Liquid repellent nanocomposites obtained from one-step water-based spray

Journal ofMaterials Chemistry A

PAPER

Publ

ishe

d on

14

May

201

5. D

ownl

oade

d by

Uni

vers

ity o

f V

irgi

nia

on 2

0/10

/201

5 05

:32:

56.

View Article OnlineView Journal | View Issue

Liquid repellent

aDepartment of Mechanical and Aerospace

Engineer's Way, 22904 Charlottesville, VA, UbDepartment of Nanochemistry, Istituto Itali

Genoa, Italy. E-mail: [email protected] Materials, Istituto Italiano di Tecnol

† Electronic supplementary information (impacting the nanocomposite coatings.ketchup and honey mustard on the10.1039/c5ta02672e

Cite this: J. Mater. Chem. A, 2015, 3,12880

Received 13th April 2015Accepted 14th May 2015

DOI: 10.1039/c5ta02672e

www.rsc.org/MaterialsA

12880 | J. Mater. Chem. A, 2015, 3, 12

nanocomposites obtained fromone-step water-based spray†

A. Milionis,*a K. Dang,a M. Prato,b E. Lotha and I. S. Bayer*ac

Significant research efforts have been directed towards the development of novel superhydrophobic and

superoleophobic coatings. However, it is still highly challenging to develop facile and environmentally

friendly methods that lead to surfaces and coatings with efficient liquid repellency. Herein, we

demonstrate for the first time, a one-step water-based spray method on heated metallic substrates for

the preparation of water and oil repellent polymer nanocomposite coatings by using hydrophilic silica

nanoparticles. These environmentally-friendly coatings exhibit static contact angles exceeding 150� and

roll off angles less than 5� for both water and oil and can also successfully repel a wide variety of other

liquids. The influence of the chemical composition and the surface texture on the wetting properties are

discussed based on measurements from white light interferometry, electron microscopy and X-ray

photoelectron spectroscopy. When the substrates were treated with an adhesive primer, the wear

abrasion resistance of the coatings was enhanced. The effect of silica nanoparticle dispersion in the

coatings on wear abrasion is also investigated. Certain nanocomposites were found to exhibit good

abrasion resistance by retaining their water repellency for up to 60 abrasion cycles under 20.5 kPa

applied pressure.

Introduction

Technologies related to liquid repellent materials have drawnsignicant attention and research efforts during the past yearsdue to their vast number of commercial, industrial and militarypotential applications. These efforts have led to a large numberof publications as well as new commercial products.1–5 Manyexamples are worth mentioning such as self-cleaning surfaces,6

water–oil separation,7 anti-icing,8 anti-fouling,9 anti-nger-print,10 stain-free protective clothing,11 drag reduction,12

sensors,13 droplet manipulation14 and robotics.15 These surfacesare referred to as superhydrophobic if they display water contactangles (WCA) >150� and very low droplet roll off angles (typically<10� or 5�), and if they have the same performance with oils,they are referred to as superoleophobic. Most superoleophobicsurfaces are also superhydrophobic because surfaces that canrepel low surface tension liquids such as oils and alcohols caneasily repel water, which possesses a higher surface tension.

Engineering, University of Virginia, 122

SA. E-mail: [email protected]

ano di Tecnologia, Via Morego 30, 16163

ogia, Via Morego 30, 16163 Genoa, Italy

ESI) available: Video 1: different liquidsVideo 2: deposition and removal ofnanocomposite coatings. See DOI:

880–12889

However, recently surfaces that are superoleophobic and yetsuperhydrophilic have also been designed and reported.16,17

Although all these studies are successful in terms of appli-cation, very few of them use materials that can be classied asenvironmentally friendly or follow green fabrication processes.Only in the last few years such approaches have becomeavailable for liquid repellent materials and coatings. Forinstance, biodegradable superhydrophobic nanocompositecoatings have been recently reported by combining a starch-based thermoplastic with hydrophobic nano-silica and lyco-podium spores.18 However, the solvent medium used waschloroform which is considered hazardous.19 In this study,coatings were developed based on an environmentallyapproved C-6 uoroacrylic compound for oil repellence. Thestatic contact angles were greater than 150� but the roll offangles (ROAs) reported were slightly higher than 15�, above theaccepted threshold for a surface to be superoleophobic. Otherbiodegradable materials that have been used for super-hydrophobic surfaces include cellulose,20 poly(lactic acid),21

hydroxycaproic acid,22 magnesium alloys23 and poly(3-capro-lactone)24 and their fabrication techniques include plasmaprocessing, chemical vapor deposition, in situ nanoparticlegrowth, electrospinning, phase separation, polymerization,spay and dip-coating using various solvents. Again, most ofthese techniques require multiple fabrication steps, are diffi-cult to scale-up, use ammable or hazardous solvents and themajority of them are surface functionalizations rather thanpolymer nanocomposite coatings.

This journal is © The Royal Society of Chemistry 2015

Paper Journal of Materials Chemistry A

Publ

ishe

d on

14

May

201

5. D

ownl

oade

d by

Uni

vers

ity o

f V

irgi

nia

on 2

0/10

/201

5 05

:32:

56.

