Journal of Colloid and Interface Science 367 (2012) 472–477

Contents lists available at SciVerse ScienceDirect

Journal of Colloid and Interface Science

www.elsevier.com/locate/jcis

Contact angles of drops on curved superhydrophobic surfaces

Goutham Viswanadam, George G. Chase ⇑
Department of Chemical and Biomolecular Engineering, The University of Akron, Akron, OH 44325, United States

a r t i c l e i n f o a b s t r a c t

Article history: Superhydrophobic surfaces have contact angles that exceed 150 degrees and are known to reduce surface
Received 17 June 2011 fouling, protect surfaces, and improve liquid–liquid separations. Electrospun sub-micron fiber mats can
Accepted 2 November 2011 perform as superhydrophobic surfaces. Superhydrophobic behavior is typically measured on planar
Available online 12 November 2011
surfaces, whereas applications may require curved surfaces. This paper discuses the measurement of
water contact angles of fiber mats formed on cylindrical surfaces to create superhydrophobic behavior
Keywords: on curved surfaces. Equations are derived that relate the radius of curvature of spherical and cylindrical
Contact angle
surfaces and drop size to the observed contact angle on the curved surfaces. Calculations from the equa-
Nanofibers
Electrospinning
tions agree well with experimental observations on spherical surfaces reported in literature and on cylin-
Superhydrophobic drical surfaces created in our lab.
Drops Ó 2011 Elsevier Inc. All rights reserved.
Droplets
Wetting

1. Introduction surface curvature in an environment friendly way to resist biofoul-

Water repellency is important in many applications [1–3], such
as the prevention of the adhesion of snow or ice to antennas, air-
craft wings, windows and self-cleaning traffic indicators. The
wettability of a liquid on a solid surface [2,4] can be evaluated
by measuring the liquid’s contact angle [1,5]. A few studies con-
sider the effect of curvature on the hydrophobicity of the surface
but most only consider planar or low curvature surfaces. Alarcon
et al. [6] studied water molecular orientations on curved surfaces
and Choi et al. [7] investigated the superhydrophobicity of silicon
nanowires, suggesting strong curvature has an effect on hydropho-
bicity. Tools to assist interpretation of experimental observation of
drops on curved surfaces are needed.
Superhydrophobic surfaces are of interest for a number of appli-
cations. As an example, fibrous depth filter media are used to
separate secondary dispersions of water droplets from oil [8]. Med-
ia are typically constructed as pleated flat sheets or as cylindrical
cartridges. If the hydrophobicity is enhanced by small radius of
curvature, then media designers may take advantage of this via
small diameter tubular media to achieve greater separation
efficiency.
In a second example, superhydrophobic surfaces exhibit low
fluid drag on aquatic surfaces and protect aquatic devices from
marine fouling. Instead of applying coatings, which may be toxic,
to repel marine organisms, these surfaces may be enhanced via
⇑ Corresponding author.
E-mail address: gchase@uakron.edu (G.G. Chase).
ing on the underwater parts of ships [1–3].
A third example application is corrosion prevention. Water is
responsible for most corrosion that occurs on materials. Superhy-
drophobic coatings to protect surfaces from corrosion are of
current research interest [9]. Electrospun nanofiber coatings are
reported for imparting superhydrophobicity and corrosion resis-
tance to surfaces [10]. Surface curvature may affect the hydropho-
bic interaction between water droplets and the surfaces in all of
these example applications. The interpretation of the observed
contact angles is important. In the extreme case of a sharp edge
of a surface, Gibbs proposed a simple geometric relation which de-
scribes the interaction of liquids with sharp edges [11]. When a
liquid droplet is placed on a flat surface with sharp edges, an intrin-
sic contact angle of h0 is observed. As liquid is added to this drop,
the drop advances to the edge of the surface. With the addition of
more liquid, the contact angle relative to the horizontal plane at
the edge increases. Eventually, the drop reaches a critical apparent
advancing contact angle (h) at the edge that any further liquid
added to the drop causes the drop to collapse. According to Gibbs,
the apparent advancing contact angle (h) increases due to the
interaction with the sharp edge according to the equation
h ¼ h0 þ ðp  UÞ ð1Þ
where U is the edge angle. In the late 1970s, Mason and colleagues
performed experiments using conical frusta which proved the
above expression correct [12].
In this paper, we derived an equation to calculate the water
contact angle on curved spherical and cylindrical surfaces. The
derived equations relate the drop size and radius of curvature of

