REVIEW | Soft Matter

Progess in superhydrophobic surface development
Paul Roach, Neil J. Shirtcliffe* and Michael I. Newton
Received 15th August 2007, Accepted 1st October 2007 First published as an Advance Article on the web 30th October 2007 DOI: 10.1039/b712575p Research into extreme water-repellent surfaces began many decades ago, although it was only relatively recently that the term superhydrophobicity appeared in literature. Here we review the work on the preparation of superhydrophobic surfaces, with focus on the different techniques used and how they have developed over the years, with particular focus on the last two years. We discuss the origins of water-repellent surfaces, examining how size and shape of surface features are used to control surface characteristics, in particular how techniques have progressed to form multi-scaled roughness to mimic the lotus leaf effect. There are notable differences in the terminology used to describe the varying properties of water-repellent surfaces, so we suggest some key definitions.

Some superhydrophobic surfaces can cause water and even oils to roll off leaving little or no residue and carry away any resting surface contamination. With more and more researchers becoming interested in this field there are a greater number of methods reported to produce such surfaces, using simpler and cheaper means to achieve high water-contact angles (WCAs) and low hysteresis (causing roll-off). The number of articles on superhydrophobic surfaces published between 2006 and 2007 is more than double the total number of articles published previously, demonstrating a high level of interest in this area for all manner of uses; from self cleaning, anti-fog or anti-corrosion surfaces to basic research. A number of reviews covering this subject area have previously been published;1 some detailing a more specialised sub-set of topics.
School of Biomedical and Natural Sciences, Nottingham Trent University, Clifton Lane, Nottingham, UK NG11 8NS. E-mail:; Fax: +44 (0)115-848-6636; Tel: +44 (0)115-848-8062/3315

Dr Michael I. Newton, is a reader in experimental physics in the Division of Physics and Mathematical Sciences at the Nottingham Trent University. Following his physics degree in 1983 (BSc, Hons) he obtained an MSc in modern electronics and a PhD in semiconductor physics at the University of Nottingham. Since becoming a lecturer at the Nottingham Trent University, he has developed lithographic facilities for surface acoustic wave device Michael I. Newton fabrication, which he has now extended to include thick-film photoresist techniques of patterning surfaces for wetting applications. His research interests include the development of surface acoustic wave systems and the investigation of wetting processes on solid surfaces.

Paul Roach

Dr Paul Roach received his first degree in chemistry (MChem) in 2002 from the Nottingham Trent University, and subsequently a PhD in chemistry in 2005 for his research of protein–surface interactions. Currently a postdoctoral research fellow at the university, Paul is investigating the development of Love wavesensing devices. His research interests include the design of surfaces for biomedical use and superhydrophobic treatments as anti-fouling surfaces.

Dr Neil J. Shirtcliffe received a BSc (Hons) in chemistry in 1993 and later a PhD in electrochemistry in 1997 from Imperial
224 | Soft Matter, 2008, 4, 224–240

College, London. His PhD involved measuring electrodeposition of colloids using a QCM. He then worked in Germany at the University of Nuremberg Erlangen, and the Max Planck Insitut fuer Eisenforchung in Duesseldorf where he developed plasma polymer coatings for the protection of steel. In 2001 Neil was appointed as a postdoctoral research fellow at the Nottingham Trent University, where he has been instrumental Neil J. Shirtcliffe in combining the chemical aspects of designing super-hydrophobic surfaces with lithographic techniques. Neil’s research interests include unusual aspects of superhydrophobicity.
This journal is ß The Royal Society of Chemistry 2008

Superhydrophobic surfaces exhibit extreme water-repellency. where filling from below or rapid internal spreading (if contact angle allows) can occur. it can be energetically favourable for a liquid to bridge across the tops of surface features so that the droplet rests upon a composite surface of flat solid tops and flat air gaps between them. and (1 2 Qs) is the fraction that corresponds to the air gaps. 1 Wetting states (a) Wenzel. strong. If they become filled with water hX becomes 0u. which we highly recommend as it focuses more on theoretical aspects and therefore compliments this review. The surface and bulk properties of the material must be taken into account when deciding on a suitable method of preparation.? Can this method be used to produce a conformal superhydrophobic coating on any sized/shaped sample? Hydrophobicity is usually determined by measuring the contact angle of a water droplet contacting a surface. coshCW = Qsr(coshe) + Qs 2 1 e (3) Additional surface roughness will therefore impact on any liquid contact angle. Various applications benefit from non-wetting surfaces with some specific products becoming commercially available. ultrahydrophobic) are biologically inspired with the lotus leaf being the classic example. Highly water-repellent surfaces (super-hydrophobic.4 that marked the start of an explosion in the of number of articles published on this topic. This second case is not the same as the Wenzel case as the roughness must be filled with liquid before the drop reaches it. such as soils or sol–gels. Extrand1j gives a concise overview in a volume of a book series. Usually surfaces following the total wetting Wenzel regime are ‘‘sticky’’ in that drops of water tend to adhere to them more than a flat surface of the same type. The contact angle is then given by a weighted average of the cosines of the contact angles on the solid and air using.6 Surface texture. This may occur on surfaces with contiguous pores. The contact angle observed on this type of surface is given by Wenzel’s equation. it was a publication in Langmuir by Onda et al.1b. The two equations can be combined to produce a more general equation (eqn (3)) to cover the case when the contacting areas themselves are not flat (Fig. respectively. In 1999 the first critical review was given by McCarthy et al. However.1c Ma and Hill in ´ ´ 20061g and most recently by Reinhoudt et al. allowing a vast range of materials to be used as substrates. 1(c)). 1(a). 2008. The principles of superhydrophobicity were first outlined by Wenzel in 19362 and then by Cassie and Baxter in 1944. 4.7 It is. This difference. The greater this difference (larger hysteresis) the more water drops will stick to the surface. Fig. 1(b). gives a measure of the surface ‘stickiness’. which was built upon by Callies and Quere in 2005. coshW = rcoshe e (1) where the roughness factor r . Is the superhydrophobic coating required to be optically transparent. and the contact angle decreases below that of a flat surface. or roughness. although the underlying principles behind and effects of superhydrophobicity are still hotly disputed. or modified to give superhydrophobic surfaces.3 Wenzel suggested that if liquid contact followed the contours of a rough surface then the effect of roughness should be to This journal is ß The Royal Society of Chemistry 2008 emphasise the intrinsic wetting tendency towards either film formation or enhanced contact angle. (b) Cassie–Baxter and (c) combined models. Fig. however. with angles greater than 90u on a flat surface always increasing (except for the unusual case of preexisting ponds in the roughness) and those less than 90u potentially moving in either direction.Whilst the basic ideas of superhydrophobicity were developed by Wenzel2 and Cassie and Baxter3 decades ago and examples were observed before that.5 Here we give a concise review of the techniques used to prepare superhydrophobic surfaces. The angle between the surface and the water meniscus near the line of contact. There are therefore many different routes reported for the production of such surfaces. termed the contact-angle hysteresis. 224–240 | 225 . electrically conductive. Currently there are massive parallel efforts to resolve both the principle understanding of wetting and to define methodologies to fabricate superhydrophobic surfaces on small and large scales. Usually low hysteresis is desired when dealing with superhydrophobic surfaces as this means water droplets will roll off extremely easily. sometimes being determined by vibrating the drop. is used to enhance the intrinsic hydrophobic chemistry of the surface. for research and mass production. tough. where they started and how they have developed over the years up to present day. possible to generate ‘‘sticky’’ surfaces in the regime of Cassie and Fig. coshC = Qs(coshe) + (1 2 Qs)coshX e (2) where Qs . There is usually a difference between the angle produced as the volume of the drop is increased (the advancing angle) and that when it is decreased (the receding angle). 1 is the fraction of the surface present at the tops of the surface protrusions. Soft Matter. those following the regime of Cassie and Baxter are ‘‘slippy’’ and allow drops of water to roll off more easily than an equivalent flat surface. hX is the contact angle on the gas in the gaps which is taken to be 180u. 1 is the ratio of the true surface area of the solid to its horizontal projection and he is the equilibrium contact angle on a smooth flat surface of the same material. with water droplets resting on them with high contact angles. hard. The contribution from the roughness is contained within r and the effect of surface chemistry in he. measured through the droplet. This always leads to an increase in contact angle when the droplet bridges the gaps. producing highly non-wetting surfaces. gives an indication of the wettability of the surface (although the size of the drop can affect the outcome). flexible etc. concentrating on the most recent publications. as described by the Cassie–Baxter relationship. Theoretical equilibrium angles lie between the advancing and receding angles.

g.8 on these surfaces it is often difficult to determine values for r and Q. The two states are usually considered to be separate energy minima with an energy barrier between.g. apart from our efforts in this regard notable contributions have been made by Patankar et al. finite element and lattice Boltzmann analysis have been used for this in some cases. (e. 2(d). Developments in theoretical approaches to problems in superhydrophobicity have occurred in the last 5–10 years. but this term and its definition have changed and many recent papers use a minimum of 150u to define a ‘‘superhydrophobic’’ surface with some requiring ‘‘low’’ contact-angle hysteresis in addition. Generally.8 usually by using a surface with high intrinsic contactangle hysteresis.11 Fig. Water condensing onto superhydrophobic surfaces usually ends up in the fully wetting state. They coined the term the lotus effect . currently there is no term for this effect.1j. and several attempts to directly imprint from these have been reported. such as vibration or electrical energy can be applied to transfer drops from one state to another. most notably the wings of the cicada (Cicada orni).7 These are just a small selection of the articles published recently. but Barthlott et al.13 There have been several attempts to directly use biological structures as templates for the production of superhydrophobic surfaces. At this stage there are many reports using whatever terminology comes to mind.9b Marmur has often focussed on the metastable states of superhydrophobicity in his calculations. Such models can reveal the height of the energy barrier between states and whether conversion at a single point will spread or not. surfaces are modelled as some kind of periodic array in 2 or 3 dimensions and the free energy (global and/or local) calculated. which is now patented as an idea and a trade mark. Currently theoretical papers are keeping pace with experimental papers. with much interest geared towards various plant leaves. If the contact angle is increased by roughness to greater than 120u the term ultrahydrophobicity should be used and if the contact angle is increased to greater than 150u and the contact-angle hysteresis is less than y10u the term superhydrophobicity should be used.6.g. this is confusing as flat surfaces can exist with this contact angle. Depending upon the shape of the surface features this energy barrier can be large enough to prevent spontaneous transfer into the primary minimum.Baxter. Biologically inspired A systematic study of water repellency in plants was undertaken by Barthlott and Neinhuis. 224–240 Originally any surface with a water-contact angle greater than the maximum observable on a flat surface (#120u for PTFE) was considered to be ‘‘ultrahydrophobic’’.14 Lichen.9e Gao and McCarthy (e. 2(a). 2c. although a transition to a bridging state is possible. transition between wetting states)9g as well as Quere and his group (e. Investigation showed that two scales of waxy surface texture gave rise to the observed effect. A similar problem exists on roughened hydrophilic surfaces with some claiming that a contact angle of 0u is sufficient to define a superhydrophilic surface. Surfaces that are patterned with pillars often show this. We would suggest that if the contact angle is increased by additional roughness but is less than 120u it should be described as positive contact-angle enhancement (as opposed to negative contact-angle enhancement towards zero). investigating contactangle hysteresis). External energy. although the last two terms are beginning to attain the meanings suggested here by popular consensus.12 Other researchers have focused on producing superhydrophobic surfaces via top-down or bottom-up generation of roughness and then using either the intrinsic hydrophobicity of the material or coating the rough surface with a hydrophobic layer. 4. mould and fungus have also been shown to have superhydrophobic character.9d Another focus of research has been to link experimental results with theory. balanced on the pillars but drops condensing on the surface remain in the Wenzel state. In this review of current literature we will therefore use the term superhydrophobicity to encompass all types of surface displaying ‘‘positive water contact angle enhancement’’. Fig. 2008. making it unclear as to exactly what type of surface is being dealt with. We have kept this section relatively short as our focus is on materials development. although most of the recent examples also have low contact-angle hysteresis.9 with particular attention being given to using theory to optimise design and to predicting whether a given surface will result in Cassie–Baxter or Wenzel wetting under different conditions. The resistance of the transition from a bridging Cassie– Baxter to a wetted Wenzel state depends on the intrinsic chemical hydrophobicity of the surface and the distance between and shape of roughness features. contact´ ´ angle hysteresis). with drops deposited on the surface remaining in the Cassie–Baxter state.5 226 | Soft Matter. but water drops falling from a height fill the roughness and may become stuck. particularly for 3D systems.9c Extrand has investigated the effect of sharp edges on features and of the contact line on superhydrophobicity. This is critical for many applications as low hysteresis is the basis of most of the useful effects of superhydrophobicity. the combined Cassie–Baxter/Wenzel state can also have large hysteresis as the interfacial area between solid and liquid can be at least as large as on a flat surface.9 For this reason contact-angle hysteresis is not a measure of the wetting state of the surface. Fig. (e. 2(b).g. Examples of this are given in the work of Kusumaatmaja and Yeomans9a and a recent paper by Li and Amirfazli.10 who highlighted the lotus leaf (Nelumbo nucifera) for its super water-repellency and selfcleaning ability. Many surfaces show Cassie–Baxter bridging and allow water to roll off if tilted slightly. Water bridges over the surface roughness in a Cassie–Baxter wetting state allowing it to roll off easily and remove any debris from TM the surface. On surfaces with simple roughness the two states are easy to define. those with a deeper interest can find longer reference lists for theory in other recent reviews. also looked at a variety of insects for similar hydrophobicity. but multilayered or complex roughness can allow a mixture of the two equations to apply. Fig. There are therefore many categories of surface where roughness enhances contact angle and not enough official names to define them. with each new experimental finding being theoretically tested and integrated into calculations. A plethora of techniques have been developed to produce a variety of shaped surface features on different This journal is ß The Royal Society of Chemistry 2008 . contact-angle hysteresis)9f and Bormashenko et al.

