Journal of Colloid and Interface Science 456 (2015) 210–218

Contents lists available at ScienceDirect

Journal of Colloid and Interface Science

journal homepage: www.elsevier .com/locate / jc is

Mechanically durable, superomniphobic coatings prepared by layer-by-layer technique for self-cleaning and anti-smudge

http://dx.doi.org/10.1016/j.jcis.2015.06.0300021-9797/� 2015 Elsevier Inc. All rights reserved.

⇑ Corresponding author.E-mail address: bhushan.2@osu.edu (B. Bhushan).

Philip S. Brown, Bharat Bhushan ⇑Nanoprobe Laboratory for Bio- & Nanotechnology and Biomimetics (NLBB), The Ohio State University, 201 W. 19th Avenue, Columbus, OH 43210-1142, USA

g r a p h i c a l a b s t r a c t

a r t i c l e i n f o

Article history:Received 20 May 2015Revised 16 June 2015Accepted 17 June 2015Available online 24 June 2015

Keywords:SuperomniphobicSuperoleophobicSuperhydrophobicWear-resistantAnti-smudgeFluorosilaneNanoparticles

a b s t r a c t

Superomniphobic surfaces are of interest for anti-fouling, self-cleaning, anti-smudge and low-drag appli-cations. Many bioinspired surfaces developed previously are of limited use due to a lack of mechanicaldurability. From a previously developed technique, an adapted layer-by-layer approach involvingcharged species with electrostatic interactions between layers is combined with an uncharged fluorosi-lane layer to result in a durable, superomniphobic coating. This technique can provide the flexibilityneeded to improve adhesion to the substrate with the addition of a low surface tension coating at theair interface. In this work, polyelectrolyte binder, SiO2 nanoparticles, and fluorosilane layers are depos-ited, providing the combination of surface roughness and low surface tension to result in a superomni-phobic coating with droplets of liquids with surface tensions from 72 to 21 mN m�1 displaying contactangles exceeding 155� with low tilt angles. The durability of these coatings was examined through theuse of micro- and macrowear experiments. These coatings currently display levels of transparencyacceptable for automotive applications. Fabrication via this novel combination of techniques results indurable, superomniphobic coatings displaying improved performance compared to existing work whereeither the durability or the repellency is compromised.

� 2015 Elsevier Inc. All rights reserved.

1. Introduction

Bioinspired, superomniphobic coatings (water and oil contactangles greater than 150�, contact angle hysteresis or tilt angle lessthan 10�) have the surface properties that can make them ideal for

Fig. 1. Schematic of layer-by-layer composite coating. Each layer is depositedseparately. Also shown are the chemical composition and charge of each layer. Thefluorosilane (FL) condenses onto the layer-by-layer stack via absorbed water.

P.S. Brown, B. Bhushan / Journal of Colloid and Interface Science 456 (2015) 210–218 211

anti-fouling, self-cleaning, anti-smudge and low-drag applications[1–9]. Such coatings could find use in the building, transportation,and electronics industries [6].

A droplet of liquid in contact with a flat, solid surface does so atan angle that is characteristic of both the liquid and the solidaccording to Young’s equation [10]

cos h ¼ csv � csl

clv; ð1Þ

where csv, csl, and clv are the solid–vapor, solid–liquid, and liquid–vapor surface tensions respectively, and h is the contact angle ofthe droplet. By introducing surface roughness, the solid–liquidinteractions can be enhanced. Increasing the surface area in contactwith the droplet leads to amplification of the solid–liquid interac-tion. As such, repellent surfaces become more repellent or liquidsspread further on non-repellent surfaces. This assumes that the sur-face is fully wet by the liquid (continuous solid–liquid interface)and is known as the Wenzel regime [11]. Alternatively, air pocketscan form, becoming trapped between the surface and the liquid andresulting in a composite interface known as the Cassie–Baxterregime [12]. Since the droplet rests partially on pockets of air, thesurface typically becomes more repellent than the correspondingflat surface. It is therefore possible for non-repellent surfaces tobecome repellent (or even super-repellent) through an increase inroughness and the formation of a composite air/solid interface.

