- Email: [email protected]
Novel method to characterize superhydrophobic coatings
Journal of Colloid and Interface Science 395 (2013) 315–321
Contents lists available at SciVerse ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Novel method to characterize superhydrophobic coatings Mohamed A. Samaha, Hooman Vahedi Tafreshi, Mohamed Gad-el-Hak ⇑ Department of Mechanical & Nuclear Engineering, Virginia Commonwealth University, Richmond, VA 23284-3015, USA
a r t i c l e
i n f o
Article history: Received 17 November 2012 Accepted 25 December 2012 Available online 10 January 2013 Keywords: Superhydrophobic coatings Fibers Aerogel Buoyancy force Effective thickness Gas volume fraction Coating’s thickness
a b s t r a c t Superhydrophobic coatings possess a strong water-repellent characteristic, which, among several other potential applications, enhances the mobility of water droplets over the surface. The coating traps air within its micropores, such that a submerged moving body experiences shear-free and no-slip regions over, respectively, the air pockets and the solid surface. This, in turn, may lead to significant skin-friction reduction. The coating maintains its superhydrophobicity as long as the air remains entrapped. It is therefore of great interest to precisely measure the amount of trapped air, which is particularly difficult to estimate for coatings with disordered microstructures. A novel method to measure the effective thickness and gas volume fraction of superhydrophobic coatings with either ordered or random microroughness is advanced. The technique is applied to both aerogel and electrospun fibrous coatings. The experiments utilize a sensitive weighing scale (down to 104 gm) and height gauge (down to 10 lm) to determine the buoyancy force on an immersed, coated glass-slide substrate. The measured force is used to calculate the volume fraction of entrapped air. The coating’s effective thickness also follows from the same calculations. The sensitivity of our particular scale enables the measuring of thicknesses down to 3 lm, which is not readily possible with conventional thickness gauges. Smaller thicknesses could be measured using more sensitive scales. Ó 2013 Elsevier Inc. All rights reserved.
1. Introduction 1.1. Superhydrophobicity Superhydrophobicity could be achieved by a combination of low surface energy (chemical hydrophobicity) and micro- or nanoscale surface roughness. Chemical hydrophobicity is a material property of repelling water, which is achieved when the chemical structure of the surface is sufficiently different from that of water. This leads to strongly reducing the ability of the surface to interact with water. The superhydrophobic phenomenon is primarily manifested by water droplets beading on the solid surface with contact angles exceeding 150°, and the droplets readily roll off when the surface is tilted at a small angle. In nature, superhydrophobic surfaces are exemplified by the lotus leaves, which enhance the mobility of rain drops on them, carrying dirt away, and creating a self-cleansing effect (so-called lotus effect). Neinhuis and Barthlott [1] obtained scanning electron microscopy (SEM) images for several water-repellent plants and reported the micromorphological characteristics of 200 species. They demonstrated that the epidermal (i.e., outermost) cells of the lotus leaf form papillae, which act as microstructure roughness. The papillae are topped off by a very dense layer of epicuticular waxes (hydrophobic wax crystals), also referred to as hair-like structures [2] or nanostructure rough⇑ Corresponding author. E-mail address: [email protected] (M. Gad-el-Hak). 0021-9797/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jcis.2012.12.066
ness [3]. Over the past two decades, biomimetic research of the lotus effect evoked several studies on manufacturing, characterizing, and applying superhydrophobic surfaces [1–9]. A superhydrophobic surface entraps air in its pores, and when a coated body is fully submersed in water, alternating water–solid and water–air interfaces are formed. The entrapped air is separated from water with a thin interface anchored on the solid walls and stretched due to surface tension forces. It has been observed that a moving body of water ‘‘slips’’ over the air–water interfaces, whereas the water ‘‘sticks’’ to the solid portions of the surface [7]. Therefore, if the percentage of the surface covered by air pockets is sufficiently high, a superhydrophobic surface can cause the so-called ‘‘slip effect,’’ resulting in a reduction in the skin-friction drag exerted on the surface [7]. As long as the air pockets exist, the surface remains superhydrophobic. In other words, the degree of hydrophobicity and the beneficial effects are diminished by the reduction of the amount of entrapped air [10–12]. The longevity of a superhydrophobic surface—how long the surface can maintain the air pockets—is critical, especially in underwater applications. An elegant experiment carried out by Sakai et al. [13] demonstrates the effect of surface microstructure on air-layer sustainability. 1.2. Fabrication Early manmade superhydrophobic surfaces were produced using the same microfabrication techniques developed for the computer industry. The coatings typically consisted of a regular ar-
316
M.A. Samaha et al. / Journal of Colloid and Interface Science 395 (2013) 315–321
ray of microposts, shallow microcavities, or microridges [14–17]. The orientation of microroughness, flow characteristics, and shape of air–water interface (meniscus) could all significantly affect the slip condition and hence the drag reduction. For example, Woolford et al. [18] demonstrated that in a turbulent flow, streamwise ridges (i.e., flow direction parallel to the microridges) could lead to drag reduction, while spanwise ridges (i.e., flow direction perpendicular to the microridges) could lead to drag increase. Additionally, Steinberger et al. [19] demonstrated that the meniscus shape influences the boundary condition, which could increase the friction. Furthermore, Feng et al. [20] indicated that superhydrophobic surfaces could be adhesive with strong contact-angle hysteresis, although the air layer exists, and the static contact angle exceeds 150°. This is called ‘‘petal effect,’’ possessed by the petals of red roses. Large-scale manufacturing of microfabricated superhydrophobic surfaces is prohibitively expensive. An alternative solution to circumvent the high cost is to produce surfaces made up of random deposition of hydrophobic particles [21–26] or electrospun fibers [27,28]. Electrospinning of superhydrophobic polymers is a simple, low-cost method that could be used to deposit micro- to nanofibrous coatings onto substrates of arbitrary material and geometry. The resulting superhydrophobic surfaces could be applied for various purposes, including self-cleaning glasses and clothes, protection against corrosion of the metallic parts of bridges, marine vehicles, underwater constructions, etc., anti-snow sticking, and, most significantly, reduction in skin-friction drag for underwater vessels such as torpedoes and submarines. A superhydrophobic coating could therefore be utilized as a passive method of flow control and may potentially become a viable alternative to the more complex and energy consuming active or reactive flow control techniques such as wall suction/blowing [29]. Yet, a series of challenges needs to be overcome before implementing such surfaces for wide use. For example, in the area of surface characterization, it is quite a challenge to measure the thickness of thin coatings approaching 1 lm using conventional methods and to estimate their gas volume fraction. Such measurements are particularly more difficult for pliable material with superimposed random microroughness. On the other hand, measurements of both the thickness and gas volume fraction of a superhydrophobic surface are needed to characterize the surface. Such measurements could be used to determine the amount of entrapped air, which in turn could be used to predict the drag reduction, slip effect, and longevity of the surface. Air is entrapped between the micro- and nanostructured roughness making a superhydrophobic surface water repellent. When a submerged object is coated with such a surface, both the coating and the entrapped air result in a larger displacement volume of water, which provides a relatively larger buoyancy force on the coated object. If the scale of the object is comparable to that of the coating, the relative force becomes significant. As demonstrated by Gao and Jiang [5], the volume of the displaced water caused by immersing a single leg of a water strider is 300 times that of the leg itself due to the entrapment of a relatively large amount of air. Furthermore, the buoyancy force developed on the immersed leg is 15 times the insect’s total body weight. The impact of buoyancy force on the performance of submerged superhydrophobic surfaces has also been studied [30–33]. In this work, we develop a novel method to estimate the thickness and gas volume fraction of superhydrophobic coatings with random microroughness. The concept of effective thickness is discussed in the next section. Section 3 describes the procedure for measuring the coating’s effective thickness and gas volume fraction. Validation of the present method is provided in Section 4. This is followed by results and discussion in Section 5. Conclusions and recommendations for future studies are given in Section 6.
