Picoliter Water Contact Angle Measurement on Polymers

Michael Taylor, Andrew J. Urquhart, Mischa Zelzer, Martyn C. Davies, andMorgan R. Alexander*

Laboratory of Biophysics and Surface Analysis, Department of Pharmaceutical Sciences,UniVersity of Nottingham, UniVersity Park, Nottingham NG7 2RD, U.K.

ReceiVed January 12, 2007. In Final Form: April 24, 2007

Water contact angle measurement is the most common method for determining a material’s wettability, and thesessile drop approach is the most frequently used. However, the method is generally limited to macroscopic measurementsbecause the base diameter of the droplet is usually greater than 1 mm. Here we report for the first time on a dosingsystem to dispense smaller individual droplets with control of the position and investigate whether water contact anglesdetermined from picoliter volume water droplets are comparable with those obtained from the conventional microlitervolume water droplets. This investigation was conducted on a group of commonly used polymers. To demonstratethe higher spatial resolution of wettability that can be achieved using picoliter volume water droplets, the wettabilityof a radial plasma polymer gradient was mapped using a 250µm interval grid.

Introduction

Wettability, the degree to which a solid may be wet by aliquid, is a property of surfaces that influences many phenomenasuch as the biological response to materials and coating adhesionand durability.1,2 Usually, wettability is assessed through themeasurement of the contact angle (CA) of a liquid droplet placedon a surface, which is a quick, economical, and relatively simpletechnique.3 The CA is one of the most sensitive of all surfaceanalytical techniques because only the top nanometer of a surfaceinfluences wettability.4 The CA of a liquid on a solid dependson both surface chemistry and roughness.5 When the contactangle is estimated by fitting a function to the profile of a droplet,shape distortion by gravity must also be taken into account forlarger droplets.6

The sessile drop method of CA measurement commonly utilizesa few microliters of a liquid (e.g.,∼2 µL giving a 2 mmbasediameter when CA) 90°), which are placed on a surface froma needle. This method is useful for relatively large, homogeneoussurfaces but lacks lateral resolution when analyzing surfaceswith chemical differences on the submillimeter scale, as a resultof the dimension of the base diameter of a droplet. Smaller regionsof heterogeneity will result in a drop shape that averages theresponse of areas along the circumference of the base.7 With theminiaturization of many areas of science (microarrays, surfacechemical gradients, and microfluidics), there is an increasingneed to characterize surfaces that are small in area, which isconsequently impractical with microliter volume drops becauseof their size.

Recently, a new method of CA measurement has beendeveloped that allows the use of picoliter volume droplets ofliquid. This allows for improved spatial resolution of wettabilityon a surface and an ability to measure CA on much smaller areas,such as microarrayed materials.8These picoliter volumes of liquid

are dispensed using a piezo dosing unit similar to those used oninkjet printers and within biological array manufacturing.9 In thecase of water as the dosing liquid, we have produced 100 pLdroplets with a base diameter of approximately 70µm (whenWCA ) 90°). Here we show that water CAs measured frompicoliter droplets are equivalent to those measured from microliterdroplets on six commonly used polymers. We then demonstratethe ability of this new technique to achieve useful spatial resolutionof wettability using a chemical gradient surface formed fromconsecutive deposition and masking of plasma polymers.

Methods and Materials

Preparation of Polymer Films.Solutions (1% w/v) of polystyrene(Mw 100 000), poly(L-lactic acid) (Mw 95 000), poly(DL-lactic acid)(Mw 95 000), poly(methyl methacrylate) (Mw 60 000), and poly-(2-hydroxyethyl methacrylate) (Mw 20 000) were prepared inchloroform. All polymers were purchased from Sigma-Aldrich.Silicon wafers were cleaned using UV light and were then sonicatedin methanol. The polymer solutions were spin coated onto the cleansilicon wafers at 3000 rpm. The polymer films were left for 24 hbefore CA measurements. The surface of a piece of poly-(tetrafluoroethylene) (Kru¨ss) was scraped clean before each CAmeasurement.

