superhydrophobic hierarchical honeycomb surfaces

8
Superhydrophobic Hierarchical Honeycomb Surfaces P. S. Brown, E. L. Talbot, T. J. Wood, C. D. Bain, and J. P. S. Badyal* ,Department of Chemistry, Science Laboratories, Durham University, Durham DH1 3LE, U.K. ABSTRACT: Two-dimensional hexagonally ordered honeycomb surfaces have been created by solvent casting polybutadiene lms under controlled humidity. Subsequent CF 4 plasmachemical uorination introduces cross-linking and surface texturing, leading to hierarchical surfaces with roughness on both the 10 μm (honeycomb) and micrometer (texturing) length scales. For microliter droplets, these display high water contact angle values (>170°) in combination with low contact angle hysteresis (i.e., superhydrophobicity) while displaying bouncing of picoliter water droplets. In the case of picoliter droplets, it is found that surfaces which exhibit similar static contact angles can give rise to dierent droplet impact dynamics, governed by the underlying surface topography. These studies are of relevance to technological processes such as rapid cooling, delayed freezing, crop spraying, and inkjet printing. 1. INTRODUCTION Well-dened surface pore arrays are of signicant interest for numerous applications including proteomics, 1 tissue engineer- ing, 2 photonics, 3,4 sensors, 5,6 and catalysis. 7 One promising approach for their fabrication is to utilize breath gures 8,9 (which are two-dimensional hexagonally packed arrays of water droplets condensed onto a cooled surface) as a means for templating polymer lm surfaces. 10 This entails dissolving a polymer into a water immiscible, volatile solvent and then lm casting onto a surface under a controlled humid environment. Subsequent solvent evaporation gives rise to cooling of the solution surface, which culminates in water condensation 11,12 and the formation of a breath gure array of hexagonally ordered water droplets, 13 Scheme 1. The coalescence of these water droplets is avoided either by the occurrence of Marangoni convection or due to the precipitation of a polymer layer at the watersolvent interface. 14 Eectively, the water droplets serve as a template for the drying polymer solution, leading to the formation of a honeycomb-like surface structure following complete evaporation of the solvent and water. In the past, such honeycomb surfaces have predominantly been prepared using block copolymers 15 or branched polymers 16 because of their ability to more readily precipitate out at the solvent/water interface, and thereby negating undesired water droplet coalescence. 1720 A few linear homopolymers with high chain densities such as polyphenylene oxide 21 and polystyrene 2225 have also been shown to form stable breath gure arrays. However, the aforementioned polymer honeycomb systems typically have limited surface functionality as well as needing a separate cross-linking step (otherwise, the honeycomb structure can be unstable toward aging, aggressive solvents, or elevated temperatures). In the past, this has been addressed by chemical-, 2629 thermal-, 30 or photo-cross-linking, 3134 which typically entail complex or harsh processing conditions (e.g., intense irradiation or toxic chemicals). In this study, a much simpler and more straightforward approach is described comprising the solvent casting of a readily available and cheap polymer, polybutadiene, under controlled humidity. The resultant hexagonal honeycomb arrays are then simultaneously functionalized (uorinated), textured, and cross-linked via CF 4 plasma treatment to yield superhydrophobic surfaces (for microliter and picoliter droplets) that are both chemically and thermally stable. Received: July 6, 2012 Revised: August 14, 2012 Published: September 11, 2012 Scheme 1. Casting of a Polymer Dissolved in a Water Immiscible Solvent under Controlled Humidity a a Solvent evaporation leads to surface cooling and water condensation to form a hexagonal breath gure array, which acts as a template for the drying polymer solution. Article pubs.acs.org/Langmuir © 2012 American Chemical Society 13712 dx.doi.org/10.1021/la302719m | Langmuir 2012, 28, 1371213719

Upload: j-p-s

Post on 16-Oct-2016

216 views

Category:

Documents


3 download

TRANSCRIPT

Page 1: Superhydrophobic Hierarchical Honeycomb Surfaces

Superhydrophobic Hierarchical Honeycomb SurfacesP. S. Brown, E. L. Talbot, T. J. Wood, C. D. Bain,† and J. P. S. Badyal*,†

Department of Chemistry, Science Laboratories, Durham University, Durham DH1 3LE, U.K.

ABSTRACT: Two-dimensional hexagonally ordered honeycombsurfaces have been created by solvent casting polybutadiene filmsunder controlled humidity. Subsequent CF4 plasmachemicalfluorination introduces cross-linking and surface texturing, leadingto hierarchical surfaces with roughness on both the 10 μm(honeycomb) and micrometer (texturing) length scales. Formicroliter droplets, these display high water contact angle values(>170°) in combination with low contact angle hysteresis (i.e.,superhydrophobicity) while displaying bouncing of picoliter waterdroplets. In the case of picoliter droplets, it is found that surfaceswhich exhibit similar static contact angles can give rise to differentdroplet impact dynamics, governed by the underlying surface topography. These studies are of relevance to technologicalprocesses such as rapid cooling, delayed freezing, crop spraying, and inkjet printing.

1. INTRODUCTIONWell-defined surface pore arrays are of significant interest fornumerous applications including proteomics,1 tissue engineer-ing,2 photonics,3,4 sensors,5,6 and catalysis.7 One promisingapproach for their fabrication is to utilize breath figures8,9