0021-9797/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcis.2011.11.004
 G. Viswanadam, G.G. Chase / Journal of Colloid and Interface Science 367 (2012) 472–477
the surface to the apparent contact angle on the curved surfaces.
Experiments are conducted with drops on superhydrophobic
curved surfaces fabricated from mats of electrospun fibers. The
theoretical predicted angles are compared with measured angles
as a function of drop volume. In most cases the experimental
results are in good agreement with the predictions.
2. Theoretical analysis
Consider a sessile liquid drop of volume V on a spherical surface
in such a way that the drop is symmetric on the y axis which passes
through the center of the sphere as well as the drop. Assuming the
drop to spread along the curved surface to produce an intrinsic
advancing contact angle (h0), as shown in Fig. 1, which would be
the angle made by the liquid drop on a flat surface of the same
material and composition. Due to the curvature of the sphere,
the apparent advancing contact angle (h), relative to the horizontal
is greater than the intrinsic value h0.
In Fig. 1, line segments CB and CD equal the radius of the sphere,
R. Line segments AB and AD equal r, the radius of the liquid drop.
In triangle FCB
a ¼ R sin a ð2Þ
and in triangle FAB
a ¼ r sin b ð3Þ
Combining Eqs. (2) and (3) gives
R sin b
¼ ð4Þ
r sin a
By inspection of the angles and the lines described in Fig. 1, the
following relations are obtained
a¼p/ ð5Þ
a þ b þ h0 ¼ p ð6Þ
Combining Eqs. (4)–(7) it is possible to derive
R sinð/  h0 Þ
¼ ð7Þ
r sin /
Fig. 1. Schematic diagram of a liquid droplet sitting on top of a solid sphere.
473
Combining Eq. (7) with Gibb’s Eq. (1) gives
R sinðhÞ
¼ ð8Þ
r sinðh  h0 Þ
Eq. (8) can be rearranged to calculate the apparent contact angle (h),
with help of a trig identity, to the form
 
sinðh0 Þ
h ¼ ArcTan ð9Þ
cosðh0 Þ  Rr
The apparent contact angle (h), due to the effect of curved sur-
face, can be calculated for a given droplet size, radius of curvature
of the sphere and the contact angle on a flat surface. Eq. (9) is de-
rived for a droplet sitting on a spherical surface. For a drop sitting
on the surface of a cylinder, the axial view is analogous to the view
shown in Fig. 1. Therefore, the equation for the apparent contact
angle for the axial view of a drop sitting on a cylindrical surface
is by analogy the same as Eq. (9).
2.1. Comparison of Eq. (9) to reported data for drops on spheres
The Gibbs equation is a consequence of the geometry and is
immaterial of the intrinsic nature of the edge. A curved surface
can be treated as a continuous edge surface with a varying angle.
To check the validity of the Eq. (9), we compare the calculated
water contact angle (WCA) with the experimental data reported
by Extrand and Moon [13] of drops of varying volumes on spherical
surfaces of Polytetrafluoroethylene (PTFE) and Polycarbonate (PC).
The intrinsic water contact angle on flat surfaces of PTFE and PC are
reported as h0(PTFE) = 108° and h0(PC) = 89° [13]. The experimental
data and values predicted by Eq. (9) are plotted in Fig. 2 for PTFE
and SPP spherical surfaces.
The plots in Fig. 2 show theoretical values have the same trend
as the experimental values. For a given radius of the spherical sur-
face the apparent contact angles relative to the horizontal axis in-
crease with increasing drop size (or radius ratio). As the drop size
increases, the drops cover more of the sphere surface resulting in a
decrease in the observed edge angle. The reasons for a deviation of
the experimental results from the calculated angles may be due to
variation in drop symmetry relative to the vertical axis through the
center of the sphere. It may also be due to our assumed spherical
drop shape in deriving Eq. (9), where physically the drop might
not be truly spherical in shape. Because Eq. (9) is derived from geo-
metrical analysis, in principle it should apply to any surface mate-
rial of the sphere.
The theoretical results for both PTFE and PC agreed reasonably
well with the experimental results reported by Extrand and Moon
[13]. Eq. (9) accounts for both the effect of the radius ratio and the
WCA for a flat surface.
θ
Fig. 2. Graphical comparison of WCA on PTFE and PC spherical surfaces calculated
theoretically and measured experimentally (data from Extrand and Moon [13]). ‘R’
is the radius of curvature of the spherical surfaces and ‘r’ is the radius of the drop in
mm.
474 G. Viswanadam, G.G. Chase / Journal of Colloid and Interface Science 367 (2012) 472–477

