JOURNAL OF COLLOID AND INTERFACE SCIENCE 187, 29–44 (1997)
ARTICLE NO. CS964507
Dynamic Film and Interfacial Tensions in Emulsion and Foam Systems
YOUNG-HO KIM, KALMAN KOCZO, 1
Chemical Engineering Department, Illinois Institute of Technology, Chicago, Illinois 60616
Received February 27, 1996; accepted July 22, 1996
In concentrated fluid dispersions the liquid films are under dy-
namic conditions during film rupture or drainage. Aqueous foam
films stabilized with sodium decylsulfonate and aqueous emulsion
films stabilized with the nonionic Brij 58 surfactant were formed
at the tip of a capillary and the film tension was measured under
static and dynamic conditions. In the stress relaxation experiments
the response of the film tension to a sudden film area expansion
was studied. These experiments also allowed the direct measure-
ment of the Gibbs film elasticity. In the dynamic film tension
experiments, the film area was continuously increased by a con-
stant rate and the dynamic film tension was monitored. The mea-
sured film tensions were compared with the interfacial tensions of
the respective single air/water and oil/water interfaces, which
were measured using the same radius of curvature, relative expan-
sion, and expansion rate as in the film studies. It was found that
under dynamic conditions the film tension is higher than twice the
single interfacial tension (IFT) and a mechanism was suggested to
explain the difference. When the film, initially at equilibrium, is
expanded and the interfacial area increases, a substantial surfac-
tant depletion occurs inside the film. As a result, the surfactant
can be supplied only from the adjoining meniscus (Plateau border)
by surface diffusion, and the film tension is controlled by the
diffusion and adsorption of surfactant in the meniscus. The results
have important implications for the stability and rheology of foams
and emulsions with high dispersed phase ratios (polyhedral struc-
ture). q 1997 Academic Press
Key Words: film tension; dynamic interfacial tension; foam film;
emulsion film; film elasticity.
INTRODUCTION
Coalescence of emulsions and rupture of foam bubbles
depend on the stability of their intervening thin liquid films.
The action of emulsifiers and demulsifiers, petroleum recov-
ery and refining processes, and defoaming are common prac-
tical examples where the stability of the emulsions and the
foam films under static or rather dynamic conditions has
paramount importance. The rheological behavior of foams
1
Currently with WITCO Corp., OSi Specialties, Inc., 777 Old Saw Mill
River Road, Tarrytown, New York 10591.
2
To whom correspondence should be addressed.
29
AID JCIS 4507 / 6g1e$$$441 02-06-97 09:21:04
AND DARSH T. WASAN 2
and emulsions depends on the response of the thin liquid
films and the Plateau borders during shear and dilation (1).
Several researchers have determined the static and dynamic
interfacial properties of gas/liquid and liquid/liquid inter-
faces in the presence of surfactants using a number of meth-
ods (2, 3). The assumption in these studies is that the behav-
ior of the film can be derived from the properties of its single
interfaces.
A versatile interfacial and film tensiometer has been devel-
oped in our laboratory (4–6). In this technique, a curved,
spherical cap-shaped fluid interface of liquid film is formed
at a capillary tip and the interfacial tension (IFT) of the
single interface or the film tension of the film can be deter-
mined by measuring the capillary pressure of the interface
or film. The IFT or film tension is related to the capillary
pressure and radius of the interface or film curvature by the
Young–Laplace equation. The IFT and film tension can be
measured not only in equilibrium but in dynamic conditions
as well. The automated apparatus makes it possible to change
the interfacial area in virtually any mode (expansion or con-
traction with various rates).
In real emulsions or foams, the size distribution of the
bubbles or drops is generally polydisperse. Hence, the thin
liquid films formed between the bubbles or drops are typi-
cally not flat, as in a homogeneous dispersion (Fig. 1a), but
have a spherical curved shape due to the capillary pressure
difference between bubbles or drops of unequal sizes (Fig.
1b). In foams or in concentrated oil-in-water emulsions, a
polyhedral structure forms as sketched in Fig. 2. The ap-
proaching drop or bubble interfaces hydrodynamically inter-
act, the gravity and the capillary forces deform their inter-
faces and aqueous lamellae form between the bubbles or
drops. Under the action of gravitational and capillary forces,
the aqueous lamellae drain and some of the formed thin
aqueous films rupture. In this way, the size and polydisper-
sity of the bubbles/drops increase and a polyhedral foam or
emulsion (biliquid foam) forms. When the films rupture, the
bubbles or drops coalesce. Note that the thin liquid films in
the polyhedral structure have adjoining menisci, or Plateau
borders. The Plateau borders have a much larger liquid vol-
ume than the films due to the very small thickness of a
0021-9797/97 $25.00
Copyright q 1997 by Academic Press
All rights of reproduction in any form reserved.
coidas
30 KIM, KOCZO, AND WASAN
FIG. 1. Formation and shape of thin liquid films between approaching
bubbles or drops.
typical liquid film. The surface area of this film, however,
is much larger than the area of the meniscus.
The rupture of a film inside the foam or emulsion acts as
a local ‘‘detonation,’’ resulting in very fast expansion in the
