View Article Online / Journal Homepage / Table of Contents for this issue

PAPER www.rsc.org/softmatter | Soft Matter

Origin of stabilisation of aqueous foams in nanoparticle–surfactant mixtures

Bernard P. Binks,*a Mark Kirklandb and Jhonny A. Rodriguesa
Received 3rd July 2008, Accepted 15th August 2008
First published as an Advance Article on the web 25th September 2008
DOI: 10.1039/b811291f
Published on 25 September 2008. Downloaded by KYUSHU UNIVERSITY on 25/01/2018 03:27:24.

Using a range of complementary experiments, a detailed investigation into the behaviour of air-in-
water foams stabilised by a mixture of silica nanoparticles and pure cationic surfactant has been made.
At high pH where particles are significantly negatively charged and surfactant is positively charged, no
foam is possible with particles alone whereas surfactant-stabilised foams break down completely within
one day at all concentrations. In particle–surfactant mixtures, a synergism occurs with respect to foam
formation and stability due to the adsorption of surfactant molecules onto particle surfaces. The
foamability of mixed dispersions is substantially reduced compared with surfactant solutions alone.
However, the foam stability passes through a maximum with respect to surfactant concentration and
these foams are remarkably stable. Based on our findings from dispersion stability measurements,
particle z potentials, the adsorption isotherm of surfactant on particles and relevant contact angles of
water in air on silica surfaces, we conclude that foams are most stable when particles are strongly
flocculated corresponding to them possessing a low charge, being maximally hydrophobic and
containing an adsorbed monolayer of surfactant. Cryo-scanning electron microscopy (cryo-SEM)
analysis of the same foams leads us to propose that foam stabilisation changes from being surfactant
dominated at low surfactant concentration to being particle dominated at intermediate concentrations
and reverting to surfactant dominated at higher concentrations.

Introduction
The use of small solid particles, frequently in the presence of oil,
as antifoam agents in surfactant-stabilised aqueous foams is well
known commercially and reasonably well understood scientifi-
cally.1–3 Here, relatively hydrophobic particles added to an
aqueous surfactant solution promote film rupture by entering
one of the film surfaces initially and then bridging both surfaces
subsequently as the film thins via drainage. Hydrophobic parti-
cles do not always destabilise foams however. Tang et al.4 and
Kumagai et al.5 both observed that hydrophobic particles of
either silica or polystyrene actually stabilised foams of sodium
dodecyl sulfonate or ovalbumin, respectively. The effect of
surfactant/protein on the in situ hydrophobicity of the particles
was not determined however. In a different application, particle–
surfactant mixtures are normally required during the froth
flotation process for recovering mineral particles from aqueous
slurries.6 In this case, one role of the surfactant (or collector) is to
modify the wettability of the chosen particles by adsorbing on
them thus enhancing the attachment of coated particles to newly
created bubble surfaces.
Recently, there has been a number of reports on the use of
colloidal particles as sole stabilisers of aqueous foams, i.e. in the
absence of surface-active molecules, since most particles are
surface-active even though they are not amphiphilic.7–12 The
hydrophobicity of the particles, as judged from the contact angle
they make with an air–water surface, is deemed to be important
a
Surfactant and Colloid Group, Department of Chemistry, University of
Hull, Hull, UK HU6 7RX. E-mail: b.p.binks@hull.ac.uk
b
Unilever Research, Colworth Laboratory, Sharnbrook, Bedfordshire, UK
MK44 1LQ
in whether a particle can adsorb onto and remain around bubbles
in water.13 In most cases, particles form a dense layer on bubble
surfaces preventing coalescence and slowing down or halting
completely disproportionation.14–17 Evidence also exists that
additional stabilisation occurs via particle network formation
between adsorbed and non-adsorbed particles reducing, in this
case, drainage of water between bubbles.18 Since many particle
types are inherently hydrophilic attaching weakly to fluid inter-
faces, the two ways to increase their hydrophobicity are to
hydrophobise the particles ex situ by chemisorption and in situ by
physisorption. For the former, Binks and Horozov8 using fumed
silica nanoparticles coated to different extents with dichlor-
odimethylsilane groups showed that particles of intermediate
hydrophobicity were the only ones capable of stabilising foams.
For the latter, Gonzenbach et al.9 achieved the required level of
hydrophobisation by addition of short chain amphiphilic mole-
cules which adsorbed on particle surfaces. However, surfactant
addition to particle dispersions does not always lead to stable
foams. Alargova et al.7 found that mixtures of rod-shaped
polymer particles and sodium dodecyl sulfate (SDS) produced
foams which were as unstable as pure surfactant ones, whereas
foams of particles alone were stable indefinitely. It was hypoth-
esised that SDS adsorption onto particles rendered them more
hydrophilic reducing their affinity for the air–water surface.
Subramaniam et al.15 also reported the destabilisation of particle-
coated bubbles in water when exposed to surfactant solutions by
ejection of particles, explaining their findings in terms of relieving
the stress in non-spherical, stable particle-coated bubbles.
The energy of attachment of a single, spherical nanoparticle to
the oil–water interface E has been shown to be related to the
stability of emulsions containing adsorbed particles,19 i.e. as
E increases, the stability to coalescence increases. In the case of