Existing examples of superomniphobic surfaces usually involvea fluorinated component providing the low surface tension [13]and a roughness component enhancing the liquid–solid interac-tions. For example, rough oxidized aluminium surfaces immersedin fluorinated monoalkylphosphates results in oil and water con-tact angles of 150� [14]. A fluorinated polyhedral oligomericsilsesquioxane has been used in numerous deposition techniques– including electrospinning and coating re-entrant structures – tocreate a superoleophobic coating [2,15]. However, few superomni-phobic surfaces with good mechanical durability have beendemonstrated [9]. Spray coating copper perfluorooctanoic acid cre-ates a rough, superoleophobic surface, however it is found to besusceptible to scratching [16]. On the other hand, chemical vapordeposition of a fluorosilane on an SiO2 aerogel results in a durablecoating, however the oleophobicity is poor with droplets of min-eral oil exhibiting high contact angle hysteresis [17]. Poor durabil-ity can be the result of using a ‘‘one-pot’’ technique, whereingredients are mixed together to form a single coating solution.This results in the low surface tension material, required only atthe air interface for oleophobicity, being distributed throughoutthe coating. This can compromise the adhesion of the coating tothe substrate and result in poor durability.

A common deposition method where components are kept sep-arate and deposited individually is known as the ‘‘layer-by-layer’’technique, which typically utilizes interactions between chargedcomponents. Layers of oppositely charged species (polymers,particles, surfactants, etc.) are deposited one after another to createa multi-layer coating bound together through electrostaticinteractions. Because different charged components can be usedwithin each layer, the technique is highly flexible and can be ofrelevance to a wide variety of applications. For example, thetechnique has been used for the creation of a layered microreactor[18], ultrathin films of conducting polymers [19], superhydrophilicsurfaces [20,21], and superoleophobic/superhydrophilic coatingsfor oil–water separation applications [22]. Due to the use ofwater-soluble polyelectrolytes, layer-by-layer coatings are typi-cally hydrophilic and oleophilic. However, deposition of a final flu-orinated layer on top of the layer-by-layer stack would increase therepellency of the coating and ensure that the low surface tensionmaterial would be present solely at the air interface and so notcompromise the adhesion of the coating to the substrate.

In this paper, the layer-by-layer technique is combined withvapor deposition of a fluorosilane to create a novel fabricationmethod resulting in a durable, super-repellent coating. This coat-ing repels both water (superhydrophobic) and oil (superoleopho-bic) and is therefore referred to as superomniphobic, unlike inprevious work where the coating repelled oil but not water (super-hydrophilic) [22]. The superomniphobic nature of this coatingmeans that it is suitable for a wider variety of applications suchas anti-fouling (where superomniphobicity and nanostructuringare of importance, especially for anti-biofouling), self-cleaning,anti-smudge and low-drag applications. The transparency of thecoatings is also assessed for the first time to determine their suit-ability in certain scenarios, including the automotive industry.

2. Experimental details

The coating described in this paper comprises four layers,deposited separately, each of which aids the creation of a mechan-ically durable, superomniphobic coating. The technique builds onthe approach previously reported for superoleophobic/superhydrophilic coatings [22], which utilized charged layers in alayer-by-layer technique. Through the addition of an unchargedlayer on top of the layer-by-layer stack (in place of the charged flu-orosurfactant [22]), it is possible to achieve superomniphobicity.Fig. 1 shows the final coating composition and the chemical struc-tures of the individual components of this novel combination oftechniques. PDDA was chosen as the polymer base layer as it hasa high cationic charge density and has been shown to bind stronglyto glass substrates [21,23] and SiO2 nanoparticles. The specificmolecular weight range (100,000–200,000) was chosen to balancemechanical properties and ease of deposition (viscosity).Untreated, hydrophilic SiO2 nanoparticles were used to enhance

212 P.S. Brown, B. Bhushan / Journal of Colloid and Interface Science 456 (2015) 210–218

the roughness of the coating. The negatively charged surface sila-nol groups ensure favorable interactions with the positivelycharged polymer layers. Additionally, SiO2 nanoparticles areknown to have high hardness [24] and wear resistance [25], whichwill aid in the creation of a mechanically durable coating [26].Particles of 7 nm in diameter were selected with the goal to createa transparent coating. The fluorosilane was selected to providerepellency because of its low surface tension fluorinated tail andability to form a self-assembled layer via vapor phase deposition.Silanes have been shown to condense on hydrophilic polymer lay-ers in the past due to the presence of absorbed water [27].