nearby lamellae. Moreover, the polyhedral system quickly
rearranges itself to a new mechanical equilibrium after the
film rupture. It can be observed that a new, thick lamella
forms during this rearrangement. At the same time, surfac-
tant diffusion and adsorption start to restore the adsorption
(and film tension) equilibrium in the films which expanded
FIG. 2. Formation of polyhedral structure in foams or concentrated oil-in-water emulsions.
AID JCIS 4507 / 6g1e$$$441 02-06-97 09:21:04
or formed during the rupture and rearrangement processes;
i.e., film tension relaxation takes place (Fig. 3a).
During film drainage, the film continuously and slowly
expands due to the liquid flow inside the film as driven
by the capillary pressure and gravity (Fig. 3b). Similarly,
continuous liquid film expansion takes place in foams or
emulsion under continuous shear or dilation.
In this paper, we report static and dynamic IFT and film
tension measurements with a foam as well as an O/W emul-
sion system stabilized by an anionic and a nonionic surfac-
tant, respectively. The relaxation process after film rupture
in a foam or emulsion system (Fig. 3a) was modeled by
stress relaxation experiments in which the film was first fast
expanded and then kept at a constant size while the dynamic
film tension was monitored. These experiments made it pos-
sible to determine the Gibbs film elasticity as well. The
continuous film expansion process accompanying film drain-
age (Fig. 3b) was modeled by dynamic film tension experi-
ments in which the films were expanded with constant and
slow rate. The film tensions and the IFTs of the respective
single interfaces measured in dynamic conditions were then
compared.
METHODS
In the film studies, a curved, spherical cap-shaped film is
formed (Fig. 4) at the tip of a glass capillary with its menis-
cus adhering to the capillary tip. In this case, it can be
assumed that the inner and outer radii of the film are the
same because the film thickness is negligible (in the range
coidas
 EMULSION AND FOAM SYSTEMS
FIG. 3. Film stress relaxation and film drainage in polyhedral foams or emulsions.
of 50–300 nm) and the film shape is spherical because the
effect of gravity can be neglected due to the very small
volume of the film. In the single interface studies, a drop or
a bubble is formed at the capillary tip along the inner rim
and is immersed into the other phase.
The pressure applied to the curved film can be directly
measured by a sensitive pressure transducer connected to
the capillary. The measured pressure consists of the pressure
difference across the curved film (pi 0 p0 ), which is the
capillary pressure (pc ), and the hydrostatic pressure (pH ):
ptotal Å pc / pH .
The hydrostatic pressure can be measured by measuring the
pressure after the film has ruptured. The film tension ( f )
can be calculated from the measured capillary pressure using
the Young–Laplace equation applied to a liquid film,
2f
pc Å pi 0 p0 Å
Rf
where R f is the film radius.
If chemical equilibrium exists, and the film radius is con-
stant, the film tension remains constant at constant tempera-
ture, corresponding to the concentrations of surfactants pres-
ent in the system. This value is the equilibrium film tension
for this system. In static conditions, for a relatively thick
AID JCIS 4507 / 6g1e$$$441 02-06-97 09:21:04
31
film (above 30 nm thickness) it can be assumed that the
effect of the disjoining pressure is negligible on the film
tension and the film tension is equal to twice the single IFT,
for a symmetrical film, because the film consists of two
single interfaces:
f Å 2s [3]
In dynamic conditions, dilational stresses caused by the film
dilational viscosity during film expansion/contraction can
also affect the capillary pressure and the calculated film
[1] tension (3),
2f 2kf d ln A
pc Å / , [4]
Rf Rf dt
where d ln A/dt is the rate of film expansion and k f is the
film dilational viscosity. If k f is less than 0.1–1.0 g/s (sur-
face poise), which is the case for common emulsion or foam
[2]
films stabilized by small molecule surfactants, the right hand
term of Eq. [4] is negligible (7) and Eq. [2] can be used
under dynamic conditions as well. The effect of gravity on
the sphericity of the film is negligible because the volume
of the film is practically negligible. The size of the film is
varied using a computer-controlled microsyringe attached
to the capillary. In the simplest experimental mode of this
technique, the equilibrium (static) film tension is directly
coidas
32 KIM, KOCZO, AND WASAN
FIG. 4. Principle of studying liquid film formed at the tip of a capillary.
calculated from the equilibrium capillary pressure and film
radius. The direct measurement of the pressure removes the
need for any calibration. Further details of the apparatus
have been published elsewhere (6).
The capillary used in our experiments had an internal
diameter of 0.6 mm and a wall thickness of 0.05 mm. The
small wall thickness was chosen to minimize the volume of
the meniscus because a large meniscus profile could deform
the film (Fig. 4). This way the effect of the meniscus on
the film sphericity was negligible. In the experiments, the
film-to-meniscus surface area ratio was about 8:1 for the
foam film, which is close to the ratio in typical foams (i.e.,
5:1–10:1), and about 2.5:1 for the emulsion film. Note that
the surface area ratio of the foam film was much higher than
that of the emulsion film. In these conditions, the film radius
can be calculated from the inner capillary radius (Rci ) and
