Effect of Sessile Drop Volume on the Wetting Anisotropy Observed onGrooved Surfaces

Jing Yang,† Felicity R. A. J. Rose,† Nikolaj Gadegaard,‡ and Morgan R. Alexander*,†

School of Pharmacy, UniVersity of Nottingham, UniVersity Park, Nottingham NG7 2RD, U.K., and Centrefor Cell Engineering, Department of Electronics and Electrical Engineering, Glasgow UniVersity,

UniVersity AVenue, Glasgow G12 8LT, U.K.

ReceiVed December 2, 2008. ReVised Manuscript ReceiVed January 19, 2009

This study reports experimental measurements of the water contact angle (WCA) measured on surfaces with groovesof different widths using drop volumes ranging from 400 pL to 4.5 µL. These measurements were carried out on bothrelatively hydrophobic and hydrophilic surface chemistry formed using a conformal plasma polymer coating oftopographically embossed poly(methyl methacrylate) (PMMA). Anisotropic wetting of the grooved surfaces wasfound to be more marked for larger drops on both the hydrophilic and hydrophobic surfaces. Above a certain dropbase diameter to groove width ratio, topography had no effect on the measured WCA; this ratio was found to bedependent on the water drop volume. The WCA measured from the direction perpendicular to the grooves usingsubmicroliter water drops is found to be a good indicator of the WCA on the flat surface with equivalent wettabilities.To the best of our knowledge, this is the first study on the phenomenon of anisotropic wetting using picoliter waterdrops.

Introduction

In recent years, the phenomenon of “anisotropic wetting” onmicropatterned surfaces has attracted a lot of scientific interest.This is mainly driven by sophisticated technologies involvingwell-defined chemically and topographically micropatternedsurfaces where wettability is an important design parameter.1-3

Such phenomena have practical applications in self-cleaningsurfaces and can be applied in microfluidic systems.4-6 On idealsolid surfaces (i.e., flat, chemically homogeneous, rigid, insoluble,and nonreactive), the three-phase contact line of a drop is acircle, and the water contact angle can be predicted by Young’sequation.7 However, for chemically heterogeneous materials orsurfaces with topography, the shape of the three-phase contactline and the drop are distorted under the influence of the chemicalor topographic patterns.

Anisotropic wetting on surfaces with regular chemical patternshas been studied both theoretically and experimentally. Drelichet al. reported contact angles on surfaces consisting of alternatingand parallel hydrophilic/hydrophobic strips of 2.5 µm width.The advancing and receding contact angles, when measured withthe drop edge normal to the strips, were found to be ∼2-10°lower than those measured when viewing the drop edge parallelto the strips.8 Morita et al. also studied anisotropic wetting onalternating and parallel hydrophilic/hydrophobic strips with fourdifferent widths.9 The distortion of drop shape was found to be

increased by increasing the drop size from 0.2 to 5 µL. Zhao etal. reported anisotropic wetting on submicrometer-scale (groovewidths of 318, 396, and 513 nm with varying nanometer-scaledepth) periodic topographic structures consisting of parallelgrooves. It was found that the water contact angle measuredusing sessile drops of 3 µL from the direction perpendicular tothe grooves increased with increasing groove depth, which wasassigned to the increasing energy barrier to drop spreading causedby pinning of the groove edge.10 Long et al. used a thermodynamicmodel to predict the water contact angle on rough surfaces withdifferent topographic geometries. The contact angle with thethree-phase contact line parallel to the groove was found to bein a metastable state. The predicted apparent contact angle wasin a relative lower energy state and was trapped by higherneighboring energy states.11 Chen et al. numerically quantifiedthe apparent contact angles on a hydrophobic surface with parallelgrooves.12 It was found that there are multiple equilibrium shapesfor a drop on rough surfaces with parallel grooves. A particularequilibrium shape was obtained by fixing the number of grooveson which the drop resided.

