ARTICLE pubs.acs.org/Langmuir
Design and Fabrication of Teflon-Coated Tungsten Nanorods for Tunable Hydrophobicity Khedir R. Khedir,† Ganesh K. Kannarpady,*,† Hidetaka Ishihara,† Justin Woo,† Charles Ryerson,‡ and Alexandru S. Biris*,† †
Nanotechnology Center, University of Arkansas at Little Rock, 2801 South University Avenue, Little Rock, Arkansas 72204, United States ‡ Terrestrial and Cryospheric Sciences Branch Cold Regions Research & Engineering Laboratory Engineer Research and Development, Center U.S. Army Corps of Engineers, Hanover, New Hampshire 03755-1290, United States ABSTRACT: The nature of water interaction with tungsten nanorods (WNRs) fabricated by the glancing-angle deposition technique (GLAD)—using RF magnetron sputtering under various Ar pressures and substrate tilting angles and then subsequent coating with Teflon—has been studied and reported. Such nanostructured surfaces have shown strong water repellency properties with apparent water contact angles (AWCA) of as high as 160°, which were found to depend strongly upon the fabrication conditions. Variations in Ar pressure and the substrate tilting angle resulted in the generation of WNRs with different surface roughness and porosity properties. A theoretical model has been proposed to predict the observed high AWCAs measured at the nanostructure interfaces. The unique pyramidal tip geometry of WNRs generated at low Ar pressure with a high oblique angle reduced the solid fraction at the water interface, explaining the high AWCA measured on such surfaces. It was also found that the top geometrical morphologies controlling the total solid fraction of the WNRs are dependent upon and controlled by both the Ar pressure and substrate tilting angle. The water repellency of the tungsten nanorods with contact angles as high as 160° suggests that these coatings have enormous potential for robust superhydrophobic and anti-icing applications in harsh environments.
1. INTRODUCTION Geometrical modification of surfaces and subsequent surface treatment with low surface energy materials is a conventional strategy for producing hydrophobic [AWCA (apparent water contact angles) > 90°] or superhydrophobic (AWCA > 150°) surfaces from hydrophilic surfaces [WCA (water contact angle) < 90°] such as metals.15 However, with the high degree of surface roughness, beyond the threshold roughness4 and controlled topography,6 the conversion of high surface energy materials to exhibit water repellency properties710 with AWCA as high as 178° has also been reported.7 Similarly, the anti-icing property was observed on superhydrophobic surfaces with contact angles in the range of 150160° on polymer composites11 and Si nanostructures.12 Nevertheless, the robustness and longevity of the modified surfaces still pose major challenges for the overall development of superhydrophobic metallic surfaces to be used in harsh environmental applications. The modification of metallic surfaces such as W to exhibit superhydrophobic properties that could prevent ice formation11 in aerospace applications would have enormous impact both economically and from a safety perspective. In order to achieve such materials, the main challenge is to develop metallic surfaces, such as tungsten, with extremely high surface energy of more than 3000 mN m1. Reduction in the surface area’s exposure to the r 2011 American Chemical Society
environment is the key to minimizing the possibility of degradation of the water repellency property as a result of corrosion and wearing of the coating. Therefore, fabrication of metallic surfaces with both a low degree of roughness and controlled geometrical morphologies may advance their overall water repellency abilities. As directly related to wettability, roughness in both micro- and nanoscale and a combination of both can be imparted to the metallic surfaces using various methods,10,1315 including glancingangle deposition (GLAD),1 electrodeposition,9 chemical etching,8 femtosecond laser2 treatment, and solgel process.4 From the application point of view, fabrication cost and ease, scaling up abilities, and surface robustness are the measures of reliability of the generated rough surfaces produced by such methods. In order to address all of these considerations, the growth of morphologically controlled WNR surfaces by GLAD16 is proposed, since this method is capable of producing one-dimensional nanostructures with tunable surface morphology and porosity. To enhance the water repellency of metallic surfaces by generating nanorods using the GLAD technique, we have previously reported an AWCA of 138° for both Al and W nanorod Received: December 8, 2010 Revised: February 21, 2011 Published: March 15, 2011 4661
dx.doi.org/10.1021/la104891u | Langmuir 2011, 27, 4661–4668
