Locking Carbon Nanotubes in Confined Lattice Geometries − A Route

ARTICLE pubs.acs.org/JPCB

Locking Carbon Nanotubes in Confined Lattice Geometries
A Route to Low Percolation in Conducting Composites Izabela Jurewicz,* Patnarin Worajittiphon, Alice A. K. King, Paul J. Sellin, Joseph L. Keddie, and Alan B. Dalton* Department of Physics, Faculty of Engineering & Physical Sciences, University of Surrey, Guildford GU2 7XH, United Kingdom

bS Supporting Information ABSTRACT: A significant reduction in the electrical percolation threshold is achieved by locking carbon nanotubes (CNTs) in a predominantly hexagonally close-packed (HCP) colloidal crystal lattice of partially plasticized latex particles. Contrary to other widely used latex processing where CNTs are randomly distributed within the latex matrix, for the first time, we show that excluding CNTs from occupying the interior volume of the latex particles promotes the formation of a nonrandom segregated network. The electrical percolation threshold is four times lower in an ordered segregated network made with colloidal particles near their glass transition temperature (Tg) in comparison to in a random network made with particles at a temperature well above the Tg. This method allows for a highly reproducible way to fabricate robust, stretchable, and electrically conducting thin films with significantly improved transparency and lattice percolation at a very low CNT inclusion which may find applications in flexible and stretchable electronics as well as other stretchable technologies. For instance, our technology is particularly apt for touch screen applications, where one needs homogeneous distribution of the conductive filler throughout the matrix.

1. INTRODUCTION Over the past decade CNTs have fascinated the scientific and industrial communities due to their outstanding electrical and mechanical properties.1
3 Such characteristics open up numerous potential applications particularly in electronics, sensors,4 antistatic packaging, conductive thin films, and coatings.5,6 However, the development of CNT-reinforced composites has been mostly impeded by the difficulty in dispersion within the polymer matrix and by a need for high inclusion by weight/ volume to promote percolating network formation. This, in turn, limits the mechanical performance of the composite as well as decreases the transparency in the visible spectral range. Generally speaking, it is well-known that depending on composite fabrication methods and the nature of constituents, random or segregated composite microstructures can be produced. In the case of CNTs or any other conductive particles distributed in a polymer matrix in a random fashion, there is a threshold concentration, called the percolation threshold, at which electrical current can flow through the matrix via interconnected filler particles.7 The percolation transition is governed by the now well-established power-law of percolation theory8
11 as in eq 1 σ ¼ σ0 ðφ
φc Þt for φ > φc

ð1Þ

where σ is the conductivity of the composite, σ0 is a proportionality constant (which usually resembles the intrinsic conductivity r 2011 American Chemical Society

of the filler), t is a critical exponent, whereas φ is the filler and φc is the percolation threshold concentrations, respectively Since Kusy first formalized the term “segregated network” to explain the effect of “particle size ratio on the continuity of aggregates”,12 new methods to produce high transparency conductive materials with reduced percolation thresholds have been developed. By restricting the volume that the filler can occupy, thus leading to a segregated spatial distribution of the percolating phase,13 significant reduction of the electrical percolation threshold can be accomplished. As a result, several examples wherein fillers such as carbon black14
19 and metallic powder20,21 were incorporated at the interfaces of a multicomponent blend by hot compaction or sintering have been shown. Another approach to the segregated network concept is to use latex processing.22 Aqueous film-forming latex dispersions are environmentally friendly (i.e., nonemitting of organic solvents), easy to process, and are widely used for adhesive, latex coatings, and laminates.22 The segregated network approach has been used by Grunlan and co-workers previously.19,23
25 They incorporated carbon-black into a polydisperse poly(vinyl acetate) (PVAc) latex where the filler occupies only the space between polymer particles and obtained electrical percolation thresholds Received: December 17, 2010 Revised: February 24, 2011 Published: March 24, 2011 6395

dx.doi.org/10.1021/jp111998p | J. Phys. Chem. B 2011, 115, 6395–6400

The Journal of Physical Chemistry B

ARTICLE

Table 1. Composition and Properties of Pristine Latex monomer composition

solids contentc [wt.%]

particle sized [nm]

hard latex

BA:MMA: MAA:AAEMa

49.5

270

28

soft latex

2EHA:BA: EA:MMAb

60

190


50

Tge [°C]

a

Butyl acrylate:methyl methacrylate:methacrylic acid:acetoacetoxyethyl methacrylate. b 2-Ethyl hexyl acrylate:butyl acrylate:ethyl acrylate:methyl methacrylate. c Determined gravimetrically. d Measured by dynamic light scattering. e Experimentally measured by DSC.