Thus far, all experimental studies on anisotropic wetting havebeen conducted using microliter volume drops; consequently,the diameter of the drop base relative to the striped chemical orgeometric feature size was large. With the increasing use ofmicroarrays, gradient samples, and microelectromechanicalsystem devices for biological studies,13-15 there has been a driveto reduce the dimensions on which surface wettability may becharacterized to improve spatial resolution, which has been metby developing systems utilizing smaller water drops. Using apiezo dispenser, a commercial system capable of routinely dosing

* Corresponding author. Email: Morgan.Alexander@nottingham.ac.uk.† University of Nottingham.‡ Glasgow University.(1) Kumar, A.; Biebuyck, H. A.; Whitesides, G. M. Langmuir 1994, 10, 1498–

1511.(2) Martines, E.; Seunarine, K.; Morgan, H.; Gadegaard, N.; Wilkinson,

C. D. W.; Riehle, M. O. Nano Lett. 2005, 5, 2097–2103.(3) Wouters, D.; Schubert, U. S. Angew. Chem., Int. Ed. 2004, 43, 2480–2495.(4) Blossey, R. Nat. Mater. 2003, 2, 301–306.(5) Grunze, M. Science 1999, 283, 41–42.(6) Zhao, B.; Moore, J. S.; Beebe, D. J. Science 2001, 291, 1023–1026.(7) Young, T. Philos. Trans. R. Soc. London 1805, 95, 65–87.(8) Drelich, J.; Wilbur, J. L.; Miller, J. D.; Whitesides, G. M. Langmuir 1996,

12, 1913–1922.(9) Morita, M.; Koga, T.; Otsuka, H.; Takahara, A. Langmuir 2005, 21, 911–

918.

(10) Zhao, Y.; Lu, Q. H.; Li, M.; Li, X. Langmuir 2007, 23, 6212–6217.(11) Long, J.; Hyder, M. N.; Huang, R. Y. M.; Chen, P. AdV. Colloid Interface

Sci. 2005, 118, 173–190.(12) Chen, Y.; He, B.; Lee, J. H.; Patankar, N. A. J. Colloid Interface Sci.

2005, 281, 458–464.(13) Gallant, N. D.; Lavery, K. A.; Amis, E. J.; Becker, M. L. AdV. Mater.

2007, 19, 965–969.(14) Anderson, D. G.; Levenberg, S.; Langer, R. Nat. Biotechnol. 2004, 22,

863–866.(15) Urquhart, A. J.; Anderson, D. G.; Taylor, M.; Alexander, M. R.; Langer,

R.; Davies, M. C. AdV. Mater. 2007, 19, 2486–2491.

2567Langmuir 2009, 25, 2567-2571

10.1021/la803942h CCC: $40.75 2009 American Chemical SocietyPublished on Web 02/03/2009

and measuring contact angles of picoliter volume water dropshas been developed. The contact angle measured using picoliterdrops on flat homogeneous polymer surfaces has been found toagree well with that measured by using microliter drops.16

However, the effect of drop size on the contact angle measuredon surfaces with anisotropic topography using submicroliter dropshas not yet been studied.

Here, we report water contact angles measured using dropsize ranging from picoliters to microliters on surfaces with parallelgrooves. PMMA surfaces embossed with grooves were coatedwith plasma polymers to achieve a relatively hydrophobic surfaceand a more hydrophilic surface. Micrometer-width grooves offive different dimensions were fabricated on PMMA substratesusing hot embossing. The relationships between the water contactangle and the drop volume is investigated, with comparison madeto predictions by Wenzel and Cassie equations.

Experimental SectionTopography Preparation. Topography with equally sized

alternating grooves and ridges where the grooves are 3 µm deepwith five different widths (5, 10, 25, 50, and 100 µm) was fabricatedusing hot embossing. Silicon masters with designed topography weremade by photolithography and reactive ion etching. PMMA sheetswith a diameter of 10 cm were compressed against the silicon masterunder a pressure of 0.48 MPa at 180 °C for 3 min. The pressure wasallowed to decrease to 0.24 MPa over a period of approximately 5min, at the same time as the temperature decreased to 70 °C. Themaster was then released from the PMMA replica, which was allowedto cool to room temperature.

Plasma Polymerization Coating. Plasma polymerization wasconducted in a T-shaped borosilicate glass chamber. Plasma wasinitiated by a 13.56 MHz radio frequency power source (CoaxialPower System Ltd.). Plasma polymerization was carried out at apower of 20 W under a working pressure of 300 mTorr. The amountsof deposited plasma-polymerized allylamine (ppAAm) and hexane(ppHex) were controlled by reading a quartz crystal microbalanceabove the sample in the plasma chamber. The plasma power sourcewas switched off when the reading of the thickness of plasma-polymerized polymer reached a thickness of 50 nm. Allylamine andhexane were used as monomers and were obtained from Sigma-Aldrich.