the distance between the capillary tip and the film apex, h
(Fig. 4):
R 2ci / h 2
Rf Å . [5]
2h
AID JCIS 4507 / 6g1e$$$441 02-06-97 09:21:04
Figure 5 depicts the procedure for preparing an aqueous
emulsion or foam film. First, the capillary tip was immersed
into a cuvette containing the aqueous phase at the bottom
and the oil or air phase at the top. Then, an oil drop was
formed at the capillary tip. For the study of a single interface,
this oil drop surrounded by the aqueous phase was used
(Fig. 5a). To form an aqueous emulsion film, the level of
the oil/water interface was lowered by slowly removing the
aqueous phase using a syringe (Fig. 5b). As the oil/water
interface passed the tip of the capillary, a thick aqueous
lamella remained there. The aqueous lamella drained due to
the buoyancy force (that is why the capillary is inverted in
Fig. 5) and after a while (in less than an hour) a thin aqueous
film was obtained. After film drainage, the level of the water/
oil interface was lowered again to isolate the film with its
meniscus adhering to the capillary tip (Fig. 5c) and the
isolated thin aqueous film was used for subsequent film rheo-
logical measurements.
In the emulsion film studies, the capillary was filled with
the oil phase (dodecane). In this case, the oil phase inside
the capillary is incompressible. To maximize the sensitivity
of the measurements, the initial film was set to hemispherical
conditions (R f Å 0.3 mm, h Å 0.3 mm in Fig. 4), where
the capillary pressure is at a maximum. In these conditions,
the surface areas of the film and the meniscus were 0.37
mm2 and 0.02 mm2 , respectively with 0.028 mm3 meniscus
volume. The flow rates used to expand the film were in the
7.0 1 10 05 –1.0 1 10 03 mm3 /s range corresponding to 3.5
1 10 04 –6.0 1 10 03 mm2 /s surface expansion rates of one
film interface.
In the foam film studies, the phase inside the capillary
was air which is highly compressible during film expansion
and contraction. Due to the compressibility of the air, the
flow rate of film expansion/contraction is not constant at a
constant feeding flow rate. Beyond the hemispherical posi-
tion, the film expansion rate is higher than the feeding flow
rate due to the expansion of the air inside the capillary as
caused by the decreasing capillary pressure. Preliminary ex-
periments, however, showed that the film expansion is practi-
cally linear if the initial film radius was set to 0.32 mm
(somewhat above hemisphere), which corresponds to h Å
0.44 mm in Fig. 4. In the foam film case the surface area
of the film is 0.70 mm2 , which is a much larger area com-
pared to that of the emulsion film. The rates of surface
expansions for one foam film surface used were in the 0.70
1 10 02 –2.5 1 10 02 mm2 /s range.
MATERIALS
In the foam film studies, an ionic surfactant, sodium decyl-
sulfonate (SDeS) (from Research Plus Inc., Bayonne, NJ)
was used at 3.2 1 10 02 mol/dm3 concentration, which
coidas
 EMULSION AND FOAM SYSTEMS
FIG. 5. Procedure of formation of aqueous foam or oil/aqueous/oil emulsion film.
is slightly above the CMC (3.0 1 10 02 mol/dm3 ) of this
surfactant.
In the studies of an oil-in-water emulsion film (O/W/O
film), a commercial nonionic surfactant, Brij 58, hexadecyl
polyoxyethylene ether with an average of 20 ethoxy groups
(from Sigma), was used. This surfactant, which is highly
water soluble but practically insoluble in oil and has a CMC
around 2.5 1 10 06 mol/dm3 (4), was used without further
purification. The surfactant concentrations in the emulsion
film studies were 2.0 1 10 06 mol/dm3 ( õCMC) and 12
CMC. It is known that this surfactant shows diffusion con-
trolled kinetics (6, 8) in water/oil immiscible systems. n-
Dodecane (Gold Label, from Aldrich) was used as the oil
phase. The oil was purified by treating it with activated
charcoal for several days. The water phase used was deion-
ized water. Purity of the water and the oil was ascertained
by checking the water/oil IFT. Prior to measurements, the
aqueous and oil phases were preequilibrated.
Note that the ionic surfactant (sodium decylsulfonate)
used in the foam film study has about four orders of magni-
tude higher CMC than the nonionic surfactant (Brij 58) used
in the emulsion film study. Also, the concentration of the
ionic surfactant studied in the foam film was much higher
than the nonionic surfactant in the emulsion film.
All the measurements were performed at 257C.
RESULTS AND DISCUSSION
As was previously mentioned, the coalescence process of
a polyhedral lamella in emulsions or foams can be divided
AID JCIS 4507 / 6g1e$$$441 02-06-97 09:21:04
33
into two processes: film drainage and film rupture. To char-
acterize the film rupture process, film stress relaxation exper-
iments were conducted. In these experiments, the film was
quickly expanded (as it happens in a real foam or emulsion)
and then the relaxation of the film was measured. To charac-
terize the film drainage process, dynamic film tension mea-
surements were conducted, in which the film was continu-
ously and slowly expanded while the film tension was moni-
tored.
Single interfaces were also studied by forming a drop at
the capillary (Fig. 5a). In order for results to be comparable
to the film studies, the capillary pressure of the drop was
measured under the same experimental conditions (size, ra-
dius of curvature, surface expansion rate, surface expansions,
concentration) as in the respective film experiment.
Equilibrium Film Tension
Prior to the dynamic film experiments, equilibrium film