Langmuir surfaces coated with a thin layer of silane.1 In addition, Fan et al.17 used the GLAD technique to fabricate Si nanorods with varying diameters. Hydrophobic surfaces with an AWCA of 142.7° were obtained on Si nanorods with heights superior to 500 nm, after treating the prepared surfaces with hydrofluoric acid. To our best knowledge, very limited research has been conducted using the GLAD technique to alter the metallic surface properties, such as roughness and porosity, by controlling the characteristics of the metal nanorods with the ability to produce scalable and robust superhydrophobic surfaces. In this work, WNRs with various surface morphologies and porosities were generated by RF magnetron sputtering, using the GLAD technique, under different Ar pressures and with various substrate tilting angles. The details about the influence of Ar pressure on the morphological control of WNRs were presented in our previous work.18 After an additional coating with a thin Teflon layer, the WNRs of various morphologies have shown water repellency properties with AWCAs ranging from 122.7° to 160°. Such findings indicate the ability to tailor the surface energy and the hydrophobic properties of such nanostructural films based on their morphologies and structural characteristics that can be directly correlated to the growth parameters. Scanning electron microscopy and atomic force microscopy were used to characterize the surface morphology of the generated WNR surfaces and were further used in the development of a theoretical model proposed to predict the observed high AWCA for the WNRs. To our knowledge, this excellent degree of control over the morphology and structural properties of such metal nanorods during the growth process and their direct impact on the corresponding surface energy of the films, resulting in tunable interactions with the water droplets, has never been shown before. The novel results highlighted by this work could find significant applications in the area of water or ice mitigation, biomedical implants, high-sensitivity sensing, and/or aeronautics.
2. EXPERIMENTAL PROCEDURE 2.1. Tungsten Nanorod Fabrication. First, glass substrates (2.5 cm 2.5 cm) were cleaned with both acetone and methanol in a sonication bath for 10 min each. The cleaned substrates were blown with nitrogen gas and mounted on a rotatable substrate holder, 150 mm away from the target. Then the chamber was pumped down using a cryopump supported by a mechanical pump until it reached the base pressure of 5.0 107 Torr. The W target with purity of 99.9% purchased from ACI Alloys, Inc., was used. The depositions were carried out under a constant power density of 7.64 W cm2, using RF magnetron sputtering. In the GLAD technique, the atoms are sputtered onto the substrate with a high oblique angle between the substrate surface normal and the target surface normal, in addition to the substrate rotation with an angular speed of 30 rpm around its surface normal. The Ar gas was introduced into the chamber at working gas pressure through an injection ring just above the target with a constant flow rate of 10 sccm. In this work, two distinct factors of deposition were manipulated to generate WNR surfaces with significant differences in surface morphology and porosity. First,WNR thin films were deposited on the glass substrate under various Ar pressures of 0.38, 0.5, 0.7, 1.0, 5, 10, and 20 mTorr and a constant substrate tilting angle of 85° for 60 min. Second, WNRs were generated at two different oblique angles of 85° and 89°, but constant Ar pressure of 1.0 mTorr, with three different thicknesses of 200, 400, and 600 nm for each tilting angle. All of the depositions were carried out with no intentional heating of the substrates.
ARTICLE
2.2. Surface Modification Using Teflon Nanolayer. The surface energy of the as-deposited WNR samples prepared under various Ar pressures and substrate tilting angles was modified by coating with a thin layer of Teflon AF2400 using an effusion cell. The coating was carried out with no substrate holder tilting (normal incidence) under the chamber pressure of less than 3 107 Torr. In order for the samples to be chemically modified under the same conditions, the whole set of the as-deposited WNRs was coated with Teflon AF2400, simultaneously. 2.3. Contact Angle Measurements. The AWCA was measured by the drop method using an EasyDrop (DSA1) device (Kruss Co.). For each sample, distilled water droplets of 2 μL were gently dispensed, using a computer-controlled automated syringe, in five random places. The water droplet images were digitally captured via a CCD camera with the capability of recording 60 fps at a 780 580 resolution. For each sample, five corresponding contact angles of the water droplet images were averaged and presented with minimum standard deviation possible. This process was repeated for selected samples at intervals of several days, and consistent results were observed. All of the measurements were carried out at room temperature. 2.4. Surface Morphology. A scanning electron microscope (JEOL SEM7000FE) was used to characterize the surface topography of the asdeposited WNRs on the glass substrate under various Ar pressures and substrate tilting angles. The uniformity of the WNRs was investigated by taking images from three different locations chosen randomly over the surface of the WNR films. In addition, the surface roughness and surface morphology analyses were carried out using a Nanoscope 3100 atomic force microscope. The tapping mode was utilized to scan the surface of the WNRs. The roughness factors (the ratio of increase in the surface area) of the three scanned locations were averaged, and a reasonable correlation among the results was achieved.