AFM and Water Contact Angle Measurements. A NanoscopeIIIa AFM (Digital Instruments) operating in tapping mode was usedto examine the surfaces of ppAAm- and ppHex-coated PMMA. Theroughness mean value was measured at five different places on thesample. Water droplets with four different sizes, namely, 0.4, 4, 12,and 40 nL, were dispensed by a piezo doser onto each polymersample using a DSA100 contact angle measurement machine (Kruss,Germany). Measurements were taken over five areas of 10 × 10µm2 for each polymer sample, from which average and standarddeviation values were calculated. A CAM200 instrument (KSVInstruments, Ltd.) was used to dispense ∼1.3 and ∼4.5 µL volumewater droplets onto each polymer sample. Again, five WCAmeasurements were taken for each polymer sample over differentareas. Ultrapure water was used for all WCA measurements (18.2MΩ resistivity at 25 °C). The WCAs were measured using a circleand a Young-Laplace fitting function for water drops with volumesof less than and more than 1 µL, respectively. The camera for recordedimages of drops worked at a speed of 107 frames/s.

Because picoliter drops evaporate much faster than microliterdrops, to minimize the effect of evaporation on the accuracy ofWCA measurements, all WCAs were measured from the first imageof the drop profile on the surface. It has been previously reportedthat when measured from the first image of the drop profile on flatsurfaces, the WCAs measured using picoliter drops were close tothose measured using microliter drops.16 No oscillation of drops

was observed after being dosed by the DSA100 small drop dispensingunit. This is supported by the video in the Supporting Information.The oscillation-free behavior of picoliter drops might be attributedto their small volumes and correspondingly small momentum.

Results and Discussion

The WCAs measured on planar ppAAm- and ppHex-coatedsurfaces using drops varying from 0.4 nL to 4.5 µL are presentedin Figure 1 as the cosine of the WCA (θ) versus the drop sizerepresented by the base diameter of the drop. The WCAs wereobserved to increase for larger drops on the ppAAm surface. Incontrast, WCAs on ppHex-coated surfaces were approximatelyunchanged within the range of drop volumes used (0.4 nL to 4.5µL). The systematic increase in the WCA with increasing dropvolume may be caused by the line tension of the three-phasecontact line.17-19 The line tension was predicted by Gibbs to bea 1D analog of the surface tension.20 A modified Young equationwas proposed to account for the effect of line tension on thecontact angle21

cos θ) cos θ∞ -γSLV

γLVr(1)

where θ ) θ∞ for r f ∞.Equation 1 indicates that there is a linear relationship between

the cosine of the contact angle (cos θ) and the reciprocal of thedrop base radius (1/r). The line tension can be calculated fromthe slope of a plot of cos θ versus 1/r as shown in Figure 1. ThelinetensionsmeasuredforppAAm-water-airandppHex-water-airsystems are 3 × 10-5 and -4 × 10-7 J m-1, respectively.Theoretical analysis of the three-phase contact line indicates thatthe line tension value ranges from 1 × 10-12 to 1 × 10-10 Jm-1.18 However, experimental line tension values are much largerbecause of the limit of resolution of instrumental techniques andthe limit of accuracy of measurements. In the case of measuringthe line tension involving a solid phase, difficulties arise fromimperfections in the solid surface with respect to the heterogeneity,roughness, elasticity, and reactivity with the contacting liquid.22

Because the average roughness values of ppAAm and ppHex are

(16) Taylor, M.; Urquhart, A. J.; Zelzer, M.; Davies, M. C.; Alexander, M. R.Langmuir 2007, 23, 6875–6878.

(17) Amirfazli, A.; Kwok, D. Y.; Gaydos, J.; Neumann, A. W. J. ColloidInterface Sci. 1998, 205, 1–11.

(18) Drelich, J. Colloid Surf., A 1996, 116, 43–54.(19) Drelich, J.; Miller, J. D.; Good, R. J. J. Colloid Interface Sci. 1996, 179,

37–50.(20) Gibbs, J. W. The Collected Works of J. Willard Gibbs, Yale University

Press: London, 1957; p 288.(21) Boruvka, L.; Neumann, A. W. J. Chem. Phys. 1977, 66, 5464–5476.(22) Drelich, J.; Miller, J. D. J. Colloid Interface Sci. 1994, 164, 252–259.

Figure 1. Water contact angle measured on a flat surface coated withppAAm and ppHex. Values are averages of five measurements. Errorbars are the measured standard deviation. The average roughness (Ra)values are 2.8 ( 0.2 nm (mean ( standard deviation) and 3.3 ( 0.8 nmfor ppAAm- and ppHex-coated surfaces, respectively.

2568 Langmuir, Vol. 25, No. 5, 2009 Letters

similar, the differences in the sign and magnitude of the linetension may be attributed to surface factors such as elasticity andreactivity.

