You are on page 1of 8

COLLOIDS

AND
',""~iioids and Surfaces SURFACES A
ELSEVIER A: Physicochemical and Engineering Aspects 136 ( 1998 ) 81 88

Line tension effect on contact angles:


Axisymmetric and cylindrical systems with
rough or heterogeneous solid surfaces
Abraham Marmur *
Department of Chemical Engineering, Technion - Israel histitute of Technology, 32000 Ha(fa, Israel
Received 9 January 1997; accepted 5 September 1997

Abstract

Equations for the apparent, intrinsic, and actual contact angles are presented, which account for the effect of line
tension in axisymmetric and two-dimensional (cylindrical) systems with rough or chemically heterogeneous solid
surfaces. The central point in the present theory is the realization thai line tension cannot be considered constant.
The dependence of line tension on the geometry of the system may be extremely complicated. Therefore, the simplifying
assumption that line tension is a function only of the position of the contact line is made. The physical meaning of
this assumption is that line tension affects only the contact angle and not the shape of the liquid-fluid interface. In
two-dimensional (cylindrical) systems, this geometrical dependence is responsible for the existence of a line tension
effect, in contrast to the currently held view. The implications of the theory are discussed in terms of the relationships
between the Young's, actual, and intrinsic contact angles. ~! 1998 Elsevier Science B.V.

Kevwords: Line tension: Contact angle: Wetting

I. Introduction the apparent contact angle, and the actual contact


angle. The tbrmer is the angle measured between
Interfacial phenomena in solid-liquid-fluid sys- the nominal solid surface (which, under low magni-
tems are usually described in terms of contact fication, appears to be smooth) and the liquid-fluid
angles. Depending on the nature of the solid interface [Fig. l(b)]. The latter is the local angle
surface, various definitions of contact angles have between the actual solid surface and the
to be considered. The intrinsic contact angle is the liquid-fluid interface [Fig. l(c)]. Under some cir-
angle between an ideal, i.e. smooth, chemically cumstances, the actual contact angle equals the
homogeneous, rigid, insoluble, and non-reactive intrinsic one; however, this is not always the case.
solid surface and the liquid-fluid interface. In The first expression proposed for the intrinsic
reality, ideal surfaces hardly exist, and most solid contact angle is the well-known Young's equation
surfaces are rough and heterogeneous to some [1]. Nonetheless, it is well established that this
extent. Consequently, there is a need to distinguish equation needs to be corrected for the effect of the
between two additional types of contact angles: three-phase molecular interactions at the contact
line region [2-14]. This is so, since the presence
* Corresponding author. Fax: + 972-4-8293088:
of a third phase in the close vicinity of each inter-
e-mail: marmurt2i!tx.technion.ac.il face may augment the interfacial energies. One

0927-7757/98 S 19.00 ~,c~1998 Elsevier Science B.V. All rights reserved.


PII S0927-7757197 )01)300-2
82 A. Marmur / Colloids Surfaces A: Physicochem. Eng. Aspects 136 (1998) 81-88

Y this idea, the following equation was suggested for


the intrinsic contact angle, 0i, of an axisymmetric
drop on an ideal solid surface [3,4,9]:

...... I " - l cos Oi =cos Oy - ~ ( 1)


