carbon nanotube based-thin film (CNT-TF). The CNT-TF is a two dimensional array of CNTs with thicknesses varying from submonolayer to hundreds of microns. This two dimensional array can be configured in a variety of ways upon a substrate. CNTs can be aligned perfectly perpendicular to the substrate, perfectly parallel to the substrate, or in some combination of the two and control of this alignment is due to the film fabrication/deposition parameters. For example, both vertically aligned forests and flat networks of CNT-TFs can be fabricated using carbon vapor deposition methods, but under very different conditions [3,4] which will be discussed later on in the chapter. Figure 1 shows a schematic of a vertically aligned CNT-TF as well as a flat network of nanotubes both perpendicular to and along the axis of the substrate.

By control of the microstructure through thin film fabrication techniques, tailored films with specific properties can be developed for a variety of applications including gas sensors[5-10], stress-strain sensors[11,12], thin film transistors [13-22], thin film electrodes [23-27], and nanoporous storage devices[28-30], among others.

There are two types of thin film gas sensors. The pure CNT-TF sensor takes advantage of the hole conducting nature of CNTs to detect adsorption of gas onto the surface of the film. In one such example, exposing a thin film of semiconducting SWNTs to NO2 increased the conductance of the film.

Figure 1. Schematic of vertically aligned (left) and a flat network (right) of carbon nanotubes with views corresponding to the side (top) and top-down (bottom).

The authors suggest that adsorption of NO2 onto the film causes SWNTs to donate electrons which increases the number of charge carries (holes) in the film [5]. Functionalized CNT –based sensors exploits the ease of chemical modification for specific species detection. For example, when SWNTs are functionalized with poly(m-aminobenzene sulfonic acid)

A Review on the Synthesis of Carbon Nanotube Thin Films 109

(PABS), NH3 gas can be detected via a decrease in film conductivity [9]. Stress-strain sensors take advantage of the dependence of electrical properties of CNT-TFs on mechanical deformation (increased strain causes an increase in resistance).

While the above applications rely on sensing a change in the electrical properties, many applications are a direct result of the electrical properties CNT-TFs display. Thin film transistors made of carbon nanotube thin films have shown mobilities of up to 2,500 cm2/Vs [20]. CNT-TF-based transistors can also be used in flexible electronics as they can hold current during repeated bending and relaxation of the film [16]. Transparent electrodes of CNT-TFs are another useful application as relatively low resistance and optically transparent electrodes can be fabricated[23-27]. Finally, nanoporous storage devices take advantage of carbon nanotube thin films that are highly entangled for forming an electrical double layer of high surface area and capacitance [29].

Because of the wide variety of applications requiring specific properties of CNT-TFs (e.g. high surface area, high conductivity), it is important to have an understanding of the entire process of thin film fabrication starting with how the nanotubes themselves are created and ending with film formation as well as all the steps in between. Each step in the process will contribute to the final properties of the film.

Figure 2. Schematic of the Arc Discharge Process, Reprinted with permission from [33].

FACTORS INFLUENCING CARBON NANOTUBE

THIN FILM PROPERTIES While the deposition method for CNT-TFs plays a very large role in the film’s final

properties, each fabrication step from the way the CNTs are created (for all thin film fabrication methods) to the way solutions are prepared (for solution-based methods)


influences those final properties. Understanding the impact of each step during processing will help explain why CNT-TFs behave as they do.

Synthesis-Determined Carbon Nanotube Properties

Arc-Discharge The CNTs first discovered by Iijima[1] were produced by varying parameters in a DC arc

discharge apparatus, which had been initially used to synthesize fullerenes. During the arc discharge process, shown in Figure 2 below, a DC current source is connected to a graphite anode and cathode. The chamber is surrounded by a helium atmosphere at temperatures which reach ~4000 K. The electrodes are vaporized during the passage of current and carbon nanotubes are built up in the system on the electrodes [31]. For MWNTs, pure graphite electrodes are required; while SWNTs can be formed by adding a metallic catalyst/precursor on one of the electrodes. For the traditional arc discharge process, adding a Ni-Y metallic catalyst has been shown to yield 90% SWNT of 1.4 nm in diameter in gram quantities per day [31]. Changing the system setup by employing an arc discharge plasma jet method can produce quantities of up to 1-2 g/min [32].

While under very specific conditions highly pure CNTs can be obtained, different atmospheres in the unit can widely change the morphology of the CNT. In a He/methane atmosphere, thick CNTs covered in carbon nanoparticles can be obtained under high currents while long MWNTs with a diameter of 6 nm can be obtained under the same atmosphere but at lower pressures and currents [34]. Similarly, in a hydrogen atmosphere, long and small diameter MWNTs can be created along with sheets of graphene [35]. Highly crystalline SWNTs can be created under a hydrogen atmosphere when the carbon anode contains 1% Fe [36]. As can be seen by just the few examples above, altering the conditions within the arc discharge apparatus results in CNTs with very different properties.

Laser Ablation/Vaporization The laser ablation process is another high temperature CNT fabrication technique in

which a high power laser is pulsed at a graphite target sealed in a silica tube with an argon atmosphere, and similarly to arc discharge. MWNTs are created with a normal target, while SWNTs require the graphite target to contain metallic catalysts [33]. However, unlike the arc discharge method which leads to a repeatable experiment under one set of conditions, the laser ablation method fails in repeatability studies. In one study, a majority of (10, 10) SWNTs were created using graphite-Co-Ni targets with a diameter of 13.8 nm [34], while a second study under the same conditions found far fewer (10,10) nanotubes and found diameters of 12.2 nm [35]. In one study which tested the effect of the metallic catalyst on CNT growth, researchers found that a few, isolated SWNTs were grown when either cobalt or nickel was mixed with graphite. They also found that high amounts of SWNTs can be grown when graphite is doped with nickel and yttrium, when there was more nickel than yttrium [37]. Growth of extremely short SWNTs can be done using pulsed laser vaporization during


which time-delayed pulses of a Nd:YAG laser are sent at a graphite target. SWNTs with lengths between 35 and 240 nm were found depending on the temperature of the system during the process [38].

Electrolysis While not widely used because of the difficulty in controlling yield and diameter of

CNTs, electrolysis might be a promising, low cost technique for CNT fabrication. During electrolysis-assisted fabrication of CNTs, graphite electrodes are immersed in a liquid salt (LiCl, KCl, NaCl, LiBr, etc) under argon and a voltage is applied between the electrodes [39,40]. A schematic of this process in a condensed-phase apparatus is shown in Figure 3 [33] Varying the salt and temperature of the electrolyte changes the production of CNTs.

Figure 3. Schematic of the Condense-Phase Electrolysis Apparatus, Reprinted with permission from [33].

Flame Pyrolysis

Flame pyrolysis is a recently developed fabrication technique in which the process of combustion is utilized to create carbon nanotubes. Varying the setup, the flame type, the fuel, and the catalyst will vary the type of CNT created. Figure 4 shows the various pyrolysis techniques and the type of CNTs found from these techniques [41].


Burner Configuration

Flame Type

Fuel

Catalytic Material

CNT Type

Premixed

C3H8 Air

Stainless Steel

Carbonaceous Matter

Premixed

C6H6 C2H2 C2H4 O2

None

Fullerenes and MWNT

Co-flow diffusion

CH4 C2H4/ air

Stainless steel Ni-Cr-Fe

Entangled and aligned MWNTs

Pyrolysis flame

C2H2/ air

Nebulizer Carbon Nanofibers (CNF)



Flame Type

Fuel

Catalytic Material

CNT Type

Co-flow diffusion

C2H2/N2/air C2H4/N2/air

Metal nitrate and TiO2

MWNT

Counter-flow diffusion

CH4+C2H2/N2+O2

Ni alloy MWNT, CNF


CH4+C2H2/N2+O2

None MWNT and soot

Inverse Diffusion

C2H4/Air

Stainless Steel and Nickel nitrate

MWNT and CNF



Flame Type

Fuel

Catalytic Material

CNT Type

Co-flow diffusion

C2H4/ air

Cobalt Aligned MWNT

Counter-flow diffusion Co-flow diffusion

CH4/ air

Ni alloy and silicon substrates

MWNTs, entangled and aligned

Premixed C2O6/Air

Stainless steel with nickel nitrate

Aligned CNTs

Co-flow diffusion

C2H4/ air

Fe Ni Pt

Entangled CNTs Aligned CNTs, CNFs

Co-flow diffusion

CH4/ air

Transition metal alloys

Entangled and aligned MWNT


CH4+C2H2/N2+ O2

Ni alloy Coiled, helical, branching and aligned CNTs

Figure 4. Parameters for Flame Pyrolysis Techniques and Resulting Structures. Reprinted with permission from [41].


Recently, a V-type pyrolysis flame technique was employed to create CNTs. In this technique, a V-type pyrolysis flame burner is welded to stainless steel so that synthesis happens inside the V while heat is created outside the V [42]. During this process, carbon monoxide (carbon source) along with helium and hydrogen (helium gas helps to dilute the concentration of combustibles which lowers the yield of non-CNT carbon [43], and hydrogen acts as a promoter for carbon precipitation [44]) are pumped into the V while combustion of acetylene in air is pumped outside the V and combusted to created heat. A catalyst of pentacarbonyl iron is carried into the V-type burner by helium and decomposes into iron in the burner. CNTs formed on these iron catalysts [42]. Temperature, catalysts particle concentration, and sampling time all determine whether carbon nanotubes are formed. The most favorable temperature condition for this technique is between 850°C and 920°C. Above 920°C, catalyst deactivation occurs, while below 850°C not enough carbon is available for CNT production. With respect to catalyst particles, when the reaction is run without any, only carbon black and amorphous carbon are found in the chamber. Also, it is shown that after 5 minutes of this process at 850°C, the maximum yield of carbon nanotubes and length will be reached [45]. Controlling these parameters will help control the length, purity, and diameter of the resulting CNTs.

HiPco A method discovered by Richard Smalley’s group at Rice University, HiPco (high

pressure carbon monoxide) is one of the more popular methods for the growth of CNTs. In this method, CO and Fe(CO)5 are sent into a heated reactor. Thermal decomposition of the iron containing phase results in gas phase iron clusters which act as a catalyst for CNT nucleation and growth. Pressures of up to 10 atm and temperatures up to 1200°C are required for this method. Also, by controlling the pressure at 1200°C, a variety of CNT diameter distributions can be obtained. The smallest diameter CNT made was 0.7 nm: the known diameter for a C60 molecule [46].

Carbon Vapor Deposition Chemical vapor deposition (CVD), particularly catalytic CVD (CCVD), is one of the few

methods allowing for large-scale production of CNTs. In CCVD, decomposition of hydrocarbons or carbon monoxide in the presence of an inert gas is aided by the presence of metallic catalysts on a substrate at temperatures ranging from 500°C to 1100°C (thermal enhanced- CVD, TCVD) or in the presence of plasma (plasma enhanced-CVD, PECVD) . Other CVD methods can be used, including water-assisted, oxygen-assisted, hot-filament, microwave-plasma, and radio-frequency, but the yield of CNTs is less than those in the CCVD techniques [47]. The resulting CNTs can either be left on the substrate on which they were grown or cleaved off for other applications.

A variety of factors can influence the type and properties of the CNTs grown with this method: temperature, metallic catalysts, carbon-containing gas, and substrates. In one study, the effect of temperature on CNT growth was studied on quartz and silicon substrates with iron catalysts[48]. The researchers found that at 800°C only catalyst particles were found.


When the temperature was increased to 850°C, short CNTs were nucleated from the particles. At temperatures greater than 900°C, large quantities of well-aligned CNTs were grown. On the silicon substrate, no CNTs were found at either 800°C or 850°C, and large quantities were found at temperatures greater than 900°C. These researchers also tested the difference in iron and cobalt catalysts for CNT growth. They found that regardless of the catalyst or substrate (quartz or silicon) well aligned CNTs were grown. However, the CNTs grown with an iron catalyst on quartz were slightly longer (16-17 µm) than those grown with the iron catalyst but on silicon (11-12 µm) [48].

In another study, researchers determined the effect of gold, nickel, and gold/nickel hybrid catalysts on a SiO2 substrate on the growth of CNTs from an acetylene source[49]. Gold and nickel thin films with varying gold content (from 0-1 mol fraction of gold) were grown on silica by physical vapor deposition. MWNTs were present on the pure nickel catalysts, while nothing was present on the pure gold catalysts at 520°C. On the hybrid catalyst with 0.1 to 0.2 mol fraction gold, the researchers found an increased number of tubes. At 0.2 mol fraction gold, the tubes were of a much narrower size distribution than those grown at 0.1 mol fraction gold [49].

Another catalyst that can be used is a Cobalt-Molybdenum hybrid nanoparticle. The researchers found that at 700°C with a CO carbon source and the Co-Mo hybrid catalyst three kinds of MWNTs were grown: long CNTs of lengths between 0.1 and 10µm, short CNTs of lengths less than 50 nm, and onion shaped graphene sheets with a catalyst particle at the center. They determined that catalyst particles of 5-15% Co produce long CNTs, the 40-45% Co produce short CNTs, and the 85-95% Co produce the onion morphology [50].

It has also been shown that the size of the metal catalyst influences nanotube growth. Nickel nanoparticles of varying sizes were grown on an amorphous carbon substrate[51]. For nickel nanoparticles with a narrow size distribution around 16 nm, CNTs with an average diameter of 46 nm were found at 700°C and 35 nm at 800°C. Nanoparticles with a wide size distribution with a peak around 60 nm grew CNTs with an average diameter of 52 nm at 700°C and 35 nm at 800°C. Finally, a bimodal size distribution of nanoparticles with a peak at 20 nm and another that extends to 150 nm shows that at 700°C, the CNTs have the same bimodal distribution and at 800°C, the CNTs have an average diameter of 42 nm [51].

It is possible to grow SWNTs using the CVD approach. In one experiment, a 10 µm x 10 µm square array pattern was created on a SiO2/Si substrate using photo-lithography followed by evaporation of a Ni catalyst followed by the deposition of another SiO2 layer[52]. This Ni/SiO2 double layer was lifted off the original substrate. These samples were placed in a thermal pyrolysis CVD at 900°C under an argon atmosphere. A flow of methane and hydrogen gas was passed through the chamber for the growth of SWNTs. Smaller diameter SWNTs were grown when the SiO2 was deposited at slower rates [52].