View Article Online

On the other hand, superoleophobicity is even more chal-lenging to achieve with sustainable materials since there is nonatural surface that can completely repel oil (excluding under-water systems). In fact, only underwater oil repellency has beenreported by natural surfaces like sh-scales25 or the lower part ofthe lotus leaf.26 Re-entrant surface curvature, chemical compo-sition and nanoscale texture are crucial parameters fordesigning articial superoleophobic surfaces.3

Previously, superoleophobic surfaces have been fabricatedby electrospinning,3,27 plasma etching,3 spray-coating15,28–32 anddip-coating.31,32 Among these techniques, spray-coating isattractive since it is fast, can be easily automatized in industrialscale and can be adapted to coat large surface areas. There arevery few studies on the production of liquid repellent surfacesbased on water sprays as many hydrophobic or uorinatedpolymers are not soluble in water. The use of water as solventmedium is very attractive since it is non-ammable while at thesame time does not impose any additional health risks to theusers. Motornov et al. prepared superhydrophobic surfaces bycasting waterborne suspensions of hybrid, organic/inorganicnanoparticles, evaporating the water and heating above theglass transition temperature of the organic component.33

Recently, Mates et al.34 and Schutzius et al.35 reported super-hydrophobic coatings by spraying water emulsions of a hydro-phobic polyolen. In some other cases, water has been used as aco-solvent in nanocomposite dispersions.30,36 However, the useof water as solvent is reported to cause problems of coatingadhesion to the substrate due to dewetting effects duringdrying.37 As such, it is quite challenging to produce super-oleophobic coatings using a single-step water-based spray coatthat does not require post surface texturing or chemical treat-ment, and to the best of our knowledge, this has not beenpreviously reported. Particularly, in the case of metal substrates,most of the superoleophobic treatments include the use ofperuorinated or highly corrosive acids for etching andtexturing and functionalization with uorosilanes. Many ofthese substances are ammable, volatile, corrosive andtoxic.38–42 In particular, peruorinated acids with C-8 chemistryor longer are susceptible for bio-accumulation by degrading viaperuorooctanoic acid routes.43,44

Herein, we present a simple, one-step, water-based spraycoating process to fabricate superhydrophobic and super-oleophobic nanocomposite coatings on metals comprisinghydrophilic silica nanoparticles and uoroacrylic polymer withC-6 chemistry. Use of water as the main solvent ensures thatmany environmental concerns related to solvent use are elimi-nated while the C-6 chemistry is approved for non-toxic degra-dation products hence minimizing abovementioned bio-accumulation risks associated with C-8 uoro-compounds. Inorder to avoid surface dewetting and accelerate atomizeddroplet evaporation, the substrates were kept at elevatedtemperatures during spray coating. Water and oil repellency ofthe composites were tuned by varying the ratio of the constit-uents in solution and the inuence of different particle sizedistribution was also investigated. Moreover, the coatings wereable to withstand linear mechanical abrasion for 55 cycles with20.5 kPa applied pressure, preserving their superhydrophobic

This journal is © The Royal Society of Chemistry 2015

characteristics. This environmentally benign method is solelybased on water as the carrier solvent for both the polymer andthe nanollers and due to its simplicity; it is expected to be anattractive alternative for a wide range of applications requiringgreen fabrication routes.

MethodsMaterials

Spherical and porous silicon dioxide nanopowder with particlesize 5–15 nm (nanollers A) was purchased from Sigma-Aldrich.Hydrophilic fumed silica, Aerosil® 200 (nanollers B), wasprovided by Evonik Industries, Germany. The waterborneacrylic uorochemical dispersion, Capstone® ST-110 (WAF) waspurchased from DuPont™, USA. More details about the envi-ronmental prole of this material are given elsewhere.45 Toimprove surface adhesion of the coatings the metal surfaceswere pre-treated with a primer, Rust-Oleum™ 209460 PlasticPrimer (White), USA. Multipurpose aluminum alloy 6061surfaces were purchased from McMaster Carr, USA and wereused as substrates. Mineral oil, light (CAS 8042-47-5) waspurchased from Fisher Scientic and used as received for use inoil wetting characterization.

Sample preparation

Initially, the nanoparticles (either nanollers A or B) weredispersed in 42 ml of deionized water using ultrasonic pro-cessing (Sper Scientic: 100004) for 8 min. Subsequently 3 ml ofCapstone® ST-110 was added and the mixture was stirred for 1minute. Five varying concentrations for each type of particleswere used in order to prepare different nanocomposite disper-sions. Namely, the concentrations were adjusted by varyingweight of the nanoparticles in solution from 0.1 and 0.5 g withincrements of 0.1 g. The substrates were 1 mm thick aluminum6061 alloy thin plates of rectangular shape (2.5 � 5.0 cm2) thatwere cleaned by using various solvents prior to spray-deposition.

The coatings were prepared by spray-coating the nano-composite dispersions. For the spray coating, a VL doubleaction, internal mix, siphon feed airbrush was used (Paasche,USA). The spray distance was held constant at approximately 15cm and the air pressure was set at 200 kPa. The samples weresprayed in 4 steps of 5 s each, with an interval of 30 s betweeneach cycle in order to ensure that there is sufficient time for thewater to evaporate. During the spraying process, the substrateswere placed on a hotplate maintained at a constant temperatureof 300 �C. The nal thickness of the coatings was about 5 mm.Thicker versions of the lms were made for wear abrasionexperiments. Both sets of lms had similar initial wettabilityand surface topology.

Wettability and surface topology

Water and oil contact angles (WCAs, OCAs) and droplet roll-offangles (ROAs) on the surfaces were measured by a video basedoptical contact angle measuring instrument Rame-Hart, USA.Ten microliter (10 ml) of deionized water and mineral oil

J. Mater. Chem. A, 2015, 3, 12880–12889 | 12881

Journal of Materials Chemistry A Paper

Publ

ishe

d on

14

May

201

5. D

ownl

oade

d by

Uni

vers

ity o

f V

irgi

nia

on 2

0/10

/201

5 05

:32:

56.