This journal is ª The Royal Society of Chemistry 2008 Soft Matter, 2008, 4, 2373–2382 | 2373

air–water surfaces relevant to foams, for contact angles q<90
,
this energy E is given by13
E ¼ pr2g(1  cosq)2 (1)
in which r is the particle radius and g is the air–water surface
tension. Very hydrophilic particles exhibiting low q do not
stabilise foams8,9 whereas very hydrophobic particles with q>90

can act as foam breakers.2,3 Inspection of eqn 1 reveals that
Published on 25 September 2008. Downloaded by KYUSHU UNIVERSITY on 25/01/2018 03:27:24.
lowering the value of g should decrease E, whereas increasing the
value of q should increase E. There is thus a competing effect of
the tension and the contact angle on the magnitude of E upon
initially increasing the surfactant concentration at a fixed particle
concentration.
Despite some activity in this area, systematic studies of the
behaviour and properties of foams in mixtures of surfactant and
surface-active particles are lacking. Do the two foaming agents
act synergistically or antagonistically with respect to foam
stabilisation? As a model system, we investigate here aqueous
foams stabilised by either pure di-chain cationic surfactant,
hydrophilic silica nanoparticles or their mixture. This system,
when mixed with oil, triggered the phase inversion of emulsions
caused by changes in the wettability of particles by adsorption of
surfactant.20 Additional experiments have been conducted to
shed light on the findings including determination of the
adsorption of surfactant onto particles, z potentials of particle
dispersions, appropriate contact angles and cryo-SEM of frozen
foam samples. The key foam findings from a similar system of
larger silica particles and a single chain cationic surfactant are
also presented for comparison.
Experimental
Materials
Water was first passed through an Elga reverse osmosis unit and
then a Milli-Q reagent water system. Its surface tension at 25
C
was 71.9 mN m1. Sodium hydroxide (BDH, 99.5%) was used to
effect the aqueous phase pH. The pure cationic surfactant
di-decyldimethylammonium bromide, di-C10DMAB, was
purchased from Aldrich with a purity of 98%. The critical micelle
concentration, cmc, in water at 25
C is 1.9 mM. Cetyl-
trimethylammonium bromide, CTAB, from Fluka (99%) has
a cmc of 0.9 mM. The silica particles were those of Ludox HS-30
(Aldrich), received as a 30 wt% aqueous dispersion at pH ¼ 9.8.
As seen in the transmission electron micrograph (TEM) in Fig. 1,
the particles are spherical and relatively monodisperse with an
average diameter of approximately 15 nm and a specific surface
area of 220 m2 g1. The particle size distribution determined by
dynamic light scattering using a Malvern Zetasizer 3000HS
instrument yields an average particle diameter of 13 nm, in
agreement with that quoted by the manufacturer of 12 nm. For
comparison, a limited set of experiments was conducted using
monodisperse hydrophilic silica particles of diameter equal to 0.6
mm (Blue Helix, UK).
Methods
(a) Preparation and characterisation of aqueous dispersions.
We have investigated silica particle dispersions in water (2 wt%)
2374 | Soft Matter, 2008, 4, 2373–2382
View Article Online
Fig. 1 TEM image of 0.1 wt% Ludox HS-30 particles in water at pH ¼
9.8. Scale bar ¼ 50 nm.
in the presence of di-C10DMAB surfactant at high pH. At
pH ¼ 9.8, aliquots of aqueous di-C10DMAB were added to
a volume of diluted, as-received dispersion at this pH. For high
di-C10DMAB concentrations, concentrated NaOH was added
subsequently to adjust the pH to 9.8. The extent of sedimentation
of the dispersions was determined by measuring the height of the
sediment as a function of time. The z potentials of 1 cm3 of
a 2 wt% particle dispersion at high pH in the presence of
increasing amounts of di-C10DMAB were determined using
a Malvern Zetasizer 3000HS instrument employing a dip cell,
using the Smoluchowski equation for converting measured
mobilities. At least 3 separate measurements were made for each
surfactant concentration. For dispersions which sediment, the
supernatant was taken for these measurements. For TEM, an
aqueous particle dispersion was evaporated on a carbon coated
copper grid and imaged using a JEOL-2010 electron microscope.
(b) Determination of the adsorption isotherm of surfactant on
particles in water. The method employed to determine the
adsorption of di-C10DMAB surfactant onto Ludox HS-30 silica
particles dispersed in water at pH ¼ 9.8 was via depletion from
solution. 10 cm3 of equilibrated particle (1 wt%)–surfactant
dispersion prepared in plastic vessels was ultracentrifuged
(Beckman L8-M) at 30 000 rpm for 60 min. The concentration of
free surfactant in the clear supernatant was determined using the
Epton titration21 in plastic vessels with SDS as titrant.
The amount of surfactant adsorbed was then calculated from the
difference between the initial concentration and that measured in
the solution after equilibration with particles.
(c) Air–water–silica contact angles. The contact angles
through water of sessile drops of aqueous surfactant solution at
pH ¼ 9.8 in air on silica-coated planar silicon wafers were
determined using a Krüss DSA 10 instrument. A 20 ml drop was
placed on the substrate contained in a thermostatted metal
chamber which was sealed to saturate the vapour space with
water. Silicon wafers (Plasmos, 1.6  1.6 cm2) were treated with
This journal is ª The Royal Society of Chemistry 2008