2.2. Experimental description for drops on cylindrical surfaces
To extend this work to cylindrical superhydrophobic surfaces,
we experimentally observe WCA on Syndiotactic Polypropylene
(SPP) sub-micron fiber mat surfaces. Cylindrical surfaces were
made by electrospinning [14–16] SPP onto stainless steel rods of
different diameters, whereas for flat surfaces the fibers were spun
onto flat glass slides.
into a 5 ml syringe. The needle of the syringe was positioned
horizontally on the syringe pump such that the needle was
oriented perpendicular to the rod at a distance of about 25 cm. A
high voltage power supply (Gamma High Voltage, Ormond Beach,
FL) was used to generate a potential difference of about 25 kV be-
tween the syringe needle and the grounded aluminum foil.
3.2. Contact angle measurement

3. Materials Contact angles were measured using a Drop Shape Analyzer

Syndiotactic Polypropylene (SPP, melt index: 2.20 g/10 min at
230 °C with a load 2.16 kg, ASTM D 1238) used in this study was pur-
chased from Aldrich (Sigma Aldrich, St. Louis, MO). The weighted
average (MW) and number-average molecular weight (MN) were
174,000 and 75,000 g/mol, respectively (MW/MN = 2.32). SPP was
dissolved in a solvent mixture of cyclohexane (99%+), acetone and
dimethylformamide (DMF) in a weight ratio of 80/10/10 at 70 °C
at a weight concentration of 2% of SPP to solvent mixture [17]. This
polymer mixture was electrospun to produce fiber mats. All of the
chemicals were purchased from Thermo Fisher Scientific Inc. (Wal-
tham, MA) and used without further modification.
3.1. Electrospinning setup
Stainless steel rods were chosen as collectors to make the
curved surfaces of electrospun fiber mats. As shown in Fig. 3, one
end of the steel rod is fixed to a motor and rotated at 10 rpm while
the other end is suspended horizontally in the air in front of a
grounded sheet of aluminum foil. The polymer solution was loaded
(DSA20E, KrüssUSA, Matthews, NC). In The sessile drop fitting
method was used for both flat and curved surfaces. The water
drops were placed using a micro-syringe pointed vertically down-
ward onto the sample surfaces.
4. Results and discussion
The flat surfaces of sub-micron fiber mats were made by elec-
trospinning SPP onto a glass slide with similar electrospinning con-
ditions as that of the curved surface.
Example SEM images of fibers produced by electrospinning are
shown in Fig. 4. From the SEM images, we observed beads as well
as fibers for both flat and curved surfaces. The average diameters of
the fibers were calculated using length weighted method [18]. The
Flat surface had an average fiber size of 0.534 lm, the 3.96 mm OD
rod had fibers with an average diameter of 0.450 lm, and the
7.57 mm rod had 0.561 lm diameter fibers, indicating the fiber
size ranges were similar due to the similar electrospinning condi-
tions. The beads contribute to the hydrophobic nature of the fiber
mat surface but because of their irregular sizes and shapes there is

Fig. 3. Electrospinning setup to make SPP cylindrical surfaces. (A) Overall setup. The polymer solution in the syringe was pumped at a flow rate of 25 ml/h through the
electrically charged needle. The motor and backdrop aluminum foil were grounded. Jets launching from the needle were attracted toward the grounded surfaces. (B) Close-up
view of motor and the steel rod collector. The stainless steel rod was positioned at a distance of 25 cm from the tip of the needle. The rod was rotated by the motor at 10 rpm.
The motor is covered with aluminum foil to protect it from fibers.
 G. Viswanadam, G.G. Chase / Journal of Colloid and Interface Science 367 (2012) 472–477 475

Fig. 4. Sample SEM images of SPP electrospun fibers on (A) flat surface (B) 3.96 mm OD steel rod (C) 7.57 mm OD steel rod.