of 2.5 vol.%. In later work they also investigated the electrical conductivity and mechanical properties of monodispersed PVAc latex, compared to the polydispersed case, as a function of drying temperature.26 They found that the coalescence of the polymer particles is significantly improved at elevated temperature, and as a result the film porosity is greatly reduced. However, both systems exhibit increases in the percolation threshold with increasing temperature of drying due to breakage in the segregated network pathways. Much of the past work has centered on replacing spherical fillers with high aspect ratio rods in the hope of further reducing the filler content required to reduce the percolation threshold. Grossiord et al.27 incorporated SDS-stabilized SWNTs into polystyrene (PS) by hot compaction and obtained electrical percolation thresholds as low as 0.3 wt.%. Regev and co-workers28 used the same system and compared it to Gum Arabic (GA) stabilized SWNT-PS composite. They obtained electrically percolating networks at 0.28 wt.% in the SWNT-PS-SDS system. Consequently, by replacing SDS dispersant with GA, an upward shift in the percolation threshold was observed due to the electrostatic and steric repulsions between the polymer particles and GA-dispersed nanotubes, both negatively charged. In their study of a latex with a low glass transition temperature (Tg), Wang et al. reported an electrical percolation threshold of 0.3 wt.% functionalized CNTs.29 These examples are representations of composites with random nonorganized segregated network microstructures. The polymer matrix acts as an excluded volume in which the conducting filler cannot be present. Although our findings are built on the work of others who used the random segregated network concept to manipulate the percolation threshold, we show that using a colloidal crystal lattice of partially plasticized latex to assemble CNTs in a highly controlled manner is a very reproducible and up-scalable way to obtain highly transparent conducting and stretchable thin films at much lower CNT inclusion.

2. EXPERIMENTAL DETAILS As-received SWNTs made by a high-pressure carbon monoxide process (HiPco) purchased from Carbon Nanotechnologies Inc. were dispersed in deionized water, containing 1 wt.% of nonionic surfactant Triton X-100 (Union Carbide Corp.) resulting in a CNT concentration in water of 1 mg mL
1. The dispersion was sonicated using a probe ultrasonicator (Branson Sonifier model 150D) for 10 min at a power of 20 W to produce surfactant-stabilized dispersion of CNTs. Two monodispersed, polymer latex samples (called here “hard” and “soft”) prepared by emulsion polymerization provided by DSM NeoResins (Waalwijk, The Netherlands) and Cytec Surface Specialties, respectively, were used and are summarized in Table 1. Depending on the desired nanotube concentration in the final composite mixture (ranging from 0.01 wt.% to 20 wt.%), both latex dispersions (“soft” and “hard”) were combined with

aqueous CNT-surfactant solution. All samples were then sonicated in an ice-cold water bath for 10 min using both the continuous and discontinuous modes to ensure good mixing. Calculation of the amount of CNT in weight percent was based on the assumption that all CNTs are in solution and on the basis of the weight of the latex solids content. Electrical properties of composites were investigated as a function of CNT concentration. A specially designed, gold-plated ceramic substrate consisting of a large gold electrode in the middle and four smaller contacts (one on each side) was used. For measurements, polymer films were prepared by casting the composite dispersion using a metallic cube applicator with a nominal 150 μm gap width onto a ceramic substrate and dried for 24 h at room temperature. The gradual evaporation of water brings polymer particles close to each other resulting in the locking of CNTs in interstitial voids of the resulting latex crystal. After drying, two 10 nm thick top gold electrodes of 3.5  3.5 mm area each were prepared by thermo-evaporation of gold using an Edwards Evaporator (see Supporting Information
Figure S1). Dark current
voltage characteristics have been obtained using a Keithley 487 picoammeter/voltage source. The final specific conductivity was calculated from the resistance and thickness of the film. All sample thicknesses were measured using a profilometer (Veeco, Dektak 8). For topographic studies an atomic force microscope (NT-MDT, Moscow, Russia), using semicontact mode, was employed. The SEM investigations of composite materials were done using a Hitachi S-4000 scanning electron microscope at an accelerating voltage of 15 kV. Samples were imaged without sputtering a metal onto their surface. The Tg was determined using a differential scanning calorimeter (DSC) (TA Instruments Q1000, New Castle, USA). DSC thermographs of heat flow as a function of temperature were typically obtained in the temperature range from
80 to 80 °C. A standard heating rate of 10 °C/min and cooling rate of 20 °C/min were used for all samples.