Pulsed laser deposition of thin Teflon1 films were also shown to convey additional nanometre-sized granular roughness to cellulose fabrics. Copyright 1996 and (d) Elsevier. Fig.16 (d) cloth surface impregnated with gold particles. have been produced to give an extra degree of roughness to the fibre surface which. 2 Superhydrophobic surfaces in biology a) lotus leaf (Nelumbo nucifera).17 Michielsen and Lee15 have recently reported the surface modification of woven nylon (Fig.21 further developed the silane-coating method. Gao and McCarthy have recently demonstrated that this method can be used for the production of an ‘artificial lotus leaf’ using polyester textiles as substrates. as well as woven materials. In 2006. Copyright 2007.20 reported the production of a superhydrophobised cotton fabric by plasma coating it with a perfluorocarbon layer. along with various hydrophobic chemical coatings. (Melaleuca hypericifolia).2 Superhydrophobic surfaces can be produced simply by making woven or non-woven cloth hydrophobic without altering its roughness. using various materials from polymers to metals. some efforts have been made to generate smaller scale structures to increase the superhydrophobicity and pressure resistance of the structures. allowing it to hydrolyse and condense to form a hydrophobic layer.15 (b) droplet resting on surface shown in (a). was first critically examined by Wenzel in 1936. Copyright 2007.16 Cotton fibres were shown to be covered with nanoscale roughness after being dipped into a CNT suspension. specifically for fabrics.Fig.18 The silane reacts with moisture in the fibrous material. Fabric surfaces treated with either coating were found to be superhydrophobic. which length scales. but also having a glycidyl functional group allowing the formation of a strong bond between the This journal is ß The Royal Society of Chemistry 2008 Fig. 3(c).19 although the term ‘superhydrophobic’ (or any variant thereof) does not appear in the original patent. In wash cycle tests these plasma-coated samples TM performed better than others treated with Scotchguard . Fibres and textiles The problem of waterproofing surfaces. Zhang et al. (c) CNT-treated cotton fibre.10 c) middle of upper side of a common pond skater (Gerris lacustris)11. Daoud et al. (c) Blackwell Publishing. 3 Woven superhydrophobic surfaces (a) multifilament woven fabric. although some surfaces can be fabricated fairly easily. 3(d)). 3(a)). Single fibres. A patent published in 1945 detailed the use of an alkyl silane to hydrophobise paper or fabrics. ii) CNTtreated woven cotton sheet shown in (c) and iii) poly(butyl acrylate)– CNT-treated woven cotton sheet. As the roughness of the cloth is usually on a relatively large scale. 224–240 | 227 . sometimes expensive equipment.12 Images reprinted with permission from (a) and (b) Oxford University Press. reported artificial lotus leaf structures on cotton textiles using carbon nanotubes (CNT). 3(e). The main concern in the area of fabric-related superhydrophobic/selfcleaning surfaces is the retention of their properties with use. 2008. again using a silane with an alkyl group to present a hydrophobic character. achieving an apparent WCA of 168u.22 and similarly the impregnation of cloths with gold particles (Fig. Soft Matter. Here we describe how the preparative methods of superhydrophobic surfaces. This gave a superhydrophobic surface with increased durability. Images reprinted with permission from (a) and (b) American Chemical Society.10 b) hillock bush leaf. Many routes to prepare superhydrophobic surfaces require complicated. which were further functionalised. Copyright 2006. and d) the lichen Lecanora conizaeoides showing high roughness with inset showing water drop WCA 155 ¡ 4u. gave superhydrophobic surfaces. 4. coating and the surface groups of the cotton fibres. both the substrates and any applied over-layers. Copyright 1997. have increased in number and been developed over the past years. with poly(butyl acrylate)-modified CNTs showing another level of roughness. has afforded superhydrophobic surfaces. Surfaces which do not have inherent hydrophobic properties or multiscale roughness require additional coatings.17 and (e) water droplets on16 i) untreated woven cotton sheet. Liu et al. Fig. (c) (d) and (e) The Royal Society of Chemistry.

but tend only to be available as non-woven mats. the fibres were treated with a layer-by-layer deposition of poly(acrylic acid) and TiO2 particles and the resulting rough fibre structure was fluorinated to give a superhydrophobic surface.25 (d) cellulose acetate fibrous membrane. Copyright 2007. but also to the distinct phase segregated nature of the fibres.30 In 2004. this method of fibre production has received much interest due to the fibre size. 224–240 This journal is ß The Royal Society of Chemistry 2008 . Fig.34 Trifluoroethoxy polyphosphazene was reported to be one of the most hydrophobic polymers. Ma et al.32 The self-cleaning properties of these surfaces were discovered by Menceloglu et al.38 The as-spun fibres were smooth Fig.28 In 1994.36 Nanostructured pores were created in the fibres. Similar morphologies were observed for polystyrene spun fibres. Zhang et al.35 These types of fibres.27 This problem was solved by spinning a secondary layer of fibres over the initial fibre-particle surface.26 After electrospinning.29 These fibres had diameters ranging from 50 nm to several microns. and (e) Elsevier. 4(b). Copyright 2005. reported the superhydrophobicity of an electrospun block copolymer (poly(styrene-block-dimethylsiloxane)).23 Agarwal et al. 4(e). Ma et al. with a WCA of 172u and very low hysteresis being reported. reported the electrospinning of a range of fluorinated homoand co-polymers to produce a variety of textured surfaces.24 (c) porous electrospun fluorinated fibres.and stain-resistant surfaces were suggested using poly[bis(trifluoroethoxy)phosphazene] spun fibres. resistant to oils and having high radiation stability they had obvious applications in biomedical and advanced materials.23 (b) electrospun fluoropolymer mat.33 Dust. Reneker and Dzenis proposed a new method for the production of continuous oriented fibres of both synthetic and natural polymers by electrospinning. may be useful in tuning material properties whilst controlling the external surface chemistry.26 (e) micro-bead connected fibres by elecrospinning. showing that the water repellency was due not only to the chemistry of the surface and the structure of the fibrous mat. it was found that the wettability of mats made from smaller diameter electrospun fibres was decreased. electrospun surfaces showed excellent water repellancy. However. 4(d). although it was observed that some of the connecting particles were often removed from the surfaces by water. although some have shown that hydrophilic polymers can directly form a superhydrophobic surface.24 By changing the process conditions fibres ranging from uniform cylinders to globules were formed. having internal self-assembled structures. Fig. Block copolymer electrospinning has also been shown to result in fibres having concentric ring or aligned core–shell microphases.25 Ogawa et al. surface texture and composition along the fibre length and the wide variety of polymers that can be used. with some nano-fibres being connected by micro-particles. by electrospinning PVA and ferrous acetate. Electrospun fibres can be very small. 2008.31 As might be expected for rough and hydrophobic surfaces.37 Such surfaces have high contact angles but are not usually ‘‘slippy’’ as the fibres penetrate into the water more than halfway and prevent it from spreading purely by being discontinuous. The technique has been developed to produce smaller fibres. magnetic and superhydrophobic carbon nano-fibres. Secondary coatings to hydrophobise materials are often required. Micro-/nano-porous fibres biologically inspired to imitate the self-cleaning properties of the silver ragwort leaf were formed by carefully controlling the solvent during electrospinning polystyrene. with fibrous mats having WCAs of nearly 160u. 4(a). Fig. Copyright 2007. (d) Institute of Physics. Other work has led to the inclusion of additives in the fibre-spinning process to vary morphology or to add other properties to the fibres. 4 Superhydrophobic fibre surfaces (a) water droplets on a block copolymer electrospun fibre mat. with more and more materials being used and various crosssectional shapes being reported. Fig. 4(c). reported the use of an additional particle layer on cellulose acetate fibres. have also shown higher contact angles and lower hysteresis using fibres ‘decorated’ with nanometre-sized pores or particles. Fig. (b) (c) and (f) Copyright Wiley-VCH Verlag GmbH & Co. 4.27 Images reprinted with permission from (a) American Chemical Society. KGaA. Production of synthetic filaments with the aid of electrostatic forces is known as electrospinning. The reduced degradability of electrospun fibre mats compared to cast polymers also made them appealing to the medical/biomaterials industry.may become damaged or diminish with frictional wear or repeated washing cycles. described the preparation of conductive. 228 | Soft Matter.