The substrate plays an important role in the deposition of CNTs due to its interaction with the catalyst. There are a variety of substrates which promote the growth of CNTs: SiO2, MgO, Al2O3, CNTs, stainless steel, and zeolite among others [47]. Pure Si substrates when mixed with catalysts tend to form a metal-silicide which inhibits the formation of the CNT. In this case, buffer layers of a non-silicide forming metal or SiO2 are used to prevent the silicide formation [53]. Aligned SWNTs can be grown on single crystal substrates such as sapphire and quartz along certain crystallographic directions. A- and R- planes of sapphire show a narrower diameter distribution than those on the C-plane of sapphire and those grown on quartz [54]. Conical CNTs can be grown on flat graphite and tungsten foil substrates using a


variety of catalytic metals: nickel, iron, and platinum in the microwave-plasma CVD [55]. It has been suggested that when the substrate is a metal oxide, there is an electronic reaction between the substrate and the metal catalyst. If the interaction between the substrate and the catalyst results in the donation of electrons to the catalyst, this can help promote the catalytic activity and, therefore, nanotube growth. This holds true even for a copper catalyst. Copper, which tends to be non-catalytic, can become catalytic when paired with a substrate that is an electron donor [56].

Porous substrates can also be used in CVD to grow well-aligned and ordered films of CNTs. In one such study, a mesoporous silica substrate was prepared by a sol-gel method resulting in aligned pores on a flat surface with diameters of 30 nm and a pore-pore distance of 100 nm[57]. Iron nanoparticles were fabricated within the pores, and finally, CNTs were formed by decomposition of acetylene at 700°C. The carbon nanotube film formed was aligned perpendicular to the substrate with a thickness of ~50 nm and with a diameter of 30 nm, or the same as the pore diameter. This result suggests that the pore size limits the size of the metallic catalyst which therefore limits the diameter of the CNT [57].

Finally, the type of carbon-based gas precursor will affect the final properties of the CNTs grown. It has been suggested that growth from hydrocarbons happens due to reactive species, free radicals, and intermediates. In order to determine which intermediates result in growth, Iijima’s group performed a systematic study of a variety of hydrocarbon agents. First, they tested the difference in CNT growth for C2H4 and C2H6, the former containing sp2 carbons and the latter containing sp3 carbons under the same experimental conditions[58]. Interestingly, both resulted in CNTs of the same type. However, when the flow rates were moved from the optimum flow rate for each precursor gas, these two gases performed differently. The yield of C2H6 decreased as compared to that of C2H4 when the flow rate was decreased; likewise, the yield of C2H6 increased as compared to that of C2H4 when the flow rate was increased. The authors suggest that this is due to C2H6 being converted into C2H4 at the beginning of the reaction since sp2 carbons are required for production of CNTs [58].

Purification Since the growth of CNTs often requires the use of metallic catalysts and/or organic

precursors, it is necessary to go through a purification step in order to remove particulates which may decrease the properties of the CNTs. Purification can be done at multiple points in the fabrication of the thin films: post-CNT fabrication, post-CNT-TF fabrication, or both.

Ultrasonication-assisted filtration can be done to remove carbon impurities and metal particles from SWNTs. In this process, SWNT-containing soot was suspended in toluene and filtered to extract soluble fullerenes[59]. The resulting filtered liquid (toluene-unsoluble particles) was resuspended in methanol. The suspension was added to a filtration funnel and an ultrasonic horn was inserted into the funnel and placed above a polycarbonate filter membrane. Ultrasonication during filtration keeps the material suspended and prevents build-up on the filter. Methanol was continuously added to maintain a constant volume for 2.5-6 hours. Following this process, the material was washed with sulfuric acid to remove any metal added by the ultrasonic horn. Before sonication, SWNTs accounted for 50% of the soot material; however, after this process, 90% of the resulting solid contained SWNTs. However, this process can cause defects in the material including cut SWNTs and nanorope ends [59].


Another way to purify CNTs is to use a wet etch/oxidation process. One such method is a three step purification process can produce SWNTs with > 98 wt% purity[60]. The first step consists of placing the soot in 3 M nitric acid for 16 hours at 120°C. Next, filtration of the liquid occurred and showed an 82 wt% yield where the loss is determined to be the metallic catalyst particles and some of the crude material. Finally, the carbon matrix is removed by oxidation in air at 550°C for 30 minutes which leaves behind pure SWNTs, 20 wt% of the original yield. Using TGA, the authors concluded that the final yield is >98% pure [60].

While the above methods were described for SWNTs, MWNTs can also be purified. First, MWNT-containing soot is selectively oxidized at 600°C for three hours. Next, the resulting powder is split into two[61]. The first pile is treated with 12 M HCl and vortexed for a short time. Following the vortex, the liquid is left for 24 hours. After 24 hours, centrifugation and rinsing is done several times. The leftover powder is then treated with 6 M HCl in an ultrasonic bath for two hours, centrifuged, and rinsed. The second pile, however, is dispersed in an organic polymer solution of poly(m-phenylenevinylene-co-2,5-dioctyloxy-p-phenylenevinylene) (PmPV) and toluene, ultrasonicated for two hours, and then left to sit for 48 hours. After 48 hours, a trilayered solution was found: the bottom layer was black carbonaceous impurities; the middle layer was a green-black suspension of MWNT and PmPV; the top layer was a clear yellow solution of residue polymer in the solvent. Pipetting out the middle layer and dissolving it in toluene to reduce the interaction between MWNT and PmPV allowed for a separation of the two materials. While the acid treated MWNTs were dispersed, the MWNTs gathered from the polymer dispersion were of equal diameter [61].

SWNTs can be either metallic or semiconducting depending on the chirality of the tube. For some applications, such as field effect transistors which rely on semiconductors, it may be necessary to use only one of these types, so separation of these tubes is necessary. Metallic SWNTs can be etched away using a gas-phase reaction. SWNTs were exposed to methane plasma at 400°C, followed by a 600°C annealing under vacuum. This caused a removal of the metallic nanotubes [62]. Sorting CNTs by electronic structure is another way to collect tubes of different types. By using bile salts as an encapsulating agent, SWNTs can be sorted by diameter, bandgap, and electronic type. SWNTs encapsulated by sodium cholate show that density of the tubes increase with tube length and band gap. This was also seen with sodium deoxycholate and sodium taurodeoxycholate. Using centrifugation and techniques for fractionization allows for the removal of SWNTs of specific lengths and bandgaps from solution [63].

Thin films of SWNTs dispersed in a surfactant were spray dried onto a PET substrate (discussed later on in the chapter). Exposure of the film to 12 M HNO3 for 60 minutes decreased the sheet resistance of the thin film via removal of residual surfactant since no significant doping was observed (oxidation of the SWNTs did not take place). In addition to the removal of the surfactant, exposure to the HNO3 led to densification of the film which improved the electrical behavior [64].

Dimensionality Effects of Single Tubes Another factor influencing the final properties of the thin film are the dimensions of

individual CNTs. For the most part, the diameter of CNTs tends to influence what type of CNT one has (metallic vs semiconducting, SWNT vs MWNT), and as discussed earlier in the


chapter, diameter is influenced by the size of the catalyst particles as well as conditions during growth. Length, on the other hand, is determined by reaction time (exposure to more carbon-containing gases or favorable conditions for growth within the chamber).

The electrical properties of CNTs are one such property influenced by the length of the tubes. Electrical transport within a single MWNT was determined using standard lithography methods to create electrode connections between the ends of one tube[65]. Using this technique, the authors determined that as the diameter increased, the resistivity decreased for tubes of similar curvature. However, they also determined that comparing tubes of different degrees of bending show that an increase in the curvature increases the resistivity of the tubes. They suggest that this is due to an increase in the number of defects in bent tubes [65]. The effect of length on the bandgap of SWNTs with armchair (6, 6) chirality was studied using quantum chemistry based computational techniques [66]. The authors show that for tubes of length < 100 Å, a band gap is calculated. This band gap oscillates in a decreasing manner with increasing length. In the range of 10-20 nm, the authors predict a transition from semiconducting to metallic SWNTs [66]. In another study, the effect of length on the resistance of SWNTs of the same diameter (~2 nm) was studied using AFM electrical and force modes. It is shown that the resistance nonlinearly increases with the length of the tube (from 350 to 2100 nm) [67]. Computational modeling confirmed the outcome of this experiment suggesting that the nonlinear behavior is an intrinsic property of CNTs [68]. The mechanical behavior of CNTs is also influenced by the dimensionality of the tubes. Using a nanoindentation study of a forest of aligned MWNTs, the effective bending modulus of a wall as well as the effective axial modulus of a single CNT can be determined. Three MWNT forests of varying areal densities of 25, 27, and 40 number of tubes/µm2 were probed and shown to have a bending modulus of 1.24, 0.91, and 1.14 TPa, respectively. Individual CNTs of average lengths of 570, 930, 1150 nm were found to have an effective axial modulus of 1.11, 1.90, and 1.23 TPa. These results suggest that there may be a critical density/length above which properties start to decrease [69]. The optical properties of CNTs show a length dependence, as well [70]. When optical absorbance is probed, there is an increasing degree of blue shift in the absorption peaks with decreasing SWNT size. Likewise, the peaks in the photoluminescence intensity spectra also experience a more of a blue shift with decreasing SWNT length. These effects are likely due to quantum confinement effects along the length of a short SWNT [70].

Dimensionality Effects Within Carbon Nanotube Thin Films Similar to how the properties of individual CNTs change with dimensionality, CNT-TFs

are also affected by dimensionality effects. For example, the electrical properties of CNT-TFs can be highly dependent on their microstructures. Vertically aligned forests of CNTs, since they are not perfectly dense, have a preferential axis for electrical conduction. Measuring the electrical properties perpendicular to the substrate on which the forest is grown shows a much higher conductivity than parallel to the substrate [54]. This is to be expected as conduction along tubes would be easier than from tube to tube of a non-perfectly dense forest.

However, in the case of the CNT-TF network where the tubes lie horizontally on the substrate, electrical conduction isn’t as cut and dry. CNT-TFs can be modeled as being made up of CNT bundles, and junctions which connect the CNTs. Figure 5 shows a 3D AFM


topography image of a SWNTs network on a glass substrate in order to describe the concept of bundles and junctions. In this image, junctions are present where CNTs cross one another. Bundles are CNTs (either single or multiples) between two junctions.

For electrical transport in these materials, it is expected that there will be a separate resistance for the bundles and another for the junctions, where the resistance of the junctions is much higher than that of the bundles. A variety of studies have been done to understand the behavior in CNT films and how to change the properties of the junctions in order to improve the conductivity of CNT-TFs. It has been shown that improving the contact between two CNTs by increasing the force on the junction can increase the film’s conductivity by orders of magnitude [71, 72]. By placing two nanotubes perpendicular to one another, the resistance on the interconnect can be measured. The resistance across this junction decreases from 3.36 MΩ to ~100 kΩ under an applied load in the nano-Newtons range [72]. Also, for SWNTs, tube type can also affect the junction resistance [73]. Using C-AFM to probe the resistance along bundles and junctions, the authors determined that metallic-metallic SWNT junctions and semiconducting-semiconducting SWNT junctions both show linear I-V curves due to each tube having similar states for electron transport. However, metallic-semiconducting SWNT junctions show a nonlinear response and a resistance about 2 orders of magnitude higher than for junctions of the same type [73]. Junction resistance can also be decreased with a nitric acid treatment of the CNTs before film preparation [74]. Exposure to nitric acid causes electrical doping in the individual tubes, which decreases the bundle resistance due to the availability of charge carriers [74]. Finally, increasing the length of the bundles increases the conductivity of the thin film. As average bundle length increases (as determined by AFM), dc conductivity scales with the length by a power law with an exponent of 1.46[75]. Since dc conductivity measurements are done between two electrodes parallel to the film, longer nanotube networks can percolate between the electrodes with fewer junctions than for shorter nanotubes [75].

Figure 5. 3D AFM topography image describing bundles and junctions in a CNT network. Image taken by chapter author, unpublished. SWNT sample deposited on glass received from Matthew Garrett of University of Tennessee, Knoxville, and Oak Ridge National Lab.


While these previous experiments used specialized experimental techniques (such as C-AFM or by systematically placing two CNTs on top of each other) to determine the electrical behavior of junctions, it is possible to simultaneously determine both junction and bundle electrical behavior using impedance spectroscopy [76]. By applying an AC voltage and measuring the current flow at different frequencies, the electronic transport behavior of materials can be determined. Using this technique to probe SWNT thin films allows for the direct evaluation of the resistance associated with the bundle and the resistance associated with the junction. Using this method to probe the electrical properties of SWNT-TFs of concentrations 0.4-1.6 µg/cm2, the authors found that there is a three times increase in resistance from the bundles to the junctions [76].

Solutions and Surfactants The majority of CNT-TFs are made using some type of solution-based method which

requires that CNTs be suspended in some sort of liquid. However, it is not enough that the powder containing CNTs just be added to a liquid; there must be some sort of interaction between the solvent and the CNTs to separate bundles of CNTs and keep them dispersed in the liquid. This can be done with electrostatic type interactions between either the solvent itself or with a surfactant in an aqueous solution.

Both organic and acidic solvents have been used to disperse CNTs. In the case of organic solvents, a systematic study was done to determine the solubility of SWNTs in sixteen different solvents to determine which would be the better solvent. The results of this study are reprinted below in Table 1. A star (* ) denotes a solubility under 1 mg/L [77].

Table 1: Solubility of SWNT in a Variety of Organic Solvents,

Reproduced by permission of The Royal Society of Chemistry [77]

Solvent Solubility (mg/L) 1,2-Dichlorobenzene 95 Chloroform 31 1-Methylnaphthalene 25 1-Bromo-2-methylnaphthalene 23 N-Methylpyrrolidinole 10 Dimethylformamide 7.2 Tetrahydrofuran 4.9 1,2-Dimethylbenzene 4.7 Pyridine 4.3 Carbon disulfide 2.6 1,3,5-trimethylbenzene 2.3 Acetone * 1,3-Dimethylbenzene * 1,4-Dimethylbenzene * Ethanol * Toluene *


Common to all solvents with a solubility over 20 mg/L is some sort of an electronegative end group. With the exception of chloroform, the other three chemicals contain a benzene ring. Therefore, it is reasonable to assume that in an organic solvent, pi-pi interactions are useful (but not 100% necessary) for interactions with CNTs. The electronegative groups may act as a charge barrier to keep agglomeration from occurring. Likewise, in a study testing the use of alkyl amide solvents on the dispersion of SWNTs, it was found that highly polar pi systems and alkyl group geometry of the solvent molecules promote dispersion- a combination of electrostatic and steric interactions[78]. It is also worth noting that organic solvents are easily removed through evaporation, so they should not contribute to the properties in film form, although they may impact the CNTs while in solution [78].