3. RESULTS AND DISCUSSION We have focused our studies on two latex-based composites that represent segregated and random types of microstructures. In the former, SWNTs are dispersed in water by means of a nonionic surfactant, Triton X-100, and mixed with a so-called “hard” latex that has a Tg of 28 °C (so that it is in the glassy state at room temperature) and particle size of 280 nm; and in the latter, a so-called “soft” latex with a Tg of
50 °C and particle size of 190 nm is used. The resulting composites are deposited as wet coatings on a substrate and allowed to film form as shown in Figure 1. The physics of such film formation is now well understood. The initial stages for both systems are the same. For both hard and soft latex, as water evaporates, individual latex particles generally self-assemble into hexagonally close-packed (HCP) or face-centered cubic (FCC) crystals in which the spheres 6396

dx.doi.org/10.1021/jp111998p |J. Phys. Chem. B 2011, 115, 6395–6400

The Journal of Physical Chemistry B

ARTICLE

Figure 1. A schematic representation of the formation of segregated and random conductive networks. The composite dispersion is deposited onto a substrate. The evaporation of water brings individual particles into contact to form a colloidal crystalline phase. CNTs are locked in the crystal lattice of the plasticized latex particles. Particles are progressively deformed to fill the voids left by water evaporation and take on the shape of rhombic dodecahedra in a segregrated network. In the soft latex composite, polymer diffusion across particle
particle boundaries takes place. CNTs partially diffuse into the latex particles to form a continuous and random network.

occupy 74% of the volume. It should be pointed out that one of the issues with the evaporation-driven template-assisted selfassembly of latex particles is the formation of dislocations or particle vacancies during film formation. Thus the crystal structure also comprises a number of defects. For the next stage of film formation, in the case of the “hard” latex based composite, the close-packed spheres deform into rhombic dodecahedra but do not fully coalesce due to the relatively low temperature of drying in comparison to the polymer’s Tg. As has been shown elsewhere,30 the organization of CNTs at the interface of polymer particles is adjusted via the plasticization of the latex particle shell and allows for improved interfacial bonding between CNTs and the matrix, retarding phase separation. During water evaporation, CNTs tend to be pushed away from the first contact area of touching polymer particles. Consequently, as the polymer particles deform into the rhombic dodecahedra, CNTs can only be positioned at the edges of the resulting well-ordered 3D array of submicrometer size latex particle array as shown in Figure 2a. Such templateassisted self-assembly leads to the formation of a highly ordered honeycomb-like segregated network of CNTs (Figure 2b). As is evident, the ordering is not perfectly HCP/FCC in nature. Due to the presence of defects and dislocations in the latex crystal, occasionally 4-fold symmetry organization is observed, where each particle has only four nearest neighbors in the surface plane (body-centered cubic-BCC) instead of six as shown in Figure 2c. However, as the number of dislocations is very small and their distribution wide, they do not significantly affect the overall packing density of polymer particles. Even so it is interesting to note that regardless of particle ordering, CNT assembly is driven by local particle-packing, and as a consequence it is possible to see CNTs positioned at BCC lattice sites in certain places. Figure 2d,e shows the coexistence of the two types of domains. No such ordering is evident for the soft composite. In contrast, as the film formation process consists of an additional stage involving the diffusion across particle
particle boundaries resulting in full particle coalescence, it is possible for the CNTs to partially diffuse into the latex particles forming a continuous and homogeneous composite material as schematically represented in Figure 1 and shown in Figure 2f,g. The percolation behavior can be effectively altered by changing the topological state of the microstructure. Therefore, in this work we compare two types of systems
a random versus a