The optical images of a water drop placed on a grooved surfaceare shown in Figure 2 from above (a) and in profile parallel (b)and perpendicular (c) to the grooves. The WCAs measured fromthe parallel direction increased significantly with larger drops(0.4 nL to 1.3 µL) as shown in Figure 3a,b. In contrast, WCAmeasured from the perpendicular direction only increased slightlyfor most drop sizes (Figure 3c,d). The increase in WCAs measuredfrom the parallel direction with increasing water drop volumeon the ppHex-coated surface was much less than those measuredon the ppAAm-coated surfaces. WCAs measured from the paralleldirection on 5 and 10 µm grooves were higher than those measuredon the other three grooves when 0.4 and 4 nL water drops wereused (Figure 3b). There was a notable increase in WCAs measuredon 25, 50, and 100 µm surfaces when the drop volume wasincreased from 4 to 40 nL (Figure 3b). Viewed in the planeperpendicular to the grooves, WCAs on all grooved surfaceswere similar to those on flat surfaces except for the largest dropvolume of 4.5 µL (Figure 3d). The upper and lower dashed linesin Figure 3 represent the predictions of the Cassie and Wenzelmodels, respectively. The Wenzel and Cassie equations are usedto calculate WCAs on rough surfaces. The Wenzel equation isrecommended when the drop wets the whole surface, and theCassie equation applies when the drop/solid surface is partiallywet (i.e., the drop/solid surface is an air/solid composite). Thesetwo models are usually used to calculate WCAs of drops withcircular three-phase contact lines on isotropic surfaces and areused here to provide a theoretical comparison with the experi-mental data. The equations for the two models are shown below:

Cassie equation: cos θ) f1 cos θ1 + (1- f1)cos θ2 (2)

Wenzel equation: cos θ) r cos θi (3)

θ is the apparent contact angle, and f1 is the area fraction of thesolid surface of a composite wetting stage (solid/air). In thiscase, f1 ) 0.5 because the grooves and ridges are equal in size.θ1 and θ2 are the intrinsic contact angles on surface 1 (solidsurface) and surface 2 (air, θ2 ) 180°), respectively, and r is theratio of the real contact area to the projected area:

r) w+ dw

w and d are the width and depth of the groove, respectively. Thevalue of WCA predicted by the Wenzel model in Figure 3 is

calculated using w ) 5 µm, d ) 3 µm, and θi ) 59.1 (WCAmeasured using 1.3 µL water droplets on the flat ppAAm surface).The WCA predicted by the Cassie model in Figure 3 is calculatedusing θi ) 94.4° (WCA measured using 4.5 µL water dropletson the flat ppHex surface).

Interestingly, the WCA measured from the perpendiculardirection using 4.5 µL water droplets on ppHex surfacesapproached the prediction by the Cassie model with decreasinggroove width (Figure 3d). We postulate that as grooves becomesmaller it is more difficult for the microliter droplet to intrudeinto the recessed areas, and the contact area becomes an air/solidcomposite state on the hydrophobic surface (i.e., the wettingmode of drops on hydrophobic surfaces changed from thatdescribed by the Wenzel model to the Cassie description whilethe grooves become smaller).23,24 However, except the data fromthe hydrophobic grooved substrate viewed from the perpendicularplane (Figure 3d), a large deviation of measured WCAs comparedwith predictions by the Wenzel and Cassie equations was found.This indicates that the Wenzel and Cassie models are not suitablefor predicting WCA on anisotropic surfaces because they do notconsider the pinning effect of the surface patterns.

One particularly interesting aspect of these findings is thatWCAs measured from the perpendicular direction were similarto those on the flat surface when submicroliter drops were used(Figure 3c,d). In addition, WCAs using subnanoliter drops (0.4nL) showed good agreement with those on flat surfaces exceptfor that measured from the parallel direction on the hydrophobicppHex surface with 5 µm grooves.

Further interrogation of Figure 3a,b indicated that the WCAsmeasured on ppHex and ppAAm parallel to the grooves reacheda plateau with increasing drop size. This suggested that the effectof topographic feature size on WCA may be dependent on therelative water drop and topographic feature size. To take thisinto account, we investigate the relationship between the WCAdifference (between flat and topographically patterned surface)and the ratio of the water drop base diameter to the groove width.Because the contact line is not circular in this case, the diameterrefers to either the long or the short axis of the distorted dropdepending on which is viewed. The difference between the WCAmeasured on the grooved surface and that measured on the flatsurface is plotted against the ratio of drop base diameter/groovewidth in Figure 4.