2.1. Samples

Glass slides (Fisher Scientific) cut to dimensions of 25 by 10 mmwere used as substrates. Polydiallyldimethylammonium chloride(PDDA, MW 100,000–200,000, Sigma Aldrich) was dissolved in dis-tilled water (DS Waters of America Inc.) at various concentrations.Silica nanoparticles (NP, 7 nm diameter, Aerosil 380, EvonikIndustries) were dispersed in acetone (Fisher Scientific Inc.) usingan ultrasonic homogenizer (Branson Sonifier 450A, 20 kHz fre-quency at 35% amplitude) at various concentrations. Coatings weredeposited via spray gun (Paasche) operated with compressed air at210 kPa. The gun was held 10 cm from the glass slide at all times.First, PDDA solution (52 mg mL�1, 2 mL) was spray coated and anyexcess was removed from the surface via bursts of compressed airfrom the spray gun. Second, the SiO2 NP solution (variousconcentrations, 3 mL) was spray coated. Third, a second PDDAlayer was deposited (8 mg mL�1, 1 mL). After this, the sampleswere transferred to an oven operating at 140 �C for 1 h. Finally,the fluorosilane layer (FL) was deposited via chemical vapor depo-sition under atmospheric conditions. One drop of trichloro(1H,1H,2H,2H-perfluorooctyl) silane (Sigma Aldrich) was deposited nextto the samples which were covered and left for 6 h.

2.2. Contact angle and tilt angle

For contact angle data, droplets of 5 lL volume were depositedonto samples using a standard automated goniometer (Model 290,Rame-Hart Inc.) and the resulting image of the liquid–air interfaceanalyzed with DROPimage software. Tilt angles were measured byinclining the surface until the 5 lL droplet rolled off. Contact anglehysteresis was measured by tilting the substrate until the dropletwas observed to move and the advancing and receding angles wererecorded. These numbers were found to be comparable to the tiltangles and are not reported. All angles were averaged over at leastfive measurements on different areas of a sample. Oils testedinclude n-hexadecane (99%, Alfa Aesar), n-tetradecane (P99%,Sigma Aldrich), n-dodecane (P99%, Sigma Aldrich), n-decane(P99%, Sigma Aldrich), and n-octane (P99%, Sigma Aldrich).

2.3. Surface topography and coating thickness

The surface topography of each sample was determined using aD3000 Atomic Force Microscopy (AFM) with a Nanoscope IVcontroller (Bruker Instruments). A Si, n-type (Si3N4) tip with anAl coating (resonant frequency f = 66 kHz, spring constantk = 3 N m�1, AppNano) operating in tapping mode was used. Thescanning area was 1 � 1 lm2 to determine roughness on thenanoscale. Root mean square roughness (RMS) andPeak-to-Valley (P–V) distance values were obtained.

Coating thickness of each individual layer and the compositecoating was measured with a step technique. One half of the sub-strate was covered with a glass slide using double-sided sticky tapebefore coating and then removed after the coating procedure

resulting in a step. An area including the step was imaged byAFM to obtain the coating thickness.

2.4. Wear experiments

The mechanical durability of the surfaces was examinedthrough wear experiments using an AFM and a ball-on-flattribometer [28]. An established AFM micro-wear procedure wasperformed with a commercial AFM (D3000, Nanoscope IV con-troller, Bruker Instruments). Surfaces were worn using a borosili-cate ball with radius 15 lm mounted on a rectangular cantileverwith nominal spring constant of 7.4 N m�1 (resonant frequencyf = 150 kHz, All-In-One). Areas of 50 � 50 lm2 were worn for 1cycle at a load of 10 lN so as to be later imaged within the scan-ning limits of the AFM. To analyze the change in morphology ofthe surface before and after the wear experiment, height scans of100 � 100 lm2 in area were obtained using a Si, n-type (Si3N4)tip with an Al coating (resonant frequency f = 66 kHz,k = 3 N m�1, AppNano) operating in tapping mode. Root meansquare roughness (RMS) values before and after wear experimentswere obtained.

Macrowear experiments were performed with an establishedprocedure of using a ball-on-flat tribometer [26]. A sapphire ballof 3 mm diameter was fixed in a stationary holder. A load of10 mN was applied normal to the surface, and the tribometerwas put into reciprocating motion. Stroke length was 6 mm withan average linear speed of 1 mm s�1. Surfaces were imaged beforeand after the tribometer wear experiment using an optical micro-scope with a CCD camera (Nikon Optihot-2) to examine anychanges [24].

Contact pressures for both AFM and tribometer wear experi-ments were calculated based on Hertz analysis [26]. The elasticmodulus of PDDA, 0.16 GPa (Ref. [29]), was used to estimate theelastic modulus of the composite coating, and a Poisson’s ratio of0.5 was used (estimated). The elastic modulus of final coating isexpected to be higher, so an underestimated pressure will beobtained with the selected modulus. The elastic modulus of70 GPa and Poisson’s ratio of 0.2 were used for the borosilicate ballused in the microscale wear experiments [30]. The elastic modulusof 390 GPa and Poisson’s ratio of 0.23 were used for sapphire ballused in the macroscale wear experiments [31]. The mean contactpressures were calculated as 4.87 MPa and 2.26 MPa for the AFM(micro) and ball-on-flat tribometer (macro) experimentsrespectively. Microscale wear experiments were performed for 1cycle while macroscale wear experiments were performed for100 cycles. Therefore, the macroscale wear experiments can causea relatively high degree of damage to the coating even though themean contact pressures are comparable to the microscaletechnique.