View Article Online

droplets were gently placed on the surfaces. Measurements wereconducted on three different locations and averaged for eachsample. In order to measure ROAs, the substrates were tilteduntil the droplets started to roll off the surfaces. The substrateangle at the onset of droplet roll-away was recorded. All ROAvalues were averaged over three different measurements oneach sample. The standard deviation was �4� for all CAmeasurements and �1� for ROA measurements. All measure-ments were performed in ambient conditions.

The morphology of the textured surfaces was characterizedby scanning electron microscopy (SEM; Model: FEI Quanta 650,USA) and surface roughness parameters were extracted by usinga white light interferometer (Zygo, USA). To eliminate chargingeffects, the samples were coated with a very thin (�20 nm) layerof Au/Pd by sputtering. The size distribution of the two differenttypes of particles tested was investigated by transmission elec-tron microscopy (TEM; Model: JEOL 2000FX, Japan). All thesamples for TEM analysis were prepared by immersing carbon-coated 200 mesh, 50 mm copper grids in the nanoparticledispersions, and then allowing to dry overnight under inertatmosphere.

X-ray photoelectron spectroscopy (XPS)

Analysis of the chemical composition of the samples' surfaceswas performed by X-ray photoelectron spectroscopy (XPS). Inparticular, XPS was performed with a Kratos Axis Ultra DLDspectrometer, using a monochromatic Al Ka source, operated at15 kV and 20 mA. Survey scan analyses were carried out with ananalysis area of 300 � 700 microns and pass energy of 160 eV.High-resolution analyses (performed on the energy regionstypical for F 1s, C 1s and Si 2p levels) were carried out with thesame analysis area and at pass energy of 10 eV. The Kratoscharge neutralizer system was used on all specimens. Spectrahave been charge corrected to the main line of the C 1s spec-trum (C–C bond) set to 284.8 eV and then analysed usingCasaXPS soware (version 2.3.16).

Fig. 1 Photograph of glass vials containing waterborne fluoroacrylicsolution (Capstone ST-110) and various concentrations of silicananoparticles.

Wear abrasion tests

Wear abrasion experiments were performed with a linearabrader (Taber Industries, USA) under 20.5 kPa applied pres-sure. Aer every 5 abrasion cycles CA and RA measurementswere taken during wear abrasion. The abradant surface ofchoice was a piece of crocking cloth (Taber abradant) and thespeed used for the mechanical abrasion was held constant at 15cycles per min. The abradant during one cycle covered adistance of 5.08 cm by performing a back and forth linearperiodical motion. In general, thicker samples were used inorder to maintain wear similarity as a result of material removaldue to wear. In particular, 0.4 g of particles were dispersed in 20ml of deionized water, as described earlier, and subsequently 4ml of Capstone ST-110 were added. The spray conditions werethe same with the only difference being that the coating thick-ness was increased by 20 spray passes. The resulting coatingthickness in this case was approximately 80 mm.

12882 | J. Mater. Chem. A, 2015, 3, 12880–12889

ASTM D3359 adhesion test

The coating adhesion to the substrate was tested by the ASTMD3359 standard. Cut marks on the coatings were made with ablade in a cross-cut shape (4 � 4 lines). The distance from oneline to the other was 3 mm. Subsequently a 3M 250 maskingtape (adhesion to steel: 85 N/100 mm) was placed at an angle of45� with the lines and pressed continuously on the surface (toavoid air pocket formation underneath). The tape was le incontact with the coating for 45 s in order for the adhesive tosettle over the coating, then the tape was removed by peelingand the sample was examined for damages.

Results and discussionImpact of deposition process

Two different types of hydrophilic (untreated) silicon dioxidenanoparticles were used in the present work, denoted bynanollers A and nanollers B (see Methods section for addi-tional information). The nanoparticles were dispersed in waterand mixed in different ratios with commercially availablewaterborne acrylic uoropolymer dispersion (WAF) known asCapstone® ST-110. Fig. 1 shows photographs of the polymer-nanoparticle dispersions in water with different concentrationof nanollers by weight.

It can be seen that the original slightly yellow and trans-parent colour of the uoropolymer dispersion with no nano-llers gradually turns to opaque white as nanoparticleconcentration increases. The coatings were directly applied onhot aluminium plates by spraying. During the spraying process,the substrates were placed on a hotplate maintained at aconstant temperature of 300 �C. Fig. 2 depicts optical images ofthe WAF coating morphology on aluminium with and withoutsubstrate heating. Upon substrate heating, very uniform trans-parent uoroacrylic coatings were obtained on the metalsubstrates. As shown in the schematic of Fig. 2, when thesubstrates were heated to 300 �C, spray generated water droplets

This journal is © The Royal Society of Chemistry 2015

Fig. 2 Waterborne fluoroacrylic polymer film sprayed on aluminium (a) without thermal and (b) with thermal treatment. Heated substratesubstantially improves the coating homogeneity. The schematics next to the photographs represent the two different mechanisms taking placeduring the drying of the coatings.

Paper Journal of Materials Chemistry A

Publ

ishe

d on

14

May

201

5. D

ownl

oade

d by

Uni

vers

ity o

f V

irgi

nia

on 2

0/10

/201

5 05

:32:

56.