concentrated nitric acid and rinsed with water before use
rendering them hydrophilic. Advancing angles were determined.
A mean contact angle of at least 3 drops was determined for each
surfactant concentration.
(d) Preparation and characterisation of foams. For systems
containing either surfactant or particles alone, foams were
prepared by hand shaking 20 cm3 of liquid in 140 cm3 graduated
Published on 25 September 2008. Downloaded by KYUSHU UNIVERSITY on 25/01/2018 03:27:24.
glass measuring cylinders for 30 s (120 shakes). For systems
containing a mixture of surfactant and particles, surfactant was
added to a particle dispersion in water and left 24 h to equili-
brate. Foaming was the same as described above. Immediately
after preparation, the initial volume of foam (foamability) is
recorded and the foam volume is measured as a function of time
(foam stability). Photographs of the vessels were taken with
a Canon IXUS I7 digital camera. A Nikon Labophot microscope
equipped with a QICAM 12-bit Mono Fast 1394 camera
(QImaging) was used to observe foam samples. A small sample
of foam was placed on a microscope slide. The images were
processed using Image-Pro Plus 5.1 software (Media Cyber-
netics). Freeze fracture SEM measurements were also carried out
on selected foam samples in the following way, as we developed
for emulsions previously.22 A small volume of fresh foam was
mounted onto a 0.1 cm diameter aluminium stub and plunged
into nitrogen slush at 196
C. The frozen sample was trans-
ferred to a Gatan Alto 2500 cryo preparation chamber and
fractured with a scalpel blade at 175
C under high vacuum
(1  107 mbar). It was then coated with a Au/Pd sputter at low
vacuum (1  102 mbar) after which the vacuum was increased
again and the sample was transferred to the chamber of a JEOL
JSM-6301F SEM at 150
C. Image acquisition was obtained
using the software Scandium (Soft Imaging System).
All measurements were carried out at room temperature being
22  2
C.
Results and discussion
We first describe the effect of adding di-C10DMAB surfactant to
aqueous dispersions of Ludox HS-30 particles at high pH (where
silica is most charged) and the subsequent foaming of these
mixed dispersions. A limited data set is then presented for
a different system of larger particles and a single chain surfac-
tant.
(a) Aqueous dispersions of silica nanoparticles and cationic
surfactant at pH ¼ 9.8
Since the isoelectric point for silica particles is between pH ¼ 2
and 3, at pH ¼ 9.8 particle surfaces are appreciably negatively
charged due to dissociation of surface silanol SiOH groups.
Addition of an oppositely charged cationic surfactant like
di-C10DMAB leads to electrostatic adsorption of surfactant
monomer onto particle surfaces. As a monolayer forms, particles
become increasingly hydrophobic (alkyl chains exposed to
solution). Such a change in the wettability of coated particles
could be sufficient to enable particles to adsorb and remain at air
bubble surfaces stabilising the foam. Further adsorption of
surfactant can occur via chain–chain interaction resulting in
a bilayer, now with headgroups exposed to the solution rendering
This journal is ª The Royal Society of Chemistry 2008
View Article Online
Fig. 2 Photograph of vessels after 24 h containing aqueous dispersions
of Ludox HS-30 particles (2 wt%) at pH ¼ 9.8 as a function of initial di-
C10DMAB concentration (given above the vessels). The free surfactant
concentration is given below the vessels (later).
the particles more hydrophilic again. At a fixed particle
concentration of 2 wt%, addition of di-C10DMAB leads to
a dramatic change in the stability of the aqueous dispersions. In
Fig. 2, it can be seen that the initial stable dispersion begins to
sediment increasingly up to 10 mM di-C10DMAB (5 times the
cmc), after which the sedimentation falls. This is reversible. The
sedimentation is complete after 24 h. Between 0 and 0.5 mM
di-C10DMAB, dispersions are stable and bluish since surfactant
adsorption is low and repulsion between particles dominates. At
intermediate concentrations between 1 and 10 mM
di-C10DMAB, coagulation of particles resulting in their sedi-
mentation occurs progressively as van der Waals attraction
between alkyl chains on neighbouring particles dominates.
Above 10 mM di-C10DMAB, the sedimentation extent decreases
again and a fraction of the particles are re-stabilised as a surfac-
tant bilayer forms on their surfaces leading to charge repulsion.
It is likely that the slight flocculation observed here at high
surfactant concentration is due to depletion of micelles between
particles on close approach, reported in a similar system.23
Samples of the dispersions were taken for TEM and cryo-SEM
analysis. In the absence of surfactant (Fig. 1), discrete particles
are visible. For 1 mM di-C10DMAB [Fig. 3(a)], particles are
more attractive forming small chains, although the degree of
coagulation is not sufficient to induce sedimentation. By 5 mM
di-C10DMAB [Fig. 3(b)], extensive flocculation has occurred
leading to large flocs of micron size dimensions. Do such flocs
adsorb to air–water surfaces on aeration?
In order to quantify the extent of surfactant adsorption on silica
particles in water, we have determined the adsorption isotherm
using the depletion from solution method. This involves
measuring the concentration of non-adsorbed surfactant left in
water after equilibration with particles. The isotherm derived
from such data is given in Fig. 4. There is a gentle increase in the
adsorbed amount initially followed by a more significant increase
at higher concentrations, ending close to the cmc. In a detailed
study of alkylpyridinium cationic surfactants on quartz particles,
Fuerstenau and Jia24 describe four regions in their isotherms. In
region I, surfactant ions adsorb individually with their head-
groups down. In region II, adsorbed ions associate into patches or
hemi-micelles on the surface through chain–chain interactions.
This results in an adsorbed monolayer. In region III, a second
layer begins to adsorb on the first through hydrophobic interac-
tions of the chains. In region IV occurring close to the cmc, the
bilayer is complete and the adsorbed amount reaches a plateau
Soft Matter, 2008, 4, 2373–2382 | 2375
Published on 25 September 2008. Downloaded by KYUSHU UNIVERSITY on 25/01/2018 03:27:24.