which phase separate internally.39 (b) sol–gel foam produced using acetone as co-solvent. such as temperature or pressure.41 (d) and (e) phase-separated block copolymer films. Phase separation is of interest for the generation of superhydrophobic materials due to the low cost and ease of production. highlighting other potential properties40 or combining the technique with others such as electrospinning36. and (f) The Royal Society of Chemistry.52 Materials produced generally have features of a single size so WCAs are not always as high as those observed on surfaces with multi-scaled roughness. A liquid phase could be removed and the solid does not fall apart because each phase is continuous i. Many efforts are focused on moving this method towards technical production due to its simplicity compared to most of the others described in this report and the ease in forming a conformal coating. this was followed up shortly afterwards by Rao and Kulkarni45a and Shirtcliffe et al. The structure is initially very fine.21 A few researchers diverge from this. 224–240 | 229 . The transparency and hardness of the materials generally decreases as the structures become larger. 4. but after calcination.42 (f) water droplet on an organic xerogel (scale bar = 1 mm)43 Images reprinted with permission from (a) American Physical Society. (d) and (e) from American Chemical Society.41 polycarbonate47 and polystyrene48 as well as some fluoropolymers.45b Another group cooled polypropylene solutions in solvent mixtures to produce the polymer equivalent. However. Another advantage of having a bicontinuous structure is that it can be cut or abraded to Fig. Copyright 2005.44 producing rough films using a sol-gel. the solid is a single piece and the liquid can evaporate. If a stable mixture becomes unstable due to some change. but coarsens over time as interfacial surface area is reduced. Phase separation Rough surfaces can also be formed by the phase separation of a multi-component mixture. The WCA on the fibre mats changed dramatically from y27u on the PVA based as-spun fibres to y157u when calcined. which can be deposited from solvents.50 Block copolymers have also been used.51 Phase-separated structures usually have microscale structure. often resulting in structured surfaces. one of which might be a solid or become solid at some point in the process. A recent paper describes a method for producing films with small enough structure that they are optically transparent using silica sol–gel and poly(acrylic acid). 5. such as cooling.42.43 allowing various surface chemistries and topographies to be investigated.40 (c) superhydrophobic PVC film. Fig. Fig. due to a change in chemistry and topography of the surface. where each of the two new phases forms a 3-dimensional network and these networks interpenetrate. may result in remixing if the conditions revert.49 and micro-contact printing. Bicontinuous structures have been used for many years as filters and chromatography stationary phases. Copyright 2007 and 2006. the structure eventually collapsing to form two layers (or possibly more if there are more immiscible components). 4(f). although smaller features are possible these materials are more prone to collapse during drying as the size is reduced. respectively. Copyright 2006. 2008. A notable example is the coating of cotton fibres with a sol–gel-derived material. 5 Examples of phase separation. (a) model of a bicontinuous structure. (b) and (c) from Elsevier. Usually phase separation induced by an environmental change.e.with an average diameter of 180 nm. 5(a). and the possibility of creating various shaped substrates by casting and coating. 20– 30 nm diameter Fe3O4-filled carbon nano-fibres were formed with ball-like nano-textured surfaces. Their use as superhydrophobic surfaces began later with Nakajima et al. if one component solidifies at the point when a bicontinous structure is formed the second phase may be removed to create a solid 3D porous network. Copyright 2001. it will begin to separate into two phases. these films require a coating of a hydrophobic material to render them superhydrophobic. Fig.46 Recent publications in this area have increased the polymers available to include polyvinyl chloride. The size of the pores in the solid structures formed can be controlled if the rate of phase separation can be altered relative to the rate of solidification. This journal is ß The Royal Society of Chemistry 2008 Soft Matter. One of the possible ways that separation can occur is through a bicontinuous structure.

but are also conductive and 60% transparent in the visible range.57 Amphiphilic inorganic materials Inorganic materials can also crystallise to form. which show the same switchability. with one of the simplest routes to a rough superhydrophobic surface reported by Han et al. More importantly. who controlled the cooling rate of an alkyl ketene dimer (AKD) to form a fractal crystalline surface with a water contact angle of 174u. The first report of an inorganic material demonstrating reversible switching between superhydrophobic and superhydrophilic states made use of densely packed ZnO nanorods. During crystallisation various parameters can be altered to influence the size and shape of crystals including the rate of cooling and solvent evaporation or addition. such ZnO films display superhydrophobicity even if they are sparsely spread over the surface. ZnO surfaces regain their superhydrophobic character after being left in the dark for a few days. and aluminium. increasing WCA to a maximum. 224–240 This journal is ß The Royal Society of Chemistry 2008 .63 A recent publication suggested a plasma-enhanced chemical vapour deposition route to form superhydrophobic surfaces of silica Fig. or a hydrophilic material can be additionally coated.54 These showed high WCAs (154u) with a tilt angle Crystal growth Complex patterns can be generated by crystal growth. Copyright 2007. becoming wetted after the UV-generated creation of free electrons or holes on the material surface.58 A recent study found that the AKD surface develops further surface roughness over a few days. from assemblies of simple crystals to fractal snowflakes. and (e) Institute of Physics. respectively.61 Other materials also form rough crystalline surfaces. many semiconductors are (super)hydrophobic in the dark but become (super)hydrophilic when exposed to light. which were both hard and transparent. Rough surfaces can be formed on top of a crystallising liquid or by adding a surface to the crystallising system.54 (c) CuS-coated copper oxide. Copyright 2005 and 2006. Fractal aluminium oxide surfaces have been shown to be super amphiphilic.reveal a ‘fresh’ superhydrophobic surface if it becomes contaminated. 4.65 The low surface energy of the (001) plane at the nano-rod surface combined with the feature size means that high WCAs can be achieved.66 with ZnO being more unstable in UV light compared to other photocatalytic metal oxides.8u. Copyright 2005 and 2004. 6(e). formed by anodic oxidation and then coated with hydrophobising agents.56 and (e) transparent superhydrophobic alumina–silica composite film. and some of them are hydrophobic enough to become superhydrophobic when rough. although it has been reported that photo-corrosion can be a problem. enlargement shows nanostructure.57 Images reprinted with permission from (a) and (b) American Chemical Society.67 Aligned SnO2 nano-rod surfaces have also been reported for use as superhydrophobic surfaces. Silver nanostructures prepared via a galvanic cell reaction using AgNO3 and HF have shown superhydrophobic properties after subsequent functionalisation with an akanethiol. (c) and (d) The Royal Society of Chemistry.62 By stretching a thin sheet of poly(tetrafluoroethylene) Teflon1 fibrous crystals with a large fraction of void space between them were formed.4 The formation of a fractal triglyceride surface has recently been reported.64 However. in 2004. 6(b). Superhydrophobic surfaces are then afforded if the material is intrinsically hydrophobic. This property may be attributed to the material being naturally hydrophobic. 2008. having a water-contact angle of 165u. giving water-contact angles as high as 171.53 (b) silver aggregates deposited on a silicon wafer.68 These oxide films not only show inherent water-repellent properties that are UV switchable. 230 | Soft Matter.55 (d) flower-like tin oxide structure.59 Fractal surfaces of wax achieved by natural cooling have also been documented. Another explanation is that the materials are in fact hydrophobic due to contamination. 6 Rough surfaces by crystal growth (a) cobalt hydroxide crystalline nano-pins (brucite-type) with diameter of 6. Crystal growth was first used to prepare superhydrophobic surfaces by Tsujii et al. which is photodegraded making the material hydrophilic when exposed to UV (and oxygen).. Fig.60 Another recent report in this area makes use of the random crystallisation of n-hexatriacontane. Fig. having a WCA of 110u upon initial deposition and becoming superhydrophobic as the fractal surface grows.5 nm. such as TiO2. respectively.

Nano-rod arrays are of particular interest because superhydrophobic surfaces generated from them are highly resistant to pressure. depending on the method used. Deposition can occur from dissolved species or from the gas phase. The first report by Gleason et al. achieving inexpensive superhydrophobic surfaces on a variety of substrates.72 This journal is ß The Royal Society of Chemistry 2008 Catalysed growth.56 These were prepared via thermal oxidation of a tin organometallic precursor. Polystyrene nanotubes formed in this way show high water-contact angles but also cause droplets to stick to their surface.76 Porous aluminium oxide Aluminium oxide layers can be grown on aluminium metal under anodic potentials in acid. Since visible light is in the range 350–750 nm a surface would have to have features less than 100 nm to be completely transparent to visible light. Usually gas-phase reagents condense on the catalyst particles and form a solid or liquid solution. nanostructured films of WSe2 and MoO3/MoO271 have been reported and. 6(a).69 A two-stage preparation of an inorganic surface which mimics the features of a lotus leaf has been carried out. Eventually a fibre or tube starts to grow from each particle. Nano-columns can also grow from screw dislocations. with a water droplet being supported with a contact angle of 178u on a dense array of dodecanoic acid-coated. Roughness is generated on the order of the crystallite size or smaller. etc. Once this type of growth begins it can generate structures with very high aspect ratio. visible transparency. or formed on the surface.5 nm. They can also often be applied conformally to complex shapes. Nano-columns can also be formed using a sputtered array of metal particles to control the growth. which depend on a material being transparent. Fig.lower than 5u..73 Other coatings. There are many optical applications for superhydrophobic surfaces. It is often necessary to add a monomer to the gas phase during plasma etching to repair any Soft Matter. The porous arrays have been used to template aligned nano-columns of various materials including carbon. rigidity. Zinc oxide crystals can grow to form arrays of oriented nano-columns either by vapour deposition or from a solution phase. Carbon fibres can be grown in this manner using transition-metal catalysts. The screw dislocation is usually visible in the centre of the column. such as thiol SAMs and ZnO. although these only have a water-contact angle of 125u. have extremely high specific surface areas and can be transparent to optical wavelengths. 6(c).77 Titanium oxide also forms similar patterns and can be used to generate superhydrophobic surfaces when coated.74 More recently. 4. depending on the substrate used. Usually growth takes place at high temperatures as diffusion through the metal must occur. which then grow in spirals to form cones. the ability to tune the film structure during vapour deposition could lead to a range of surface wettabilities upon further chemical modification. This is often due to differences in the relative rates of etching of different crystal planes or of the matrix compared to crystalline regions. Early etching methods utilised plasma and ion etching or laser ablation of polymers. such as zinc and cobalt. have been used. SnO2 ‘nano-flowers’ have also been grown and demonstrated to have superhydrophobic character. described how a water droplet could be supported almost as a sphere on PTFE-coated nanograss. If distributed nuclei are present. The simplest mechanism is where conditions are arranged so that the crystal face parallel to the surface grows fastest. either on top of the particle or below it. Such films are often termed ‘nanograss’ or ‘nanoforests’ and can be chemically modified to present a rough and hydrophobic surface.75 A general review of the wetting properties of carbon-nanotube films has been given by Liu and Jiang. these show a high degree of hydrophobicity when hydrophobised70 Recently. either straight away or after treatment with a hydrophobising agent. This mechanism can be used to generate suitably rough surfaces to cause superhydrophobicity.55 Progression of materials and techniques used to fabricate superhydrophobic surfaces has led to the investigation of other properties that are desirable for many applications: their strength. hardness. Carbon nanotube arrays are the most popular of these surfaces. firstly forming micro-globular CuO which is subsequently treated with sulfur gas to produce CuS nanoroughness.78 Differential etching Although etching can polish surfaces it generally increases the roughness of the substrate. initially PTFE.53 The gas-phase coating of polymethylsilsesquioxane nano-filaments at atmospheric pressure without the need of a carrier gas was described by Seeger et al. therefore diminishing the transparency of a material. The growth of the structure is determined by the size mismatch between aluminium oxide and the underlying metal as well as electrostatic repulsion between the pore walls mediated by the electric potential. the latter having the advantage of being switchable from superhydrophobic to superhydrophilic as discussed above. carbonnanotube coatings have been applied as a secondary coating on top of micro-machined posts to give 2-tier roughness – mimicking the lotus leaf. As the scale of surface roughness is increased the scattering of light also increases.83 but other polymers are also suitable such as polypropylene. but the window of conditions allowing organised growth is often small. cobalt hydroxide crystalline nano-pins (brucite-type) with diameter of 6. this leads to the growth of pillars. polymers and metals. Usually a lattice misfit between the adlayer and the substrate gives rise to distributed dislocations. There are a few mechanisms for the growth of these materials. The oxide forms nano-pores in a hexagonal array with sizes determined by the potential used. 224–240 | 231 .1b This results in considerably roughened surfaces and can be carried out on a relatively large scale.. Fig. Fig. 2008. Nanostructured flower-like crystals can also be prepared by the controlled crystallisation of polyethylene from xylene by the addition of a non-solvent (cyclohexanone). 6(d). Nanostructured crystals Crystal growth. One of the highest water-contact angles achieved so far was reported by Zhou et al. lifting the particle as it grows. but silicon oxide and other materials have also been generated. This most commonly occurs in oxides of metals.

