Asymmetric Wetting Hysteresis on Hydrophobic Microstructured


Asymmetric Wetting Hysteresis on Hydrophobic Microstructured...

0 downloads 84 Views 2MB Size

pubs.acs.org/Langmuir © 2009 American Chemical Society

Asymmetric Wetting Hysteresis on Hydrophobic Microstructured Surfaces Craig Priest,* Trent W. J. Albrecht, Rossen Sedev, and John Ralston Ian Wark Research Institute, ARC Special Research Centre for Particle and Material Interfaces, University of South Australia, Mawson Lakes, South Australia 5095, Australia Received December 23, 2008. Revised Manuscript Received February 9, 2009 The wetting behavior of hydrophobic, microstructured surfaces containing arrays of pillars or holes has been investigated. The size of the surface features was fixed (20 μm), while their separation was varied to adjust the area fraction (0-80%). The wettability of structured surfaces for liquids resting in the Cassie state is strongly dependent on the continuity of the solid component. Microstructured square pillars and holes showed distinct, asymmetric wetting hysteresis, consistent with our previous observations on flat, chemically heterogeneous surfaces. Furthermore, clear trends for the magnitude of contact angle hysteresis versus area fraction for the two types of microstructured surfaces are evident. The pinning energy associated with these surface features is estimated.

Introduction The wettability of heterogeneous surfaces is of practical importance, due to the inherently nonideal surfaces encountered in a variety of processes, including coating, water repellency, mineral flotation, and capillary flow. For this reason, many studies have been devoted to determining a relationship between surface features (whether physical or chemical) and the contact angle.1-15 As the relationship between surface features and the contact angle becomes clearer, a surface’s properties may be increasingly tailored to fit the desired application, for example, repellency11,16-19 responsive surfaces,20-23 *To whom correspondence should be addressed. E-mail: craig. [email protected]. (1) Wenzel, R. N. Ind. Eng. Chem. 1936, 28, 988. (2) Cassie, A. B. D. Discuss. Faraday Soc. 1948, 3, 11. (3) Johnson, R. E. J.; Dettre, R. H. J. Phys. Chem. 1964, 68, 1744. (4) Neumann, A. W.; Good, R. J. J. Colloid Interface Sci. 1972, 38, 341. :: (5) Raphael, E.; de Gennes, P. G. J. Chem. Phys. 1989, 90, 7577. (6) Shanahan, M. E. R.; Di Meglio, J. M. J. Adhes. Sci. Technol. 1994, 8, 1371. (7) Naidich, Y. V.; Voitovich, R. P.; Zabuga, V. V. J. Colloid Interface Sci. 1995, 174, 104. (8) Imabayashi, S.-I.; Gon, N.; Sasaki, T.; Hobara, D.; Kakiuchi, T. Langmuir 1998, 14, 2348–_2351. (9) Rousset, E.; Baudin, G.; Cugnet, P.; Viallet, A. J. Imaging Sci. Technol. 2001, 45, 517. (10) Kanoufi, K.; Combellas, C.; Shanahan, M. E. R. Langmuir 2003, 19, 6711. :: (11) Dorrer, C.; Ruhe, J. Langmuir 2006, 22, 7652. (12) Kusumaatmaja, H.; Yeomans, J. M. Langmuir 2007, 23, 6019. (13) Priest, C.; Sedev, R.; Ralston, J. Phys. Rev. Lett. 2007, 99, 026103. (14) Yeh, K.-Y.; Chen, L.-J.; Chang, J.-Y. Langmuir 2008, 24, 245. (15) Priest, C.; Stevens, N.; Sedev, R.; Skinner, W.; Ralston, J. J. Colloid Interface Sci. 2008, 320, 563. (16) Lafuma, A.; Quere, D. Nat. Mater. 2003, 2, 457. (17) Wagterveld, R. M.; Berendsen, C. W. J.; Bouaidat, S.; Jonsmann, J. Langmuir 2006, 22, 10904. (18) Tuteja, A.; Choi, W.; Ma, M.; Mabry, J. M.; Mazzella, S. A.; Rutledge, G. C.; McKinley, G. H.; Cohen, R. E. Science 2007, 318, 1618. (19) Nosonovsky, M.; Bhushan, B. Ultramicroscopy 2007, 107, 969. (20) Abbott, S.; Ralston, J.; Reynolds, G.; Hayes, R. Langmuir 1999, 15, 8923. (21) Sun, T.; Wang, G.; Feng, L.; Liu, B.; Ma, Y.; Jiang, L.; Zhu, D. Angew. Chem., Int. Ed. 2004, 43, 357. (22) Lake, N.; Ralston, J.; Reynolds, G. Langmuir 2005, 21, 11922. (23) Dhindsa, M. S.; Smith, N. R.; Heikenfeld, J.; Rack, P. D.; Fowlkes, J. D.; Doktycz, M. J.; Melechko, A. V.; Simpson, M. L. Langmuir 2006, 22, 9030.