2.5. Self-cleaning experiment

The self-cleaning characteristics of the surfaces were examinedusing an experimental setup previously reported [32]. Coatingswere contaminated with silicon carbide (SiC, Sigma Aldrich) in aglass chamber (0.3 m diameter and 0.6 m high) by blowing 1 g ofSiC powder onto a sample for 10 s at 300 kPa and allowing it to set-tle for 30 min. The contaminated sample was then secured on astage (45� tilt) and water droplets (total volume 5 mL) weredropped onto the surface from a specified height. Once dried,images were taken using an optical microscope with a CCD camera(Nikon, Optihot-2). The removal of particles by the water dropletswas compared before and after tests. The ability for the waterstream to remove particles was quantified using image analysissoftware (SPIP 5.1.11, Image Metrology A/S, Horshølm, Denmark).

P.S. Brown, B. Bhushan / Journal of Colloid and Interface Science 456 (2015) 210–218 213

2.6. Anti-smudge experiment

The anti-smudge characteristics of the surfaces were examinedusing an experimental setup previously reported [33]. Coatingswere contaminated as reported above. The contaminated samplewas then secured on a stage and a hexadecane-impregnated micro-fiber wiping cloth was glued to a horizontal glass rod (radius0.5 mm) fixed on a cantilever above the sample. As the cloth wasbrought in contact with the sample, the microfiber cloth was setto rub the contaminated sample under a load of 5 g for 1.5 cm ata speed of about 0.2 mm s�1. Images were taken using an opticalmicroscope with a CCD camera (Nikon, Optihot-2). The removaland transfer of particles by the cloth was compared before andafter tests.

2.7. Transparency measurements

A line-of-sight light extinction apparatus was assembled using adiffractive spectrometer (Acton, Princeton Instruments), an inten-sified CCD camera and an incandescent light bulb as a point source,which emitted a black-body type spectrum across the 400–700 nmbandwidth of interest. The sample slides were placed within1–2 mm of the incandescent light source. A pair of 50-mm diame-ter, 100-mm focal length plano-convex lenses was used to collectemission from the light source and focus it onto the entrance slitsof the spectrometer. For a single camera exposure, the spectrome-ter bandwidth was approximately 80 nm, so the grating wasstepped at �60 nm intervals to sample the entire bandwidth with

Fig. 2. Hexadecane and water droplets (5 lL) deposited on flat and layer-by-layercomposite coating with SiO2 NP concentration of 15 mg mL�1.

Table 1Comparison of various liquids deposited on flat (PDDA/FL) and layer-by-layercomposite coatings.

Liquid Surface tension [35](mN m�1)

Flat PDDA/FLcoating

Layer-by-layercomposite coating

Contact angle(�)

Contactangle (�)

Tiltangle (�)

Octane 21.14 62 ± 2 155 ± 1 9 ± 1Decane 23.37 69 ± 2 159 ± 1 7 ± 1Dodecane 25.35 76 ± 1 159 ± 1 5 ± 1Tetradecane 26.13 78 ± 1 158 ± 1 4 ± 1Hexadecane 27.05 83 ± 1 157 ± 1 4 ± 1Water 71.99 109 ± 1 163 ± 1 2 ± 1

Fig. 3. AFM surface height maps with RMS and P–V roughness values for glass,PDDA/FL, and layer-by-layer composite coating with SiO2 NP concentration of15 mg mL�1.

Fig. 4. (a) Static contact angle and tilt angle for hexadecane as a function of SiO2 NP concentration and (b) static contact angle and tilt angles for various liquids on theoptimized surface as a function of liquid surface tension. Oils used: octane (21.14 mN m�1), decane (23.37 mN m�1), dodecane (25.35 mN m�1), tetradecane (26.13 mN m�1)and hexadecane (27.05 mN m�1).

Fig. 5. (a) Surface height maps and sample surface profiles (locations indicated by arrows) before and after AFM wear experiment with 15 lm radius borosilicate ball at a loadof 10 lN for flat and optimized layer-by-layer composite coatings. RMS roughness values are displayed, and (b) optical micrographs before and after wear experiments usingball-on-flat tribometer at 10 mN for flat and layer-by-layer composite coatings.