and (f) submicron pillar structures in p-type silicon after buffered oxide etching.93 A femtosecond laser has also been used to create micro/nanoscale roughness on a silicon wafer. (c) and (f) American Chemical Society. which can be treated with a fluorocarbon to afford superhydrophobic character.1b Nanostructured poly(ethylene terephthalate) has been prepared by domain-selective oxygen plasma etching.79 Recent developments using etching to engineer materials have widened the range of metals used to include titanium materials. to prepare nano-pitted aluminium. preferentially dissolving the dislocation sites in the metal grains to give a rough surface. (d) and (e) Elsevier. have suggested a nitric acid and hydrogen peroxide wet chemical etch solution for the treatment of steel and copper alloys.85 Oxygen plasma etching has also been used to prepare optically transparent poly(ethylene naphthalate) and polystyrene. Later. with the laser fluence being varied to control the induced morphology. 7. and the large size of samples – depending on the size of the plasma chamber used. and (b). wet etching of metals was investigated. 4.80 (d) silicon wafer/ photoresist layer over-etched by an inductively-coupled SF6 plasma before cleaning. 232 | Soft Matter.92 This is a two-stage process wherein small pores are formed first. oxygen plasma processing has been used to produce transparent poly(methyl methacrylate) surfaces.8 Aluminium alloy has been used more recently to form surfaces with micro-protrusions and nano-particles. 224–240 This journal is ß The Royal Society of Chemistry 2008 . then an additional wet etching step modifies the porous layer to a pillared structure. 2008. to produce multi-scale roughness.84 Higher WCAs have been achieved by argon plasma etching polypropylene in the presence of poly(tetrafluoroethylene). Fig.hydrophilic damage that occurs. Qu et al. with an additional spun-coated layer of either perfluorononane or vinylterminated poly(dimethysiloxane) to give superhydrophobic surfaces.80 Another way to achieve 2-tier roughness of silicon was reported by Kim et al. they were then used as a surface mask during plasma etching. Aluminium. forming nanostructures which were then treated with organosilanes to produce superhydrophobic surfaces.90 The TiO2 layer was etched using a RF plasma using CF4 as etchant.95 Aluminium discs were formed on a polyimide film by a standard lift-off technique. using a combination of CF4 glow-discharge etching and masking with copper nano-dots.81 and (e) after ultrasonication to remove residual photoresist. These substrates were etched by a dislocation etchant. Fig.. 7 Etching (a) roughened aluminium alloy. Ion etching followed by oxygen glow-discharge was used to roughen FEP-Teflon forming a superhydrophobic surface. 7(b) and (c).86 More recently. a technique which does not require high temperatures and so can be used for the preparation of roughened surfaces using a variety of polymeric materials.2 kJ m22 and (c) using 5 kJ m22. achieving WCAs up to 150u. the investigation of superhydrophobic surfaces for the reduction of ice-crystal formation steered Narita et al.89 Copper etching was also used by Shirtcliffe et al. The rough surfaces were then chemically modified to give high WCA and low hysteresis surfaces. In 2000. The crystallisation of metal alloys can also be controlled to produce rodlike crystals that can be revealed by selective etching of the eutectic. zinc and copper are polycrystalline metals which have been used to form superhydrophobic surfaces by wet chemical etching. Copyright 2006 and 2007. Copyright 2006 and 2005.88 Large pits were formed by electrolytic etching followed by smaller pits introduced by anodic oxidation.79 (b) laser-etched silicon surface in SF6 3. and hydrofluoric acid/hydrogen peroxide for titanium alloys. An article examining the adhesive properties of nanostructured surfaces – mimicking gecko foot hair also uses a deposited metallised mask to form nano-features. with a total processing time of a few minutes. allowing many materials to be made superhydrophobic with relative ease.82 Images reprinted with permission from (a).91 Silicon can also be roughened using ethanolic hydrofluoric acid and anodic etching.94 These surfaces were treated with hexamethyldisiloxane to give high WCAs and low hysteresis.87 The advantage of this reported technique is the speed of the method. After functionalisation with a fluoroalkylsilane the surfaces showed superhydrophobic character. Fig. etching time determining the height of the features exposed.

Research publication on these has slowed as the technique has matured. so were not investigated until somewhat later. As the structures get larger the same effect occurs on their sides. bond permanently or desorb. The appearance of such surfaces and their branching characteristics depend upon the growth process. either in solution or in gas phase. One report highlights the ability to transform a cheap polymer film into a transparent superhydrophobic film with a plasma polymer by ensuring that the roughness scale remains small. but they usually resemble cauliflower florets.107 The early reported are still referred to in papers. 2008. Initially this has little effect upon the surface morphology.105 Other plasma techniques being used are cheaper and cover larger areas.100 Fluorocarbons were often used. they may later move around on the surface. Early research was carried out using plasma-deposited polymers. as plasma-deposited polymers are relatively hard and the process is often used to deposit technical coatings on high value or small components. some of them are quite strong.8 (d) an electrodeposited amorphous TiO2 thin film. Copyright 2007. Rough surfaces are most efficiently formed when the approaching species bond instantly to the surface.101 These must usually be coated with a thin layer of hydrophobising agent to make them superhydrophobic. many of these have the useful property of being switchable from conducting and hydrophilic to non-conducting and hydrophobic.96 (b) electrochemically deposited copper at 100 mA cm22. but any protruberation gathers more material than the surrounding area. This roughness growth can occur during electrodeposition and gas-phase deposition. but most are easily damaged and few are transparent as the fractal patterns have structures on many length scales. Copyright 2005.96. respectively. 2005 and 2001. 224–240 | 233 .97 and (e) HMDS plasma-deposited polymer. and (c) 200 Ma cm22.98 Images reprinted with permission from (a). Initially metals and metal oxides were used. but silanes were also popular. approach the surface and adsorb. such as an example using electroless deposition onto areas of a surface selected by chemical Fig. and (b) and (c) American Chemical Society. including expanding arc plasma using a low vacuum and completely atmospheric pressure techniques. often as standard superhydrophobic surfaces.102 being joined by copper. This journal is ß The Royal Society of Chemistry 2008 Soft Matter. generating a branching structure with some fractal character. Others add different techniques together to produce more complex surfaces. 8.108 Also apparent is a general move towards generating nanoscale roughness. An example is silver nanostructures prepared via galvanic replacement.Diffusion-limited growth processes The formation of roughness during deposition of layers of material is common and is often considered to be an unwanted side effect. The superhydrophobicity of these surfaces was recognised around 1997.106 Although the methods are different the properties of the films and their structures are similar to those initially studied. Growth is concentrated at these points so when any roughness is generated by chance it increases rapidly.104 The surfaces.99. 8 Diffusion-limited growth on surfaces (a) plasma-deposited Teflon structures. Electrodeposition is more accessible to most research departments and allows the formation of similar fractal structures if conditions are set so that the deposition is diffusion limited. (d) and (e) Elsevier. being fractal. Most of the more recent publications concentrate on methods more suitable for industrial realisation as most basic research has been covered. with zinc oxide being used in 2003.99 The method showed early promise. Electrochemical techniques using electroless deposition also show greater suitability to industrial application. are usually very highly hydrophobic. Conducting polymers have also been used.54 which showed high water-contact angles and low hysteresis after hydrophobisation. Generally some species. 4. which means that the concentration of unbound species at the surface is effectively zero and the rate of deposition is only dependent upon the flux of material – diffusion-controlled deposition. Fig.8 gold103 then titania97 in 2005.

although it is possible to generate very smooth layers using this technique.. It is also possible to use a laser or particle beam to etch the surface directly or to expose a photoresist layer.115 (d) photolithographic towers.122 SU-8 features have been investigated further by Suzuki et al. A similar silicon-processing technique was reported by Zhu et al. for instance in photoresist. 9(d). X-ray. (d) Elsevier. but this is relatively slow.114 (c) diced silicon wafer.8b (h) silicon islands and (i) silicon nanowires grown on those silicon islands. and (e) and (g) Institute of Physics. Fig. 2008. respectively. Copyright 2006 and 2004. although more recently direct applications have been suggested.112.109 The combination of diffusion limited growth and etching has been reported for the production of a nanoneedle array110 Another example is a growth pattern of an organic salt where crystalline structures are affected by diffusion limitations. including square posts from 20–140 mm height and side lengths 2–128 mm. 9(a) as well as staggered rhombus. 2002. leaving a positive or negative image of the mask on the surface. 9(g). who looked at the wetting of hexamethyldisiloxane plasma-coated features with various size Fig.120 although these features were of low aspect ratio (height/width) they did show a change in wettability with respect to feature height.111 Plasma polymers are also used as thin hydrophobising coatings on other roughened surfaces. The patterned surface is then either used as is. Fig. allowing multiple copies to be made. with micrometre-sized features being standard and the newly established nano-imprint lithography (NIL) allowing smaller patterns to be produced. Shirtcliffe et al. Lithographic processes are often used to produce master surfaces.113 Lithographic techniques Lithography encompasses many different types of surface preparation in which a design is transferred from a master onto a substrate surface. 2006 and 2007. (b). 9 Lithographic surface modification (a) photolithographic towers and (b) indented square posts. it is often unclear how smooth these layers are by themselves. Oner and McCarthy114 produced a larger range of feature sizes with patterns etched in silicon. 9(b). Copyright 2000. For this reason it has mostly been employed to generate surfaces allowing theories of superhydrophobicity to be tested. or used as a mask on the substrate for deposition or etching. 234 | Soft Matter. Circular pillars were produced with diameters from 2 to 40 mm and up to 80 mm in height Fig.118 (g) SU8 towers. growing an extremely rough surface. which are then used as templates for the casting of the desired surface features in another material. Fig. (h) and (i) American Chemical Society. Similar surfaces were used in a later study to assess electrowetting on model superhydrophobic surfaces.patterning. Copyright 2006.116 giving square pillars in the range 10–85 mm. etc. 224–240 This journal is ß The Royal Society of Chemistry 2008 .119 Images reprinted with permission from (a).116 (e) silicon nano-towers. Lithography is useful for generating superhydrophobic surfaces where the shape of the features and the pattern is well-defined.117 (f) laser-modified SU8 surface. (f). In photolithography a photoactive polymer layer is irradiated through a mask followed by developing stages where either the exposed or unexposed polymer is removed.and star-shaped structures. (c). Photolithography can de sub-divided into different categories depending on the radiation used: UV. e-beam.8 reported a technique using patterns in SU-8 photoresist. and more recently by Dorrer and Ruhe121 to generate smaller posts. 4. One of the first reports using photolithography to produce 3D surface features for the investigation of wetting was by Kawai and Nagata in 1994. These types of lithographic templating are covered in a later section. Methods closest to the original meaning of the term involve contact between an inked stamp and the substrate.