Fig. 5. DSA20E photographs of water droplets of different volumes on flat SPP surface. (A) 4 ll (r = 0.985 mm), (B) 5 ll (r = 1.061 mm), (C) 6 ll (r = 1.128 mm), (D) 7 ll
(r = 1.187 mm). The images are at different magnifications and hence are not comparable in scale.

not a convenient way to characterize their coverage on the electro- indicating the cylindrical curvature affects the drop shape from the
spun fiber mats. side view.
WCA for different drop sizes were measured on flat SPP electro- The calculated angles for the axial view and observed apparent
spun fiber surfaces using the DSA20E. Example images of drops on angles are plotted in Fig. 9. The experimental results agree reason-
the surfaces are shown in Fig. 5. We do not have data for WCA of ably well with the calculations of Eq. (9). They show the same
drops less than 4 ll as they could not be detached from the syringe.
The WCA on the fiber surfaces depends upon the fiber diameter
and pore size due to surface roughness as described by Cassie–
Baxter theory [19], hence the contact angle varies with the size
of the drop, especially for small drops [20]. Fig. 6 shows the varia-
tion of the measured WCA for the water droplets on the flat SPP
θ

surface. The drops with WCA greater than 150 deg show the
surface has superhydrophobic properties.
The images in Fig. 7 are examples of axial views of drops of
different sizes on the curved cylindrical surfaces. Fig. 8 shows
examples of the side views of drops sitting on the cylindrical
surfaces. The axial view shows the drops look similar to those sit-
ting on the spherical surfaces and we expect Eq. (9) to give an esti-
mate of the apparent contact angle. Measurement of WCA for the Fig. 6. WCA on flat SPP electrospun fiber surface. The larger drops have contact
side views do not give the same values as for drops on flat surfaces, angles exceeding 150° indicating superhydrophobic behavior.
476 G. Viswanadam, G.G. Chase / Journal of Colloid and Interface Science 367 (2012) 472–477

Fig. 7. DSA photographs of axial views of water droplets of different volumes on 0.156 inch. outer diameter SPP cylindrical surface. (A) 4 ll (r = 0.985 mm), (B) 5 ll
(r = 1.061 mm), (C) 6 ll (r = 1.128 mm), (D) 7 ll (r = 1.187 mm). The images are at different magnifications and hence are not comparable in scale.

Fig. 8. DSA photographs of side view of water droplets of different sizes on cylindrical surfaces R = 1.98 mm (A) 4 ll (r = 0.985 mm), (B) 5 ll (r = 1.061 mm), (C) 6 ll
(r = 1.128 mm), (D) 7 ll (r = 1.187 mm). The images were at different magnifications and hence are not comparable in scale.

trend of increasing contact angles with increasing drop to cylinder
radius ratio. However, unlike the data plotted in Fig. 2, the data for
different cylindrical radii do not collapse into one curve because
the flat surface contact angles vary with drop radius, hence the
contact angles are dependent on both drop size (r) and the radius
ratio (r/R). Unlike Fig. 2, the curves in Fig. 9 calculated for Eq. (9)
are not extended to zero drop radius because the flat surface con-
tact angles for the smaller drop sizes are not known.
5. Experimental challenges
Measurements on curved surfaces were experimentally
challenging. One challenge was to attach drops to the superhydro-
phobic surface. The most common way to attach drops of a mea-
sured volume to a surface is to produce the drop on the tip of a
syringe needle and to touch the drop to the surface. For most mate-
rials the drop detaches from the needle and attaches to the surface
 G. Viswanadam, G.G. Chase / Journal of Colloid and Interface Science 367 (2012) 472–477 477

to study the effect of curvature on the contact angle. The effect of

liquid drop volume on the contact angle for a curved surface can
also determined. Eq. (9) can be useful to estimate the effects of sur-
face curvature on apparent contact angles.
θ