214 P.S. Brown, B. Bhushan / Journal of Colloid and Interface Science 456 (2015) 210–218

an overlap of about 40 nm between each grating position. A singlecamera exposure was acquired at each grating position. The spec-tra were then background subtracted and divided by the spectrumacquired from an uncoated glass slide and the data plotted as afunction of wavelength (400–700 nm).

3. Results and discussion

The final, optimized composite coating comprises four separatelayers (total thickness ca. 625 nm) each deposited individually,Fig. 1. The first layer comprises PDDA (thickness ca. 200 nm) andacts as an anchor layer to the glass substrate. The second layer con-tains SiO2 nanoparticles (NP, thickness ca. 350 nm) and acts as theroughness layer, enhancing the overall liquid–solid interactions.

Third is a second polymer layer (PDDA (2), thickness ca. 50 nm),which helps to bind the nanoparticle layer, improving mechanicaldurability. The final layer is the fluorosilane layer (FL, thickness ca.25 nm), which condenses onto the hydrophilic PDDA (2) layer andprovides the oil-repellency. Deposition of a separate fluorosilanelayer ensures the correct functionality at the air interface (super-omniphobicity) without compromising the durability of the bulkcoating.

3.1. Wettability of coated surfaces

Initially, two layer coatings of spray coated PDDA (52 mg mL�1,thickness ca. 200 nm) followed by vapor deposited FL (PDDA/FL) onglass substrates were created to determine the oil-repellency of the

P.S. Brown, B. Bhushan / Journal of Colloid and Interface Science 456 (2015) 210–218 215

flat polyelectrolyte–fluorosilane coating. Fig. 2 shows representa-tive pictures of the contact angles for droplets of water and hex-adecane deposited on this coating. Table 1 provides a summaryof all contact angle data. The flat coating was found to result infinite contact angles for all liquids tested, in contrast to bare glassand PDDA coatings, which were wet by all the liquids (contactangle < 5�). The PDDA/FL coating repels oils to some extent (finitecontact angles) in addition to being hydrophobic with a water con-tact angle of 109 ± 1�.

To enhance the repellency, roughness was introduced via spraydeposition of a SiO2 nanoparticle layer (15 mg mL�1). Fig. 3 showsthe AFM surface height maps, RMS and P–V values determinedfrom several 1 � 1 lm2 scan areas of the various surfaces. Thisincrease in roughness results in an increase in the water and oilcontact angles (Fig. 2 and Table 1) due to the formation of a com-posite air/solid interface typically known as the Cassie–Baxterstate.

To ensure optimal repellency, the concentration of the SiO2 NPcoating solution was varied and the hexadecane contact and tiltangles measured. The concentrations of the PDDA (1) andPDDA (2) anchor layers were kept constant (52 mg mL�1 and8 mg mL�1 respectively) as these were found to have no effect onthe repellency of the final coating. The data shown in Fig. 4a indi-cates that, across the range of NP concentrations studied, hexade-cane contact angles remained above 150� indicating that dropletswere in the Cassie–Baxter regime on all coatings. It was found that,at higher NP concentrations, the tilt angle for hexadecane dropletsremained high. This is thought to be due to an increased solid con-tribution to the composite air/solid interface, leading to a decrease

Fig. 6. Images of hexadecane droplets (5 lL) on the optimized layer-by-layercomposite coating before and after wear and scratching. Droplets were dragged ortilted across wear track (centered) in direction of arrows. For the worn samples(center column), droplets either rolled over the wear track as normal at 4 ± 1� tiltangle (when placed to the right of the defect) or rolled from the wear track aftertilting the sample 6 ± 1� (when placed directly over the defect). Droplets on thescratched sample (control) were pinned at the defect site until 30 ± 2� tilt angleregardless of starting location.

in the number of air pockets underneath the droplet. Increasing thearea of solid surface in contact with the droplet will better pin thedroplets, which causes an increase in hysteresis [6]. By reducingthe NP concentration, the hexadecane tilt angle is reduced due toan increase in the air contribution to the interface and a decreasein pinning. Below a concentration of 15 mg mL�1, the tilt anglesbegin to rise again due to a decrease in the amount of air thatcan be effectively trapped between the surface and the liquid.This leads to a greater adhesive force between the two and a cor-responding increase in hysteresis.