tensions were measured by forming a film at the capillary
tip and maintaining it at that size. If surface active materials
are present, the film tension and the capillary pressure de-
crease as the surfactants adsorb at the interfaces of the film.
After equilibrium is reached, that is, the capillary pressure
is constant, the equilibrium film tension is calculated from
the recorded capillary pressure and radius of film curvature.
The equilibrium film tensions for the foam and emulsion
films at the given surfactant concentrations are shown in
Table 1. For comparison with the film, equilibrium single
IFTs are also presented, which were measured with the Wil-
helmy plate technique. The table shows that the film tension,
coidas
34 KIM, KOCZO, AND WASAN
in both the foam and emulsion film studies, is equal to two
times the single interfacial or surface tension within experi-
mental error. This result indicates that the film is relatively
thick (above 30–50 nm), so the effect of disjoining pressure
on the film tension is really negligible, as was assumed
in Eq. [3].
Film Stress Relaxation
To model the film rupture process of polyhedral emulsions
or foams, film stress relaxation experiments were carried out
for emulsion and foam films.
Emulsion film. Figure 6 shows a film stress relaxation
experiment with an emulsion film formed from Brij 58 solu-
tion at 12 CMC concentration between dodecane phases.
The film was suddenly expanded by 22% and then the film
size was kept constant. The film stress relaxation curve pro-
vides information about the kinetics of surfactant adsorption
in the film. When the film expands from the initial radius
to a certain radius, the surfactant concentration on the film
surfaces drops and thus the film tension as well as the capil-
lary pressure suddenly ‘‘jumps’’ to a high value. As a result,
the surfactant inside the film starts to adsorb and thus, the
capillary pressure and film tension decrease until equilibrium
is restored. Figure 6 also shows that the reproducibility of
the film stress relaxation experiments was good.
In Fig. 7 film stress relaxation curves are plotted for emul-
sion films of the same system as in Fig. 6 using various
initial film expansions. Both the initial film tension and the
initial rate of relaxation (initial slope of the relaxation curve)
increase with relative film expansion, but the relaxation
curves have similar shapes and all of them finally reach the
same equilibrium film tension.
Significance of surfactant depletion in the meniscus.
Note that in our film experiments the film formed at the
capillary tip is isolated from the rest of the aqueous phase
(Fig. 5c) and thus the total liquid volume and the amount
of surfactant in the film and in the adjoining meniscus are
constant during the film experiment. When the film surface
area is increased in the stress relaxation experiment, the
moles of surfactant adsorbed at equilibrium on the film inter-
TABLE 1
Equilibrium Film Properties and IFTs of Some
of the Systems Studied
Foam film Emulsion film
Surfactant and concentration
(mol/dm3) SDeS, 3.2 1 1002 Brij 58, 2.0 1 1006
Film tension, f, (mN/m) 75.2 { 0.3 27.8 { 0.3
IFT (mN/m) 37.5 { 0.2 13.8 { 0.2
2 1 IFT (mN/m) 75.0 { 0.3 27.6 { 0.3
AID JCIS 4507 / 6g1e$$$441 02-06-97 09:21:04
FIG. 6. Reproducibility of the film stress relaxation experiment for the
aqueous emulsion film stabilized with 3.0 1 10 05 mol/dm3 (12 CMC)
Brij58.
faces increase. The increased surfactant adsorption results
in a decrease in the bulk of the film and its meniscus, that
is, a surfactant depletion takes place. The relaxation curves
of the various film sizes (relative expansions) in Fig. 7 have
similar shapes and reach the same equilibrium film tension.
This result implies that the surfactant depletion in the bulk,
i.e., in the meniscus, was negligible in the conditions of our
experiments. (The increased surface area must have resulted
in the dissociation of micelles into monomers and thus the
concentration of micelles decreased without, however, caus-
ing measurable decrease in the monomer concentration.) The
significance of depletion in the film and meniscus was also
checked by another experiment, as shown in Fig. 8. The
film containing 12 CMC Brij58 was first equilibrated, then
abruptly expanded with a 10.2% relative film expansion,
and then the relaxation of the film was measured. After
equilibrium had set in at the expanded size, the film was
expanded again with the same expansion ratio (10.2%) and
the relaxation was measured again. These relaxation experi-
ments were repeated another two times with the same expan-
sion on the same film. Figure 8 shows that the relaxation
curves of the four experiments were identical within experi-
mental error. This result again shows that the change in bulk
surfactant concentration due to increased surfactant adsorp-
tion was negligible under the surfactant concentration and
film geometrical conditions of our experiments. The volume
and area of the meniscus do not significantly change when
the film is expanded in these experimental conditions be-
cause the volume of the film is negligibly small compared
to the meniscus volume. In our experiments, assuming a 100
nm film thickness, the volume of the film was about 6 1
10 05 mm3 and the volume of the meniscus was about 5 1
10 03 mm3 . Thus, when the film area was expanded by 30–
50%, the volume of the film increased by about 1.5 1 10 05
to 3 1 10 05 mm3 assuming a constant film thickness (100
coidas
 EMULSION AND FOAM SYSTEMS 35