O'lfR
(a)
In this equation, 0v is Young's contact angle, ~ is
the line tension, a~r is the If interfacial tension, and
R is the radius of the base of the drop.
Various attempts have been made to either esti-
mate or measure line tension, in order to quantify
the significance of its effect [ 14-24]. Some of these
values differ by as much as about five orders of
Co) magnitude from each other. The main reason for
these huge differences seems to be related to lack
of sufficient theory for proper interpretation of
experimental results. Based on a recent estimate
[ 14], which agrees with some of the measurements,
the line tension effect is important only when the
radii of curvature in the system are of the order
of magnitude of microns or less. In other words,
(0
it seems to be important for micron size (or
Fig. I. The geometry of the system: (a} definition of phases and smaller) drops as well as for larger drops with
coordinates; (b) the apparent contact angle; (c) the actual con- local, micron size radii of curvature due to rough-
tact angle.
ness or heterogeneity of the solid surface. From a
practical point of view then, the effect of line
approach that was studied in order to develop the tension may be mostly important for wetting of
necessary correction was to assume the interfacial rough or heterogeneous surfaces. However, theory
tensions to depend on the local thickness of the has so far been concerned only with smooth and
liquid layer and on the local slope of the homogeneous surfaces.
liquid-fluid (If) interface [ 12, 13]. This approach Moreover, the line tension in Eq. ( 1) is consid-
led to a very complicated equation for the shape ered constant. This assumption is unrealistic [14],
of the drop, and to an implicit, but relatively as explained in the following. As mentioned above,
simple, equation for the intrinsic contact angle. line tension is introduced in order to account for
Another approach, which was developed many the effect of the presence of a third phase on each
years ago [3,4,9,16], followed a suggestion by interfacial tension. This correction may, in prin-
Gibbs [2] to introduce the concept of linear (line) ciple, be calculated from the integral of all the
tension. The purpose is to account for the differ- molecular interactions of all the molecules in the
ence between the internal energy of the actual three phases. It is quite obvious that this integral
system and that of a simplified model of the must depend on the spatial distribution of the
system, for which each interfaciai tension is consid- three phases. Therefore, in general, Sine tension
ered to be unaffected by the presence of the must depend on the geometries of the solid-liquid,
adjacent third phase. Line tension is then defined liquid-fluid and solid-fluid interfaces, as well as
as this difference, per unit length of the three- on the contact angle. However, if such a general
phase contact line. In other words, the line tension dependence of line tension on the geometry of the
concept can be considered as the one-dimensional system were introduced into the fundamental equi-
analog of the surface tension concept. Based on librium condition, the result would be an extremely
A. Marmur / ColloMs Surfatz, s A: Physicochen~. Eng. Aspects 136 (1998) 81--88 83

complicated integral-differential equation. Such in comparison with Eq. (!), to account for the
an equation will not be analytically solvable, and dependence of line tension on the contact angle.
will most likely be of limited value both in terms The combination of Eqs. (2) and (3) leads to an
of added insight and from a practical point of view. implicit equation for the intrinsic contact angle
Consequently, it makes sense to study the prob- [141.
lem under some simplifying assumptions, though As mentioned above, corrections to Young's
less extreme than assuming constant line tension. equation seem to be of practical importance mainly
The fact that the line tension approach enables for systems in which the liquid contacts rough or
various levels of approximation, as discussed in heterogeneous surfaces. Therefore, the purpose of
the following, constitutes a practical advantage the present paper is to develop an equation that
over the variable interfacial tension approach accounts for the effect of line tension on contact
which was previously mentioned [i 2,13]. Recently, angles in such systems. Since it is important to
an approximate theory has been developed for distinguish between the intrinsic, actual, and
calculating line tension in the specific case of an apparent contact angles in these systems [25], the
axisymmetric drop on an ideal solid surface, in the following discussion addresses the relationship
absence of gravity [14]. This theory assumes that between these angles when line tension effects are
the shape of the drop is not affected by the meaningful. Due to mathematical difficulties, the
existence of line tension, and calculates line tension present treatment refers only to axisymmetric or
directly from the van der Waals interaction two-dimensional (cylindrical) systems.
between the solid and the liquid. It leads to the
following approximate equation for the line ten-
sion in a solid-liquid-vapor system (which is valid 2. Theory
for contact angles that are not too low and not
too close to 90 ~): In general, equilibrium is achieved when the
internal energy of a global isolated system (which
r = 46V~s~rl cot 0i (2) in the present case consists of the solid-liquid-fluid
In this equation, 6 is a typical distance between system together with the environment) reaches its
molecules, trt is the surface tension of the liquid, minimum. From this general requirement, three
and trs is the surface tension of the solid. Eq. (2) equilibrium conditions emerge: (a) uniformity of
demonstrates that line tension is a function of the temperature throughout the system; (b) uniformity
intrinsic contact angle, as expected from the gene- of the generalized chemical potentials; and (c) the
ral considerations mentioned above. In this case, mechanical equilibrium condition, from which the
the intrinsic contact angle is the only geometrical equations for the shape of the If interface and for
parameter that matters, since the shape of the drop the contact angle are derived (e.g. Ref. [25]). The
is a priori fixed. latter takes the following form when line tension
When Eq. (2) was introduced into the expression is introduced to account for the three-phase inter-
for the energy of the ~ystem, and the latter was actions at the contact line:
minimized to get the expression for the intrinsic
contact angle, the following equation resulted [ 14]: .'~i ( P f - PI)V~(Y')'dNixk dx
r ~(~2 +3) dr
cos 0i = cos 0v - - - + ---
o'lfR trlfR( 1 + ~2) d:~
(3)
+d "
I
+~lWl+(Y')i+[as,(x)-Gs~x)]r(x) }
where
x x k dx+[d(Xkr)].~=x2-[d(Xkz)]x=x=O (5)
1- cos 0 i
~_- (4) In this equation, the subscripts s, 1, f indicate the
sin 0i solid, liquid, and fluid phase, respectively, P~ ( j =
Eq. (3) shows that an additional term is needed, f,l ) is the pressure in phase j, dNt is the displace-
84 A. Marmur / ColloMs Surfitces A: Physicochem. Eng. A,Tmcts 136 (1998) 81-88

