Self-cleaning materials that can be cleaned by simply being rinsed with water, can be potentially applied in many areas. As exterior surfaces, such materials will be cleaned just with regular rain significantly reducing maintenance costs, due to less labor required and no need for detergents, with minimal impact on the environment.1 Therefore, self-cleaning surfaces are intensively studied and commercialized for such applications as windows,2 paints,3 satellite dishes, solar energy panels,4 and automotive windshields.3 Incorporating the self-cleaning functions in textile materials makes them wash-free5 or at least reduces the number of washings,6 offers convenience for everyday use and provides additional value to the product.5 Fibrous self-cleaning surfaces offer some particular benefits such as simple and potentially large-scale fabrication from a wide range of raw materials, mechanical integrity, self-supporting, low vapor transport resistance.7, 8 Self-cleaning textile materials also have many potential applications in apparel (self-cleaning necktie, shirts, blouses, skirts), and outdoor textiles.5, 8
In general, there are two routes to produce self-cleaning surfaces both of which utilize specific surface design and chemistry to control wettability.1 In the superhydrophilic self-cleaning approach, water completely covers a surface with a continuous film and washes away any dirt.2 Usually, the complete wettability of the surface is achieved by incorporating photocatalytic chemicals, for example, TiO2, which has dual action. Under UV irradiation, these photocatalytic chemicals form a very low contact angle with water (<1°),9 and at the same time, they generate activated oxygen that decompose organic materials on the surface.2
The second approach utilizes the opposite side of the surface wettability scale, and the self-cleaning property is achieved with help of high water-repellency or superhydrophobicity of a surface. This approach was inspired by the natural world where over 200 species of plants benefit from the proper combination of surface chemistry and morphology to stay clean.1 This property was initially discovered by Neinhaus and Barthlott on the surface of Lotus leaves and was subsequently patented in 1998 as the Lotus-effect.1 The superhydrophobic surfaces have a very high water contact angle (WCA) (>150°) and a very low roll-off angle allowing water droplet to roll at a very low tilt angle of a surface.10 In contrast to a sliding motion, rolling droplets easily pick up dust particles and remove them from a surface, realizing the self-cleaning effect.11 In addition to being self-cleaning, such surfaces have stain resistance properties, due to the high contact angle and limited contact area of contaminated water droplets.3
The superhydrophobicity of textile materials has been studied for a long time. However the development of new fabrication techniques, which allow manipulating of surface morphology on the micro and nano levels, has significantly increased scientific interest to this problem in recent years.7 Artificial superhydrophobic surfaces can be produced in many ways, including template synthesis, phase separation, crystallization control, etching, sol-gel processing, layer by layer deposition, and electrospinning.4, 12 Electrospinning is a very promising technique for the production of superhydrophobic surfaces. This process is easy to set up for laboratory research, it offers many controllable parameters that can be used to obtain the required surface morphology, it can produce micro and nanofibers from a wide range of materials, and fibers produced by electrospinning are very similar to those used in nature to generate superhydrophobic self-cleaning properties.7
This review will discuss the results published in the scientific literature concerning fabrication of superhydrophobic surfaces with electrospinning. A similar study can be found in ref.7, but our study considers about 25 papers published after 2007, which were not discussed in ref.7. The article consists of three main sections. In the first two sections, the theory and main principles of superhydrophobicity and the electrospinning process will be discussed. However, due to a great number of publications, these topics will be discussed somewhat briefly just to give the reader a basic understanding. The references to more specific literature will be given in each section. The third section will review about 40 papers concerning electrospun superhydrophobic materials. The papers will be categorized in several groups based on the raw materials used and the production process utilized (one step or multistep).
SELF-CLEANING AND SUPERHYDROPHOBICITY
Because the self-cleaning effect on superhydrophobic surfaces is achieved by a rolling motion of water droplets, there are three basic requirements to such surfaces. First, droplets should have a high static (WCA >150°). Second, these droplets should not be strongly attached to a surface and easily move even at low inclination (<10°). Finally, water droplets should actively pick up dust particles, so the adhesion between dust and a solid surface should be lower than the adhesion between dust and water.13
Wettability of solid surfaces has been studied for a very long time. The first fundamental equation that quantified the static contact angle of a liquid droplet on a flat surface was proposed by Young in 1805.14 The droplets form a three-phase contact line where solid–liquid, solid–vapor, and liquid–vapor interfaces meet. Forces created by the surface tensions at each interface are pulling droplet in solid plane and define its shape.15 According to the Young's equation, the cosine of the WCA in the equilibrium state (θY) is directly proportional to the difference of the interfacial forces per unit length of solid–vapor (γSV) and solid–liquid interfaces (γSL), and inversely proportional to the interfacial force of the liquid–vapor (γLV) interface.16 When the contact angle for liquid is higher than 20°, it can be assumed that γSV ≈ γS and γLV ≈ γL.17
This relationship is shown graphically in Figure 1. It can be seen from the figure and the eq 1 that the static WCA can be increased by lowering the surface tension of a solid material,13 since the surface tensions of water and surrounding medium (usually air) are fixed and defined. The chemical groups with low surface energies can be ordered as > > > > > .18 The lowest surface energy of 6.7 mJ/m2 was obtained for regularly aligned closest-hexagonal-packed CF3 groups.19 However, the static WCA of a smooth surface produced from this material can reach only 119°,7 which is far away from the requirements of superhydrophobic surfaces.
Through working on the water resistance of textile surfaces, Wenzel20 found that the roughness of a surface can significantly change an equilibrium WCA on such surface. He assumed that water follows all topological variations of a surface as shown in Figure 2(a), thus increasing the water–liquid interface.21 Wenzel's assumption is also called homogeneous wetting. Based on this assumption, the author derived an equation that relates surface roughness (r) and Young's contact angle (θY) to the actual WCA on a rough surface (θW).22
The roughness factor (r) is the ratio of the water–solid contact area to the area of geometrical projection of water–solid interface (excluding surface topology).23 If the value of the roughness factor is equal to one, which corresponds to a flat surface, Wenzel's eq 2 transforms into Young's eq 1. However, for rough surfaces this factor is always greater than unity (assuming that water penetrates into surface cavities) leading to an increase in the contact angle for hydrophobic surfaces (θY > 90°) and a decrease in the contact angle for hydrophilic surfaces (θY > 90°).13
Later Cassie and Baxter24 proposed another model to explain the effect of roughness on droplet behavior on a surface. In contrast to Wenzel, they assumed heterogeneous wetting when air is entrapped by water in surface cavities as shown in Figure 2(b). In such case, the contact area between water and the solid is minimized, and the area between water and air is maximized, thus forcing water to form spherical droplets. In addition, roughness of wet areas also influences the contact angle, and in general, Cassie's equation can be described as:
where f is the ratio of the solid–liquid contact area to the area of geometrical projection of water droplet.13 Therefore, when f decreases due to an increase in roughness (more air is trapped), the WCA increases.10
Wenzel's and Cassie's models give a fundamental explanation of how a static WCA can be increased with the help of surface roughness. However, these models do not consider the dynamics of a water droplet when it moves on a surface. Not all surfaces with high static WCAs have low roll-off angles.12 Öner and McCarthy25 showed that the structure of three-phase contact line is an important factor that defines the dynamic behavior of a water droplet on a solid surface. Water molecules on the three-phase contact line are the only solid–water interface molecules that move during the motion of a droplet on a solid surface. When a droplet is in the equilibrium state, the contact line is fixed and energy barriers for advancing and receding of the contact line cause hysteresis. Surfaces where the contact line can be easily destabilized have lower hysteresis and a droplet on such surfaces moves better.