Fig. 3 SEM images of frozen aqueous dispersions of Ludox HS-30 particles (2 wt%) at pH ¼ 9.8 in the presence of (a) 1 and (b) 5 mM di-C10DMAB.
Scale bars ¼ 100 nm.
Fig. 4 Adsorption isotherm of di-C10DMAB surfactant on Ludox HS-
30 particles in water (1 wt%), pH ¼ 9.8.
value. The last point on our isotherm corresponds to an area per
di-C10DMAB molecule (for a bilayer) on particle surfaces equal to
0.46 nm2. This is the same as that reported by Esumi et al.25 for
di-C12DMAB on 0.3 mm silica particles at pH ¼ 6.
Since the z potential of particles directly reflects adsorption in
the Stern plane, electrophoretic measurements were conducted
with Ludox HS-30 dispersions as a function of di-C10DMAB
concentration and are given in Fig. 5. In the absence of surfac-
tant at this high pH, the z potential is 48 mV. Addition of
di-C10DMAB gradually decreases the magnitude of the potential
up to 10 mM after which there is a reversal of the sign and
magnitude to yield positively charged particles of appreciable
charge. The initial di-C10DMAB concentration at which particles
possess a z potential of zero corresponds to an equilibrium (free)
concentration of z 0.37 mM, which is in the second increase in
the isotherm. The picture emerging is that on addition of
di-C10DMAB, initially hydrophilic and negatively charged
2376 | Soft Matter, 2008, 4, 2373–2382
View Article Online
Fig. 5 Variation of z potential of aqueous Ludox HS-30 dispersions (2
wt%) with initial di-C10DMAB concentration at pH ¼ 9.8.
particles become hydrophobic and uncharged leading to floccu-
lation and then positively charged and hydrophilic again.
The above scenario is backed up with contact angle
q measurements of aqueous surfactant drops in air on silicon
wafers which possess a layer of silica on their surface. Since
determination of these angles using nanoparticles themselves is
not possible, we chose a planar substrate of similar chemical
composition to that of the particles. The close agreement
between contact angles on the two substrate types has been
shown previously.26 In Fig. 6, it can be seen that an initially
hydrophilic surface in the absence of surfactant exhibits a contact
angle through water of 8
at this high pH. As related by Young’s
equation, adsorption of surfactant to the air–water, solid–water
and, importantly, air–solid interfaces results in an increase in q to
a maximum of 63
followed by a significant decrease at higher
concentrations.27–29 Silica surfaces thus undergo a transition
from hydrophilic to more hydrophobic to hydrophilic again.30 As
This journal is ª The Royal Society of Chemistry 2008
Published on 25 September 2008. Downloaded by KYUSHU UNIVERSITY on 25/01/2018 03:27:24.
Fig. 6 Advancing contact angles of aqueous di-C10DMAB drops in air
on silicon wafers, pH ¼ 9.8, as a function of surfactant concentration.
The angle without surfactant is 8
.
shown by Yaminsky and Yaminskaya,28 at low surfactant
concentrations, adsorption at the air–solid surface is larger than
at the surfaces involving water such that an increase in contact
angle results. The significant adsorption at the solid–water
interface which occurs at higher concentrations reverses this
inequality and contact angles decrease. Although it is not easy to
compare surfactant concentrations for the z potential and
contact angle data (due to different solid surface areas and
volumes of solution), we anticipate that the condition corre-
sponding to maximum hydrophobicity is also that for which
particles are least charged.31
(b) Foams stabilised by surfactant or nanoparticles alone at pH
¼ 9.8
At high pH, silica particles are very hydrophilic and, alone, poor
emulsifiers of oil and water.19 Likewise, they are poor foamers of
air and water. A series of foams prepared at various particle
concentrations (0.1–30 wt%) was extremely unstable to coales-
cence with no bubbles remaining 3 min after aeration. The
foamability of these dispersions was also very low, with less than
1 cm3 of foam being generated. By contrast, the surfactant
di-C10DMAB stabilises air-in-water foams in which the initial
foam volume is dependent on surfactant concentration. In
Fig. 7(a), the foam volume is plotted as a function of time, where
it can be seen that the foamability initially increases with
concentration and then decreases at higher concentrations.
Maximum foamability occurs just above the cmc in water. Up to
the cmc, the monomer concentration increases allowing a larger
volume of air bubbles to be stabilised by adsorption. Above the
cmc, although the monomer concentration remains unchanged,
micelles which form need to dissociate into monomers which
then adsorb at freshly created air–water surfaces. In an anionic
surfactant system, Oh and Shah32 have shown that there is
a minimum in foamability of solutions well above the cmc and
This journal is ª The Royal Society of Chemistry 2008
View Article Online
Fig. 7 (a) Foam volume vs. time for di-C10DMAB-stabilised foams at
different concentrations (given) at pH ¼ 9.8. (b) Volume of water sepa-
rated vs. time from selected foams in (a). The dotted line represents the
initial volume of solution.
that this corresponds to a maximum in the lifetime of the
micelles. This however does not explain why the foamability does
not remain constant above the cmc. A possible reason for the
decreased foamability above the cmc is the increase in the ionic
strength of the solution with increasing surfactant concentration,
since it is well known that ionic surfactants foam less in the
presence of salt.2
The surfactant-stabilised foams break down with time via
a combination of mechanisms including the drainage of water
between bubbles, bubble coalescence and bubble dispropor-
tionation. The foam volume decreases gradually at first and then
more markedly at longer times [Fig. 7(a)]. After 15 h or so, no
foam remains in all but the highest concentration system. At 100
mM surfactant, the apparent increase in foam volume up to 10
min is due to the fact that the surfactant solution is more viscous
and, after aeration, some air bubbles are trapped in a thin layer
adjacent to the vessel wall. With time, they are slowly dragged on
top of the main foam layer increasing its volume. The drainage of
Soft Matter, 2008, 4, 2373–2382 | 2377