ment of the If interface normal to itself in the tions with respect to the x-axis for the same
outward direction from phase 1, and aj,, (j,m = variation in the geometry near the contact line. In
s,l,f) are the interfacial tensions. The local rough- the axisymmetric case the fourth term is zero.
ness ratio, r(x), is defined as the local ratio of the As mentioned in the Introduction, these terms
actual solid surface area to its projection on the should, in principle, depend on the spatial distribu-
horizontal plane. For a two-dimensional system tion of the three phases; however, admitting such
(cylindrical If interface), which is indicated in the a dependence will lead to severe mathematical
equation by setting k = 0 , x is the horizontal difficulties. Therefore, the central assumption of
Cartesian coordinate, Xt and X., (X2 > XI) are the the present study is that the geometrical depen-
x values corresponding to the contact lines. For dence of line tension involves only quantities which
an axisymmetric system, which is indicated by are defined at the contact line, such as the actual
setting k = 1, x is the radial horizontal coordinate, contact angle. Thus, for example, r can be
Xt =0, and X, is the x value of the contact line. expressed as r[y'(X)] [for a given geometry of the
In both cases, y is the vertical coordinate of the If solid surface, J's(x), the actual contact angle can
interface, and the prime indicates derivation with be written in terms ofy'(X)].
respect to x [the geometry of the system is shown However, even this simplified dependence of line
in Fig. I(a)]. tension leads to mathematical difficulties. This is
The first integral in Eq. (5) accounts for the so, since the variations considered in order to
work done by the pressure difference between the search for the minimum in energy include varia-
liquid and fluid phases due to the perturbations in tions of the shape of the If interface, in addition
the position of the If interface. It is important to to variations in the position of the contact line.
realize that the values of the pressures which Consequently, the coordinate of the contact line,
appear in Eq. (5) alread2y include the effect of X, cannot by itself uniquely determine the slope.
gravity: Only after the shape is known will y'(X) be
uniquely defined (for example, if it is known that
P~=Pjo-P~gY ( j = f,l ) (6)
the shape is circular and tile volume of the liquid
In this equation, Pjo is the pressure that exists or is given, then the coordinate X uniquely determines
would have existed in phase j at v=0, g is the the details of the If interface). Therefore, an addi-
gravitational acceleration, and pj is the density of tional simplification is made: line tension is
phase j. assumed to be a function of the position of the
The second integral in Eq. (5) expresses the contact line only:
interfacial energies. If the solid surface is chemi-
cally heterogeneous, the interfacial tensions that r=r(X) (7)
involve the solid, i.e. a~ and a,,r, depend on x. The The physical interpretation of this assumption is
_+ sign accounts for the fact that, depending on that line tension affects only the contact angle and
the inclination of the If interface, an increase in not the shape of the If interface.
the If interfacial area may be associated with either The first two terms of Eq. (5) may be treated as
an increase or a decrease in x. previously reported [25] to yield:
The third and fourth terms in Eq. (5) represent
the line tension correction. They replace the gene-
ral term d ~ ~r [12,13], which is the differential of
the excess internal energy of the system due to the
d {[ ]
A P t , - t~v2Apg + trlrV ! + (.v') 2

three-phase molecular interactions. This excess + [rr~l(x)- o',,r(x)lrlx)} v k d.,t-


energy is expressed here, by definition of the line
tension concept, as the product of the line tension
and the length of the co~atact line. These terms + [d(X~r)]., .~x: - [d(Xkr)]., = x, = 0 (8)
have opposite signs, since the two contact lines (in where AP-Pr-PI, v-y(x)--ys(x),y,(x)is the
the two-dimensional case) move in opposite direc- profile of the solid surface [see Fig. l(a)] and
A. Marmur / Colloids Surfaces A: Physicociwm. Eng. Aspects 136 (1998) 81:88 85