2. Effective thickness When a superhydrophobic surface is immersed, it could entrap air in its micro- and nanostructured roughness resulting in a surface with both air–water and solid–water interfaces. A pressure force could be exerted on the air–water interface by the column of water above the surface. If the pressure is sufficiently high, water penetrates into the pores on the surface and replaces the air. The surface is then said to transition from the nonwetted Cassie state (with entrapped air in the pores) [34] to the wetted Wenzel state (with water replacing the air and filling the pores) [35]. This transition is interpreted in two ways: the first is based on minimizing the thermodynamic free energy [36,37], and the second using a balance of forces across the interface [38–40]. Lee and Kim [41] used the latter interpretation to develop an equation to compute the maximum allowable hydrostatic pressure without wetting transition (critical pressure) in terms of the surface microstructure for aligned or staggered arrangement of posts. The heterogeneity of the surface could result in mixed wetting and nonwetted states [25,42]. In the present experiment, we fabricate two types of superhydrophobic coatings. The first uses aerogel particles with different controlled particle sizes [25]. Hydrophobic aerogel beads made of amorphous silicon dioxide were ground and filtered through four sieves with mesh sizes of 43, 104, 150, and 210 lm. Starting with the smallest mesh, the aerogel particles were sieved to separate the finest particles. The remaining particles in the first sieve were then filtered in the next one, and the procedure was repeated until four categories of aerogel particles were obtained. From each category, aerogel particles were deposited onto a glass substrate coated with a double-sided tape for adhesion. Clearly, the thinnest produced aerogel coating could not be thinner than 43 lm—the smallest mesh size of the available sieves. Fig. 1a shows a scanning electron microscope (SEM) image of the 43 lm sample. The deposited particles provide the surface roughness and porosity necessary to entrap air when the surface is submerged in water. The inset in the figure shows a 3-mg water droplet on the coating with a static contact angle of 155°. An average contact-angle hysteresis of about 3° for a 9-mg water droplet is observed, further demonstrating the coating’s superhydrophobicity. Electrospun fibrous coating is the second type of superhydrophobic material that we previously fabricated in our laboratory [28]. By controlling the electrospinning conditions, one could produce a coating with thicknesses from the micro- to the nanoscale. Coatings were made with 18% concentration of polystyrene using a modified electrospinning method in which the substrate on the rotating drum was grounded, and the rest of the apparatus was isolated. A constricted electric field resulted in a larger proportion of fibers being deposited onto the substrate, which minimized the amount of wasted fibers during the electrospinning process. The substrate in this case was also a glass slide. Fig. 1b shows a SEM image of a typical fibrous coating. The deposited mat of hydrophobic polymer fibers provides the surface roughness and porosity necessary to entrap air. The inset in the figure shows a 3-mg water droplet on the coating with a static contact angle of 157°. An average contact-angle hysteresis of about 11.5° for a 9-mg water droplet is also observed. 2.1. Buoyancy force We measure the buoyancy force for both coated and uncoated immersed glass slides (see the next section). The difference between the two forces, DFb, is the increase in buoyancy due to the presence of the coating. Therefore, the volume of the displaced water due to the presence of the coating is
M.A. Samaha et al. / Journal of Colloid and Interface Science 395 (2013) 315–321
317
different possibilities of the interface for the same aerogel coating. We use aerogel coatings for validation because it is less pliable than electrospun fibers. The images in Fig. 2 are captured using an optical microscope made by Olympus Optical Co., LTD. (model BX60F5). From the figure, it is clear that the volume of the displaced water in case (a) is higher than that in case (b). The applied pressure plays a significant role in forming the shape of the interface. As the pressure increases, the water penetrates deeper into the coating’s micropores, which leads to decreasing the effective thickness. The minimum thickness that could be measured is the one corresponding to the complete wetting transition (Wenzel state), that is, the water fills all the micropores on the surface, which minimizes the volume of the displaced water. Therefore, the measured effective thickness takes into account both the effect of elevated pressure and shape of the air– water interface. The ratio of entrapped air volume to total volume of the coating is called gas volume fraction, , and is determined herein only for the electrospun fibrous coatings. It is quite a challenge to determine for the aerogel coatings because the density is needed for the calculations, and the density of aerogel particles changes after the grinding process. Accurate estimate of this density is difficult because of the micro- and nanoporosity of the particles. The thicker and less soft aerogel coatings are used only to validate the present technique, which is utilized to measure the thickness as well the gas volume fraction of the softer and potentially thinner electrospun fibrous coatings. The mass of the glass slide with and without the fibrous coating is measured using the sensitive scale. The difference between the two masses is approximately the mass of the fiber, mf, as the entrapped air mass is negligible. The solid volume of fibers could then be determined from
Vs ¼
Fig. 1. Two types of superhydrophobic coatings: (a) aerogel and (b) polystyrene fibers. Upper right inset in each figure shows a small water droplet on top of the respective coating.
DV b ¼
DF b
qw g
ð1Þ
noindent where qw is the deionized water density at the testing temperature (22 °C), and g is the gravitational acceleration. We then define an effective thickness as
te ¼
DV b As
ð2Þ
where As is the coating’s surface area. Validation of the thickness measurements is quite a challenge because the shape of the air–water interface is not known. For example, Fig. 2 shows two
mf
qp
ð3Þ
where qp is the density of the polystyrene (1040 kg/m3). The gas volume fraction is simply
¼
DV b V s DV b
ð4Þ
3. Experimental technique Fig. 3a shows the setup used to measure the effective thickness of superhydrophobic coatings, while Fig. 3b shows a schematic drawing of the same setup. A glass slide (52 mm 25.5 mm 1 mm) is used as a substrate on which aerogel particles are deposited. A smaller glass slide (25 mm 26.5 mm 1 mm) is used for the electrospun fibrous coatings. A double-sided tape with known thickness (70 lm) is used for adhesion for both types of coating. The slide is fastened to a holding arm using epoxy. The arm is attached to a precise height gauge having a resolution of 10 lm
Fig. 2. Two possible shapes of air–water interface for aerogel coating: (a) at low pressure and (b) at high pressure.
318
M.A. Samaha et al. / Journal of Colloid and Interface Science 395 (2013) 315–321
coating is immersed, and Ws is the corresponding reading when the glass with only the tape is immersed. Using Eqs. (1) and (2), the measured force DFb is employed to determine the effective thickness of the coating, te. 4. Validation In order to validate the results, we use aerogel coatings as explained earlier. We measure the coating nominal thickness using a thickness gauge (J-40-T) made by Electromatic Equipment Company. The thickness of the slide and the tape with and without the aerogel coating is measured, and the difference between the two readings is the coating’s nominal thickness. Fig. 4 shows the effective thickness as computed from Eq. (2) versus the coating’s nominal thickness measured using the thickness gauge. Four samples with different thicknesses are tested. The thickness is changed by changing the size of the deposited aerogel particles [25]. The figure indicates a good agreement between the two results. Note that the thickness gauge is composed of two parallel plates between which the sample is placed to measure its thickness. Thus, the gauge measures the distance between the top and bottom edges of the coating. The agreement between the results shown in the figure indicates that the estimated effective thickness is carried out for case (a) shown in Fig. 2, where there is no protrusion of water into the micropores of the coating. In order to justify this hypothesis, we calculate the maximum allowable pressure without wetting transition (critical pressure) for each coating and compare it to the actual applied pressure head (1 cm of water). In 2006, Liu and Lange [43] provided a theoretical model to predict the critical pressure at which a superhydrophobic surface starts departing from the Cassie state. Several other numerical and analytical studies have been performed to predict the critical pressure. Some of these studies predicted the meniscus stability for surfaces comprised of ordered [36,39] and/or random roughness [26,44–46]. Other studies have been performed to determine the meniscus shape at different pressures up to the critical pressure [47–49]. In the present experiments, we calculate the critical pressure for the coatings using the equation provided by Lee and Kim [41], which could be used for randomly rough surfaces, as explained in our previous publication [45]: Fig. 3. Setup used to estimate effective thickness and gas volume fraction of superhydrophobic coatings: (a) photograph of actual setup and (b) schematic drawing.