FIG. 7. Stress relaxations at various relative film expansions ( D A/A) of emulsion films stabilized with 3.0 1 10 05 mol/dm3 (12 CMC) Brij58.

nm) resulting in a negligible, 0.3–0.5% decrease in the me-
niscus volume.
Foam film. In the foam film experiments, an anionic
surfactant at higher molar concentration (10 3 to 10 4 times
higher than in the emulsion film studies) was used. Due to
the high surfactant concentration, the surfactant concentra-
tion in the meniscus can be expected to decrease only negli-
gibly after film expansion. The relaxation curves of a foam
film formed from 3.2 1 10 02 mol/dm3 sodium decylsulfo-
nate are shown in Fig. 9. Various relative film expansions,
starting from the same initial, equilibrium state were used.
Similar to the emulsion film experiments, the relaxation
curves with different expansions have similar shape and all
of them approach the same equilibrium film tension. This
indicates again that there is indeed no measurable depletion
in the meniscus. Figure 9 also provides information on the
stability of the stressed film. Higher expansion (higher
stress) resulted in lower film stability indicated by the film

FIG. 8. Effect of film area on film stress relaxation for repeated consecutive film expansions at a constant relative film expansion for an emulsion
film, in the presence of 3.0 1 10 05 mol/dm3 (12 CMC) Brij58.