(which are two-dimensional hexagonally packed arrays of waterdroplets condensed onto a cooled surface) as a means fortemplating polymer film surfaces.10 This entails dissolving apolymer into a water immiscible, volatile solvent and then filmcasting onto a surface under a controlled humid environment.Subsequent solvent evaporation gives rise to cooling of thesolution surface, which culminates in water condensation11,12

and the formation of a breath figure array of hexagonallyordered water droplets,13 Scheme 1. The coalescence of thesewater droplets is avoided either by the occurrence ofMarangoni convection or due to the precipitation of a polymer

layer at the water−solvent interface.14 Effectively, the waterdroplets serve as a template for the drying polymer solution,leading to the formation of a honeycomb-like surface structurefollowing complete evaporation of the solvent and water.In the past, such honeycomb surfaces have predominantly

been prepared using block copolymers15 or branchedpolymers16 because of their ability to more readily precipitateout at the solvent/water interface, and thereby negatingundesired water droplet coalescence.17−20 A few linearhomopolymers with high chain densities such as polyphenyleneoxide21 and polystyrene22−25 have also been shown to formstable breath figure arrays.However, the aforementioned polymer honeycomb systems

typically have limited surface functionality as well as needing aseparate cross-linking step (otherwise, the honeycomb structurecan be unstable toward aging, aggressive solvents, or elevatedtemperatures). In the past, this has been addressed bychemical-,26−29 thermal-,30 or photo-cross-linking,31−34 whichtypically entail complex or harsh processing conditions (e.g.,intense irradiation or toxic chemicals).In this study, a much simpler and more straightforward

approach is described comprising the solvent casting of areadily available and cheap polymer, polybutadiene, undercontrolled humidity. The resultant hexagonal honeycombarrays are then simultaneously functionalized (fluorinated),textured, and cross-linked via CF4 plasma treatment to yieldsuperhydrophobic surfaces (for microliter and picoliterdroplets) that are both chemically and thermally stable.

Received: July 6, 2012Revised: August 14, 2012Published: September 11, 2012

Scheme 1. Casting of a Polymer Dissolved in a WaterImmiscible Solvent under Controlled Humiditya

aSolvent evaporation leads to surface cooling and water condensationto form a hexagonal breath figure array, which acts as a template forthe drying polymer solution.

Article

pubs.acs.org/Langmuir

© 2012 American Chemical Society 13712 dx.doi.org/10.1021/la302719m | Langmuir 2012, 28, 13712−13719

Page 2: Superhydrophobic Hierarchical Honeycomb Surfaces

2. EXPERIMENTAL SECTION2.1. Sample Preparation. Control sample preparation comprised

spin coating polybutadiene (Sigma-Aldrich Inc.,Mw = 420 000, 36% cis1,4 addition, 55% trans 1,4 addition, 9% 1,2 addition) dissolved intoluene (BDH, +99.5% purity) at a concentration of 5% (w/v) ontopolished silicon (100) wafers (Silicon Valley Microelectronics, Inc.)using a photoresist spinner (Cammax Precima) operating at 3000 rpm.These polymer films were then annealed at 90 °C under vacuum for60 min.For the honeycomb surfaces, polybutadiene dissolved at varying

concentrations in dichloromethane (Fisher Scientific, +99.9% purity)was cast onto clean glass slides (Smith Scientific Ltd.) undercontrolled humidity conditions. This entailed placing the glasssubstrate onto a wire mesh so that it was suspended above a saturatedsalt solution contained within a 25 mL glass bottle fitted with a rubberseptum. The salt solutions used were magnesium chloride (Sigma-Aldrich Inc., +98%), potassium carbonate (Sigma-Aldrich Inc., +99%),magnesium nitrate (Sigma-Aldrich Inc., +99%), sodium bromide(Sigma-Aldrich Inc., +99.5%), strontium chloride (Sigma-Aldrich Inc.,+99%), sodium chloride (Sigma-Aldrich Inc., +99.5%), and potassiumchloride (Sigma-Aldrich Inc., +99.5%) which gave relative humiditiesof 33, 43, 54, 59, 73, 76, and 85%, respectively.35 For each film, 0.1 mLof polymer solution was deposited onto the glass slide using amicrosyringe. Subsequently, these polymer films were annealed at 90°C under vacuum for 60 min.Plasmachemical fluorination, texturing, and cross-linking of the

polybutadiene films were undertaken in a cylindrical glass chamber (5cm diameter, 470 cm3 volume) connected to a two-stage rotary pumpvia a liquid nitrogen cold trap (4 × 10−3 mbar base pressure and a leakrate better than 6 × 10−9 mol s−1). An L-C matching unit was used tomaximize power transmission between a 13.56 MHz radio frequencygenerator and a copper coil externally wound around the glass reactor.Prior to each plasma treatment, the chamber was scrubbed withdetergent, rinsed in propan-2-ol, and further cleaned using a 50 W airplasma for 30 min. Next, a piece of polybutadiene coated substrate wasplaced into the reactor at ambient temperature (either in the glowregion for textured surfaces or 8 cm downstream for smooth surfaces),followed by evacuation to base pressure. CF4 gas (+99.7% purity, AirProducts) was then admitted into the system via a needle valve at apressure of 0.2 mbar and 2 cm3 min−1 flow rate, and the electricaldischarge ignited at a power of 30 W for 5 min duration for texturedsurfaces or 60 s for smooth surfaces. Upon completion of surfacefunctionalization (and texturing), the gas feed was switched off and thechamber vented to the atmosphere.2.2. Surface Characterization. The obtained honeycomb surfaces

were visually examined using an optical microscope (Olympus BX40)fitted with a digital camera and a Euromax fiber optic light source. Poresize distribution, surface coverage, and lattice parameters weremeasured using image analysis software (ImageJ, public domain,http://rsbweb.nih.gov/ij/).A VG ESCALAB spectrometer equipped with an unmonochrom-

atized Mg Kα X-ray source (1253.6 eV) and a concentrichemispherical analyzer (CAE mode pass energy = 20 eV) was usedfor X-ray photoelectron spectroscopy (XPS). The C(1s) XPS spectrawere referenced to the (−CxHy) hydrocarbon peak at 285.0 eV andfitted to a linear background and equal full-width-at-half-maximum(fwhm) Gaussian components using a Marquardt optimizationalgorithm.36 Elemental compositions were calculated using sensitivityfactors derived from chemical standards, C(1s):O(1s):F(1s) =1.00:0.34:0.26.AFM images were acquired using a Digital Instruments Nanoscope

III scanning probe microscope. Damage to the tip and sample surfacewas minimized by employing tapping mode AFM. Root-mean-square(rms) roughness values were calculated over 50 μm × 50 μm scanareas.Microliter sessile drop contact angle analysis was carried out with a

video capture system (VCA2500XE, AST Products Inc.) using a 1.0μL dispensation of deionized water. Advancing and receding contact

angle values were determined by respectively increasing or decreasingthe liquid drop volume by a further 1.0 μL.37