Langmuir 2009, 25(10), 5655–5660

and microfluidics.24-29 However, despite advancements in wetting research, contact angle hysteresis has remained a central wetting phenomenon without substantial predictive theory. In this paper, we study the wettability of rough, microstructured surfaces where the liquid rests on top of the surface structure (Cassie state). We pay particular attention to the asymmetry of the contact angle hysteresis on these surfaces and compare this behavior with that observed on flat, chemically heterogeneous surfaces. It is well-known that, at thermodynamic equilibrium, the contact angle (equilibrium angle), θ0 can be related to the liquid-vapor, γLV, solid-liquid, γSL, and solid-vapor, γSV, interfacial tensions, according to Young’s equation, provided the surface is rigid, flat, homogeneous, and inert to both fluid phases.30 γLV cos θ0 ¼ γSV -γSL

ð1Þ

Surfaces typically exhibit inhomogeneity to various degrees. In order to account for roughness, Wenzel used a roughness factor, r, given by the ratio of the actual surface area to the projected surface area of a rough surface.1 cos θ ¼ r cos θ0

ð2Þ

where θ is the observed contact angle on the rough surface and θ0 is the Young angle on the planar surface. Wenzel’s equation assumes that the liquid completely fills the underlying surface structure, a situation termed the Wenzel state.16 According to eq 2, θ > θ0 whenever θ0 > π/2 and θ < θ0 whenever θ0 < π/2. A special case exists on rough surfaces when air is trapped in the surface structure. In this case, the surface is effectively (24) Zhao, B.; Moore, J.; Beebe, D. J. Science 2001, 291, 1023. (25) Zhang, J.; Kwok, D. Y. Langmuir 2006, 22, 4998. (26) Choi, C.-H.; Ulmanella, U.; Kim, J.; Ho, C.-M.; Kim, C.-J. Phys. Fluids 2006, 18, 087105. (27) Besson, E.; Gue, A.-M.; Sudor, J.; Korri-Youssoufi, H.; Jaffrezic, N.; Tardy, J. Langmuir 2006, 22, 8346. (28) Fidalgo, L. M.; Abell, C.; Huck, W. T. S. Lab Chip 2007, 7, 984. (29) Baret, J.-C.; Decre, M. M. J. Langmuir 2007, 23, 5200. (30) Young, T. Philos. Trans. R. Soc. London 1805, 95, 65.