The dynamic behavior of a water droplet on an inclined surface is shown in Figure 3.26 A water droplet on a tilted surface has a different contact angle on each side of the drop.12 The angle in front of the droplet motion is called the advancing angle (θA) and the angle on the other side of the drop is called the receding angle (θR). The difference between these two angles is called the contact angle hysteresis (H = θR−θA). Hysteresis creates force (F) that attaches a droplet to a surface and that have to be overcome to initiate and support droplet motion:12
where γLV is the interfacial energy of the liquid–vapor interface, and cosθR(min)−cosθA(max) is a contact angle hysteresis in a cosine form.
On the inclined surface, the force that interacts against the pinning force is created by gravity and depends on the tilt angle. Therefore, the critical tilt angle (αC), that is, required to initiate droplet motion depends on the contact angle hysteresis10 and can be derived from the next equation:
where m is the weight of the droplet, g is the gravitational acceleration, and d is the diameter of the water droplet on an incline.
It can be seen from eq 5 that a reduction in the contact angle hysteresis leads to a reduction in the critical tilt angle. Although both Wenzel's and Cassie's models predict high static WCAs on rough hydrophobic materials, WCA hysteresis in both these states are completely different.14 In Wenzel's state, water follows all surface cavities and is strongly pinned to the surface, which results in high hysteresis and consequently, higher force (higher tilt angle) is required to move a droplet. On the other hand, in Cassie's state, a contact area between the water and solid are reduced, which creates lower hysteresis and a droplet moves at smaller inclinations. Therefore, Cassie's state is preferable to create the superhydrophobic self-cleaning effect on a solid surface.
In addition, it was shown in ref.27 that an increase in surface roughness leads to a transition from Wenzel's to Cassie's state. When a water droplet is in the Wenzel's state, an increase in roughness increases both the static WCA and the hysteresis. However, when the roughness factor exceeds a certain level (∼1.7),27 it is no longer efficient for water droplet to fill all holes on the surface. Therefore, the transition to Cassie's state occurs and some air becomes entrapped beneath water droplet. In this state, further increase in roughness makes the static WCA bigger, but reduces the hysteresis.19
However, even in the Cassie's state, the hysteresis may be high enough to prohibit easy movement of a droplet. The factor f in the Cassie's eq 3 may be equal for the surfaces with different topologies, which results in the same static contact angle.25 However, since the equation does not account for the structure of the contact line at these surfaces, which may be different for each surface, advancing and receding contact angles and hysteresis may also differ, which results in dissimilar dynamic behavior of a water droplet. If the contact line is continuous, the droplet is pinned to the surface and is difficult to move. In contrast, discontinuous contact line is unstable and easily advances or recedes moving the droplet.
Surface roughness also has a significant effect on the adhesion of dust particles to the surface. It minimizes the contact area between the surface and particles (assuming that particles are bigger than surface cavities), so particles can be easier picked-up by water.13
Fundamental studies discussed previously clearly show that the combination of proper surface chemistry and topology can be utilized to produce surfaces with superhydrophobic self-cleaning properties. These results agree with experimental works and biological findings. In 1997, Barthlott and Neinhuis28 examined the micro topology of lotus leaves and found that their surface consists of micrometer bumps covered with many nanosized waxy crystals. Due to this double-scale roughness and the low surface tension of crystals, water easily rolls off and picks up dust. Thus lotus leaves are always clean. The authors named this phenomenon the “Lotus effect”. However, later the same property was discovered in many plants and animals.13
The fundamental theories of superhydrophobicity and self-cleaning were discussed in this section. The development of nanoscience, new fabrication methods, and findings from the natural world spurred an increased interest in this area in recent years leading to the development of many new theories as well as preparation methods of self-cleaning superhydrophobic surfaces. A more detailed discussion of this topic can be found in references.16, 29–31
In the next section, electrospinning, a promising technique for fabrication of self-cleaning surfaces will be discussed. By producing fibers with diameters in nano and micro ranges and permitting control of fiber chemical composition and surface morphology, electrospinning enables the surface structures from the natural world to be mimicked.
ELECTROSPINNING: BRIEF REVIEW
Since nonwoven webs composed of small fibers are very similar to the superhydrophobic surfaces of plants leaves, they can be utilized to fabricate artificial self-cleaning surfaces. There are several technologies widely used in the textile industry to produce fibers with diameter in a submicron and nano ranges, including splitting bi-component fibers or dissolving one of its components, melt-blowing and electrospinning.32 The first two techniques enable fibers with diameters of 250-300 nm to be prepared at high production rates.32 However, only thermoplastic polymers can be used as raw materials in these techniques.33 In contrast, electrospinning can generate fibers with diameter as low as 10 nm34 from any conductive viscoelastic liquid that can solidify (e.g., polymer solutions, melts, or sol-gels34), and allows functionalization of fibers with a wide range of additives.35 In addition to flexibility in raw material selection, electrospinning is a method, that is, cheap and simple to implement36 that has many options to control fiber diameter,33 morphology, surface topology,35 and fiber arrangements,37 and has been successfully scaled up for mass production (e.g., Nanospider™).36
Electrospinning is a process that utilizes electrical forces to produce continuous filaments from polymer solution, melt or other viscoelastic liquids, like sol-gels. The resulting fibers can be randomly dispersed on a collecting device to form a nonwoven web or can be aligned to form uniaxial or biaxial nanofiber assemblies.37 The nano-dimension of electrospun fibers and the simple and convenient production process makes electrospinning a very attractive candidate to use in such applications as filter membranes, biomedical scaffolds, wound dressing, nano-sensors and protective clothing.38
Despite the fact that the number of scientific papers concerning electrospinning significantly increased in recent years, this process has been known for almost 100 years, with some related studies dating back to the 18th century.33 The first devices to electrospray liquids were patented by Cooley and Morton at the very beginning of the 20th century (US Patents 692,631; 705,691; 745,276), and the electrospinning process of polymers with detailed description was patented by Formhals in 1934 (US Patent 1,975,504).33 However, real interest in electrospinning was induced by the development of nanoscience and nanotechnology34 and the number of publications in this field increased from 40 papers for the period 1994-2000 to 2200 papers in 2010 (the number of papers were obtained from the SciFinder Scholar database searching for “electrospinning”). Together with a high number of papers, many reviews of different aspects of the electrospinning process have been published in recent years, including general reviews in refs.33–35, 39–45, polymers used in electrospinning in refs.33, 34, 40, 42, mathematical modeling of electrospinning in refs.34 and39, electrospinning apparatus and techniques in refs.36–38, nanofibers assemblies in ref.37, coaxial electrospinning in ref.46, and electrospinning from melt in ref.47, 48. Therefore, the electrospinning process will be discussed briefly in this section. More detailed information about different aspects of this process can be obtained from the literature reviews previously mentioned.
The Electrospinning Process
The principles of electrospinning can be explained based on the schematic diagram shown in Figure 4. The basic electrospinning set-up consists of three main components: a feeding system, a high-voltage power supply and a grounded collector. The feeding system consists of a container with a liquid precursor (polymer solution, melt, sol-gel etc.), a spinneret (usually a thin metallic needle with an inner diameter of about 10 μm)33 and a pump that can inject the precursor at a constant rate. The precursor droplet at the needle tip is charged by connecting the needle to the high voltage (usually 1-30 kV)43 supply.
The force created by the electrical field acts in the opposite direction to the surface tension of the solution and elongates the droplet to a conical shape called a Taylor cone.34 When the electrical force is strong enough to overcome the surface tension, a thin jet of liquid precursor ejects from the tip of the Taylor cone. A jet is straight and stable close to the needle tip. At some point, due to instabilities in the electrical field, the jet begins to oscillate and move chaotically. This region of the jet path is called the bending instability region, and the oscillation of the jet is called the whipping motion. During this chaotic motion, the jet elongates and becomes very thin and solid due to evaporation of the solvent or solidification of the melt. Finally, reaching the collector, the jet deposits in a random manner creating a nonwoven mat composed of fine fibers.