and made by Mitutoyo Corp. (model HDS-H12 ‘‘C). A petri dish filled with deionized water is placed over a sensitive mass scale having a resolution of 104 gm and made by Fisher Scientific (model accuSeries, accu-124). The height gauge and the holding arm are used to immerse the slide in the water at a precise location at which the water column is 1 cm over the slide. The water evaporates during the experiment, which could affect the measurements. We compensate for that by adding water to return back to the original weight before recording any data. After zeroing the scale (to exclude the weight of the water and the petri dish from the calculations), the indicated reading is the buoyancy force generated on the slide divided by the gravitational acceleration, since the buoyancy force on the slide is equal and opposite to the force on the water. The slide with the tape is immersed with and without the coating. The difference between the forces, DFb, for the two cases can be expressed as
DF b ¼ W t W s
Pmax /g 6
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 2c pð1 /g Þ cos h L
ð6Þ
where Pmax is the maximum allowable pressure without wetting transition, /g is the maximum gas area fraction, c is the surface
ð5Þ
where Wt is the scale reading multiplied by the gravitational acceleration when the glass with the tape and the superhydrophobic
Fig. 4. Comparison between effective thickness and measured coating nominal thickness. Solid line is linear best fit of experimental data.
M.A. Samaha et al. / Journal of Colloid and Interface Science 395 (2013) 315–321
319
of water head), which is not sufficient to deform the fibers. This is another advantage of the present method over the conventional thickness gauge. 5. Results 5.1. Coating thickness and gas volume fraction The present method is utilized to precisely measure both the microscale thickness and the gas volume fraction of randomly rough superhydrophobic coatings. This type of coatings is much harder to characterize than those made with organized microroughness. Superhydrophobic electrospun fibrous coatings [28] were produced for 18% polystyrene concentration dissolved in two solvents (1:1 toluene to Dimethylformamide, DMF). The electrospinning parameters are 4.8 kV DC-biased 11 kV AC terminal voltage, 570 Hz frequency, 50% duty cycle, 33 ll/min pump-infusion rate, and 7 cm nozzle–shaft distance. Several coatings with different thicknesses are produced by keeping all the electrospinning conditions constant except the collection time of the fiber deposition on the substrate. Optical microscope images of the resulted coatings show that the fiber’s diameter has a range of 6– 12 lm. Fig. 5a and b, respectively, show the effective thickness and the mass per unit area of the coatings versus the collection time. The data indicate a linear increase in thickness and coating’s mass with time. This is attributed to insulating the collector to minimize wasted fibers during the electrospinning process. The gas volume fraction is an important parameter that influences the longevity of a superhydrophobic coating. Determining the gas volume fraction is especially difficult when the surface is randomly rough. The method presented in this work is effectively used to determine the gas volume fraction of such a coating, as shown in Fig. 5c. To obtain these results, Eq. (4) is used together with the estimated solid volume of fibers (Vs calculated from Eq. (3)) and the volume of the displaced water DVb. The figure shows that is approximately constant, independent of the collection time. This is expected considering the linear relation between the mass and thickness, as shown in Fig. 5a and b. 5.2. Measuring thickness down to the microscale Fig. 5. (a) Influence of collection time on effective thickness; (b) coating mass per unit area versus collection time; and (c) gas volume fraction, , versus effective thickness. Fiber’s diameter has a range of 6–12 lm. Solid line is linear best fit of experimental data.
tension of the liquid (72 103 N/m in case of water), h is the contact angle, and L is the pitch (maximum distance between two particles). The data for the gas area fraction are imported from Samaha et al. [25]. In order to obtain the maximum pitch L, we compared several images of each coating taken by the optical microscope at several spots. The contact angle for a smooth aerogel surface is about 130° [50]. The results show that the applied pressure is always less than the critical pressure. Therefore, wetting transition is not reached for all coatings tested herein and that explains the agreement of the results shown in Fig. 4. The comparison between the effective and the nominal thickness is carried out for only the aerogel coating not for the fibrous one because the latter is made of softer material, which might be deformed when tested using the thickness gauge. When using conventional thickness gauges, the deformation of the fibers could lead to a significant error in the measurements. On the other hand, the buoyancy method advanced herein maintains its fidelity because the coating is subjected to a weak pressure of 98 Pa (1 cm
In this subsection, we demonstrate the application of our technique to measure coating’s thickness down to few micrometers, which is not possible with conventional thickness gauges. From Fig. 5a, it is clear that the smallest thickness that could be measured for this type of coating is about 15 lm (30 s collection time). We attempted to produce a coating with a smaller thickness by reducing the collection time to 15 s, but the error in the measured effective thickness becomes large. We reason that the packing of the fibers decreases too much at some locations for the latter case, as shown in Fig. 6. This results in bursting of the air–water interface at those locations, which yields significant errors in the effective thickness measurements. The interface fails at some areas but not at others owing to the randomness of the fibers. To produce thinner but more homogenous coatings, we reduced the fiber diameter by decreasing the pump-infusion rate [51] to 16 ll/min during the electrospinning process. This resulted in coatings with fiber’s diameter in the range of 0.8–3 lm, as determined from optical microscope images. Fig. 7 shows the effective thickness of the coating versus the collection time. It is clear that the present method can be used to determine coating effective thickness down to 3 lm. This is a significant achievement considering the difficulties involved in measuring the thickness of materials as thin and soft as electrospun nanofiber
320
M.A. Samaha et al. / Journal of Colloid and Interface Science 395 (2013) 315–321
Fig. 7. Effective thickness versus collection time for electrospun fibrous coatings. Fiber’s diameter has a range of 0.8–3 lm. Solid line is linear best fit of experimental data.
readings would be too time consuming to yield integrated, global measurements over extended surface areas. 6. Conclusions We developed a novel method to characterize both the thickness and the gas volume fraction of randomly rough superhydrophobic coatings. Using conventional laboratory scales, we demonstrated that our method could be applied to coatings as thin as 3 lm and as soft as fibrous coatings. Such a thickness is not measurable using conventional micrometers. Clearly, the present technique could be used to characterize coatings much thinner than 3 lm if one uses a more sensitive scale than the one used here. For example, scales as sensitive as 106 gm do exist, and therefore, coating’s thicknesses as low as 30 nm could be measured using the same methodology described herein. The technique is applicable for superhydrophobic coatings comprised of either ordered or random roughness. This opens a pathway to precisely characterize micro/nanoporous materials for industrial purposes, which is quite a challenge using conventional methods. Acknowledgments We are grateful to Professor Gary C. Tepper for the use of his laboratory. This research is sponsored by the Defense Advanced Research Projects Agency (DARPA), Contract Nos: W91CRB-10-10003 and W911QX-12-1-0001, technical sponsor Captain Christopher Warren, USN. The content of this paper does not necessarily reflect the position or the policy of the Government, and no official endorsement should be inferred. References
Fig. 6. Optical images of superhydrophobic electrospun fibrous coatings produced at different deposition collection times. (a) 60 s; (b) 30 s; and (c) 15 s. Pumpinfusion rate is 33 ll/min.