AID JCIS 4507 / 6g1e$$$441 02-06-97 09:21:04 coidas

36 KIM, KOCZO, AND WASAN
FIG. 9. Film stress relaxations at various relative film expansions for foam film stabilized with 3.2 1 10 02 mol/dm3 sodium decylsulfonate.
rupture time. The easier film rupture at higher stress was
probably caused by the lower surface concentration on the
film surface after expansion and the larger film area after
the film is expanded (9). The larger film area has a higher
probability of rupture at a given mechanical shock produced
by self-vibration or mechanical disturbance.
Comparison of film tension and IFT of single interface.
In order to investigate if a foam or emulsion film behaves
in dynamic conditions simply as the sum of the two film
interfaces, stress–relaxation experiments were also con-
ducted with single interfaces (Fig. 5a) under the same condi-
tions (same surfactant concentration, expansion ratio, radius
of capillary and interface curvature), as in the respective
film experiments.
Figures 10–12 show comparisons of stress relaxation of
the film tension as well as two times the respective single
IFT in the emulsion system with 12 CMC Brij58 at various
initial expansions. As was shown in the previous section, at
equilibrium, the film tension is equal to two times the IFT
of the corresponding single interface. At a relatively small
expansion (30.2%), the film tension was equal to two times
the single IFT during the entire stress relaxation process, as
shown in Fig. 10. At higher film expansions, as shown in
Figs. 11 and 12, the initial film tension is equal to two
times the single IFT; however, the film tension relaxation is
significantly slower than the relaxation of the single inter-
face. Thus, in dynamic conditions, before the film tension
reaches equilibrium, the film tension is greater than two
times the single IFT. Moreover, higher expansion results in
a greater difference in the rate of relaxation and dynamic
AID JCIS 4507 / 6g1e$$$441 02-06-97 09:21:04
tension between the film and the corresponding single inter-
face. This difference can be expressed in terms of relaxation
time (taken as the time it takes to approach a steady state
closer than {0.3 mN/m): 420 s for the film and 190 s for
the single interface at 38.9% expansion (Fig. 11), and 580
s for the film and 220 s for the single interface at 48%
expansion (Fig. 12), respectively.
For the foam system similar results were obtained. Figures
13 and 14 show comparisons of stress relaxations of foam
film tension and two times the respective single surface ten-
sion, respectively, at various film expansions with 3.2 1
10 02 mol/dm3 SDeS. The film stress relaxation is slower
than the relaxation of the single surface tension and higher
expansion results in larger differences in the rate of relax-
ation between the film and the single surfaces.
Film elasticity. The initial (maximum) film tension ( fi )
after the expansion in the film stress relaxation experiments
can be used to determine the film elasticity. The film elastic-
ity (Ef ) (10) can be expressed as
dfi
Ef Å 0 , [6]
d ln Geq
where Geq is the surfactant excess on the film interfaces in
equilibrium ( Geq Å nads /A, where nads is the excess moles of
surfactant on the film and A is the film interfacial area).
During the fast film expansion there is practically no time
for surfactant adsorption, thus the excess moles of surfactant
at the film interfaces do not change. Moreover, if the molecu-
coidas
 EMULSION AND FOAM SYSTEMS 37