The Electrospinning Parameters
In addition to the raw material selected, the final properties of electrospun fibers greatly depend on the processing parameters. Fibers with different diameters, morphology, secondary structures on a surface, and different fiber arrangements can be produced by varying electrospinning conditions. Uniform nanofibers are preferred for most applications, although the formation of beads or microparticles is usually considered a defect.49 However, in the field of the superhydrophobic materials, beaded fibers and particles are often favored. Therefore, the effect of processing parameters will be discussed further with respect to not only fiber diameter but also formation of beads and secondary structures.
Electrospinning parameters can be categorized into several groups, including solution properties, process parameters, equipment design and ambient conditions. Detailed discussions of the effects of electrospinning parameters can be found in refs.33, 34, 41–43, 50. Some of the most important parameters for each category are summarized in Table 1.
Table 1. Most Important Electrospinning Parameters
The morphology of resulting fibers is affected by a combination of all electrospinning parameters, so the effect of some individual parameter can be controversial. However, some general trends can be derived from the literature.
The viscosity of precursor solution should be selected in a specific range to allow formation of fibers.41 If the viscosity is too low, a jet collapses into droplets before all solvent evaporates.41 Thus such a viscosity prohibits electrospinning of uniform fibers. A solution with too high a viscosity cannot overcome its viscoelastic forces, which leads to droplet formation instead of fibers.42 Within these boundaries, an increase in solution viscosity generates fibers with a larger diameter,41, 50 but these fibers are more uniform with less or no beads.33, 40 The solution viscosity depends on solvents and polymers used and can be controlled by adjusting the polymer concentration34 and the solution temperature.40 Although higher polymer concentration increases solution viscosity, higher temperature reduces it.
The effect of polymer molecular weight is similar to the effect of viscosity. An increase in polymer molecular weight increases fiber diameter,50 but reduces the number of beads. At concentrations greater than the critical chain overlap concentration,51 uniform fibers are produced regardless of polymer molecular weight.34
Surface tension of a solution depends on its chemical composition and acts to minimize surface area trying to convert a jet into droplets, which is called Rayleigh instability.43 Therefore, a solution with high surface tension tends to form more beads and/or droplets.40, 50
Another important property of a precursor solution is its conductivity. Higher conductivity increases the influence of the electrical field on jet elongation, thus leading to finer and more uniform fibers with fewer beads.42 This property can be induced by addition of some conductive materials to the solution (e.g., ionic salts).34
Solvent volatility affects both fiber diameter and fiber surface roughness. Fast solvent evaporation reduces jet drying time and therefore, generates coarser fibers.42 In addition, usage of a combination of two solvents with different evaporation rates leads to formation of pores on the fiber surface due to phase separation effect.42 This effect occurs because fast evaporation of one solvent from a jet creates polymer rich and polymer lean regions. The first type of regions solidify quickly due to the low content of solvent and form a matrix, although the polymer lean regions form pores.52
Among process variables, the characteristics of the electrical field, solution flow rate and the distance to collector are the most important. The voltage of the electrical field should be set high enough to overcome surface tension of the solution and initiate jet formation. A further increase in voltage has a twofold effect on a jet. It increases the flow rate of the solution, which leads to an increase in fiber diameter, but at the same time, it creates more wiping instabilities in the jet motion resulting in higher elongation.50 Therefore, it can be concluded that voltage should not have any significant effect on fiber diameter.50 However, some papers reported both an increase40 and decrease42 in fiber diameter with an increase in the voltage applied. Such ambiguity can be attributed to the difference in electrospinning parameters and solutions used during experiments. An increase in the voltage also results in higher level of beads formation.41, 42 Electrospinning of a positively charged solution leads to fewer beads and more uniform fibers, compared to the negative polarity.53
Although a high solution flow rate increases the deposition rate of electrospun fibers and consequently the productivity of electrospinning, it has negative effects on fibers quality. An increase in the flow rate increases fiber diameter and the size of pores on the fiber surface.42 In addition, a high flow rate does not allow a jet to solidify completely resulting in the collection of wet fibers.50 Too high flow rate results in beaded fibers or droplets.42, 50
The fiber collection distance also affects fiber properties. At short distances, a jet does not have enough time to elongate and therefore too thick or beaded fibers are created.42 In addition, not all of the solvent evaporates and collected fibers are wet and therefore may change their shape from round to flat.42 An increase in the collection distance reduces bead density and fiber diameter.41
High relative humidity during electrospinning induces pore formation on the fiber surface due to breath figures effect.42 Evaporation of solvent cools down a jet surface leading to condensation of moisture from air, which leaves imprints in the form of pores on a fiber surface. In addition, high humidity can increase or reduce fiber diameter depending on the polymer used.50 Too much moisture in the surrounding air can inhibit solvent evaporation.34 For solution electrospinning, high temperature of the spinning area increases solvent evaporation rate, thus leading to fibers with a larger diameter.50 However, control of ambient temperature is more important for melt electrospinning to give a polymer melt enough time to elongate before solidification.54
Among the equipment parameters, the inner diameter of the nozzle has the most influence on fiber morphology. An increase in the needle diameter leads to the release of more solution, thus leading to larger fiber diameter.50 Specifically designed needles are used for coaxial electrospinning to produce core-sheath or hollow fibers.
In most laboratory experiments, a static grounded collector is used. However, electrospun fibers can also be collected on rotating drums, which is important for continuous production. An increase in the rotational speed of a collection drum up to some point can improve the mechanical properties of fibers.34 In addition, specifically designed collectors, like parallel electrodes, disc, tube with knife-edge electrodes below, collection through water bath, etc., can be used to obtain aligned fibers, patterned webs or yarn made of electrospun fibers.37
Modifications of Electrospinning Apparatus and Techniques
Many modifications of basic single-needle solution electrospinning were developed to extend the range of polymers used, produce multicomponent fibers or webs and increase process productivity. Melt, coaxial, multi-jet and needless electrospinning techniques are the most discussed in the literature.
Some important polymers for the nonwoven industry, like polyethylene and polypropylene, cannot be properly dissolved at room temperature to use in solution electrospinning. One possible way to use such polymers is electrospinning from a melt. A detailed review of melt electrospinning can be found in ref.54. Here some major points concerning melt electrospinning will be discussed. In contrast to conventional solution electrospinning, melt electrospinning requires additional heating systems to melt the polymer and to control fiber formation and deposition. In addition to a polymer container, temperature control equipment can be used at a needle, as well as in a spinning zone and/or a collector to facilitate fiber formation and deposition.36, 55 There are two main issues with melt electrospinning: very low conductivity of melts and higher viscosity of melts compared to polymer solutions. An absence of conductivity does not allow the wiping motion that elongates the jet in solution electrospinning. In addition, high viscosity leads to an increase in fiber diameter and requires higher voltage to initiate a jet.35 However, despite these limitations, scientists were able to produce fibers from melt spinning with a diameter as small as 300 nm.54 The advantages of melt electrospinning are the ability to spun nonsoluble polymers, no need for solvents33 (which is an issue for solution electrospinning), and consistency of the fiber diameter.54 In addition, melt electrospinning has higher productivity, because there is no loss of solvents, and all materials released from the feeding system are deposited on the collector.36
Combining useful properties of two polymers allows producing fibers with multiple functionalizations. Coaxial electrospinning technique was developed to utilize this advantage in producing core-sheath fibers with different properties (e.g., good mechanical properties of the core and hydrophobicity of the sheath). A detailed review of this technique can be found in ref.46. The only difference in coaxial electrospinning equipment and the basic set-up is in the design of the spinneret and feeding system. Usually two needles, one inside another, are used and each needle is fed from separate solution containers. Such set-up allows the fabrication of not only core-sheath fibers from two electrospinnable polymers, but also it allows electrospinning when a core solution cannot be electrospun by itself due to low molecular weight, low conductivity or other reasons.37 In this case, the outer solution should be electrospinnable and guide the core material. After removal of the sheath, fibers formed from nonelectrospinnable material can be obtained.46 In the opposite situation, when a core material is electrospinnable and the sheath material is not, this technique offers a one step process for fiber formation and simultaneous coating of fibers with functionalized materials.46 Also, coaxial electrospinning can be utilized to produce hollow fibers, fibers with microencapsulated compounds46 or systems with discontinuous drop-shape inclusion.33 Similar to coaxial electrospinning, a spinneret with two side-by-side nozzles can be used to produce bi-component fibers.38
In contrast to traditional fiber extrusion methods (i.e., melt spinning and melt-blowing), electrospinning relies on the electrical force that elongates a jet very slowly36 Therefore, the productivity of single-nozzle electrospinning set-up is very low (typically 0.1-1 g/h).56 The obvious way to increase productivity is to use multiple nozzles. However, in such a configuration, jets are pushed away from each other due to the same polarity of jets and resulting nonwoven webs are highly nonuniform.36 A possible solution to this problem is to use secondary electrodes to control jets trajectories.57 In addition to a higher production rate and coverage area, multinozzle approach allows production of multicomponent nonwovens, when different nozzles are fed with different polymer solutions.