Oil repellency of the layer-by-layer composite coatings was fur-ther investigated using a series of straight-chain alkane oils and thedata is presented in Table 1 and Fig. 4b. The coating was found toremain superoleophobic for tetradecane, dodecane, decane, andoctane, with slight increases in tilt angles for the latter two, dueto their lower surface tensions. This repellency over a range of oils,in addition to water repellency, means the coatings can be consid-ered superomniphobic.

3.2. Wear resistance of coated surface

The mechanical durability of the coatings was investigatedthrough the use of AFM and tribometer wear experiments andthe resulting images are shown in Fig. 5. AFM images show a100 � 100 lm2 scan area with the wear location (50 � 50 lm2) in

Fig. 7. Optical micrographs of contaminated coatings before and after self-cleaningtest on flat and optimized layer-by-layer composite coatings. Dark spots on coatingsand cloth indicate silicon carbide particle contaminants. Image analysis suggestsa > 90% removal of particles on composite coating.

Fig. 8. Optical micrographs of contaminated coatings and oil-impregnated micro-fiber cloth before and after smudge test on flat and optimized layer-by-layercomposite coatings. Dark spots on coatings and cloth indicate silicon carbideparticle contaminants.

Fig. 9. Photographs of flat and optimized layer-by-layer composite coatings. Theflat coating appears transparent. Any reduction in transparency for the compositecoating compared to the flat coating is due to the SiO2 NP layer.

216 P.S. Brown, B. Bhushan / Journal of Colloid and Interface Science 456 (2015) 210–218

the center of each image. The optical images show a portion of thewear track from the tribometer experiments. For the soft PDDA/FLcoating (ca. 225 nm thick), there is significant wear with both AFMand tribometer experiments causing observable damage to the sur-face. In contrast, the layer-by-layer composite coating survived theAFM wear experiment with no observable defects. For the tribome-ter experiment, there is some noticeable burnishing to the coating,however it is minimal when compared to the PDDA/FL coating.This suggests that the hard SiO2 nanoparticle layer (underneathca. 75 nm thick PDDA/FL layers) helps improve the durability ofthe coating, while the oppositely charged PDDA binder layers helpanchor the particles to the glass substrate via an electrostatic bond.This is in contrast to other polymer-nanoparticle coatings wherethe interfacial adhesion is not aided by this electrostatic attraction.

To further investigate the durability of the coating, hexadecanetilt angles were measured before and after tribometer wear exper-iments, Fig. 6. Despite a noticeable wear track under optical

microscope, droplets of hexadecane tilted or dragged over the loca-tion of the wear experiment were not impeded in their motion.When deposited directly onto the wear track, there was a smallincrease in tilt angle before the droplet was observed to roll away.This suggests that any damage caused to the coating is minimaland the fluorosilane layer remains intact. In contrast, controlexperiments featuring coatings deliberately destroyed with tweez-ers resulted in a large increase in tilt angles with the hexadecanedroplets pinning when deposited directly on the scratch or whentilted to roll over it.

3.3. Self-cleaning property of coated samples

To examine the self-cleaning properties, the coatings were con-taminated with silicon carbide particles, Fig. 7. A stream of waterdroplets was then used to clean the surface. On the flat PDDA/FLcoating this resulted in an incomplete removal of the particles withthe surface remaining contaminated. For the layer-by-layer com-posite coating, the vast majority of the particles were removedfrom the coating by the action of water droplets rolling acrossthe repellent surface, collecting particles in the process.

3.4. Anti-smudge property of coated samples

To examine the anti-smudge properties, a hexadecane-soakedcloth was then used to wipe a contaminated surface, Fig. 8. Onthe flat PDDA/FL coating this resulted in incomplete removal of

P.S. Brown, B. Bhushan / Journal of Colloid and Interface Science 456 (2015) 210–218 217

the particles with the surface remaining contaminated. For thelayer-by-layer composite coating, the particles were transferredto the cloth with no observable particles remaining on the coating.

3.5. Transparency of coated samples

Many applications of self-cleaning, anti-smudge surfaces relyon the transparency of the coating. When placed directly behindthe layer-by-layer composite coating sample, text remains legible,suggesting that the coating displays characteristics of trans-parency, Fig. 9. The transmission of visible light through thelayer-by-layer composite coating was found to be 65 ± 5% of thatof uncoated glass. The flat PDDA/FL coating exhibited similar trans-parency (99 ± 1%) to bare glass suggesting that any decrease intransparency for the composite coating is due to the NP layer.Further improvement in transparency, potentially by decreasingthe thickness of the NP layer or reducing particle agglomeration,will be investigated in the future. A level of 70% visible light trans-mittance is acceptable for certain automotive applications [34].

