pubs.acs.org/Langmuir © 2010 American Chemical Society
Nonaligned Carbon Nanotubes Partially Embedded in Polymer Matrixes: A Novel Route to Superhydrophobic Conductive Surfaces Mao Peng,* Zhangjie Liao, Ji Qi, and Zhi Zhou MOE Key Laboratory of Macromolecular Synthesis and Functionalization, Department of Polymer Science and Engineering, Zhejiang University, Hangzhou 310027, China Received May 7, 2010. Revised Manuscript Received June 25, 2010 A new method for transforming common polymers into superhydrophobic conductive surfaces, with both a high static water contact angle (∼160°) and a low sliding angle (2.0°-4.5°), and a low sheet resistance on the order of 101-103 Ω/sq is presented. A layer of multiwalled carbon nanotubes (MWNTs) is first distributed on the surface of a polymer substrate, then by a single step of pressing, the MWNTs are partially embedded inside the substrate surface and form a superhydrophobic coating with a “carpet-” or “hair”-like morphology. The infiltration of polymer melts into the porous MWNT layer follows Darcy’s law, and the pressing time greatly influence the morphology and superhydrophobicity. Moreover, the coating can be electrically heated by 20-70 °C with a voltage as low as 4-8 V at an electric energy density below 1.6 J/cm2 and therefore can be used for deicing applications. Hydroxylation and fluoroalkylsilane treatment can greatly improve the stability of the superhydrophobicity of MWNTs. This method is convenient and applicable to a variety of thermoplastic polymers and nonpolymer substrates coated by silicone rubber.
Introduction Preparation of superhydrophobic surfaces with a water contact angle (WCA) above 150° has attracted great research interest over the past decade in both scientific and engineering communities.1-6 Superhydrophobic surfaces are generally prepared by combining surface roughness at both the micro- and nanoscales with lowsurface-energy materials. For commercially available and widely used polymers, superhydrophobic and self-cleaning surfaces are also highly desirable; therefore, several approaches have been successfully developed to transform common polymers into superhydrophobic surfaces. For example, Erbil et al.4 prepared superhydrophobic isotactic polypropylene (i-PP) by solvent casting. With a suitable selection of solvents and temperature to control the surface roughness, the i-PP coating has a water contact angle (WCA) of 160°. Pakkanen et al.7 prepared superhydrophobic polyolefin surfaces by injection molding using a microstructured anodized aluminum oxide mold insert to pattern the surfaces. A WCA of about 165° and a sliding angle (SA) of about 2.5° were achieved for the optimized microstructure. Feng et al.8 prepared hydrophobic polyethylene films with a WCA of about 154° by hot pressing the films in vacuum onto polydimethylsiloxane stamps replicated from lotus leaves. More recently, Frechet et al.9 successfully prepared polymer porous coatings that can impart superhydrophobicity to various polymer and metal substrates. On the other hand, electrically conductive *To whom correspondence should be addressed. Fax: þþ86-57187953712. Tel.: þþ86-571-87953712. E-mail address:
[email protected]. (1) Neinhuis, C.; Barthlott, W. Ann. Bot. 1997, 79, 667. (2) Lafuma, D. Quere Nat. Mater. 2003, 2, 457. (3) Pastine, S. J.; Okawa, D.; Kessler, B.; Rolandi, M.; Llorente, M.; Zettl, A.; Frechet, J. M. J. J. Am. Chem. Soc. 2008, 130, 4238. (4) Erbil, H. Y.; Demirel, A. L.; Avci, Y.; Mert, O. Science 2003, 299, 1377. (5) Sun, T. L.; Feng, L.; Gao, X. F.; Jiang, L. Acc. Chem. Res. 2005, 38, 644. (6) Nakajima, A.; Fujishima, A.; Hashimoto, K.; Watanabe, T. Adv. Mater. 1999, 11, 1365. (7) Puukilainen, E.; Rasilainen, T.; Suvanto, M.; Pakkanen, T. A. Langmuir 2007, 23, 7263. (8) Feng, J.; Huang, M.; Qian, X. Macromol. Mater. Eng. 2009, 294, 295. (9) Levkin, A. P.; Svec, F.; Frechet, J. M. J. Adv. Mater. 2009, 21, 1.