View Article Online

evaporated immediately upon contact with the substrate leavinga homogeneous coating of organic/inorganic material deposit.The temperature of the substrates was maintained well abovethe melting point of the uoroacrylic polymer (�100 �C) inorder to avoid aking of the polymer once the water dropletsevaporated but rather partially or fully melt and conform to thesubstrate. Note that at 300 �C acrylic polymers are found at theirmolten glassy state.46 As such atomized drops containing thepolymer are expected to melt upon impact on the heated metalsurface. On the other hand, when the substrates were notheated during spray deposition, the slow evaporation of waterled to the formation of wet islands on the aluminium substratedue to severe de-wetting, which resulted in a non-uniform andirregular dry coating. Uniform coatings could also be obtainedwith lower temperatures. However, the spray distance and thepressure should be adjusted in order to obtain similar andconsistent surface roughness characteristics since the evapo-ration rate of the sprayed droplets and the viscosity of themolten polymer are different under varying temperature. As wasshown in a previous study, spray distance and pressure caninduce signicant changes in the coating's liquid-repellentcharacteristics and its durability as well.47 Since the aim of thepresent work is to develop non-wetting polymer compositecoatings for metals, a range of high temperatures could bemaintained to optimize the coating conditions. As such,substrate temperature of 300 �C was found to be ideal foruniform coating deposition.

Water and oil wettability results

Fluoroacrylic polymer sprayed onto heated aluminiumproduced a hydrophobic coating with water contact angle(WCA) of 113.0� 2.3� and oil contact angle of (OCA) 85.2� 1.4�.Interestingly, although the WCA is greater than the OCA, whenthe coatings were tilted the water droplets remained adhered onthe substrate even for 90� tilt angles while the oil droplets slid ata tilt angle of 14.3 � 1.2� indicating good self-cleaning proper-ties against non-polar liquids. Since the uoroacrylic polymer is

This journal is © The Royal Society of Chemistry 2015

already dispersed in water, it can be easily mixed with a separatewater dispersion of hydrophilic silica nanoparticles. Increasingthe amount of silica nanoparticles in the nal polymer/particledispersion inuences the nal wetting characteristics of thecoatings. Fig. 3 shows the water and oil contact anglesmeasurements (Fig. 3a) and droplet roll-off or sliding angles(Fig. 3b) against the weight percent of the nanoparticles. Fivedifferent concentrations were tested for both types of hydro-philic silica nanoparticles. Spray dispersions containing 32.5wt% nanoparticles was the threshold for the coatings to exhibitsuperhydrophobic and superoleophobic properties, with CAsfor water and oil greater than 150� and ROAs less than 10�.These concentrations also resulted in the formation of oilrepellent surfaces with OCA 159� and oil roll-off angle (OROA)down to 3�, particularly for the nanocomposites containingnanollers B. Nanollers A, however, produced a slightly lesssuperoleophobic coating with OCA 153� and OROA 9� with thesame concentration. The threshold of superhydrophobicity wasfound to be lower. Coatings loaded with 19 wt% nanollers wererendered superhydrophobic with WCA 161� and water roll-offangle (WROA) 3�. On the other hand, the superhydrophobicitythreshold with nanollers B was found to be 26.6 wt% andresulted in WCA and WROA 164� and WROA 1.7�. Althoughcoatings with these silica concentrations did not exhibitsuperoleophobicity, oil droplets maintained high nite contactangles on these surfaces indicating that at these nanoparticleconcentration levels coatings were oleophobic.

Impact of topology on wettability

Different nanoparticle loadings resulted in different surfaceroughness and the effect of surface topology on the non-wettingproperties was investigated next. Four different statisticalroughness parameters were calculated for all the differentconcentrations of nanoparticles studied. In particular, thearithmetic mean height (Sa), the root mean squared (Sq), theskewness (Ssk) and the kurtosis (Sku) values were measured. TheSa and Sq parameters are given by the following equations:

J. Mater. Chem. A, 2015, 3, 12880–12889 | 12883

Fig. 4 Surface roughness parameters plotted as a function of thenanofiller concentration by weight in the sprayed dispersions. Theresults for the two different types of particles are presented in bothgraphs. Sa (average mean height), Sq (root mean squared), Ssk (skew-ness) and Sku (kurtosis).

Fig. 3 Contact angles (a) and roll off angles (b) for oil and waterdroplets plotted against the total nanofiller concentration by weight inthe sprayed dispersions. The results for the two different types ofparticles are presented in both graphs.

Journal of Materials Chemistry A Paper

Publ

ishe

d on

14

May

201

5. D

ownl

oade

d by

Uni

vers

ity o

f V

irgi

nia

on 2

0/10

/201

5 05

:32:

56.

View Article Online

Sa ¼ 1

A

ð ðA

jzðx; yÞjdxdy (1)

Sq ¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi1

A

ð ðA

z2ðx; yÞdxdys

(2)

where A is the measured area and z the height of the surfacefeatures across the surface. The Sa and Sq parameters representan overall measure of the texture comprising the surface. Sa andSq are insensitive in differentiating peaks, valleys and thespacing of the various texture features. Once a surface type hasbeen established, the Sa and Sq parameters may be used toindicate signicant deviations in the texture characteristics.48 Sqrepresents the standard deviation of the distribution of surfaceheights, so it is also an important parameter to describe thesurface roughness. In Fig. 4a the Sa and the Sq values are plottedagainst the amount of nanoparticles in the coatings. As it can beseen, both Sa and Sq increase suddenly aer the nanoparticleconcentration exceeds 26.6 wt% in the dispersions. As perFig. 3, this concentration threshold also approximately corre-sponds to the increase in the oil droplet contact angles. Thesudden increase in the surface roughness indicates a wettingtransition to a Cassie–Baxter state by which the liquid droplet

12884 | J. Mater. Chem. A, 2015, 3, 12880–12889

can no longer penetrate into the surface asperities leading tovery low liquid adhesion at the surface.