A p - p , - pf. Eq. (8) is valid for the general case of In these equations [25]:
a volatile liquid in contact with a rough and/or
chemically heterogeneous solid surface. F = {APv--~v2Apg 4._5"
alfV 1 + (),')z
The minimization of the integral in Eq. (8),
without considering line tension terms, was pre-
viously performed by standard techniques of calcu- + --asr(x)]r(x)}x k (11 }
lus of variations [25,26]. The result was a system
of two independent equations: (a) the Euler equa- In the absence of constraints on the motion of the
tion, which transforms into the well known contact lines, dXt is independent of dX2, and so
Young-Laplace equation for the shape of the If are the d[(Xkr)].,: x terms. Therefore, each pair of
interface; and (b) the transversality condition, terms in Eq. (10) that relates to one of the contact
which accounts for the fact that the position of a lines independently equals zero. This can be written
contact line is a priori unknown, and yields the in the following form:
equation(s) tbr the contact angle(s).
However, the present discussion focuses on the
F + (y~ - y ' ) d X + d [Xkz(X)]., =x = 0
line tension effect, therefore the minimization needs t'l' dx = x
to be extended to include the line tension terms in
Eq. (8). Since these terms depend only on the (12)
position(s) of the contact line(sL as stated in This equation yields the expression for the appar-
Eq. (7), they simply need to be added to the ent contact angle, 0. As mentioned in the
transversality condition, and do not affect the Introduction, 0 is defined as the angle measured
Euler equation. This is so, because the transversal- from the x direction to the tangent to the If
ity condition includes all the terms that deal with interface which is normal to the contact line [see
the contact line(s). This conclusion can be formally Fig. 1(b)]. In the unconstrained case:
proven by repeating the development of the Euler
equation and the transversality condition [25,26], cos 0 =
however, the details are omitted here since it simply V I + [.),'(x)] 2
involves adding the line tension terms throughout
k~ d'r,
all the steps of lhe development. From a physical r(X) cos Or(X)
point of view this is clearly understandable, since ),~lr alctX
(13)
the basic assumption is that the line tension l + y~(X)y'(X)
depends only on the position of the contact line
and has therefore no effect on the shape of the If Eq. (13) is the most general equation under the
interface. above-stated assumptions. A specific example
Thus, the Euler equation remains intact: which was previously studied [14] is represented
by Eq. (3), which can be directly derived from
d t~F t'.'[" Eq. ( 13 ) under the conditions of axisymmetry and
=0 (9) a smooth surface. Eq. (13) is relatively simple in
dx ?y' Pv form, although it should be realized that it is an
implicit equation (on the right-hand side there are
and the transversality condition for the present three terms which depend on the contact angle:
case is the two terms in the nominator that include line
tension, and the term in the denominator that
includes v ' = t a n 0). It should also be mentioned
[F+(.v~-.v')g.v J.~:x, dX, + d[Xkr(X)].,= x,_ that Eq. (13) can easily be generalized to include
the effects of constrained contact lines [25], how-
ever this will not be shown here since it is not the
-[F+ty;-y') ?vJ,:: X!
dXl - d[Xkr(X)].,: x,(I0)
main point of interest in the present discussion. It
86 A. Ahmmw Colh~MsSm:hwes A." Phvsicochcm. Eng. A.wects !36 (1998) 81-88