Published on Web 3/10/2009

DOI: 10.1021/la804246a

5655

Article

Priest et al.

a composite one, comprising distinct regions of air (θ0 = 180) and solid. This configuration is called the Cassie state.16 The observed contact angle for liquids wetting chemically composite surfaces can be related to the weighted average of the surface components according to the Cassie equation: cos θ ¼ φ1 cos θ1 þ ð1 -φ1 Þ cos θ2

ð3Þ

where θ is the observed contact angle on the chemically heterogeneous surface, φ1 is the surface area fraction of component 1, θ1 and θ2 are the equilibrium contact angles for components 1 and 2, respectively.2 Cassie’s equation is frequently applied to superhydrophobic surfaces, where air is trapped in the surface structure and the contact angle is elevated above 150, accompanied by little hysteresis.16 Superhydrophobicity has become a central theme in wetting research for its great potential in many applications, including microfluidics.25,26 The majority of work has focused on the fabrication and architecture of superhydrophobic surfaces; however, the stability of Cassie and Wenzel states16,21,23,31,32 and contact angle hysteresis 11,12,14,17,33 have also been addressed. Wetting hysteresis is closely related to surface heterogeneity, although neither the Wenzel nor the Cassie equation account for contact angle hysteresis because they are based on Young’s equation. Contact angle hysteresis is due to free energy barriers associated with surface heterogeneity, which pin the contact line as it advances or recedes over a surface. This process is described in detail in some of the earliest wetting studies, with respect to various periodic surface features, for example, ref 3. Other studies have focused on the pinning behavior at individual defects/domains due to the complexity of heterogeneous surfaces.3,5,10,34-36 These studies have provided pertinent evidence which has informed recent discussions regarding contact angle hysteresis on superhydrophobic surfaces.11,12,14,17,33 However, little is known about the specific relationship between surface features and any departures from Cassie’s equation. One type of departure is asymmetric wetting hysteresis, where either the advancing or receding contact angle deviates strongly from the equilibrium theory.7,13,37 For chemically micropatterned plates with individual defects and defect arrays, we have previously observed remarkable asymmetry in contact angle hysteresis, based upon Cassie predictions.13 We related the observed asymmetric hysteresis directly to the type of defects present, that is, whether they were hydrophobic or hydrophilic, with respect to the wettability of the matrix. In the present work, we study the wetting behavior of water on microstructured, hydrophobic surfaces in the Cassie state. For both microscopic pillars and holes, asymmetric hysteresis is manifested on hydrophobic microstructures and is related to the type of surface features present. This behavior is analogous to that previously observed on flat, chemically heterogeneous surfaces and has implications for the design of superhydrophobic surfaces and functional surfaces in general.

(31) Zheng, Q.-S.; Yu, Y.; Zhao, Z.-H. Langmuir 2005, 21, 12207. (32) Liu, B.; Lange, F. F. J. Colloid Interface Sci. 2006, 298, 899. (33) McHale, G.; Shirtcliffe, N. J.; Newton, M. I. Langmuir 2004, 20, 10146. (34) Nadkarni, G. D.; Garoff, S. Europhys. Lett. 1992, 20, 523. (35) Marsh, J. A.; Cazabat, A. M. Europhys. Lett. 1993, 23, 45. (36) Marsh, J. A.; Cazabat, A. M. Phys. Rev. Lett. 1993, 71, 2433. (37) De Jonghe, V.; Chatain, D. Acta Metall. Mater. 1995, 43, 1505.