13572 DOI: 10.1021/la101827c
superhydrophobic surfaces are of great interest because of their capability to remove static charges accumulated on the surfaces and potential applications as electromagnetic interference (EMI) shielding materials.10,11 Carbon nanotubes (CNTs) with promising conductivity, thermoconductivity, and mechanical properties have been extensively investigated since early 1990s. Superhydrophobic conductive surfaces have also been successfully constructed from CNTs/silane sol mixtures,12 CNTs/polymer composite coatings13 or free-standing films,10,14 a self-assembled nanocomposite coating of CNTs and oligo(p-phenylenevinylene)s,15 nonaligned CNT coatings directly grown on the surface of stainless steel,16 and well-aligned CNT forests on a variety of substrates.17,18 Although the targeted superhydrophobic properties of various substrates have been successfully obtained, convenient methods for imparting both superhydrophobicity and conductivity to common polymer substrates are still rare. In this article, we report a simple but effective approach for the construction of superhydrophobic and highly conductive composite coatings on the surface of a variety of polymers and siliconecoated glass and metals by simply pressing a layer of multiwalled carbon nanotubes (MWNTs) on the surface of polymer melts or silicone oligomers. In other words, the polymer melts or oligomers are infiltrated under pressure into the MWNT mat on the surface of the substrates. Under appropriate conditions, the MWNTs are partially embedded inside and partially exposed outside the surface of the coatings, forming a “carpet-” or “hair”-like morphology and imparting superhydrophobicity and conductivity to (10) Zou, J. H.; Chen, H.; Chunder, A.; Yu, Y. X.; Huo, Q.; Zhai, L. Adv. Mater. 2008, 20, 3337. (11) Zhu, Y.; Hu, D.; Wan, M.; Jiang, L.; Wei, Y. Adv. Mater. 2007, 19, 2092. (12) Han, J. T.; Kim, S. Y.; Woo, J. S.; Lee, G. W. Adv. Mater. 2008, 20, 3724. (13) Luo, C.; Zuo, X. L.; Wang, L.; Wang, E. G.; Song, S. P.; Wang, J.; Wang, J.; Fan, C. H.; Cao, Y. Nano Lett. 2008, 8, 4454. (14) Ji, J.; Fu, J.; Shen, J. Adv. Mater. 2007, 18, 1441. (15) Srinivasan, S.; Praveen, V. K.; Philip, R.; Ajayaghosh, A. Angew. Chem., Int. Ed. 2008, 47, 5750. (16) Sethi, S.; Dhinojwala, A. Langmuir 2009, 25, 4311. (17) Lau, K. K. S.; Bico, J.; Teo, K.B. K.; Chhowalla, M.; Amaratunga, G. A. J.; Milne, W. I.; McKinley, G. H.; Gleason, K. K. Nano Lett. 2003, 3, 1071. (18) Zhang, L.; Resasco, D. E. Langmuir 2009, 25, 4792.
Published on Web 07/14/2010
Langmuir 2010, 26(16), 13572–13578
Peng et al.
Article
Figure 1. Schematic diagram of the preparation of MWNTs/ polymer composite coatings by pressing and a typical optical photograph of a thin slice of MWNT-coated HDPE sample, showing a relatively uniform MWNTs/HDPE composite coating on the HDPE substrate.
the substrates. The WCA values of the composite coatings are about 160°, and the SA values are about 2.0-4.5° for different polymer substrates. The surfaces are also highly conductive with an average sheet resistance below 103 Ω/sq. The low stability of hydrophobicity of untreated MWNTs was improved by the surface modification of MWNTs with a fluoroalkylsilane.
Experimental Section Materials. MWNTs with an average diameter of 10-20 nm, a purity above 95 wt %, were supplied by Shenzhen NanoTechnologies Port Co., Ltd., Shenzhen, China, and used as received. High-density polyethylene (HDPE) (2480) was purchased from Qilu Petro., Shandong, China. Polypropylene (PP) (T300) was purchased from Sinopec Shanghai Petrochemical Co., Ltd., China. Ethylene-propylene diene rubber (EPDM) (4045) was purchased from Jilin Chemical Works, China. Polystyrene (PS) (535) was purchased from Zhanjiang Zhongmei Chemical Industries Co., Ltd., Guangdong, China. Poly(methyl methacrylate) PMMA (IF850) was produced by LG, Korea. All the thermoplastics were made into rectangular plates with a size of 20 mm 50 mm and thickness of 2 mm. The room-temperature-vulcanizing type silicone rubber (RTV-SR) Dow Corning 3140 RTV (viscosity = 30 000 mPa 3 s, Dow Corning, Shanghai. China) is a moisture-curable, flowable, semitransparent, and one-part solventless liquid. (Heptadecafluoro-1,1,2,2-tetrahydrodecyl) triethoxysilane was purchased from Alfa Aesar. Preparation of MWNT Composite Coatings. The fabrication steps for superhydrophobic conductive MWNTs/polymer coatings are schematically presented in Figure 1. The powder of pristine MWNTs (0.1 g) was distributed on the surface of polymer plates and pressed at a pressure of about 4 MPa. The mold was then heated to above the melting temperature or viscous flow temperature of the polymers, so that the polymer melts are infiltrated into the internanotube space, and the nanotubes were embedded inside the coating surfaces. For