Preparation of a Radial Plasma Polymer Gradient.The radialwettability gradient was prepared by plasma polymer deposition ofallylamine (ppAAm) through an aperture onto a glass substrate coatedwith plasma-polymerized hexane (ppHex). The radio-frequencyplasma (13.56 MHz) was driven at a power of 20 W while themonomer pressure was kept at 300 mTorr for both hexane andallylamine. The chemicals were supplied by Sigma-Aldrich anddegassed prior to use. The glass substrate was cleaned with ultrasound,washed with acetone, and treated with an oxygen plasma for 3 minbefore the deposition of ppHex. The radial gradient was obtainedby the diffusion-controlled deposition of ppAAm through a 1.2 mmhole in a nonconductive mask that was placed at a distance of 0.1mm over the ppHex-coated substrate. This is a development of apreviously reported patterning technique.10,11

CA Measurements.Images of the droplet profile were recordedfrom which the CA was determined using the angle of intersectionbetween a baseline and a circle or the Young-Laplace function fit

* Corresponding author. E-mail: morgan.alexander@nottingham.ac.uk.(1) Garbassi, F.; Morra, M.; Occhiello, E.Polymer Surfaces: From Physics

to Technology; John Wiley & Sons Ltd: Chichester, U.K., 1998.(2) Barnes, G.Interfacial Science: An Introduction; Oxford University Press:

Oxford, U.K., 2005.(3) Whitesides, G. M.; Laibinis, P. E.Langmuir1990, 6, 87-96.(4) Bain, C. D.; Whitesides, G. M.Langmuir1989, 5, 1370-1378.(5) Wenzel, R. N.J. Phys. Colloid Chem.1949, 53 (9), 1466-1467.(6) Herzberg, W. J.; Marian, J. E.J. Colloid Interface Sci.1970, 33, 161-164.(7) Pease, D. C.J. Phys. Chem.1945, 49, 107-110.(8) Urquhart, A. J.; Anderson, D. G.; Taylor, M.; Alexander, M. R.; Langer,

R.; Davies, M. C.AdV. Mater., submitted for publication.

(9) Gustavsson, P.; Johansson, F.; Kanje, M.; Wallman, L.; Linsmeier, C. E.Biomaterials2007, 28, 1141-1151.

(10) Barry, J. J. A.; Howard, D.; Shakesheff, K. M.; Howdle, S. M.; Alexander,M. R. AdV. Mater. 2006, 18, 1406-1410.

(11) Whittle, J. D.; Barton, D.; Alexander, M. R.; Short, R. D.Chem. Commun.2003, 1766-1767.

6875Langmuir2007,23, 6875-6878

10.1021/la070100j CCC: $37.00 © 2007 American Chemical SocietyPublished on Web 05/16/2007

to the drop profile. The Young-Laplace function models the dropletshape using two radii of curvature. The CA was also determinedusing a tangent placed at the intersection of the liquid and solid. Awater droplet with a volume of 100 pL was dispensed by a piezodoser onto each polymer sample using a DSA100 (Kru¨ss). Measure-ments were taken over 10 areas for each polymer sample from whichaverage and standard deviation values were calculated. A CAM200instrument (KSV Instruments, Ltd) was used to dispense∼2-12µLvolume water droplets onto each polymer sample. Again, 10 CAmeasurements were taken for each polymer sample over differentareas. Ultrapure water was used for all CA measurements (18.2 MΩresistivity at 25°C). To map the wettability of the polymer gradient,the DSA100 was used to deposit 625 pL volume droplets onto theradial plasma polymer gradient in a 6 mm× 6 mm square grid. Thiswas achieved with an automated stage and took 6 h for automateddosing and fitting. The CAs of these droplets were fitted using acircle fitting function, and the resulting CAs were plotted to givea 2D map of the gradient’s wettability.

