Physical Texturing for Superhydrophobic Polymeric Surfaces: A

Jun 18, 2017 - Although new wetting theories continue to emerge, there is not yet a set of design rules to guide the selection of surface topographies...
1 downloads 22 Views 9MB Size
Article pubs.acs.org/Langmuir

Physical Texturing for Superhydrophobic Polymeric Surfaces: A Design Perspective Jing Yang Quek,† Christopher L. Magee,‡ and Hong Yee Low*,† †

Engineering Product Development, Singapore University of Technology and Design, 8 Somapah Road, 487372, Singapore IDSS and Department of Mechanical Engineering, Massachusetts Institute of Technology, 77 Massachusetts Avenue, Cambridge, Massachusetts 02139, United States



S Supporting Information *

ABSTRACT: Surface wetting on the textured surface is classically explained by the theories of Cassie−Baxter or Wenzel. However, in recent years, an increasing number of complex surface topographies with superhydrophobic properties have been achieved without prediction or simulation using these theories. One example is biomimetic surfaces. In many instances, theories were used to explain surface properties found in nature but have not led to or predicted the complex topographies. Although new wetting theories continue to emerge, there is not yet a set of design rules to guide the selection of surface topographies to achieve superhydrophobicity. By grouping known surface topographies into common geometrical descriptions and length scale, this paper suggests a set of surface topography classifications to provide selection guidelines for engineering superhydrophobic surfaces. Two key outcomes emerged from the design analysis: first, categorization of frequently reported surface patterns shows that there exists a set of commonly used descriptions among diverse designs; second, the degree of hydrophobicity improvement within a class of topography design can be used to predict the limit of improvement in superhydrophobicity for a given material. The presentation of topography descriptors by categories of design and performance may serve as a prologue to an eventually complete set of design guidelines for superhydrophobic performance.

1. INTRODUCTION Material innovation has gone through phases of change from the discovery mode in the medieval age to the modern materials by design mode. The rapid advancement of chemical knowledge in conjunction with the development of novel processing techniques has been the key contributor to the successful synthesis of a vast number of new chemicals with exciting properties. However, there has also been an increasing interest in the use of nonchemical-based knowledge for materials design. Physical structuring or more specifically micro- and nanoscale patterning is an important example of using nonchemical approaches to manipulate material properties. This is done through altering of surface topography or roughness that can influence material properties such as optical reflectivity, surface wetting, and friction. Among these surface properties, extreme nonwetting surfaces or superhydrophobic performance has been a topic of intense studies by both academic and industrial researchers. Searches carried out using Web of Science show a continuous increase in the number of publications and citations on this topic. Examples of search data using various combinations of keywords are shown in Figures S1−S3. The intensive interests in superhydrophobic surfaces stem from its potential uses in a myriad of applications ranging from consumer products (including appliances and electronics) © XXXX American Chemical Society

to biomedical devices to transport vehicles to buildings and even water treatment technologies. Superhydrophobic surfaces lead to functions such as self-cleaning or easy clean,1−5 antifogging,6,7 anti-icing,8−10 and antifouling.11−13 Several existing reviews provide tremendous insights into the fundamental and technological aspect of superhydrophobic surfaces.14−17 These reports mostly cover the advancements made in the fabrication technologies that have enabled a wide range of topography designs, from simple two-dimensional structure to complicated three-dimensional and hierarchical structure with length scales that span from micrometer to nanometer on a variety of materials. From these reports, we see an increasing number of complex and sophisticated topography designs that point to a nearly open-ended set of design possibilities. Although a high degree of design freedom offers the opportunity for custom tuning to specific application requirements, the adoption of this technology may be hampered in the absence of design guidelines. The two most basic theories that have been widely used to predict the wetting property of a surface are the equations of Wenzel and Cassie− Received: April 18, 2017 Revised: June 17, 2017 Published: June 18, 2017 A