Picoliter drop impact studies were carried out using a piezo-typenozzle (MicroFab MJ-ABP-01, Horizon Instruments Ltd.) with anaperture diameter of 50 μm. Water drops of 50 μm diameter (65 pL)were generated by using a drive voltage of 30 V and a pulse waveformconsisting of a rise time of 13 μs, a dwell width of 13 μs, a fall time of38 μs, an echo of 30 μs, and a final rise time of 13 μs. The separationbetween the nozzle tip and the substrate surface was set at 0.4 mm.Typical impact speeds were measured to be in the range 0.8−1.2 ms−1. The nozzle temperature was maintained at 30 °C. A high-speedcamera (FASTCAM APX RS, Photron Europe Ltd.) in conjunctionwith a 20× magnification microscopic objective lens (M Plan, NikonU.K. Ltd.) and a backlighting system (HPLS-30-02, Thorlabs Ltd.)were used to observe the droplet impact. By using 90 000 frames persecond, an image every 11 μs was obtained with the shutter speed setto 2 μs. Individual frames consisted of 128 × 96 pixels with 0.73 μmpixel size. The jetting driver was triggered by the camera.

3. RESULTS3.1. Honeycomb Formation. Polybutadiene honeycomb

surfaces were created across a range of relative humidities(RH), Figure 1. It was found that the average pore diameter

increased and pore density decreased with rising humidity,Figure 2. As previously reported for other polymer systems,honeycombs did not form at 100% RH23 or below 40% RH.38

The dimensions of the honeycomb arrays could also bevaried by changing the concentration of the polybutadienesolution, Figure 3. The polymer solution concentration hadlittle effect on the average pore diameter at constant humidity,while the flat polymer bridging regions in between the poresincreased in width with rising polymer concentration (albeitstill dilute),39 leading to a concurrent decrease in overall poresurface area and a corresponding drop in average surface poredensity, Figures 3 and 4.The breath figure templating process produces an approx-

imate hexagonally ordered two-dimensional array of surfacepores. The lattice parameter of these hexagonal arrays increasedwith both RH and concentration, Figure 5.

3.2. Plasmachemical Fluorination and Surface Textur-ing. XPS analysis of the polybutadiene honeycomb surfaces

Figure 1. Optical microscope images of honeycomb surfaces cast from1% w/v polybutadiene solution dissolved in dichloromethane undercontrolled RH of (a) 43%, (b) 54%, (c) 73%, and (d) 85%. The poresincrease in size with rising RH. Scale bar = 50 μm.

Langmuir Article

dx.doi.org/10.1021/la302719m | Langmuir 2012, 28, 13712−1371913713

Page 3: Superhydrophobic Hierarchical Honeycomb Surfaces

confirmed complete coverage of the glass slides with no Si(2p)signal detected from the underlying substrate, Table 1. The

measured oxygen signal can be attributed to aerobic oxidationlocalized at the outer surface of the polymer film duringannealing,40 which disappears upon subsequent CF4 plasmafluorination.41

At room temperature, the as-prepared honeycomb structuresgradually disappeared over a period of 48 h due to polymerchain relaxation,42 whereas a short exposure to the CF4 plasmawas sufficient to lead to VUV-assisted subsurface cross-linking40,43,44 so as to stabilize the honeycomb structure,Figure 6.Surface texture could be varied by altering the location of the

polybutadiene substrate within the CF4 plasma, Figure 7. Arougher surface was observed for the plasma glow region (dueto ion bombardment45,46) as compared to the downstreamregion (absence of ion bombardment45,46), without anynoticeable difference in surface chemistry as verified by XPSanalysis, Table 1.

3.3. Water Droplet Impact. For sessile drops withmicroliter volumes, the water contact angles increaseapproximately linearly with average pore diameter for thesolvent cast polybutadiene honeycomb surfaces, Figure 8. CF4plasma smooth (30 W, 60 s, downstream) honeycomb surfacesdisplayed enhanced hydrophobicity, with water contact anglesrising to 172° for average pore diameters exceeding 20 μm,Figure 8. However, there remains significant contact anglehysteresis. Whereas for the case of CF4 plasma textured (30 W,

Figure 2. Average pore diameter and average pore density in polymerfilms cast from 1% w/v polybutadiene dissolved in dichloromethane asa function of controlled RH.

Figure 3. Optical microscope images of honeycomb surfaces formedunder 54% RH from polybutadiene dissolved in dichloromethane withconcentrations of (a) 0.5% w/v, (b) 1% w/v, (c) 2% w/v, and (d) 3%w/v. The raised plateaus (lighter regions) encircling the pores (darkerareas) expand in width with increasing polymer concentration. Scalebar = 25 μm.

Figure 4. Overall surface coverage of honeycomb pores as a functionof polybutadiene concentration (dissolved in dichloromethane).Samples were prepared under 54% RH. (The average pore size acrossthe polymer concentrations shown remains constant within the range12−14 μm.)

Figure 5. Lattice parameter of the hexagonally ordered two-dimensional honeycomb array as a function of (a) RH (1% w/vpolybutadiene concentration) and (b) polybutadiene concentration(54% RH).

Table 1. XPS Elemental Compositions for PolybutadieneHoneycomb Surfaces: (a) Untreated; (b) CF4 PlasmaSmooth (30 W, 60 s, Downstream); (c) CF4 PlasmaTextured (30 W, 5 min, Glow)

XPS elemental composition (±0.5%)

honeycomb polybutadiene % C % F % O

(a) untreated theoretical 100.0 0.0 0.0experimental 87.8 0.0 12.2

(b) CF4 plasma fluorinated(smooth)

experimental 40.9 57.1 2.0

(c) CF4 plasma fluorinated(textured)

experimental 41.1 58.9 1.6

Langmuir Article

dx.doi.org/10.1021/la302719m | Langmuir 2012, 28, 13712−1371913714

Page 4: Superhydrophobic Hierarchical Honeycomb Surfaces

5 min, glow) honeycomb surfaces, the very low contact anglehysteresis values are indicative of superhydrophobicity47