simplicity, all samples were prepared by pressing at 190 °C for 10 min in this study. After cooling, the samples were demolded, soaked in ethanol, ultrasonically cleaned in a sonicator (KQ-100B, Kunsan, China) for 10 min, and then rinsed with ethanol. The sonication and rinsing procedures were repeated for at least three times until no visible MWNTs appeared in ethanol after sonication; therefore, the residual MWNTs in the composite coatings were firmly embedded inside the substrates. The samples were dried in a vacuum oven at 70 °C until constant weight. To prepare MWNTs/RTVSR coatings on metal and glass substrates, the substrates were first cleaned by isopropyl alcohol and then coated by a layer of RTVSR primer by brushing. The thickness of the RTV-SR coating is Langmuir 2010, 26(16), 13572–13578
about 0.1 mm. The samples were transferred into a mold, and MWNTs were distributed on the surface. The samples were then pressed at a pressure of about 0.05 MPa for 72 h under ambient temperature. During this process the RTV-SR coating was moisture-cured. Finally, the samples were demolded and purified by repeated sonication and rinsing in ethanol three times to remove the MWNTs not firmly embedded inside the substrates. Functionalization of the Composite Coatings. The HDPE plates with MWNTs/HDPE composite coating were immersed in hydrogen peroxide (100 mL) and heated to 60 °C for 72 h. The samples were then rinsed repeatedly in deionized water, until the solution reached a pH value of 7. The samples were vacuum-dried at 70 °C for 24 h to obtain hydroxylated surfaces and then immersed in an ethanol solution (10 mL) of (heptadecafluoro1,1,2,2-tetrahydrodecyl) triethoxysilane (0.1 mL), and the sol-gel reaction was allowed to continue for 72 h at 60 °C. The samples were then rinsed in ethanol for several times and vacuum-dried to obtain fluorinated surfaces. Characterization. The morphologies of MWNTs/polymer composite coatings were observed by field emission scanning electron microscopy (FE-SEM) (Hitachi S-4800, Japan) at an accelerating voltage of 5 kV after the coatings were spray-coated with gold. The optical images were taken with a digital camera (Kodak, P850). The WCA and SA values were measured with a contact angle meter (DATA Physics System, OCA20, Germany). The volume of the water droplets for WCA measurements was fixed to 4 μL. The WCA values were the average of five measurements made at various positions of each sample. The SA values were measured by dropping a water droplet (10 μL) onto the tilted surfaces from about 4 mm height, and determined as the tilting angles at which the water droplets rolled off the surfaces. Sheet resistances were measured by the standard four-probe technique using a RTS-4 four-probe conductivity meter (Guangzhou 4 Probes Tech., China). Surface area measurements of the composite coatings were carried out using the Brunauer-EmmettTeller (BET) nitrogen adsorption method with an Autosorb-1-C volumetric adsorption analyzer (Quantachrome Instruments, USA) after the sample was degassed at 473 K for 2 h in a vacuum. The BET specific surface area was obtained over a relative pressure range of 0.003-0.081. It should be noted that the total BET surface areas of the HDPE plates with the composite coatings were first measured, then, the specific surface areas of the composite coatings were calculated as the ratio between the surface areas of the HDPE plates and the accurate weight of the composite coatings. The accurate weight of the composite coatings was calculated as the ratio between the weights of MWNTs in the coatings and the content of MWNTs in the composite coatings. The content of MWNTs in the coatings was determined by the TG measurements. Thermal infrared images were recorded by a DL-500EM thermal infrared camera (Zhejiang DALI technology, Hangzhou, China)
Results and Discussions Superhydrophobic Conductive Composite Coatings. Pressure infiltration is a process in which a liquid is infiltrated into a porous media when the pressure becomes high. This process has been extensively investigated both theoretically19 and in the preparation of ceramic/metal composites,20 and fiber21 or nanotube22 reinforced polymer composites, in which the filler performs work as the porous media and the molten metals or polymers/oligomers work as the liquids. As shown in Figure 1, in our experiment, a mat of pristine MWNTs distributed on the surface of polymer substrates works as a nanoporous media. To prepare the MWNTs/polymer composite coatings, the mold was (19) (20) (21) (22)
Han, J.; Qiao, Y. J. Am. Chem. Soc. 2006, 128, 10348. Garcia-Cordovilla Acta Mater. 1999, 47, 4461. Pang, J. W. C.; Bond, I. P. Compos. Sci. Technol. 2005, 65, 1791. Qiu, J. J.; Zhang, C.; Wang, B.; Liang, R. Nanotechnology 2007, 18, 275708.