FIG. 10. Comparison of the relaxation of film tension and single IFT with 30.2% relative expansion of the emulsion system with 3.0 1 10 05 mol/
dm3 (12 CMC) Brij58.

lar rearrangements on the film interfaces occur quickly, the Thus the response of the film will be purely elastic and
film response to the expansion is perfectly elastic, the maximum film elasticity will be measured. From Eqs.
[6] and [7],
d ln Geq Å 0d ln(A/A0 ), [7]

where A and A0 are the expanded and the initial film areas, dfi
Ef Å . [8]
respectively. d ln(A/A0 )

FIG. 11. Comparison of the relaxation of film tension and single IFT with 38.9% relative expansion of the emulsion system with 3.0 1 10 05 mol/
dm3 (12 CMC) Brij58.

AID JCIS 4507 / 6g1e$$$441 02-06-97 09:21:04 coidas

38 KIM, KOCZO, AND WASAN

FIG. 12. Comparison of the relaxation of film tension and single IFT with 48.0% relative expansion of the emulsion system with 3.0 1 10 05 mol/
dm3 (12 CMC) Brij58.

The measured maximum film elasticity is equal to the Gibbs
elasticity at zero film thickness. A plot of the initial film
tension vs the logarithm of the relative film expansion is
shown in Fig. 15 for emulsion films stabilized by 2 1 10 06
mol/dm3 (below CMC) Brij 58. For comparison, the initial
single IFTs obtained in the experiments with the respective
single oil/water interface are also plotted. The analogous
results with the foam films formed from 3.2 1 10 02 mol/
dm3 SDeS are shown in Fig. 16. The initial film tensions in
Figs. 15 and 16 indeed show a linear relationship with the
logarithm of relative interfacial expansion, as predicted by
Eq. [8]. In both the emulsion and foam systems, the film
elasticity obtained from the slope of the curve is equal,
within experimental error, to twice the interfacial elasticity

FIG. 13. Comparison of the relaxation of film tension and single surface tension with 25.0% relative expansion of the foam system with 3.2 1 10 02
mol/dm3 SDeS.

AID JCIS 4507 / 6g1e$$$441 02-06-97 09:21:04 coidas

 EMULSION AND FOAM SYSTEMS 39

FIG. 14. Comparison of the relaxation of film tension and single surface tension with 35.7% relative expansion of the foam system with 3.2 1 10 02
mol/dm3 SDeS.

(Es f ) of the corresponding interface, which was obtained in
a similar way as the film elasticity. This result indicates that
the film and surface expansion rates were indeed high
enough so that only negligible surfactant adsorption occurred
during the expansion. The points corresponding to the high-
est expansions in Fig. 15 are slightly below the straight line,
which indicates that during the highest expansions a little
surfactant adsorption occurred.
Dynamic Film Tension
When the lamellae of a polyhedral foam or emulsion
drain, due to the capillary and gravitational forces, the
liquid flow continuously shears the film interfaces causing
the film to slowly and continuously expand ( Fig. 1b ) . To
model this process, dynamic film tension measurements
were carried out.

FIG. 15. Initial film tension and IFT in the stress relaxation experiments as a function of ln( A/A0 ) for the emulsion system, in the presence of 2.0
1 10 06 mol/dm3 Brij58 ( õ1 CMC).

AID JCIS 4507 / 6g1e$$$441 02-06-97 09:21:04 coidas

40 KIM, KOCZO, AND WASAN

FIG. 16. Initial film tension and IFT in the stress relaxation experiments as a function of ln( A/A0 ) for the foam system in the presence of 3.2 1
10 02 mol/dm3 SDeS.

Figure 17 shows the results of dynamic film tension mea-
surements with emulsion films stabilized by Brij 58 at 12
CMC concentration. The films were slowly and continuously
expanded with constant rate from an initial equilibrium state
while the film radius, the capillary pressure and the film
tension were monitored. The dynamic film tension depends
on the rate of the film expansion and the rate of surfactant
adsorption during the expansion. As the film was expanded,
the film tension increased because the surface adsorption of
the surfactant decreased. As shown in Fig. 17, a higher film
expansion rate results in higher dynamic film tensions be-
cause the surfactant adsorption gets more and more delayed
at increasing expansion rates. Similar trends were obtained
for foam films (SDeS, 3.2 1 10 02 mol/dm3 ) as shown in
Fig. 18.
Figure 19 shows a comparison between the dynamic film
tension for the emulsion film and two times the correspond-
ing single IFT, at 3.0 1 10 05 mol/dm3 (12 CMC) Brij58

FIG. 17. Dynamic film tension of emulsion films with various film expansion rates in the presence of 3.0 1 10 05 mol/dm3 (12 CMC) Brij58.