(especially for pore sizes exceeding 12 μm), Table 2. This isconsistent with CF4 plasma textured flat polybutadiene surfaces(i.e., in the absence of pores) being sufficient to yield highwater contact angle and low contact angle hysteresis values.41 Itshould be noted that an increase in the pore diameter does leadto a slight decrease in contact angle hysteresis value due tothere being a greater amount of air trapped within the largerpores combined with a more irregular contact line.Previous work has shown that the wetting of picoliter drops

(the size delivered by modern inkjet printers) on plasmafluorinated surfaces can be quite different from microliterdrops41 and consequently that the impact and spreading ofpicoliter drops cannot be extrapolated from studies on themicroliter scale. When a picoliter droplet strikes these(super)hydrophobic surfaces, the liquid first spreads outwardto a maximum diameter and then oscillates about its staticposition until the excess energy is lost by viscous dissipation.The amplitude and decay of the oscillations can be observed inthe height or width of the drop or in the diameter of thecontact line,48 Figures 9 and 10.This oscillatory motion of picoliter water droplets during

impact was compared for the four different types of plasmafluorinated polybutadiene surfaces: smooth, textured, smooth

honeycomb, and textured honeycomb, Table 2. The pore sizeof the honeycomb (12−14 μm) was chosen to be comparableto, but smaller than, the diameter of the water droplet (50 μm).The CF4 plasma parameters for textured surfaces were chosenso as to give similar picoliter contact angles to those found forthe smooth honeycomb surface. No oscillation was observed onthe CF4 plasma fluorinated smooth spin coated sample, as allthe excess surface free energy was dissipated during the initialspreading of the contact line. Water droplets impacting uponthe textured spin coated polybutadiene oscillated at a higherfrequency compared to those on the smooth honeycombsurfaces despite both surfaces exhibiting similar static picolitercontact angles and the same mode of oscillation (movingcontact line), Table 2. Picoliter droplets striking the texturedhoneycomb surfaces bounced straight off in all cases (for 5−30μm pore size range in the present study).The relative size of the droplet and the pores is important in

determining the impact dynamics. For the 12−14 μm diameterpores and 50 μm diameter droplets described above,reproducible oscillations and high contact angles (126°) wereobserved on the smooth honeycomb surfaces. For pores <10μm in diameter, the droplets spread out to lower contact angleswith no noticeable oscillations. For pores ≥20 μm in diameter,the impact behavior was variable and dependent on the locationof the drop impact in relation to the surrounding pores;droplets bounced more frequently on these surfaces. A detailed

Figure 6. Optical microscope images of the honeycomb polybutadienesurface: (a) solvent cast; (b) after storage for 48 h; (c) CF4 plasmatreatment of (a) (30 W, 60 s, downstream); and (d) following storageof (c) for 48 h. Scale bar = 50 μm.

Figure 7. AFM height images of flat spin coated polybutadiene: (a) untreated; (b) CF4 plasma smooth (30 W, 60 s, downstream); and (c) CF4plasma textured (30 W, 5 min, glow).

Figure 8. Microliter sessile drop (a) static water contact angle and (b)contact angle hysteresis as a function of average pore diameter foruntreated and CF4 plasma treated smooth (30 W, 60 s, downstream)honeycomb surfaces. Polymer films were cast from 1% w/vpolybutadiene in dichloromethane (by variation in RH as describedin Figure 2). Dashed lines are added for guides to the eye.

Langmuir Article

dx.doi.org/10.1021/la302719m | Langmuir 2012, 28, 13712−1371913715

Page 5: Superhydrophobic Hierarchical Honeycomb Surfaces

study on the effect of droplet size, feature size, and impactvelocity on droplet dynamics is underway and will be publishedsubsequently.

4. DISCUSSIONThe pore sizes of these polybutadiene honeycomb arrays (up to30 μm) are significantly larger than those previously reportedfor breath figure templating (typically 0.2−10 μm).10 This

difference may be attributed to the lower chain density ofpolybutadiene,49 which leads to slower solidification around thewater droplets and therefore increased droplet coalescence.14

The polybutadiene solution concentration (and thereforeviscosity) had little effect on this precipitation behavior(average pore diameters remain similar) across the rangestudied, Figure 4. In contrast, earlier studies on other polymersystems showed that higher polymer concentrations stabilizedsmaller droplet arrays;14 again, the lower chain density ofpolybutadiene might be a plausible contributing factor.A key drawback encountered in prior studies of breath figure

templating is the instability of the honeycomb structures in thepresence of solvents or at elevated temperatures.38 Indeed, thehoneycomb polybutadiene surfaces formed in the present studyare also seen to completely disappear at room temperature in48 h, Figure 6. Stabilization of these honeycomb surfaces iseasily accomplished by CF4 plasmachemical subsurface cross-linking while concurrently lowering the surface energy viasurface fluorination.40 This should be contrasted with sulfurmonochloride vulcanization which is commonly employed forother honeycomb systems and suffers from drawbacks such aschemical entrapment and prolonged cross-linking times(typically 5 h).28,29 Similar honeycomb surfaces can also beprepared using polyisoprene by the same methodology.Plasmachemical surface fluorination of polybutadiene follows

earlier predicted structure−behavior relationships.40,41,50 Theextent of plasma-induced surface roughening (texturing) can bedecoupled from plasma fluorination by placement of thepolybutadiene surfaces either in the electrical discharge glowregion (plasma sheath bombardment45,46) or downstream (noion bombardment,45,46 thus smooth), Figure 7. The honey-comb structures, combined with nontexturing CF4 plasmafluorination (30 W, 60 s, downstream), lead to an increase inhydrophobicity as observed by placing microliter water dropletsonto the surface, Figure 8. An average pore diameter of at least20 μm is required to achieve contact angles greater than 170°.Smaller pores may be too shallow (assuming constantinterfacial behavior during pore formation51) to be capable oftrapping air,52−56 which is key to achieving a Cassie−Baxterstate,57 and therefore lead to lower contact angles moreindicative of a Wenzel state of wetting.58 Such smoothhoneycomb samples also exhibit high contact angle hysteresis,Figure 8. This can be lowered if the polybutadiene samples areplaced in the CF4 plasma glow region (which generates atextured surface), which is consistent with the rationale thathierarchical surfaces (two length scales of roughness, in thiscase honeycomb structure plus plasma-induced surface rough-