Chlorosulfonic acid is considered a ‘superacid’ for SWNT dispersion due to its ability to dissolve high concentrations of SWNT ( > 10%) as well as its ability to dissolve these concentrations in minutes. without oxidation occuring on the SWNTs[79]. It is suggested that chlorosulfonic acid reversibly protonates the surface of SWNTs, and the positive charge on each CNT repels other CNTs, so agglomeration does not occur. Interestingly, when both short and long (> 100 µm) SWNTs. as well as MWNTs at large concentrations are dissolved in this solvent, a liquid-crystalline phase forms, which can be useful for large scale fluid processing of CNTs [79].

When CNT solutions are made using water as the solvent, water alone cannot de-bundle CNTs as the organic and acid solutions discussed above can. Therefore, it is essential to add what is referred to as a ‘surfactant’ to the CNT-water solution. The role of the surfactant is to adsorb onto the surface of the CNTs and promote electrostatic repulsion between tubes while breaking apart bundles. A schematic of this is shown below in Figure 6. As surfactant particles move between CNTs in a bundle, they promote separation by individual tubes which should not react again with the bundle[80].

Figure 6. Surfactant adsorption breaking up a bundle of CNTs. Reprinted with permission from [80].


Surfactants can be polymeric or ionic (both cationic and anionic). Polymeric surfactants, for example Gum Arabic (GA), physically adsorb onto CNTs and separate entangled bundles into individual tubes. These long chain molecules diffuse between CNTs and act as a steric barrier between agglomeration due to the large volume occupied by GA around each tube [81]. Other polymers such as polyvinyl pyrrolidone (PVP) and polystyrene sulfonate (PSS) will wrap around individual nanotubes to promote dispersion. Other polymers such as cross-lined copolymers will encapsulate each tube to enhance dispersion in water as well as in a variety of other solvents [80].

Ionic surfactants will change the surface charge of CNTs to either positive or negative, which promotes separation in water. Figure 7 shows the adsorption of ionic surfactants on the surface of tubes [82]. The end groups will be either positive (cationic) or negative (anionic). The optimum concentration for anionic surfactants is approximately six times higher than that for cationic concentrations at the same molarity [82]. For the most part, in the case of ionic surfactants, anionic types are very widely used in research. In a study researching the maximum concentration of CNTs that can be dispersed, the authors found an enhancement in the maximum concentration when benzene-containing anionic surfactants (sodium dodecylbenzene sulfonate) interact with CNTs compared to non-benzene-containing anionic surfactants (sodium dodecyl sulfate). Likewise, the authors found that neutral surfactants containing benzene groups (Triton-X100) can disperse CNTs better than certain anionic surfactants which do not contain benzene groups. While surface charge is important for solution stability, pi-pi interactions promote stability to higher concentrations [83].

Figure 7: Separation of Individual SWNTs by Ionic Surfactants. Reprinted with permission from [82]

While anionic surfactants have better concentration stability, in some CNT-TF fabrication techniques, they actually hinder formation of the film. Thin films made using spin-coating (which will be discussed later on in the chapter) found their thicknesses limited when SDS and SDBS type surfactants were used to stabilize the solution [84]. Since the


charge for each layer during spin coating would be the same, there is an electrostatic repulsion between layers; therefore, it is important to use a surfactant which will not repel layers in order to increase thickness which will increase the properties of the film. Using a non-ionic surfactant with a low molecular weight such as quinquethiophene-terminated poly(ethylene glycol) (5TN-PEG) can promote film creation using the spin coating technique since electrostatic repulsion is less of a factor. Also, because the surfactant is of low molecular weight, it is easy to remove from the final film without the creation of large porosities [84].

THIN FILM DEPOSITION There are two broad categories which can be used to categorize the large number of

CNT-TF fabrication techniques: direct deposition and transfer. During direct deposition, the CNT-TF is formed on the substrate required for its application. These CNT-TFs can be fabricated using either CVD techniques or wet deposition methods. Transfer-type CNT-TFs are first made on a dummy substrate and then transferred to either another substrate or removed completely as a free standing film. There can be a huge difference in the properties and the topography of CNT-TFs formed using different techniques, so it is important to understand how processing changes the thin films in order to aid in development of tailored films for specific applications.

Direct Deposition

Carbon Vapor Deposition Carbon vapor deposition for thin film growth follows the same procedure as discussed

earlier in this chapter; however, rather than removing the CNTs from the substrate as a powder, CVD for thin film growth creates films on their final substrates with orientations either ordered perpendicularly, ordered parallel, or randomly networked parallel to the surface. Note that this is just a short preview of the topographies created from the large amount of research being done based on CVD for CNT-TF growth. A simple literature search yields thousands of papers on this topic.

Vertically aligned tubes can be reproducibly obtained if the density alloy precursor exceeds a critical value. In one such study, a Co-Ni alloy precursor grown to 100 nm in thickness on a p-type Si (100) substrate was etched for 100-200 s in HF solution and then pretreated using NH3 gas to selectively etch some of the Ni away to expose more Co nucleation sites. The large number of nucleation sites promotes steric hinderance between CNTs such that they only grow vertical to the substrate. Figure 8 shows CNTs with a diameter of 200 nm and height of 5 µm using scanning electron microscopy (SEM) at both 5,000x (a) and 10,000x (b) magnification [85].

Field effect transistor (FET) properties were tested for these vertically grown CNTs. The authors found a turn on electrical field of 1.2 V/µm with a current of 10 nA [85]. It is also


possible to vertically align CNTs under an applied electric field during growth. These vertically aligned CNTs showed a turn on electric field for FET at 4 V/µm with a current of 0.5 µA [86].

Horizontally aligned CNTs can also be grown using CVD. A pure Co precursor was applied on a Si wafer and hydrogen plasma etched to become well-distributed and nanosized[84]. In order for horizontally aligned CNTs to be grown, a guided flow of the gas must be done parallel to the substrate while the substrate is negatively biased. Figure 9 shows the horizontally aligned CNTs. The substrate is the back of the image and is tilted ~45° to show the tube ends [87].

Figure 8. SEM images for Vertically Aligned CNTs grown using PECVD at a) 5000x; b)10000x magnifications. Reprinted with permission from [85].

Figure 9. SEM picture of guided flow CVD deposition of horizontally aligned CNTs. Reprinted with permission from [87].

Comparing the FET properties of horizontally aligned CNTs to vertically aligned CNTs grown under the same conditions but without guided flow, the authors show that the field


effect properties are 2 V/µm at 1 µA/cm2 for horizontally aligned CNTs and 3V/µm at 1 µA/cm2 for vertically aligned CNTs. There is an increase in FET properties for horizontally aligned films due to an increase in body emission for horizontally aligned CNTs [87].

It is possible to grow horizontally aligned CNTs on iron patterned silicon nitride substrates. In one such experiment, strips of Fe were deposited into striped arrays separated by a width of ~500 nm[88]. CVD was used to grow CNTs between each stripe so that each stripe edge is considered an initiation and termination edge for CNTs. Using an external electric field aligned perpendicular to the gap helps to align the CNTs so that they touch nearly perpendicular to each Fe strip; likewise, using an external electric field oriented 45° relative to the gap results in growth aligned 45° between stripes. Figure 10 shows the horizontally aligned CNTs without (a) and with an external electric field (b,c). A single metallic SWNT aligned between Fe stripes shows a linear I-V curve resulting in a resistance around 10 kΩ; conversely, a single semiconducting SWNT aligned between Fe stripes shows a non-linear I-V curve and a resistance of 100 MΩ [88].

Using a combination of these CVD methods in addition to common lithography techniques allows for the deposition of both horizontally and vertically aligned CNTs in a patterned assembly. Figure 11 shows a periodic array from combination growth forming a flower like pattern[89]. The authors surmise that this technique may allow for electromechanical devices made of CNTs which require varied nanotube structures in different orientations on a single substrate [89].

Figure 10: CNTs grown parallel to the substrate, between Fe stripes. A) Randomly; B) Aligned perpendicular to stripes; C) Aligned 45° between stripes. Reprinted with permission from [88].


Figure 11. Combined vertically and horizontally aligned CNTs forming a flower-like pattern. The scale bar shows 50 µm. Reprinted with permission from [89].

Random networks of CNTs are much more commonly grown using CVD due to the sheer number of parameter combinations that will form a random network over an aligned network. A representative random network structure is shown below in Figure 12 [4]. These networks can be used in a variety of applications including gas sensing due to open porosity [90], thin film transistors grown directly on glass substrates, which cannot be used at high temperature CVD techniques due to these temperatures exceeding the low strain temperature [91], field effect transistors due to the large amount of side walls available for body emission [92].

Figure 12. AFM topography image of a high density random network of CNTs grown using CVD techniques. Reproduced with permission of [4].


Direct Deposition Wet Assembly With the exception of thin films made using CVD techniques, most CNT-TFs require

solution dispersion of CNTs followed by wet deposition onto a substrate. These techniques are often low temperature (< 100°C), highly compatible with a variety of substrates including polymer-based ones, low cost, and high speed [93]. However, these thin films may have repeatability issues. While CVD grown CNT-TFs can be made with the same morphology and properties as long as the same parameters are used each times, solution based techniques show more variation stemming from slight differences in solutions, such as stability of the solution, and differences in the environment of deposition.

However, because of the positives for wet deposition assembly, these are extremely popular techniques that are constantly being tweaked for better performance. Figure 13 shows a representative topography image for all wet deposition techniques that will be discussed in the next few pages[93-102].

Figure 13. Images of the surfaces of CNT-TFs for all wet deposition techniques. A) SEM image of an electrophoresis-deposited CNT-TF, Reprinted with permission from [94]; B) 3D AFM topography image for CNTs deposited using tethering/self-assembled monolayers, Reprinted with permission from [95]; C) AFM Topography Image of CNT-TF deposited by Langmuir Blodgett technique, Copyright 2003 The Japan Society of Applied Physics [96]; D) AFM topography image of CNT-TF deposited via dropcasting, reprinted with permission from [97], Copyright 2006 American Chemical Society; E) AFM topography image of SWNT-TF on PET deposited via dipcoating, reprinted with permission from [98], Copyright 2004 American Chemical Society; F) AFM topography image of SWNT-TF deposited by spray coating, Reprinted with Permission from [99], Copyright 2005 American Chemical Society; G) AFM topography image (2x2 μm) of CNT-TF deposited via spin coating, Reprinted with permission from [100]; H) SEM image of CNT-TF created via inkjet printing, Reprinted with permission from [101]; I) SEM image of flash dry deposition of CNT-TF using the Meyer Rod technique, Reprinted with permission from [102].


Electrophoretic Deposition Electrophoresis or electrophoretic deposition (EPD) is a technique born out of ceramic

processing. Electrodes are placed within a stable suspension of charged powders or colloidal particles, size less than 30 μm, and an electric field is applied between the electrodes. This causes the charged particles to flow with the field and collect near the electrode. Next, these particles will deposit on the electrode surface, but only if the surface is conducting [94]. It has been suggested that the deposition occurs in spite of electrostatic repulsion between the similarly charged particles due to the acceleration of other particles toward the collection of particles. This speed causes displacement toward the surface and eventually film formation [103]. A schematic of the EPD setup with CNT solution is shown in Figure 14.

Figure 14. Diagram of Electrophoretic Deposition for CNT-TFs.

EPD is a low cost and relatively simple technique which can provide uniform, homogeneous films with control of film thickness on a variety of substrates, including those which are oddly shaped [94]. It is also scalable since increasing the area of deposition only requires a larger volume of material. Varying the solution parameters (solvent, CNT type, surfactant) for EPD can be used to tailor the film properties.

SWNTs were dispersed in dimethylformamide (DMF) at a concentration of 0.4 mg/mL after 2 hours of sonication. Adding sodium hydroxide to the solution during sonication causes the surface charge of the SWNTs to become negative which allows for deposition on the cathode. After 30 seconds at an electrical field of 20 V/cm, the 0.4 mg/mL solution of CNTs in deposited on the cathode. After 10 minutes, a 1 μm thick film was formed. Changing the salt from sodium hydroxide to magnesium chloride causes the CNTs to be deposited on the anode. No deposition occurred when the CNTs were not surface charged [104]. MWNTs were dispersed in solvents of acetone and ethanol with a ratio of 2:1, 1:1, 1:2 (acetone:ethanol), and pure acetone [105]. In pure acetone, the MWNT film deposited was


highly porous due to the evolution of hydrogen during the process which gets trapped in the film before it can be released. Release of the embedded hydrogen from the film leaves behind large pores (pore size = ~ 70 μm). At the 2:1 ratio, a similar pore size developed. At the 1:2 ratio, the deposition of the film is extremely slow and removal from the suspension causes the film to separate from the substrate. However, at the 1:1 ratio, pore sizes reduce to 1 μm due to a decrease in hydrogen evolution. These pore sizes affected the final electrical properties of the films. The authors determined a film resistance in the kΩ range which they suspect is due to hydrogen adsorption on the CNT walls [105]. In surfactant containing solutions, the electrical resistance of the CNT-TF is much higher [106]. SWNTs were dispersed in water with both SDBS and poly(maleic acid/octyl vinyl ether) (PMAOVE) surfactants. Relatively high contact resistances (GΩ) were found for both types of surfactants, with the PMAOVE being an order of magnitude higher. However, annealing these films in a nitrogen atmosphere greatly reduces the contact resistance closer to that of the non-surfactant containing solutions due to the removal of the surfactant and densification of the film [106].

Electrode distance during EPD also changes properties of the CNT-TFs. Uniformity of field emission from CNT-TFs is a function of the distance between electrodes during the deposition process. For deposition when the electrode distance was either 1.1 or 1.8 cm, non-uniform field emission was observed and only on the edges of the CNTs[107]. At 0.3 mm, a uniform field emission current was achieved. The authors suggested that larger distances create non-uniform electric fields around the electrodes, which lead to non-uniform deposition. Shrinking the distances between electrodes causes the electric field to become more uniform and therefore, the deposition is more uniform [107].

Surface oxidation of CNTs from acid treatment allows for deposition from aqueous solution without the need of a surfactant. Surface oxidation of CNTs leads to deposition on the anode. At low concentrations of solution (0.25 mg/mL) only a small amount of solution deposited and in a very inhomogeneous manner regardless of any of the parameters during EPD. At 0.55 mg/mL, high quality and homogeneous films were deposited with thicknesses of up to 10 μm with low porosity [108].