DOI: 10.1021/la101827c
13573
Article
Peng et al. Table 1. Water Contact Angle, Sliding Angle, and Sheet Resistance (Rs) of MWNTs/Polymer Composite Coatings on Various Substances substrates
WCA (deg)a
WCA (deg)b
HDPE 97 ( 2 165 ( 3 PP 100 ( 2 166 ( 4 EPDM 107 ( 2 167 ( 3 PS 86 ( 1 159 ( 2 PMMA 83 ( 2 160 ( 3 RTV-SR 108 ( 1 157 ( 3 a Pure polymer. b Composite coating.
SA (deg)b
Rs (Ω/sq)b
∼2.4 ∼2.2 ∼2.3 ∼4.5 ∼3.6 ∼2.0
10.2 ( 0.2 10.3 ( 0.7 322.9 ( 5.5 91.5 ( 4.2 1214 ( 8.0 652.7 ( 61.6
Figure 2. (a) A water droplet (4 μL) on the surface of MWNTs/ HDPE composite coating showing a WCA of 165 ( 3°. (b) A water droplet (10 μL) rolls off the substrate showing a SA of 2.4°. (c) A jet of water bounces off a MWNTs/HDPE sample (dimension of 20 mm 50 mm 2 mm). (d) A superhydrophobic HDPE rectangular plate with a dimension of about 105 mm 105 mm 4 mm.
heated rapidly to 190 °C and pressed for 10 min under a pressure of 4 MPa, so that the polymer melts were infiltrated into the internanotube space, that is, the micropores or microchannels in the MWNT mats. After cooling and demolding, the samples were extensively cleaned by ultrasonication to remove the nanotubes that are not firmly embedded inside the surfaces. The infiltration of polymer melts into the MWNT mats is confirmed by the optical image in Figure 1, in which, a uniform and black coating on the surface of high density polyethylene (HDPE) sample is observed. The composite coating imparts superhydrophobicity to HDPE. As shown in Figure 2a, a water droplet (4 μL) forms a high WCA of 165 ( 3°. It was noted that the water repellency of the sample was so remarkable that it was impossible to measure the WCA if the water droplets were not kept bound to the dispensing needle. Besides WCA, SA is another important criterion for superhydrophobic surfaces. Figure 2b shows a water droplet (10 μL) that rolled off the substrate rapidly and left the surface completely dry when the substrate was slightly tilted (2.4°). When a jet of water was applied on the horizontal composite coating, it bounced off without leaving a trace (Figure 2c). In contrast, a pure HDPE surface has a low WCA of about 97°. This confirms that the superhydrophobicity is caused by the MWNTs/HDPE composite coating. Moreover, by appropriate design of the metal mold, it is possible to coat MWNTs onto polymer objects of any reasonable size, for example, Figure 2d shows a superhydrophobic sample with a relatively large area (105 105 mm2). The water droplets on the surface are bright because of the reflection of light from the interface between the water droplets and the air trapped at the substrate surface.13 On the other hand, the composite coatings also have excellent conductivity. The typical sheet resistance of the MWNTs/HDPE coating is 10.2 ( 0.2 Ω/sq. Other than HDPE, we also succeeded in preparing superhydrophobic MWNT coatings on a wide variety of polymer substrates, including PP, EPDM, PS, and PMMA. For simplicity, all the samples were prepared by pressing at 190 °C for 10 min under 4 MPa. The WCA values of the pure polymers and corresponding MWNTs/polymer composite coatings are summarized in Table 1, which shows that the average WCA values of all the composite coatings are above 159° and the SA values are below 3°. The substrates also have a good conductivity with sheet 13574 DOI: 10.1021/la101827c
Figure 3. SEM images of (a) the top-view surface morphology and (b) the cross sectional morphology of MWNTs/HDPE composite coating at low magnifications. The top-view surface morphology at a high magnification (c). The cross sectional morphology of the top edge (d) and the interface region (e) of the composite coatings at a high magnification. (f) A zoom-in SEM image of the coating surface shows some polymer granules on the top surface of the coating.
resistances ranging from 101 to 103 Ω/sq. Therefore, this study has presented a versatile and general single-step approach for the transformation of common polymers into superhydrophobic conductive surfaces. Morphology. Figure 3 shows the surface and cross sectional morphology of the MWNTs/HDPE composite coating. The field emission scanning electron microscopy (FE-SEM) images of a low magnification (Figure 3a,b) shows that the composite coating is rather uniform. At a higher magnification (Figure 3c,d), it can be observed that a large number of nanotubes are exposed outside the substrate surface, showing a carpet- or hair-like morphology. The MWNTs are highly curved, leading to a high porosity and surface roughness, so that a large amount of air can be trapped at the coating surface. It is believable that this unique carpet- or hairlike structure is responsible for the superhydrophobic nature of the composite coatings, because its surface contact area available with water is very low. Figure 3e presents a typical cross-sectional SEM image of the interface region of the MWNTs/HDPE composite coating, which Langmuir 2010, 26(16), 13572–13578
Peng et al.