DOI: 10.1021/acs.langmuir.7b01175 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

covering, whereas the troughs including the sloping sides are covered by a microstructure coated in wax. The microstructure consists of flattened hemispheres 10 μm in diameter arranged in a regular hexagonal array similar to the lotus leaf.35 Water striders are insects that live on the surfaces of ponds, slow streams, and quiet waters and that can stand and walk upon a water surface without getting wet. The leg surface was reported to have special hierarchical structures consisting of numerous oriented, needle-shaped microscopic keratinous hairs known as setae. On each micro seta, there are elaborate nanoscale grooves. The needle-shaped micro setae are roughly 50 μm in length and have diameters ranging from 3 μm down to several hundred nanometers. The micro setae are arranged at an inclined angle of approximately 20° from the leg surface. The maximal force of a single leg was found to be 1.52 mN or approximately 15 times the total body weight of the insect, and the corresponding volume of water displaced is roughly 300 times that of the leg itself. It was further reported that the leg does not pierce the water surface until a dimple of 4.4 mm depth is formed.36 Lizards, like geckos, are known to climb walls and ceilings via van der Waals forces on their footpad; this ability is attributed to the hierarchical structure of their feet.37 Their foot is made up of well-aligned micro setae with an average diameter of 5 μm and a length ranging from 30 to 130 μm. The setae split into hundreds of smaller nanoscale ends called spatula that are 0.2− 0.5 μm in diameter.38 Contact between the gecko feet and an opposing solid surface generates van der Waals forces that allow them to climb vertically or across ceilings.39 In addition to reversible adhesion, the gecko foot also exhibits superhydrophobicity and self-cleaning properties.40,41 1.2. Design Perspective. Today, there are many synthetic topography designs known to have superhydrophobic performance, that is, with WCA near or above 150°. However, there is no systematic guideline for superhydrophobic topography designs. The selection of a topography design for a desired hydrophobic performance depends on many factors; besides the topography design, the choice of materials and the fabrication methods are two other key factors. The selection of material in most instances is governed by the application requirements. For an example, considering selection from a broad perspective, a high-temperature stability requirement for the superhydrophobic surface leads to metals and ceramics as suitable candidate materials. If lightweight and mechanical flexibility are important application requirements, then polymeric materials are more suitable candidates. Furthermore, the choice of topography designs and the materials are strongly dependent on the availability and cost of fabrication methods. The interdependence of these four key design considerations is shown schematically in Figure 1. An example of a design process is one that begins with a set of application requirements that are followed by the material selection. Once a material is identified, the selection of topography designs will be done in close consideration with the choices of fabrication methods. Often, the type of material limits the choice of the fabrication method, which leads to a narrower set of topography designs. As pointed out by several review articles,14,16,19,42 certain topographies can only be produced by specific fabrication methods. As an example, if polymeric material is the candidate material for the intended application, then the choice of fabrication methods can be down-selected to either etching or embossing techniques. To go beyond this simplistic design approach, we should recognize