2008. Electrostatic repulsion can be exploited to generate films of spaced particles. with a nickel relief being electroformed around the original. dip-coating or reverse-dipcoating methods. The arrays of particles are also photonic crystals and therefore display useful and attractive optical properties.. Electron-beam etching. Silica particles are often used to form hexagonally close packed arrays with particle sizes ranging from a few nanometres to a few hundred micrometres. 9(e). the stamp was then used to create replica leaf-type structures in photoresist by UV-NIL with WCAs being measures as high as 168u. allowing surfaces to have tailored hydrophobicity.and shape as well as inter-pillar spacing. 9(h) and (i).115 in 2002 employed a mechanical dicing saw to produce stripes and square pillar structures in silicon wafers. 9(c).136 one of the most recent articles on this subject details the use of a colloidal array of polystyrene beads coated with carbon nanotubes. adhering sparsely packed CaCO3-loaded hydrogel spheres onto a surface to act as a template for later adsorption (and aggregation) of silica or polystyrene spheres. although these are generally less ordered. Water-contact angles on these surfaces coated with a hexafluoropropene layer were measured at 165u. The potential for industrial applications This journal is ß The Royal Society of Chemistry 2008 requires surfaces to be processed rapidly and to be reproducible and cheap. More recently. Fig. Polymer spheres can also be used to make ordered superhydrophobic surfaces131 with a recent article describing the use of functionalised silica/latex particle suspensions as superhydrophobic coatings. thus Soft Matter.123 An alternative technique used by Yoshimitsu et al. this has been achieved in several ways including. After dissolution of the leaf. the impressive fabrication of a dense array of nano-sized silicon structures has been reported over a large sample area (4 cm22). Fig.134 The superhydrophobicity of these arrays can be increased by adding a second layer of roughness to the particles. Feature sizes as small as 50 mm were achieved although smaller features would be very difficult to prepare using this technique. Rosa and Nelumbo leaves.118 used high-energy excimer laser pulses to create microstructures in SU-8 by photochemical laser ablation and. can be applied to fairly large surface areas and does not require any specialised equipment. Aggregation/assembly of particles Colloidal particles can form close-packed assemblies on surfaces by either spin-coating.127 From the first report in 1994. lithographic techniques have been developed to reduce feature sizes and yield larger structured areas. Cao et al. Investigation into the shape and curvature of the feature edges suggested that these play an important role in determining advancing WCAs.128 More random arrays of particles are also useful for creating superhydrophobic surfaces and also often give rise to higher contact angles than organised structures. Such arrays can be functionalised by silanes to produce superhydrophobic surfaces.113 Wagterveld et al. a cycle that could be repeated after the sample was dried.124 The height/depths were in the range of 116–792 nm and the diameters 105–157 nm.137 Particle aggregation was then controlled with solution acidity forming a range of surfaces having different aggregate density. A nano-imprint lithographic process has been used to fabricate an ordered array of grooves into a silicon substrate. nanoscale roughness was introduced as a result of the generation of debris.135 sputtering gold on top and heating to form gold nano-clusters. however. 4.132 Aggregated particle surfaces with tuneable wettability have also been demonstrated. By utilising the attractive Van der Waals forces between particles a tightly packed layer (if a monodisperse colloid is used) can be formed. A recent article describes the grafting of pH and temperature-sensitive amphiphilic block copolymer brushes to silica spheres to produce switchable surfaces. introducing ordered roughness on the order of the particle size used. 9(f). allowing samples to be produced having overhang structures. using a poly(styrene-n-butyl acrylate–acrylic acid) polymer-sphere array. SU-8 has also recently been used to produce patterned micro-arrays with hydrophilic areas surrounded by superhydrophobic zones. by first imprinting a photoresist layer with a stamp followed by ozone treatment and wet etching. Siliconpillared structures 1–4 mm tall with 1–2 mm diameters were prepared using X-ray lithography by Furstner et al. Another approach was reported by Wang et al. Another colloidal crystal. the angle was not dependent on the silica particle size. These surfaces were superhydrophobic to water droplets even though the intrinsic contact angle of the silicon is only 74u (as were those of Shirtcliffe et al. Fig. 10(b). nano-imprinting has been used for the production of replica plant leaves.125 This article is particularly relevant when considering biologically inspired superhydrophobic surfaces as lithographically prepared surfaces are compared to microstructured copper and aluminium foils and replicates of Alocasia. Fig.138 The template spheres act by displaying a markedly different hydrophilicity compared to the bare substrate. fabricated from polystyrene-blockpoly(methyl methacrylate)-block-poly(acrylic acid) was reported to have similar characteristics. Upon immersion into water the wetting mode switches from Cassie– Baxter to Wenzel and the surfaces become fully wetted. Fig. with the tips of these features being further ‘sharpened’ by thermal oxidation and removal of the oxide. as a side effect of the ablation process.126 Water-contact angles as high as 167u were measured on these surfaces after treatment with octadecyltrichlorosilane.130 Here advancing WCAs were measured at 150u. Recently.133 Here the wettability of the material is affected by surface segregation of the polymers (n-butyl acrylate–styrene ratio). Subsequent growth of silicon nano-wires via chemical vapour deposition afforded hierarchical roughness. 224–240 | 235 . This method is comparatively low cost.8 in SU-8 with a WCA of 81u). has been successfully used to direct write pits and pillar structures into silicon.117 Here a combination of deep reactive ion etching and interference (holographic) lithography was used to produce pillar structures in silicon with feature size of the order of tens of nanometres. aggregating different sized particles together to produce ‘‘raspberries’’. giving higher contact angles of 160u. Heating above the coating polymer’s glass transition temperature or treatment in acidic water was shown to switch the surfaces between superhydrophobic and hydrophilic.119 reported a siliconprocessing technique using a combination of photolithographic patterning and various etches.

replicas of superhydrophobic surfaces such as a lotus leaf have been used with amazing success. Copyright 2007. and although a direct replica of the original structure was not achieved the surfaces were superhydrophobic. Templating A pattern or shape. Copyright 2007.143 The spikes. The attraction between electrostatically charged species is often used to build up multi-layer structures by sequential dipping in positive and negative polyelectrolytes – a method termed layer-by-layer assembly.129 Images reprinted with permission from (a) Institute of Physics. can be replicated using a templating method. Templating surfaces is often fast. this produced a superhydrophobic surface with slightly hydrophilic chemistry. Indeed there is much interest in natural superhydrophobic surfaces. A progression of this work to produce direct replica of various leaves was reported recently by Suh et al. Fig. cast a lotus leaf in PDMS to form a master template then used this to produce a PDMS replica. As with some of the lithographic surfaces. If ordered structures are required with small length scales. the spikes gave an advancing contact angle of 170u compared to 118u for the unstructured surface. leaving the inverse of its pattern.129.26 (b) CNT-coated polystyrene-sphere array..152 Lee et al. later described the use of a template-based extrusion method for the preparation of much smaller structures via templating. very low cost and reproducible and so is a widely used method for the preparation of polymeric surfaces. give a concise overview of the recent developments of polyelectrolyte multilayer fabrication.141 TiO2 particles have also been aggregated around cellulose fibres using PAA to achieve a WCA of 162u after modification with a fluorosilane. which was subsequently used to cast silica features onto a silicon wafer. The advantage of this method is that the particle coating is conformal and readily controlled. Fig. such as colloidal. The resulting hierarchical roughness was then gold-coated and thiol-functionalised to afford a superhydrophobic surface. Often the template is then removed. Zhu et al.150 These surfaces were used to investigate the effects of surface structure on wetting. Recent developments in this field have seen particle assemblies being produced in the shape of larger structures. Any surface can be used as a template. lithography of some sort is often used to realise the master templates.151 The randomly arranged micron-sized surface protrusions as well as the nanoscale architecture created by wax crystalloids on the original lotus leaf were clearly copied. Copyright 2007.128 (c) silica-sphere array with additional smaller sphere aggregates (scale bar = 5 mm) and (d) micron-sphere array produced from 300 nm particles silica nano-spheres (scale bar = 5 mm). Fig. bamboo and lovegrass. some of these masters maybe reused and some may be intentionally destroyed to reveal the replica surface. He et al.140 Surfaces with gradients from superhydrophilicity to superhydrophobicity have been prepared by graded UV exposure of such layer-by-layer deposited structures. with both surfaces exhibiting superhydrophobicity and an advancing WCA of 160u. Fig.26 Shiratori et al. this can be used as a template to achieve a replica of the original.Fig. 4. Fig. pressed or 236 | Soft Matter. produced a master surface via photolithography having features in the micron range. with researchers focusing mainly on leaf structures.129 with several patents on this type of technology being granted. subsequent adhesion of polystyrene or silica spheres occurs only on the bare substrate. 10(d). Similarly an AAO template was used to for perfluoropolyether derivative nano-pillars having superhydrophobic character.153 Here an inverse of the leaf structure was fabricated in a photoreactive polymer via nano-imprint lithography. 11(f). lithographic and woven material surfaces. Sun et al.139 Layerby-layer assembly can also be used to control of the arrangement and aggregation of particles on a surface. which was subsequently coated with silica nano-particles and further silylated to form a superhydrophobic surface.142 Particle aggregation can be used to form surfaces with ordered or non-ordered roughness on scales from nanometres to micrometres. either 2D or 3D. Following a hydrophobic self-assembled monolayer coating. have investigated the feasibility of producing superhydrophobic surfaces by templating the micro/nano surface features of various leaves: tulip tree.147 Porous aluminium oxide has also been used as a nano-imprint template. In 1999.148 An anodic aluminium oxide (AAO) membrane was used as the template to produce a poly(vinyl alcohol) aligned nano-fibre surface. wherein a material is printed. 10(a). 11(a). More recently. using alternating polymer and charged particle layers. A simple layer-by-layer structure using only polymers: polyallylamine hydrochloride (PAH) and poly(acrylic acid) (PAA) have been shown to afford a rough micro-porous structure after acid treatment. pressed into heated polystyrene substrates and removed by dissolution of the aluminium oxide to produce large area polystyrene nanopillars or nanotubes. 224–240 grown against the voids of a template. who cast a PDMS negative and then used this to create a moulded PDMS positive of lotus and This journal is ß The Royal Society of Chemistry 2008 . looking particularly how transition from Cassie–Baxter to Wenzel states occurs. 10(c). described the use of PDMS template surfaces produced from a micro-machined master having fairly large micron-sized features. 10 Particle aggregation (a) layer-by-layer deposition of TiO2 particles on fibres. (b) American Chemical Society. silver maple. 2008.149 Another approach to template-based surfaces is to build the negative master directly. (c) and (d) Elsevier. shallow cavities and stripes of the master were then replicated using an elastomeric mould. Bico et al.