Results and Discussion

If picoliter volume droplets are to be routinely used for CAmeasurements, it is useful to confirm experimentally that the CAdata acquired from them is equivalent to that acquired frommicroliter droplets. There are two major differences betweensmall and large dropletssthe influence of gravity on the dropletand the rate of drop size reduction due to evaporation.6,12

To investigate the effect of droplet size, water droplets ofdifferent volumes were placed on the PMMA surface bycontrolling the amount dispensed by the syringe. The dropletprofile was fit using either a circle or Young-Laplace functionand a tangent fitted by eye to the point of intersection with thesurface. It is clear that the CAs estimated using the circle functiondecrease as the droplet volume increases in Figure 1. The CAdetermined between the tangent and the surface did not varysignificantly (69( 1°) with droplet volume (not shown). Thisis considered to be the actual contact angle, and below we discussthe estimates based on fitting a function to the profile.

Theory states that when a droplet is placed onto a surface athree-phase equilibrium exists between the liquid, solid, and vaporphases, which is described by the classical Young equation.13

However, this does not include droplet size, which can influencethe CA measured when fitting a function to a droplet profile.6

If the droplet is small enough for the influence of gravity to beinsignificant, then the free energy of the system at equilibrium

is minimized for a truncated sphere shape; thus its profile canbe fitted to a segment of a circle. If it is large enough for thedistortion of the shape by gravity to be significant, then it isinstead better fit using the Young-Laplace equation. The Young-Laplace equation describes the curved profile of a droplet usinga two radius of curvature solution.14 The Young-Laplace fit ofthe droplet shape resulted in a small increase in CA with time(∼0.5° between 0.2 and 12µL as seen in Figure 1), which suggeststhat this method does not completely compensate for the increasein droplet size. Compared with the measured variance of the CA,this is insignificant. In contrast, the circle fit provides anincreasingly large underestimate of the CA compared to the actualCA (69 ( 1°) as the droplet volume increased. As the dropletvolume increases, the greater the effect of gravity, the lessspherical the droplet becomes, causing the circle-fitting modelto become increasingly inaccurate. Comparison with the tangentmethod suggests that for droplets in this size range gravity isinfluencing the shape of the droplet but not the CA itself. If theBond number (a numerical expression of the ratio of gravitationalto surface tension forces) is calculated for 2µL and 100 pLdroplets, then we obtain values of 1.51 and 6.58× 10-4,respectively.14 Supporting Information Figure 1 shows therelationship between the Bond number of a droplet and its CAon PMMA. It can be seen that at Bond numbers of less than∼1gravitational forces no longer influence the shape of the droplet.The Bond number for 100 pL droplets is much smaller than 1,which indicates that surface tension forces dominate overgravitational forces. Thus, for the remainder of this letter theYoung-Laplace method will be used for the 2µL droplets, anda circle fit will be used for the 100 pL droplets. Although theYoung-Laplace method would also be suitable for 100 pLdroplets, its increased complexity compared to that of the circle-fitting method makes the circle method preferable.

To investigate the influence of evaporation, droplet imageswere collected using a high-speed camera triggered to record asthe droplets were released (Figure 2a,b). It was possible toaccurately record the droplet image at time intervals very muchshorter than the time scale on which the droplets evaporated forboth large and small droplets as seen in Figure 2c,d. The WCAhas been plotted against time for both picoliter and microlitervolume droplets on five polymers. The differently sized dropletsshow quite distinctly different CA profiles over time after dosing.The WCA of 2 µL droplets decreases slightly with time in alinear manner (Figure 2c), and the decrease appears to be morepronounced with PHEMA, which has the lowest WCA. Therapid evaporation of the picoliter droplets causes a rapid decreaseof 10 to 20° in CA within approximately the first 0.5 s (Figure2d). Subsequently, a second stage was observed where the WCAdecreases more slowly. Both stages are essentially linear, andagain the WCA of PHEMA appears to decrease more rapidlythan the other polymers.

The observation of two distinct stages of WCA for picolitervolume droplets raises the question as to which WCA to measure.For example, is it correct to take the WCA at time zero or at thestart of the second stage. An examination of the videos revealedthat as the picoliter droplets evaporate they initially decreasedin height without movement of the perimeter and then decreasedin diameter with a contraction of the perimeter. Very similarbehavior has been reported for microliter droplets, although overmuch greater time periods (∼30 min), and has been rationalizedin terms of contact angle hysteresis.12,15,16Thus, an initial constant

(12) Bourgesmonnier, C.; Shanahan, M. E. R.Langmuir 1995, 11, 2820-2829.