3. RESULTS AND DISCUSSIONS 3.1. Tungsten Nanorods Synthesis and Their Water Hydrophobicity Properties. 3.1.1. WNRs Generated under Various Ar Pressures. In the first part of this work, the as-deposited WNRs
fabricated by the GLAD technique under various Ar pressures exhibited various morphologies and surface roughness, in addition to the variation in their porosity, as shown in Figure 1. The top-view SEM images for WNRs deposited at Ar pressures lower than 1 mTorr are not shown here; however, they are comparable to the images of WNRs obtained at 1 mTorr in terms of surface morphology and porosity. The increase in Ar pressure caused a reduction in the lateral size of the WNRs and a visible increase in their density-forming large agglomerates.18 The WNRs generated by the GLAD technique were observed to present naturally occurring pyramidal tips at their top ends. These pyramidal ends are more significant and visible for the isolated WNRs fabricated at low Ar pressures but were found to gradually lose their geometrical structures as the Ar pressure increased, as shown in Figure 1a. This fact has been reported by other researchers, as well.1,19 After chemical modification of the as-deposited WNRs by coating them with a thin layer of Teflon, tunable water repellent metallic surfaces with AWCAs ranging from 122.7° to 150.7° were obtained. The graphic representation of the steps followed to fabricate WNRs with hydrophobic properties is presented in Figure 2. The systematic increase in the water repellency of the prepared surfaces was due to the control of surface morphology and porosity of the WNR films by varying the Ar pressure during deposition. As can be seen from the top view SEM images shown in Figure 1A, the as-deposited WNRs generated at relatively low Ar pressures possess larger lateral sizes but lower densities with isolated nanopillars.18 However, the opposite scenario occurred 4662
dx.doi.org/10.1021/la104891u |Langmuir 2011, 27, 4661–4668
Langmuir
ARTICLE
Figure 1. (a) Top-view SEM images of WNRs deposited on glass substrate at various Ar pressures. The insets are the corresponding AFM images of the surface morphology of WNRs. The dimensions in AFM images are X = Y = 1 μm and Z = 100 nm. (b) Top-view SEM images of WNRs deposited on glass substrate at constant Ar pressure of 1 mTorr and different substrate tilting angle of 85° and 89° with height profile of 200, 400, and 600 nm, respectively. The dimensions in AFM images are X = Y = 1 μm and Z = 100 nm.
Figure 2. Schematic representation of steps to generate WNRs with water repellent properties. 4663
dx.doi.org/10.1021/la104891u |Langmuir 2011, 27, 4661–4668
Langmuir
Figure 3. (a) The Wenzel model and (b) the Cassie model for θ < 90° (hydrophilic), θ = 90°, and θ > 90° (hydrophobic).
ARTICLE
pyramidal tips of WNRs can be seen in Figure 1b. As shown in this figure, the films deposited at an 89° tilt angle showed much better porosity and sharpness in the pyramidal tips than did those deposited at an 85° tilt angle. Eventually, after coating the surfaces of the prepared samples with a thin layer of Teflon AF2400, a significant enhancement in their water repellent behavior was observed with AWCAs ranging from 150° to 160°. The highest contact angle of 160° was observed for the nanorods with a film thickness of 600 nm deposited at an 89° tilt angle due to the reduction in the density of nanorods, as seen in the corresponding SEM picture in Figure 1b. The samples prepared at a substrate tilting angle of 85° showed no significant decrease in the density of nanorods (no solid fraction decrease) with the increase in the height. As a result, their water repellency properties remained almost the same, with an AWCA of around 150°. In this work, the variation of deposition pressure and substrate tilting angle enabled us to generate WNRs with various surface morphologies and consequent controllable water repellency starting from the moderately hydrophobic to the superhydrophobicity after chemical treatment of their surfaces. In order to understand the mechanism of water interaction with these nanomodified surfaces and their nature in adopting a specific regime, the predicted AWCAs were analyzed by both the Wenzel and Cassie methods as the classic proposed models for rough surfaces. Such studies can be carried out in terms of two parameters: the increase in surface area (roughness) and the ratio of the top solid fraction of the nanopillars. 3.2. Theoretical Wenzel and CassieBaxter Models Overview. The study of the interaction of liquid with the rough surfaces was fundamentally proposed by both Wenzel21 and CassieBaxter.22 In the Wenzel model, the rough surface area is in direct contact with the liquid, and the wetabillity of the surface, AWCA, can be predicted using the following equation: cos θW ¼ r cos θY