Table 2. Comparison of Microliter and Picoliter Water Droplet Behavior on CF4 Plasma Treated Spin Coated versusHoneycomb Polybutadiene Surfaces (1% w/v Concentration in Dichloromethane, 54% RH, pore size = 12−14 μm), WhereSmooth Corresponds to (30 W, 60 s, Downstream) and Textured to (30 W, 5 min, Glow)

microliter picoliter

polybutadieneCF4 plasmatreatment

static contact angle(±5°)

contact anglehysteresis (±1°)

static contact angle(±5°)

contact anglehysteresis (±1°)

impact oscillation frequency(±0.5 kHz)

spin coated (flat,control)

smooth 110 37 71 44 spreading

textured >170 4a 126b 5 22.8honeycombc smooth 150 23 126b 28 17.9

textured >170 1 droplet bouncingaLower microliter contact angle hysteresis values (<1°) can be achieved be employing different CF4 plasma conditions (30 W, 10 min, glow).41 bTheCF4 plasma parameters for textured surfaces were chosen so as to give similar picoliter contact angles to those found for the CF4 plasma treatedsmooth honeycomb surface. cA 12−14 μm pore size was selected because it is sufficiently smaller than the diameter of the picoliter droplets (50 μm).

Figure 9. Typical high-speed video images of picoliter water dropletimpact upon a superhydrophobic CF4 plasma fluorinated polybuta-diene surface (showing lower reflection as well). White scale bar = 20μm.

Figure 10. Typical damped oscillating curve fitted to the experimentaldata for picoliter water droplet fluctuation following impact.

Langmuir Article

dx.doi.org/10.1021/la302719m | Langmuir 2012, 28, 13712−1371913716

Page 6: Superhydrophobic Hierarchical Honeycomb Surfaces

ening) can lead to true superhydrophobicity (high contactangles, low hysteresis).41,59,60 Both of these observations aresupported by theoretical studies, which have predicted largetrough widths (pore widths) as well as two length scales ofsurface texturing to be beneficial to superhydrophobicity (highcontact angles and low contact angle hysteresis, respec-tively).54,55,61

For the case of picoliter droplet impact on smooth plasmafluorinated polybutadiene, no oscillation was observed due toincreased movement of the contact line during spreading,leading to an increase in the dissipation of the excess surfacefree energy of the droplet. Droplets impacting upon hierarchicalplasma-textured honeycomb surfaces bounced due to the highcontact angles and low hysteresis observed on these surfaces.A comparison between CF4 plasma fluorinated smooth

honeycomb surfaces and CF4 plasma fluorinated and texturedspin coated polybutadiene films (with identical picoliter staticcontact angle values) shows that the picoliter droplet impactbehavior onto these two surfaces is markedly different. Thedroplet oscillation frequency is found to be much higher for thelatter, Table 2. This is in disagreement with previous theoreticalmodels, which suggest that droplets with similar contact anglesshould oscillate at similar frequencies.62,63 This discrepancymay be due to the fact that the static contact angle is not anappropriate contact angle to use when predicting oscillationfrequencies, and that contact angle hysteresis and the motion ofthe contact line should be taken into account.These CF4 plasma fluorinated polybutadiene honeycomb

surfaces provide a quick and easy route to stable hierarchicalsuperhydrophobicity. They are more superhydrophobic thanconventional honeycomb arrays prepared from fluorinatedpolymers.64 Also, shorter CF4 plasma exposure times arerequired to achieve superhydrophobicity compared to thetreatment of flat polybutadiene40,41 due to the trapping of airwithin the pore structures giving rise to the creation of acomposite interface. In addition, by utilizing a solvent with alower density than water, it should be feasible to create a 3Dhoneycomb structure,22 resulting in a highly porous polymerlayer. Such low energy surface porous layers could find use inconfined crystallization,65 transportation,66 templating,67 orhigh surface area scaffolds.68 Furthermore, droplet impactonto these hydrophobic surfaces is of relevance to technologicalprocesses including rapid cooling,69−71 delayed freezing,72−75

crop spraying,76 and inkjet printing (for microelectronics,77−80

pharmaceutical dosing or screening,81−83 tissue engineer-ing,84,85 and optics86,87).

5. CONCLUSIONS

Solvent casting of linear polybutadiene under controlledhumidity gives rise to the formation of two-dimensionalhexagonally ordered honeycomb arrays. Pore aperture sizeand surface coverage can be independently controlled byvarying the humidity and polymer concentration, respectively.CF4 plasmachemical modification imparts low surface energyfunctional groups in combination with surface texturing andsubsurface cross-linking of the honeycomb structures to yieldsuperhydrophobicity (high contact angles and low hysteresis formicroliter droplets and bouncing for picoliter droplets).

■ AUTHOR INFORMATION

Corresponding Author*E-mail: [email protected].

Author Contributions†J.P.S.B. and C.D.B. have made equal contributions to thiswork.

NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTS

We thank the Engineering and Physical Sciences ResearchCouncil (EPSRC) for financial support (Grant Reference EP/H018913/1).