Another approach to increase productivity of electrospinning is needleless electrospinning. Probably the most successful design of electrospinning for mass production to date was developed by Jirsak et al.58 In his apparatus, a slowly rotating horizontal cylinder is immersed into a polymer solution. The solution is picked up and carried to the top of the cylinder where multiple jets are formed under the influence of an electric field. Fibers are deposited on the collector above the cylinder.59 This technology was subsequently commercialized by Elmarco Company (http://www.elmarco.com), which produces needleless electrospinning equipment under the “Nanospider” trade mark. According to the company web site, the equipment has a significantly higher production rate compared to single-nozzle electrospinning, for example, it allows the production of 50 million square meters of PVA nonwoven web with basis weight of 0.03 g/m2 and an average fiber diameter of 200 nm,60 which corresponds to the production rate of about 170 g/h.
ELECTROSPUN SELF-CLEANING SURFACES
As was discussed in Section 1, two factors are required to create the superhydrophobic effect on any surface. This includes low surface tension of a material itself and a high degree of a surface roughness. Electrospinning allows the production of fibers with a diameter ranging from several nanometers to hundred micrometers, depending on the polymer characteristics and electrospinning parameters used. A low fiber diameter introduces one degree of roughness to the electrospun mats. In addition, the tuning of electrospinning parameters, using some additives in a polymer solution or posttreatment steps, allows creating the second scale of roughness to further induce superhydrophobicity of a surface.
Due to small fiber diameters, simple implementation, and a wide range of polymers and additives that can be used with electrospinning, this technique has attracted great deal of attention in the scientific community associated with producing superhydrophobic and self-cleaning surfaces. In this section, about 40 papers in the field of creating superhydrophobic electrospun webs will be discussed.
There are several dimensions that can be used to distinguish different approaches used to achieve the superhydrophobic effect on electrospun mats, including a type of electrospinning, web posttreatment methods, polymers and additives used. Considering the type of electrospinning, conventional one-jet electrospinning from solution, melt electrospinning, multi-jet electrospinning, and coaxial electrospinning were utilized by authors. Table 2 contains the number of papers and references devoted to each type of electrospinning for formation of superhydrophobic webs.
Table 2. Applications of Different Types of Electrospinning to Produce Superhydrophobic Surfaces
In addition to electrospinning, some authors used multistep procedures to induce hydrophobicity of a surface. In the case of polymers with fluorinated compounds, annealing was used to reorient the perfluorinated groups to a fiber surface.61, 88 Nonwoven webs produced from hydrophilic polymers were coated with water repellent materials using the sol-gel method,68, 73 plasma treatment77, 83 and initiated chemical vapor deposition (iCVD).65, 72 Also Ma et al.72 and Ogawa et al.73 utilized three-step processes using layer-by-layer (LBL) deposition of nanoparticles before iCVD and sol-gel coating, correspondingly, to increase fiber surface roughness. Menini and Farzaneh76 and Asmatulu et al.86 blended different kinds of nanoparticles (PTFE,76 TiO2, and graphene86) with a polymer solution before spinning to increase fibers surface roughness.
There is a wide range of polymers used for electrospun superhydrophobic mats in the literature reviewed. The authors discussed more than 30 different polymers. Among those, polystyrene (PS) and polystyrene based copolymers were the most frequently used due to the low surface energy of PS, its low cost and simplicity to use in electrospinning.76 Other polymers used vary from conventional polymers, like PET, Nylon 6 and cellulose acetate to complicated copolymers, block-polymers, and grafted-polymers. A detailed list of polymers used with corresponding references is shown in Table 3.
Table 3. Polymers Used for Electrospun Superhydrophobic Mats Production
A detailed discussion of the work done in the field of superhydrophobic electrospun surfaces will be divided into several subsections. The first subsection will be devoted to the polystyrene, where the results from conventional electrospinning, electrospinning with additives and multi-jet electrospinning will be discussed. Then superhydrophobic webs produced from other nonfluorinated polymers with a one-step process will be considered. In the third subsection, complex polymers with fluoro compounds will be discussed. Finally, the combination of electrospinning with different coating techniques will be described.
Superhydrophobic Polystyrene Webs
The polystyrene (PS) homopolymer was widely discussed in the literature to produce highly water repellent electrospun webs because this material is cheap and easy to use in electrospinning.76 It has relatively low surface energy. The static WCA on the smooth PS surface is 95°.62 However, this angle is far below the 150° threshold required for superhydrophobic materials. Therefore, authors utilized electrospinning to increase the roughness of polystyrene substrates, usually creating two-level micro/nanostructures. Several authors62, 63, 67, 70, 74 studied the influence of electrospinning parameters on the surface roughness and hydrophobicity of the resulting webs. In addition, some papers discussed the effect of additives, like nanoparticles,76, 86 ionic liquid75 and polyaniline (PANI),71 blended with a solution before electrospinning. To improve the mechanical properties of the PS web, mono- or multipolymer composite webs were produced using multi-jet electrospinning.70, 89–91
Imitating the surface structure of lotus leaves, Jiang et al.62 used solutions of polystyrene in dimethylformamide (DMF) with different concentrations keeping tip-to-collector distance and voltage constant and adjusting the feed rate. The solution with high viscosity (25% of PS by weight) allowed producing a web composed of smooth fibers with an average diameter of 420 nm, although the substrate produced from diluted solution (5% of PS by wt) was composed of microparticles (2–7 μm) covered with nanostructures (50–70 nm). The secondary structures created two-level surface roughness (microparticles/nanostructures), so the static WCA on such film was 162° compared to 139° on the nanofiber web. However, the microparticle substrate had low mechanical integrity. Therefore, the authors used the 7% by weight solution of PS, which allowed the production of porous microparticles (3–7 μm) interlinked by nanofibers (60-140 nm). This substrate demonstrated a high static WCA of 160° and was more mechanically stable. The authors did not report dynamic behavior of water droplets on this surface. Therefore, it is unknown if it possessed self-cleaning properties or not.
Zheng et al.70 conducted similar experiments with surface morphology of electrospun PS substrates. The authors varied the type of solvent, polymer concentration and molecular weight, solution conductivity, and other parameters to produce electrospun materials composed of microparticles, bead-on-string structures or nanofibers. The static WCAs for the produced microparticle substrates varied from 153° to almost 160°. The highest WCA was obtained on a substrate composed of microparticles covered with nanostructures that was electrospun from the PS/DMF solution with 5% concentration by weight. The WCAs for the bead-on-string structures were 146°–151°. The most water-repellent bead-on-string substrate was produced from the 10% by weight solution of PS in the blend of DMF and tetrahydrofuran (THF) in proportion of 75% to 25% by weight. It was composed of the beaded fibers with a fiber diameter of 50-200 nm and beads with dimensions of 5 × 3 μm. The webs made of bead-free fibers exhibited WCAs from 143° to 153°. The highest WCA was measured on the web composed of fibers with an average diameter of 1.7 μm that was obtained from the 30% by wt PS/DMF solution electrospun with increased flow rate (45 μL/min) and at elevated temperature (46 °C). Therefore, confirming the results of Jiang et al.,62 the authors found that microparticle/nanostructured substrates had the highest WCA. To increase the mechanical integrity of this substrate they covered it with a thin layer of 50 nm fibers and obtained a web with a WCA of 158°. However, the authors did not report any measurement of the dynamic water behavior on this material.