Aluminum and titanium coated cathodes can be used during EPD of CNTs to limit porosity. During this process, oxygen generation occurs at the electrode which passivates the metallic species to either aluminum oxide or titania. The metallic films absorb the oxygen that is generated during electrolysis which limits the porosity of the film. Once total passivation occurs, EPD of CNTs on the electrodes stops due to the drop in electrical current in the cell. These films have surface resistances between 1000 and 10000 Ω/sq with transmittance of 60-90% at 550 nm [109].

Self-Assembled Monolayers The self-assembled monolayers (SAM) technique is a classical organic chemistry

technique which allows for tethering of organic molecules onto, traditionally, gold surfaces. This technique is achieved in one of two ways, which is shown below schematically in Figure 15 for CNT-TFs. On the left, the gold substrate surface is functionalized with a molecule that has a head group of sulfur covalently bonded to the surface of the gold and a functional group, known as X, at the tail. The interaction between X and CNTs end-functionalized with a Y group will react to form a covalent bond which will form a layer of thickness


approximately equivalent to the height of the CNTs (short organic molecule height) [110]. Also, if the other end of the CNT is end-functionalized with another X group, continuous film growth will occur from the reaction of X and Y. On the right, the CNT is end-functionalized with a group, referred to as Z, which is terminated by S-H. The S-H will react with the gold surface, bonding the gold and sulfur together and releasing hydrogen, and form a layer with thickness approximately equal to the height of the CNTs [111].

Carboxyl functionalized MWNTs of length between 10-30 μm self assembled onto a gold surface functionalized with a sulfur head group and an NH2 tail group similar to what is shown on the left of figure 15[111]. Studying adsorption using cyclic voltametry, the authors found that as incubation time increased (time for CNTs to tether to the surface), the response is modified due to the increasing number of CNT molecules attached to the surface. This is also shown by an increase in current flow with increasing incubation time. Using the AFM, the authors also found that an increased incubation time with an increased number of tethered surface-molecules created a film with greater length distribution [110]. Similar to the scheme on the right of figure 15, SWNTs functionalized with cysteamine, -NH2CH2CH2SH, is reacted with a gold surface to create a SAM of SWNTs on Au (111). The authors found that at short incubation times, short SWNTs are the dominant molecule. Conversely, at high incubation times, the length distribution shifts to greater lengths [111].

More complex structures and substrates can be fabricated using the SAM technique. An ITO substrate was silanized with NH2-(CH2)x-SiO2-OH with oxygen tethered on the surface[112]. This was reacted with C60 and then SWNTs to form a complex structure containing alternating strands of SWNT-C60-ITO and SWNT-ITO. This structure provides blocking of redox reactions at the surface as well as a more dense film due to π-π interactions between SWNTs and the C60 molecule [112].

Figure 15. Traditional Self-Assembled Monolayer techniques adapted for CNT-TFs. Left: surface tethered gold surface reacting with functionalized CNT. Right: Pure gold surface reacting with functionalized CNT.


Another complex structure that takes advantage of π-π interactions for film construction is a SAM of aromatic molecules on Au surfaces[113]. CNTs interact with the aromatic rings and create a film which is highly oriented and easily removable under pH change due to non-tethering of the CNTs to the substrate. This allows it to be used for applications such as drug delivery at pH sensitive sites [113]. While this mechanism is not a traditional CNT-TF based on SAM, it delivers similar structures and densities as the others, while only employing π-π interactions to place the CNTs to the substrate.

Langmuir-Blodgett

The Langmuir-Blodgett (LB) film making technique is a “two solution” process. CNTs are dispersed in an organic solution (typically chloroform) or in a surfactant-containing aqueous solution. This solution is then placed on top of another solvent which will not interact with the CNT-containing solution. Film deposition can take place by either horizontally placing the substrate on top of the CNT dispersion, “placing,” and lifting the substrate back out (Figure 16a) or by submerging the substrate in a vertical arrangement, “dipping,” at a constant velocity allowing for deposition to occur on both sides of the substrate (Figure 16b) [96]. The LB method relies on both Van der Waals interactions between the substrate and the CNT dispersion as well as on the phobicity of the substrate. Hydrophobic substrates are more hydrophobic typically when water is the non-interacting solution, as these substrates will push away the water during deposition. When other media or other dispersing agents are used, the substrate phobicity may not have an effect on film formation [111,112]. Orientation of LB films depends upon compression and flow pressures. Horizontally deposited LB films have an orientation due to compression only which is parallel to the barrier of the LB container. Vertically deposited LB films must take into account both compression and flow, and therefore, the tubes will be aligned with the dipping direction due to the upward and downward motion of the substrate [96].

Figure 16. a) Horizontal; b) Vertical Langmuir-Blodgett Film Formation techniques.


Vertically dipped LB thin films of SWNTs were made on single crystal silicon substrates[116]. Multiple dips were required for CNT deposition. On the first dipping step, CNTs only deposited onto the substrate during the lifting step, not the submerging step. Subsequent dips found CNT deposition during both submersion and lift. The topography of this film was a network of interconnected bundles of nanotubes. This film showed a linear current-voltage behavior with a resistance near 500 Ω after only a single dipping [116]. The topography of horizontally deposited LB films depends upon the surface pressure[117]. At low surface pressures, CNTs deposit in a loosely packed manner over a large area. At too high surface pressures, the CNTs collapse into the non-interaction solution, forming aggregates on the substrate. At ideal surface pressures (in this study 30 mN/m), SWNTs are aligned orderly on the substrate (in one direction) [117].

Dropcasting The dropcasting technique for thin film formation works by placing a drop or by covering

an entire surface with the CNT-containing solution. The drop can then deposit the nanotubes via evaporation of the solvent (Figure 17a) [118], rolling down an angled substrate such that it leaves behind CNTs as it rolls (Figure 17b) [97,119], or through laminar flow of the drop during which a stream of air blows the drop across the surface and deposits the CNTs as it is blown away (Figure 17c) [120]. Both single and multiple layers of CNTs can be deposited using this technique.

Patterned substrates will also allow for deposition in certain areas and with a certain CNT alignment. Patterning a substrate with alternating areas of polar and nonpolar groups will cause the CNT aqueous solution to segregate into the hydrophilic rows and avoid the hydrophobic rows [118]. The droplet will form into multiple cylindrical droplets in these polar rows. Directional evaporation of the liquid through the channel will align the CNTs parallel to the rows [118]. Similarly, placing the substrate coated with alternating hydrophobic and hydrophilic SAMs at an angle and letting a droplet roll down will orient the CNTs into aligned rows along the hydrophilic SAM lines [119]. This method can also be exploited for liquid crystalline processing. Solutions of 0.1-0.2 mg/mL were dropcasted onto a NH2-terminated SAM on a silicon wafer. These formed nematic oriented topographical textures of monolayered CNTs. Increasing the evaporation rate of the solvent decreased the size of liquid crystalline domains [97]. Alignment without SAMs can also be done with dropcast-fabricated CNT-TFs. Droplets of CNT suspensions are put on a silicon substrate and nitrogen is flowed across the substrate. The CNTs align along the flow direction. Multiple dropping steps will increase the density as well as the alignment of the CNT-TF. The increased number of steps forces the CNTs to align nearly perfectly along with the flow direction, whereas few steps show more randomness with a majority of the tubes aligned along the flow direction [120].


Figure 17. Schematics for the dropcasting process. A) Drop and evaporation; B) Drop and roll; C) Drop and air blow. Modified from references [97, 118-120].

Dipcoating

The dipcoating method is an extremely simple and cost effective, but potentially time

consuming method for CNT-TF fabricating. In this method, a substrate is submerged into the CNT dispersion, as shown in figure 18, and removed after a given amount of time for coating to take place. Multiple submersion steps can be done to build the thickness of the CNT-TF. A rinse step after each layer deposition is required to remove impurities from the surface [121].

The dipcoating technique has been used to coat poly(ethylene terphthalate) (PET) for use in flexible transparent electronics[98]. The PET substrate is placed in a SWNT-containing dispersion with surfactant for 30 seconds, air dried for 5 minutes, and rinsed in toluene to remove excess surfactant. The sheet resistance of these films decreases from 100 kΩ/sq to ~ 100 Ω/sq with an increase in film thickness (increase in number of deposition steps). However, the transmittance drops from > 95% to ~82% with increasing film thickness. The film resistance remains relatively constant under bending [98]. This technique can also be done to create oriented CNT-TFs. A silicon wafer was vertically submerged in a surfactant-contain CNT aqueous dispersion, and the solvent was allowed to evaporate. After 2 days and an approximately 50% reduction in solvent, a thin film was formed on the substrate. AFM analysis showed oriented domains of CNT bundles on the surface [122]. Forming a CNT-TF on a poly(ethylene naphthalate) (PEN) substrate shows an improved adhesion of CNTs compared to a CNT-TF on PET. This is likely due to the presence of aromatic rings in PEN. This improved adhesion leads to a resistance of 100-300 Ω with a transparency of 88% after 3 dipping steps [123].

The surface roughness of a substrate impacts the density of the CNT networks in the CNT-TF during dip-coating. Increasing the surface roughness of glass using plasma etching causes the density of CNT networks to increase due to an increase in surface free energy which is minimized by film formation. An increase in density causes a decrease in sheet resistance of the CNT-TF; therefore, it is possible to create thin films with tailored electrical properties by controlling the time of plasma exposure for the glass [124].


Figure 18: Substrate submersion in CNT-containing solution for the dip coating process.

Changing the surface charge of the SWNTs by changing the surfactant present (either negative with SDS or positive with dodecyltrimethylammonium bromide (DTAB) will aid in the deposition of CNTs on a silicon wafer coated with polymer to enhance the first layer adhesion. Alternatively dipping of the wafer in the positively surface charged SWNTs and in the negatively surface charged SWNTs allows for the buildup of many layers of SWNTs without the aid of polymer layers which lower overall conductivity of the film. The researchers indicated that the through-plane (perpendicular to the film) conductivity arises from tunneling of charges between each layer of nanotubes [125].

Spray Coating The spray coating technique is an extremely simple and fast method for CNT-TF

fabrication. In this method, a CNT solution is sprayed at a substrate, which can either be heated with a temperature appropriate for evaporation of the solvent or unheated. This process is shown below in Figure 19. Using a diluted solution along with multiple spraying steps will lead to a relatively homogeneous film. While this method is fast and simple, defects the size of the droplets leaving the spray bottle are found in the film, especially for thinner films. Thicker films which result from a large number of spray steps tend to mitigate the defects, although greater time is required [93].

For films prepared by spray coating from CNTs dispersed in dimethylformamide on unheated glass substrates, there is a sharp increase in conductivity with respect to the thickness of the film. By the time the thickness reaches 100 nm, CNT-TFs show an increase in conductivity from 1x10-3 to 100-1000 S/cm. Before the critical thickness is reached, the CNT-TFs have a low connectivity; beyond the critical thickness, the connectivity is high which causes the dramatic increase in conductivity [99]. For CNT-TFs sprayed from aqueous solutions containing surfactants, nitric acid treatment of the film causes the removal of residual surfactant particles which increases the conductivity of the film by ~4 times [64]. To improve adhesion, it may be necessary to apply a polyimide on top of the network which percolates through the network. The use of the polyimide increases the resistance by 10% [126].


A spray coated CNT-TF can also be patterned for electrowetting devices. There are two ways to make a patterned spray coated film. In the first case, CNTs are sprayed onto glass and PET substrates. Next, the film is covered with a patterned photoresist and exposed to a plasma etching step. During plasma etching, the exposed CNTs are removed. Removal of the photoresist leaves behind a patterned film. In the second, the patterned photoresist is put down on top of the substrate first, and the film is sprayed second. The photoresist is removed which removes the film on top of it, leaving behind only the CNTs on the exposed substrate. Both types of films will change the properties of a water drop under an applied electrical field [127].

Combining spray drying and dipcoating techniques will allow for the formation of dense CNT networks and increase the conductivity of the film. CNT-TFs are formed via a large number of dipping steps followed by a single spray dry step. The addition of a single spray coating step leads to an increase in conductivity from 25-55% depending on the thickness of the structure prior to spraying with a 10% reduction in transparency. It is suggested that the spray step fills in regions not as occupied by CNTs during the dipcoating steps which leads to better connections and a more dense thin film [128].

Figure 19. Schematic for spray coating technique and apparatus for CNT deposition.

Figure 20. Spin coating technique: a substrate (white) with a CNT suspension on it (gray) rotating counterclockwise (arrows) to spread the suspension over the substrate and deposit a thin film.


Spin Coating During the spin coating technique, a substrate with an excess amount of solution is

rotated at a high speed to coat the substrate with a thin film by centrifugal force. The excess fluid is spun off the edges of the substrate, and the solvent is evaporated, leaving behind a thin film. Spin speed and acceleration, length of spin coating process time, concentration of the solution, and the solvent-substrate interactions will all impact the thickness of the film. Multiple spin steps can be done to create thicker films [129]. Figure 20 shows a crude schematic of a substrate with excess CNT solution (gray) spinning in a counterclockwise direction demonstrating the spin coating technique.

Spin coating of CNTs onto substrates is not a widely done process because there is a limit in the thickness of the film that can be achieved. Solutions of CNTs often rely on electrostatic repulsion to keep the CNTs from agglomerating, either by addition of a charged particle which wraps the CNT or by using a solvent which creates a surface charge on the CNT. The repulsion between CNTs often makes thicker layers impossible to deposit, so overcoming this obstacle must be done to create homogenous films. This can be achieved by using neutral solutions [100,130] or by modifying the substrate to allow for better adhesion so fewer layers are required for dense network formation [100,130,131].

Neutral solution deposition on hydrophilic glass will show a drop in sheet resistance to 200 Ω/sq until reaching 15 layers [100]. Beyond 15 layers of CNTs, the drop in resistance is negligible. Treating the film in a mix of nitric acid and isopropyl alcohol after deposition will drop the sheet resistance to 120 Ω/sq after five layers, so subsequent layers are unnecessary. It is suggested that the exposure to the alcohol-acid solution etches away any amorphous carbon and also pulls the network closer together [100,130].

Addition of a layer of PEDOT:PSS to the glass substrate before CNT deposition will allow for faster adherence of the CNT film to the substrate. Approximately 10 layers are required for equal sheet resistance as that for the uncoated glass substrate. The authors suggest that this is due to increased surface energy provided by the PEDOT:PSS layer [100]. Similar to the response of uncoated glass, the alcohol-acid treatment decreases the number of layers required to reach an equivalent resistance. Using the PEDOT:PSS layer and the alcohol-acid treatment leads to the same sheet resistance of 120 Ω/sq, but after 2 or 3 layers [100]. Similarly, on hydrophilic PEN and PES flexible substrates, the sheet resistance will drop from about 400 Ω/sq to less than 100 Ω/sq upon alcohol-acid treatment [130]. Functionalizing a silicon surface can also improve specific SWNT deposition via spin coating. Placing phenyl type groups on the surface will cause a stronger deposition of metallic SWNTs than semiconducting SWNTs. Conversely, functionalizing the surface with amines will cause an increase in the deposition of semiconducting SWNTs than metallic SWNTs. The presence of these groups also allow for alignment of the tubes in a single direction rather than a random network [131]. By modifying the substrate to only accept one type of CNT, devices which require specific kinds of electrical properties can be tailored.