nonrandom segregated microstructure. In the composite system comprised of randomly distributed, insulating polymer particles and conducting CNTs, as in the case of the soft latex composite, the electrical percolation can be explained by a continuum percolation theory.31 As the concentration of CNTs increases, the conducting cluster grows in size to form at least one continuous cluster that connects the opposite ends of the matrix for electrical current flow. In the hard latex composite representing the nonrandom segregated structure, however, the configuration of particle boundaries dramatically influences the passage of current due to the crystallographic constraints on the network topology. Therefore, it is more relevant to discuss the percolation behavior in terms of a classical lattice bond percolation model, where it has been shown theoretically that the percolation threshold is strongly dependent on the lattice type (square, triangular, cubic, BCC, FCC).32 Because, in real systems, one must also consider other parameters such as filler aspect ratio, filler/matrix interactions, and the intrinsic electrical properties of the filler, a theoretical analysis using a bond percolation theory is beyond the scope of this paper. The electrical conductivity measurements were performed on both types of composites over a wide range of CNT concentrations. Both types of composites exhibit a significant conductivity enhancement at specific CNT loadings, indicating the formation of a percolating CNT network. By fitting the measured conductivity data with a power law (eq 1), it is found that the percolation threshold, φc, of the hard latex composite appears at approximately 0.12 wt.%, while for the soft latex composite it is almost four times higher and occurs at 0.4 wt.% (Figure 3). In order to investigate the dimensionality of the conductive nanotube network, a standard least-squares linear regression of log-transformed data was performed. Because the density of SWNTs can only be approximated, the mass fraction is used rather than the volume fraction for fitting the experimental data. Values of the critical exponent, t, for the hard and soft latex composites were obtained as 1.19 and 1.65, respectively. According to the literature, a theoretical value for the critical exponent is t = 2.0 for a three-dimensional percolating network33,34 and t = 1
1.33 for a two-dimensional system.8 Thus it is seen that the value of t for the hard composite in the present work falls in the range of the theoretical predictions for a two-dimensional network. A similarly low critical exponent has been reported 6397

dx.doi.org/10.1021/jp111998p |J. Phys. Chem. B 2011, 115, 6395–6400

The Journal of Physical Chemistry B

ARTICLE

Figure 3. (a) Electrical conductivity of soft latex composite (red triangles) and hard latex composite (blue squares) as a function of CNT weight fraction (defined on the mass of the polymer). SEM micrographs of the hard latex composite containing (b) 0.12 wt.%, (c) 0.5 wt.%, and (d) 1 wt.% of CNTs.

Figure 2. (a) AFM topography (b) and corresponding low magnification SEM micrograph of the surface of a hard latex composite showing highly ordered latex particles with CNT bundles present at interstitial sites assembled into a hexagonal pattern. Dislocations and cracks are also observed. (c) AFM height image of the same composite showing the coexistence of FCC and BCC domains (d,e) corresponding SEM micrographs showing assembly of CNTs being driven by the particles’ arrangement. AFM (f) height and (g) phase images of soft latex composite showing uniform and random distribution of CNTs within the film.

previously in the literature.19,35
39 For example, Gubbels et al.35 investigated the conductive network dimensionality based on the selective localization of CB particles in multiphase polymeric materials. By comparing three systems: where CB is a) incorporated into a polyethylene (PE) matrix
amorphous system, b) dispersed in the PE phase in the blend of PE and polystyrene (PS), and c) at the interface of a PE/PS blend
segregated system, they observed a decrease in the critical exponent value from 2.0 to 1.3 as the conditions for CB localization changed from amorphous through an intermediate situation to the segregated state. Similarly, the CNTs in the hard latex composite are located in the interstitial spaces between the polymer particles, and the associated low value of the critical exponent clearly suggests that

a 2D conductive network is formed and that the charge must flow along surfaces and interfaces.34,40 In contrast, the value of the critical exponent of the soft latex composite is higher (t = 1.65) and is thus closer to a 3D conducting percolating network’s theoretical value. This is because CNTs partially diffuse into the polymer matrix during the film formation process to create a nonsegregated structure. Our explanation is drawn from AFM studies, where clear boundaries between polymer particles are not observed (Figure 2f,g). The reduction of the percolation threshold by a factor of 4 resulting from the incorporation of CNTs into a colloidal crystal lattice can be simply explained by hard sphere packing theory. In an ideal situation monosized hard spheres fill 74 vol.% of space when packed in an FCC crystal. The interstitial volume that CNTs can occupy is therefore fixed at 26 vol.%, which is approximately four times less than the 100% of volume available for CNTs in a random composite system. This marked reduction in the fractional free volume means appreciably fewer CNTs are required to form the first conductive percolation pathway (Figure 3b). Thus the electrical percolation threshold is also approximately reduced by a factor of 4. Following the initial formation of the conductive network, adding more CNTs results in a further slight increase in the conductivity that is attributed to the improvement of the conductive network quality (Figure 3c). As a result of introducing more CNTs into the system, above 0.5 wt.%, significant alterations to the structure occurred. Owing to the large bundle diameter and higher CNT content, disorder in the latex structure is induced as shown in Figure 3d. As can be seen in Figure 3, in both systems “double percolation” phenomena are present.41 This term refers to a structure in which there are two types of percolation in the same composite material. The initial electrical percolation centered around 0.01 wt.% of CNTs (estimated from the first derivative of the best fit 6398