Baxter for water contact angle (WCA) on a surface (The details of these two theories have been extensively covered in existing review reports14,18,19 and will not be repeated here. A brief description of these theories is available in the Supporting Information). Newer design theory for superhydrophobic surfaces, the reentrant profile, is increasingly being incorporated into the design of superhydrophobic topography.20−22 Another approach to designing superhydrophobic surfaces is biomimicry. Indeed, many reported superhydrophobic topographies either mimic or are inspired by natural topographies. 1.1. Biomimetic Superhydrophobic Surface Patterns. Topography designs for superhydrophobic surfaces can be mimicked from nature as many biological systems in plants and animals have natural superhydrophobic surfaces that exhibit remarkable water-repellent properties.23−25 Among these, the lotus leaf is the most well-known and studied. Barthlott and Neinhuis highlighted the lotus leaf (Nelumbo nucifera) for its super water repellence and self-cleaning ability.11,26 Water can bridge over the surface roughness in a Cassie−Baxter wetting state. This allows water droplets to roll easily across and off the surface, carrying away dirt and debris concurrently.5 This phenomenon was given the term lotus effect and has been patented as an idea and trademark. The superhydrophobic and self-cleaning properties of the lotus leaf are attributed to its epicuticular wax and micronanoscale hierarchical architectures made up of cilium-like nanostructures superimposed on the micrometer-scale papillae on their surfaces. Using the self-cleaning effect of lotus leaves as an inspiration, a wide variety of other artificial superhydrophobic self-cleaning surfaces have been fabricated by creating appropriate surface chemical composition and hierarchical surface geometrical structures.27,28 Like the lotus leaf, the rice leaf surface is also covered with hierarchically structured micropapillae and nanobumps but in a sinusoidal groove array. This arrangement gives the rice leaf surface anisotropic wettability where the water droplets can roll more easily along the direction parallel to the rice leaf edge than along the perpendicular direction.29,30 As for animals, studies of superhydrophobic surfaces have been carried out on mosquito eyes, cicada wings, butterfly wings, the back of the Namib beetle, and the feet of water striders and geckos.4,24,31 The mosquito eye is composed of spherical microstructures called ommatidia that act as individual sensory units. The ommatidia have an average diameter of 26 μm and are organized in a hexagonal closed packed array. At high magnification, the surface of each ommatidia is covered with nanobumps that have an average diameter of 101 nm with a pitch of 47 nm arranged in a nonclosed packed array. The mosquito eye has antifogging and antireflection properties in addition to being water-repellent.6 In cicada wings, the surface consists of hexagonally closed packed nanopillars with a height ranging from 225 to 250 nm and an average pitch of 110−140 nm. The cicada wing is waterrepellent and has antireflection property.32,33 The butterfly wing consists of overlapping microscales that are separated by nanoscale ridges. This enables the butterfly wing to exhibit anisotropic wetting and directional adhesive properties to water.30,34 The Namib beetle can collect drinking water from the fog onto its back. On a macroscopic scale, the elytra of the beetle are covered in random arrays of bumps with an average diameter of 0.5 mm, which are 0.5−1.5 mm apart. At the microscopic level, the peak of the bumps is smooth with no B

DOI: 10.1021/acs.langmuir.7b01175 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

The hydrophobicity of a surface is determined by either static WCA or advancing WCA. It is important to consider WCA of the flat pristine surface because simply studying the WCA of the patterned surface would mix the effect of topography on hydrophobicity with material effects. By comparing the types of topography, we have chosen to evaluate the change in the WCA between the nontextured surface and the textured surface. Although more elaborate performance criteria can be incorporated such as sliding angle and contact angle hysteresis, these properties were not as extensively reported in the collected publications as was the WCA.

3. DESCRIPTORS OF TOPOGRAPHIES In the field of surface engineering, surface topographies originated mostly from roughened surfaces. A roughened surface is often described as “peak and valley” of certain heights or amplitude and quantified by the root mean square (RMS) of the roughness. A high roughness surface would have a high RMS value. Following the above convention, we separated the description of surface topography into protruding features (peaks) or recessed features (valleys). By comparing the protruding and recessed structures, protrusions exhibit a wider range of geometrical variations, which we categorized as pillarlike structures (high aspect ratio structure) and low aspect ratio structures that are roughly spherical in shape. One classification of topography that is not traditionally considered as a texturing effect is fibrous structure. A number of reports have described fibrous structure as a geometrical factor that affects surface wetting; hence, in this analysis, fibrous structure is included as a topography category. The flowchart in Figure 2 shows the classification of topographies as just described: the three main classes of topographic designs are the protrusion, recession, and fibrous geometries. For labeling of the surface patterns found in publications, specific letters were used to classify them according to their topographic designs while they were numbered accordingly. Recessed structures are labeled “R” (with each literature example given a number as well, e.g., R4), whereas pillar structures are labeled “P”. On the other hand, low aspect ratio structures are labeled “S” and fiber structures are labeled “F”. The influence of silane coating was also studied for some physical textures, and these surface patterns were labeled “C”. Figure 3 shows a summary of literature values for the increase in WCA organized into the four major classes of topography designs. Within each of the design groups, there are further variations in the surface pattern geometries. The WCAs for the surface patterns consolidated are all hydrophobic (i.e., >90°) with some in the superhydrophobic range (i.e., ≥150°). As shown in the figure, the increase in WCA ranges from about 5° to about 75°, except for the fibrous structure, which shows