with the WNRs fabricated at high Ar pressures: their density increased to the extent of agglomeration—eventually leading to the formation of thin films at Ar pressure of 20 mTorr. Moreover, the side view SEM images that are not shown here revealed that the height of the fabricated WNRs was less than 200 nm. To our knowledge, high AWCAs in the vicinity of superhydrophobicity on nanopillar surfaces with height profiles of less than 200 nm have not been reported previously. 3.1.2. WNRs Generated at Different Substrate Tilting Angles with Different Thicknesses. In the second part of this work, another set of samples was generated by changing the substrate tilting angle (85° and 89°) under a constant Ar pressure of 1.0 mTorr. Since the WNRs generated at 1 mTorr showed a higher degree of alignment and isolation, we chose this pressure to control the nanostructure further by varying the tilting angle. In addition, the thin films were deposited with different thicknesses of 200, 400, and 600 nm at both substrate tilting angles. An increase in the substrate tilting angle causes the lateral component of the flux to shadow more and more and hence the vertical component of the flux to grow more selectively.20 Also, with the increase in the height of WNRs, more premature WNRs failed to develop fully during deposition. Consequently, taller nanorods grow at the expense of shorter nanorods. The two factors of extreme substrate tilting angle and the increasing heights of the nanorods promoted the fabrication of WNRs with significant enhancement in the porosity and sharpness of their pyramidal tips. The gradual variation in the porosity and sharpness in the
ð1Þ
where θW is the AWCA and θY the intrinsic WCA for the corresponding flat surfaces, which can be deduced using the Young equation. The roughness factor, r, is the ratio of the real over apparent area of the surface. The homogeneity of the interaction at the interface of both the liquid drop and rough surface has limited the validity of the Wenzel model to surfaces with hydrophilic and mild hydrophobic properties. For a better illustration of the Wenzel model and the influence of roughness on the AWCA, Figure 3a shows how hydrophilic surfaces with θ < 90° become more hydrophilic with increase in roughness, while hydrophobic surfaces with θ > 90° tend to have enhanced hydrophobicity. At a specific degree of roughness, which is also known as the threshold roughness, air pockets may be trapped between the solid voids under the liquid, promoting composite interfaces of liquidsolid and liquidair. In that case, the CassieBaxter model will be applied: cos θCB ¼ rs fs cos θY þ fa cos θa
ð2Þ
where fs and fa are the fraction of solidliquid and airliquid interfaces, respectively, and rs is the roughness factor over the top of the solid fraction that is in contact with water droplet. Also fa = 1 fs and θa = 180° as the contact angle at the airliquid interface. In the case of total wetting, fs = 1 and rs = r, and the Wenzel model will be recovered. From eq 2, it can be seen that the solid fraction fs, which is the counterpart of the roughness in the 4664
dx.doi.org/10.1021/la104891u |Langmuir 2011, 27, 4661–4668
Langmuir
Figure 4. The variation of AWCA on the surfaces of chemically modified WNRs with different roughness factors generated at various Ar pressures of 1.0, 5.0, 10, and 20 mTorr versus AWCAs predicted by the Wenzel model.
Figure 5. Illustration of image analysis steps in the developed Matlab program to estimate the solid fraction of WNR surfaces: (a) top-view SEM image for WNRs deposited at 1.0 mTorr and (b) a binary image, showing white areas (objects) and black areas (spaces), after background subtraction and contrast adjustment.