Fig. Fig. waterresistant materials and superhydrophobicity are among the Soft Matter. Fig. Copyright 2007 and 2006.159 the tendency for impacting drops to remove contamination160 and the prevention of conversion between Cassie–Baxter and Wenzel states of wetting161.129 The paper describes two routes.147 Images reprinted with permission from (a) EDP sciences.127 Here nickel was electroformed around a gold-coated leaf to afford a nickel mould. This technique has also been used by Jia et al. the ease that drops roll off surfaces.. which is then used as a mould to produce a crystalline solid of nanosized spheres. wherein gold was electrodeposited between the spheres.156 The template was dissolved leaving an inverse hemispherical array. such as diffusion-controlled electrochemical deposition. One of the first reported uses of sacrificial templates to generate superhydrophobic surfaces was by Whitesides et al. 2008. 11 Replica surfaces produced by templating (a) micro-posts. Hydrophobic fabrics. These can be achieved with single techniques. 10(d). dissolving the spheres to give a macroporous solid.136 combining templating and particle arrays129. (b) and (c) Elsevier. several steps or methods are combined. Copyright 2006. Fig 2.158 More recent theoretical contributions support the earlier suggestions that multiple layers of roughness enhance several aspects of superhydrophobicity. giving rise to: growth of crystals during etching.144 (c) a polyvinylidene fluoride inverse opaline structure.146 and (f) polymer hairs grown through an AAO template with insert showing water droplet resting on surface. although the longevity of the PDMS negatives is questionable. allowing the drops to run off easily.Fig. who used a monolayer of close-packed polystyrene particles as a template and filled the gaps between the particles with silica via a sol–gel route. the first coating an array of micron-sized silica spheres with nano-sized spheres.156 or particle arrays and lithography.163 assembly of colloidal particles with a second layer of roughness on top by adding another layer of smaller particles128 or by crystallisation on the particles.127 (e) water droplet resting on a polymer hot-press transferred pattern. who fabricated sub-micron half-shell features by depositing metals onto a silica-colloidal array. the science has been around for much longer. 224–240 | 237 . Copyright 2007. 11(d).143 (b) PMMA replica of Colocasia-like leaf surface.146 This journal is ß The Royal Society of Chemistry 2008 Multiple scale roughness An issue that has returned to recent discussion is the effect of multiple scales of roughness.146 a combination of templating and etching also provides dual-scale roughness.75 Recently there has been a general trend towards producing surfaces with multiple scale roughness. giving dual scale roughness and a WCA of 160u. with the surface showing superhydrophobicity. Colocasia leaves. Fig. 11(b). Copyright 1999.10 This was investigated previously theoretically157 and experimentally. Fig.145 (d) photoresist replica of lotus leaf by UV-NIL. Experimental evidence suggests that surfaces with sufficient multiple roughnesses can condense water in the Wenzel state and then convert to Cassie–Baxter. Similarly fluoropolymeric films have been templated around a silica-sphere array. respectively. and the second templating an array of micron-sized spheres in PDMS. and (e) and (f) American Chemical Society. A similar result was obtained by sandwiching an aggregated polymer-beaded surface against a crimped stamp. 11(c). 4. A similar approach using a close-packed silica-sphere array as a template was adopted more recently.94 Outlook Superhydrophobicity is a fairly recent term used to describe the extreme water repellency of rough hydrophobic surfaces.154 After dissolution of the template the nanometre-thick metal cups were aggregated on surfaces and treated with alkanethiols to afford superhydrophobic surfaces.144 These methods have shown excellent reproducibility of the original template leaves. typically with features of several micrometres with nanometre-sized roughness too.145 A recent article has shown how a surface with ordered multi-scale roughness can be created by templating a colloidal array. (d) Institute of Physics.. 11(e). A metallised negative replicate of a lotus leaf has been reported that has obvious advantages for mass production.155 Removal of the template led to an hexagonal array of partial spherical pores. ultrahydrophobic surfaces.162 Often. The original biological surfaces were often found to have roughness structures on several scales. however. This method is very useful to prepare dual scale features. however.

D. 2007. Cassie and S. K. Fadeev. 546–551. Ma and R. 2006. 87. 20 (a) F. Feng. such as abrasive wear and washing cycles. Phys. 2006. in Encyclopedia of Surface and Colloid Science. Lee and N. 2006. Germany. D.. van Ooij. Rev. Yeomans. 2004. (j) C. Pogreb. 16 Y. Bormashenko. M. Xin and X. 21(3). Bathlott.. Jiang and D. 2003.most used terms. 17. He. F. This area is well-established with coatings continually being developed and improved. J. L. Onda. Li and A. Superhydrophilic surfaces. Langmuir. Extrand. 33. Lafuma and J. G. Tao. Gao and L. We have suggested standardising terms in the introduction. Li. (f) L. (b) W. 4.. 202(1). R. Li. Mater.g.. Jiang. 19 L. opening new avenues for potential applications. Shiratori. 1071–1078. Eng. 12. (h) A. 7991–7999. H. J. Baxter. 18 D. Adv. Hill. US Pat. Shirtcliffe. Langmuir. J. 213–225. Langmuir. Langmuir. Wagner. L. Zhai. 1–8. Langmuir. 1999. Nakajima. Mater. 2006.. Barthlott. Soc. J. 592–601.g. C. 1996. Communication and radar antennae and receivers are sometimes coated with superhydrophobic paint to reduce water/iceinduced distortion (e. 1945... Gleason and R. Fridrikh and G.. L. Langmuir. Bico. Eng. J. Some applications do not involve abrasion or contact with oils that may cause contamination. Michielsen and H. 2002. Feng. 36. Song. Pyatt. 4 T. 14 G. Zhu. H. 2000. C. 1. Nat. J. 2 386 259. 2007. 2004. 12 N.. 988–994. Cohen. or even a hydrophobic one. Chen. Zhang. Xie. Newton. Tsujii. are also being used as coatings for roof tiles (e. P. Qi and J. Nanotechnology. pp. 1999. van Ooij. Neinhuis and W. Amirfazli. 8. 1996. 132(2). Quere. 18. 16. 23(11). 5549–5554. 14. N. Raindrops rolling off a superhydrophobic glass surface. McCarthy. 2007. 2nd edn. 2002. 18. 2 613 850. Trans. can be more distracting than if the surface is hydrophilic. Lowery. J. The main barrier preventing self-cleaning surfaces being used in many areas is that the surfaces are easily damaged or contaminated during normal use. Xin. 6019–6032. 291. 15. M.. Adv.2. Colloid Interface Sci. Z. L. Langmuir. V. (g) E. Sone and S. Reinhoudt and M.. I. Langmuir. (d) L. Langmuir. L. 14(10). Res. 2006. 515. Japan).. 1849–1851. Wang. C. Zhang. Liu. McCarthy. Neinhuis. (i) D. 2006. H. Chem. M. 3517–3519. 1193–1197. Adv. Despite this over 200 patents have been granted that use or produce superhydrophobicity in some form and products are becoming commercially available. 1384–1389. 2007. 28.. B. Ma. which use the plastron effect of superhydrophobic surfaces. 1999. Daoud. Y. 2004. 1473–1481. 2007. 39 958 441. X.. 2. (c) A. 1440–1443. Korea). A. M. H. an apparently successful product. 6234–6237. Zhao and W. 193–202. 13 (a) US Trademark. Quere. Although the dew point is reduced on superhydrophobic surfaces condensation still occurs if the temperature difference and humidity are great enough. Shirtcliffe. 644–652. J. 1350–1368. Horst and M. Colloid Interface Sci. S. Commun. ed. Hill. L. Planta. 11. 11. 2007. C. 101–104. Lu. G. Daoud. Opin. Appl. particularly as many 238 | Soft Matter. STO corp. 8 (a) N. J. K. This journal is ß The Royal Society of Chemistry 2008 . A. New York. Zhu. (d) C. B. Y. (f) X. 937–943. Extrand. 5 X. 31–41. 5969–5973. Am. G. F. 6 W. Li. particularly when describing hysteresis character. N. Mater. Lee. I. Feng and L. 15 S.165 A recent paper by Gao and McCarthy highlights this by showing that a technique designed in 1945 produces superhydrophobic fabrics. F. 2006. Small. Sci. Perry and P. 14. A recent report highlights the possibilities for superhydrophobic surfaces to reduce bio-adhesion. A. H. or will be. J. Acta Zool. B. Blossey. Soc. 2125–2127. Langmuir. Roach. For this reason most coatings for windows and mirrors are superhydrophilic. Micromech. 23(12). applied to LCD screens and CDs to protect them. S. Patankar. 23 M. 2495–2532. Jiang. J. 9. Gupta. 1997. Plant Physiol. Youngblood and T. G. Zhang. Shibuichi. Taylor & Francis. (c) M. 6. I. 18(21). 2003. C. Early products include breathable fabrics and some membrane filters. L. Ann. 2 R. Li.166 One would imagine that superhydrophobic surfaces would be useful on glass surfaces to keep them clear of condensation and water drops. Acc.. Z. Bot. 3. Somasundaran. 2. 24 S. Dongbei Daxue Xuebao. Watanabe. 10 C.. 255–259. The number and scope of techniques to generate superhydrophobic surfaces has greatly increased in recent years with suggested applications expanding with them. McHale. 224–240 of the drawbacks are now well-understood and some can be minimised by choosing specific coatings. B. E. C. 23. Sun. S. Y. 17 T. Marmur. (b) German Trademark. 7 D. Xin. J. Kong. 21. J. C. Perry. Newton and G. H. McHale. Japan Hydrotech Coatings Co. 40. Microeng. J. 1997. Y. 301–306. 667–677. A. Crego-Calama. Hashimoto and T. and have made use of these super-water-repellent surfaces. 2005. Newton and C. Chem.g. although it should not be confused with those displaying superhydrophobicity. L. Neinhuis and W. 5854–5868. 2007. 19. 2004. 3395–3399. although the exact meaning of each is rather vague. Ceram. B. Shao. which is a problem in a huge range of areas. 2005. 2002. Li. 835–837. Li. (b) N. 79. Kusumaatmaja and J. J. Mater. Parker. Aqil. Barthlott and C. Datta. H. Radomyselskiy. 6501–6503. 165607. J. Ogawa. Langmuir. (USA) is a manufacturer of high-quality building materials which has been marketing an exterior paint since 1999 under the trade name Lotusan1. 9 (a) H. Mak.164 but these are rarely referred to as being superhydrophobic as their use predates the term (e. S. 2003. P. Callies and D. B. 88. M. Gao and T. Erlich. Mater. which is not easily visible when viewed perpendicularly (e. 25 M. 1109–1112. Thin Solid Films. 248(1–3). 21 W. France. Satoh and K. Japan). Chem. McHale. Y. 55–61. Rutledge. 3063–3078. Zhang and C. R. Mater. 20(9). 22. X. Liu and H. M. Chem.19 the term superhydrophobicity arose around 1991. J. Curr. 2005.. C. Macromol. 2007. Zhai. H. methods of production and materials used to form superhydrophobic surfaces broaden. Ind. (b) W. Bognitzki. being produced by TDK and Sony (Japan). M. A. Colloids Surf... Y. causing condensation and drops of water to form a film. J. Wenzel. J.. H. S.g. Chem. Hu and S. Adv. Prog. 22(14). M. products by Erlus AG. D. 2008.g. 2005. 26 T. 2005. products by Ventree Co. Dong. 11.. J. 163. L. Hirec by NTT-AT corp. 51–68. 68. 77. S. Faraday Soc. S. Zhang. Wang. R. Quere. 14(10). 1857–1860. As coatings improve and prices fall the use of superhydrophobic surfaces is likely to increase. 28(4). (g) M. A.. K. Oner. Ma. As research continues. Liu. J. L. Ziran Kexueban. Whyman and M. Hsieh. J. 1782–1784. 6004–6010. Rep. J. Ding. Y. Monatsh. McCarthy. 38. Y. particularly those with photo-oxidative properties. 2006. Agarwal. Shirtcliffe. Q. Chem. (e) T. Gore-Tex patented in 19721). 1944. Nanotechnology. 22–27. Zhang and W. Evans. Y.. 11 T.. Other general coatings are available from Degussa GmbH (Germany) and some further recent products are more resistant to scratching and are. Gao and T. 2007. References 1 (a) R. and by Deutsche Steinzeug Cremer & Breuer AG. 2001.. Germany) and in paints (e. Soft Matter. 1936. (e) B. K. (b) J. Polym. 3 A. 22 W. 132.