is also worth mentioning that the definition used normal to the contact line [ Fig. I (c)]. The intrinsic
here for the apparent contact angle follows the contact angle is the actual contact angle on an
mathematical convention for defining angles, and ideal solid surface. It has recently been concluded
is somewhat different from the conventional defi- [25,27] that if line tension effects are ignored, and
nition used in the contact angle literature [25]. in the absence of constraints, the actual contact
angle equals the intrinsic one. This conclusion
needs to be tested for the effect of line tension.
3. Discussion The actual contact angle is, in fact, the angle
measured from a line of slope ),~ (the local tangent
As a starting point for the discussion, Eq. (13) to the solid) to a line of slope y' (the local tangent
for the apparent contact angle should be compared to the if interface). Based on analytical geometry
with Eq. ( 1 }. The latter has been in use for many (regardless of any thermodynamic considerations],
years tbr axisymmetric drops on an ideal solid the actual contact angle is given by
surface (for which the intrinsic contact angle is
I +y'.r~ I +y'y~
also the apparent one). Two main differences are COS 0ac = =
illustrated by this comparison: (a] an additional + V[ l + [.v')2 ][ I + (j'~)2] "!-rX/I + (.v') 2
term is required to account for the fact that the
(14)
line tension is not constant: and (b) correction
factors are required to account for the roughness The second equality in this equation is derived
of the surface. It is easy to show that Eq. (13) from the definition of the local roughness ratio:
turns into Eq. 11) if the surface is smooth ( r - I , r - X/! + (y~)2. Eq. (14) is used below to develop
and .v~=0) and the line tension is constant. It is the thermodynamic expression for the actual con-
interesting to notice that chemical heterogeneity tact angle, in order to compare this contact angle
does not explicitly affect the form of the equation. with the intrinsic one.
Its effect is implicit, through the dependence of the The latter can be derived from Eq. (13) by
line tension itself on the local chemical nature of substituting r = 1 and v~ = 0 (i.e. a smooth surface):
the solid surface. The effect of the geometry of the
kr I dr
contact line itself is explicitly determined by the cos 0 i = cos 0 v ( 15 )
value of k and, in addition, implicitly expressed Xo tr d X
through the dependence o f d r / d X on this geometry.
When Eq. ( 131is introduced into Eq. ( 15j in order
The parallel of Eq. (I) for two-dimensional
to eliminate cos 0v, the following equation results
(cylindrical) systems does not exist in the literature,
for the intrinsic contact angle:
since it has been believed that line tension has no
effect on such systems. In contrast, Eq. (13) shows ! +.v'j'~ I(~,
that line tension may affect two-dimensional (cylin- cos 0, . . . . 7 + -
+rV[l+Iy')2] a
drical ) systems. This is so because of the depen-
dence of line tension on the local position of the (16)
contact line [indeed, Eq. ( 13} shows that if the line
tension is assumed constant, i.e. dr/dX=0, there Since the fiist term on the right-hand side of this
is no line tension effect in two-dimensional sys- equation equals, by Eq. (14), the actual contact
tems). It is worth noticing that this result holds angle, Eq. (16) can be rewritten as
whether the solid surface is ideal or not.
Another important outcome of Eq. (13) is cos 0,.,c = cos 0 i - -- + - 1 ( 171
O" ,
related to the possible difference between the
intrinsic and the actual contact angle. As men- Eq. [17) can now be discussed for a variety of
tioned in the Introduction, the latter is defined as situations. First, as previously concluded [25,27],
the angle measured from the tangent to the solid the actual contact angle equals the intrinsic one
surface to the tangent to the If interface, both when line tensica is ignored (and in the uncon-
,4. Marmur ,, (~dh, ids Smfaces A: Physicochem. Eng. Aspects 136 { 1998) 81.-88 87

strained case, for which these equations apply). In is meaningful [Eq. ( 15)]. It should be empha-
addition, Eq.(17) show~ that these angles are sized that this conclusion is true not only for
always equal when the surface is smooth ( r = 1). axisymmetric systems, as previously believed,
However, the important point to be emphasized is but also for two-dimensional (cylindrical)
that the actual contact angle is different from the systems.
intrinsic one for rough surfaces, if the effect of line (3) The actual contact angle is different from the
tension is meaningful. This conclusion holds for Young's contact angle whenever the line ten-
the axisymmetric case even if the line tension is sion is meaningful [Eq. ( 18)].
constant: it has not been previously noted, since (4) The actual contact angle is different from the
the effect of roughness has not been hitherto intrinsic contact angle whenever line tension
studied in conjunction with line tension. is meaningful and the solid surface is rough
It is also useful to directly compare the actual [Eq. (17)]. If the solid surface is smooth
contact angle with Young's contact angle. For this (regardless of the line tension), or if line ten-
purpose, Eq.(15) needs to be introduced into sion can be ignored (regardless of the rough-
Eq. (17) to yield ness) the actual contact angle equals the
intrinsic one.
cos (l,,¢ - c o s 0 v - - - + ( 181
OT

This equation shows that the actual contact angle Acknowledgment


equals Young's contact angle only if the line
tension can be completely ignored. This conclusion The author thanks Dr. G. Wolansky for a very
holds for ideal as well as for rough or hetero- useful discussion.
geneous surfaces.