5656

DOI: 10.1021/la804246a

Experimental Section Arrays of microscopic pillars or holes were prepared in SU8 photoresist (MicroChem Corporation, USA) using UV-photolithography. SU8 is an epoxy-based negative photoresist with excellent structural integrity and chemical resistance (after hard bake; 200 C, 5 min).38 The SU8 monomer is an aromatic hydrocarbon with eight epoxy functional groups available for cross-linking.38 The rms roughness of the flat SU8 layer was less than 1 nm (1 μm2 scan size), as measured by atomic force microscopy (Digital Instruments Nanoscope III) using NSG10 cantilevers (NT-MDT). Prior to forming the microstructures, the substrates (silicon wafer, Æ100æ, Si-Mat, Germany) were spincoated with a 30 μm thick, unstructured, and hard-baked SU8 layer. Microstructures were formed in a subsequent 30 μm layer of photoresist. The width, w, of the square pillars/holes was fixed at 20 μm, and the height, h, was 30 μm, that is, defined by the photoresist thickness (cf. Figure 1). As the pillar/hole dimensions are fixed, the area fraction of the features is determined only by the lattice spacing, d (cf. Figure 1b). Every sample included an unstructured flat region, which was used to monitor the inherent wettability of the photoresist. All samples were hard-baked after photolithography to enhance the structural integrity and chemical resistance of the microstructures.38 Figure 1c,d shows scanning electron microscopy (SEM) images of the microstructured surfaces. To avoid surface charging during SEM analysis, the surfaces were rendered conductive by sputtering a thin gold layer (20 nm). The thickness of the layer was sufficiently small to be negligible with respect to the microscopic features. The SEM images of the microstructured surfaces clearly show the square pillars or holes formed in the photoresist (Figure 1). The area fraction of the pillars/holes, defined in Figure 1, ranged from zero (i.e., no pillars/holes) to 0.8. The quality of the walls and edges of the pillars/holes was assessed using optical microscopy. The walls of the features were vertical with sharp edges (radius of curvature ∼ 1 μm); surfaces of lower quality were discarded. The SU8 surfaces were hydrophobized by immersing the samples in a 10% wt aqueous solution of a commercial hydrophobizing agent (Granger’s Extreme Wash In) for 1 h.39 The hydrophobizing solution containing the samples was degassed under vacuum to ensure that air was not trapped in the microstructure during hydrophobization. After removing the samples from the hydrophobizing solution, they were rinsed with pure water (18 MΩ cm) and cured at 50 C overnight. Static advancing and receding contact angles were determined from the profile of small water droplets (DataPhysics, OCAH 200).40 Contact angles were measured along the rows of pillars/ holes. The contact line was typically pinned along a single row of features. Droplets (90. Alternatively, where samples gave contact angles less than 90, for example, in some receding measurements, the coverslip was not required (the contact line was viewed directly). Contact line movement was induced by gradual translation of the coverslip parallel to the sample.

Results and Discussion In this Article, we only discuss results where the droplet rests in the Cassie state. The Cassie state was confirmed by microscopic observation of the solid-liquid interface through the liquid phase, when the droplet was confined between a coverslip and the structured surface. By slowly moving the liquid droplet, we were able to map the contact line as it advanced or receded over the surface. We observed that the contact line periodically pinned on the structured surfaces (Figure 2), reminiscent of that observed for chemically heterogeneous surfaces. On chemically heterogeneous surfaces, the three-phase contact line is perturbed by regions of contrasting wettability. In the advancing case, the contact line may be pinned by a more hydrophobic region of the surface or wet preferentially a more hydrophilic region, thus increasing the local curvature of the contact line. The opposite is observed for the receding case. This behavior is a major contributor to contact angle hysteresis, which is observed on almost every surface. We note that the photoresist is more hydrophilic than the regions of air present under the droplet in the Cassie state. Figure 2 shows the contact line pinned while (a) advancing and (b) receding over the microstructured pillars. In each case, the liquid phase is to the left of the contact line. The images of the advancing and receding contact lines on the pillars (Figure 2a) and the advancing contact line on the holes (Figure 2b) show the perspective through the liquid phase (via a glass coverslip), so that the observed pillars are wet by the droplet (solid-liquid interface). Conversely, the image of the receding contact line on the holes shows the perspective Langmuir 2009, 25(10), 5655–5660