G. 2005. H. C. S. Jiang. V. 5626–5631. B. Zhou. Weng. 10624–10627. T. 2006. G. Chem. 2004. Chem. Europhys. 71 (a) I. 18. 30 (a) G. Takano. Feng. A. T. M. 43 M. H. Ivannikov. N. H. 35 (a) M. Ren. R. 71. Rossi. J. X. C. American Chemical Society. Yamamoto. Palgrave. N. J. USA. Zhang and Y. J. J. Technol. F. Jeong. Perry and P. Onda. Mazur. 21(20). 182–188. Chem. 2005. Nishimoto. G. E: Stat. Mendez. C. H. 15. Proceedings of the 206th Meeting of the Electrochemical Society. L. Takano. Jiang and J. Fujishima. Zhao. 2000. X. S. T. C. Colpo and F. 2004. S. Liu and L. Philadelphia. Y. K. O. 2008. Perry. Macromol. Mater. Q. Kulkarni.. 68 W. 103. 5115–5118. Soc. M. M. W. Nonlinear. C.-Y. Zhang. Langmuir. Jung. 75 C. Palmese and R. Tee. J. Song. Polym. L. 173108. 58 S.-H. 74 (a) C. Soc. M. Li and L. 2004. 90 X. Golverdingen and J. Wang. Zheng. 7095–7102. H. 26. Phys. Fujishima. 22. 2001. H. Nature. Wang. K. A. Pap. Chem. K. Reneker. 33 K. Honma and H. Chem. Wessling. Phys. W. J. H.. Chem. M. 40. Chem. B. ECS. 2006. 46 H. Ashraf. Amirfazli. Int. 5. Avci and O. 47. 109. 15. J. Shibuichi.. M. Zachariades. 37 M. Steely. L. 77 M.27 J. Chem. H. L. Raghunathan and B. Kormunda. J. United States. Rutledge.. Narita. H-P. Zhao. F. L. A. Polym. Feng. Douglas. X. 271–275. A. 228th ACS National Meeting. 1011–1012. Newton and C. (Am. 47 N. 37(9). H. Chen. N. 1(3). 28(3). K. 23. U. 3575–3582. 125304. USA. Washington. Wassink and A. Balaur. Kamperman and Y. 1991. (b) S. 56 A. Appl. 1964–1965. C. Neumann. Dzenis and D. Macromol. Phys. 62–63. 25. Langmuir. Mayama and K. T. 50(11). L. 29 Y. Krikorian. J. 17. Lacroix. 3089–3092. Roach. Demirel. Chem. Vargel and E. Polymer News. (b) X. H. Han. O. Onda and K. 2006. Z. J.. 53 E. Angew. B. Ceriotti. 3125–3130. 84 H. K. M. M. 2006. D. POLY–504. Prawer. H. Raghunathun and B. Zhang. Feng and L. G. E. Srinivasan. 17. 127(39). J. L. Polym. 431–432. J. G. Fang. 9. 6(3). Rapid Commun. 19(25). Y. 19(14). H. McHale. Langmuir. J. Schakenraad. Watanabe. Soc. 55 J. Chem. Lett. Jiang. Y. J. Mater. 86 K. Angew. 36(2). Chikuni. 54 F. 40 N. L. Robinette. Zhang. 2006. Gogolides. Zhou and E. Langmuir. Zhou. Hashimoto and T. Y Menceloglu. 192–197. Zuo. Jiang and D.. 2969. 44 A. A. Zhu. 2005. 1365–1368. Bocquet. J. Langmuir. 59 R. 2004. Rapid Commun. Acatay. Xu and K. H. 5151. H. Yang and Y. A.. Langmuir. American Chemical Society. This journal is ß The Royal Society of Chemistry 2008 Soft Matter. Zheng and D. 2006. Ishizaki. Res. H. Y. Miller. D. 1995. 594–597. J. 73 K.. Jin. 50 L. 206–207. 1995. E. Funct. Leong. (Am. 506–509. Q. 2006. 1998. Y. H. South Pacific. X. D. Feng. Phys. Newton. J.. S. Langmuir. Han. Phys. 2003. Vogelaar. Fujihara. Song. Qian and Z. 2005. 85 K. J. Nano Lett. Thin Solid Films. Bico. DC. McKinley and K. P. 592(1–3). Y. Sun. Inoue. T. G. Kiyotani. 43(33). 21.-F. 38 Y. 46(2). Liang. Ma. 2005. Lan. 70 X. M. E. Mater. Bekke and O. 82 M. T. 2007. 4917–4919. P. Park.. 4488–4491. I.: Condens. Chem. Matter. S. Coat. Commun. J. 140–143. Chem. Ed. S. Proc. Zhai. Zimmermann.. H. H. Steely and H. Wang and X. Cao. H. Yu and Z. J. Lau. Allcock. 2006. Adv. B. C. L. Prepr. 657–65. 19512–19517. Stokroos. 347–356. G. Soc. 2006. Chem. W. J. 811–21. (b) S. X. Kojima. Shimohigoshi and T. Ma. Yuan. Langmuir. 69 G. H. Schmuki. Shibuichi. 2007. Ed. S. Inoue. J. S. Div. 32 H. Milne.. Soc. Onda and S. J. L. M. X. 5556–5559. Langmuir. Xiong.. Song and D. 44.. Langmuir. R. Surf. Menceloglu. T. 3. 2000. Yang. 46(2). Adv. C.. Hou and Y. Chen. Adv.. Int. Kalra. Pap.. Cai. R. K. L. 62 J. Biomater. Chem. Wang. 12. J. 2007. 2005. Li. Yu. Abs. Inc... 298–305. Feng.. V. Shi. Mater. Phys. 2005. Piskin. X. Zhang.. Chem. Jin. Polymer. 2003. 524–527. J. 7746–7748. Irie and K. Chem. 2005. Yu. (b) V.. Engl. Feng and H. Parkin and R. Giuseppe. Wang. Rev. D. Huang. M. 388. 34 A. 61 H. Technol. 36 Y. ChemPhysChem. Technol. B. 224–240 | 239 . J.. Stokroos. 36. R. Busscher. 25. Mohammadi. M. Jang. S. 2005. Y. M.. Mater. C. T. Xu and C. 21. G.. I. Y. 1689–1695. Jiang. S. 76 H. 43(1). 22. 287–294. Zhang and X. Z. 16. K. Li. Langmuir. 2003. 2300–2303. Zhai and L. A. K. 2005. Mater. 22(7). Purcell and L.. Singh and L. Xu. Artus. Thomas and G. Yu and S. Porter. A. Murakami and A. 23. Lu. K. 2006. Wu. Jin. 100. Yamamoto. Rapid Commun. 65 Y. Zhang and L. Miyata. 2004. Tsuchiya and P. 39. Takai and A. Macak. Soft Matter. N. L. Han. H. Martys and J. 1997. Chem. H. L. Marquardt and S. 2005. 72 X. M. 63. T. 9477–9479. P. Reneker and H. 91 M. C. 67 (a) R. Zhai. Lammertink and M. Am. J. 2007. 303. W. 9007–9009. J. (b) L. Macromol. San Diego. Sugimura and O. CA. 81 L. Tserepi and E. S. R. 2005. Han. Q. L. 1606–1610. Polymer. Mert. Chen. 13458–13459. J. 1701–1705. Vandermei. Lee. Reneker. 18. Ed. 2003. 7. X. M. 5441–5446. Saito. L. B. J. Cleva and Z. Feng. Chem. Qu. Ziaie. 2. 2007. J. K. T. 195–201. Wang. Nanotechnology. Rao and M... 2004. L. 17. E. 2007. 126. A. Li.. J. 111(3). Nakano. S. 45 (a) A. L. Y.. G. Hyomen Gijutsu. Tavana. T. Hanabusa and K. Journet. 42 J. S. 15. 17. Mater. Tsujii. Tsai. P. Hao and W. F. 1105–1108. Y. Chem. T. 031205. Sada. Nano Lett. C. 2665–2667. 2248–2250. M. 2005. J. J. J. Z.. 51 P. 3. Mater. Yamanaka. 2005. D: Appl. Acatay. Nanotechnology. Science. Yang. D. A. Shen. F. Shirtcliffe. Hosono.. Fundamental Science and Technology of Photofunctional Interfaces.. S. Kitamura. Sci. Zhai. Koczkur and B. 63 (a) K.. Cheng and C. 2300–2303. H. Polym. Ding and S. Shiratori. J.. Mater.. 2006.. McHale. Y.. Rouxhet and J. Feng. Y. Tang. C. J. Busscher. 41 X. T. 2006. X. M. T. 48 Z. Erbil. Sol–Gel Sci.. Mater. He. 9657–9662. 376(1–2). C. J. S. 28 N. Hashimoto. A. F. 88 M. 2006. Amaratunga. H. J. Abe... Xie. Macromol. Blackman. 60 W. Macromolecules.). 41. 5453. Watanabe. Z. 2007. 66 S. R. Moulinet. J. Washington. S. 1977–1981. Y. Compos. Adhes. Int. 1. Z. 16. Ban. Q. 20(7). 2758–2762. Y. Commun. 564–571. I. Chen. Han. Niu. Wang. Tsujii. 201. 2007. 2006. C. Tech. 4. Y. Zhu. Akada. 92 M. 31 E. D. 6662–6665. 2007. I. Cho. Tsujii. Peng. Soft Matter Phys. Teshima. 243–249. Saito. Chen. Zhang. 2753–2755. Zhu. 52 H. Ed. Hyang. 2004. L. Chen. Phys.. E. Zhai. Tung. Zhao and Z. Kasuga and A. Shi. 80 T. W. 1994. Gleason. I. (b) A. Chhowalla. Sugimura. 2005. D. Feng and L. R. Ziaie. Takai. M. G. Kuo. Y. J. Abs. J. J. 1997. 90(17). 593–596. 1537–1555. Gautschi. Commun. Nakajima.. Singh. G. Div.. P.. 2006. Colloid Interface Sci. Parkin. J. I. Prepr. Am. Chen. 1667–1677. Zhu. Shirtcliffe. Jiang. L. M. 2006. Huang. H. F. Vourdas. 78 E. K. Liu. Teshima. Phys. H. 4338–4341. Ybert.. 1996. 2005. 104–9. J. J. 208. Mater. Chen and G. 603. H. Rapid Commun. C. I. W. J. S. Carey. 56(9). J. Y. Eren. 599–600. 2005.. Miyauchi. Hashimoto. P. J. J. 64 X. G. 1992. 87 N. Meziani. 21(15). L. 1075–1080. 22(23). M. G.. S. Keikinzoku. Shibata. Allcock. Cells Mater. 79 Z. 2005. 299(5611). Amirfazli and A. J. Kazim. Polym. H. 2007. 20. Baldacchini. Lett. 229th ACS National Meeting.. Kakad. Joo. 17. 3793–3797. Chen. Guo. K.. Am. Angew. Shibuchi. 83 H. Teo. Chen. Int. C. M.). Seegar. K. PMSE–536. Li and Y. Xia. Sci. 12. (b) N. Liu. Phys. DC. T. 2002. Sci. Langmuir. Colloid Interface Sci. Zheng. Bolgen. 6. W. 53–57. 2004. Lejeune. Darrell. 112–117. 1. 18. P. 57 T. Takai Osamu. S. Hyett and I. Polym. Surf. Sci. A. 49 L. 2005. Doshi. 336–341. Bull. J. 2006. B. 39 N. Schakenraad. Chem. 7138–7142. Srinivasan and D. P. 2006. Guo. C. Wu. Satoh and K. 2007. 89 B. Chem. L. 1377–1380. Inoue. Tsujii. Conf. 47.