References
4. Summary
[I ] T. Young, Philos. Trans. Roy. Soc.(London) 9511805) 65.
The following points summarize time results of [2] J.W. Gibbs, The Scientific Papers of J. Willard Gibbs,
Vol. I, Dover Publications, New York, 1961, p. 288.
the present analysis of line tension effects on [3] V.S. Veselovski, V.N. Pertchov, Zh. Fiz. Khim. 8 (1936)
contact angles, which takes into account a simpli- 245.
fied dependence of line tension on the geometry of [4] B.A. Pethica, Rep. Progr. Appl. Chem. 46 ( 1961 ) 14.
the,solid-liquid-fluid system. All conclusions refer [5] C.A. Miller, E, Ruckenstein, J. Colloid Interface Sci. 48
to axisymmetric as well as two-dimensional (cylin- 11974) 368.
[6] E. Ruckenstein. P.S. Lee, Surface Sci. 52 11975) 298.
drical) systems. For axisymmetric systems, these [71 G.J. Jameson. M.C.G. del Cerro. J. Chem. Soc., Faraday
conclusions also hold under the extreme assump- Trans. 72 (1976) 883.
tion of constant line tension. [81 L.R. While, J. Chem. Soc.. Faraday Trans. 1 73(1977) 390.
( l ) A new equation for the apparent contact angle B.A. Pcthica, J. Colloid Interlace Sci. 62 (1977) 567.
was developed [Eq. ( 13)]. Thi:: equation gener- [t0l E. Ruckenstein, J. Colloid Interface Sci. 86 11982) 573.
alizes previous equations which took into [ill A. Marmur, Adv. Colloid Interface Sci. 19 (1983) 75.
[12l A. Marmur, J. Colloid Interface Sci. 148 (1992) 541.
account either the effects of roughness and [13] A. Marmur, J. Adhesion Sci. Technol. 6 11992) 689.
heterogeneity of the solid surface or the effects [141 A. Marmur, J. Colloid Interlace Sci. 186 (1997) 462.
of a constant line tension. The new equation [151 W.D. Harkins, J. Chem. Phys. 5 (1937) 135.
includes all of these effects as well as the {16l J.S. Rowlinson and B. Widom, Molecular Theory of
Capillarity. Oxford University Press. Oxford, 1984.
observation that line tension cannot, in gene- [171 B.V. Toshev, M.Z. Avramov, Colloids and Surfaces A 100
ral, be constant. 1995) 203.
(2) The intrinsic contact angle is different l¥om [18] A. Scheludko. B.V. Toshev, D.T. Bojadjiev, J. Chem. Soc,,
Young's contact angle whenever line tension Faraday Trans. I 72 (1976) 2815.
88 A. Marmur / Coil,lids Surfitces A: Physicochem. Eng. Aspects 136 (1998) 81-88

[19] R.J. Good, M.N. Koo, J. Colloid Interface Sci. 71 [23] J. Drelich, J.D. Miller, J. Colloid Interface Sci. 164
(1979) 283. { 1994) 252.
[20] J. Gaydos, A.W. Neumann, J. Colloid Interface Sci. 120 [24] D. Duncan, D. Li, J. Gaydos, A.W. Neumann, J. Colloid
(1987) 76. Interface Sci. 169 (1995} 256.
[21] D. Li, A.W. Neumann, Colloids and Surfaces 43 {1990} [25] A. Marmur, Langmuir 12 (1996) 5704.
195. [26] F.B. Hildebrand, Methods of Applied Mathematics, 2rid
[22] L. Boruvka, J. Gaydos, A.W. Neumann, Colloids and edn., Prentice-Hall, Englewood Cliffs, N J, 1968, Chap. 2.
Surfaces 43 ( 1990} 307. [27] G. Wolansky and A. Marmur, in preparation.

You might also like