Article
Figure 4. Influence of infiltration time on (a) the thickness of the composite layer, (b) the weight ratio of HDPE in the composite coatings as determined by TG analysis, and the specific surface area of the composite coatings, (c) the WCA and SA values, and (d) the sheet resistance of the MWNTs/HDPE composite coatings. The inset in panel a is the plot of log(d) vs log(t).
confirms that MWNTs are partially embedded in the HDPE matrix (as indicated by the black arrow) and the “roots” of the nanotubes pulled out of the HDPE substrate are still encapsulated by a layer of polymer because the thickness is somewhat larger than that of clean MWNTs (as indicated by the white arrows). Because the MWNTs are mechanically embedded inside the polymer surfaces, the superhydrophobicity and conductivity of the composite coating exhibit satisfactory durability. For example, the WCA, SA, and conductivity were almost invariant after the samples were heat treated at 120 °C or cooled to -20 °C for 1 h, and for foldable substrates such as EPDM (an elastomer) and thin HDPE films, the values were also invariant after the samples were curved for several tens of times. Furthermore, a magnified FE-SEM image in Figure 3f reveals that there exist some grape-like granules on the surface of the sample. The granules are submicrometer sized and tightly entangled with MWNTs. This morphology should result from the inhomogeneous infiltration of polymer melts into the internanotube space of the MWNT mats, as is widely accepted that, during a pressure infiltration process, the liquid is infiltrated into the largest pores first and progressively into smaller pores.19 Because the MWNTs are randomly stacked in the MWNT mats and the size of the microchannels in the mats is not uniform, the polymer melts preferentially penetrate into the larger microchannels in the mats, so a part of the melts can penetrate through the MWNT layer and appear on the top surface. This morphology obviously has great influence on the superhydrophobicity of the composite coatings. It should be noted that the samples in this study exhibit both high WCA and low SA values, while previous studies have shown that untreated aligned carbon nanotube forest (ACNT) films usually have relatively high SA values because of the high surface energy of untreated CNTs as a graphite material. For example, Chen et al. reported an untreated ACNT film with a WCA of 146°, but the difference between advancing WCA and receding WCA was 33°. Jiang et al. also reported an ACNT film with a WCA of 158.5 ( 1.5° and a relatively high SA of above 30°,23 but a patterned ACNT film with a honeycomb-like structure had a (23) Li, H.; Wang, X; Song, Y; Liu, Y.; Li, Q.; Jiang, L.; Zhu, D. B. Angew. Chem., Int. Ed. 2001, 40, 1743.
Langmuir 2010, 26(16), 13572–13578
much smaller SA of about 5°.24 It is well-known that the water contact angle for flat graphite25,26 and pure HDPE is about 85° and 97°, respectively. Therefore, the high WCA and low SA values of our samples are most likely related to the fact that the composite coatings are highly porous,2 because the MWNTs are nonaligned and form heterogeneous networks on the top surface of the coatings. On the other hand, the surface morphology of the composite coatings on other polymer substrates were also observed by FESEM (see Supporting Information), which showed that all samples are covered by a layer of nonaligned MWNTs and some submicrometer-sized polymer granules can also be observed on the coating surfaces. Generally speaking, the samples have similar morphology for different polymer substrates. The polymer granules exposed on the surfaces are most likely responsible for the relatively smaller WCA and larger SA values (Table 1) of PS and PMMA substrates, because their surface free energy values are somewhat higher than those of HDPE, PP, and EPDM. Influence of Infiltration Time. It has been shown that the polymer melts can penetrate through the MWNT mats and appear on the top surface of the composite coating. This phenomenon significantly affects the surface morphology and superhydrophobicity of the composite coatings. The one-dimensional infiltration of polymer fluids into the MWNT mats in the through-thickness direction may be described by Darcy’s law which describe the dynamics of a fluid flow through a porous medium:27 h2 ¼
2kt ΔP ηð1 - VCNTs Þ
ð1Þ
where t is the infiltration time, VCNTs is the volume of CNTs in the mat, η is the viscosity of polymer fluids, h is impregnation distance (through-thickness direction), ΔP is the pressure drop in the fluid, and k is the permeability coefficient, which depends on many factors including the porosity of the medium, the configuration of (24) Li, S. H.; Li, H. J.; Wang, X. B.; Song, Y. L.; Liu, Y. Q.; Jiang, L.; Zhu, D. B. J. Phys. Chem. B 2002, 106, 9274. (25) Fowkes, F. M.; Harkins, W. D. J. Am. Chem. Soc. 1940, 62, 3377. (26) Morcos J. Chem. Phys. 1972, 57, 1801–1802. (27) Cassie, B. D.; Baxter, S. Trans. Faraday Soc. 1944, 40, 546.