Figure 2. Pinning behavior on hydrophobized microstructured surfaces observed through a droplet sandwiched between a partially hydrophobized coverslip and the structured surfaces, as described in the Experimental Section. In all of the images, the liquid phase is to the left of the contact line. Images for the pillars (advancing and receding) and the holes (advancing only) were taken through the liquid phase, while the image for the holes (receding) was taken through the vapor phase. (a) On the pillars, the meniscus is suspended between adjacent pillars and the contact line pins such that the liquid wets the top surface of the pillars irrespective of whether the liquid is advancing or receding. (b) Over the holes, the advancing contact line is pinned at the edge of the holes with liquid fingers extending between the holes, while the receding case is quite different: The edge of the holes pins the contact line and the liquid partially recedes from between the holes (on the matrix) until rapid depinning from the edge of the hole occurs. Whenever liquid entrainment was observed, the liquid remained on top of the matrix (black arrow) rather than in the holes, indicating that the Cassie state was being observed. through the vapor phase, since θ < 90 , so that the observed holes are outside the perimeter of the droplet (solid-vapor interface). For the pillars, both the advancing and receding contact lines wet the tops of the “hydrophilic” pillars. No three phase contact line exists between the pillars, but, instead, the liquid-vapor interface is continuous due to the air present under the droplet, and the curvature of the liquid-vapor interface in the plane of the surface is negative. For the holes, the situation is quite different. The advancing contact line wets between the holes on the “hydrophilic” matrix and is pinned at the front edge of the holes. In this case, the small regions of liquid-vapor interface are formed as the liquid covers the holes; that is, the liquid-vapor interface is discontinuous. For each of these cases, the pinning behavior mirrors that observed on flat, chemically heterogeneous surfaces,13 which suggests that any penetration of liquid into the structure is too small to be relevant. However, the receding contact line on the holes shows quite different pinning behavior from that observed on flat, chemically heterogeneous surfaces. Rather than spontaneously dewetting the holes, the contact line pins strongly on the front edge of the holes, possibly aided by the finite curvature of the edge of the structure. In this case, the three-dimensional geometry of the microstructured sample appears to be important and can no longer be approximated by a flat, chemically heterogeneous surface. We have shown previously that wetting hysteresis is asymmetric on micropatterned, chemically heterogeneous self-assembled monolayers (SAMs). Alkane thiol SAMs containing two different terminal functionalities were photopatterned on gold substrates, and the hysteretic behavior depended strongly on which component was discontinuously distributed.13 Where the discontinuous component (i.e., the “defects”) was more hydrophilic, there was a strong deviation from Cassie’s equation for the advancing contact angle, DOI: 10.1021/la804246a

5657

Article

Priest et al.

Figure 3. Contact angle measurements for microstructured pillars and discontinuous, hydrophilic chemical defects13 compared with Cassie’s equation (solid black lines). The results for the pillars (area fraction of pillars = φ) and the hydrophilic defects (area fraction of defects = φ) show deviation from Cassie’s equation for the advancing contact angles, while only the receding contact angles for the pillars diverge from Cassie’s equation. The broken lines are guides to the eye only. whereas the receding contact angles were in close agreement with theory. The opposite was observed for surfaces with a discontinuous more hydrophobic component.13 These findings resemble the results obtained in this study. The size and arrangement of the features as well as the range of area fractions were similar for both data sets, and thus, the comparison is viable. The static advancing and receding contact angles as a function of the area fraction, φ, are shown in Figures 3 and 4 for the (a) microstructured surfaces and the (b) chemically heterogeneous surfaces.13 The area fraction, φ, refers to the more hydrophilic surface (i.e., solid fraction on the microstructured surfaces and hydrophilic regions on the chemically heterogeneous surfaces). There is a remarkable similarity between the wetting behavior on the two types of surfaces for both pillars and hydrophilic defects (Figure 3) and holes and hydrophobic defects (Figure 4). Table 1 is a general summary of our results for the two types of surfaces with respect to defect type (pillars or holes; hydrophilic or hydrophobic), contact line direction (advancing or receding), and Cassie’s predictions. When discontinuous, the hydrophilic regions (tops of the pillars) induce a distinct departure from Cassie’s equation for the advancing measurements, where the contact angle is almost unchanged up to φ ∼ 0.6 (cf. Figure 3a). Beyond this value, the liquid penetrated the microstructures, and the Wenzel state was established (details not shown). In contrast, the receding contact angles decrease with φ, following a linear trend, similar to the Cassie prediction (cf. Figure 3a). The quantitative difference between Cassie’s equation and the receding data may be attributed to the slightly rounded edges of the pillars. This may cause a partial penetration of the liquid 5658