dx.doi.org/10.1021/jp111998p |J. Phys. Chem. B 2011, 115, 6395–6400

The Journal of Physical Chemistry B

ARTICLE

100 nm) shown in the inset of Figure 4. It has to be noted that in the case of latex films, the conductivity has to be compromised to achieve high transmittance. Nevertheless, the latex films still conduct electricity and can be used in applications where high transparency and relatively low conductivity is required, such as in antistatic coatings as well as in capacitive touchscreen materials.

Figure 4. Optical transmission spectra of a representative “hard” pristine latex film and composites containing “hard” latex and 0.12 wt.% and 0.5 wt.% of CNTs, respectively. The inset shows a comparative transmission spectrum of an ITO-coated thin glass substrate (adapted from the “www.pgo-online.com” Web site).

to the data in Figure 3a) is dominated by tunneling and other transport mechanisms (trapping, hopping) between adjacent CNT aggregates.40 We are aware of the fact that the percolation threshold of 0.12 wt.% is not the lowest presented in the literature. There are reports of the electrical percolation being as low as 0.002% by volume42 or 0.0025% by weight.43 In fact, in our system, ultimately low electrical percolation thresholds could be achieved by maximizing polymer particle packing density by changing the particle size distribution. The main requirements for conducting composites to be considered suitable for applications in electronics are high transparency and a relatively high conductivity. Most of the composite systems fulfill these requirements. However, if used for transparent electrode coatings for flexible electronics, they are inevitably subjected to deformation, which can be detrimental to electrical stability. For instance, for many applications where one wants the benefit of both the electrical and mechanical attributes of the CNT inclusions, it is beneficial to have nanotubes evenly distributed throughout the composite. In contrast to other systems, the segregated network composites that we have fabricated are flexible and highly stretchable in the plane of the sheet and still conduct electricity if strained beyond 200%.30 Moreover, uniform distribution of CNTs, as was shown here, is also beneficial for up-scalable and reproducible device fabrication. Additionally, if CNTs are to be used in transparent coatings, it is critical that they are uniformly dispersed to avoid aggregates leading to excessive light scattering.44 It is also important that the particles do not reaggregate during the film formation process. Therefore, template-assisted self-assembly of CNTs is a perfect way to obtain highly transparent conductive thin films at a very low CNT content. For example in Figure 4, an 800 nm thick polymer film containing 0.12 wt.% CNTs (percolation threshold concentration) obtained by spin coating (at 3000 rpm) onto a glass substrate has a transmittance of 93% (at 600 nm). In comparison, a film with the same thickness but having a fully developed honeycomb-like network of CNTs (at 0.5 wt.%) has a transmittance of 91% (at 600 nm). The optical transmittance of the representative latex films used in this study is higher than that of a representative commercial ITO film (which has been deposited on a glass substrate and having a thickness of approximately

4. CONCLUSIONS To summarize, we have shown a unique, up-scalable, and facile method to obtain conductive thin films using environmentally friendly latex processing. The high transparency and low electrical percolation threshold are derived by locating CNTs continuously and homogenously at the submicrometer scale in a selforganized way. The composite material that we have produced permits precise amounts of CNTs to be reproducibly and uniformly dispersed over arbitrarily large areas. This simple approach facilitates the reduction of the percolation threshold by almost a factor of 4 in comparison to what is found for a random network in a latex host. The application of these composite coatings is relatively straightforward and can be potentially achieved by several conventional paint application methods,45 including brushing, spraying, or rolling, making them ideal candidates for, as an example, large-area antistatic coatings. As the CNTs are homogenously distributed in the colloidal crystal lattice, in a controlled fashion, easy dissipation of static electric charge build-up can be achieved. In addition, this composite is soft, flexible, and elastically deformable in the plane, while still conducting electricity when strained to 200%,30 which is very important for usage in flexible electronic applications. Moreover, the manipulation of the colloidal crystal structure could be ultimately used to control ordering of CNTs (or other particles) at the submicrometer level in a range of application areas from photonics to mechanical actuators. ’ ASSOCIATED CONTENT

bS

Supporting Information. A schematic representation of the setup used for measuring the electrical conductivity of thin composite films. This material is available free of charge via the Internet at http://pubs.acs.org.

’ AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected] (A.B.D.), izabela.jurewicz@ surrey.ac.uk (I.J.).

’ ACKNOWLEDGMENT We thank the following funding sources for their support of our research activities: the Royal Society, Sharp Corporation, Human Frontier Science Program, and the Royal Thai Government. We acknowledge NeoResins (Waalwijk, The Netherlands) and Cytec Surface Specialties (Drogenbos, Belgium) for providing latex samples and Violeta Doukova for general laboratory assistance. ’ REFERENCES (1) Odom, T. W.; Huang, J. L.; Kim, P.; Lieber, C. M. Nature 1998, 391, 62–64. 6399

dx.doi.org/10.1021/jp111998p |J. Phys. Chem. B 2011, 115, 6395–6400

The Journal of Physical Chemistry B (2) Dalton, A. B.; Collins, S.; Razal, J.; Munoz, E.; Ebron, V. H.; Kim, B. G.; Coleman, J. N.; Ferraris, J. P.; Baughman, R. H. J. Mater. Chem. 2004, 14, 1–3. (3) Yu, M. F.; Lourie, O.; Dyer, M. J.; Moloni, K.; Kelly, T. F.; Ruoff, R. S. Science 2000, 287, 637–640. (4) McNally, T.; Potschke, P. Polymer Carbon Nanotube Composites Preparation, Properties and Applications; Woodhead Pub Ltd.: 2011. (5) Sangermano, M.; Pegel, S.; P€otschke, P.; Voit, B. Macromol. Rapid Commun. 2008, 29, 396–400. (6) Ng, M. H. A.; Hartadi, L. T.; Tan, H.; Poa, C. H. P. Nanotechnology 2008, 19, 1–5. (7) Bunde, A.; Dieterich, W. J. Electroceram. 2000, 5, 81–92. (8) Stauffer, D. Introduction to percolation theory; Taylor & Francis: London, Philadelphia, 1985. (9) Kulshreshtha, A. K.; Vasile, C. Handbook of polymer blends and composites; Rapra Technology Lt.: Shawbury, Shewsbury, Shropshire, 2002. (10) Sahimi, M.; Hughes, B. D.; Scriven, L. E.; Davis, H. T. J. Phys. C: Solid State Phys. 1983, 16, L521–L527. (11) Rintoul, M. D.; Torquato, S. J. Phys. A: Math. Gen. 1997, 30, L585–L592. (12) Kusy, R. P. J. Appl. Phys. 1977, 48, 5301–5305. (13) Johner, N.; Grimaldi, C.; Maeder, T.; Ryser, P. Phys. Rev. E: Stat., Nonlinear, Soft Matter Phys. 2009, 79, 020104. (14) Mamunya, Y. Macromol. Symp. 2001, 170, 257–264. (15) Narkis, M.; Srivastava, S.; Tchoudakov, R.; Breuer, O. Synth. Met. 2000, 113, 29–34. (16) Feng, J. Y.; Chan, C. M. Polym. Eng. Sci. 1998, 38, 1649–1657. (17) Chan, C. M.; Cheng, C. L.; Yuen, M. M. F. Polym. Eng. Sci. 1997, 37, 1127–1136. (18) Zhang, C.; Ma, C. A.; Wang, P.; Sumita, M. Carbon 2005, 43, 2544–2553. (19) Grunlan, J. C.; Gerberich, W. W.; Francis, L. F. J. Appl. Polym. Sci. 2001, 80, 692–705. (20) Thommerel, E.; Valmalette, J. C.; Musso, J.; Villain, S.; Gavarri, J. R.; Spada, D. Mater. Sci. Eng., A 2002, 328, 67–79. (21) Alvarez, M. P.; Poblete, V. H.; Pilleux, M. E.; Fuenzalida, V. M. J. Appl. Polym. Sci. 2006, 99, 3005–3008. (22) Keddie, J. L.; Routh, A. F. Fundamentals of latex film formation: processes and properties; Springer: Dordrecht, The Netherlands, 2010. (23) Grunlan, J. C.; Kim, Y. S.; Ziaee, S.; Wei, X.; Abdel-Magid, B.; Tao, K. Macromol. Mater. Eng. 2006, 291, 1035–1043. (24) Kim, D.; Kim, Y.; Choi, K.; Grunlan, J. C.; Yu, C. H. ACS Nano 2010, 4, 513–523. (25) Yu, C.; Kim, Y. S.; Kim, D.; Grunlan, J. C. Nano Lett. 2008, 8, 4428–4432. (26) Grunlan, J. C.; Gerberich, W. W.; Francis, L. F. Polym. Eng. Sci. 2001, 41, 1947–1962. (27) Grossiord, N.; Miltner, H. E.; Loos, J.; Meuldijk, J.; Van Mele, B.; Koning, C. E. Chem. Mater. 2007, 19, 3787–3792. (28) Regev, O.; ElKati, P. N. B.; Loos, J.; Koning, C. E. Adv. Mater. 2004, 16, 248–251. (29) Wang, T.; Lei, C. H.; Dalton, A. B.; Creton, C.; Lin, Y.; Fernando, K. A. S.; Sun, Y. P.; Manea, M.; Asua, J. M.; Keddie, J. L. Adv. Mater. 2006, 18, 2730–2734. (30) Jurewicz, I.; King, A. A. K.; Worajittiphon, P.; Asanithi, P.; Brunner, E. W.; Sear, R. P.; Hosea, T. J. C.; Keddie, J. L.; Dalton, A. B. Macromol. Rapid Commun. 2010, 31, 609–615. (31) Bug, A. L. R.; Safran, S. A.; Webman, I. Phys. Rev. Lett. 1985, 54, 1412. (32) Torquato, S. Random heterogeneous materials: microstructure and macroscopic properties; Springer: New York, 2002. (33) Potschke, P.; Dudkin, S. M.; Alig, I. Polymer 2003, 44, 5023–5030. (34) Kilbride, B. E.; Coleman, J. N.; Fraysse, J.; Fournet, P.; Cadek, M.; Drury, A.; Hutzler, S.; Roth, S.; Blau, W. J. J. Appl. Phys. 2002, 92, 4024–4030. (35) Gubbels, F.; Jerome, R.; Teyssie, P.; Vanlathem, E.; Deltour, R.; Calderone, A.; Parente, V.; Bredas, J. L. Macromolecules 1994, 27, 1972–1974.