AID JCIS 4507 / 6g1e$$$441 02-06-97 09:21:04 coidas

 EMULSION AND FOAM SYSTEMS 41

FIG. 18. Dynamic film tension of foam films with various film expan- FIG. 20. Comparison of dynamic film tension and 21 single surface
sion rates in the presence of 3.2 1 10 02 mol/dm3 sodium decylsulfonate. tension at two different expansion rates, for the foam system, in the presence
of 3.2 1 10 02 mol/dm3 sodium decylsulfonate.

concentration. Similarly to the case of the stress relaxation
experiments, the film tension is greater than two times the
single IFT in dynamic conditions. In the case of the foam
film experiment, as shown in Fig. 20, similar trends can be
observed at two different film expansion rates.
The dynamic film tension experiments gave information
on film stability as well. In Fig. 21, the dynamic film tension
of emulsion films with various expansion rates is plotted at
a low Brij58 concentration (6 1 10 06 mol/dm3 , õCMC).
It can be observed that the film ruptures at a critical film

FIG. 19. Comparison of dynamic film tension and 21 single IFT at a constant expansion rate, for the emulsion system, in the presence of 3.0 1
10 05 mol/dm3 (12 CMC) Brij 58.

AID JCIS 4507 / 6g1e$$$441 02-06-97 09:21:04 coidas

42 KIM, KOCZO, AND WASAN
FIG. 21. Dynamic film tensions for the emulsion film at a low surfactant concentration, Brij58 6.0 1 10 06 mol/dm3 ( õCMC). The film ruptured
during expansion at a critical film tension (38 mN/m).
tension (about 38 mN/m) which does not depend on the
rate of film expansion. The constant film tension value (38
mN/m) implies that there is a critical surface excess of
surfactant on the film interfaces which is necessary to stabi-
lize the emulsion film. Based on the IFT isotherm of Brij58
(4) the critical surface adsorption corresponding to the above
critical tension was found to be 2.8 1 10 010 mol/cm2 , which
would be in equilibrium with 4.0 1 10 07 mol/dm3 bulk
Brij58 concentration.
Interfacial Tension and Film Tension in Dynamic
Conditions
The most interesting result of our film rheological studies
is that under dynamic conditions the film tension is not equal
but higher than twice the respective IFT. This difference can
be explained by the following mechanism. When the film
or single interface is expanded, the surfactant adsorption on
the interface(s) decreases. As a result, surfactant diffusion
and adsorption start to restore the adsorption equilibrium.
In the case of a single interface (Fig. 22a), the surfactant
molecules diffuse and adsorb from the phase containing the
surfactant (the aqueous phase in our experiments) along the
entire interface.
Figure 22b is a diagram of the film stress relaxation pro-
cess. In a stable emulsion the surfactant is typically soluble
in the film phase only, that is, in the continuous phase, as
in the systems studied. In this case, the surfactant molecules
can diffuse from inside the film as well as from the meniscus.
(If the surfactant is soluble in the dispersed phase only, then
the film and the emulsion are generally unstable (11).) It
was shown in the stress relaxation experiments (Figs. 7,
8) that there was no detectable surfactant depletion in the
AID JCIS 4507 / 6g1e$$$441 02-06-97 09:21:04
meniscus. Surfactant depletion, however, occurred inside the
film, which is related to the delayed film stress relaxation
and higher dynamic film tension, as will be discussed below.
The ratio of the total moles of surfactant adsorbed on the
two film interfaces (nads ) and the moles of surfactant inside
the film (nms ), neglecting the difference between surface
concentration and surface excess, is
nads 2G
Å , [9]
nms ch f
where h f is the film thickness and c is the surfactant concen-
tration in the bulk of the continuous phase. With G Å 2 1
10 010 mol/cm3 and h f Å 10 Ç 100 nm (which are typical
values in real emulsion or foam films) and at 3 1 10 08 mol/
cm3 (12 CMC Brij58), surfactant concentration nads /nms is
10 3 to 10 4 . This means that, in the emulsion films studied,
the total amount of surfactant inside the film was negligible
compared to the amount of surfactant adsorbed. Therefore,
strong surfactant depletion inside the film happened even at
a small film expansion. In the case of the foam film, an
ionic surfactant (SDeS) was used at a much higher molar
concentration than the nonionic surfactant in the emulsion
film and thus the ratio of the total amount of surfactant
inside the film and the amount which is adsorbed on the film
surfaces (Eq. [9]) was lower, about 2–13. Therefore, a
smaller depletion occurred in the foam films studied.
The strong surfactant depletion in the film can be also
shown by calculating the diffusion penetration depth, d Å
( pDt) 1 / 2 (8). Using D Å 5 1 10 06 cm2 /s and t Å 1000 s,
d Å 0.125 cm, which is orders of magnitude larger than the
film thickness.
coidas
 EMULSION AND FOAM SYSTEMS
FIG. 22. Schematic showing (a) single IFT and (b) film tension in dynamic conditions.
When the film is expanded several processes take place.
Diffusion and adsorption (Ifa ) from inside the film to the
film surfaces (Fig. 22b). The amount of surfactant is not
enough to restore equilibrium, as was shown above, and
therefore other processes should be also considered: surfac-
tant diffusion and adsorption from the bulk of the meniscus
to the meniscus surface (Ima ), diffusion from inside the me-
niscus to the inside of the film (Imd ), and surface diffusion
from the meniscus surfaces to the film surfaces (Isd ). The
surface diffusion from the meniscus to the adjoining film
occurs because of the surfactant depletion in the expanded
film which results in a lower surfactant concentration on the
film than on the meniscus surface. Imd can be neglected be-
cause the film cross-sectional area is very small (the film
thickness is very small). Moreover, the surface diffusion
from the meniscus is generally fast (12) and so the rate
controlling steps are the diffusion and adsorption in the me-
niscus (Ima ). Ima is proportional to the interfacial area of the
meniscus and thus the relaxation of the film will be slower
than in the respective single interface because the meniscus
area is smaller than the film interfacial area. Comparison of
our relaxation curves for the emulsion system (Figs. 10,
11) and the curves for the foam system (Figs. 13, 14),
respectively, show that the deviation between the film and
the single interface was more pronounced for the foam sys-
tem than for the emulsion system at a similar relative expan-
sion. The reason for the slower relaxation in the foam film
was that the film area in the foam film experiments was
about three times larger than in the emulsion film measure-
AID JCIS 4507 / 6g1e$$$441 02-06-97 09:21:04
43
ments. The area of the meniscus, which is controlled by the
wall thickness of the capillary, was about the same in both
systems; thus, the ratio of meniscus and film areas was also
about three times lower in our foam film experiment re-
sulting in a slower film tension relaxation.
Our dynamic film experiments may represent the dynamic
processes in the respective concentrated foam and emulsion
systems. The dynamic film tension, as controlled by the
surfactant diffusion and adsorption in the meniscus (Plateau
borders), becomes higher with increasing dispersed phase
ratio (and thus with decreasing ratio of the film area to the
Plateau border areas) and with decreasing surfactant concen-
tration.
SUMMARY
A method and apparatus developed earlier in our labora-
tory (6) were used to understand the dynamic behavior of
liquid films. It was found for both emulsion and foam sys-
tems that the dynamic properties of liquid films and thus
the behavior of concentrated fluid dispersions, cannot be
predicted only from the dynamic IFT of the corresponding
air/liquid or liquid/liquid interfaces. The film tension in an
expanding film is higher than twice the single (dynamic)
IFT due to a surfactant depletion inside the film. It is sug-
gested that in the film the surfactants diffuse and adsorb not
at the film surfaces but at the meniscus followed by a surface
diffusion from the meniscus to the film surface. The effect,
that is, the difference between the film tension and twice
coidas
44 KIM, KOCZO, AND WASAN
the respective IFT, is greater if the film/meniscus (Plateau
border) area ratio is higher and if the surfactant concentra-
tion is lower.
The lamellae of a foam or concentrated emulsion continu-
ously expand under shear or dilation. Thus, the process as
the film tension increases by film expansion should affect
not only the stability but also the rheological properties of
concentrated fluid dispersions. The results of this study indi-
cate that an expanding foam or concentrated emulsion should
be less stable if the rate of expansion is higher (Fig. 21) or
if the volume fraction of the dispersed phase (and thus the
area ratio of the liquid films to the Plateau borders) is higher.
Work is in progress in our laboratory to model the dynamic
film behavior based on our results.
ACKNOWLEDGMENTS
This study was supported in part by the Department of Energy and the
National Science Foundation. The authors gratefully acknowledge the help-
ful comments provided by Dr. A. D. Nikolov.
AID JCIS 4507 / 6g1e$$$441 02-06-97 09:21:04
REFERENCES
1. Edwards, D. A., and Wasan, D. T., J. Rheol. 32(5), 429 (1988).
2. Defay, R., and Petre, G., 1979, ‘‘Dynamic Surface Tension,’’ Surface
and Colloid Science, Vol. 3 (Matijevic, Ed.). Wiley, New York.
3. Edwards, D. A., Brenner, H., and Wasan D. T., ‘‘Interfacial Transport
Process and Rheology.’’ Butterworth–Heinemann, 1991.
4. Nagarajan, R., and Wasan, D. T., J. Colloid Interface Sci. 159, 164
(1993).
5. Nagarajan, R., Koczo, K., Erdos, E., and Wasan, D. T., AIChE J. 41(4),
915 (1995).
6. Soos, J., Koczo, K., Erdos, E., and Wasan, D. T., Rev. Sci. Instrum.
65(11), 3555 (1994).
7. Joly, M., 1972, ‘‘Rheological Properties of Monomolecular Films,’’
Surface and Colloid Science, Vol. 5 (Matijevic, Ed.). Wiley, New
York.
8. Van Hunsel, J., and Joos, P., Colloids Surf. 25, 251 (1987).
9. Prins, A., in ‘‘Foams’’, Proc. Symp. at Brunel Univ., 1975, ed. R. J.
Akers, p. 51.
10. Adamson, A. W., Physical Chemistry of Surfaces, 5th ed. Wiley, New
York, 1990.
11. Bancroft, W. D., J. Phys. Chem. 17, 501 (1913).
12. Edition, P., 1989, ‘‘Diffusion in Solids,’’ 2nd ed. TMS.
coidas