DOI: 10.1021/la101827c
13575
Article
voids, and wettability of polymer fluids on the surface of MWNTs, etc. From eq 1, it can be speculated that the distance of polymer melts infiltrated into the MWNT mats is determined by several factors including pressure, temperature (which determines the viscosity of the polymer melts), and time for infiltration. Because the time for pressure infiltration is easiest to control, we held a constant temperature and pressure for the pressing, and investigated the influence of infiltration time on the morphology, superhydrophobicity, and conductivity of the MWNTs/HDPE composite coatings. Therefore, the relationship between the impregnation distance h and the infiltration time t should follows h µ t1/2 according to eq 1. As shown in Figure 4a, when the infiltration time was increased from 5 to 30 min, the thickness of the composite coatings d, which should be close to the impregnation distance h of the HDPE melts into the nanoporous MWNT mats, increased rapidly from 60.5 to 145.3 μm. The inset in Figure 4a shows a logarithmic plot of the infiltration time t against the coating thickness d, which shows that when the pressing time is below 30 min, log(d) is linearly dependent on log(t) with a slope of 0.49, suggesting that d µ t1/2, in good agreement with the prediction of the Darcy’s law. This result is also in good agreement with the previously reported experimental result about the impregnation behavior of thermoplastics into porous carbon fiber performs.28 When the infiltration time is above 30 min, the thickness became invariant because the thickness of the MWNT mats was preset to about 150 μm at the beginning of the experiment. On the other hand, the result of thermogravimetric (TG) analysis showed that the fraction of HDPE in the composite coatings gradually increased from 9 wt % to above 30 wt % when the infiltration time is prolonged from 5 to 90 min, as shown in Figure 4b. This phenomenon demonstrates that the polymer melts penetrated into large pores inside the MWNT mats first and then into smaller size pores for the increased infiltration time. At the same time, the specific surface area of the composite coatings decreases rapidly from 92.4 to 3.6 m2/g when the infiltration time is prolonged from 5 to 90 min, which strongly affect the surface structure and properties of the composite coatings. As shown in Figure 4c, when the infiltration time was below 30 min, the WCA values were above 160°, but when the infiltration time was prolonged to 40 min, WCA dropped rapidly to 148 ( 3°, and for an infiltration time of 120 min, WCA decreased to 116 ( 5°. The infiltration time has a more dramatic impact on the SA values. As to SA values, it increased from about 2.1° to 3.8° (Figures 4c and 5a,b) when the infiltration time was increased from 5 to 30 min, but it increased abruptly to above 34° (Figure 4c and 5c) when the infiltration time was 40 min, although the sample still had a relatively high WCA of 148 ( 3°. For the infiltration time above 40 min, the samples exhibited strong water pinning behavior. Even when the substrates were tilted 180°, the water droplets were still pinned (Figure 5d). Different from the WCA and SA values, the sheet resistances were only slightly decreased with the increase of infiltration time for the MWNTs/HDPE composite coatings, as shown in Figure 4d. This is because the contents of MWNTs in the composite coatings are rather high, so that MWNTs form a continuous conductive network even for a short infiltration time. The SEM images in Figure 6 shed light on the effect of the infiltration time on the surface morphology and superhydrophobicity of the MWNTs/HDPE composite coatings. When the infiltration time is 30 min the surface is hairy. But with the (28) Jespersen, S. T.; Wakeman, M. D.; Michaud, V.; Cramer, D.; Manson, J.-A. E. Compos. Sci. Technol. 2008, 68, 1822.
13576 DOI: 10.1021/la101827c
Peng et al.
Figure 5. Influence of infiltration time on the SA values of the MWNTs/HDPE composite coatings: (a) 20, (b) 30, and (c) 40 min (volume of water droplets is 10 μL). (d) When the infiltration time is above 40 min, the samples exhibit water pinning.