Wenzel model, controls the contact angle variation. With the same strategy that we followed above in the case of the Wenzel model, the variation of both a hydrophilic surface and a hydrophobic surface with the decrease in solid fraction is shown in Figure 3b. Unlike the Wenzel model, with both types of surfaces, the decrease in solid fraction fs caused an increase in the measured AWCA. Therefore, by controlling only the solid fraction, the hydrophilic surfaces can be modified to exhibit water repellency phenomena without chemical modification of the surface. 3.3. Proposed Theoretical Model for the Prediction of the Experimentally Measured AWCAs. As mentioned above, the Wenzel model is largely a reliable approach to predict the AWCAs for surfaces with hydrophilic or moderate hydrophobic properties. Therefore, it would be more reasonable to start first with the predictions that can be made by this model with the associated roughness obtained on the prepared WNR surfaces. The roughness factors of the WNRs generated under various Ar pressures and the associated AWCAs predicted by the Wenzel model against the deposition pressures are plotted and shown in Figure 4. The observed AWCAs for the WNRs generated by the onestep GLAD technique are much higher compared to the ones predicted by the Wenzel model (θY = 120°), for a flat surface coated with Teflon), except for the samples deposited at 20 mTorr. Also, from Figure 3a, according to the Wenzel model, a
ARTICLE
more than 70% increase of surface area (r > 1.7) is required to reach the limit of θ ∼ 150°. AWCAs of 150° on the chemically modified WNRs deposited at 1 mTorr in both the cases of substrate tilting angles (85° and 89°) with the thickness of 200 nm were obtained with only a 30%50% increase in the surface area (1.31.5 roughness factor). Therefore, the increase in the height of WNRs, which definitely increases their surface roughness, would not promote the Wenzel state. To the contrary, that would enhance the stability of the metastable Cassie state by increasing the energy barrier between the two states. As a result, it would be insufficient to explain the nature of interactions between the water droplets and the nanostructural WNR surfaces with the Wenzel model only. It can be deduced, therefore, that there is a strong possibility of complex interactions (excluding the films deposited at 20 mTorr) that could also be explained by the Cassie model. In the Cassie model, the solid fraction fs is the major factor predicting the AWCA values and “stickiness” of water droplet adhesion to the surface. With the progress in understanding the interaction of the liquid with the surface, the Cassie regime is also subdivided into two categories: the lotus effect (high AWCA but low adhesion)23 and the petal effect (high AWCA and also high adhesion to the surface).24 The limits of stability and metastability of this state are still not very well understood; however, although very important, neither the stickiness of water droplets to the top surfaces of the prepared WNRs nor measurement of the contact angle hysteresis (different between receding and advancing contact angle) were addressed in this study. For investigating the validity of the Cassie state occurrence, the solid fraction factor must be determined. For this purpose, image analysis technique was utilized to determine the solid fraction factors of the WNR surfaces. This was carried out by converting the grayscale top-view SEM images to binary images after subtracting the background and adjusting the image contrast. In the binary image, the white areas represent the flat topview of WNRs, and the black areas represent spaces among the objects. Eventually, the solid fraction factor was obtained by the summation of all the areas covered by the objects (white areas) divided by the entire area of the image. Moreover, small areas with poor brightness were excluded, because those areas represent nanorods that, during the growth process, were terminated prematurely and did not reach the surface. This procedure was accomplished by developing a Matlab program using functions from the image processing toolbox available in Matlab 7.9.0 (R2009b). Fan et al.17 followed a similar strategy to calculate the image threshold of the top-view SEM images in order to estimate the solid fraction factor of the silicon nanorods fabricated by the GLAD technique. A typical example of the image conversion steps obtained through the use of the Matlab program developed in this work is shown in Figure 5. To examine the possibility that the Cassie regime might explain the results obtained for the WNR surfaces, the calculated solid fractions obtained by the described image processing approach were substituted in eq 2. The AWCAs predicted by the Cassie model associated with their solid fraction factors are shown in Figure 6. It can be noticed that the predicted AWCAs are still lower than those observed on the surface of WNRs— particularly for the WNRs fabricated at low Ar pressures, from 0.38 to 5.0 mTorr, and an extreme oblique angle of 89° with various thicknesses, as well. This indicates that the water droplet sits partially on the very top portion of the WNRs’ tips with a solid fraction less than the one obtained through the image 4665
dx.doi.org/10.1021/la104891u |Langmuir 2011, 27, 4661–4668
Langmuir
Figure 6. (a) The observed AWCAs on the surfaces of chemically modified WNRs with different solid fraction factors generated at various Ar pressures of 1.0, 5.0, 10, and 20 mTorr versus AWCAs predicted by both the Cassie model and modified Cassie model. (b) The observed AWCAs on the surfaces of chemically modified WNRs with different solid fraction factors generated at constant Ar pressure of 1.0 mTorr and substrate tilting angle of 89° with thicknesses of 200, 400, and 600 nm versus AWCAs predicted by both the Cassie model and modified Cassie model.