■ REFERENCES(1) Nishikawa, T.; Nishida, J.; Ookura, R.; Nishimura, S.-I.; Wada, S.;Karino, T.; Shimomura, M. Mesoscopic patterning of cell adhesivesubstrates as novel biofunctional interfaces. Mater. Sci. Eng., C 1999,10, 141.(2) Shastri, V. P.; Martin, I.; Langer, R. Macroporous polymer foamsby hydrocarbon templating. Proc. Natl. Acad. Sci. U.S.A. 2000, 97,1970.(3) Wijnhoven, J. E. G. J.; Vos, W. L. Preparation of photonic crystalsmade of air spheres in titania. Science 1998, 281, 802.(4) Imada, M.; Noda, S.; Chutinan, A.; Tokuda, T.; Murata, M.;Sasaki, G. Coherent two-dimensional lasing action in surface-emittinglaser with triangular-lattice photonic crystal structure. Appl. Phys. Lett.1999, 75, 316.(5) Perez, J. M.; O’Loughin, T.; Simeone, F. J.; Weissleder, R.;Josephson, L. DNA-based magnetic nanoparticle assembly acts as amagnetic relaxation nanoswitch allowing screening of DNA-cleavingagents. J. Am. Chem. Soc. 2002, 124, 2856.(6) Nam, J.-M.; Thaxton, C. S.; Mirkin, C. A. Nanoparticle-based bio-bar codes for the ultrasensitive detection of proteins. Science 2003, 301,1884.(7) Boker, A.; Lin, Y.; Chiapperini, K.; Horowitz, R.; Thompson, M.;Carreon, V.; Xu, T.; Abetz, C.; Skaff, H.; Dinsmore, A. D.; Emrick, T.;Russell, T. P. Hierarchical nanoparticle assemblies formed bydecorating breath figures. Nat. Mater. 2004, 3, 302.(8) Rayleigh, L. Breath figures. Nature 1911, 86, 416.(9) Rayleigh, L. Breath figures. Nature 1912, 90, 436.(10) Widawski, G.; Rawiso, M.; Francois, B. Self-organizedhoneycomb morphology of star-polymer polystyrene film. Nature1994, 369, 387.(11) Beysens, D; Knobler, C. M. Growth of breath figures. Phys. Rev.Lett. 1986, 57, 1433.(12) Reiss, H. The growth of uniform colloidal dispersions. J. Chem.Phys. 1951, 19, 482.(13) Gray, J. J.; Klein, D. H.; Bonnecaze, R. T.; Korgel, B. A.Nonequilibrium phase behavior during the random sequentialadsorption of tethered hard disks. Phys. Rev. Lett. 2000, 85, 4430.(14) Stenzel, M. H.; Barner-Kowollik, C.; Davis, T. P. Formation ofhoneycomb-structured, porous films via breath figures with differentpolymer architectures. J. Polym. Sci., Part A: Polym. Chem. 2006, 44,2363.(15) Karthaus, O.; Maruyama, N.; Cieren, X.; Shimomura, M.;Hasegawa, H.; Hashimoto, T. Water-assisted formation of micrometer-size honeycomb patterns of polymers. Langmuir 2000, 16, 6071.(16) Francois, B.; Ederle, Y.; Mathis, C. Honeycomb membranesmade from C60(PS)6. Synth. Met. 1999, 103, 2362.(17) Pitois, O.; Francois, B. Formation of ordered micro-porousmembranes. Eur. Phys. J. B 1999, 8, 225.(18) Pitois, O.; Francois, B. Crystallization of condensation dropletson a liquid surface. Colloid Polym. Sci. 1999, 277, 574.(19) Stenzel-Rosenbaum, M. H.; Davis, T. P.; Fane, A. G.; Chen, V.Porous polymer films and honeycomb structures made by the self-organization of well-defined macromolecular structures created byliving radical polymerization techniques. Angew. Chem., Int. Ed. 2001,40, 3428.

Langmuir Article

dx.doi.org/10.1021/la302719m | Langmuir 2012, 28, 13712−1371913717

Page 7: Superhydrophobic Hierarchical Honeycomb Surfaces

(20) Hernandez-Guerrero, M.; Davis, T. P.; Barner-Kowollik, C.;Stenzel, M. H. Polystyrene comb polymers built on cellulose orpoly(styrene-co-2-hydroxyethylmethacrylate) backbones as substratesfor the preparation of structured honeycomb films. Eur. Polym. J. 2005,41, 2264.(21) Tian, Y.; Jiao, Q.; Ding, H.; Shi, Y.; Liu, B. The formation ofhoneycomb structure in polyphenylene oxide films. Polymer 2006, 47,3866.(22) Srinivasarao, M.; Collings, D.; Philips, A.; Patel, S. Three-dimensionally ordered array of air bubbles in a polymer film. Science2001, 292, 79.(23) Peng, J.; Han, Y.; Yang, Y.; Li, B. The influencing factors on themacroporous formation in polymer films by water droplet templating.Polymer 2004, 45, 447.(24) Ferrari, E.; Fabbri, P.; Pilati, F. Solvent and substratecontributions to the formation of breath figure patterns in polystyrenefilms. Langmuir 2011, 27, 1874.(25) Karthaus, O.; Cieren, X.; Maruyama, N.; Shimomura, M.Mesoscopic 2-D ordering of inorganic/organic hybrid materials. Mater.Sci. Eng., C 1999, 10, 103.(26) Yabu, H.; Tanaka, M.; Ijiro, K.; Shimomura, M. Preparation ofhoneycomb-patterned polyimide films by self-organization. Langmuir2003, 19, 6297.(27) Kabuto, T.; Hashimoto, Y.; Karthaus, O. Thermally stable andsolvent resistant mesoporous honeycomb films from a crosslinkablepolymer. Adv. Funct. Mater. 2007, 17, 3569.(28) Li, L.; Zhong, Y.; Gong, J.; Li, J.; Chen, C.; Zheng, B.; Ma, Z.Constructing robust 3-dimensionally conformal micropatterns: vulcan-ization of honeycomb structured polymeric films. Soft Matter 2011, 7,546.(29) Li, L.; Zhong, Y.; Gong, J.; Li, J.; Huang, J.; Ma, Z. Fabricationof robust micro-patterned polymeric films via static breath-figureprocess and vulcanization. J. Colloid Interface Sci. 2011, 354, 758.(30) Erdogan, B.; Song, L.; Wilson, J. N.; Park, J. O.; Srinivasarao,M.; Bunz, U. H. F. Permanent bubble arrays from a cross-linkedpoly(para-phenyleneethynylene): picoliter holes without microfabrica-tion. J. Am. Chem. Soc. 2004, 126, 3678.(31) Karikari, A. S.; Williams, S. R.; Heisey, C. L.; Rawlett, A. M.;Long, T. E. Porous thin films based on photo-cross-linked star-shapedpoly(d,l-lactide)s. Langmuir 2006, 22, 9687.(32) Yabu, H.; Kojima, M.; Tsubouchi, M.; Onoue, S.; Sugitani, M.;Shimomura, M. Fabrication of photo-cross linked honeycomb-patterned films. Colloids Surf., A 2006, 284−285, 254.(33) Li, L.; Chen, C.; Li, J.; Zhang, A.; Liu, X.; Xu, B.; Gao, S.; Jin, G.;Ma, Z. Robust and hydrophilic polymeric films with honeycombpattern and their cell scaffold applications. J. Mater. Chem. 2009, 19,2789.(34) Lim, D.-K.; Do, H.-S.; Kim, H.-J.; Bang, J.-S.; Yoon, G.-H.Preparation of SIS/SBS-based UV-cross-linkable pressure-sensitiveadhesives using the thiol-ene reaction. J. Adhes. Sci. Technol. 2007, 21,589.(35) Greenspan, L. Humidity fixed points of binary saturatedaqueous solutions. J. Res. Natl. Bur. Stand. (U.S.) 1977, 81A, 89.(36) Evans, J. F.; Gibson, J. H.; Moulder, J. F.; Hammond, J. S.;Goretzki, H. Angle resolved ESCA analysis of plasma modifiedpolysterene. Fresenius' J. Anal. Chem. 1984, 319, 841.(37) Johnson, R. E., Jr.; Dettre, R. H. Wetting of low-energy surfaces.In Wettability; Berg, J. C., Ed.; Marcel Dekker, Inc.: New York, 1993;Chapter 1, p 13.(38) Bunz, U. H. F. Breath figures as a dynamic templating methodfor polymers and nanomaterials. Adv. Mater. 2006, 18, 973.(39) Maruyama, N.; Koito, T.; Nishida, J.; Sawadaishi, T.; Cieren, X.;Ijiro, K.; Karthaus, O.; Shimomura, M. Mesoscopic patterns ofmolecular aggregates on solid substrates. Thin Solid Films 1998, 327−329, 854−856.(40) Woodward, I.; Schofield, W. C. E.; Roucoules, V.; Badyal, J. P. S.Super-hydrophobic surfaces produced by plasma fluorination ofpolybutadiene films. Langmuir 2003, 19, 3432.