Another parameter known as Ssk is used to measure thesymmetry of the surface height distribution. It is mathemati-cally dened as the third central moment of the prole height(z) distribution, measured over the assessment area.

Ssk ¼ 1

Sq3

" ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi1

A

ð ðA

z3ðx; yÞdxdys #

(3)

This parameter indicates a positive Ssk for a surface withpredominant high peaks. On the other hand, a surface withpredominant deep valleys will have negative Ssk value. A surfacewith equal presence of peaks and valleys will have a normaldistribution of heights with an Ssk of zero. If eqn (3) is modiedslightly with Sq

4, the surface kurtosis (Sku) can be also calcu-lated. As shown in eqn (4), it is the fourth central moment of theprole height (z) distribution, measured over the assessmentarea.

Sku ¼ 1

Sq4

" ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi1

A

ð ðA

z4ðx; yÞdxdys #

(4)

This journal is © The Royal Society of Chemistry 2015

Fig. 6 TEM images of the nanofillers A (a) and nanofillers B (b).

Paper Journal of Materials Chemistry A

Publ

ishe

d on

14

May

201

5. D

ownl

oade

d by

Uni

vers

ity o

f V

irgi

nia

on 2

0/10

/201

5 05

:32:

56.

View Article Online

A surface with inordinately high peaks and deep valleys willhave Sku > 3 while a gradually varying surface, free of extremepeaks or valleys will have Sku < 3. In Fig. 4b the Ssk and the Skuare plotted against the amount of nanoparticles in the drycoatings. From the values obtained it is observed that there is asteady increase in the surface peaks when nanoller Aconcentration is increased in the coatings. However, this is notthe case for the nanoller B. In this case, a gradual decrease inthe values of the skewness and kurtosis is observed aer 26.6wt% nanoparticles loading. Although the increase in nanollerconcentration for both types of particles promoted water and oilrepellence (Fig. 3), skewness and kurtosis do not follow asimilar trend. Interestingly, differences in surface topographywere also evident from the analysis of SEM images (Fig. 5) fortwo different types of nanoparticles with a concentration of 32.5wt% on dry basis. In Fig. 5, high (Fig. 5b and d) and lowmagnication (Fig. 5a and c) SEM micrographs of the twodifferent surfaces are presented. It is clear that in the case ofnanollers A (Fig. 5a and b), the nanoparticles are disperseduniformly within the polymer forming a sub-micrometer textureover the micro-scale roughness features. On the other hand,nanollers B (Fig. 5c and d) assemble into separated andagglomerated islands. TEM analysis of both nanoparticlesdeposited from water dispersions was also performed. Fig. 6aand b display TEM images of nanollers A and nanollers B,respectively. Silica nanoparticles designated as nanollers Aexhibit a closely packed structure, whereas, nanollers Bdemonstrate a chainlike structural agglomeration. This can beexpected since colloidal fumed silica nanoparticles are knownto form a range of such aggregates (closely packed, chainlike ora blend of both) depending on the solvent and temperatureused during deposition.49 As such, the colloidal state of fumed

Fig. 5 High (b and d) and low (a and c) magnification SEM images ofthe superhydrophobic/superoleophobic nanocomposites with nano-fillers A (a and b) and nanofillers B (c and d). Optical images of water (e)and oil (f) droplets on the surface (c and d).

This journal is © The Royal Society of Chemistry 2015

silica nanoparticles as well as high substrate temperaturesduring spray deposition strongly inuence the surface textureparameters such as skewness and kurtosis.

Surface chemistry

XPS measurements were performed on four different repre-sentative samples, namely nanollers B 26.6 wt%, nanollers B32.5 wt%, nanollers A 37.5 wt% and nanollers A 19.4 wt%. Insummary, for all the samples the Si 2p peak is found at thebinding energy of (103.3 � 0.2) eV (Fig. 7a), indicative of SiO2 asnanollers. The F 1s peak is at (689 � 0.2) eV for all the samples(Fig. 7b), a typical value reported in literature for uorinatedpolymers and indicative of CF2 and CF3 moieties. For all thesamples, the C 1s prole (Fig. 7c) is characterized by four peaks:a broad and intense one at approximately 285 eV and otherthree at higher binding energy values, in the range between 288and 296 eV. In order to identify interacting chemical speciesassociated with the observed proles, its deconvolution intocomponents was performed. As an example, for the nano-composite with 37.6 wt% nanollers A, the best t was obtainedusing six components as reported in Fig. 7c.

The same peak deconvolution calculation has been per-formed for all the other samples. In Table 1, these six compo-nents are listed, together with the chemical bond assignmentsand relative intensities. It is interesting to observe that the CF2/CF3 ratio is close to 5. The rst 4 components are those typicallyobserved when dealing with acrylic polymers, even withdifferent relative intensities (for reference, in pure PMMA weshould observe 40% component 1 and 20% each of the other 3components). The deviation with respect to pure PMMA stoi-chiometry could be assigned to the presence of adventitiouscarbon species on top of the sample (always present on samplesexposed to air). It is worth considering the two samplesprepared with nanoller A. The nanocomposite containing 19.4wt% nanollers A is obtained with a relatively low loading of thenanoller and it is characterized by relatively low oil contactangle and high oil roll-off angle; the nanocomposite containing37.6 wt%, nanollers A, on the other hand, is characterized bysuperoleophobicity. The difference in ller loading shouldreect in different Si concentrations between the two samples;moreover, we could expect that the Si concentration for samplenanollers A – 37.6 wt% to be approximately 2 times higherthan that of sample nanollers A – 19.4 wt%.