water from the foam is shown as a function of time in Fig. 7(b)
for several concentrations. In general, foams are more stable with
increasing concentration but most of them drain nearly 100% of
water in 10 min. The noticeable increase in stability to drainage
at 100 mM is due to the increased viscosity of the aqueous phase.
Using the data in both (a) and (b), the calculated volume fraction
of air in the creamed foam layer decreases with time as opposed
to it increasing if only drainage occurred. This decrease is mainly
Published on 25 September 2008. Downloaded by KYUSHU UNIVERSITY on 25/01/2018 03:27:24.
due to the frequent coalescence events visible by eye.
As with other surfactant-stabilised foams, the bubbles in these
foams are initially spherical throughout of diameters between 0.3
and 2 mm. As drainage and disproportionation proceed, bubbles
nearer the top of the foam layer increase in size to around 5 mm
and become polyhedral in shape. The more dramatic decrease in
foam volume after a certain period of time corresponds to the
rupture of the thin aqueous film between neighbouring poly-
hedral bubbles. A selection of these bubbles is shown in Fig. 8[(a)
and (b)], where it is noticed that all the surfaces appear smooth.
(c) Foams stabilised by a mixture of surfactant and
nanoparticles at pH ¼ 9.8
Aqueous foams were prepared from mixtures of silica particles
(2 wt%) and increasing concentrations of surfactant at high pH.
Their foamability is compared to that of solutions of surfactant
alone in Fig. 9(a). In mixtures (crosses), the initial foam volume
increases slightly with surfactant concentration and then
decreases to a minimum around 5 mM, increasing continuously
Fig. 8 Optical microscopy images of bubbles stabilised by di-C10DMAB alone at (a) 1 and (b) 10 mM (10 s after formation) and by surfactant–particle
(2 wt%) mixtures at surfactant concentrations of (c) 0.1 and (d) 5 mM (1 h after formation), pH ¼ 9.8. Scale bars are given separately.
2378 | Soft Matter, 2008, 4, 2373–2382
View Article Online
for surfactant concentrations above this. This is very different to
the maximum in foamability discussed earlier for surfactant-only
foams (circles). Apart from at very high surfactant concentra-
tions, it is clear that the addition of silica particles significantly
reduces the foamability of the surfactant solutions. Those of
lowest foamability arise from mixed dispersions which are floc-
culated and which sediment (Fig. 2). The reduction in foaming is
partly due to the loss of monomeric surfactant from solution as it
adsorbs on particle surfaces modifying their dispersion state and
wettability. Using the adsorption isotherm data, we can calculate
the concentration of free surfactant after adsorption on particles
(2 wt%) for different initial concentrations of surfactant; see
values below the vessels in Fig. 2. We find that the data in
Fig. 9(a) can now be divided into three regions. For initial
concentrations less than z 2 mM, the free surfactant present in
mixtures produces much less foam if present alone than the
mixtures do. This implies that the partially coated particles are
responsible for a significant percentage of the foamability of
these mixtures. For initial concentrations between 5 and 10 mM,
although between 98 and 99% of the surfactant becomes adsor-
bed on the particles, the foamability of the mixtures is about the
same as that of surfactant solutions alone at the equivalent free
concentration. The foamability could thus be due entirely to the
presence of free surfactant, although adsorbed particles
contribute to foam stability. Above z 20 mM surfactant
initially, the foamability of surfactant solutions alone containing
the same concentration of non-adsorbed surfactant as in the
mixtures is higher than that of the mixtures. Although we do not
This journal is ª The Royal Society of Chemistry 2008