increase of infiltration time, more and more HDPE gets to the top surface, so that many nanotubes are bounded together by HDPE and the surface becomes less hairy, leading to a decreased surface roughness and porosity. For the infiltration time of 120 min, HDPE even fuses into a continuous flat film (Figure 6e1), leading to a WCA of 116 ( 5° and a strong water pinning effect as aforementioned. This result is in good agreement with the result of TG analysis and BET surface area measurement which show that the amount of HDPE in the composite coatings increases and the BET surface area decreases apparently with the increase of infiltration time. Therefore, the superhydrophobicity of the composite coatings is strongly dependent on the composition and morphology of the surfaces, and thus can be tailored by the infiltration time and other factors such as temperature and pressure for infiltration. Composite Coatings on Glass and Metals. Besides polymeric substrates, this method can also be applied to glass, metal, and other solid surfaces coated by a layer of polymers. To demonstrate this, we coated glass, stainless steel, and aluminum rectangular plates with layers of commercially available RTV-SR first and then coated them with MWNTs by a pressure induced filtration process similar to that described in Figure 1. The mold was kept at ambient temperature and the pressure was only about 0.05 MPa in considering the low viscosity of RTV-SR. The MWNTs are cemented by RTV-SR, forming a highly rough and porous surface, as illuminated by the SEM images in Figure 7a,b. Figure 7c shows that no matter on what substrates, the MWNTs/RTV-SR coatings are superhydrophobic. The average WCA is 157 ( 3° and the SA is about 2°. Figure 7d shows that a steel plate with dimensions of 40 mm 40 mm 0.5 mm and both sides coated by the MWNTs/RTV-SR composite coating can float on the surface of water, which mimics water striders or water spiders that can float on the surface of water. Durability of the Superhydrophobicity. Previous studies have shown that because of the high surface energy of the graphitic surface of CNTs, the superhydrophobicity of untreated MWNTs films exhibits instability for prolonged time. For example, Gleason et al.17 reported that an ACNT film prepared by plasma-enhanced chemical vapor deposition, in which the nanotubes were perfectly aligned and untangled, had an initial WCA of 161°, but the WCA deceased rapidly and the droplets seeped into the forest voids after a few minutes. As to the MWNTs/HDPE coatings in this study, it was found that WCA decreased from Langmuir 2010, 26(16), 13572–13578
Peng et al.
Article
Figure 6. Top-view (a1-e1) and corresponding side-view (a2-e2) SEM images of the MWNTs/HDPE composite coatings prepared by
pressure infiltration at 170 °C for (a1, a2) 30, (b1, b2) 40, (c1, c2) 60, (d1, d2) 90, and (e1, e2) 120 min, respectively. The insets are the optical pictures of water droplets (10 μL) on the top of the surfaces, which show that the WCA values decrease apparently with the increase of infiltration time.
Figure 8. FT-IR spectra of the MWNTs/HDPE composite coatings before and after being treated by the fluorosilane.
Figure 7. Morphology of the cross sectional (a) and top view (b) of a MWNTs/RTV-SR composite coating on the surface of glass. The inset shows a water droplet on the surface showing a WCA of 157 ( 3°. (c) MWNTs/RTV-SR composite coatings on a variety of substrates: (1) aluminum, (2) glass, and (3) stainless steel. (d) A stainless steel plate (40 mm 40 mm 0.5 mm) with its both sides coated with MWNTs/RTV-SR coating can float on the surface of water.
165 ( 3° to 148 ( 2° after the samples were immersed in deionized water for only 10 min, and SA increased from about 2.4° to about 5.6°. Although the stability of our samples is somewhat better than that of ACNT forests, possibly due to the difference between the morphology of the nonaligned MWNT network and the untangled ACNT forest, such instability is negative to the applications of the superhydrophobic surfaces and should be avoided. In the literature, to improve the stability of superhydrophobicity, CNTs were coated with materials of low surface free energy, such as poly(tetrafluoroethylene),17 ZnO,29 oligo(p-phenylenevinylene)s,15 silica particles with fluorinated groups,12 and poly(3-hexylthiophene)-block-polystyrene10 or chemically modified by alkylation30 or silylation with a fluoroalkylsilane.31 In this study, the composite coatings were first treated with hydrogen peroxide to attach hydroxyl groups onto the sidewall of MWNTs according to a method previously reported12 and then treated with (heptadecafluoro-1,1,2,2-tetrahydrodecyl) triethoxysilane. Figure 8 (29) Huang, L.; Lau, S. P.; Yang, H. Y.; Leong, E. S. P.; Yu, S. F. J. Phys. Chem. B 2005, 109, 7746. (30) Xu, D. H.; Liu, H.; Yang, L.; Wang, Z. G. Carbon 2006, 44, 3226. (31) Georgakilas, V.; Bourlinos, A. B.; Zboril, R.; Trapalis, C. Chem. Mater. 2008, 20, 2884.