analysis technique, which considers the top of nanorods as a flat surface with (rs = 1). From the SEM images of the WNRs generated under various Ar pressures and substrate tilting angles resulting in nanorods with consequent different surface morphologies (shown in Figure 1), the natural pyramidal tip of WNRs can be clearly seen, particularly for the isolated WNRs deposited at low Ar pressures of 1.0 mTorr. However, the sharpness of the pyramidal tip is more noticeable with the increase in the height of WNRs deposited at the substrate tilting angle of 89°. Therefore, in addition to the increase in the degree of porosity of the WNRs deposited at an extreme oblique angle, the naturally sharp pyramidal tip of WNRs with the height of 400 and 600 nm has contributed significantly in decreasing the effective solid fraction and consequently increasing the AWCA to as high as160°. For WNRs deposited at high Ar pressures of more than 1 mTorr, the pyramidal tip has started to diminish and almost disappears at 20 mTorr. It is also noticeable that the AWCAs predicted by both the Wenzel and Cassie models are
ARTICLE
comparable, which means that the solid fractions determined by image analysis techniques are very reasonable. 3.4. Effect of the WNRs Pyramidal Tips on the AWCAs. The AWCAs predicted by the Cassie model, assuming the top of nanopillars as flat triangular areas, are still lower than the measured AWCAs over the surfaces of prepared samples. Therefore, we believe that the natural pyramidal tip of WNRs is responsible for the higher AWCAs by changing the effective solid fraction in contact with water droplets. In a geometrical study, Patankar25 has suggested the design of periodic pillars with inclined side walls to lower the top solid fraction of the pillars. A significant decrease in the solid fraction of the proposed geometrical design can be noticed from the model that has been proposed to predict their solid fraction, compared to the same pillars with vertical sides. The effect of solid fraction due to the top geometrical design has been reported in some other studies, as well. Oner and McCarthy26 have shown that the pillars with a star cross-section pin more efficiently in the liquid meniscus than pillars with circular cross sections. Gao and McCarthy27 have reported perfectly hydrophobic surfaces with both advancing and receding contact angles of 180° on the surface of a loose network of thin fibers. They argue that the elasticity of the fabricated network might have contributed to the unexpectedly high AWCA. To estimate the solid fraction of the WNRs with a pyramidal tip, we used the geometrical approach that has been followed in other studies for various geometries, such as square pillars28 and circular pillars.14 For simplicity, we assumed that the pyramid has a simple tetrahedral shape with four equilateral triangles with dimensions very close to the diameter d of the nanorod, as illustrated in Figure 7a. Therefore, the solid fraction for pillars with triangular base tips and spaces S in between can be represented as follows: pffiffiffi 3 2 d 4 ð3Þ fs ¼ pffiffiffi ! 2 3 Sþ d Sþ d 3 2 The surface area of the tetrahedral pyramidal tip is 3 times the surface area of its base, so the solid fraction determined for the WNRs, generated at different Ar pressures, by using an image analysis technique is for nanopillars with flat triangular tips. Therefore, less than one-third of the pyramidal tip would be occupied by the water droplet to predict the observed AWCAs on the surface of prepared samples. To determine the reduction in the solid fraction due to the pyramidal tip of WNRs, which may represent the effective solid fraction occupied by the water droplet and promoting higher AWCA, the eq 3 can be rewritten as follows: pffiffiffi 2h 2 3 d tan R 4 fs ¼ pffiffiffi ! ð4Þ 3 2 Sþ d Sþ d 2 3 The most practical way to predict the effective solid fraction occupied by the water droplet on the surface area of the pyramidal tip is to gradually decrease the water droplet occupation on its surface area by adjusting the parameter h, which 4666
dx.doi.org/10.1021/la104891u |Langmuir 2011, 27, 4661–4668
Langmuir
ARTICLE
Figure 7. Schematic illustration of (a) the unit area of flat triangular nanopillars, with the two situations (b and c) of water occupation on the pyramidal tip.