ARTICLE

(36) Hu, J. W.; Li, M. W.; Zhang, M. Q.; Xiao, D. S.; Cheng, G. S.; Rong, M. Z. Macromol. Rapid Commun. 2003, 24, 889–893. (37) Wu, S. H.; Masaharu, I.; Natsuki, T.; Ni, Q. Q. J. Reinf. Plast. Compos. 2006, 25, 1957–1966. (38) Ha, M. L. P.; Grady, B. P.; Lolli, G.; Resasco, D. E.; Ford, W. T. Macromol. Chem. Phys. 2007, 208, 446–456. (39) Gao, J. F.; Li, Z. M.; Meng, Q. J.; Yang, Q. Mater. Lett. 2008, 62, 3530–3532. (40) Calberg, C.; Blacher, S.; Gubbels, F.; Brouers, F.; Deltour, R.; Jerome, R. J. Phys. D: Appl. Phys. 1999, 32, 1517–1525. (41) Pionteck, J.; Wypych, G. Handbook of antistatics; ChemTec Pub.: Toronto, 2007. (42) Wakabayashi, A.; Sasakawa, Y.; Dobashi, T.; Yamamoto, T. Langmuir 2007, 23, 7990–7994. (43) Sandler, J. K. W.; Kirk, J. E.; Kinloch, I. A.; Shaffer, M. S. P.; Windle, A. H. Polymer 2003, 44, 5893–5899. (44) Vandervorst, P.; Lei, C.; Lin, Y.; Dupont, O.; Dalton, A. B.; Sun, Y.-P.; Keddie, J. L. Prog. Org. Coat. 2006, 57, 91–97. (45) Weldon, D. G. Failure analysis of paints and coating; Wiley: Chichester, 2009.

6400

dx.doi.org/10.1021/jp111998p |J. Phys. Chem. B 2011, 115, 6395–6400