Langmuir 2010, 26(16), 13572–13578
Figure 9. Relationship between the WCA values and immersing time in water for the MWNTs/HDPE composite coatings before and after the fluorosilane treatment.
presents the FTIR spectra of the composite coating before and after the fluorosilane treatment. It is shown that after the surface treatment, some new bands appear in the FTIR spectra. The band at 1210 cm-1 is attributed to the stretching vibration of C-F bonds of the fluorosilane, and those at 1149 and 1069 cm-1 are characteristic for the formation of Si-O-C bonds.31 It was found that after the surface modification, the coatings had a WCA of 166 ( 2° and a SA of about 2.4°. Moreover, after being immersed in deionized water for 10 min, WCA became 164 ( 2° and SA was invariant (see Supporting Information). As shown in Figure 9, the samples remain superhydrophobic even after being immersed in water for 24 h; in contrast, the WCA of pristine MWNTs decreased rapidly to below 150°. This demonstrates that coating MWNTs with fluorosilane can significantly improve the stability of superhydrophobicity. Meanwhile, the conductivity is only slightly increased from 10.2 ( 0.2 to 20.0 ( 0.6 Ω/sq after the surface modification, indicating that the DOI: 10.1021/la101827c
13577
Article
Peng et al.
boiling temperature of water or even the melting temperature of polyethylene. In the later case, the coatings deformed apparently. To demonstrate the application of this composite coating for deicing purposes, some water was placed on the surface of the composite coating and frozen at -10 °C for 2 h. The sample was tilted at an angle of about 25°, and electrically heated under a voltage of 5 V at an environmental temperature of about -8 °C. After being heated for only about 28 s, the ice slides off the coating because the ice was melted at the ice/coating interface, moreover, the coating remains superhydrophobic after the ice was removed from the coating surface (see video in the Supporting Information)
Figure 10. (A) The thermal infrared images of the composite coating heated at a voltage of 6 V and (B) the increase of temperature at the center of the composite coating under different voltages.
conductivity of the composite coatings was not deteriorated by surface modification. Electric Heating Behavior. It is well-known that carbon nanotubes have good electrical conductivity and can be electrically heated by a current flowing through them.32 The composite coatings in this study also exhibit good conductivity and can be easily heated by applying a direct electric current. Figure 10 presents the relationship between surface temperature and heating time at various voltages measured by thermal infrared imaging technique. The composite coating can be electrically heated by about 20-70 °C within 5 min when the voltage is increased from 4 to 8 V at an ambient temperature of 10 °C. It was found that the WCA values decreased somewhat when voltage was applied, and the temperature of the coatings was increased. The higher the voltage, the lower the WCA values. Fortunately, the WCAs were still above 150°. For example, for the pristine CNT coatings, the WCA decreased from about 161° to about 151° after the coating was heated at the voltage of 8 V, and the WCAs of the coatings treated by fluorosilane decreased from about 165 o to about 158° when the voltage was increased to 8 V. For higher voltages, the temperature was increased to above the (32) Deshpande, V. V.; Hsieh, S.; Bushmaker, A. W.; Bockrath, M.; Cronin, S. B. Phys. Rev. Lett. 2009, 102, 105501.
13578 DOI: 10.1021/la101827c
Conclusions We have presented a facile and versatile method for the preparation of superhydrophobic conductive surfaces with MWNTs/polymer composite coatings. Under appropriate conditions, the nanotubes are partially embedded inside the coatings, offering superhydrophobicity and a high conductivity to the substrates. Surface modification of MWNTs with fluoroalkylsilane can significantly improve the stability of superhydrophobicity of the composite coatings. This method is convenient, inexpensive, and easy to scale up. Moreover, previous studies have shown that CNTs can be covalently attached with organic compounds or polymers through chemical functionalizations or decorated with a variety of quantum dots or nanoparticles of metals, ceramics, and semiconductors.33-35 Therefore, the carpet-like layer of MWNTs would provide a versatile platform for the creation of functionalized surfaces that are useful in many applications such as electrodes for chemical sensors, biosensors, batteries, microwave absorption,36 or EMI shielding materials, and catalyst carriers. Acknowledgment. We greatly appreciate financial support from the National Natural Science Foundation of China (Grants 20574060 and 50773066). Supporting Information Available: Optical pictures, SEM images, scheme for composite coating on nonpolymer surfaces (PDF), and a video. This material is available free of charge via the Internet at http://pubs.acs.org. (33) Banerjee, S.; Hemraj-Benny, T.; Wong, S. S. Adv. Mater. 2005, 17, 17. (34) Ravindran, S.; Chaudhary, S.; Colburn, B.; Ozkan, M.; Ozkan, C. S. Nano Lett. 2003, 3, 447. (35) Haremza, J. M.; Hahn, M. A.; Krauss, T. D. Nano Lett. 2002, 2, 1253. (36) Deng, L; Han, M Appl. Phys. Lett. 2007, 91, 023119.
Langmuir 2010, 26(16), 13572–13578