DOI: 10.1021/la804246a

Figure 4. Contact angle measurements for microstructured holes and discontinuous, hydrophobic chemical defects13 compared with Cassie’s equation (solid black lines). The results for the holes (area fraction of holes = 1 - φ) and the hydrophobic defects (area fraction of defects = 1 - φ) show deviation from Cassie’s equation for the receding contact angles. For the advancing measurements, good agreement with Cassie’s equation was observed, despite significant scatter in our measurements on the microstructured holes. The broken lines are guides to the eye only. Table 1. Comparison of Contact Angles on Heterogeneous Surfaces Compared to Cassie’s Equation contact line direction surface type

advancing

receding

hydrophilic/pillars deviation agreementa hydrophobic/holes agreement deviation a Agreement with Cassie’s prediction for pillars is qualitative only.

into the structure, and we speculate that this could increase the depinning energy and reduce the measured receding contact angles. The wetting behavior observed on the pillars mirrors that observed for flat surfaces exhibiting hydrophilic defects (Figure 3b), where the receding contact angles maintain the linear trend predicted by Cassie, while the advancing data depart significantly. The advancing contact angles on the chemically patterned surfaces are unchanged at low area fractions, φ < 0.4, after which the contact angle decreases toward Cassie’s prediction. When discontinuous, the hydrophobic regions (microstructured holes) induce the opposite behavior. The receding contact angles deviate from Cassie’s equation, while the advancing angles follow the theory (Figure 4a). The departure from Cassie’s equation for the receding measurements differed slightly from that observed on chemically heterogeneous surfaces (Figure 4b). For the microstructured surfaces, the receding contact angle was generally reduced below that for a flat homogeneous surface. This reduction in contact angle can be explained by considering the pinning behavior shown Langmuir 2009, 25(10), 5655–5660

Priest et al.