Published on 25 September 2008. Downloaded by KYUSHU UNIVERSITY on 25/01/2018 03:27:24.
Fig. 9 (a) Initial foam volume vs. initial surfactant concentration for
di-C10DMAB foams (circles) and Ludox HS-30 (2 wt%) + di-C10DMAB
foams (crosses) at pH ¼ 9.8. (b) Foam stability after 24 h vs. initial
surfactant concentration for systems in (a) with same data point types.
know the reason for this, we can speculate that diffusion of free
surfactant to bubble surfaces is hindered possibly due to the
slightly increased viscosity of the aqueous phase in the presence
of hydrophilic, bilayer-coated particles. We plan to probe this
further with surface tension measurements. Finally, the amount
of foam formed in mixtures eventually recovers to that formed in
surfactant solutions alone as expected ($50 mM).
The stability of the foams in particle–surfactant mixtures also
varies with surfactant concentration. The most stable foams are
formed at intermediate concentrations from dispersions in which
the particles are most aggregated. As seen from the z potentials in
This journal is ª The Royal Society of Chemistry 2008
View Article Online
Fig. 5, this is where particles are of low charge and most
hydrophobic (Fig. 6). The adsorption of flocculated, partially
coated particles to air bubble surfaces may be responsible and we
noticed that such foams were sensitive to gentle shear or vibra-
tion of the vessel causing a fraction of such aggregates to detach
from the foam layer and sediment to the bottom of the vessel. In
order to compare the foam stabilities from dispersions of
different foamabilities, we define
volt¼0  volt¼t
stability ¼
volt¼0
where the vol terms refer to foam volumes initially or after
a certain time t. If the stability value is zero, the foam is
completely stable to drainage and coalescence, whereas a value of
unity corresponds to complete foam breakdown. Fig. 9(b) shows
values of the stability after 24 h of surfactant-containing foams in
the absence (circles) and presence (crosses) of silica particles. At
this timescale, most of the surfactant-only foams have
disappeared with only those at very high concentrations
remaining due to the reduction in drainage as a result of the
increased viscosity of the aqueous phase. By contrast, the
stability of foams in mixtures passes through a well defined
maximum with surfactant concentration (stability value is
minimum). At low surfactant concentrations, dispersions are
stable and the foam stability is similar to that of surfactant-sta-
bilised foams alone. At intermediate concentrations (2–10 mM),
particles become coated with a surfactant monolayer causing
them to aggregate as their hydrophobicity increases. Foams are
most stable at 5 mM surfactant, which corresponds to the
concentration at which aqueous dispersions are most sedi-
mented. The enhanced stability arises from a combination of the
adsorption of coated particles around bubbles preventing
coalescence and disproportionation as well as the reduction in
the extent of drainage between bubbles due to the increased
aqueous phase viscosity of a flocculated dispersion. It may be
that adsorbed and non-adsorbed particles are contiguous in the
form of a network, as shown previously.18 At high concentra-
tions, a bilayer of surfactant forms on particle surfaces rendering
them hydrophilic again and reducing their tendency to adsorb at
air bubble surfaces (eqn 1). Foam stability decreases in this range
as monomeric surfactant replaces particles around bubbles.
Foams formed from optimal particle–surfactant mixtures
remained at these stabilities for up to 9 months (longest
monitored).
The sizes and shapes of bubbles in foams of particle–surfactant
mixtures differ from those of surfactant-only systems. In Fig. 8(c)
and (d), some bubbles are non-spherical and all bubbles possess
a rippled surface texture. The first observation is a result of the
jamming of particles at the fluid interface preventing its relaxa-
tion to a sphere.33 The second observation is due to the
compression of a particle layer increasing its surface concentra-
tion. Eventually, since such particles are irreversibly adsorbed,
further compression causes the air–water surface to undulate
giving it the rough appearance seen.8 In order to probe the
arrangement of surfactant/particles around air bubbles, we have
developed the technique of cryo-SEM of freshly prepared foams
using larger silica particles. These particles are monodisperse of
diameter equal to 600 nm and behave in the same way with
Soft Matter, 2008, 4, 2373–2382 | 2379
Published on 25 September 2008. Downloaded by KYUSHU UNIVERSITY on 25/01/2018 03:27:24. View Article Online