[1] [2] [3] [4] [5] [6] [7] [8] [9] [10]
mats. Accurate measurement of thickness has been a major challenge for many other industries that take advantage of nanofiber materials such as aerosol filtration and membranes [52,53], among several others. The method advanced herein is global and is suited for heterogeneous coatings. Optical profilometers and atomic force microscopes, for example, which lead to localized, accurate
[11] [12] [13] [14]
C. Neinhuis, W. Barthlott, Ann. Bot. 79 (1997) 667–677. Y.T. Cheng, D.E. Rodak, C.A. Hayden, Nanotechnology 17 (2006) 1359–1362. K. Koch, B. Bhushan, Y.C. Jung, W. Barthlott, Soft Matter 5 (2009) 1386–1393. D.L. Hu, B. Chan, J.W.M. Bush, Nature 424 (2003) 663–666. X. Gao, L. Jiang, Nature 432 (2004) 36. D.L. Hu, M. Prakash, B. Chan, J.W.M. Bush, Exp. Fluids 43 (2007) 769–778. J.P. Rothstein, Annu. Rev. Fluid Mech. 42 (2010) 89–109. N.J. Shirtcliffe, G. McHale, S. Atherton, M.I. Newton, Adv. Colloid Interface Sci. 161 (2010) 124–138. M.A. Samaha, H.V. Tafreshi, M. Gad-el-Hak, C.R. Mec. 340 (2012) 18–34. M. Bobji, S.V. Kumar, A. Asthana, R.N. Govardhan, Langmuir 25 (2009) 12120– 12126. R. Poetes, K. Holtzmann, K. Franze, U. Steiner, Phys. Rev. Lett. 105 (2010) 166104. M.A. Samaha, F.O. Ochanda, H.V. Tafreshi, G.C. Tepper, M. Gad-el-Hak, Rev. Sci. Instrum. 82 (2011) 045109. M. Sakai, T. Yanagisawa, A. Nakajima, Y. Kameshima, K. Okada, Langmuir 25 (2009) 13–16. J. Bico, C. Marzolin, D. Quéré, Europhys. Lett. 47 (1999) 220–226.
M.A. Samaha et al. / Journal of Colloid and Interface Science 395 (2013) 315–321 [15] N.J. Shirtcliffe, G. McHale, M.I. Newton, G. Chabrol, C.C. Perry, Adv. Mater. 16 (2004) 1929–1932. [16] J. Davies, D. Maynes, B.W. Webb, B. Woolford, Phys. Fluids 18 (2006) 087110. [17] C. Lee, C-H. Choi, C-J. Kim, Phys. Rev. Lett. 101 (2008) 064501. [18] B. Woolford, J. Prince, D. Maynes, B.W. Webb, Phys. Fluids 21 (2009) 085106. [19] A. Steinberger, C. Conttin-Bizonne, P. Kleimann, E. Charlaix, Nat. Mater. 6 (2007) 665–668. [20] L. Feng, Y. Zhang, J. Xi, Y. Zhu, N. Wang, F. Xia, L. Jiang, Langmuir 24 (2008) 4114–4119. [21] S.D. Bhagat, Y.-H. Kim, K.-H. Suh, Y.-S. Ahn, J.-G. Yeo, J.-H. Han, Microporous Mesoporous Mater. 112 (2008) 504–509. [22] H. Yang, Y. Deng, J. Colloid Interface Sci. 325 (2008) 588–593. [23] S. Lee, D. Kim, W. Hwang, Curr. Appl. Phys. 11 (2011) 800–804. [24] H.S. Hwang, N.H. Kim, S.G. Lee, D.Y. Lee, K. Cho, I. Park, ACS Appl. Mater. Interfaces 3 (2011) 2179–2183. [25] M.A. Samaha, F.O. Ochanda, H.V. Tafreshi, G.C. Tepper, M. Gad-el-Hak, Colloids Surf. A 399 (2012) 62–70. [26] B. Emami, T.M. Bucher, H.V. Tafreshi, D. Pestov, M. Gad-el-Hak, G.C. Tepper, Colloids Surf. A 385 (2011) 95–103. [27] M. Ma, Y. Mao, M. Gupta, K.K. Gleason, G.C. Rutledge, Macromolecules 38 (2005) 9742–9748. [28] F.O. Ochanda, M.A. Samaha, H.V. Tafreshi, G.C. Tepper, M. Gad-el-Hak, J. Appl. Polym. Sci. 123 (2012) 1112–1119. [29] M. Gad-el-Hak, Flow Control: Passive, Active, and Reactive Flow Management, Cambridge University Press, London, United Kingdom, 2000. [30] X.-Q. Feng, X. Gao, Z. Wu, L. Jiang, Q.-S. Zheng, Langmuir 23 (2007) 4892–4896. [31] Q. Pan, J. Liu, Q. Zhu, ACS Appl. Mater. Interfaces 2 (2010) 2026–2030.
[32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53]
321
B. Su, M. Li, Q. Lu, Langmuir 26 (2010) 6048–6052. G. McHale, M.R. Flynn, M.I. Newton, Soft Matter 7 (2011) 10100–10107. A.B.D. Cassie, S. Baxter, Trans. Faraday Soc. 40 (1944) 546–551. R.N. Wenzel, Ind. Eng. Chem. 28 (1936) 988–994. N.A. Patankar, Langmuir 20 (2004) 7097–7102. L. Barbieri, E. Wagner, P. Hoffmann, Langmuir 23 (2007) 1723–1734. C.W. Extrand, Langmuir 20 (2004) 5013–5018. C.W. Extrand, Langmuir 22 (2006) 1711–1714. Q.S. Zheng, Y. Yu, Z.H. Zhao, Langmuir 21 (2005) 12207–12212. C. Lee, C-J. Kim, Langmuir 25 (2009) 12812–12818. A. Marmur, Langmuir 19 (2003) 8343–8348. B. Liu, F.F. Lange, J. Colloid Interface Sci. 298 (2006) 899–909. M. Sbragaglia, A. Prosperetti, J. Fluid Mech. 578 (2007) 435–451. M.A. Samaha, H.V. Tafreshi, M. Gad-el-Hak, Phys. Fluids 23 (2011) 012001. T.M. Bucher, B. Emami, H.V. Tafreshi, M. Gad-el-Hak, G.C. Tepper, Phys. Fluids 24 (2012) 022109. B. Emami, H.V. Tafreshi, M. Gad-el-Hak, G.C. Tepper, Appl. Phys. Lett. 98 (2011) 203106. B. Emami, H.V. Tafreshi, M. Gad-el-Hak, G.C. Tepper, Appl. Phys. Lett. 100 (2012) 013104. B. Emami, H.V. Tafreshi, M. Gad-el-Hak, G.C. Tepper, J. Appl. Phys. 111 (2012) 064325. B.N. Nguyen, M.A.B. Meador, M.E. Tousley, B. Shonkwiler, L. McCorkle, D.A. Scheiman, A. Palczer, ACS Appl. Mater. Interfaces 1 (2009) 621–630. http://www.che.vt.edu/Wilkes/electrospinning/electrspinning.html. S.A. Hosseini, H.V. Tafreshi, Chem. Eng. Sci. 65 (2010) 2249–2254. S.A. Hosseini, H.V. Tafreshi, Powder Technol. 201 (2010) 153–160.