Acknowledgments

This work was financially supported by the Coalescence Filtra-

tion Nanofibers Consortium: Donaldson, Ahlstrom, Parker Hanni-
fin, Cummins Filtration, Elmarco, Hollingsworth and Vose, and
SNS Nanofiber Technology.
Fig. 9. Comparison of WCA on SPP cylindrical surfaces calculated using Eq. (9) and
experimentally measured. ‘R’ is the radius of the cylindrical surface and ‘r’ is the References
radius of the drop.
[1] X. Zhang, F. Shi, J. Niu, Y. Jiang, Z. Wang, J. Mater. Chem. 18 (2008) 621.
[2] M. Nosonovsky, B. Bhushan, Curr. Opin. Colloid Interface Sci. 14 (2009) 270.
due to the attractive forces between the drop and the surface being [3] X. Yao, Y. Song, L. Jiang, Adv. Mater. 23 (2011) 719.
[4] C. Dorrer, J. Ruhe, Soft Matter 5 (2009) 51.
larger than the attractive forces between the drop and the needle. [5] C.R. Crick, I.P. Parkin, Chem. Eur. J 16 (2010) 3568.
However, with superhydrophobic surfaces the smaller drops pre- [6] L.M. Alarcon, D.C. Malaspina, E.P. Schulz, M.A. Frechero, G.A. Appignanesi,
ferred to stay attached to the needle. Only when the drop exceeded Chem. Phys. 388 (2011) 47.
[7] C. Choi, Y. Yoon, D. Hong, K.S. Brammer, S. Karla, K. Noh, Y. Oh, S. Oh, F.E. Talke,
a certain volume (about 4 ll) could the drop be shaken off of the S. Jin, Electron. Mater. Lett. 6 (2010) 59.
tip of the needle with help of gravity and inertial forces. A second [8] C. Shin, G.G. Chase, AIChE J. 50 (2004) 343.
challenge was to obtain symmetry of the drop about the vertical [9] X.H. Xu, Z.Z. Zhang, J. Yang, X.T. Zhu, Colloids Surfaces, A: Physicochem. Eng.
Aspects 377 (2011) 70.
axis shown in Fig. 1. Once a water drop was positioned on the [10] B. Grignard, A. Vaillant, J. de Coninck, M. Piens, A.M. Jonas, C. Detrembleur, C.
cylindrical surface, the orientation of the cylinder had to be rotated Jerome, Langmuir 27 (2011) 335.
to orient the vertical axis through the center of the liquid drop, [11] J.W. Gibbs, On the equilibrium of heterogeneous substances, The Collected
Works of J. Willard Gibbs, vol. 1, Yale University Press, New Haven, CT, 1961, p.
without the drop rolling off. This was difficult for the larger drops 326.
because the WCA increases with drop size on superhydrophobic [12] J.F. Oliver, C. Huh, S.G. Mason, .J. Colloid Interface Sci. 59 (1977) 568.
fiber surfaces. Hence, many experimental attempts were made to [13] C.W. Extrand, S.I. Moon, Langmuir 24 (2008) 9470.
[14] N. Bhardwaj, S.C. Kundu, Biotechnol. Adv. 28 (2010) 325.
obtain the images shown in Figs. 5, 7 and 8.
[15] D.H. Reneker, A.L. Yarin, Polymer 49 (2008) 2387.
[16] G.C. Rutledge, S.V. Fridrikh, Adv. Drug Deliv. Rev. 59 (2007) 1384.
6. Conclusion [17] K.H. Lee, O. Ohsawa, K. Watanabe, I.S. Kim, S.R. Givens, B. Chase, J.F. Rabolt,
Macromolecules 42 (2009) 5215.
[18] J.S. Varabhas, G.G. Chase, D.H. Reneker, Polymer 49 (2008) 4226.
The derived Eq. (9) agrees reasonably well with the experimen- [19] X.M. Li, D. Reinhoudt, M.C. Calama, Chem. Soc. Rev. 36 (2007) 1350.
tal data for both cylindrical and spherical surfaces and can be used [20] R.J. Good, M.N. Koo, J. Colloid Interface Sci. 71 (1979) 283.