As part of the more broad work, Agarwal et al.67 also studied the hydrophobic properties of electrospun polystyrene webs. They varied the concentration of PS in DMF:THF (50:50 by volume) solution and obtained results similar to results in refs.62 and70: at higher concentrations (here 10% by wt), a web of uniform fibers was produced with a WCA of 130°. At lower concentrations (2–5% by wt), substrates composed of microparticles and nanofibers were fabricated with a WCA of 150°. In addition to the WCAs, the authors also measured the roll-off angles. For both types of surfaces, these angles were higher than 90°, indicating that water droplets stick to such surfaces and therefore the self-cleaning effect cannot be achieved.
More mechanically stable fiber-only webs with WCAs higher that 150° were fabricated by Gu et al.63 and Kang et al.74. Gu et al.63 obtained a WCA of 156° using 23% by wt of PS in THF. Kang et al.74 reported protuberances created on the fiber surface when PS is electrospun from a 35% by wt solution in DMF. This resulted in a static WCA of 154°.
Miyauchi et al.69 studied the effect of a composition of solvents with different evaporation rates on hydrophobic properties of PS webs. For this purpose, they used 30% by wt concentration of PS in THF:DMF mixture. Reducing a percentage of THF, which has higher evaporation rate compared to DMF, the researchers were able to reduce fiber diameter from 15 μm (pure THF) to about 3 μm (pure DMF) as shown in Figure 5. Hydrophobic properties that correspond to each solvent composition are shown in Figure 6.69
In addition, the phase separation effect allowed creating secondary structures on fiber surface, when two solvents were used in the solution as shown in Figure 7(b–d).69 It can be seen from Figure 7(a,e) that when only one solvent was used fiber surface was smooth. Roughening fiber surface created double scale roughness of the web and improved its hydrophobic properties. Figure 6 shows that webs that satisfied requirements to superhydrophobic surfaces (roll-off angle ≤ 10°, WCA ≥ 150°) were obtained when THF:DMF ratios were 2:2 and 1:3. In the latter case, the web was composed of microfibers with an average diameter of 4.7 μm covered with grooves having dimensions of 1.43 μm × 158 nm [see Fig. 7(d)]. This resulted in the WCA of almost 160° and roll-off angle of 8°. The calculated ratio of air–liquid interface (f2) for DMF:THF = 2:2 was almost the same as was achieved by Jiang et al.62 using the combination of microparticles and nanofibers, although the integrity of the web was significantly higher.
Two types of additives were used in the literature reviewed to produce superhydrophobic mats from polystyrene, including conductive substances and nanoparticles. Zhu et al.71 and Lu et al.75 mixed PS solution with polyaniline (PANI) and 1-butyl-3-methylimidazolium hexafluorophosphate (BMIPF6), correspondingly, to produce superhydrophobic conductive substrates. An increase in conductivity of solution allowed reducing fibers diameter and eliminating beads formation. In addition, these additives formed secondary rough structures on fibers surface and increased overall roughness of the resulted webs. Electrospun PS/PANI composite film consisted of submicron PANI spheres (300-820 nm) covered with nanoknots (40-100 nm) and interconnected by nanofibers (20-40 nm) (see Fig. 871). These substrates exhibited static WCAs of 167° (0° on the flat surface) with very low roll-off angles (<5°) showing self-cleaning properties. In addition, self-cleaning properties did not change significantly over a wide range of pH levels.71
The addition of 0.5% of ionic liquid (BMIPF6) to the PS solution allowed Lu et al.75 to produce a PS/BMIPF6 composite web of fibers covered and interconnected by branch-like structures (see Fig. 975). These structures increased the surface roughness and reduced surface energy due to the presence of CF3 groups in the ionic liquid, which resulted in a static WCA of 153° and a water roll-off angle <5°. Therefore, the web exhibited self-cleaning properties.
Another approach to increase surface roughness of PS webs was the addition of nanoparticles to PS solutions. Menini and Farzaneh76 used polytetrafluoroethylene (PTFE) nanoparticles. Their results showed that nanoparticles increased the water-repellency of the PS film compared to the electrospinning without nanoparticles. However, the highest static WCA obtained was only slightly higher than the 150° threshold. Asmatulu et al.86 electrospun PS blended with TiO2 nanoparticles. In contrast to Menini and Farzaneh,76 static WCAs for samples with different concentrations of nanoparticles were higher than 155° and a WCA increased with an increase in nanoparticles in the solution, reaching 177.5° at an 8% concentration of nanoparticles. The authors attributed such high WCAs to the nanoscale structures on the fiber surface that were formed due to the presence of TiO2 nanoparticles.
The polystyrene substrates already discussed exhibited highly hydrophobic properties. However, these webs had low mechanical properties.89 Therefore, there are several papers that discussed reinforcement of superhydrophobic polystyrene webs. In two consecutive papers, Ding et al. manufactured composite electrospun webs of polystyrene with Nylon 689 and polyacrylonitrile (PAN).90 For this purpose, they used an electrospinning device with four nozzles arranged in parallel that allowed spinning several different polymers simultaneously. By changing the number ratio of jets the authors were able to produce self-cleaning webs with improved mechanical characteristics. For the case of Nylon 6, this was achieved by spinning the 30% by wt PS solution in DMF:THF (3:1) from two jets and the 20% by wt Nylon 6 solution in formic acid from another two jets.89 The static WCA on this surface was 150° and the contact angle hysteresis (CAH) was 10°, although the tensile strength of the mat was three times higher compared to the pure PS web. Similar results were obtained for the PS/PAN mats with a threefold increase in tensile strength, a WCA of 150° and hysteresis of about 8°. This was achieved by spinning a 30% by wt PS solution in DMF:THF (3:/1) and a 12% by wt solution of PAN in DMF with a number ratio of jets of 3:1. Images of PS webs reinforced with Nylon 6 and PAN are shown in Figures 10 and 11.
Zhan et al.91 proposed a different configuration of a multinozzle electrospinning device, locating two nozzles on the opposite sides of a rotating collection drum and isolating them from each other. This configuration allows mixing fibers from two jets more homogeneously. Using this device the authors formed a composite polystyrene web of different fiber types: bead-on-string fibers (4% by wt solution of PS in DMF) to induce hydrophobicity and microfibers (20% by wt solution of PS in DMF) to improve mechanical properties of the mat. Optimizing the weight ratio of each fiber type, they improved the fabric tensile strength by about two times with a static WCA slightly above 150°. The optimum weight ratio of the microfibers to beads-on-strings fibers was 5:3. Unfortunately, the authors did not report the water roll-off angle for this surface.
In this subsection, the applications of polystyrene as a basic polymer to form superhydrophobic and self-cleaning surfaces were discussed. The electrospinning conditions and properties of resulting PS webs are summarized in Table 4.
Table 4. Hydrophobic Properties of Some Electrospun PS Mats
As can be seen from Table 4, the self-cleaning surfaces produced from polystyrene by electrospinning were only reported in a few papers. Miyauchi et al.69 found the optimal ratio of DMF:THF to produce fiber-only mechanically stable PS webs with self-cleaning properties. Later the same group utilized the multi-jet electrospinning set-up to reinforce the web obtained in ref.69 with Nylon 689 and PAN90 fibers. Another approach that led to webs with self-cleaning properties was to use some conductive additives in the solution of PS in DMF, like PANI71 or ionic liquid.75
Superhydrophobic Webs From Other Nonfluorinated Polymers
The superhydrophobic substrates of many other polymers were discussed in the literature in addition to polystyrene. In this section, the results of these studies with respect to nonfluorinated polymers will be overviewed.