Contents lists available at SciVerse ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Novel method to characterize superhydrophobic coatings Mohamed A. Samaha, Hooman Vahedi Tafreshi, Mohamed Gad-el-Hak ⇑ Department of Mechanical & Nuclear Engineering, Virginia Commonwealth University, Richmond, VA 23284-3015, USA
a r t i c l e
i n f o
Article history: Received 17 November 2012 Accepted 25 December 2012 Available online 10 January 2013 Keywords: Superhydrophobic coatings Fibers Aerogel Buoyancy force Effective thickness Gas volume fraction Coating’s thickness
a b s t r a c t Superhydrophobic coatings possess a strong water-repellent characteristic, which, among several other potential applications, enhances the mobility of water droplets over the surface. The coating traps air within its micropores, such that a submerged moving body experiences shear-free and no-slip regions over, respectively, the air pockets and the solid surface. This, in turn, may lead to significant skin-friction reduction. The coating maintains its superhydrophobicity as long as the air remains entrapped. It is therefore of great interest to precisely measure the amount of trapped air, which is particularly difficult to estimate for coatings with disordered microstructures. A novel method to measure the effective thickness and gas volume fraction of superhydrophobic coatings with either ordered or random microroughness is advanced. The technique is applied to both aerogel and electrospun fibrous coatings. The experiments utilize a sensitive weighing scale (down to 104 gm) and height gauge (down to 10 lm) to determine the buoyancy force on an immersed, coated glass-slide substrate. The measured force is used to calculate the volume fraction of entrapped air. The coating’s effective thickness also follows from the same calculations. The sensitivity of our particular scale enables the measuring of thicknesses down to 3 lm, which is not readily possible with conventional thickness gauges. Smaller thicknesses could be measured using more sensitive scales. Ó 2013 Elsevier Inc. All rights reserved.
1. Introduction 1.1. Superhydrophobicity Superhydrophobicity could be achieved by a combination of low surface energy (chemical hydrophobicity) and micro- or nanoscale surface roughness. Chemical hydrophobicity is a material property of repelling water, which is achieved when the chemical structure of the surface is sufficiently different from that of water. This leads to strongly reducing the ability of the surface to interact with water. The superhydrophobic phenomenon is primarily manifested by water droplets beading on the solid surface with contact angles exceeding 150°, and the droplets readily roll off when the surface is tilted at a small angle. In nature, superhydrophobic surfaces are exemplified by the lotus leaves, which enhance the mobility of rain drops on them, carrying dirt away, and creating a self-cleansing effect (so-called lotus effect). Neinhuis and Barthlott [1] obtained scanning electron microscopy (SEM) images for several water-repellent plants and reported the micromorphological characteristics of 200 species. They demonstrated that the epidermal (i.e., outermost) cells of the lotus leaf form papillae, which act as microstructure roughness. The papillae are topped off by a very dense layer of epicuticular waxes (hydrophobic wax crystals), also referred to as hair-like structures [2] or nanostructure rough⇑ Corresponding author. E-mail address: [email protected] (M. Gad-el-Hak). 0021-9797/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jcis.2012.12.066
ness [3]. Over the past two decades, biomimetic research of the lotus effect evoked several studies on manufacturing, characterizing, and applying superhydrophobic surfaces [1–9]. A superhydrophobic surface entraps air in its pores, and when a coated body is fully submersed in water, alternating water–solid and water–air interfaces are formed. The entrapped air is separated from water with a thin interface anchored on the solid walls and stretched due to surface tension forces. It has been observed that a moving body of water ‘‘slips’’ over the air–water interfaces, whereas the water ‘‘sticks’’ to the solid portions of the surface [7]. Therefore, if the percentage of the surface covered by air pockets is sufficiently high, a superhydrophobic surface can cause the so-called ‘‘slip effect,’’ resulting in a reduction in the skin-friction drag exerted on the surface [7]. As long as the air pockets exist, the surface remains superhydrophobic. In other words, the degree of hydrophobicity and the beneficial effects are diminished by the reduction of the amount of entrapped air [10–12]. The longevity of a superhydrophobic surface—how long the surface can maintain the air pockets—is critical, especially in underwater applications. An elegant experiment carried out by Sakai et al. [13] demonstrates the effect of surface microstructure on air-layer sustainability. 1.2. Fabrication Early manmade superhydrophobic surfaces were produced using the same microfabrication techniques developed for the computer industry. The coatings typically consisted of a regular ar-
316
M.A. Samaha et al. / Journal of Colloid and Interface Science 395 (2013) 315–321
ray of microposts, shallow microcavities, or microridges [14–17]. The orientation of microroughness, flow characteristics, and shape of air–water interface (meniscus) could all significantly affect the slip condition and hence the drag reduction. For example, Woolford et al. [18] demonstrated that in a turbulent flow, streamwise ridges (i.e., flow direction parallel to the microridges) could lead to drag reduction, while spanwise ridges (i.e., flow direction perpendicular to the microridges) could lead to drag increase. Additionally, Steinberger et al. [19] demonstrated that the meniscus shape influences the boundary condition, which could increase the friction. Furthermore, Feng et al. [20] indicated that superhydrophobic surfaces could be adhesive with strong contact-angle hysteresis, although the air layer exists, and the static contact angle exceeds 150°. This is called ‘‘petal effect,’’ possessed by the petals of red roses. Large-scale manufacturing of microfabricated superhydrophobic surfaces is prohibitively expensive. An alternative solution to circumvent the high cost is to produce surfaces made up of random deposition of hydrophobic particles [21–26] or electrospun fibers [27,28]. Electrospinning of superhydrophobic polymers is a simple, low-cost method that could be used to deposit micro- to nanofibrous coatings onto substrates of arbitrary material and geometry. The resulting superhydrophobic surfaces could be applied for various purposes, including self-cleaning glasses and clothes, protection against corrosion of the metallic parts of bridges, marine vehicles, underwater constructions, etc., anti-snow sticking, and, most significantly, reduction in skin-friction drag for underwater vessels such as torpedoes and submarines. A superhydrophobic coating could therefore be utilized as a passive method of flow control and may potentially become a viable alternative to the more complex and energy consuming active or reactive flow control techniques such as wall suction/blowing [29]. Yet, a series of challenges needs to be overcome before implementing such surfaces for wide use. For example, in the area of surface characterization, it is quite a challenge to measure the thickness of thin coatings approaching 1 lm using conventional methods and to estimate their gas volume fraction. Such measurements are particularly more difficult for pliable material with superimposed random microroughness. On the other hand, measurements of both the thickness and gas volume fraction of a superhydrophobic surface are needed to characterize the surface. Such measurements could be used to determine the amount of entrapped air, which in turn could be used to predict the drag reduction, slip effect, and longevity of the surface. Air is entrapped between the micro- and nanostructured roughness making a superhydrophobic surface water repellent. When a submerged object is coated with such a surface, both the coating and the entrapped air result in a larger displacement volume of water, which provides a relatively larger buoyancy force on the coated object. If the scale of the object is comparable to that of the coating, the relative force becomes significant. As demonstrated by Gao and Jiang [5], the volume of the displaced water caused by immersing a single leg of a water strider is 300 times that of the leg itself due to the entrapment of a relatively large amount of air. Furthermore, the buoyancy force developed on the immersed leg is 15 times the insect’s total body weight. The impact of buoyancy force on the performance of submerged superhydrophobic surfaces has also been studied [30–33]. In this work, we develop a novel method to estimate the thickness and gas volume fraction of superhydrophobic coatings with random microroughness. The concept of effective thickness is discussed in the next section. Section 3 describes the procedure for measuring the coating’s effective thickness and gas volume fraction. Validation of the present method is provided in Section 4. This is followed by results and discussion in Section 5. Conclusions and recommendations for future studies are given in Section 6.