Inkjet Printing The inkjet printing technique for CNT-TFs is exactly as it sounds: an ink composed of a

dispersion of CNTs is placed on a substrate similar to how a printer works. The ink can be


loaded into a printer cartridge and used to print on paper or other soft substrates that can be run through a printer. For hard substrates, the ink can be loaded into a pen type apparatus, as shown below in figure 1, which can be used to print CNT containing lines and patterns, depending on how the apparatus is made. For figure 21, nearly homogenous lines can be printed in a controlled fashion by placing a pen containing CNT ink on a substrate and moving it at a constant speed [101]. The inkjet printing technique is cost-effective and highly scaleable [132].

One such way to make a CNT aqueous based ink without the aid of a surfactant for a homogeneous dispersion is to chemically modify the CNTs with carboxyl groups in order to keep the tubes from agglomerating in solution. This ink is loaded inside a desktop bubble-jet printer and could print lines with a width of ~70 µm after a single print. Multiple prints caused a loss in the narrow lines that can be achieved due to a spreading of the ink. For this apparatus, as the number of print steps increases, the sheet resistivity drops by an order of magnitude from 1 to 0.1 MΩ/sq for CNTs on paper [132]. Using N-methyl-2-pyrrolidone as the solvent rather than water, unmodified CNTs can be printed using an inkjet piezoelectric dispenser for field effect transistors. These prints show semi-ordered domains of CNTs on silicon substrates. Field effect properties were measured; however, due to the high number of metallic CNTs, the field effect properties measured are poor, and a better separation of CNTs is necessary for this technique to be validated [133].

The substrate properties impact the final morphology of the CNT-TF. The wetting properties of the glass changed the deposition of the film. When the contact angle was larger (greater degree of hydrophobicity), the ink drops were smaller and rapidly decreased in size during evaporation. This led to non-uniform film formation. Low contact angle glass showed the opposite effect. When the substrate is heated, the ink rapidly evaporated and formed a uniform network; however, heating the substrate too high led to an increase in the temperature in the air between the printer and the substrate[134]. This caused the ink to evaporate in the air. A temperature of 60°C led to an ideal network formation for SWNT in DMF. Allowing the film to dry slowly at room temperature resulted in the formation of so-called ‘coffee rings’, where dense networks were found on the inside and outside of the ring but not in the intermediate area [134].

Figure 21. Schematic for printing on hard substrates using an ink-filled pen and a control arm which moves in the direction of the arrow. Reprinted with permission from [101].


Meyer Rod With Flash Dry The Meyer Rod coating technique is a popular technique from industry for thin film

deposition that has been adapted for large scale CNT-TFs. In this technique, a CNT dispersion is dropped on a substrate and a coiled coating rod is used to spread the coating. After, a heating element is passed over the film to allow for even drying of the film [102, 135]. This process is shown schematically in figure 22 including a zoomed in region near the coil where the rod is also the heating element [135]. IR lamps which are moved in a controlled manner across the substrate can also be used once the dispersion is spread [102].

It is necessary for the fluid to have specific rheological and wetting behaviors. The solution must flow well enough to bring the solids along for uniform deposition, so it must have a low surface tension to avoid defects. Once the film is rolled, the viscosity must be high enough to allow for drying to occur without flow. The as-rolled film has a wavy texture to it, which is a cause of capillary effects between the rod and the solution, and in order for the wavy texture to dry flat an uniform, the viscosity must be reasonably high [135]. This method can be done on a variety of substrates, such as glass and PET, so long as the wetting behavior between the liquid and the substrate is sufficient for deposition [102]. There is a limitation, however, of the surfactants that can be used in this method. The SDBS surfactant mixed with CNTs ends up providing a very low viscosity, which causes a poor deposition [135]. Adding Triton X-100 to the dispersion increased the viscosity by 3 orders of magnitude and led to highly uniform CNT-TFs with sheet resistances of 100-1000 Ω/sq. Acid treating these films decreased the sheet resistance by an order of magnitude [135].

Film Transfer While the solution based methods are popular because of their ease in deposition and

relatively low cost to make, for many of the methods, specific substrates (for example, gold for Langmuir-Blodgett deposition) are required for film deposition to occur. However, certain applications require specific substrates which may not be the one used in deposition; therefore, film transfer methods were developed. Additionally, the film transfer method allows for much easier formation of both patterned thin films and aligned thin films without the use of harsh chemicals/etchants or CVD techniques [93].

In film transfer, the CNT-TF is first formed on a dummy substrate. This dummy substrate can either be a filter through which the CNT dispersion is filtered through or a hard substrate on which CNTs are grown or deposited. From the dummy substrate, the CNT-TF is moved to another substrate. The methods and science behind the transfer from the dummy to the final substrate will be discussed in this section.

Single Step Surface Energy Exploitation The first set of transfer methods involves exploiting surface energy differences between

the dummy and final substrate in order to move the film. Vacuum filtration is a popular method for forming dense, flat films of random networks of CNTs on a dummy substrate. A filter is attached to a vacuum apparatus and the solution is placed on top of the filter. The


vacuum pulls the solution through the pores, leaving behind a network of CNTs on top of the filter [136].

Figure 22. Schematic of the Meyer Rod coating technique, reprinted with permission from [135].

In the most basic of the transfer methods, vacuum filtration is done on a ceramic filter with a pore size of 0.2 µm. The formed film is pressed against a plastic surface. The surface energy between CNTs and inorganics is much higher than between CNTs and organics; therefore, the CNT-TF will move from the filter to the substrate. A schematic of this is shown in figure 23 where the ceramic filter (gray) in initially covered with a CNT-TF (black). It is moved onto the plastic substrate (white) and pressed together with some force, F, and rubbed in a single direction. When the ceramic filter is removed, the film has transferred to the polymer substrate. During transfer, the CNTs have aligned in the direction that the surfaces were rubbed together [136]. While this method is extremely easy, it is not always repeatable since the same amount of film may not transfer each time. The schematic shows a simplified version where the entire film switches substrates; however, that is rarely the case.

Figure 23. Steps for CNT-TF transfer under direct pressing. In step 1, the CNT-TF (black) is on the ceramic substrate (gray) and is moved such that the CNT-TF is sandwiched between the ceramic and the polymer (white) substrates. In step 2, a force is applied between the two substrates. In step 3, the substrates are separated and the CNT-TF has moved from the ceramic substrate to the polymer substrate.


Figure 24. Water-Assisted Transfer Process. Reprinted with permission from [137].

Another way to exploit surface energy differences is to employ a water-assisted transfer of CNT-TFs from filter to substrate. The process can be shown below in figure 24[137]. The CNT-TF is created via vacuum filtration on an aluminum oxide filter with 0.2 µm pore size. It is pressed onto a PET substrate and submerged in water. Due to water flow through the filter, release of the film occurs. Since the surface energy between the filter and film is much higher than the surface energy between the substrate and film, the film settled on the substrate. Upon removal from the water, the film was stable on the PET substrate. The structure found after transfer was that of a random network and a sheet resistance between 160 and 2000 Ω/sq was found for films of increasing thickness [137].

Multistep Surface Energy Exploitation Transfer Another way to remove CNT-TFs from the substrate they are made on is to peel the film

off using tape. The CNT-TFs will easily detach from the substrate they are on and stick to the tape. From the tape, you can move the film to the final substrate. In one case, a CNT-TF is made on a cellulose nitrate filter with a pore size of 0.2 µm under vacuum filtration[138]. A piece of adhesive tape with a rectangular hole cut in the middle was applied to the film surface. The tape picked up the film, and the hole in the tape allowed for movement from the tape to another substrate, in this case a metallic grid for imaging. This technique essentially forms a free-standing film which can then be affixed to whatever substrate is needed. A dense, random network of CNTs is formed [138].

Tape can also be used to transfer aligned CNTs grown from CVD on quartz. In this process, the CVD thin film of vertically aligned CNTs is coated with a 100 nm gold film. A thermal tape is place on top of the film and peeled off slowly to remove the aligned film. The tape is affixed to the target substrate, PET, and heated above 120°C, the temperature at which this tape loses its adhesive ability. The gold film is etched away, leaving behind the aligned thin film on the target substrate. Since the aligned film cannot be grown on the target substrate under CVD conditions, this method allows for the vertically aligned forest structure to be easily moved to other target substrates [139].


Pdms Assisted Transfer Poly(dimethylsiloxane) (PDMS) is an elastomeric polymer with a hydrophobic surface

which can be treated to become hydrophilic. It will not absorb water based solutions, and therefore, it will not swell upon exposure to inks and can maintain its shape and mechanical properties [140]. Because of its unique properties and the ease in creating a pattern on its surface for patterned ink transfer, it is often used as a stamp to transfer ink from one substrate to another substrate.

PDMS assisted transfer or microcontact printing can be done using a variety techniques. Common to all techniques, however, is the use of the PDMS as a stamp for CNT-TF transfer. In the first technique, the stamp itself serves as the first substrate shown below in Figure 25a. A solution of CNTs is applied to the PDMS using a spin coating technique. The surface containing the ink is contacted against the receiving substrate and removed slowly in a peeling kind of fashion leaving behind the CNT-TF on the receiving substrate [141]. PDMS can also be used as an intermediate step between two substrates. The initial CNT-TF can be made using a variety of methods including vacuum filtration [142,143] and Langmuir-Blodgett techniques [144]. In the second step, a PDMS stamp is placed on top of these initial substrates to remove the formed film. The PDMS stamp is then pressed onto a final substrate to transfer the film. This technique is shown schematically in Figure 25b. Lastly, films can be made on substrates and transferred to PDMS which, in this case, is the final substrate via stamping and shown below in figure 25c. In this case, the initial substrate has a patterned photoresist on it and CNTs are deposited on top. The regions with the photoresist are etched away and the resulting pattern is transferred. This allows for very intricate patterns to be moved to the PDMS substrate [145]. Just like the other surface energy exploitation techniques, PDMS assisted transfer allows for films to be formed on substrates which could not be used initially in the film making process.

Figure 25. PDMS Assisted Transfer Techniques. A) The CNT film (black) is deposited on the PDMS (white) and pressed onto the final substrate (gray) for transfer. B) CNT thin film is grown on a substrate (gold) and, through pressing, is transferred to PDMS. Finally, the PDMS is pressed against another substrate for complete transfer. C) CNT thin film is fabricated on a substrate. Through pressing, it is transferred to PDMS where it is used in applications.


This method is helpful for depositing patterns onto substrates without any etching or extra processing. The surface of the PDMS can be shaped such that some regions are raised with respect to the rest of the top surface. When ink is deposited on it and the PDMS is then pressed on the substrate, only the ink on the raised areas is moved to the substrate. Very fine patterned regions can be printed [141-143].

Other Transfer Techniques The transfer techniques already described take advantage of a surface energy difference

allowing from the movement of a CNT-TF from one surface to another. However, surface energy isn’t the only way to transfer a film. High energy sources will aid in the movement of films. A Nd:YAG Laser with a pulse of 10 ns can be used to transfer a CNT-TF on glass to another substrate[146]. The laser is pulsed from the back of the substrate and the absorption of the energy by the film causes the film to be lifted from the support substrate to the target substrate [146].

Microwaves can also assist in the transfer of vertically grown CNTs by CVD from silicon to a polymeric substrate. The polymeric substrate is placed on top of the forest of CNTs and the system is exposed to microwave energy. The microwaves cause the region between the carbon and silicon to heat while the polymeric substrate remains at room temperature. The regions where the CNTs touch the polymeric substrate undergo localized heating which causes the CNTs to transfer directly to the polymeric substrate [147].

Buckypaper Buckypaper, so named because it is a carbon nanotube network that has an appearance

similar to that of paper under a microscope, is normally made by a vacuum filtration method. Double-walled carbon nanotubes (DWNTs) are used to make networks that are thick enough to be a mechanically stable free standing film[148]. The dispersion which undergoes vacuum filtration contains no surfactant and leaves behind a DWNT network that is hexagonally packed and has very similar diameters between CNTs [148]. A buckypaper with a thickness of 30 µm is flexible and mechanically stable enough to be folded into origami without breaking, as shown below in figure 26. SWNT and MWNT buckypapers can be made as well.

The mechanical properties of buckypaper are outstanding. The mechanical properties arise from the interactions between junctions in the network as well as from the microstructure of the buckpaper. Reported elastic moduli range from 0.6 to 4.2 GPa and strength to breakage ranges from 6.3 to 33 MPa [150]. These properties can be slightly raised by oxidizing the tubes before the buckypaper is made. The addition of the oxygen to the CNTs aids in the solution dispersion which decreases the bundling and increases the entanglement [150]. Chemical force microscopy done on SWNT buckypaper which contained substitutions of alkane and arylthiol molecules on the side walls showed an increased interfacial interaction for the doped SWNT buckypaper when compared to non-substituted forms. Increasing the interfacial interactions decreases the number of bundles which increases the mechanical properties [151]. The reported electrical conductivites range from 103 to 105


S/m for SWNT buckypapers [152] and 106 S/m for DWNT buckypapers [153]. MWNT buckypapers have shown a conductivity of 104 S/m.

Figure 26. Images of buckypaper; a) image of the formed buckypaper. The inset is an SEM micrograph showing the thickness of the paper; b) TEM image of the microstructure of the buckypaper; c) origami folded buckypaper. Reprinted with permission from [149].

Carbon Nanotubes on Paper Substrates

The use of paper as a substrate is a relatively new idea for CNT-TFs but is already a

promising technology for energy-storage devices as well as for field emission. Some of these studies focus on making the paper conductive by inserting CNTs into the paper network. Highly conductive paper with a sheet resistance of 10 Ω/sq was created using the Meyer Rod method to coat the paper. Such a low resistance is possible due to the conformal coating of the paper fibers with the CNTs[154]. In addition to the low resistance, the CNTs on paper are highly stable and do not peel off during a scotch tape test nor do they wash away in water [154]. In another study, incorporating a conductive polymer with the MWNTs in solution and using unidirectional drying to help pull the solution into paper after coating results in sheet resistances between 10 and 100 for the conductive polymer/CNT paper and between 20 and 500 for the paper containing only CNTs[155]. Higher sheet resistances correspond to smaller concentrations of CNTs. It has also been shown that increasing the surface roughness of paper coated with CNT inks leads to a better threshold field voltage; therefore, flexible and geometrically-enhanced field effect transmitters can be made [156].