(13) Kwok, D. Y.; Neumann, A. W.AdV. Colloid Interface Sci.1999, 81,167-249.

(14) Adamson, A. W.Physical Chemistry of Surfaces, 6th ed.; John Wiley &Sons: New York, 1997.

(15) Yu, H. Z.; Soolaman, D. M.; Rowe, A. W.; Banks, J. T.ChemPhysChem2004, 5, 1035-1038.

Figure 1. Water CA versus droplet volume on PMMA. Each waterdroplet was fitted using both circle and Young-Laplace functionsto model the drop shape. Linear regression fits have been providedto guide the eye.

6876 Langmuir, Vol. 23, No. 13, 2007 Letters

contact area results from the pinning of the perimeter, causingthe droplet height to decrease on evaporation and resulting indecreasing CA. When the CA reaches the receding value, thecontact area decreases while the CA stays constant.15 Thisdescription is valid for the size changes seen for picoliter droplets,although on shorter time scales because the decrease in volumeoccurs more rapidly for smaller drops.

It is apparent that this description of WCA with time after thedroplet is dispensed is valid for PLLA, PDLLA, PMMA, and PS,whereas the PTFE and PHEMA WCA values do not enter astable second phase. PHEMA is a mobile hydrogel that is knownto modify its surface structure to minimize its interfacial energyby exposing methyl groups (hydrophobic) when in contact withair and hydroxyl groups (hydrophilic) when in contact withwater.17 This could explain why its WCA decreases so rapidlyinitially, possibly as a result of the reorientation of the polymer’ssurface upon contact with the water droplet, and why the WCAcontinues to decrease. It is also possible that water sorption bythe PHEMA could be responsible for the continued decline inWCA. A rapid decrease in the WCA on PTFE was observed thatmay be attributed to either droplet or surface instability. Despitethis, within observed experimental variance, the initial contactangle is close to literature values.18

The WCAs of picoliter droplets taken from the first image ofthe droplet on the surface with the same image for the microliterdroplets are presented in Table 1. A remarkably good correlationbetween the two techniques was observed. The average differencebetween microliter and picoliter measurements for the sixpolymers is only 1.5°, and in most cases the values were withinthe variance between measurements on the respective instruments.The precision of the WCA measurement is often quoted as(1°,and accuracy measurements are rarely provided. With dueconsideration of precision and accuracy, the results gained withthe two methods are comparable. There is much debate over theinfluence of line tension on the CA of droplets, particularly smaller

droplets where the influence of line tension (and other surfaceforces) is theoretically greater.19 Line tension is defined as alinear tension at the boundary where three phases meet at theperimeter of a droplet. It is interesting to observe that over thedroplet size range of 100 pL to 2µL there is no significantchange in CA, suggesting that the influence of line tension onthe contact angle in droplets of this size range is negligible.

The polymer with the largest difference in WCA betweenmicroliter and picoliter volume droplets (3.6°) was poly(L-lacticacid), whereas the difference for poly(DL-lactic acid) was negli-gible (0.1°). To investigate this difference, both polymers wereimaged in air using tapping mode atomic force microscopy. Itwas observed that the poly(L-lactic acid) was considerably rougher(ra ) 29 nm) than poly(DL-lactic acid) (ra ) 7 nm) (SupportingInformation, Figure 2). Contact angles have been observed todecrease on rougher surfaces for CAs below 90°.5Picoliter volumedroplets with a base diameter of 70µm may be more sensitiveto this surface topography than microliter droplets, hence the3.6° lower WCA measured on PLLA. This explanation issupported by the similarity of the WCA measured using themicroliter and picoliter droplets on the smoother variant of thepolymer, poly(DL-lactic acid). A full study of the relative effectsof roughness on WCA measurements with differently sizeddroplets is warranted but is beyond the scope of this short letter.