2. Effective thickness When a superhydrophobic surface is immersed, it could entrap air in its micro- and nanostructured roughness resulting in a surface with both air–water and solid–water interfaces. A pressure force could be exerted on the air–water interface by the column of water above the surface. If the pressure is sufficiently high, water penetrates into the pores on the surface and replaces the air. The surface is then said to transition from the nonwetted Cassie state (with entrapped air in the pores) [34] to the wetted Wenzel state (with water replacing the air and filling the pores) [35]. This transition is interpreted in two ways: the first is based on minimizing the thermodynamic free energy [36,37], and the second using a balance of forces across the interface [38–40]. Lee and Kim [41] used the latter interpretation to develop an equation to compute the maximum allowable hydrostatic pressure without wetting transition (critical pressure) in terms of the surface microstructure for aligned or staggered arrangement of posts. The heterogeneity of the surface could result in mixed wetting and nonwetted states [25,42]. In the present experiment, we fabricate two types of superhydrophobic coatings. The first uses aerogel particles with different controlled particle sizes [25]. Hydrophobic aerogel beads made of amorphous silicon dioxide were ground and filtered through four sieves with mesh sizes of 43, 104, 150, and 210 lm. Starting with the smallest mesh, the aerogel particles were sieved to separate the finest particles. The remaining particles in the first sieve were then filtered in the next one, and the procedure was repeated until four categories of aerogel particles were obtained. From each category, aerogel particles were deposited onto a glass substrate coated with a double-sided tape for adhesion. Clearly, the thinnest produced aerogel coating could not be thinner than 43 lm—the smallest mesh size of the available sieves. Fig. 1a shows a scanning electron microscope (SEM) image of the 43 lm sample. The deposited particles provide the surface roughness and porosity necessary to entrap air when the surface is submerged in water. The inset in the figure shows a 3-mg water droplet on the coating with a static contact angle of 155°. An average contact-angle hysteresis of about 3° for a 9-mg water droplet is observed, further demonstrating the coating’s superhydrophobicity. Electrospun fibrous coating is the second type of superhydrophobic material that we previously fabricated in our laboratory [28]. By controlling the electrospinning conditions, one could produce a coating with thicknesses from the micro- to the nanoscale. Coatings were made with 18% concentration of polystyrene using a modified electrospinning method in which the substrate on the rotating drum was grounded, and the rest of the apparatus was isolated. A constricted electric field resulted in a larger proportion of fibers being deposited onto the substrate, which minimized the amount of wasted fibers during the electrospinning process. The substrate in this case was also a glass slide. Fig. 1b shows a SEM image of a typical fibrous coating. The deposited mat of hydrophobic polymer fibers provides the surface roughness and porosity necessary to entrap air. The inset in the figure shows a 3-mg water droplet on the coating with a static contact angle of 157°. An average contact-angle hysteresis of about 11.5° for a 9-mg water droplet is also observed. 2.1. Buoyancy force We measure the buoyancy force for both coated and uncoated immersed glass slides (see the next section). The difference between the two forces, DFb, is the increase in buoyancy due to the presence of the coating. Therefore, the volume of the displaced water due to the presence of the coating is
M.A. Samaha et al. / Journal of Colloid and Interface Science 395 (2013) 315–321
317
different possibilities of the interface for the same aerogel coating. We use aerogel coatings for validation because it is less pliable than electrospun fibers. The images in Fig. 2 are captured using an optical microscope made by Olympus Optical Co., LTD. (model BX60F5). From the figure, it is clear that the volume of the displaced water in case (a) is higher than that in case (b). The applied pressure plays a significant role in forming the shape of the interface. As the pressure increases, the water penetrates deeper into the coating’s micropores, which leads to decreasing the effective thickness. The minimum thickness that could be measured is the one corresponding to the complete wetting transition (Wenzel state), that is, the water fills all the micropores on the surface, which minimizes the volume of the displaced water. Therefore, the measured effective thickness takes into account both the effect of elevated pressure and shape of the air– water interface. The ratio of entrapped air volume to total volume of the coating is called gas volume fraction, , and is determined herein only for the electrospun fibrous coatings. It is quite a challenge to determine for the aerogel coatings because the density is needed for the calculations, and the density of aerogel particles changes after the grinding process. Accurate estimate of this density is difficult because of the micro- and nanoporosity of the particles. The thicker and less soft aerogel coatings are used only to validate the present technique, which is utilized to measure the thickness as well the gas volume fraction of the softer and potentially thinner electrospun fibrous coatings. The mass of the glass slide with and without the fibrous coating is measured using the sensitive scale. The difference between the two masses is approximately the mass of the fiber, mf, as the entrapped air mass is negligible. The solid volume of fibers could then be determined from
Vs ¼
Fig. 1. Two types of superhydrophobic coatings: (a) aerogel and (b) polystyrene fibers. Upper right inset in each figure shows a small water droplet on top of the respective coating.
DV b ¼
DF b
qw g
ð1Þ
noindent where qw is the deionized water density at the testing temperature (22 °C), and g is the gravitational acceleration. We then define an effective thickness as
te ¼
DV b As
ð2Þ
where As is the coating’s surface area. Validation of the thickness measurements is quite a challenge because the shape of the air–water interface is not known. For example, Fig. 2 shows two
mf
qp
ð3Þ
where qp is the density of the polystyrene (1040 kg/m3). The gas volume fraction is simply
¼
DV b V s DV b
ð4Þ
3. Experimental technique Fig. 3a shows the setup used to measure the effective thickness of superhydrophobic coatings, while Fig. 3b shows a schematic drawing of the same setup. A glass slide (52 mm 25.5 mm 1 mm) is used as a substrate on which aerogel particles are deposited. A smaller glass slide (25 mm 26.5 mm 1 mm) is used for the electrospun fibrous coatings. A double-sided tape with known thickness (70 lm) is used for adhesion for both types of coating. The slide is fastened to a holding arm using epoxy. The arm is attached to a precise height gauge having a resolution of 10 lm
Fig. 2. Two possible shapes of air–water interface for aerogel coating: (a) at low pressure and (b) at high pressure.
318
M.A. Samaha et al. / Journal of Colloid and Interface Science 395 (2013) 315–321
coating is immersed, and Ws is the corresponding reading when the glass with only the tape is immersed. Using Eqs. (1) and (2), the measured force DFb is employed to determine the effective thickness of the coating, te. 4. Validation In order to validate the results, we use aerogel coatings as explained earlier. We measure the coating nominal thickness using a thickness gauge (J-40-T) made by Electromatic Equipment Company. The thickness of the slide and the tape with and without the aerogel coating is measured, and the difference between the two readings is the coating’s nominal thickness. Fig. 4 shows the effective thickness as computed from Eq. (2) versus the coating’s nominal thickness measured using the thickness gauge. Four samples with different thicknesses are tested. The thickness is changed by changing the size of the deposited aerogel particles [25]. The figure indicates a good agreement between the two results. Note that the thickness gauge is composed of two parallel plates between which the sample is placed to measure its thickness. Thus, the gauge measures the distance between the top and bottom edges of the coating. The agreement between the results shown in the figure indicates that the estimated effective thickness is carried out for case (a) shown in Fig. 2, where there is no protrusion of water into the micropores of the coating. In order to justify this hypothesis, we calculate the maximum allowable pressure without wetting transition (critical pressure) for each coating and compare it to the actual applied pressure head (1 cm of water). In 2006, Liu and Lange [43] provided a theoretical model to predict the critical pressure at which a superhydrophobic surface starts departing from the Cassie state. Several other numerical and analytical studies have been performed to predict the critical pressure. Some of these studies predicted the meniscus stability for surfaces comprised of ordered [36,39] and/or random roughness [26,44–46]. Other studies have been performed to determine the meniscus shape at different pressures up to the critical pressure [47–49]. In the present experiments, we calculate the critical pressure for the coatings using the equation provided by Lee and Kim [41], which could be used for randomly rough surfaces, as explained in our previous publication [45]: Fig. 3. Setup used to estimate effective thickness and gas volume fraction of superhydrophobic coatings: (a) photograph of actual setup and (b) schematic drawing.