represents the height of a triangular face of the tetrahedral pyramid, so we start from the base of the triangle at h = 0, which indicates that only one-third of the surface area of the pyramid (tetrahedral pyramid has three equilateral triangles on the surface area) has been occupied by the water droplet (Figure 7b), and adjust to the top of the pyramid at h = 31/2/2d, which represents almost zero effective solid fraction (Figure 7c). Eventually, the modified CassieBaxter model due to the proposed solid fraction formula can be represented:
cos θCB
pffiffiffi 2h 2 3 d tan R 4 ¼ pffiffiffi !ð1 þ cos θY Þ 1 2 3 Sþ d Sþ d 3 2
ð5Þ
After examining the modified Cassie model for different values of h, we learned that, with a 30% decrease in the solid fraction of the triangular base, which represents a 77% decrease in the surface area of pyramidal tip, the predicted AWCAs were in good agreement with observed AWCAs for the two rough films prepared under deposition pressures of 1.0 and 5.0 mTorr, as demonstrated in Figure 6A. Therefore, only 23% of the pyramidal tip was occupied by the water droplet, which means suspending the water droplet at the very top of the pyramidal tip. However, the model cannot be applied for the samples generated at 10 mTorr and above, where there are no isolated nanorods with significant pyramidal tips and relatively well-defined space/diameter ratios. Meanwhile, the high OAWCA on the sample generated at 10 mTorr may be due to the large size of the voids among relatively large bunches of WNRs, which generally promotes the coexistence of the two states.29 With the second variable used to control the growth of the nanostructures, the extreme tiling angle of 89°, the increase in the thickness of the films facilitated the growth of taller nanorods at the expense of shorter ones. This increased the porosity of the films with the decrease in the density of the nanorods that reached the surface of the film. However, the solid fractions of the WNR surfaces with heights of 200, 400, and 600 nm determined by image analysis technique were much larger, as indicated by predicted Cassie contact angles that were lower than the observed contact angle (Figure 6b). As shown in the figure, the predicted AWCAs with the modified Cassie model were in excellent agreement due to well-isolated WNRs and their noticeable pyramidal tip that was promoted with the increase in the height of prepared surfaces. It turned out that the effective solid fractions, the average pyramidal surface area occupied with water,
were around 23%, 20%, and 18% of the surface area of geometrical pyramidal tip for WNR surfaces with height profiles of 200, 400, and 600 nm, respectively. For the samples deposited with different thicknesses but at a constant deposition pressure of 1.0 mTorr and a substrate tilting angle of 85°, a very slight difference in the OAWCAs of 150° 151° was observed. This fact can be understood simply by looking at the SEM images shown in Figure 1b, where the first row of images show no significant decrease in solid fraction (decrease in porosity) or increase in the sharpness of the nanorods’ pyramidal tips. The quasiconstant OAWCAs on these surfaces—despite a significant increase in their surface roughness due to the increase in the height of WNRs from 200 to 600 nm— is another concrete piece of evidence that these samples have adopted the composite Cassie state without being affected by the increase in their roughness. As we have demonstrated in this study, the nanoscale WNRs with pyramidal tips exhibited OAWCAs as high as 160o by utilizing the two factors of Ar pressure and substrate tilting angle. Also, the proposed model was able to predict the reduction in solid fraction due to the pyramidal tips of WNRs; consequently, the predicted AWCAs were very consistent with the measured values. Therefore, the possibility of water droplets touching the sides of WNRs, which may not be coated with Teflon, was very low. Otherwise, there would be a change in the liquid/metal interface that would reduce the AWCAs and increase the possibility for the transition from composites to the homogeneous state. Therefore, the geometric design of the tops of pillars would have a significant impact on the reduction of effective solid fraction and consequent robust hydrophobic surfaces.
4. CONCLUSIONS In this work, nanoscale WNRs generated under various Ar pressures and substrate tilting angles by using GLAD technique have shown tunable hydrophobic properties after their surfaces were modified with a thin layer of Teflon. The natural pyramidal tips and variation in the WNRs’ surface morphology and porosity contributed to the enhancement of water repellency for these metallic surfaces with AWCAs ranging from 122.7° to 160°. A proposed theoretical model was able to predict the high AWCAs measured for the WNRs generated particularly at low Ar pressures and extreme substrate tilting angles due to their significant pyramidal tip and isolation. It is concluded that controlling both the space/diameter ratio of nanorods and lowering their solid fractions, along with the presence of pyramidal tips, would have a significant impact on the fabrication of robust superhydrophobic metallic surfaces. 4667
dx.doi.org/10.1021/la104891u |Langmuir 2011, 27, 4661–4668
Langmuir
ARTICLE
’ AUTHOR INFORMATION Corresponding Author
*G.K.K.: e-mail,
[email protected]; tel, 501-569-8067; fax, 501-683-7601. A.S.B.: e-mail,
[email protected]; tel, 501-5519067; fax, 501-683-7601.