4. Conclusions

A durable, superomniphobic coating has been fabricatedthrough the use of a novel combination of deposition techniquesutilizing the charged layer-by-layer method for durability plusthe addition of an uncharged fluorosilane layer on top for super-omniphobicity. The fluorosilane layer provides repellency and,when combined with the layer-by-layer stack, results in a durable,superomniphobic coating with water and hexadecane contactangles exceeding 155� and tilt angles of 4� or less. The coating isrepellent to liquids with surface tensions ranging from 72 to 21mN m�1 and can be considered superomniphobic.

The coating is found to be mechanically durable with micro-and macrowear experiments not causing any considerable damagedue to the hard SiO2 nanoparticles and the electrostatic interactionbetween the base layers. After wear tests, hexadecane droplets stillroll off from the wear location suggesting any damage is minimaland the coating remains super-repellent. The coatings were foundto be self-cleaning with a stream of water droplets easily removingcontaminants from the surface. The coatings were also found to beanti-smudge with a hexadecane-soaked cloth removing all con-taminants from the surface.

Finally, these surfaces were found to display characteristics oftransparency with 65 ± 5% transmission and text remaining visiblethrough the coating. This level of transparency is acceptable forcertain automotive applications. It is believed that the trans-parency can be improved in the future by reducing the thicknessof the composite coating, layer-by-layer coatings are typicallymuch thinner than reported here.

The coating could find use in anti-fouling, self-cleaning,anti-smudge and low-drag applications. In particular, it could beuseful in anti-biofouling, where superomniphobicity and nanos-tructuring can both contribute to reducing microorganismattachment.

Author contributions

P.S.B. performed the experiments and analyzed the data. P.S.B.wrote the main text and P.S.B. and B.B. participated equally in plan-ning, execution, and review of the manuscript.

Competing financial interests

The authors declare there are no competing financial interests.

Acknowledgment

The authors thank Kraig Frederickson for performing the trans-parency measurements.

References

[1] C.-T. Hsieh, J.-M. Chen, R.-R. Kuo, T.-S. Lin, C.-F. Wu, Influence of surfaceroughness on water- and oil-repellent surfaces coated with nanoparticles,Appl. Surf. Sci. 240 (2005) 318–326.

[2] A. Tuteja et al., Designing superoleophobic surfaces, Science 318 (2007) 1618–1622.

[3] Y.C. Jung, B. Bhushan, Wetting behavior of water and oil droplets in three-phase interfaces for hydrophobicity/philicity and oleophobicity/philicity,Langmuir 25 (2009) 14165–14173.

[4] T. Darmanin et al., Superoleophobic behavior of fluorinated conductivepolymer films combining electropolymerization and lithography, Soft Matter7 (2011) 1053–1057.

[5] Z. He et al., Fabrication of a transparent superamphiphobic coating withimproved stability, Soft Matter 7 (2011) 6435–6443.

[6] B. Bhushan, Biomimetics: Bioinspired Hierarchical-Structured Surfacesfor Green Science and Technology, Springer-Verlag, Heidelberg, Germany,2012.

[7] G.D. Bixler, B. Bhushan, Fluid drag reduction and efficient self-cleaning withrice leaf and butterfly wing bioinspired surfaces, Nanoscale 5 (2013) 7685–7710.

[8] G.D. Bixler, A. Theiss, B. Bhushan, S.C. Lee, Anti-fouling properties ofmicrostructured surfaces bio-inspired by rice leaves and butterfly wings, J.Colloid Interfaces Sci. 419 (2014) 114–133.

[9] Y. Wang, B. Bhushan, Wear-resistant and antismudge superoleophobic coatingon polyethylene terephthalate substrate using SiO2 nanoparticles, ACS Appl.Mater. Interfaces 7 (2015) 743–755.

[10] T. Young, An essay on the cohesion of fluids, Philos. Trans. R. Soc. 95 (1805)65–87.

[11] R.N. Wenzel, Resistance of solid surfaces to wetting by water, Ind. Eng. Chem.28 (1936) 988–994.

[12] A.B.D. Cassie, S. Baxter, Wettability of porous surfaces, Trans. Faraday Soc. 40(1944) 546–551.

[13] H.W. Fox, W.A. Zisman, Some advances in techniques for the study ofadsorbed monolayers at the liquid–air interface, Rev. Sci. Instrum. 19 (1948)274.

[14] K. Tsujii, T. Yamamoto, T. Onda, S. Shibuichi, Super oil-repellent surfaces,Angew. Chem. Int. Ed. Engl. 36 (1997) 1011–1012.

