Langmuir 1994,10, 1777-1779 1777

Thermodynamics of Microemulsions Revisited

Eli Ruckenstein
Department of Chemical Engineering, State University of New York at Buffalo,
Buffalo,New York 14260
Received January 18,1994. In Final Form: March 24,1994"

By comparing two different thermodynamic formulations, which consider a microemulsion either as a

multicomponentmixture or as a dispersionof globules in a continuousmedium, one demonstrates, because
they must be equivalent, that (i) the pressure in the continuous medium of a microemulsion differs from
the external pressure and (ii) the Laplace equation for a microemulsion contains an additional term due
to the entropy of dispersion of the globules in the continuous phase. On this basis, a choice is made
between two thermodynamicapproachesto microemulsions. The derived equations explain the occurrence
of a microemulsionphase in equilibrium with an excess dispersed phase, as well as the equilibriumbetween
a microemulsion phase and the two excess phases, and the change in the structure of the microemulsion
at the transition from two to three phases. The thermodynamic equations suggest that the middle-phase
microemulsion can be described by the zero mean average curvature of the interface between the two
media.

The thermodynamics of microemulsions was treated in
somewhat different ways by Ruckenstein' and by Overbeek
et a1.2 In what follows, the approach of Ruckenstein will
be denoted by I and that of Overbeek et al. by 11. The
scope of this paper is to make a choice between the two.
In both approaches the Helmholtz free energy F of a
microemulsion is decomposed in the sum of a frozen free
energy Fo, which does not include the entropy of dispersion
of the globulesin the continuous medium, and a free energy
AF due to the latter entropy of dispersion. The following
main differences exist between the two: (i) In treatment
I one demonstrates that the pressure in the continuous
phase of the microemulsion differs from the external
pressure, while in treatment I1 it is equated to the latter.
(ii) In treatment I a modified Laplace equation, which
contains, in addition to the usual terms, one of entropic
origin, is derived. In treatment I1the conventional Laplace
equation valid for a single droplet in a continuum is
extended to microemulsions.
The above differences have implications, since they
affect the expressions of the chemical potentials and hence
the conditions of equilibrium between a microemulsion
phase and the excess dispersed phase, and between a
microemulsion phase and both excess phases. In fact, only
because the pressure in the continuous medium differs
from the external pressure can treatment I explain the
occurrence of the latter equilibrium as well as the change
in the structure of the microemulsion associated with the
transition from two to three phases. The choice between
the two approaches will be made on the basis of formal
thermodynamics, by comparing two equivalent thermo-
dynamic equations: one in which the microemulsion is
considered a single multicomponent mixture phase, and
another one which models a microemulsion as a dispersion
of globules of one medium in another medium.
From the point of view of traditional thermodynamics,
a microemulsion is a single phase which consists of the
components oil, water, surfactant, cosurfactant, and an
@Abstractpublished in Advance ACS Abstracts, May 15, 1994.
(1) Ruckenstein, E. Ann. N. Y. Acad. Sci. 1983,404,224;Fluid Phose
Equilib. 1985,20,189. For a review, see: Ruckenstein,E. In Progress in
Microemdsions;Martellucci,S., Cheater,A. N., Eds.;PlenumPresa: New
York, 1989, based on the proceedings of Progress in Microemulsions, Oct
26 to Nov 1, 1985, Erice, Italy; pp 3-29.
(2) Overbeek,J. Th. G.;Vemoeckx, G. J.;DeBruyn,P. L.;Lekkerkerker,
H. N. W. J. Colloid Interface Sci. 1987,119, 422.
electrolyte. The change a t constant temperature of the
Helmholtz free energy F of such a system can consequently
be written as
where pi* and Ni are the chemical potential and the number
of moles of species i, respectively,p is the external pressure,
Vis the volume of the microemulsion, and the summation
is extended over all the components.
Let us now use a more detailed representation of a
microemulsion, by considering that it contains globules of
water (oil) dispersed in oil (water), and that the other
components are distributed among the dispersed and
continuous media of the microemulsion and the interface
between them. The Helmholtz free energy is the ap-
propriate thermodynamic potential because the pressures
in the globules and in the continuous medium are different.
For the sake of simplicity, the globules are assumed
spherical and of the same radius r. Denoting the volume
fraction of the dispersed phase by 4, the interfacial area
A between the two media per unit volume of microemulsion
is given by
A = 34/r (2)
The Helmholtz free energy of the microemulsion is written
as the sum
F=F,,+AF (3)
where AFis the free energy due to the entropy of dispersion
of the globules in the continuous phase.
For curved interfaces, characterized by the principal
curvatures c1 and c2, the Gibbs thermodynamics3provides
the following expression for d F 0 :
where y is a generalized interfacial tension since it includes
the effect of the interactions among the globules, C1
and Cp are bending stresses associated with the
(3) The Collected Works of J. Willard Gibbs; Yale University Press:
New Haven, CT, 1948; Vol. I, p 225.