Bico. D. T. Cao. Barthlott and C. 2003. 1359–1362. 165 W. Feng. Herminghaus. Liu. Y. Ed. 130. Wang. 2091–2094. 95 A. 101 R. J. Larmour. R.. Wang. 7. 105 J. 130–131.. T. Macromolecules. 97 S. L. Jain and S. Patankar and J. F. P. Seunarine. 201–207. B. Sun. Thin Solid Films. 152 (a) P. 2007.. Nature. Jia. S. S. 2007. Yu and Y. C. Surf. Jiang. S. Marzolin and D. Wen.. Liu. V. 16(45). Hozumi and O. L. G. Zhu. Bewilogua. Eng. 4686–4690. F. 2004. 341–345. 2000. 27(3). 1929–1932. Int. 6009–6012. Lee. Yang. C. 2006. I. 145 J. K. Wagner. Y. B. 137 M. 12. 118 R. Sheparovych.. B. M. Walzel. Shirtcliffe. 2007. Colloids Surf. Heo. 46(10). Q. L. 22. Xue and Y. Neinhuis. 1295–1303. X. 142 T. A. Love. M. Hashimoto. C. S. 9676–9681. H. C. Cohen and M. Pozzato. N. Sun and J. Nanotechnol. Phys. Sci. Chapel and F. 5462–5466. Coat. T. I. 156 Y. J. Ding. Neinhuis and P. J. Q. Yan. Z. 415–419. 52. Suzuki. S. 2007. Sens. Europhys. Chem. 154 J. 2005. H. F. J. Chem. Coat. Sci. 2007. Liu and J. He and F. Li. 17. 89(10). Feng. B. L. Phys. 2006. 3157–3161. H. Sci. 98 N. 240 | Soft Matter. Surf. Wilson.Shirtcliffe. Zhang. Song and L. 253. A. Chen. C. 2002. Saito. W. Kim and T. 595–600. Langmuir. Song. Li. 2006. Han. Choi and C. 132 L. Int. 100 J. Langmuir. Tsuruta. MacWilliams and S. 54. H. 113 S. Lee. P. J. Jiang. Kim. Z. Technol. J. C. H. J. N. Phys. Langmuir. 102 M. 188–192. 121 C. 875–879. Karim. Y. 7665–7669. 22(5). 2005. K. J. A. 2005. 4. 222–225. Hirotsu and H. V. 2006. 4. 2007. P. 140 (a) J. 2003. Nanotechnology. Cong. 1121–1128. Langmuir. Q. Gates. Yu. 2007. Jonsmann. Surf. Y. 22. D. Liu. M. 2006. Ball. M. Gu. Cai. Kang and H. Mater. Oh. M. D. 22(7). Y. Li. Zhang. Ingmar. Zhang. 189–193. 2007. Lee. 238–243. Y. Grundmeier. 33L. Furstner. Wang. J. Tsai. Minko. 149 W. 431–434. Mater. P. 9982–9985. Chem. Langmuir. 131 J. Sone and S. 1753–1757. C. K. L. 5(4). 104106. 115 Z. Technol. 2006. C. Macromol. 2005. K. Lupitskyy. X. K. Sun. K. Langmuir. Bormashenko and R. N. Rapid Commun. A. 111(1). Mater. O. M. W. 47. E. E. T. 17. Z. J. 2005. 2966–2967. (b) L. Gao and T. Adv. 2006. 6800–6806. Ed. T. 108 Y. Song and L. 141 J. Chen. 309. Natali. J. 11(6). 26. A. W. Whitesides. Luo. J. X. Y. Q. Opin. 23(6). 253. J. 17. 2006. Barthlott. McCarthy. Paul and G. Qi. C. Langmuir. 2007. Wu. Jiang. Langmuir. H. Nanotechnology. Faupel. 23(13). P.. 561–564. Kim. Ye and Z. Shi. Y. 225–229. K. Chen and C. Riehle. 1994. 116 L. Langmuir. Z. 3432–3439. Cho. Chen.. 2169–2174. 884–888. Ming. ¨ Bot. D. Liu. 21(5). H. Schurmann. Surf. 12213–12217. X. Nano Lett. K. R. Gao. 20. C. Zaporojtchenko. (b) X. Angew. L. G. Langmuir. Reuben and J. Sens. Yoon. Ge and Z. Deposition. Belotti. 5296–5305. Zhu. F. s. M. F. H. Funct. 165. 8502–8505. Jiang. Hong and H. Oner and T. He. Y. Y. Stratmann and G. Bhushan and Y. 201. Xu. Dorrer and J. S. J. 111 G. Hu. H. Phys. 219–225. Jpn. L. 2006. 4999–5003. B. Langmuir. 44(37). Greve. U. Nagata. 2007. Perry. L. Xiu. H. 2006. S. Coat. C. Gadegaard. Langmuir. R. Abdelsalam. Z. 3710–3714. Ruhe. J. Wagterveld. Chem. Langmuir. M. Zhu. C. J. Adv. Ed. J. 687–692. J. Efimenko. Cao. 1999. Shen. B. McHale. J. Y. K. 9143–9148. McHale. Xu. Microelectron. A. 24. Perry and F. 17. Mater. Feng. Sun. J. 1999.. 94 T. Y. 2007. 199–203. J. Wu. Chem. 1673. Zhou. 2008. Chem.. 150 B. 112 S. Gu and H. 015103. 144 R. L Cui. 138 G. Zhang. 23(4). E. Huang. 1997. Zhai. 166 J. Liang and H. Bravo. Phys. McCarthy. Jpn. W. Yang. Zhou. 93. 4. Eng. Shiu. L. S.. Li.. 7777–7782. Nakajima. 9954–9957.. M. Z. Nema. Kim. S. J. 5. 45(10B). Zhukov. L. Martines. Xu. 83(4–9). Xu and X. D. E. Jeong and K. S. 19.-Y. 22. 2007. Shirtcliffe. C. H. T. 159 L. Y. Sci. Cao. 8. M. B. 2007. Li. Y. T. 84. Chem. 2006. 220–226. 252(6). Wu. Xiao. S. Fois. W.. Newton. F. B. Novoselov and A. C. Sci. 136 Y. Chem. Bartlett. Kim and A. Nae. I. B.. 18(1). Europhys. 2005. Combellas and Y. Thin Solid Films. 2006.. B. Microeng. Brady and R.. Chem. He. Wolfe. Zhai.. Bell and G. 151 H. Guo. S. 124 E. Langmuir. Jiang and D. 324–329.. Int. W. Wong and C. 2006. 106 S. M. Mater. Chen. 2007. 11... J. Wen. 385. Micromech. W. Little. Y.. C. 232–237. Cheng. Y. 155 M. 1986–1990. G. Colloid Interface Sci. Berendsen. Adelung and F. Appl. Milenkovic. Surf. W. Y. C. B. Newton. 16(21). Y. Lett. 125 R. P. J. P. Kelf and J. Jin. Saunders. 22(23). S. Y. Huang. A. Phys. 103 Y. Mulchandani and Y. 2097. W. Cho. E. P. Rubner. 2006. Li. P. Zhou. Appl. 135 (a) W. Morimoto. P. 2005. 2003. 948–953. J. Whyman. Y. O. Mistura. 2005. X. Zhang. 162 S. 2007. Li and L. Dal Zilio. 2006. H. 2229–2234. Vendramin. 107(37). 4310–4314. K. D. Zhu. Uhm. Stein. 23. G. E. 1984. R. Choi. M. US Pat. Ji. Langmuir. 110 Y. L. H. K. J. 2002. G. Takai. 5326–5333. H. 2. Hess and C. Z. A. W. X. 2006. N. 128 Y. Ogawa. 27. 2007.. Y. Quere. Jiang and D. Mater. R. Youngblood and T.. 19.. Appl. B. Song. H. J. 21. 2. C. Y. K. Samulski. Song and L. 2006. L. H. W. G. Mou. Yoo and J.. Fu. Appl. 96 A. 1972. Nano Lett. Chem. Watanabe and K. 133 J. Takai. J. G. Bouaidat and J. H. 17(5). 5660–5665. D. 2298–2301. Gore. 122 D. Phys. J. 2004.. 2004.. 224–240 This journal is ß The Royal Society of Chemistry 2008 . 1997. B. J. E. X. Xu. 107 I. Curr. Chen and W.. Langmuir. Rodak. Cai and S. W. Langmuir. 21. Mohwald. Eng. Bai. R. S. 4427–4434. W. 2003. 481–488. Nosonovsky. Fang. Mater. Han. P. 10904–10908. C. Kim. 3 664 915. 2001. B. Langmuir. Exp. Rubner. Duan.. Jiang and D. V. Surf. Xu. 891–894.. Angew. Langmuir. 2007. 130 P. 129 C. A. 164 N. 800–802. H. Angew. L. 2608–2614. D. Chen and M. Hwan and K. 20. Ji. Jin. Genzer and K. 163 J. L. C. R. A. 8576–8580. E. Lee. J. Wong. Baumberg. 5818–5822. Newton. L. M. Lett. 147 L. 2007. 314(1). 2000. Appl.. 2007. 272(1–2). E. C. L. DeSimone and E. Kwon. Furstner. 553–559. 127 S. 17(3). Bonnar. F. Y. 2005. 18. Yoshimitsu. H. D. J. Shiratori. 142. 7293–7298. 461–463. Bormashenko. Langmuir. Hayden. Biofouling. R. Z.. 600(18). Pyatt. (b) M. C. 22(13). Yang and Y. 21. H. 148 L. Herbertson. Hu and D.. G. 1349. T. Hassel. 2005. P. 2006. J. 42. 2006. W. X. D. A. I. A.. 310(2). B. Nanotechnology. L. Vap.. Chem. Wang. Satyaprasad. (b) B. 139 L. Jeong. 2006. X. F. 1283–1285. Lee. H. 126 A. Zhai. Cheng. Inoue and O. Colloid Interface Sci. 22. Poncin-Epaillard.. Langmuir. 123 S. 109 L. Lee and T.. H. 104 L. N. 515. T. J. J. Tang. 114 D.-R. McHale and M. A. Colloid Interface Sci. Kawai and H. Actuators. 1710–1712. Zhai. X. Pogreb. C. Nanotechnology. Evans. Mater. 160 Y. L. X. Lee. Y. Technol. Singh. C. Cebeci. 2004. Uhm. Wilkinson and M. 2007. Microelectron. Y. H. Lett. 143 J. J. McCarthy. G. 15. M. Cohen and M. Sun M. 153 S. Jung. Kim. J. de With. Suh. E. 16. Grigorieva. 32(20). 120 A. 63–67. Shirtcliffe. 146 E. Sun. 7652–7657. Lee. 3(4)..-M. 16. Actuators. C. 2007. Appl. 117 C. Nano Lett. 2006. 251–254. C. Wu. 2006. V. M. Surf. 21(5). Kim. Y.. L.. Morgan. K. 17(2). 23(4). Y. Nat. 157 S. Burnside. IEEE Trans. Lett. 2758–2772. Mater. Shi. C. 303(1–2). S. 99 (a) A. 2006. Kuo. 21(20). 2006. Zhu. 436. Y. 134 J. D. B. O. Karuppuchamy and J. F. 2003. 5(11). 158 N. S. Chabrol and C. Zhang. Langmuir. Zhu and Y. Sulieman. Fresnais. Langmuir. J. 119 L. Geim. Huang. 21(26). Chem. Y. Thiemann. 956. 200(18–19).. Nano Lett. Nanotechnology. 2006. X. Kwon. Q. 8978–8981. 339–360. 23(5).. S. Y. Zhu.93 A. Motornov. Wang. van Benthem and G. I. 2006. Dubonos. 2(8).. R. 161 M.