Fig. 10 Cryo-SEM images of frozen aqueous foams at pH ¼ 9.8 stabilised by: (a) 1 mM di-C10DMAB alone; a mixture of 2 wt% of 0.6 mm silica
particles and (b), (c) 0.05, (d) 0.1, (e) 0.5, (f) 1 and (g) 10 mM di-C10DMAB. Scale bars ¼ 1 mm.

2380 | Soft Matter, 2008, 4, 2373–2382 This journal is ª The Royal Society of Chemistry 2008
 View Article Online

respect to cationic surfactant addition. Thus, 2 wt% aqueous

dispersions at pH ¼ 9.8 undergo extensive sedimentation due to
flocculation between 0.02 and 0.2 mM di-C10DMAB, with the
particle z potential increasing from 41 mV without surfactant
to 1 mV at 0.1 mM surfactant and then to +57 mV at 1 mM
surfactant. As before, foams are most stable at intermediate
surfactant concentrations, particularly around 0.1 mM. The
SEM image in Fig. 10(a) of the surface of a bubble stabilised by
Published on 25 September 2008. Downloaded by KYUSHU UNIVERSITY on 25/01/2018 03:27:24.

surfactant alone shows no features as expected. In mixtures, we

see that the surface concentration of particles around bubbles Fig. 11 Appearance of foams 48 h after formation in mixtures of 0.6 mm
decreases progressively with increasing surfactant concentration, silica particles (3 wt%) and CTAB surfactant (initial concentration given)
Fig. 10(b)–(g). Thus we can conclude that foam stabilisation at pH ¼ 10.
changes from surfactant dominated at low concentration
(unstable) to particle dominated at intermediate
concentration (very stable) to surfactant dominated at high Table 1 Foamability (V0) and a measure of foam stability measured
concentration (unstable). Other features are worthy of comment. after 24 h. (V24) cm3 of 10 cm3 aqueous CTAB solutions alone and mixed
In Fig. 10(b), the wavy nature of a particle-coated bubble surface CTAB/0.6 mm silica dispersions (3 wt% particles) at pH ¼ 10. No foam
could be formed from 3 wt% silica dispersions alone. The stability to
is evidenced and, in the same system, an image of the air–water flocculation of the mixed dispersions before aeration is also given
surface edge-on reveals the contact angle of the particles
measured into the aqueous phase of around 90
[Fig. 10(c)]. Surfactant + Particles
A monolayer of close packed particles on a bubble surface seen
[CTAB]/mM V0 V24 V0 V24 Mixed dispersions
clearly in Fig. 10(d) is responsible for the highest foam stability in
this series. Above this surfactant concentration, the coverage by 0.01 0 0 2.5 1.4 Stable
particles decreases progressively [Fig. 10(e) and (f)] until at 10 0.05 5 0 8.0 8.0 Flocculated
mM surfactant bubble surfaces are more or less devoid of 0.09 11 0 8.0 8.0 Most flocculated
0.50 23 0 8.0 7.3 Flocculated
particles [Fig. 10(g)] and foams are completely unstable to all 0.90 30 0 12.0 0 Stable
processes within a couple of days. The maximum in foam 5.00 60 0 22.0 0 Stable
stability as a function of surfactant concentration is directly
linked to the changes in particle wettability from initially
hydrophilic (bare) to hydrophobic (monolayer coated) to
hydrophilic again (bilayer coated).

(d) Comparison with larger silica particles and CTAB

surfactant

In order to establish if the findings reported above for a nano-
particle/di-chain cationic surfactant mixture are general, we
investigated in less detail the mixture containing 0.6 mm mono-
disperse silica particles and the single chain cationic surfactant
CTAB at pH ¼ 10. As before, no foam could be generated by
hand shaking aqueous dispersions of particles alone (up to
5 wt%), and foams prepared from surfactant solutions alone
collapsed completely within one day at all concentrations (0.01 to
10 mM). For particle–surfactant mixtures at 3 wt% particles, the
stability to flocculation of aqueous dispersions changed from
stable to unstable to stable with increasing surfactant concen-
tration. Their foamability increased progressively with surfactant
concentration. However, the foam stability measured after 48 h
(and monitored up to 6 months after) again passed through
a maximum, with the position of the maximum corresponding to
the most flocculated precursor dispersions as seen in Fig. 11.
A summary of these findings is given in Table 1. The most stable
foams formed between 0.05 and 0.5 mM contained all the
particles originally present as none were visible in the aqueous
phase following drainage. These results are in close agreement
with those described previously. The picture emerging regarding
foam stabilisation in particle–surfactant mixtures of opposite
charge from the results of the present investigation is given
schematically in Fig. 12. The change in the particle charge and
Fig. 12 Schematic representation of stabilisation of aqueous foams by
silica particle–cationic surfactant mixtures at high pH. At low [surfactant]
(left), relatively unstable bubble surfaces contain mainly surfactant
molecules with a low percentage of partially coated anionic particles. At
intermediate [surfactant] (centre), very stable bubbles are coated with
uncharged, hydrophobic particles containing an adsorbed surfactant
monolayer. At high [surfactant] (right), very unstable bubbles contain an
adsorbed surfactant layer; cationic particles coated with surfactant
bilayers and surfactant micelles remain dispersed in the aqueous phase.
hydrophobicity induced by adsorption of surfactant triggers the
drastic change in foam stability as particles replace surfactant on
bubble surfaces at intermediate concentrations.
It is worth pointing out that some of our findings relating to the
formation of stable bubbles in water are consistent with those by
Alexandrova et al.29 in the field of flotation. Their study was on the
stability of the aqueous film between a silica substrate and an air
bubble in the presence of a cationic surfactant (or collector). At very
low surfactant concentrations, the film is stable and so contact
between the solid and vapour does not exist since both surfaces are
negatively charged and repel each other. At intermediate concen-
trations, preferential adsorption of surfactant at the air–water
surface making it positively charged leads to attraction between the
two surfaces (heterocoagulation) and an unstable aqueous film. At

This journal is ª The Royal Society of Chemistry 2008 Soft Matter, 2008, 4, 2373–2382 | 2381