(41) Brown, P. S.; Berson, A.; Talbot, E. L.; Wood, T. J.; Schofield,W. C. E.; Bain, C. D.; Badyal, J. P. S. Impact of picoliter droplets onsuperhydrophobic surfaces with ultralow spreading ratios. Langmuir2011, 27, 13897.(42) Edward, G. H.; O’Donnell, R. G. Relaxations in lightlycrosslinked polymers. J. Mater. Sci. 1986, 21, 958.(43) Clark, D. T.; Dilks, A. ESCA applied to polymers. XVIII. RFglow discharge modification of polymers in helium, neon, argon, andkrypton. J. Polym. Sci.: Polym. Chem. Ed. 1978, 16, 911.(44) Woodward, I. S.; Schofield, W. C. E.; Roucoules, V.; Bradley, T.J.; Badyal, J. P. S. Micropatterning of plasma fluorinated super-hydrophobic surfaces. Plasma Chem. Plasma Process. 2006, 26, 507.(45) Harshbarger, W. R.; Porter, R. A.; Miller, T. A.; Norton, P. Astudy of the optical emission from an rf plasma during semiconductoretching. Appl. Spectrosc. 1977, 31, 201.(46) Grill, V.; Walder, G.; Scheier, P.; Kurdel, M.; Mark, T. D.Absolute partial and total electron impact ionization cross sections forC2H6 from threshold up to 950 eV. Int. J. Mass Spectrom. Ion Processes1993, 129, 31.(47) Oner, D.; McCarthy, T. J. Ultrahydrophobic surfaces. Effects oftopography length scales on wettability. Langmuir 2000, 16, 7777.(48) Kim, H.-Y.; Chun, J.-H. The recoiling of liquid droplets uponcollision with solid surfaces. Phys. Fluids 2001, 13, 643.(49) Fetters, L. J.; Lohse, D. J.; Richter, D.; Witten, T. A.; Zirkel, A.Connection between polymer molecular weight, density, chaindimensions, and melt viscoelastic properties. Macromolecules 1994,27, 4639.(50) Hopkins, J.; Badyal, J. P. S. Nonequilibrium glow dischargefluorination of polymer surfaces. J. Phys. Chem. 1995, 99, 4261.(51) Wan, L.-S.; Ke, B.-B.; Zhang, J.; Xu, Z.-K. Pore shape ofhoneycomb-patterned films: modulation and interfacial behavior. J.Phys. Chem. B 2012, 116, 40.(52) Bico, J.; Marzolin, C.; Quere, D. Pearl drops. Europhys. Lett.1999, 47, 220.(53) Luo, B. H.; Shum, P. W.; Zhou, Z. F.; Li, K. Y. Surfacegeometrical model modification and contact angle prediction for thelaser patterned steel surface. Surf. Coat. Technol. 2010, 205, 2597.(54) Li, W.; Amirfazli, A. A thermodynamic approach for determiningthe contact angle hysteresis for superhydrophobic surfaces. J. ColloidInterface Sci. 2005, 292, 195.(55) Li, W.; Amirfazli, A. Microtextured superhydrophobic surfaces: athermodynamic analysis. Adv. Colloid Interface Sci. 2007, 132, 51.(56) Tuteja, A.; Choi, W.; McKinley, G. H.; Cohen, R. E.; Rubner,M. F. Design parameters for superhydrophobicity and super-oleophobicity. MRS Bull. 2008, 33, 752.(57) Cassie, A. B. D.; Baxter, S. Wettability of porous surfaces. Trans.Faraday Soc. 1944, 40, 546.(58) Wenzel, R. N. Resistance of solid surfaces to wetting by water.Ind. Eng. Chem. 1936, 28, 988.(59) Miwa, M.; Nakajima, A.; Fujishima, A.; Hashimoto, K.;Watanabe, T. Effects of the surface roughness on sliding angles ofwater droplets on superhydrophobic surfaces. Langmuir 2000, 16,5754.(60) Gao, L.; McCarthy, T. J. The “lotus effect” explained: tworeasons why two length scales of topography are important. Langmuir2006, 22, 2966.(61) Li, W.; Amirfazli, W. L. A. Hierarchical structures for naturalsuperhydrophobic surfaces. Soft Matter 2008, 4, 462.(62) Strani, M.; Sabetta, F. Free vibrations of a drop in partial contactwith a solid support. J. Fluid Mech. 1984, 141, 233.(63) Strani, M.; Sabetta, F. Viscous oscillations of a supported dropin an immiscible fluid. J. Fluid Mech. 1988, 189, 397.(64) Yabu, H.; Shimomura, M. Single-step fabrication of transparentsuperhydrophobic porous polymer films. Chem. Mater. 2005, 17, 5231.(65) Ha, J.-M.; Wolf, J. H.; Hillmyer, M. A.; Ward, M. D. Polymorphselectivity under nanoscopic confinement. J. Am. Chem. Soc. 2004, 126,3382.