In order to understand the role of the paper structure and the drying method on the final properties of the deposited CNT-TFs, a systematic study was done by the authors varying the pore size of the same type of filter paper. Four layers of 1 mg/mL MWNT (0.5-2 µm in length and 6-15nm in diameter), dispersed in water with 10 mg/mL of SDBS were deposited on filters with pore sizes of 1 µm, 5 µm, 25 µm, and 40 µm. The films were deposited by vacuum filtration, dropcasting and drying in a single temperature environment, and dropcasting and drying in a heated environment with a higher temperature ceramic heating board underneath to promote unidirectional drying[155]. It was found that during vacuum filtration, there is a huge effect of the pore size of the initial paper on the final electrical properties. The sheet resistance at 1 µm and 5 µm pore sizes was 8.3 and 6.1 kΩ, respectively. The slightly larger pore size seems to aid in the enhancement of the electrical


properties due to more filtering of surfactant and better surface coverage from the entangled network. At 1 µm pore size, the surface has regions that are uncovered and contains more surfactant. At pore sizes much greater than the MWNT length, the sheet resistance increases by 6 orders of magnitude. For films that were dried, the determined sheet resistance was similar for all pore sizes and both drying techniques (between 10 and 40 kΩ). These drying techniques retain a greater amount f MWNTs on the surface at the larger pore sizes but do not reach the sheet resistances of the vacuum filtration films. Figure 27 shows SEM micrographs of the lowest resistance and one of the high resistance vacuum filtered films. In figure 27a, the network is interconnected and spans the entire image. Conversely, in figure 27b, there are regions of dense networks surrounded by regions devoid of nanotubes leading to the high resistance of these films [157].

Figure 27. CNT networks corresponding to sheet resistances of a) 6.1 kΩ; b) 5.1 GΩ.

CONCLUSION

A large number of techniques can be employed in the fabrication of CNT-TFs. Direct

deposition of the CNT-TFs by CVD techniques allow for the greatest control for film topography; however, this technique is limited by the cost as well as the limitations in the substrates that can be used. Wet direct deposition methods are helpful in lowering the cost of the films; however, control of the film topography is more difficult and is often different from one deposition method to another. Transfer methods allow for films to be created on one substrate and moved to a more appropriate substrate or used to form useful patterns on the final substrate. Transfer methods are simple to employ but generally rely on surface energy manipulations to be effective.

Thin films based on carbon nanotube networks provide the ability to make CNT-TFs with varying microstructures and properties that may be used in a wide variety of applications. However, the reader is reminded that the final properties of any CNT-TFs depend on many different factors such as CNT type (whether SWNT, DWNT, MWNT), electronic properties( whether metallic, semiconducting or combination of both), physical dimensions of the nanotubes (diameter, length, aspect ratio), arrangement of the CNT network (vertically aligned, horizontal deposition, whether aligned or random), method of deposition, thermal history and what substrate is used to deposit the film or to transfer it to. Therefore, in spite of the abundant literature in this field, there is still room to make additional contributions to the


science and technology of carbon nanotubes and how to make films that have the properties that are needed for a given application.

REFERENCES

[1] S. Iijima, Helical Microtubes of Graphitic Carbon, Nature, 354, 1991, pp 56-58. [2] S. Iijima, T. Ichinashi, Single-Shell Carbon Nanotubes of 1-nm Diameter, Nature,

363, 1993, pp 603-605. [3] W. Zhou, L. Ding, S. Yang, J. Liu, Orthogonal Orientation Control of Carbon

Nanotube Growth, J. Am. Chem. Soc.,, 132, 2010, pp 336-341. [4] J.P. Edgeworth, N.R. Wilson, J.V. Macpherson, Controlled Growth and

Characterization of Two-Dimensional Single-Walled Carbon-Nanotube Networks for Electrical Applications, Small, 3(5), 2007, pp. 860-870.

[5] J. Kong, N.R. Franklin, C. Zhou, M.G. Chapline, S. Peng, K. Cho, H. Dai, Nanotube Molecular Wires as Chemical Sensors, Science, 287, 2000, 622-625.

[6] J.P. Novak, E.S. Snow, E.J. Houser, D. Park, J.L. Stepnowski, R.A. McGill, Nerve Agent Detection Using Networks of Single-Walled Carbon Nanotubes, Applied

Physics Letters, 83(19), 2003, 4026-4028. [7] E. Bekyarova, M. Davis, T. Burch, M.E. Itkis, B. Zhao, S. Sunshine, R.C. Haddon,

Chemically Functionalized Single-Walled Carbon Nanotubes as Ammonia Sensors, J.

Phys. Chem. B, 108, 2004, pp. 19717-19720. [8] L. He, Y. Jia, F. Meng, M. Li, J. Liu, Sensors for Ammonia Detection Based on

Polyaniline-Coated Multi-Wall Carbon Nanotubes, Materials Science and

Engineering B, 163, 2009, pp 76-81. [9] T. Zhang, S. Mubeen, E. Bekyraova, B.Y. Yoo, R.C. Haddon, N.V. Myung, M.A.

Deshusses, Poly(m-aminobenzene sulfonic acid) Functionalized Single-Walled Carbon Nanotubes Based Gas Sensor, Nanotechnology, 18(16), 2007, 165504.

[10] P. Qi, O. Vermesh, M. Grecu, A. Javey, Q. Wang, H. Dai, Toward Large Arrays of Multiplex Functionalized Carbon Nanotube Sensors for Highly Sensitive and Selective Molecular Detection, Nano Letters, 3(3), 2003, pp 347-351.

[11] Z. Li, P. Dharap, S. Nagarajajah, E.V. Barrera, J.D. Kim, Carbon Nanotube Film Sensors, Advanced Materials, 16(7), 2004, pp. 640-643.

[12] C.C. Hsu, R.M. Chao, C.W. Liu, S.Y. Liang, Evaluation of the gauge factor for single-walled carbon nanonets on the flexible plastic substrates by nano-transfer printing, J. Micromech. Microeng., 21, 2011, 075012.

[13] S. Kumar, B.A. Cole, R. Jackson, S. Graham, A Review of Carbon Nanotube Ensembles as Flexible Electronics and Advanced Packaging Materials, Journal of

Electronic Packaging, 133, 2011, 020906. [14] E.S. Snow, J.P. Novak, P.M. Campbell, D. Park, Random Networks of Carbon

Nanotubes as an Electronic Material, Applied Physics Letters, 82(13), 2003, pp. 2145-2147.

[15] S. Kumar, G.B. Blanchet, M.S. Hybertsen, J.Y. Murthy, M.A. Alam, Performance of carbon nanotube-dispersed thin-film transistors, Applied Physics Letters, 89(14), 2006, 143501.


[16] B. Chandra, H. Park, A. Maarof, Glenn J. Martyna, G.S. Tulevski, Carbon nanotube thin film transistors on flexible substrates, Applied Physics Letters, 99(7), 2011, 072110

[17] S. Liu, Q. Shen, Y. Cao, L. Gan, Z. Wang, M.L. Steigerwald, X. Guo, Chemical functionalization of single-walled carbon nanotube field-effect transistors as switches and sensors, Coordination Chemistry Reviews, 254, 2009, pp. 1101-1116.

[18] Q. Cao, H. Kim, N. Pimparkar, J. Kulkami, C. Wang, M. Shim, K. Roy, M.A. Alam, J.A. Rogers, Medium-scale carbon nanotube thin-film integrated circuits on flexible plastic substrates, Nature Letters, 454, 2008, pp. 495-500.

[19] K. Bradley, J.-C.P. Gabriel, G. Gruner, Flexible Nanotube Electronics, Nano Letters, 3(10), 2003, pp. 1353-1355.

[20] S.J. Kang, C. Kocabas, T. Ozel, M. Shim, N. Pimparkar, M.A. Alam, S.V. Rotkin, J.A. Rogers, High-performance electronics using dense, perfectly aligned arrays of single-walled carbon nanotubes, Nature Nanotechnology, 2, 2007, pp. 230-236.

[21] Javey, H. Kim, M. Brink, Q. Wang, A. Ural, J. Guo, P. McIntyre, P. McEuen, M. Lundstrom, H. Dai, High-K dielectrics for advanced carbon-nanotube transistors and logic gates, Nature Materials, 1, 2002, pp. 241-246.

[22] J.P. Novak, M.D. Lay, F.K. Perkins, E.S. Snow, Macroelectronic applications of carbon nanotube networks, Solid-State Electronics, 48, 2004, pp. 1753-1756.

[23] K.H. An, W.S. Kim, Y.S. Park, J.-M. Moon, D.J. Bae, S.C. Lim, Y.S. Lee, Y.H. Lee, Electrochemical Properties of High-Power Supercapacitors Using Single-Walled Carbon Nanotube Electrodes, Advanced Functional Materials, 11(5), 2001, pp. 387-392.

[24] C. Du, J. Yeh, N. Pan, High power density supercapacitors using locally aligned carbon nanotube electrodes, Nanotechnology, 16, 2005, pp. 350-353.

[25] Q. Cao, Z.-T. Zhu, M.G. Lemaitre, M.-G. Xia, M. Shim, J.A. Rodgers, Transparent flexible organic thin-film transistors that use printed single-walled carbon nanotube electrodes, Applied Physics Letters, 88(11), 2006, 113511.

[26] J.L. Blackburn, T.M. Barnes, M.C. Beard, Y.-H. Kim, R.C. Tenent, T.J. McDonald, B. To, T.J. Coutts, M.J. Heben, Transparent Conductive Single-Walled Carbon Nanotube Networks with Precisely Tunable Ratios of Semiconducting and Metallic Nanotubes, ACS Nano, 2(6), 2008, pp. 1266-1274.

[27] D.S. Hecht, L. Hu, G. Irvin, Emerging Transparent Electrodes Based on Thin Films of Carbon Nanotubes, Graphene, and Metallic Nanostructures, Advanced Materials, 23, 2011, pp. 1482-1513.

[28] J.B. Cui, R. Sordan, M. Burghard, K. Kern, Carbon nanotube memory devices of high charge storage stability, Applied Physics Letters, 81(17), 2002, pp. 3260-3262.

[29] E. Frackowiak, F. Beguin, Carbon materials for the electrochemical storage of energy in capacitors, Carbon, 39, 2001, pp. 937-950.

[30] K.H. An, W.S. Kim, Y.S. Park, Y.C. Choi, S.M. Lee, D.C. Chung, D.J. Bae, C. Lim, Y.H. Lee, Supercapacitors Using Single-Walled Carbon Nanotube Electrodes, Advanced Materials, 13(7), 2001, pp. 497-500

[31] C. Journet, W.K. Maser, P. Bernier, A. Loiseau, M. L. de la Chapelle, S. Lefrant, P. Deniard, R. Lee, J.E. Fischer, Large-scale production of single-walled carbon nanotubes by the electric-arc technique, Nature, 388, 1997, pp 756-758.


[32] Y. Ando, Z. Zhao, K. Hirahara, K. Suenaga, S. Bandow, S. Iijima, Mass production of single-wall carbon nanotubes by the arc plasma jet method, Chemical Physical

Letters, 323(5-6), 2000, pp. 580-585. [33] M. Terrones, Carbon nanotubes: synthesis and properties, electronic devices and other

emerging applications, International Materials Reviews, 49(6), 2004, pp. 325-377. [34] X. Zhao, M. Wang, M. Ohkohchi, Y. Ando, Morphology of Carbon Nanotubes

Prepared by Carbon Arc, Japanese Journal of Applied Physics, 35, 1996, pp. 4451-4456.

[35] X. Zhao, M. Ohkohchi, M. Wang, S. Iijima, T. Ichihashi, Y. Ando, Preparation of high-grade carbon nanotubes by hydrogen arc discharge, Carbon, 35(6), 1997, pp. 775-781.

[36] B. Chen, S. Inoue, Y. Ando, Raman spectroscopic and thermogravimetric studies of high-crystallinity SWNTs synthesized by FH-arc discharge method, Diamond and

Related Materials, 18(5-8), 2009, pp. 975-978. [37] W.K Maser, E. Munoz, M.T. Martinez, A.M. Benito, G.F. de la Fuente, Study of

parameters important for the growth of single walled carbon nanotubes, Opt. Mater. 2001, 17, pp. 331–334.

[38] A.A. Puretzky, H. Schittenhelm, X. Fan, M.J. Lance, L.F. Allard, Jr. D.B. Geohegan, Investigations of single-wall carbon nanotube growth by time-restricted laser vaporization, Physical Review B,65, 2002, 245425,

[39] W.K. Hsu, J.P. Hare, M. Terrones, H.W. Kroto, D.R.M. Walton, P.J.F. Harris, Condensed-phase nanotubes, Nature, 377, 1995, pp. 687.

[40] G.Z. Chen, I. Kinloch, M.S.P. Shaffer, D.J. Fray, A.H. Windle, Electrochemical investigation of the formation of carbon nanotubes in molten salt,. High Temperature

Material Processes, 2, 1998, pp. 459- 469. [41] W. Merchan-Merchan, A.V. Saveliev, L. Kennedy, W.C. Jimenez, Combustion

Synthesis of Carbon Nanotubes and Related Nanostructures, Progress in Energy and

Combustion Science, 36(6), 2010, pp. 696-727. [42] Y. Liu, N. Zheng, J. Huang, B. Sun, Synthesis of Carbon Nanotubes with typical

structure from the pyrolysis flame, Advanced Materials Research, 221, 2011, pp. 99-103.

[43] Y. Liu, Effect of helium on synthesis of carbon nanotubes from the V-type pyrolysis flame, Advanced Materials Research, 221, 2011, pp. 540-544.

[44] Y. Liu, B. Sun, Z.-Y. Ding, Effect of Hydrogen on the V-type pyrolysis flame synthesis of carbon nanotubes, Advanced Materials Research, 221, 2011, pp. 545-549.

[45] Y. Liu, B. Sun, Z.-Y. Ding, Influence analysis of sampling time for synthesis of carbon nanotubes in the V-type pyrolysis flame, Advanced Materials Research, 221, 2011, pp. 235-239.

[46] P. Nikolave, M.J. Bronikowski, R.K. Bradley, F. Rohmund, D.T. Colbert, K.A. Smith, R.E. Smalley, Gas-phase catalytic growth of single-walled carbon nanotubes from carbon monoxide, Chemical Physical Letters, 313(1-2), 1999, pp. 91-97.