Zhu et al.49 studied the influence of the surface morphology on the hydrophobic properties of web electrospun from hydrophilic polymers. The authors selected poly(3-hydroxybutyrate-co-3-hydroxyvalerate) (PHBV) for this purpose, and found that varying the polymer concentration in the chloroform (CHCl3) produced webs with different morphologies. Confirming the results from the polystyrene studies, the authors found that a web composed of rough microparticles has the highest water repellency. In the case of PHBV, electrospinning of 0.75% by wt solution in chloroform generated a film made of popcorn-like nanoparticles that created a surface with two-level roughness. This material had excellent hydrophobic and self-cleaning properties (static WCA = 158°, roll-off angle = 7°), but due to microparticle composition, its mechanical properties were very low. The most mechanically stable beads-on-strings structure produced from the 4% by wt solution had a WCA of 110° only and a sliding angle of 80°.
A similar study was conducted by Shao et al.80 However, in this case the authors synthesized the highly hydrophobic polymer poly(γ-stearyl-L-glutamate) (PSLG). The static WCA on a smooth surface made of this polymer was 119°. This polymer has a stiff helical backbone, but at the same time, it is soluble in common organic solvents due to long flexible side chains. In addition, the side chains are highly saturated with CH2 and CH3 groups that increases water repellency. The polymer concentrations in the chloroform from 2 to 16% by wt allowed electrospinning microparticle substrates with different diameters. More concentrated solutions changed the web morphology to beads-on-strings, and the concentration of 22% by wt resulted in a fibers-only web. For beaded surfaces, the static WCAs varied from 155 to 170°. Although for substrates formed of fibers, the WCAs were from 150° to 156°. For both types of substrates the WCA decreased with an increase in beads or fiber diameters. The authors also reported the roll-off angles for these surfaces. However, they were lower than 10° for substrates formed of microparticles only. The beads-on-strings and fibers-only webs had significantly higher roll-off angles. Thus these webs did not exhibit self-cleaning properties.
Duzyer et al.87 conducted an extensive study of properties of electrospun PET webs. Using three different concentrations of PET in trifluoroacetic acid (TFA) and dichloromethane (DCM) solution, they produced webs of fibers with different diameters. Higher concentrations led to fibers with a larger diameter, but lower molecular orientation in the amorphous regions. Therefore, coarser fibers had lower mechanical properties, but at the same time, only the web composed of the larger fibers (average diameter 3 μm) had a static WCA of 160.42°. Their results confirmed the conclusion made by Schonhorn and Ryan93 that the surface tension of PET increases with increase in surface crystallinity.
The hydrophobic properties of solvent and melt electrospun isotactic polypropylene (iPP) nonwoven webs were studied by Cho et al.55 For solution spinning, they dissolved PP in decalin at elevated temperature, and for both types of electrospinning they optimized and carefully controlled the temperatures of solution or melt at the needle, the fiber forming region and the collector. They measured the WCAs for a smooth PP film, a commercially available PP nonwoven web (15 μm fiber diameter), and for four samples of electrospun nonwoven mats (finer and coarser fibers from solution and melt electrospinning). The smooth PP surface had a static WCA of 100°, and the nonwoven mat had a WCA of 104°. On the other hand, all four electrospun samples had WCAs close to 150°. Therefore, the authors concluded that a diameter of electrospun PP fiber under 10 μm is sufficient to create a superhydrophobic surface. The highest WCA of 151° was obtained for solution-spun fibers with an average diameter of 0.8 μm.
Lim et al.85 polymerized methyltriethoxysilane (PMTES) in a sol-gel and subsequently electrospun this sol-gel to obtain fibrous webs. The main parameter that influenced the morphology of the resulting web was gelation time. Freestanding webs with WCAs higher than 150° composed of microfibers were produced after gelation during 672 hours at room temperature or 168 hours at 50 °C. In addition, this fabric was oleophilic, which can be utilized for oil water separation, and its hydrophobic properties were maintained even at 500 °C.
Other silicone based polymers were synthesized by Ma et al.64 and Xue et al.82 and only these two papers in this section reported high WCA, low hysteresis and produced mechanically stable webs. Ma et al.64 utilized the low surface tension of polydimethylsiloxane (PDMS), but because this polymer cannot be made into solid fibers, the authors synthesized the PS-PDMS di-block copolymer and electrospun it from a mixture with pure PS in the THF:DMF solution. The resulting web was formed of fibers with a diameter of 150-400 nm with minor beads (see Fig. 12). Due to lower surface energy of PDMS, blocks composed of this material reorients to fiber surface reducing web surface tension. The highest WCA observed on this material was 163° with hysteresis of 15°. Xue et al.82 fabricated a mechanically stable, self-cleaning web by electrospinning a copolymer of polyhedral oligomeric silsesquioxanes (POSS) and polymethylmethacrylate (PMMA). Fibers with average diameters of 2.16 μm were produced from 5% by wt solutions of polymers in THF:DMF. As in the previous case, POSS blocks reoriented to fiber surface, but in addition to lowering surface energy, these block created rough structures due to cube-like shape of POSS molecules (see Fig. 13). As a result, such webs exhibited better hydrophobic properties compared to the previous paper with the static WCA of about 169° and roll-off angle of 6°, thus showing self-cleaning properties.
In addition to the electrospinning of a PS solution mixed with TiO2 nanoparticles (that was discussed in the previous subsection), Asmatulu et al.86 also studied the effect of nanoparticles in the solution on the hydrophobicity of polyvinyl chloride (PVC) webs. Similar to the PS results, the static WCAs of PVC mats increased with an increase of amount of TiO2 or graphene nanoparticles in the solution. The highest WCAs for TiO2 (8% by wt) and graphene (4% by wt) inclusions were 165° and 166.3°, respectively.
A brief summary of hydrophobic surfaces produced from nonfluorinated polymers is shown in Table 5. It can be seen from the table that only webs produced in ref.82 and to some extent in ref.64 can be considered as stable and self-cleaning. Other authors fabricated mechanically unstable webs, webs with high water roll-off angles, or did not report measurements of dynamic behavior of water droplets on the surface.
Table 5. Hydrophobic Properties of Electrospun Mats from Nonfluorinated Polymers
Superhydrophobic Webs From Fluorinated Polymers
Highly hydrophobic properties of fluorine-containing groups were frequently utilized in the literature to electrospin superhydrophobic nonwoven webs. As was shown in Table 3, about one half of the discussed polymers contained some form of fluorine groups. Most fluoropolymers are too insoluble to use in electrospinning.66 Therefore, in most cases authors combined fluorinated compounds with soluble polymers in the form of fluoro side groups or copolymers.
For this purpose, Singh et al.66 synthesized poly [bis(2,2,2-trifluoroethoxy)phosphazene] (PBTP), a polymer that consists of inorganic backbone (NP) and fluorinated side groups (OCH2CF3). Due to CF3 groups, this polymer is highly hydrophobic and resistant to many chemicals, fire and radiation, but at the same time, due to the inorganic backbone, it is soluble in common organic solvents, like THF, acetone and methyl ethyl ketone (MEK). By varying the concentration of the polymer in THF, the authors found that a decrease in fiber diameter from 1.5 μm to 80 nm resulted in an increase in the static WCA from 133° to 155° (see Fig. 14). As in the previously described cases for other polymers, WCAs higher than 150° were obtained for webs with beads-on-string morphology (see Fig. 15) produced from solutions with low polymer concentrations (0.5–1% by wt). These webs also had very low contact angle hysteresis (<4°).