There is much interest in using gradients of surface chemistryand wettability to guide cellular response and biomolecular(16) Rowan, S. M.; Newton, M. I.; McHale, G.J. Phys. Chem.1995, 99,

13268-13271.(17) Holly, F. J.J. Biomed. Mater. Res.1975, 9, 315-326.(18) Ellison, A. H.; Fox, H. W.; Zisman, W. A.J. Phys. Chem.1953, 57,

622-627.(19) Amirfazli, A.; Kwok, D. Y.; Gaydos, J.; Neumann, A. W.J. Colloid

Interface Sci.1998, 205, 1-11.

Figure 2. Images of the profile of (a) 2µL and (b) 100 pL water droplets on PMMA to demonstrate the difference in scale. Graphs of WCAversus time for (c) 2µL and (b) 100 pL water droplets on five polymers- PTFE, PHEMA, PLLA, PMMA and PS.

Table 1. CAs of Six Polymers Measured from Picoliter andMicroliter Volume Drops ( (SD)

polymer

CA measuredfrom a 100 pLdroplet (deg)

CA measuredfrom a 2µLdroplet (deg)

poly(2-hydroxyethylmethacrylate)

44.9( 1.0 43.5( 1.4

poly(L-lactic acid) 67.9( 0.5 71.5( 0.4poly(DL-lactic acid) 73.5( 0.5 73.6( 0.7poly(methyl methacrylate) 67.8( 0.6 69.7( 1.0polystyrene 89.6( 0.4 90.2( 2.2poly(tetrafluoroethane) 111.2( 1.5 111.3( 0.4

Letters Langmuir, Vol. 23, No. 13, 20076877

adsorption in scaffolds, sensors, and devices.10,20Using the aboveprocedure developed to make measurements with picoliterdroplets, we investigate a radial chemical pattern to illustrate theutility of the approach. A circular plasma-polymerized allylamine(ppAAm) (hydrophilic) area was deposited through an apertureacting as a raised mask on top of a predeposited plasma-polymerized hexane (ppHex) coating (hydrophobic) to providea radial wettability gradient. WCA measurements were acquiredat 250µm intervals within a 6 mm× 6 mm area, which givesa highly resolved picture of the change in wettability over thegradient. All WCA measurements were acquired under computercontrol using a motorized stage, and the drop shapes were recordedand fitted automatically. It can be seen that the WCA reachesa minimum of 44° in the center of the radial gradient, suggestingcomplete surface coverage with ppAAm (Figure 3). The WCA

increased gradually from the center of the gradient outward,reaching a maximum of 79° at the periphery. The WCA of uniformsamples of ppAAm is∼49 ° and for ppHex is∼97°; therefore,it can be reasonably assumed that the CA increase is due to thechange in chemistry over the gradient formed by the diffusionof ppAAm from the center outward.11

Using∼2µL volume droplets with a diameter of approximately2 mm, it would be possible to make only nine CA measurementswithin the 6 mm× 6 mm area; therefore, we have achieved 70times the sample density using the picoliter volume droplets.Until now, we have been limited to making routine sessile dropWCA measurements on the millimeter scale. The developmentof this approach opens up a whole new world of applications onthe submillimeter length scale, with applicability to areas suchas microarray and gradient surface analysis.8,20

Conclusions

We have compared the use of microliter and picoliter vol-ume droplets of water for WCA measurement on six polymers:PDLLA, PLLA, PMMA, PS, PHEMA, and PTFE.

We have shown that CAs measured from picoliter volumedroplets are remarkably similar to those measured from microlitervolume droplets when an appropriate consideration of evaporationand droplet shape is used.

We have demonstrated the ability of this new approach toachieve high levels of spatial resolution of wettability on a plasmapolymer gradient.

Acknowledgment. The BBSRC funded this work with aproject grant (BBC5163791) and a studentship to M.T.

Supporting Information Available: Water CA versus dropletBond number on PMMA. Tapping mode AFM height images of PLLAand PDLLA. This material is available free of charge via the Internet athttp://pubs.acs.org.

LA070100J(20) Zelzer, M.; Majani, R.; Bradley, J. W.; Rose, F. R. A. J.; Davies, M.

C.; Alexander, M. R. To be submitted for publication.

Figure 3. Three-dimensional wettability map of a radial plasmapolymer gradient.

6878 Langmuir, Vol. 23, No. 13, 2007 Letters