J. Mater. Chem. A, 2015, 3, 12880–12889 | 12885

Fig. 7 Results of XPS measurements from a superhydrophobicnanocomposite. (a) Si 2p profile; (b) F 1s profile and (c) C 1s profilealong with convolution and curve fit.

Table 1 The six components are listed, together with the chemicalbond assignments and relative intensities

ComponentPosition –BE (eV)

Chemicalassignment Rel. amount

1 284.8 C–C, C–H 28.6%2 285.7 C–C–O 18.8%3 286.9 C–O 9.3%4 288.8 COOH 8.8%5 291.5 C–F2 29.0%6 293.8 C–F3 5.5%

Table 2 Elemental composition for different coatings

Filler type – wt% F (at%) C (at%) Si (at%)

Nanollers B – 26.6 44.7 42.6 12.7Nanollers B – 32.5 45.6 44.2 10.2Nanollers A – 37.6 46.5 48.6 4.9Nanollers A – 19.4 49.8 49.3 0.9

Journal of Materials Chemistry A Paper

Publ

ishe

d on

14

May

201

5. D

ownl

oade

d by

Uni

vers

ity o

f V

irgi

nia

on 2

0/10

/201

5 05

:32:

56.

View Article Online

As seen in Table 2, this is not the case. The scale factorbetween the two concentrations is slightly higher than 5, thussuggesting that, when the ller loading is increased over acertain threshold, the silica particles tend to cluster more at thesurface of the nanocomposite lm. This means also that, atlower silica concentrations, the surface of the nanocomposite isricher in polymer than that of the lm obtained with highersilica concentrations.

To compare the nanocomposites obtained with the differentllers, consider the samples with nanollers B (32.5 wt%) and

12886 | J. Mater. Chem. A, 2015, 3, 12880–12889

nanollers A (37.6 wt%). It is clear from Table 2 that nano-composites containing nanoller B have silica-richer surface.This therefore indicates that, for the samples containingnanoller B, the silica particles are locatedmainly at the coatingsurface while for the samples containing nanoller A, the silicadistribution along the z-axis of the nanocomposites is morehomogeneous. The uorine and carbon atom percentages asshown in Table 2 are practically constant for all the samplestested. This indicates that regardless of the type of the nano-particle concentration, uorine functionality of the surface ismaintained which is particularly important for oil repellence.

Omniphobicity and wear abrasion resistance

To investigate omni-phobicity, different liquids includingcommon liquids used as food were tested in terms of theirwettability on the coatings and their wetting behaviour wascompared with the results of water and oil. In particular,ethylene glycol, hexadecane, vinegar, milk, coffee, hot sauce andorange juice were tested. Fig. 8 shows optical images of all thesedifferent liquids placed on the fabricated coatings. The bottompanel of Fig. 8 presents CAs and ROAs of these liquids. All theliquids have very high CAs and very low ROAs and they easilyrolled off the surfaces leaving liquid traces (self-cleaning; Video1†). Additionally, highly viscous substances like ketchup andhoney mustard were tested and these liquids were againrepelled by the coatings without leaving residues (Video 2†).

For the analysis of wear abrasion of the coatings, a linearmechanical abrasion test was performed. Note that in order tomaintain a good degree of abrasion resistance, the thickness ofthe coatings was increased as described in detail in the Methodssection. The abradant material was a crocking cloth and theapplied pressure on the surface was 20.5 kPa. The coatingthickness was adjusted such that enough thickness allowancewas given to the abrasion cycles in case the coating surfaces were

This journal is © The Royal Society of Chemistry 2015

Fig. 8 Left: photograph of various liquids on the nanocomposite surface. The table on the right displays static contact angles as well as dropletroll-off angles for every liquid tested. Surface tensions: water, 72.8 mNm�1; mineral oil, 28.6 mNm�1; ethylene glycol 47.7 mNm�1; hexadecane,27.5 mN m�1; milk, 41 mN m�1; orange juice, 32 mN m�1; vinegar, 36 mN m�1 and hot sauce, 33 mN m�1 (values at 20 �C).

Paper Journal of Materials Chemistry A

Publ

ishe

d on

14

May

201

5. D

ownl

oade

d by

Uni

vers

ity o

f V

irgi

nia

on 2

0/10

/201

5 05

:32:

56.

View Article Online

scraped away immediately aer abrasion started. The initialcoating thickness was 80 mm and the removed coating layer aereach abrasion cycle was approximately 1.3 mm thick. The valuesof the WCAs and water WROAs of the two different coatings areplotted against the number of abrasion cycles in Fig. 9a. Valuesfor oil droplets are not reported since the coatings lost theiroleophobicity with the rst abrasion cycle. Nanollers A resisted55 abrasion cycles maintaining water repellence whereas nano-llers B resisted 8 abrasion cycles as shown in Fig. 9a. As such,the coatings with nanollers A were consistently found to bemuch more resistant to wear abrasion induced super-hydrophobicity loss, possibly due to the fact that they main-tained betterwear similarity (similarity of exposed surface textureto non-abraded surface texture) and better dispersionthroughout the coating thickness as was conrmed by the XPSmeasurements. Uniform and self-similar dispersion of

Fig. 9 (a) Water contact angles and roll off angles plotted against thenumber of abrasion cycles. (b and c) SEM images of the coatings afterbeing damaged due to the wear abrasion. SEM image in (b) containssilica nanoparticles designated by nanofillers A and SEM image in (c)contains silica nanoparticles designated by nanofillers B.

This journal is © The Royal Society of Chemistry 2015

nanoparticles in the bulk of the polymer matrix can lead to apolymer nanocomposite in which the exposed sub-surface layers,as a result of abrasion, can still maintain a texture similar to thepristine surface texture, enough to cause water repellence. SEMimages from the damaged coatings right aer loss of super-hydrophobicity are shown in Fig. 9b and c. Note the exposure ofthe aluminium substrate as a result of coating removal.