Agarwal et al.67 used pentafluorostyrene (PFS) as a fluorinated compound and synthesized its copolymer with polystyrene. In addition, the authors used PPFS to functionalize poly(p-xylylene) (PPX). PFS molecules are aromatic fluorinated rings that possess hydrophobic properties due to CF bonds and fluorine atoms. The results showed that webs with beads-on-strings morphology electrospun from 2 to 5% by weight solutions of PS-co-PPFS (see Fig. 16) exhibited such a high hydrophobicity that a water droplet was rolling on the horizontal surface indicating a static WCA much higher than 160° and a roll-off angle very close to zero. In addition, the researchers were looking for the minimum amount of PPFS content in the copolymer. They found that 30 mol % of PPFS is enough to generate the surface described above. For PPX functionalized with PFS, superhydrophobic properties were observed only for microparticle surfaces, although for fiber morphology, the WCA was 140° with a roll-off angle of 30–40°.
Chen et al.78 used perfluorooctyl-triethoxysilane (PFOTES) to functionalize and reduce surface energy of poly(vinylidene fluoride) (PVDF) and PVDF grafted with poly-3-(trimethoxysilyl)-propyl-methacrylate. PFOTES molecules consist of a triethoxysilane side, which is responsible for chemical reactivity during functionalization, and a fluoro side, which is saturated with CF2 and CF3, and therefore, is highly water repellent. By electrospinning of these polymers from the 18% by wt solution in N,N′-dimethyl-acetamide (DMAc), the authors were able to produce beads-free nanofiber webs with WCAs of 156° and 152°, respectively. PFOTES was also used by Islam et al.84 to functionalize pullulan (PULL), a naturally occurring polymer. Beaded fiber and beads-free fiber webs were produced from solutions with 9% and 12% concentrations of polymer. Both morphologies had high WCAs of 155° and 151°, respectively.
Valtola et al.81 studied hydrophobicity of webs made of di-block copolymers of polystyrene with different fluorinated polymers. The substrates electrospun from di-block copolymers of polystyrene with poly-tetrafluoro heptadecafluorodecaoxy styrene (PS-b-PFSF) and poly-perfluorooctyl ethyl methacrylate (PS-b-PFMA) had high WCAs of 160° and 150°, respectively but these substrates were composed of particles only. In addition, the authors used blends of the above mentioned di-block copolymers with pure PS. The resulting webs were formed from microparticles interconnected by nanofibers and had the same WCAs despite the fact that the weight of fluorinated copolymer in blends was as low as 10%. The WCAs did not change with decrease of fluorinated molecules because fluorinated chains tend to aggregate on the surface of materials due to their low surface energy. Therefore, the amount of expensive fluorinated compounds can be reduced significantly.
Acatay et al.61 synthesized poly acrylonitrile copolymer with α,α-dimethyl metasopropenylbenzyl isocyanate (P(AN-co-TMI)) crosslinked with perfluorinated linear diol (fluorolink-D) and electrospun this substance from different concentrations with subsequent annealing at elevated temperatures for 8 h. The annealing allows reorientation of fluorinated groups to the surface lowering the surface energy of the material. However, even with annealing, self-cleaning properties were obtained for the unstable film only with a high amount of beads. This film had a WCA of 166.7° with a roll-off angle of 4.3°. The fiber-only web had a WCA of 148° with a roll-off angle higher than 90°.
The results described in this subsection are summarized in Table 6. As can be seen from the table, webs with self-cleaning properties were obtained for morphologies containing beads in ref. 66 and 67. A fiber-only web with high WCA was obtained in ref.78, but the authors did not report the corresponding roll-off angle or hysteresis.
Table 6. Hydrophobic Properties of Electrospun Mats from Fluorinated Polymers
Electrospun Superhydrophobic Webs Prepared with Additional Coating
As previously discussed in this article, the authors tried to achieve hydrophobic properties by manipulating the electrospinning conditions to obtain an appropriate surface roughness. Mainly hydrophobic polymers were used for these studies. In this subsection, another approach will be discussed. This approach utilizes additional coatings of electrospun webs to reduce their surface energies and/or increase roughness. This allows utilization not only of hydrophobic, but also hydrophilic polymers to create superhydrophobic surfaces.
Coaxial electrospinning is the only method that allows producing electrospun fibers and coating them with a highly-water repellent layer in one step. All other approaches that will be discussed in this section utilized multistep procedures. Coaxial electrospinning produces fibers with sheath-core structure. The core polymer is usually responsible for the mechanical strength of the webs produced, although the sheath polymer adds some additional functionality to it. This method was utilized in two papers to create superhydrophobic webs. In both cases amorphous, Teflon AF, which has one of the lowest surface energies,79 was used to induce hydrophobicity. Teflon AF cannot be electrospun by itself because of its low dielectric constant, but in a coaxial set-up, another polymer acts as a guiding material and makes it possible to electrospun Teflon AF. Han et al.79 used polycaprolactone (PCL) for this purpose. Using the 1% by wt solution of Teflon AF and the 10% by wt solution of PCL, the authors obtained a fibers-only (1–2 μm) web with striation on the fibers (see Fig. 17), which increases hydrophobicity, but does not degrade mechanical properties. This web had a static WCA of 158° with very low hysteresis (3.8°), rolling angle (7°) and water bouncing behavior. In addition, compared to PCL-only webs, this web had higher strain at break (9.6 vs. 6.3 mm/mm) and stiffness (13 vs. 6.3 MPa), but lower ultimate tensile strength (2.3 vs. 3.1 MPa). Similar results were obtained by using PMMA instead of PCL showing that many polymers can be used as a core material.
Muthiah et al.92 also studied coaxial electrospinning for superhydrophobic application. In addition to Teflon AF, they used another fluorinated compound (PVDF) as a second polymer and considered both configurations: Teflon AF as a sheath and as a core polymer. In the first case, the authors obtained slightly lower hydrophobic properties compared to ref.79. The beaded morphology with a WCA of 153°, roll-off angle of 4° and hysteresis of 17° was obtained from a 7% by wt solution of PVDF and a 1% by wt solution of Teflon AF. The fibers-only morphology produced from 15% by wt of PVDF with the same concentration of Teflon AF had a WCA of 144° only and a higher rolling angle (6°) and hysteresis (19.5°). The average fiber diameters of these webs were 158 and 361 nm, correspondingly. Using the reverse configuration (Teflon AF as a core and PVDF as a sheath), a highly hydrophobic web with an average fiber diameter of 92 nm was obtained (WCA = 156°, Rolling angle = 1°, Hysteresis = 4°). Therefore, the authors concluded that hydrophobic properties are less influenced by the sheath polymer and more depend on the fibers dimension.
Several multistep approaches were proposed in the literature to achieve superhydrophobicity. Ding et al.68 utilized sol-gel coating of electrospun cellulose acetate by TEOS:DTMS mixture. The electrospinning of cellulose acetate from 8% by wt and 10% by wt solutions in acetone:DMAc resulted in a web composed of fibers (183 nm) with microparticles in the first case and a fibers-only (344 nm) web in the second case. Both uncoated surfaces were very hydrophilic absorbing water immediately after placement. After immersion of the samples into the TEOS:DTMS sol-gel for 5 seconds, subsequent drying, and a heat treatment (at 120 °C for 1 h), both samples became superhydrophobic with WCAs of 156° and 153° and roll-off angles of 10°–20° and 10°–30°, correspondingly. The thickness of the coating layer was about 80 nm.
In two consequent papers, Yoon et al.77, 83 coated PHBV and cellulose triacetate webs by plasma treatment with CF4 molecules to enhance hydrophobic properties of webs. By electrospinning the 26% by wt PHBV solution, the authors obtained webs with beads-on-strings morphology where beads were covered with pores creating two level roughness.77 This web exhibit a WCA of 141°. After plasma treatment for 450 seconds, the WCA was improved to 158°. In the second paper,83 the authors produced nonwoven webs from a 5% by wt solution of cellulose triacetate (CTA) in methylene chloride (MC):ethanol (EtOH) varying the solvent weight composition. The solvent composition of 90:10 (MC:EtOH) resulted in fibers with an average diameter of 1.7 μm, which were covered with pores due to phase separation created by the fast evaporation of MC. The ratio of solvents of 80:20 (MC:EtOH) generated fibers with an average diameter of 1.8 μm covered by wrinkles (see Fig. 18). The WCAs on these surfaces were 141.6° and 140.7° and the roll-off angles were 12.1° and 16.1°, correspondingly. The CF4 plasma treatment of the second web for 60 seconds improved its WCA to 153° and reduced the roll-off angle to 4°, creating a superhydrophobic web with a self-cleaning effect. Longer treatment time reduce hydrophobic properties as shown in Figure 19, due to smoothening of fiber surface by plasma.