Only small WCA differences were found when the drops wereviewed parallel to the grooves with 0.4 nL droplets on the ppAAmsurface (Figure 4a). For water drops larger than 0.4 nL, the WCAincreased prominently before reaching a plateau, indicating athreshold ratio (approximately 10) of drop base diameter to groovesize beyond which the influence of topographic on WCA becamemarginal. In contrast, WCAs measured from the perpendiculardirection on the ppAAm surface are much less dependent on theratio (Figure 4c). A similar threshold ratio was also observed forthe ppHex surface whereas the water drop was smaller than 40nL (Figure 4b). WCAs measured from the perpendicular directionon ppHex surfaces were similar to that on the flat surface, withthe exception of the largest 4.5 µL drop, for which there was asharp increase in WCA when the ratio increased from ∼20 to∼80 (Figure 4d).

The marked dependence of WCA on drop size for groovedsurfaces may be attributed to the pinning effect of the grooveedge. During the process of reaching the equilibrium state aftercontacting the surface, the drop that is initially present in the airas a sphere has to overcome a series of energy barriers caused

(23) Cassie, A. B. D. Discuss. Faraday Soc. 1948, 3, 11–16.(24) Wenzel, R. N. Ind. Eng. Chem. 1936, 28, 988–944.

Figure 2. Optical images (top view) of the PMMA substrate withalternating grooves and ridges of equal sizes. The scale bar is 100 µm.(a) PMMA substrate with 5 µm grooves. The inset is the top view ofa ∼1.3 µL water drop on a ppAAm-coated surface with 5-µm-widegrooves. Profile view of a sessile drop (1.3 µL) on a ppAAm-coatedsurface (b) parallel and (c) perpendicular to the grooves.

Letters Langmuir, Vol. 25, No. 5, 2009 2569

by the groove edge. The final state of the drop will stay at ametastable state at which its free surface energy is lower thanits neighboring states.11 We postulate that during the process ofreaching a metastable state, differently sized drops reach differentmetastable states that correspond to different WCAs. A mildincrease in the apparent contact angle (140.4 to 150.7°) has beenobserved with increasing drop volume (0.59 to 5.68 mL) on asurface with parallel grooves (groove width and depth, 25.6 and30 µm, respectively). The experimental results agreed well withpredicted WCAs based on minimizing the free energy of thesystem to obtain the equilibrium drop shape using a algorithmdeveloped by Brakke.12 The increase in WCA could be more asdrop volume increases from picoliters to microliters. However,

a numeric simulation for predicting WCA on surfaces withmicrogrooves is beyond the scope of this study.

Conclusions

Water contact angles measured using picoliter to microliterwater drops on model hydrophilic and hydrophobic groovedsurfaces have been reported for the first time. A greater WCAis measured when the drop is viewed from the direction parallelto the grooves compared with that measured in the perpendiculardirection, which is close to that on the flat surface in most cases.Interestingly, even on the more hydrophilic ppAAm-coatedsurface, WCAs measured from the parallel direction exhibit anapparently hydrophobic WCA of approximately 120°. The

Figure 3. Water contact angle vs water drop volume on grooves of different sizes. The upper and lower dashed lines in each graph represent thepredicted WCA using the Cassie and Wenzel models, respectively.

Figure 4. Water contact angle difference vs ratio of drop base diameter to groove width. The difference is calculated by subtracting the WCA onthe rough surface from that on the flat surface. θi and θ1 used in both Wenzel and Cassie equations are WCAs measured on flat surfaces using microliterdroplets.

2570 Langmuir, Vol. 25, No. 5, 2009 Letters

significant change in WCAs measured when varying the volumeof the drop from picoliters to microliters indicates that greatattention needs to be paid when measuring contact angles onrough surfaces using submicroliter drops. The authors recommendshowing drop size alongside contact angle results in future studieson anisotropic wetting because of the impinging effect ofanisotropic surfaces on drops. The WCAs measured from theperpendicular direction with submicroliter water drops on allgrooved surfaces tested were found to be a good indicator of theWCA measured on flat surfaces.

Acknowledgment. This work was financially supported bythe BBSRC (grant no. BB/E012256/1). We thank M. Robertsonfor her help with the fabrication of topographically patternedsubstrates.

Supporting Information Available: Video of a picoliter waterdrop profile after contacting the surface. This material is available freeof charge via the Internet at http://pubs.acs.org.

LA803942H

Letters Langmuir, Vol. 25, No. 5, 2009 2571