In order to enhance substrate adhesion of the nano-composites to metal substrates, the substrates were pre-treatedwith an adhesive primer as was described earlier (see Methods).Aer the application of the primer, the nanocomposite wassprayed with the same conditions. The coatings exhibitedidentical topographical and liquid-repellent characteristics asdescribed before. An adhesion peel test was performedaccording to ASTM D3359 standard. Fig. 10 shows representa-tive photographs before and aer the tape peel test. The testallows qualitative evaluation of the coating adhesion to thesubstrate. Since the cross-cut squares were intact and undam-aged and were not lied off aer the tape peel action, ASTMD3359 classies the coatings with the top rating of 5B. As such,the coatings can be successfully applied to a wide selection ofmaterials that are compatible with such type of primers andalso can handle temperatures up to 300 �C.

Fig. 10 Photographs of knife-scratched nanocomposite before andafter the ASTMD3359 adhesion test. The coating did not show any signof adhesion failure (note that no peeling related damage is seen inevery square region), which is classified as 5B according to the ASTMstandard.

J. Mater. Chem. A, 2015, 3, 12880–12889 | 12887

Journal of Materials Chemistry A Paper

Publ

ishe

d on

14

May

201

5. D

ownl

oade

d by

Uni

vers

ity o

f V

irgi

nia

on 2

0/10

/201

5 05

:32:

56.

View Article Online

Conclusions

In summary, we have described a facile, environmetally friendly(water-based) and a cost-effective method to produce large-scalesuperhydrophobic and superoleophobic polymer–silica nano-composites by using hydrophilic silica nanoparticles. Thespraying process reported herein requires maintaining thesubstrate at elevated temperatures in order to improve thequality of the coatings and eliminate dewetting effects due towater high surface tension and high boiling point. The wettingproperties could be controlled by changing the polymer tonano-silica ratio in solution. Water and oil contact anglesgreater than 159� were achieved, while droplet roll-off angleswere as low as 3� and 2� for oil and water, respectively. Theseenvironmentally-friendly coatings were also able to repel a widevariety of other liquids. In addition, the effect of particle sizeand morphology was found to inuence the surface roughnesscharacteristics, wetting and the resultant wear abrasion resis-tance of the coatings. Although more research is needed inorder to improve mechanical abrasion resistance of thesenanocomposites, such liquid repellent coatings produced by anenvironmentally benign process can open up new avenues forthe use of sustainable and green technologies in the productionof super-liquid repellent coatings.

Acknowledgements

The authors would like to acknowledge nancial support fromthe National Science Foundation.

Notes and references

1 A. K. Mohanty, M. Misra and L. T. Drzal, J. Polym. Environ.,2002, 10, 19–26.

2 D. Prat, J. Hayler and A. Wells, Green Chem., 2014, 16, 4546–4551.

3 A. K. Kota, W. Choi and A. Tuteja, MRS Bull., 2013, 38, 383–390.

4 P. Roach, N. J. Shirtcliffe and M. I. Newton, SoMatter, 2008,4, 224–240.

5 A. Tuteja, W. Choi, M. Ma, J. M. Mabry, S. A. Mazzella,G. C. Rutledge, G. H. McKinley and R. E. Cohen, Science,2007, 318, 1618–1622.

6 R. Furstner, W. Barthlott, C. Neinhuis and P. Walzel,Langmuir, 2005, 21, 956–961.

7 P. Calcagnile, D. Fragouli, I. S. Bayer, G. C. Anyfantis,L. Martiradonna, P. D. Cozzoli, R. Cingolani andA. Athanassiou, ACS Nano, 2012, 6, 5413–5419.

8 L. Cao, A. K. Jones, V. K. Sikka, J. Wu and D. Gao, Langmuir,2009, 25, 12444–12448.

9 J. Genzer and K. Emenko, Biofouling, 2006, 22, 339–360.10 G. Wang, H. Wang and Z. Guo, Chem. Commun., 2013, 49,

7310–7312.11 B. Leng, Z. Shao, G. de With and W. Ming, Langmuir, 2009,

25, 2456–2460.12 J. Ou, B. Perot and J. P. Rothstein, Phys. Fluids, 2004, 16,

4635.

12888 | J. Mater. Chem. A, 2015, 3, 12880–12889

13 A. Milionis, R. Giannuzzi, I. S. Bayer, E. L. Papadopoulou,R. Ruffilli, M. Manca and A. Athanassiou, ACS Appl. Mater.Interfaces, 2013, 5, 7139–7145.

14 A. Milionis, D. Fragouli, L. Martiradonna, G. C. Anyfantis,P. D. Cozzoli, I. S. Bayer and A. Athanassiou, ACS Appl.Mater. Interfaces, 2014, 6, 1036–1043.