[47] J. Prazek, J. Drbohlavova, J. Chomoucka, J. Hubalek, O. Jasek, V. Adam, R. Kizek, Methods for carbon nanotube synthesis- review, Journal of Materials Chemistry, online, 2011.


[48] O. Lee, J. Jung, S. Doo, S.-S. Kim, T.-H. Noh, K.-I. Kim, Y.-S. Lim, Effects of Temperature and Catalysts on the Synthesis of Carbon Nanotubes by Chemical Vapor Deposition, Met. Mater. Int, 16(4), 2010, pp. 663-667.

[49] R. Sharma, S.-W. Chee, A. Herzing, R. Miranda, P. Rez, Evaluation of the Role of Au in Improving Catalytic Activity of Ni Nanoparticles for the Formation of One-Dimensional Carbon Nanostructures, Nano Letters, 11, 2011, pp. 2464-2471.

[50] X.Z. Liao, A. Serquis, Q.X. Jia, D.E. Peterson, Y.T. Zhu, H.F. Xu, Effect of catalyst composition on carbon nanotube growth, Applied Physics Letters, 82(16), 2003, pp. 2694-2696.

[51] E.F. Kukovitsky, S.G. L’vov, N.A. Sainov, V.A Shustov, L.A. Chernozatonskii, Correlation between metal catalyst particle size and carbon nanotube growth, Chemical Physical Letters, 355, 2002, pp. 497-503.

[52] W.-Y. Lee, H. Lin, L. Gu, K.-C. Leou, C.-H. Tsai, CVD catalytic growth of single-walled carbon nanotubes with a selective diameter distribution, Diamond and Related

Materials, 17, 2008, pp. 66-71. [53] A.G Mamalis, L.O.G. Vogtlander, A. Markopoulos, Nanotechnology and

nanostructured materials: Trends in carbon nanotubes. Precis. Eng. 2004, 28, pp. 16–30.

[54] N. Ishigami, H. Ago, K. Imamoto, M. Tsuji, K. Iakoubovskii, N. Minami, Crystal Plane Dependent Growth of Aligned Single-Walled Carbon Nanotubes on Sapphire, J.

Am. Chem. Soc., 130(30), 2008, pp. 9918-9924. [55] S. Dumpala, J.B. Jasinski, G.U. Sumanasekera, M.K. Sunkara, Large area synthesis of

conical carbon nanotube arrays on graphite and tungsten foil substrates, Carbon, 49, 2011, pp. 2725-2734.

[56] R.L. Vander Wal, T.M. Ticich, V.E. Curtis, Substrate-support interactions in metal-catalyzed carbon nanofiber growth, Carbon, 39, 2001, pp. 2277-2289.

[57] S.S. Xie, B.H. Chang, W.Z. Li, Z.W. Pan, L.F. Sun, J.M. Mao, X.H. Chen, L.X. Qian, W.Y. Zhao, Synthesis and Characterization of Aligned Carbon Nanotube Arrays, Advanced Materials, 11(13), 1999, pp. 1135-1138.

[58] B. Shukla, T. Saito, S. Ohmori, M. Koshi, M. Yumura, S. Iijima, Interdependency of Gas Phase Intermediates and Chemical Vapor Deposition Growth of Single Wall Carbon Nanotubes, Chemistry of Materials, 22, 2010, pp. 6035-6043.

[59] K.B. Shelimov,R.O. Esenaliev, A.G. Rinzler, C.B. Huffman, R.E. Smalley, Purification of single-wall carbon nanotubes by ultrasonically assisted filtration, Chemical Physics Letters, 282, 1998, pp. 429-434.

[60] A.C. Dillon, T. Gennett, K.M. Jones, J.L. Alleman, P.A. Parilla, M.J. Heben, A. Simpel, A Simple and Complete Purification of Single-Walled Carbon Nanotube Materials, Adv. Mater., 11(16), 1999, pp. 1354-1358.

[61] V. Djordjevic, J. Djustebek, J. Cveticanin, S. Velicknovic, M. Verjjkovic, M. Bokorov, B.B. Stojic, O. Neskovic, Methods of purification and characterization of carbon nanotubes, Journal of Optoelectronics and Advanced Materials, 8(4), 2006, pp. 1631-1634.

[62] G. Zhang, P. Qi, X. Wang, Y. Lu, X. Li, R. Yu, S. Bangsaruntip, D. Mann, L. Zhang, H. Dai, Selective Etching of Metallic Carbon Nanotubes by Gas-Phase Reaction, Science, 314, 2006, pp. 974-977.


[63] M.S. Arnold, A.A. Green, J.F. Julvat, S.I. Stupp, M.C. Hersam, Sorting carbon nanotubes by electronic structure using density differentiation, Nature, 1, 2006, pp. 60-65.

[64] H.-Z. Geng, K.K. Kim, K.P. So, Y.S. Lee, Y. Chang, Y.H. Lee, Effect of Acid Treatment on Carbon Nanotube-Based Flexible Transparent Conducting Films, J. Am.

Chem. Soc., 129, 2007, pp. 7758-7759. [65] H. Dai, E.W. Wong, C.M. Lieber, Probing Electrical Transport in Nanomaterials:

Conductivity of Individual Carbon Nanotubes, Science, 272, 1996, pp. 523-526. [66] Rochefort, D.R. Salahub, P. Avouris, Effects of Finite Length on the Electronic

Structure of Carbon Nanotubes, J. Phys. Chem. B., 103, 1999, pp. 641-646. [67] P.J. de Pablo, C. Gomez-Navarro, P.A. Serena, J. Gomez-Herrero, Nonlinear

Resistance versus Length in Single-Walled Carbon Nanotubes, Physical Review

Letters, 88(3), 2002, 036804. [68] A.N. Andriotis, M. Menon, L. Chernozatonskii, Nonlinear Resistance Dependence on

Length in Single-Wall Carbon Nanotubes, Nano Letters, 3(2), 2003, pp. 131-134. [69] H.J. Qi, K.B.K. Teo, K.K.S. Lau, M.C. Boyce, W.I. Milne, J. Robertson, K.K.

Gleason, Determination of mechanical properties of carbon nanotubes and vertically aligned carbon nanotube forests using nanoindentation, Journal of the Mechanics and

Physics of Solids, 51, 2003, pp. 2213-2237. [70] X. Sun, S. Zaric, D. Daranciang, K. Welsher, Y. Lu, X. Li, H. Dai, Optical Properties

of Ultrashort Semiconducting Single-Walled Carbon Nanotube Capsules Down to Sub-10 nm, JACS, 130(20), 2008, pp. 6552-6555.

[71] S. Kumar, J.Y. Murthy, M.A. Alam, Percolating Conduction in Finite Nanotube Networks, Physical Review Letters, 95, 2005, 066802.

[72] Buldum, J.P. Lu, Contact Resistance between Carbon Nanotubes, Physical Review B, 63, 2001, 161403.

[73] M.S. Fuhrer, J. Nygard, L. Shih, M. Forero, Y.-G. Yoon, M.S.C. Mazzoni, H.J. Choi, J. Ihm, S.G. Louie, A. Zettl, P.L. McEuen, Crossed Nanotube Junctions, Science, 288, 2000, pp. 494-497.

[74] P.N. Nirmalraj, P.E. Lyons, S. De, J.N. Coleman, J.J. Boland, Electrical Connectivity in Single-Walled Carbon Nanotube Networks, Nano Letters, 9(11), 2009, pp. 3890-3895.

[75] D. Hecht, L. Hu, G. Gruner, Conductivity Scaling with Bundle Length and Diameter in Single Walled Carbon Nanotube Networks, Appl. Phys. Lett., 89, 2006, 133112.

[76] M.P. Garrett, I.N. Ivanov, R.A. Gerhardt, A.A. Puretzky, D.B. Geohegan, Separation of junction and bundle resistances in single wall carbon nanotube percolation networks by impedance spectroscopy, Appl. Phys. Lett., 97, 2010, 163105.

[77] J.L. Bahr, E.T. Mickelson, M.J. Bronikowski, R.E. Smalley, J.M. Tour, Dissolution of small diameter single-wall carbon nanotubes in organic solvents?, Chem Comm, 2, 2001, pp. 193-194.

[78] B.J. Landi, H.J. Ruf, J.J. Worman, R.P. Raffaelle, Effects of Alkyl Amide Solvents on the Dispersion of Single-Wall Carbon Nanotubes, J. Phys. Chem. B., 108, 2004, pp. 17089-17095.

[79] A.N.G. Parra-Vasquez, N. Behabtu, M.J. green, C.L. Pint, C.C. Young, J. Schmidt, E. Kesselman, A. Goyal, P.M. Ajayan, Spontaneous Dissolution of Ultralong Single- and Multiwalled Carbon Nanotubes, ACS Nano, 4(7), 2010, pp. 3969-3978.


[80] L. Vaisman, H.D. Wagner, G. Marom, The role of surfactants in dispersion of carbon nanotubes, Advances in Colloid and Interface Science, 128-130, 2006, pp. 37-46.

[81] R. Bandyopadhyaya, E. Nativ-Roth, O. Regev, R. Yerushalmi-Rozen, Stabilization of Individual Carbon Nanotubes in Aqueous Solutions, Nano Letters, 2(1), 2002, pp. 25-28.

[82] J.-Y. Shin, T. Premkumar, K.E. Geckeler, Dispersion of Single-Walled Carbon Nanotubes by Using Surfactants: Are the Type and Concentration Important?, Chemistry, A European Journal, 14, 2008, pp. 6044-6048.

[83] M.F. Islam, E. Rojas, D.M. Bergey, A.T. Johnson, A.G. Yodh, High Weight Fraction Surfactant Solubilzation of Single-Wall Carbon Nanotubes in Water, Nano Letters, 3(2), 2003, pp. 269-273.

[84] J.W Jo, J.W. Jung, J.U. Lee, W.H. Jo, Fabrication of Highly Conductive and Transparent Thin Films from Single-Walled Carbon Nanotubes Using a New Non-ionic Surfactant via Spin Coating, ACS Nano, 4(9), 2010, pp. 5382- 5388.

[85] C.J.Lee, D.W. Kim, T.J. Lee, Y.C. Choi, Y.S. Park, Y.H. Lee, W.B. Choi, N.S. Lee, G.-S. Park, J.M. Skim, Synthesis of aligned carbon nanotubes using thermal chemical vapor deposition, Chemical Physics Letters, 312, 1999, pp. 461-468.

[86] Y. Chen, D.T. Shaw, L. Guo, Field Emission of Different Oriented Carbon Nanotubes, Applied Physics Letters,76(17), 2000, pp. 2469-2471.

[87] C.M. Hsu, C.H. Lin, H.L. Chang, C.T. Kuo, Growth of large area horizontally-aligned carbon nanotubes by ECR-CVD, Thin Solid Films, 420-421, 2002, pp. 225-229.

[88] H.B. Peng, T.G. Ristroph, G.M. Schurmann, G.M. King, J. Yoon, V. Narayanamurti, J.A. Golovchenko, Patterned Growth of Single-Walled Carbon Nanotube arrays from a vapor deposited Fe catalyst, Applied Physics Letters, 83(20), 2003, pp. 4238-4240.

[89] B.Q. Wei, R. Vajtai, Y. Jung, J. Ward, R. Zhang, G. Ramanth, P.M. Ajayan, Organized assembly of carbon nanotubes, Nature, 416, 2002, pp. 495-496.

[90] W. Wongwiriyapan, S.-I. Honda, H. Konishi, T. Mizuta, T. Ikuno, T. Ito, T. Maekawa, K. Suzuki, H. Ishikawa, K. Oura, M. Katayama, Single-Walled Carbon Nanotube Thin-Film Sensor for Ultrasensitive Gas Detection, Japanese Journal of

Applied Physics, 44(16), 2005, pp. L482-L485. [91] E.J. Bae, Y.-S. Min, U.J. Kim, W. Park, Thin film transistors of single-walled carbon

nanotubes grown directly on glass substrates, Nanotechnology, 18, 2007, 495203. [92] X. Xu, G.R. Brandes, A method for fabricating large-area, patterned, carbon nanotube

field emitters, Applied Physics Letters,74(17), 1999, pp. 2549-2551. [93] L.Hu, D.S. Hecht, G. Gruner, Carbon Nanotube Thin Films: Fabrication, Properties,

and Applications, Chem. Rev. 110, 2010, pp. 5790-5844. [94] A.R. Boccaccini, J. Cho, J.A. Roether, B.J.C. Thomas, E.J. Minay, M.S.P. Shaffer,

Electrophoretic Deposition of Carbon Nanotubes, Carbon, 44(15), 2006, pp. 3149-3160.

[95] T. Ferri, D. Frasca, O.A. de Fuentes, R. Santucci, M. Frasconi, Spatially Oriented and Reversible Surface Assembly of Single-Walled Carbon Nanotubes: A Strategy Based on π-π Interactions, Angewandte Chemie Internation Edition, 50(31), 2011, pp. 7074-7078.

[96] Y. Kim, N. Minami, W. Zhu, S. Kazaoui, R. Azumi, M. Matsumoto, Langmuir-Blodgett Films of Single-Wall Carbon Nanotubes: Layer-by-layer Deposition and In-plane Orientation of Tubes, Jpn. J. Appl. Phys, 42, 2003, pp. 7629-7634.


[97] H. Ko, V.V. Tsukruk Liquid-Crystalline Processing of Highly Oriented Carbon Nanotube Arrays for Thin-Film Transistors, Nano. Lett, 6(7), 2006, pp. 1443-1448.

[98] N. Saran, K. Parikh, D.-S. Suh, E. Munoz, H. Kolla, S.K. Manohar, Fabrication and Characterization of Thin Films of Single-Walled Carbon Nanotube Bundles on Flexible Plastic Substrates, J. Am. Chem. Soc., 126, 2004, pp. 4462-4463.

[99] E. Bekyarova, M.E. Itkis, N. Cabrera, B. Zhao, A. Yu, J. Gao, R.C. Haddon, Electronic Properties of Single-Walled Carbon Nanotube Networks, J. Am. Chem.

Soc., 127, 2005, pp. 5990-5995. [100] S. Manivanan, J.H. Ryu, H.E. Lim, M. Nakamoto, J. Jang, K.C. Park, Properties of

Surface Treated Transparent Conducting Single Walled Carbon Nanotube Films, J.

Mater. Sci.: Mater. Electron., 21, 2010, pp. 72-77. [101] K. L. White, M. Shuai, X. Zhang, H.-J. Sue, R. Nishimura, Electrical Conductivity of

Well-Exfoliated Single-Walled Carbon Nanotubes, Carbon, 49(15), 2011, pp. 5124-5131.