high concentrations, clusters of adsorbed surfactant aggregates
lead to charge reversal of the solid surface and the re-stabilisation of
the aqueous film due to repulsion. The similarity between the
behaviour of our aqueous foams and their flotation system in
surfactant–particle mixtures is evident.
Conclusions
A detailed investigation into the behaviour of air-in-water foams
Published on 25 September 2008. Downloaded by KYUSHU UNIVERSITY on 25/01/2018 03:27:24.
stabilised by a mixture of Ludox HS-30 silica nanoparticles and
di-C10DMAB cationic surfactant at high pH has been made.
Particles alone are ineffective foamers being highly charged and
hydrophilic whereas surfactant-stabilised foams break down
completely within a day at all concentrations. In mixtures, the
synergism between the foaming agents leads to enhanced foam
stability at intermediate surfactant concentration. Complemen-
tary experiments have been undertaken in order to offer an
explanation for the latter synergy. It is shown that surfactant
addition initially transforms particles from anionic to uncharged
and hydrophobic and subsequently to cationic as a result of
adsorption. The most hydrophobic particles, possessing an
adsorbed monolayer of surfactant, yield foams which are
completely stable to disproportionation and coalescence. Cryo-
SEM analysis of frozen foams reveals the changeover from
surfactant to particle dominated bubble surfaces in line with the
dramatic increase in stability.
Note added in proof
A paper by S. Zhang, Q. Lan, Q. Liu, J. Xu and D. Sun, Colloids
Surf., A, 2008, 317, 406, was published recently on foams stabi-
lised by clay particle–cationic surfactant mixtures.
Acknowledgements
The authors thank Unilever Corporate Research for the provi-
sion of a postdoctoral grant to JAR. They also thank Ms D.A.
Braz, University of Hull, for measuring the data in Table 1.
References
1 P. R. Garrett, J. Colloid Interface Sci., 1979, 69, 107.
2 R. Aveyard, B. P. Binks, P. D. I. Fletcher, T. G. Peck and
C. E. Rutherford, Adv. Colloid Interface Sci., 1994, 48, 93.
2382 | Soft Matter, 2008, 4, 2373–2382
View Article Online
3 N. D. Denkov, Langmuir, 2004, 20, 9463, and references therein.
4 F.-Q. Tang, Z. Xiao, J.-A. Tang and L. Jiang, J. Colloid Interface Sci.,
1989, 131, 498.
5 H. Kumagai, Y. Torikata, H. Yoshimura, M. Kato and T. Yano,
Agric. Biol. Chem., 1991, 55, 1823.
6 Colloidal Science of Flotation, Surfactant Science Series, ed. A. V.
Nguyen and H. J. Schulze, Marcel Dekker, New York, 2004, vol.
118.
7 R. G. Alargova, D. S. Warhadpande, V. N. Paunov and O. D. Velev,
Langmuir, 2004, 20, 10371.
8 B. P. Binks and T. S. Horozov, Angew. Chem., Int. Ed., 2005, 44,
3722.
9 U. T. Gonzenbach, A. R. Studart, E. Tervoort and L. J. Gauckler,
Angew. Chem., Int. Ed., 2006, 45, 3526.
10 S. Fujii, A. J. Ryan and S. P. Armes, J. Am. Chem. Soc., 2006, 128,
7882.
11 K. Vijayaraghavan, A. Nikolov and D. Wasan, Adv. Colloid Interface
Sci., 2006, 123–126, 49.
12 B. P. Binks and R. Murakami, Nat. Mater., 2006, 5, 865.
13 Colloidal Particles at Liquid Interfaces, ed. B. P. Binks and T. S.
Horozov, Cambridge University Press, Cambridge, 2006.
14 Z. Du, M. P. Bilbao-Montoya, B. P. Binks, E. Dickinson, R. Ettelaie
and B. S. Murray, Langmuir, 2003, 19, 3106.
15 A. B. Subramaniam, C. Mejean, M. Abkarian and H. A. Stone,
Langmuir, 2006, 22, 5986.
16 M. Abkarian, A. B. Subramaniam, S.-H. Kim, R. J. Larsen, S.-
M. Yang and H. A. Stone, Phys. Rev. Lett., 2007, 99, 188301.
17 A. Cervantes Martinez, E. Rio, G. Delon, A. Saint-Jalmes,
D. Langevin and B. P. Binks, Soft Matter, 2008, 4, 1531.
18 E. Dickinson, R. Ettelaie, T. Kostakis and B. S. Murray, Langmuir,
2004, 20, 8517.
19 B. P. Binks and S. O. Lumsdon, Langmuir, 2000, 16, 8622.
20 B. P. Binks and J. A. Rodrigues, Angew. Chem., Int. Ed., 2007, 46,
5389.
21 V. W. Reid, G. F. Longman and E. Heinerth, Tenside, 1967, 4, 292.
22 B. P. Binks and M. Kirkland, Phys. Chem. Chem. Phys., 2002, 4,
3727.
23 K. Furusawa, A. Sato, J. Shirai and T. Nashima, J. Colloid Interface
Sci., 2002, 253, 273.
24 D. W. Fuerstenau and R. Jia, Colloids Surf., A, 2004, 250, 223.
25 K. Esumi, M. Matoba and Y. Yamanaka, Langmuir, 1996, 12, 2130.
26 R. Aveyard, B. P. Binks, P. D. I. Fletcher and C. E. Rutherford,
Colloids Surf., A, 1994, 83, 89.
27 E. A. Vogler, Langmuir, 1992, 8, 2005.
28 V. V. Yaminsky and K. B. Yaminskaya, Langmuir, 1995, 11, 936.
29 L. Alexandrova, R. J. Pugh, F. Tiberg and L. Grigorov, Langmuir,
1999, 15, 7464.
30 L. K. Koopal, T. Goloub, A. de Keizer and M. P. Sidorova, Colloids
Surf., A, 1999, 151, 15.
31 L. G. J. Fokkink and J. Ralston, Colloids Surf., 1989, 36, 69.
32 S.-G. Oh and D. O. Shah, Langmuir, 1991, 7, 1316.
33 A. B. Subramaniam, M. Abkarian, L. Mahadevan and H. A. Stone,
Nature, 2005, 438, 930.
This journal is ª The Royal Society of Chemistry 2008