A low surface energy coating using initiated chemical vapor deposition (iCVD) was studied by Ma et al.65, 72 In the first study,65 the authors obtained beaded and beads-free fibers with different diameters by varying parameters during the electrospinning of the PCL solution. The uncoated samples had the highest WCA of 139° for beaded fibers with an average fiber diameter of 110 nm and the average beads size of 2.7 μm. The highest WCA for the beads-free webs was 129° with an average fiber diameter of 580 nm. However, in both cases, WCAs decreased with time that water droplet spend on the surface, showing unstable hydrophobic properties. After the coating of these webs with a 70 nm layer of poly perfluoroalkyl ethyl methacrylate (PPFEMA), the WCAs for all samples increased to above 150° (see Fig. 20) and became stable. The highest WCAs were 175° for beaded fibers and 156° for beads-free fibers with sliding angles of 2.5° and 5° (see Fig. 21), correspondingly. Later, the authors produced a PMMA web formed of microfibers with average diameter of 1.7 μm and covered with nanopores (80 nm).72 The web was created by electrospinning PMMA from a mixture of chloroform and DMF. The web had a hierarchical, two-level structure and exhibited an unstable WCA of 147° that decreased to 74° after 25 min. When this web was treated with PPFEMA by iCVD, the WCA was 163°, which was slightly higher than for the coated web formed of smooth PMMA fibers (158°).
In the same paper,72 the authors considered the layer-by-layer self-assembly technique (LBL) to create a hierarchical two-level roughness and chemical vapor deposition (CVD) to decrease web surface energy. An electrospun nylon web of smooth fibers with an average diameter of 1.7 μm was covered with two nanoparticle bilayers by alternating immersion of the web into positively charged poly(allylamide hydrochloride) and the suspension of negatively charged 50 nm silica nanoparticles. Then the web was coated with (tridecafluoro-1,1,2,2-tetrahydrooctyl)-1-trichlorosilane by CVD to reduce the surface energy. The resulted fibers are shown in Figure 22. This multistep process improved the unstable WCA of the nylon web from 135° to a stable WCA of 168° with almost zero hysteresis.
The layer by layer deposition of nanoparticle with subsequent fluorination was also studied by Ogawa et al.73 An electrospun cellulose acetate web with an average fiber diameter of 353 nm was covered by different numbers of bilayers of positively charged TiO2 nanoparticles with a diameter of 7 nm and negatively charged poly(acrylic acid), and fluorinated by fluoroalkylsilane (FAS). The initial web was superhydrophilic absorbing a water droplet completely. The web with the highest WCA of 162° and the lowest hysteresis of 2° was achieved with ten nanoparticle bilayers (see Fig. 23). The estimated thickness of one bilayer was 14 nm. The deposition of more bilayers converted the web into a film reducing fiber surface roughness and filling web pores.
As can be seen from Table 7, special coatings of electrospun webs significantly improve superhydrophobic and self-cleaning properties. Compared with previous sections, where super-water repellency was mostly achieved with help of microparticles or beaded fibers, most papers in this section produced superhydrophobic webs with fiber-only morphology, that have much better mechanical properties compared to beads-on-strings or microparticles structures. In addition, many hydrophilic polymers were used as base materials for webs showing that additional coatings allow using a broader range of possible polymers.
Table 7. Hydrophobic Properties of Electrospun Mats with Additional Coating
The review of recent development in the field of superhydrophobic surfaces produced by electrospinning clearly indicates that this approach can be successfully applied to fabricate fibrous substrates with self-cleaning properties. Among all methods, coating of electrospun webs with some additional materials to increase roughness or reduce surface energy allowed creating webs with high WCAs and low roll-off angles (or hysteresis) in almost all cases. The main advantage of coating is that it allows utilizing both hydrophobic and hydrophilic polymers as raw material for fibers. However, with exception of co-axial electrospinning, additional coating requires several production steps.
Many authors produced webs with high WCA without additional coating. However, in most cases they did report neither roll-off angle nor contact angle hysteresis, so it remains unclear if these webs possessed self-cleaning properties or not. In addition, this set of papers is mostly limited to hydrophobic polymers. However, electrospinning allows creating such high roughness that electrospun webs made of hydrophilic polymers can be tuned to be hydrophobic without additional coating.7
It also should be noticed that many substrates produced were composed of beaded fibers or microparticles embedded into fibers. In addition, some polymers used had poor mechanical properties. Electrospun webs are weak by themselves and together with aforementioned issues these results in low mechanical stability of these webs that significantly limits their utilization in commercial applications. Therefore, electrospun surfaces with increased mechanical stability but self-cleaning properties, as in refs.89 and90, are of particular interest. In addition to strong webs, the robustness of self-cleaning properties is also very important. Testing and fabricating self-cleaning webs with improved resistance to friction and wear7 can also increase commercial interest to these surfaces.
Another limiting factor is high costs of electrospun webs. There are several factors that influence the costs of webs, including cost of raw materials, complexity of the production process and production rate. Fluorinated compounds have very low surface energy and were widely utilized in the literature reviewed. However, these materials are expensive and can significantly increase cost of resulted web. Therefore, the important topic is optimization of expensive components and defining the minimum required amount of these components to achieve self-cleaning effect, as was done by Agarwal, et al.67 Manufacturing process that includes several steps is usually more expensive compared to single-step process, due to increase in production time and requirement of additional equipment. Thus, technology that allows electrospinning self-cleaning webs without any additional treatment is more viable for commercial applications. Finally, the biggest issue is low production rate of single nozzle electrospinning. However, most studies were done utilizing this equipment. In contrast, needleless electrospinning has higher production rate and is used for mass-scale production of electrospun webs. Therefore, studying possibility of fabrication of self-cleaning surfaces using needleless electrospinning may also increase the viability of application of these webs in commercial products.
This work was supported by the US National Textile Center under the “Logistics of Closed Loop Textile Recycling” project (Project No. S09-NS04).
Iurii Sas is a Ph.D. student in the College of Textiles at NC State University with a major in Textile Technology Management. He is a member of a team that works on the “Logistics of Closed Loop Textile Recycling” project sponsored by the US National Textile Center. Iurii received his B.S. and M.S. degrees in Quantitative Economics and Econometrics from Kyiv National Taras Shevchenko University in Ukraine.
Dr. Russell E. Gorga is an Associate Professor in the Textile Engineering, Chemistry, and Science Department at NC State University and Program Director of Textile Engineering. His main interests lie in developing polymer nanocomposite nanofibers with improved mechanical and conductive properties.
Dr. Jeffrey A. Joines is an Associate Professor in the Textile Engineering, Chemistry, and Science Department at NC State University. He received a B.S. in EE and B.S. in ISE in 1990, a M.S in ISE in 1990, and Ph.D. in ISE in 1996 all from NC State University. His expertise is in supply chain optimization utilizing computer simulation and computational optimization methods.
Dr. Kristin A. Thoney is an Associate Professor in the Textile and Apparel, Technology and Management Department at NC State. She received a Ph.D. in industrial engineering and operations research from NC State in 2000. Dr. Thoney also has a M.S. in operations research from NC State and a B.S. in mathematics from Valparaiso University. Her research interests include logistics, production scheduling, inventory management, and other types of supply chain modeling involving optimization and simulation approaches.