and made by Mitutoyo Corp. (model HDS-H12 ‘‘C). A petri dish filled with deionized water is placed over a sensitive mass scale having a resolution of 104 gm and made by Fisher Scientific (model accuSeries, accu-124). The height gauge and the holding arm are used to immerse the slide in the water at a precise location at which the water column is 1 cm over the slide. The water evaporates during the experiment, which could affect the measurements. We compensate for that by adding water to return back to the original weight before recording any data. After zeroing the scale (to exclude the weight of the water and the petri dish from the calculations), the indicated reading is the buoyancy force generated on the slide divided by the gravitational acceleration, since the buoyancy force on the slide is equal and opposite to the force on the water. The slide with the tape is immersed with and without the coating. The difference between the forces, DFb, for the two cases can be expressed as
DF b ¼ W t W s
Pmax /g 6
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 2c pð1 /g Þ cos h L
ð6Þ
where Pmax is the maximum allowable pressure without wetting transition, /g is the maximum gas area fraction, c is the surface
ð5Þ
where Wt is the scale reading multiplied by the gravitational acceleration when the glass with the tape and the superhydrophobic
Fig. 4. Comparison between effective thickness and measured coating nominal thickness. Solid line is linear best fit of experimental data.
M.A. Samaha et al. / Journal of Colloid and Interface Science 395 (2013) 315–321
319
of water head), which is not sufficient to deform the fibers. This is another advantage of the present method over the conventional thickness gauge. 5. Results 5.1. Coating thickness and gas volume fraction The present method is utilized to precisely measure both the microscale thickness and the gas volume fraction of randomly rough superhydrophobic coatings. This type of coatings is much harder to characterize than those made with organized microroughness. Superhydrophobic electrospun fibrous coatings [28] were produced for 18% polystyrene concentration dissolved in two solvents (1:1 toluene to Dimethylformamide, DMF). The electrospinning parameters are 4.8 kV DC-biased 11 kV AC terminal voltage, 570 Hz frequency, 50% duty cycle, 33 ll/min pump-infusion rate, and 7 cm nozzle–shaft distance. Several coatings with different thicknesses are produced by keeping all the electrospinning conditions constant except the collection time of the fiber deposition on the substrate. Optical microscope images of the resulted coatings show that the fiber’s diameter has a range of 6– 12 lm. Fig. 5a and b, respectively, show the effective thickness and the mass per unit area of the coatings versus the collection time. The data indicate a linear increase in thickness and coating’s mass with time. This is attributed to insulating the collector to minimize wasted fibers during the electrospinning process. The gas volume fraction is an important parameter that influences the longevity of a superhydrophobic coating. Determining the gas volume fraction is especially difficult when the surface is randomly rough. The method presented in this work is effectively used to determine the gas volume fraction of such a coating, as shown in Fig. 5c. To obtain these results, Eq. (4) is used together with the estimated solid volume of fibers (Vs calculated from Eq. (3)) and the volume of the displaced water DVb. The figure shows that is approximately constant, independent of the collection time. This is expected considering the linear relation between the mass and thickness, as shown in Fig. 5a and b. 5.2. Measuring thickness down to the microscale Fig. 5. (a) Influence of collection time on effective thickness; (b) coating mass per unit area versus collection time; and (c) gas volume fraction, , versus effective thickness. Fiber’s diameter has a range of 6–12 lm. Solid line is linear best fit of experimental data.
tension of the liquid (72 103 N/m in case of water), h is the contact angle, and L is the pitch (maximum distance between two particles). The data for the gas area fraction are imported from Samaha et al. [25]. In order to obtain the maximum pitch L, we compared several images of each coating taken by the optical microscope at several spots. The contact angle for a smooth aerogel surface is about 130° [50]. The results show that the applied pressure is always less than the critical pressure. Therefore, wetting transition is not reached for all coatings tested herein and that explains the agreement of the results shown in Fig. 4. The comparison between the effective and the nominal thickness is carried out for only the aerogel coating not for the fibrous one because the latter is made of softer material, which might be deformed when tested using the thickness gauge. When using conventional thickness gauges, the deformation of the fibers could lead to a significant error in the measurements. On the other hand, the buoyancy method advanced herein maintains its fidelity because the coating is subjected to a weak pressure of 98 Pa (1 cm
In this subsection, we demonstrate the application of our technique to measure coating’s thickness down to few micrometers, which is not possible with conventional thickness gauges. From Fig. 5a, it is clear that the smallest thickness that could be measured for this type of coating is about 15 lm (30 s collection time). We attempted to produce a coating with a smaller thickness by reducing the collection time to 15 s, but the error in the measured effective thickness becomes large. We reason that the packing of the fibers decreases too much at some locations for the latter case, as shown in Fig. 6. This results in bursting of the air–water interface at those locations, which yields significant errors in the effective thickness measurements. The interface fails at some areas but not at others owing to the randomness of the fibers. To produce thinner but more homogenous coatings, we reduced the fiber diameter by decreasing the pump-infusion rate [51] to 16 ll/min during the electrospinning process. This resulted in coatings with fiber’s diameter in the range of 0.8–3 lm, as determined from optical microscope images. Fig. 7 shows the effective thickness of the coating versus the collection time. It is clear that the present method can be used to determine coating effective thickness down to 3 lm. This is a significant achievement considering the difficulties involved in measuring the thickness of materials as thin and soft as electrospun nanofiber
320
M.A. Samaha et al. / Journal of Colloid and Interface Science 395 (2013) 315–321
Fig. 7. Effective thickness versus collection time for electrospun fibrous coatings. Fiber’s diameter has a range of 0.8–3 lm. Solid line is linear best fit of experimental data.
readings would be too time consuming to yield integrated, global measurements over extended surface areas. 6. Conclusions We developed a novel method to characterize both the thickness and the gas volume fraction of randomly rough superhydrophobic coatings. Using conventional laboratory scales, we demonstrated that our method could be applied to coatings as thin as 3 lm and as soft as fibrous coatings. Such a thickness is not measurable using conventional micrometers. Clearly, the present technique could be used to characterize coatings much thinner than 3 lm if one uses a more sensitive scale than the one used here. For example, scales as sensitive as 106 gm do exist, and therefore, coating’s thicknesses as low as 30 nm could be measured using the same methodology described herein. The technique is applicable for superhydrophobic coatings comprised of either ordered or random roughness. This opens a pathway to precisely characterize micro/nanoporous materials for industrial purposes, which is quite a challenge using conventional methods. Acknowledgments We are grateful to Professor Gary C. Tepper for the use of his laboratory. This research is sponsored by the Defense Advanced Research Projects Agency (DARPA), Contract Nos: W91CRB-10-10003 and W911QX-12-1-0001, technical sponsor Captain Christopher Warren, USN. The content of this paper does not necessarily reflect the position or the policy of the Government, and no official endorsement should be inferred. References
Fig. 6. Optical images of superhydrophobic electrospun fibrous coatings produced at different deposition collection times. (a) 60 s; (b) 30 s; and (c) 15 s. Pumpinfusion rate is 33 ll/min.