Article

through the holes data gave Δ cos θ = -0.53, which compares favorably with the measured value of -0.35. The wetting hysteresis versus φ for the chemically heterogeneous surfaces is shown in Figure 5b. These linear fits are based on the low density data only (φ < 0.5 for hydrophilic defects and φ > 0.5 for hydrophobic defects). Linear extrapolation to φ = 0 (100% hydrophobic surface) gave Δ cos θ = -0.53, in excellent agreement with the measured value of -0.53, while extrapolation to φ = 1 (100% hydrophilic surface) gave Δ cos θ = -0.41, compared with the measured value of -0.30. Above and below φ ∼ 0.4, the magnitude of the hysteresis on chemically heterogeneous surfaces decreases independent of the type of heterogeneity. The gradients of the linear fits were significantly different for the microstructured and chemically heterogeneous surfaces. If we assume that the pinning energy associated with the microstructures, Epin, is proportional to the surface coverage of features, the hysteresis measured on the heterogeneous surfaces can be written in the following form: Δ cos θ ¼ Δ cos θ0 -χ Figure 5. (a) Wetting hysteresis as a function of the area fraction of photoresist (solid fraction) φ, that is, the area fraction of features is φ for pillars and 1 - φ for holes. Linear extrapolation to φ = 0 (100% air) for pillars and to φ = 1 (100% photoresist) for holes gave Δ cos θ = -0.07 and -0.53, respectively. (b) Wetting hysteresis for flat, chemically patterned surfaces, plotted against the area fraction of the hydrophilic component, φ. Solid lines are linear fits for low surface densities of holes/pillars, which are extrapolated over the full scale. At φ = 0 (100% hydrophobic), Δ cos θ = -0.53, while extrapolation to φ = 1 (100% hydrophilic) gave Δ cos θ = -0.41. in Figure 2b. As expected, the advancing contact line is always pinned at the front edge of the holes, with liquid fingers extending between the holes. However, in the receding case, the configuration is quite different. Here, the edge of the holes pins the contact line and the liquid partially recedes from between the holes (on the matrix) until depinning from the front edge of the hole occurs. Note that, even in this contact line configuration, air remains trapped in the holes. This pinning process was not observed on chemically heterogeneous surfaces,13 suggesting that it is specific to the threedimensional architecture. These subtle differences in pinning behavior may be due to partial penetration of the contact line into the holes prior to depinning due to the finite curvature of the edges of the surface features; however, high-speed, threedimensional mapping of the interface would be required to distinguish the mechanism in greater detail. In many applications, the magnitude of wetting hysteresis will influence the functionality of the surface. Wetting hysteresis is often minimized in wetting applications to allow a droplet to readily move over the surface. Figure 5a shows contact angle hysteresis versus the area fraction of “hydrophilic” surface, φ (the area fraction of the pillars is φ and for holes is 1 - φ). The contact angle hysteresis increases with increasing density of surface features, that is, pillars/holes. The relationship holds over the full scale of accessible area fractions. The linear fit is reasonably good for the pillars but is somewhat less convincing for the holes, possibly due to the more complex depinning mechanism (Figure 2). Extrapolation to φ = 0 (100% air) through the pillars data gave Δ cos θ = -0.07, in good agreement with the assumed value of zero. Extrapolation to φ = 1 (100% solid fraction) Langmuir 2009, 25(10), 5655–5660

Epin γLV

ð4Þ

where Δ cos θ0 is the contact angle hysteresis on the matrix (i.e., the continuous portion of the surface), χ is the surface area fraction of the features (i.e., χ = φ for pillars and hydrophilic defects, and χ = 1 - φ for holes and hydrophobic defects). The linear fits in Figure 5 follow the form of eq 4. From these fits, we find the pinning energy normalized to the surface tension of the liquid, Epin/γLV, is 0.5 and 0.6 for the hydrophobic and hydrophilic defects, respectively, while Epin/γLV = 0.4 and 1.2 for the holes and pillars, respectively. This difference is an indication of the different pinning energy associated with the two types of surface features. Thus, the pinning energy on the microstructured pillars is significantly greater than that observed on the holes and the chemically heterogeneous surfaces. Furthermore, the linear dependence appears to hold over a greater range of χ for the microstructures than for the flat, chemically heterogeneous surfaces. None of the measurements for the microstructured surfaces showed significant departure from these fits over the full range of data. Alternatively, for the chemically heterogeneous surfaces, the hysteresis was maximized at χ ∼ 0.4 and showed no clear dependence on the type of heterogeneity present (i.e., hydrophilic or hydrophobic defects). Where the pinning energy is lower, the barriers around the lowest free energy are reduced, allowing the contact angle to approach the equilibrium value. In this case, the contact line does not conform precisely to the surface features. As the separation between the surface features reduces, the liquid no longer distinguishes between the defect and the matrix but rather averages the surface components, as predicted by Cassie’s equation. In this case, the observed deviations from Cassie’s law reduce and so does the contact angle hysteresis. This, at least in part, explains the difference in the hysteresis on the two different types of surfaces. It is also possible that the two- and threedimensional nature of the surfaces also contributes to the difference in hysteresis energy. The ability to relate wetting hysteresis to specific surface features is a powerful tool for the design of functional surfaces. One example is the superhydrophobic surface, where the Cassie state is desirable for its ability to minimize the solid-liquid adhesion and, therefore, minimize contact angle DOI: 10.1021/la804246a

5659

Article

Priest et al.