’ ACKNOWLEDGMENT Financial support from the U.S. Army (ERDC Cooperative Agreement Number: W912HZ-09-02-0008), the Arkansas Science & Technology Authority (Grant # 08-CAT-03), and the Department of Energy (DE-FG36-06GO86072) and National Science Foundation (NSF/EPS-1003970) is greatly appreciated. The editorial assistance of Dr. Marinelle Ringer is also acknowledged. ’ REFERENCES (1) Kannarpady, G. K.; Sharma, R.; Liu, B.; Trigwell, S.; Ryerson, C.; Biris, A. S. Appl. Surf. Sci. 2010, 256, 1679–1682. (2) Wu, B.; Zhou, M.; Li, J.; Ye, X.; Li, G.; Cai, L. Appl. Surf. Sci. 2009, 256, 61–66. (3) Wang, H.; Dai, D.; Wu, X. Appl. Surf. Sci. 2008, 254, 5599–5601. (4) Rao, A. V.; Latthe, S. S.; Dhere, S. L.; Pawar, S. S.; Imai, H.; Ganesan, V.; Gupta, S. C.; Wagh, P. B. Appl. Surf. Sci. 2010, 256, 2115–2121. (5) Kuan, W.; Chen, L. Nanotechnology 2009, 20, 035605–035613. (6) Abdelsalam, M. E.; Bartlett, P. N.; Kelf, T.; Baumberg, J. Langmuir 2005, 21, 1753–1757. (7) Hosono, E.; Fujihara, S.; Honma, I.; Zhou, H. J. Am. Chem. Soc. 2005, 127, 13458–13459. (8) Chen, Z.; Guo, Y.; Fang, S. Surf. Interface Anal. 2010, 42, 1–6. (9) Bhattacharya, P.; Gohil, S.; Mazher, J.; Ghosh, S.; Ayyub, P. Nanotechnology 2008, 19, 075709–075714. (10) Hang, T.; Hu, A.; Ling, H.; Li, M.; Mao, D. Appl. Surf. Sci. 2010, 256, 2400–2404. (11) Cao, L.; Jones, A. K.; Sikka, V. K.; Wu, J.; Gao, D. Langmuir 2009, 25, 12444–12448. (12) Mishchenko, L.; Hatton, B.; Bahadur, V.; Taylor, J. A.; Krupenkin, T.; Aizenberg, J. ACS Nano 2010, 4, 7699–7707. (13) Kietzig, A.; Hatzikiriakos, S. G.; Englezos, P. Langmuir 2009, 25, 4821–4827. (14) Cha, T.; Yi, J. W.; Moon, M.; Lee, K.; Kim, H. Langmuir 2010, 26, 8319–8326. (15) Hye-Mi, Bok; Tae-Yeon, Shin; Park, S. Chem. Mater. 2008, 20, 2247–2251. (16) Robbie, K.; Friedrich, L. J.; Dew, S. K.; Smy, T.; Brett, M. J. J. Vac. Sci. Technol. 1995, 13, 1032–1035. (17) Fan, J.; Tang, X.; Zhao, Y. Nanotechnology 2004, 15, 501–504. (18) Khedir, K. R.; Kannarpady, G. K.; Ishihara, H.; Woo, J.; Ryerson, C.; Biris, A. S. Phys. Lett. A 2010, 374, 4430–4437. (19) Karabacak, T.; Mallikarjunan, A.; Singh, J. P.; Ye, D.; GwoChing, W.; Toh-Ming, L. Appl. Phys. Lett. 2003, 83, 3096–3098. (20) Zhao, Y.; Ye, D.; Wang, G.; Lu, T. Proc. SPIE 2003, 5219, 59–73. (21) Wenzel, R. N. Ind. Eng. Chem. 1936, 28, 988–994. (22) Cassie, A. B. D.; Baxter, S. Trans. Faraday Soc. 1944, 40, 0546–0550. (23) Marmur, A. Langmuir 2004, 20, 3517–3519. (24) Feng, L.; Zhang, Y.; Xi, J.; Zhu, Y.; Wang, N.; Xia, F.; Jiang, L. Langmuir 2008, 24, 4114–4119. (25) Patankar, N. A. Langmuir 2004, 20, 7097–7102. (26) Oner, D.; McCarthy, T. J. Langmuir 2000, 16, 7777–7782. (27) Gao, L.; McCarthy, T. J. J. Am. Chem. Soc. 2006, 128, 9052–9053. (28) Patankar, N. A. Langmuir 2003, 19, 1249–1253. (29) Koishi, T.; Yasuoka, K.; Fujikawa, S.; Ebisuzaki, T.; Zeng, X. C. Proc. Natl. Acad. Sci. U. S. A. 2009, 106, 8435–8440. 4668
dx.doi.org/10.1021/la104891u |Langmuir 2011, 27, 4661–4668