[102] L. Hu, G. Gruner, J. Jenkins, C.-J. Kim, Flash dry deposition of nanoscale material thin films, Journal of Materials Chemistry, 19, 2009, pp. 5845-5849.

[103] O.O. Van der Biest, L.J. Vandeperre, Electrophoretic Deposition of Materials, Annu.

Rev. Mater. Sci., 29, 1999, pp. 327-352. [104] B. Gao, G.Z. Yue, Q. Qiu, Y. Cheng, H. Shimoda, L. Fleming, O. Zhao, Fabrication

and Electron Field Emission Properties of Carbon Nanotube Films by Electrophoretic Depoision, Adv. Mater., 13(23), 2001, pp. 1770-1773.

[105] C. Du, D. Heldbrant, N. Pan, Preparation and preliminary property study of carbon nanotube films by electrophoretic deposition, Materials Letters, 57, 2002, pp. 434-438.

[106] S. Banerjee, B.E. White, L. Huang, B.J. Rego, S. O’Brien, I.P. Herman, Precise Positioning of single-walled Carbon nanotubes by AC dielectrophoresis, J. Vac. Sci.

Technol. B., 24(6), 2006, pp. 3173-3178. [107] D.A. Kurnosov, A.S. Baturin, A.S. Bugaev, K.N. Nikolski, R.G. Tchesov, E.P.

Shesin, Influence of the interelectrode distance in electrophoretic cold cathode fabrication on the emission uniformity, Applied Surface Science, 215, 2003, pp. 232-236.

[108] B.J.C. Thomas, A.R. Boccacini, M.S.P. Shafer, Multi-walled Carbon Nanotube Coatings Using Electrophoretic Deposition (EPD), J. Am. Ceram. Soc., 88(4), 2005, pp. 980-982.

[109] M.D. Lima, M.J. De Andrade, C.P. Bergmann, S. Roth, Thin, conductive, carbon nanotube networks over transparents substrates by electrophoretic deposition, Journal

of Materials Chemistry, 18, 2006, pp. 776-779. [110] M. Adabi, R. Saber, M. Adabi, S. Sar,ar, Examination of incubation time of bare gold

electrode inside cysteamine solution for immobilization of muti-walled carbon nanotubes on a gold electrode modified with cysteamine, Microchim Acta, 172, 2011, pp. 83-88.

[111] Z. Liu, Z. Shen, T. Zhu, S. Hou, L. Ying, Z. Shi, Z. Gu, Organizing Single-Walled Carbon Nanotubes on Gold Using a Wet Chemical Self-Assembling Technique, Langmuir, 16(8), 2000, pp. 3659-3573.


[112] Q. Wang, H. Moriyama, [63]-Fullerene and Single-Walled Carbon Nanotube-Based Ultrathin Films Stepwise Grafted onto a Self-Assembled Monolayer on ITO, Langmuir, 25(18), 2009, pp. 10834-10842.

[113] T. Ferri, D. Frasca, O.A. de Fuentes, R. Santucci, M. Frasconi, Spatially Oriented and Reversible Surface Assembly of Single-Walled Carbon Nanotubes: A Strategy Based on π-π Interactions, Angew. Chem., 123, 2011, pp. 7212-7216

[114] V. Krstic, G.S. Duesberg, J. Muster, M. Burghard, S. Roth, Langmuir-Blodgett Films of Matrix-Diluted Single-Walled Carbon Nanotubes, Chem. Mater., 10, 1998, pp. 2338-2340.

[115] Y. Guo, N. Minami, S. Kazaoui, J. Peng, M. Yoshida, Y. Miyashita, Multi-layer LB films of single-wall carbon nanotubes, Physica B, 323, 2002, pp. 235-236.

[116] C. Venet, C. Pearson, A.S. Jombert, M.F. Mabrook, D.A. Zeze, M.C. Petty, The morphology and electrical conductivity of single-wall carbon nanotube thin films prepared by the Langmuir-Blodgett technique, Colloids and Surfaces A: Physiochem.

Eng. Aspects, 354, 2010, pp. 113-117. [117] J. Li, Y. Zhang, Langmuir-Blodgett films of single-walled carbon nanotubes, Carbon,

45, 2007, pp. 493-498. [118] R. Sharma, C.Y. Lee, J.H. Choi, K. Chen, M.S. Strano, Nanometer Positioning,

Parallel Alignment, and Placement of Single Anisotropic Nanoparticles Using Hydrodynamic Forces in Cylindrical Droplets, Nano Letters,7(9), 2007, pp. 2693-2700.

[119] V.V. Tsukruk, H. Ko, S. Peleshanko, Nanotube Surface Arrays: Weaving, Bending, and Assembling on Patterned Silicon, Physical Review Letters, 92(6), 2004, 065502.

[120] Q. Zhang, P. Vichchulada, M.A. Cauble, M.D. Lay, Percolation in networks of aligned SWNTs formed with laminar flow deposition, J. Mater. Sci, 44, 2009, pp. 1206-1211.

[121] H. Shimoda, L. Fleming, K. Horton, O. Zhou, Formation of Macroscopically Ordered Carbon Nanotube Membranes by Self-Assembly, Physica B, 323, 2002, pp. 135-136.

[122] L. Huang, X. Cui, G. Dukovic, S.P. O’Brien, Self-organizing high-density single-walled carbon nanotube arrays from surfactant suspensions, Nanotechnology, 15, 2004, pp. 1450-1454.

[123] Rahy, P. Bajaj, I.H. Musselman, S.H. Hong, Y.-P. Sun, D.J. Yang, Coating of carbon nanotubes on flexible substrate and its adhesion study, Applied Surface Science, 255, 2009, pp. 7084-7089.

[124] E.Y. Jang, D.K. Seo, S. Jung, T. Kim, T.J. Kang, Y.H. Kim, Electrical resistance variation of carbon-nanotube networks due to surface modification of glass substrate, Phys. Status Solidi A, 207(8), 2010, pp. 1912-1917.

[125] M Palumbo, K.U. Lee, B.T. Ahn, A. Suri, K.S. Coleman, D. Zeze, D. Wood, C. Pearson, M.C. Petty, Electrical investigations of layer-by-layer films of carbon nanotubes, J. Phys. D: Appl. Phys, 39, 2006, pp. 3077-3085.

[126] Schindler, J. Brill, N. Furehauf, J.P. Novak, Z. Yaniv, Solution-depositied carbon nanotube layers for flexible display applications, Physica E, 37, 2007, pp. 119-123.

[127] L. Hu, G. Gruner, J. Gong, C.-K. Kim, B. Jornbostel, Electrowetting devices with transparent single-walled carbon nanotube electrodes, Applied Physics Letters, 90, 2007, 0903124.


[128] Y. Il Song, C.-M. Yang, D.Y. Kim, H. Kanoh, K. Kaneko, Flexible Transparent Conducting Single-Wall Carbon Nanotube Film with Network Bridging Method, Journal of Colloid and Interface Science, 318, 2008, pp. 365-371.

[129] D.W. Schubert, T. Dunkel, Spin coating from a molecular point of view: its concentration regimes, influence of molar mass and distribution, Mat. Res. Innovat, 7, 2003, pp. 314-321.

[130] S. Manivanan, J.H. Ryu, J. Jang, K.C. Park, Fabrication and effect of post treatment on flexible single-walled carbon nanotube films, J Mater Sci: Mater Elecron, 21, 2010, pp. 595-602.

[131] M.C. LeMieux, M. Roberts, S. Barman, Y.W. Jin, J.M. Kim, Z. Bao, Self-Sorted, Aligned Nanotube Networks for Thin-Film Transistors, Science, 321, 2008, pp. 101-104.

[132] K. Kordas, T. Mustonen, G. Toth, H. Jantunen, M. Lajunen, C. Soldano, S. Talapatra, S. Kar, R. Vajtai, P.M. Ajayan, Inkjet Printing of Electrically Conductive Patterns of Carbon Nanotubes, Small, 2, 2006, pp. 1021-1025.

[133] P. Beecher, P. Servati, A. Rozhin, A. Colli, V. Scardaci, S. Pisana, T. Hasan, A.J. Flewitt, F.M. Li, A. Nathan, A.C. Ferrari, W.I. Milne, Ink-jet printing of carbon nanotube thin film transistors, Journal of Applied Physics, 102, 2007, pp. 043710-1-7.

[134] J.-W. Song, J. Kim, Y.-H. Yoon, B.-S. Choi, J.-H. Kim, C.-S. Han, Inkjet printing of single-walled carbon nanotubes and electrical characterization of the line pattern, Nanotechnology, 19, 2008, 095702.

[135] B. Dan, G.C. Irvin, M. Pasquali, Continuous and Scalable Fabrication of Transparent Conducting Carbon Nanotube Films, ACS Nano, 3(4), 2009, pp. 835-843.

[136] W. A. de Heer, W.S. Basca, A. Chatelain, T. Gerfin, R. Humphrey-Baker, L. Forro, D. Ugarte, Aligned Carbon Nanotube Films: Production and Optical and Electronic Properties, Science, 268, 1995, pp. 845-847.

[137] J.-H. Shin, D.W. Shin, S.P. Patole, J.H. Lee, S.M. Park, J.B. Yoo, Smooth, transparent, conducting and flexible SWCNT films by filtration-wet transfer processes, J. Phys. D.: Appl. Phys., 42, 2009, 045305.

[138] F. Hennrich, S. Lebedkin, S. Malik, J. Tracy, M. Barczewski, H. Rosner, M. Kappes, Preparation, characterization and applications of free-standing single walled carbon nanotube thin films, Phys. Chem. Chem. Phys., 4, 2002, pp. 2273-2277.

[139] F.N. Ishikawa, H.-K. Chang, K. Ryu, P.-C. Chen, A. Badmaev, L.G. de Arco, G. Shen, C. Zhao, Transparent Electronics Based on Transfer Printed Aligned Carbon Nanotubes on Rigid and Flexible Substrates, ACS Nano, 3(1), 2009, pp. 73-79.

[140] J.C. McDonald, D.C. Duffy, J.R. Anderson, D.T. Chiu, H. Wu, O.J.A. Schuller, G.M. Whitesides, Fabrication of microfluidic systems in poly(dimethylsiloxane), Electrophoresis, 21(1), 2000, pp. 27-40.

[141] M.A. Meitl, Y. Zhao, A. Gaur, S. Jeon, M.L. Usrey, M.S. Strano, J.A. Rogers, Solution Casting and Transfer Printing Single-Walled Carbon Nanotube Films, Nano

Letters, 4(9), 2004, pp. 1643-1647. [142] Y. Zhou, L. Hu, G. Gruner, A Method of Printing Carbon Nanotube Thin Films,

Applied Physics Letters, 88, 2006, 123109. [143] D. Zhang, K. Ryu, X. Liu, E. Polikarpov, J. Ly, M.E. Tompson, C. Zhou, Transparent,

Conductive, and Flexible Carbon Nanotube Films and Their Application in Organic Light-Emitting Diodes, Nano Letters, 6(9), 2006, pp. 1880-1886.


[144] S.-W. Choi, W.-S. Kang, J.-H. Lee, C.K. Najeeb, H.-S. Chun, J.-H. Kim, Patterning of Hierarchically Aligned Single-Walled Carbon Nanotube Langmuir-Blodgett Films by Microcontact Printing, Langmuir, 26(19), 2010, pp. 15680-15685.

[145] C. Lim, D.-H. Min, S.-B. Lee, Direct patterning of carbon nanotube network devices by selective vacuum filtration, Applied Physics Letters, 91, 2007, 243117.

[146] S.-K Chang-Jian, J.-R. Ho, J.-W.J. Checng, C.-K. Sung, Fabrication of carbon nanotube field emission cathodes in patterns by laser transfer method, Nanotechnology, 17, 2006, pp. 1184-1187.

[147] E. Sundan, J.K. Moon, C.P. Wong, W.P. King, S. Graham, Microwave assisted patterning of vertically aligned carbon nanotubes onto polymer substrates, J. Vac. Sci.

Technol. B., 24(4), 2006, pp. 1947-1950. [148] M. Endo, H. Muramatsu, T. Hayashi, Y.A. Kim, M. Terrones, M.S. Dresselhaus,

‘Buckypaper’ from coaxial nanotubes, Nature, 433, 2005, pp. 476. [149] Y.A. Kim, K. Muramatsu, T. Kayashi, M. Endo, M. Terrones, M.S. Dresselhaus,

Fabrication of High-Purity, Double-Walled Carbon Nanotube Buckypaper, Chem.

Vap. Deposition, 12, 2006, pp. 327-330. [150] P. G. Whitten, G.M. Spinks, G.G. Wallace, Mechanical properties of carbon nanotube

paper in ionic liquid and aqueous electrolytes, Carbon, 43, 2005, pp. 1891-1896. [151] M.A. Poggi, P.T. Lillehei, L.A. Bottomley, Chemical Force Microscopy on Single-

Walled Carbon Nanotube Paper, Chem. Mater., 17, 2005, pp. 4289-4295. [152] G. Xu, Q. Zhang, W. Zhou, J. Huang, F. Wei, The Feasibility of Producing MWCNT

paper and strong MWNCNT film from VACNT array, Appl. Phys. A., 92, 2008, pp. 531-539.

[153] T, Gong, Y. Zhang, W.J. Liu, J.Q. Wei, C.G. Li, K.L. Wang, D.H. Wu, M.L. Zhong, A Connection of macro-sized double-walled carbon nanotube strands by bandaging with double-walled carbon nanotube films, Carbon, 45(11), 2007, pp. 2235-2240.

[154] L. Hu, J.W. Choi, Y. Yang, S. Jeong, F. La Mantla, L.-F. Cui, Y. Cui, Highly conductive paper for energy storage devices, PNAS, 106, 2009, pp. 21490-21494.

[155] L. Huang, K. Chen, C. Peng, R.A. Gerhardt, Highly conductive paper fabricated with multiwalled carbon nanotubes and poly(3,4-ethylenedioxythiophene)-poly(styrenesulfonate) by unidirectional drying, J. Mater. Sci., 46, 2011, pp. 6648-6655.

[156] S.M. Lyth, S.R.P. Silva, Field emission from multiwall carbon nanotubes on paper substrates, Applied Physics Letters, 90, 2007, 173124.

[157] R.L. Muhlbauer, R.A. Gerhardt, Understanding the Effect of Pore Size and Drying Technique on the Electrical Resistance of Multiwalled Carbon Nanotube Thin Films on Paper Substrates, unprinted, 2011.

a review on the synthesis of carbon nanotube thin...

Documents