hysteresis for liquid repellency.42 Therefore, a general approach is to prepare surfaces structured with small needlelike features or a hierarchical surface structure to minimize the solid area fraction, which typically exhibit low contact angle hysteresis. This is in line with the biomimetic approach to surface design, based on observations of surfaces found in nature, for example, the lotus leaf,43 Stenocara beetle’s back,44 and water strider’s leg.45 Relating the wetting behavior on a specific type of surface structure to that observed on other functional surfaces can be difficult, due to the diversity of surfaces encountered. The work presented here provides a systematic approach toward optimizing the wettability of a superhydrophobic surface, revealing not only the relationship between the solid fraction and the wettability but also the importance of the type of structure present. Our results show that needlelike structures at very low area fractions will maximize both the advancing and receding contact angles while minimizing the magnitude of wetting hysteresis. However, other scenarios may also be considered. For example, some applications might favor the use of continuous porous networks, for example, membranes, or simply require the structural integrity of a continuous solid fraction. In each of these cases, only the advancing contact angle would be increased by decreasing the solid fraction, while the magnitude of wetting hysteresis on the holes structure is maximized at low solid fractions. Furthermore, the crossover in Figure 5 suggests that, at high solid fractions (in excess of >0.6), the hysteresis on the holes structure may in fact be less than that observed on the pillar surfaces. Therefore, the design of superhydrophobic surfaces and other specialty surfaces must account for the interplay between absolute contact angles and the magnitude of hysteresis with respect to solid fraction and the type of surface features present. (42) Callies, M.; Quere, D. Soft Matter 2005, 1, 55. (43) Sun, M.; Luo, C.; Xu, L.; Ji, H.; Ouyang, Q.; Yu, D.; Chen, Y. Langmuir 2005, 21, 8978. (44) Zhai, L.; Berg, M. C.; Cebeci, C-. F.; Kim, Y.; Milwid, J. M.; Rubner, M. F.; Cohen, R. E. Nano Lett. 2006, 6, 1213. (45) Feng, X.-Q.; Gao, X.; Wu, Z.; Jiang, L.; Zheng, Q.-S. Langmuir 2007, 23, 4892.

5660

DOI: 10.1021/la804246a

Conclusion The wettability of structured surfaces for liquids residing in the Cassie state is strongly dependent on both the solid fraction and the continuity of the solid component. Microstructured surfaces consisting of square pillars or holes showed distinct asymmetric wetting hysteresis. For surfaces containing holes, the receding contact angle was unaffected by the solid fraction, while the advancing contact angle increased in accordance with Cassie’s prediction. Conversely, the advancing contact angles were unaffected by the solid fraction for the microstructured pillars, while the receding contact angles followed a trend similar to that predicted by Cassie’s equation (albeit with some departure). This behavior is consistent with our previous observations of the wettability of chemically heterogeneous surfaces. Furthermore, we have shown clear trends for the magnitude of contact angle hysteresis with area fraction for the two types of microstructured surfaces and estimated the pinning energy, Epin. The relation between surface features and wetting behavior outlined in this work will assist in the design of functional surfaces and contribute to our understanding of the wettability of naturally occurring rough and heterogeneous surfaces. Acknowledgment. Photolithography masks and microstructured surfaces were prepared at the Macquarie/ATP and University of South Australia nodes of the Australian National Fabrication Facility, respectively, under the National Collaborative Research Infrastructure Strategy to provide nano- and microfabrication facilities for Australia’s researchers. Financial support from the Australian Research Council (ARC) Special Research Centre Scheme, ARC Linkage and Linkage International Schemes, AMIRA International, and State Governments of South Australia and Victoria, is gratefully acknowledged. Supporting Information Available: Wettability measurements on the microstructured surfaces; advancing and receding contact angles on surfaces containing pillars (Table 1) and for holes (Table 2). This material is available free of charge via the Internet at http://pubs.acs.org.

Langmuir 2009, 25(10), 5655–5660