0 1994 American Chemical Society

0143-1463/94/2410-~177$04.50/0
1778 Langmuir, Vol. 10, No. 6, 1994 Ruckenstein
curvatures c1 and C Z , pz is the pressure in the globules, p1 Equations 10-12 coincide with those already derived in
is the pressure in the continuous medium, and pi are the a different way.' Because of eq 9d, the chemical potential
chemical potentials of species i at the pressure p1 for the of the continuous medium (oil or water) is equal to the
species present in the continuous medium and at the chemical potential pc (i = c for the continuous medium
pressure pz for those present in the dispersed medium. (oil or water)), which appears in eqs 5 and 7. The latter
For spherical globules chemical potential being expressed a t the pressure p1, one
can conclude that the pressure in the continuous medium
c1 = c, = l / r and C, = C, = C/2 is equal to pl. Equation 12 clearly demonstrates that the
pressure p1 cannot be equated to the external pressure p.
and eq 4 becomes The pressure p is a macroscopic quantity defined on the
scale of a microemulsion (hence on a scale large compared
dF, = y d(AV) + CV d(l/r) + z p i Wi- pzd[V4l- to the size of the globules),which acta over the combination
PI d[V(1- $13 (5) of the two media. In contrast, the pressure p1 acts on the
scale of a globule, being the pressure sensed in the
I t is convenient to introduce the free energy density Af continuous medium by each globule.
due to the entropy of dispersion of the globules per unit Another consequence of the above thermodynamic
volume of microemulsion: treatment concerns the conventional Laplace equation
which is no longer valid for a microemulsion. Indeed,
AF= VAf (6) eliminating the pressure p between eqs 11 and 12, one
obtains
Consequently,
dF = y d(A V) + C V d(l/r) + z p i Wi- p, d[ V41-
d[V(1- 411 + d[VAfl (7)
Denoting by
Since Af depends only on r and 4l

dAf = r$)$+ (3)

dr &#I
the number of globules per unit volume of microemulsion,
one can write that
Performing the differentials in eq 7, one obtains

dF= ( wr- p l ( l - $ ) - p 2 4 + A f ) d V +
[ Wr + p , V - p , V + V(z)]de+ [-y-y+ which, combined with eqs 10 and 13, yields

Expression 8,which is basedon a model, must be equivalent

to expression 1, which is based on the traditional ther-
modynamics of a multicomponent mixture. Consequently,
34ylr - pl(1- 4) - pZ4+ Af = -P
3r+@1-P2)+($$
r =o
and
Equation 9c can be rewritten as
In addition to the first two terms, which appear in the
case of a single droplet, eq 16 contains a third term which
is due to the entropy of dispersion of the globules in the
(gal continuous medium. Ita presence in the modified Laplace
equation is expected, because the virtual change in radius
which is involved in the derivation of the Laplace equation
changes the radius of the globules and hence their entropy
of dispersion in the continuous phase.
Finally, approach I can explain in a simple manner, in
terms of pressures, the phase behavior of microemulsions.
For a microemulsion to coexist with an excess dispersed
phase the chemical potential pd* of the dispersed medium
(water or oil) should be equal to both the chemical potential
pd@2) and that of the excess dispersed phase &e@).
Considering that the compositions of the dispersed
medium and excessdispersed phase are the same, the above
equality leads to

and solving the system of eqs 9a and 9b, one obtains, after
y is replaced with its expression given by eq 10,
and
Pz=P (17)
The expression obtained in treatment I1for the chemical
potential Pd is different from ours because of the incon-
sistencies which exist in ita basic thermodynamic equa-
tions. Similarly, the equilibrium between the continuous
medium and the excess continuous medium leads to

Consequently, a microemulsion will be in equilibrium

Thermodynamics of Microemulsions
with both excess phases when
PI = Pa = P (19)
The condition of equilibrium in eq 19 among the three
phases is responsible for the change in the structure of the
microemulsion from spherical globules to a disordered one
at the transition from two to three phases. This change
occurs because a spherical interface cannot remain stable
to thermal perturbations when the pressures inside and
outside the globules are the same.
The condition of thermodynamic equilibrium in eq 19
leads to the conclusion that the curvature at the transition
from two to three phases is zero.' In the middle-phase
microemulsion, the pressures p2 and p1 fluctuate in time
and space, and intuition suggests to replace eq 19 with its
average and hence to replace the condition of zero
curvature with the condition of zero mean average
curvature. A bicontinuous rigid sponge composed of two
interpenetrating phases can have a zero average curvature?
Langmuir, Vol. 10, No. 6,1994 1779
However, the rigid structure is expected to be unstable,
and a nonrigid sponge may better describe the real
structure.
In conclusion,the choice is in favor of approach I because
(i) the pressure in the continuous medium of a micro-
emulsion is different from the external pressure, (ii) the
conventional Laplace equation is not valid for a micro-
emulsion, and finally (iii) it can explain the occurrenceof
three phases and the change in the structure of the
microemulsion associated with the transition from two to
three phases. Approach I1 considers the pressure in the
continuousmedium equal to the external pressure, employs
the conventional Laplace equation, and cannot predict
the coexistence of three phases.
(4) Neovius, E. R.; Minimalflachen; Frenkel, J. C.; Helsingfors, 1883.
Schwartz, M. A. Gesammette Mathematkche Abhandlung; Springer:
Berlin, 1890,Vol. 1. Scriven, L. E. In Micellization, Solubilization and
Microemulsiom; Mithal, K. L., Ed.; Plenum: New York, 1977.