Figure 1. Schematic depiction of the interdependence of the key design factors for superhydrophobic surface design.

that each of the design factors shown in Figure 1 includes multiple subfactors to consider. A complete set of design principles for superhydrophobic surfaces could serve as a toolbox to shorten the development time and to increase the adoption of this technology by industries for applications or products. The establishment of a complete design guideline for superhydrophobic surfaces is beyond the scope of this paper; instead, the current effort takes an initial step toward such a goal by defining the descriptors of micro- and nanoscale topography designs for the superhydrophobic performance of polymeric surfaces. A systematic categorization of the descriptors is presented as a library of tested surface topographies.

2. METHODOLOGY The Web of Science was used to identify relevant review and report articles through keywords search. Keywords used during the searches included “superhydrophobicity”, “surface patterns”, “biomimicry”, “physical texturing”, and “superhydrophobic surfaces”. Relevant references were also extracted from review articles. The collected literature was organized according to the commonalities in pattern designs and length scales. The search was further filtered to limit the result to polymeric surfaces, which can be either a polymeric substrate or a polymer coating on a silicon (Si) substrate. Because of the photolithography process, many topography designs are fabricated on the Si substrate; however, only the patterned Si with a polymer coating is included in this study. Generally, metals and oxidebased materials possess higher surface energy compared to polymeric materials, which is an intrinsic material property that directly influences the wetting performance. The effects of topographical design on superhydrophobic properties are expected to differ between the high surface energy materials and polymeric materials.

Figure 2. Flowchart for the categorization of polymeric surface pattern designs. C

DOI: 10.1021/acs.langmuir.7b01175 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

Figure 3. Increase in WCA for all surface patterns sourced from journal articles and organized according to the category of designs, namely, recessions,43−49 pillars,50−52 low aspect ratio protrusions,31,32,53−60 and fibers.61−69 Literature and organized according to the category of designs.

an increase between 35 and 85°. In the following sections, we analyze the structural variations within each of the categories. The WCA improvement for each design is also plotted in bar graphs where different colors were used to represent the intrinsic WCA of the unpatterned substrate material. 3.1. Recessions/Pores. 3.1.1. Description of Recession/ Porous Design. In this category, the porous structure is included as a recessed structure. In the context of the classical Cassie−Baxter and Wenzel equations, these types of structures are not considered as surface roughness. In other words, pores and recessed structures are not associated with a roughness factor, and hence, they would not be predicted to be hydrophobic by the theories of Cassie−Baxter and Wenzel. There are two main differences between pores and recessed structures. First, porous structures are typically present throughout the bulk of the material, whereas recessed structures are found only on the surface of a material. As a result, recessed structures have a defined depth, whereas porous structures do not. Second, in porous structures, pore diameters are usually polydisperse and pore arrangements are disordered, whereas in recessed structures, the “holes” have well-defined sizes (diameter/width and depth) and can have a highly ordered arrangement. The scanning electron microscopy (SEM) images in Figure 4 highlight the differences.44,47 With the exception of R147 and R4,48 all porous structures analyzed consist of random pores distributed throughout the bulk of the materials. Such structures are typically fabricated via wet chemistry approaches and are not considered as a surface topography design. However, the superhydrophobic porous structures in R1 and R4 were fabricated using fluorinated copolymers with uniform pores and ordered arrangement.47,48 In this analysis, all porous structures that have been reported as having a superhydrophobic surface are grouped without further differentiating them. 3.1.2. WCA Improvement. From the perspective of liquid wetting on a surface, pores and recessed structures have airspace that coexists in the plane of the surface. Hence, in this analysis, pores and recessed structures are grouped under the same category. The amount of airspace occupied by pores and recessed structures has a direct effect on liquid wetting the surface; hence, the next comparative analysis is the effect of length scale. Figure 5 presents a further subgrouping of porous structures by the average pore size/diameter. The figure

Figure 4. (a) SEM image of random porous structure, R744 and (b) uniform porous structure, R1.47

indicates that there is no significant difference between the structures with pore size