[15] S. Pan, A.K. Kota, J.M. Mabry, A. Tuteja, Superomniphobic surfaces for effectivechemical shielding, J. Am. Chem. Soc. 135 (2013) 578–581.

[16] J. Yang, Z. Zhang, X. Men, X. Xu, X. Zhu, A simple approach to fabricatesuperoleophobic coatings, New J. Chem. 35 (2011) 576–580.

[17] H. Jin, X. Tian, O. Ikkala, R.H.A. Ras, Preservation of superhydrophobic andsuperoleophobic properties upon wear damage, ACS Appl. Mater. Interfaces 5(2013) 485–488.

[18] M. Onda, Y. Lvov, K. Ariga, T. Kunitake, Sequential reaction and productseparation on molecular films of glucoamylase and glucose oxidase assembledon an ultrafilter, J. Ferment. Bioeng. 82 (1996) 502–506.

[19] A.C. Fou, M.F. Rubner, Molecular-level processing of conjugated polymers. 2.Layer-by-layer manipulation of in-situ polymerized p-type doped conductingpolymers, Macromolecules 28 (1995) 7115–7120.

[20] X. Liu, J. He, Superhydrophilic and antireflective properties of silicananoparticle coatings fabricated via layer-by-layer assembly andpostcalcination, J. Phys. Chem. C 113 (2009) 148–152.

[21] K.K. Lee, C.H. Ahn, Superhydrophilic multilayer silica nanoparticle networks ona polymer microchannel using a spray layer-by-layer nanoassembly method,ACS Appl. Mater. Interfaces 5 (2013) 8523–8530.

[22] P.S. Brown, B. Bhushan, Mechanically durable, superoleophobic coatingsprepared by layer-by-layer technique for anti-smudge and oil–waterseparation, Sci. Rep. (Nature) 5 (2015) 8701.

[23] X. Du, X. Li, J. He, Facile fabrication of hierarchically structured silica coatingsfrom hierarchically mesoporous silica nanoparticles and their excellentsuperhydrophilicity and superhydrophobicity, ACS Appl. Mater. Interfaces 2(2010) 2365–2372.

[24] D. Ebert, B. Bhushan, Durable Lotus-effect surfaces with hierarchical structureusing micro- and nanosized hydrophobic silica particles, J. Colloid InterfacesSci. 368 (2012) 584–591.

[25] Y. Lvov, K. Ariga, M. Onda, I. Ichinose, T. Kunitake, Alternate assembly ofordered multilayers of SiO2 and other nanoparticles and polyions, Langmuir 13(1997) 6195–6203.

[26] B. Bhushan, Introduction to Tribology, second ed., Wiley, New York, 2013.[27] Y. Xie, C.A.S. Hill, Z. Xiao, H. Militz, C. Mai, Silane coupling agents used for

natural fiber/polymer composites: a review, Composites: Part A 41 (2010)806–819.

[28] B. Bhushan, Nanotribology and Nanomechanics I � Measurement Techniquesand Nanomechanics, II � Nanotribology, Biomimetics, and IndustrialApplications, third ed., Springer-Verlag, Heidelberg, Germany, 2011.

218 P.S. Brown, B. Bhushan / Journal of Colloid and Interface Science 456 (2015) 210–218

[29] P. Podsiadlo, Layer-by-Layer Assembly of Nanostructured Composites:Mechanics and Applications, Ph.D. Thesis, University of Michigan, Ann Arbor,Michigan, 2008.

[30] W.D. Callister, D.G. Rethwisch, Materials Science and Engineering � AnIntroduction, ninth ed., Wiley, New York, 2013.

[31] B. Bhushan, B.K. Gupta, Handbook of Tribology: Materials, Coatings, andSurface Treatments, McGraw-Hill, New York, 1991.

[32] G.D. Bixler, B. Bhushan, Bioinspired rice leaf and butterfly wing surfacestructures combining shark skin and lotus effects, Soft Matter 8 (2012) 11271–11284.

[33] B. Bhushan, P. Muthiah, Anti-smudge screening apparatus for electronic touchscreens, Microsyst. Technol. 19 (2013) 1261–1263.

[34] S.V. Thomsen, R. Hulme, L. Landa, K. Landa, High visible transmission andinfrared/ultraviolet radiation absorption; for automotive windows(windshields, sidelites, backlites and sunroofs) and in architectural windows,US Patent 20050020430 A1 filed 21 July 2003 (issued 27.01.05).

[35] W.M. Haynes, Handbook of Chemistry and Physics, 95th ed., CRC Press, BocaRaton, FL, 2014.