[1] [2] [3] [4] [5] [6] [7] [8] [9] [10]
mats. Accurate measurement of thickness has been a major challenge for many other industries that take advantage of nanofiber materials such as aerosol filtration and membranes [52,53], among several others. The method advanced herein is global and is suited for heterogeneous coatings. Optical profilometers and atomic force microscopes, for example, which lead to localized, accurate
[11] [12] [13] [14]
C. Neinhuis, W. Barthlott, Ann. Bot. 79 (1997) 667–677. Y.T. Cheng, D.E. Rodak, C.A. Hayden, Nanotechnology 17 (2006) 1359–1362. K. Koch, B. Bhushan, Y.C. Jung, W. Barthlott, Soft Matter 5 (2009) 1386–1393. D.L. Hu, B. Chan, J.W.M. Bush, Nature 424 (2003) 663–666. X. Gao, L. Jiang, Nature 432 (2004) 36. D.L. Hu, M. Prakash, B. Chan, J.W.M. Bush, Exp. Fluids 43 (2007) 769–778. J.P. Rothstein, Annu. Rev. Fluid Mech. 42 (2010) 89–109. N.J. Shirtcliffe, G. McHale, S. Atherton, M.I. Newton, Adv. Colloid Interface Sci. 161 (2010) 124–138. M.A. Samaha, H.V. Tafreshi, M. Gad-el-Hak, C.R. Mec. 340 (2012) 18–34. M. Bobji, S.V. Kumar, A. Asthana, R.N. Govardhan, Langmuir 25 (2009) 12120– 12126. R. Poetes, K. Holtzmann, K. Franze, U. Steiner, Phys. Rev. Lett. 105 (2010) 166104. M.A. Samaha, F.O. Ochanda, H.V. Tafreshi, G.C. Tepper, M. Gad-el-Hak, Rev. Sci. Instrum. 82 (2011) 045109. M. Sakai, T. Yanagisawa, A. Nakajima, Y. Kameshima, K. Okada, Langmuir 25 (2009) 13–16. J. Bico, C. Marzolin, D. Quéré, Europhys. Lett. 47 (1999) 220–226.
M.A. Samaha et al. / Journal of Colloid and Interface Science 395 (2013) 315–321 [15] N.J. Shirtcliffe, G. McHale, M.I. Newton, G. Chabrol, C.C. Perry, Adv. Mater. 16 (2004) 1929–1932. [16] J. Davies, D. Maynes, B.W. Webb, B. Woolford, Phys. Fluids 18 (2006) 087110. [17] C. Lee, C-H. Choi, C-J. Kim, Phys. Rev. Lett. 101 (2008) 064501. [18] B. Woolford, J. Prince, D. Maynes, B.W. Webb, Phys. Fluids 21 (2009) 085106. [19] A. Steinberger, C. Conttin-Bizonne, P. Kleimann, E. Charlaix, Nat. Mater. 6 (2007) 665–668. [20] L. Feng, Y. Zhang, J. Xi, Y. Zhu, N. Wang, F. Xia, L. Jiang, Langmuir 24 (2008) 4114–4119. [21] S.D. Bhagat, Y.-H. Kim, K.-H. Suh, Y.-S. Ahn, J.-G. Yeo, J.-H. Han, Microporous Mesoporous Mater. 112 (2008) 504–509. [22] H. Yang, Y. Deng, J. Colloid Interface Sci. 325 (2008) 588–593. [23] S. Lee, D. Kim, W. Hwang, Curr. Appl. Phys. 11 (2011) 800–804. [24] H.S. Hwang, N.H. Kim, S.G. Lee, D.Y. Lee, K. Cho, I. Park, ACS Appl. Mater. Interfaces 3 (2011) 2179–2183. [25] M.A. Samaha, F.O. Ochanda, H.V. Tafreshi, G.C. Tepper, M. Gad-el-Hak, Colloids Surf. A 399 (2012) 62–70. [26] B. Emami, T.M. Bucher, H.V. Tafreshi, D. Pestov, M. Gad-el-Hak, G.C. Tepper, Colloids Surf. A 385 (2011) 95–103. [27] M. Ma, Y. Mao, M. Gupta, K.K. Gleason, G.C. Rutledge, Macromolecules 38 (2005) 9742–9748. [28] F.O. Ochanda, M.A. Samaha, H.V. Tafreshi, G.C. Tepper, M. Gad-el-Hak, J. Appl. Polym. Sci. 123 (2012) 1112–1119. [29] M. Gad-el-Hak, Flow Control: Passive, Active, and Reactive Flow Management, Cambridge University Press, London, United Kingdom, 2000. [30] X.-Q. Feng, X. Gao, Z. Wu, L. Jiang, Q.-S. Zheng, Langmuir 23 (2007) 4892–4896. [31] Q. Pan, J. Liu, Q. Zhu, ACS Appl. Mater. Interfaces 2 (2010) 2026–2030.
[32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53]
321
B. Su, M. Li, Q. Lu, Langmuir 26 (2010) 6048–6052. G. McHale, M.R. Flynn, M.I. Newton, Soft Matter 7 (2011) 10100–10107. A.B.D. Cassie, S. Baxter, Trans. Faraday Soc. 40 (1944) 546–551. R.N. Wenzel, Ind. Eng. Chem. 28 (1936) 988–994. N.A. Patankar, Langmuir 20 (2004) 7097–7102. L. Barbieri, E. Wagner, P. Hoffmann, Langmuir 23 (2007) 1723–1734. C.W. Extrand, Langmuir 20 (2004) 5013–5018. C.W. Extrand, Langmuir 22 (2006) 1711–1714. Q.S. Zheng, Y. Yu, Z.H. Zhao, Langmuir 21 (2005) 12207–12212. C. Lee, C-J. Kim, Langmuir 25 (2009) 12812–12818. A. Marmur, Langmuir 19 (2003) 8343–8348. B. Liu, F.F. Lange, J. Colloid Interface Sci. 298 (2006) 899–909. M. Sbragaglia, A. Prosperetti, J. Fluid Mech. 578 (2007) 435–451. M.A. Samaha, H.V. Tafreshi, M. Gad-el-Hak, Phys. Fluids 23 (2011) 012001. T.M. Bucher, B. Emami, H.V. Tafreshi, M. Gad-el-Hak, G.C. Tepper, Phys. Fluids 24 (2012) 022109. B. Emami, H.V. Tafreshi, M. Gad-el-Hak, G.C. Tepper, Appl. Phys. Lett. 98 (2011) 203106. B. Emami, H.V. Tafreshi, M. Gad-el-Hak, G.C. Tepper, Appl. Phys. Lett. 100 (2012) 013104. B. Emami, H.V. Tafreshi, M. Gad-el-Hak, G.C. Tepper, J. Appl. Phys. 111 (2012) 064325. B.N. Nguyen, M.A.B. Meador, M.E. Tousley, B. Shonkwiler, L. McCorkle, D.A. Scheiman, A. Palczer, ACS Appl. Mater. Interfaces 1 (2009) 621–630. http://www.che.vt.edu/Wilkes/electrospinning/electrspinning.html. S.A. Hosseini, H.V. Tafreshi, Chem. Eng. Sci. 65 (2010) 2249–2254. S.A. Hosseini, H.V. Tafreshi, Powder Technol. 201 (2010) 153–160.