Langmuir Article

dx.doi.org/10.1021/la302719m | Langmuir 2012, 28, 13712−1371913718

Page 8: Superhydrophobic Hierarchical Honeycomb Surfaces

(66) Hulteen, J. C.; Jirage, K. B.; Martin, C. R. Introducing chemicaltransport selectivity into gold nanotubule membranes. J. Am. Chem.Soc. 1998, 120, 6603.(67) Cheng, C. X.; Tian, Y.; Qiao Shi, Y.; Pei Tang, R.; Xi, F. Porouspolymer films and honeycomb structures based on amphiphilicdendronized block copolymers. Langmuir 2005, 21, 6576.(68) Mitchell, D. T.; Lee, S. B.; Trofin, L.; Li, N.; Nevanen, T. K.;Soderlund, H.; Martin, C. R. Smart nanotubes for bioseparations andbiocatalysis. J. Am. Chem. Soc. 2002, 124, 11864.(69) Yao, S. C.; Choi, K. J. Heat transfer experiments of mono-dispersed vertically impacting sprays. Int. J. Multiphase Flow 1987, 13,639.(70) Pasandideh-Fard, M.; Aziz, S. D.; Chandra, S.; Mostaghimi, J.Cooling effectiveness of a water drop impinging on a hot surface. Int. J.Heat Fluid Flow 2001, 22, 201.(71) Chen, R.-H.; Chow, L. C.; Navedo, J. E. Effects of spraycharacteristics on critical heat flux in subcooled water spray cooling.Int. J. Heat Mass Transfer 2002, 45, 4033.(72) Tourkine, P.; Le Merrer, M.; Quere, D. Delayed freezing onwater repellent materials. Langmuir 2009, 25, 7214.(73) Kulinich, S. A.; Farzaneh, M. How wetting hysteresis influencesice adhesion strength on superhydrophobic surfaces. Langmuir 2009,25, 8854.(74) Cao, L.; Jones, A. K.; Sikka, V. K.; Wu, J.; Gao, D. Anti-icingsuperhydrophobic coatings. Langmuir 2009, 25, 12444.(75) Mishchenko, L.; Hatton, B.; Bahadur, V.; Taylor, J. A.;Krupenkin, T.; Aizenberg, J. Design of ice-free nanostructured surfacesbased on repulsion of impacting water droplets. ACS Nano 2010, 4,7699.(76) Reichard, D. L. Drop formation and impaction on the plant.Weed Technol. 1988, 2, 82.(77) Sirringhaus, H.; Kawase, T.; Friend, R. H.; Shimoda, T.;Inbasekaran, M.; Wu, W.; Woo, E. P. High-resolution inkjet printing ofall-polymer transistor circuits. Science 2000, 290, 2123.(78) Wang, J. Z.; Zheng, Z. H.; Li, H. W.; Huck, W. T. S.;Sirringhaus, H. Dewetting of conducting polymer inkjet droplets onpatterned surfaces. Nat. Mater. 2004, 3, 171.(79) Sele, C. W.; von Werne, T.; Friend, R. H.; Sirringhaus, H.Lithography-free, self-aligned inkjet printing with sub-hundred-nano-meter resolution. Adv. Mater. 2005, 17, 997.(80) Doggart, J.; Wu, Y.; Liu, P.; Zhu, S. Facile inkjet-printing self-aligned electrodes for organic thin-film transistor arrays with small anduniform channel length. ACS Appl. Mater. Interfaces 2010, 2, 2189.(81) Barbulovic-Nad, I.; Lucente, M.; Sun, Y.; Zhang, M.; Wheeler,A. R.; Bussmann, M. Bio-microarray fabrication techniquesa review.Crit. Rev. Biotechnol. 2006, 26, 237.(82) Tan, C. P.; Cipriany, R.; Lin, D. M.; Craighead, H. G. Nanoscaleresolution, multicomponent biomolecular arrays generated by alignedprinting with parylene peel-off. Nano Lett. 2010, 10, 719.(83) Arrabito, G.; Pignataro, B. Inkjet printing methodologies fordrug screening. Anal. Chem. 2010, 82, 3104.(84) Mironov, V.; Boland, T.; Trusk, T.; Forgacs, G.; Markwald, R. R.Organ printing: computer-aided jet-based 3D tissue engineering.Trends Biotechnol. 2003, 21, 157.(85) Xu, T.; Jin, J.; Gregory, C.; Hickman, J. J.; Boland, T. Inkjetprinting of viable mammalian cells. Biomaterials 2005, 26, 93.(86) Bharathan, J.; Yang, Y. Polymer electroluminescent devicesprocessed by inkjet printing: I. Polymer light-emitting logo. Appl. Phys.Lett. 1998, 72, 2660.(87) Chang, S.-C.; Bharathan, J.; Yang, Y.; Helgeson, R.; Wudl, F.;Ramey, M. B.; Reynolds, J. R. Dual-color polymer light-emitting pixelsprocessed by hybrid inkjet printing. Appl. Phys. Lett. 1998, 73, 2561.

Langmuir Article

dx.doi.org/10.1021/la302719m | Langmuir 2012, 28, 13712−1371913719