15 X. Zhang, J. Zhao, Q. Zhu, N. Chen, M. Zhang and Q. Pan,ACS Appl. Mater. Interfaces, 2011, 3, 2630–2636.

16 A. K. Kota, G. Kwon, W. Choi, J. M. Mabry and A. Tuteja, Nat.Commun., 2012, 3, 1025.

17 J. Yang, Z. Zhang, X. Xu, X. Zhu, X. Men and X. Zhou, J.Mater. Chem., 2012, 22, 2834–2837.

18 A. Milionis, R. Ruffilli and I. S. Bayer, RSC Adv., 2014, 4,34395–34404.

19 D. Prat, J. Hayler and A. Wells, Green Chem., 2014, 16, 4546–4551.

20 H. Teisala, M. Tuominen and J. Kuusipalo, Adv. Mater.Interfaces, 2014, 1, 1300026.

21 Z. Xue, Z. Sun, Y. Cao, Y. Chen, L. Tao, K. Li, L. Feng, Q. Fuand Y. Wei, RSC Adv., 2013, 3, 23432–23437.

22 S. T. Yohe, Y. L. Colson and M. W. Grinstaff, J. Am. Chem.Soc., 2012, 134, 2016–2019.

23 P. Wan, J. Wu, L. Tan, B. Zhang and K. Yang,Mater. Sci. Eng.,C, 2013, 33, 2885–2890.

24 S. Osawa, M. Yabe, M. Miyamura and K. Mizuno, Polymer,2006, 47, 3711–3714.

25 M. Liu, S. Wang, Z. Wei, Y. Song and L. Jiang, Adv. Mater.,2009, 21, 665–669.

26 Q. Cheng, M. Li, Y. Zheng, B. Su, S. Wang and L. Jiang, SoMatter, 2011, 7, 5948–5951.

27 H. Wang, H. Zhou, H. Niu, J. Zhang, Y. Du and T. Lin, Adv.Mater. Interfaces, 2015, 2, 1400506.

28 L. Xiong, L. L. Kendrick, H. Heusser, J. C. Webb, B. J. Sparks,J. T. Goetz, W. Guo, C. M. Stafford, M. D. Blanton,S. Nazarenko and D. L. Patton, ACS Appl. Mater. Interfaces,2014, 6, 10763–10774.

29 S. A. Mahadik, D. B. Mahadik, M. S. Kavale, V. G. Parale,P. B. Wagh, H. C. Barshilia, S. C. Gupta and A. V. Rao, J.Sol-Gel Sci. Technol., 2012, 63, 580–586.

30 A. Steele, I. Bayer and E. Loth, Nano Lett., 2009, 9, 501–505.31 P. Muthiah, B. Bushan, K. Yun and H. Kondo, J. Colloid

Interface Sci., 2013, 409, 227–236.32 Y. Lu, S. Sathasivam, J. Song, C. R. Crick, C. J. Carmalt and

I. V. Parkin, Science, 2015, 347, 1132–1135.33 M. Motornov, R. Sheparovych, R. Lupitskyy, E. MacWilliams

and S. Minko, Adv. Mater., 2008, 20, 200–205.34 J. E. Mates, T. M. Schutzius, I. S. Bayer, J. Qin,

D. E. Waldroup and C. M. Megaridis, Ind. Eng. Chem. Res.,2014, 53, 222–227.

35 T. M. Schutzius, I. S. Bayer, J. Qin, D. Waldroup andC. M. Megaridis, ACS Appl. Mater. Interfaces, 2013, 5,13419–13425.

36 T. M. Scutzius, I. S. Bayer, M. K. Tiwari and C. M. Megaridis,Ind. Eng. Chem. Res., 2011, 50, 11117–11123.

37 J. Z. Wang, Z. H. Zheng, H. W. Li, W. T. S. Huck andH. Sirringhaus, Nat. Mater., 2004, 3, 171–176.

This journal is © The Royal Society of Chemistry 2015

Paper Journal of Materials Chemistry A

Publ

ishe

d on

14

May

201

5. D

ownl

oade

d by

Uni

vers

ity o

f V

irgi

nia

on 2

0/10

/201

5 05

:32:

56.

View Article Online

38 Z. Yuan, J. Xiao, C. Wang, J. Zheng, S. Xing and J. Liu, J. Coat.Technol. Res., 2011, 8, 773–777.

39 X. Zhu, Z. Zhang, X. Xu, X. Men, J. Yang, X. Zhou and Q. Xue,J. Colloid Interface Sci., 2012, 367, 443–449.

40 S. Peng, X. Yang, D. Tian and W. Deng, ACS Appl. Mater.Interfaces, 2014, 6, 15188–15197.

41 A. Starostin, V. Valtsifer, V. Strelnikov, E. Bormashenko,R. Grynyov, Y. Bormashenko and A. Gladkikh, Adv. Eng.Mater., 2014, 16, 1127–1132.

42 S. Barthwal, Y. S. Kim and S. H. Lim, Langmuir, 2013, 29,11966–11974.

43 J. W. Martin, M. M. Smithwick, B. M. Braune, P. F. Hoekstra,D. C. G. Muir and S. A. Mabury, Environ. Sci. Technol., 2004,38, 373–380.

This journal is © The Royal Society of Chemistry 2015

44 F. Suja, B. K. Pramanik and S. M. Zain, Water Sci. Technol.,2009, 60, 1533–1544.

45 http://www2.dupont.com/Capstone/en_US/assets/downloads/DuPont_Capstone_Product_Stewardship_Detail_Document_072910.pdf.

46 M. C. Costache, D. Wang, M. J. Heidecker, E. Manias andC. A. Wilkie, Polym. Adv. Technol., 2006, 17, 272–280.

47 Y. H. Yeong, A. Davis, A. Steele, E. Loth and I. S. Bayer, Surf.Innovations, 2014, 2, 70–78.

48 E. S. Gadelmawla, M. M. Koura, T. M. A. Maksoud,I. M. Elewa and H. H. Soliman, J. Mater. Process. Technol.,2002, 123, 133–135.

49 X. J. Wu, Y. Wang, W. Yang, B. H. Xie, M. B. Yang andW. Dan, So Matter, 2012, 8, 10457.

J. Mater. Chem. A, 2015, 3, 12880–12889 | 12889