Journal of Colloid and Interface Science 338 (2009) 201206

Contents lists available at ScienceDirect

Journal of Colloid and Interface Science

www.elsevier.com/locate/jcis

Emulsication mechanism and storage instabilities of hydrocarbon-in-water

sub-micron emulsions stabilised with Tweens (20 and 80), Brij 96v
and sucrose monoesters
John V.L. Henry a, Peter J. Fryer a, William J. Frith b, Ian T. Norton a,*
a
Department of Chemical Engineering, University of Birmingham, Edgbaston, Birmingham B15 2TT, UK
b
Unilever Corporate Research, Colworth Park, Sharnbrook MK44 1LQ, UK

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

Article history: The inuence of both the nature of the surfactant and surfactant concentration on the processes of drop-
Received 1 April 2009 let break-up and coalescence in the formation of decane-in-water nano-emulsions in a high-pressure
Accepted 29 May 2009 homogenizer was investigated. Emulsions were produced using a Christison Scientic M110-S microui-
Available online 6 June 2009
diser with an impinging jet high-shear chamber. For all six surfactants studied (Tween 20, Tween 80, Brij
96v, sucrose monolaurate, sucrose monomyristate and sucrose monopalmate), the droplet size decreased
Keywords: with increasing surfactant concentration reaching a limiting droplet size at a surfactant concentration of
Mini-emulsion
15 mM. The limiting droplet size for the different surfactants used were; Tween 20 (250 30 nm),
Nano-emulsion
Emulsication
Tween 80 (320 40 nm), Brij 96v (200 20 nm) and the three sucrose monoesters had very similar
Impinging jets sizes of 250 20 nm. A hydrophobic uorescent dye (1-undecylpyrene) was used to establish the extent
Droplet break-up of competition between droplet break-up and coalescence in the emulsication process. For all the emul-
Coalescence siers studied, droplet coalescence in the process reduced as the amount of emulsier increased, becom-
Ostwald ripening ing zero at concentrations of about 15 mM, i.e. the same concentration as that required to produce the
limiting minimum droplet size. This shows that in the emulsication process droplet size is determined
by both break-up and re-coalescence events, and at lower surfactant concentrations (<15 mM) that the
nal droplet size is probably a consequence of multiple break-up events. Emulsion stability over 200 h
was investigated by measuring changes in the droplet size using dynamic light scattering. The increase
in droplet volume was shown to be linear with respect to time, indicating an Ostwald ripening process.
The observed ripening rate for the three sucrose monoesters (monopalmitate, monomyristate and mon-
olaurate) was approximately 20 nm3 s1, which is the ripening rate calculated using the LifshitzSlesov
Wagner (LSW) theory. This ripening rate is the change in radius that results from movement of the oil
through the continuous phase, taking into account the oil solubility in water and the diffusion coefcient
of the decane-in-water. The ripening rate for Brij 96v was about three times larger than the calculated
rate and there is an indication that the ripening rate increases slightly with increasing surfactant concen-
tration, indicating that some enhancement due to the presence of micelles has occurred. With Tween 80
and 20 the ripening rates were 20 and 40 times, respectively, larger than those calculated using the sol-
ubility and diffusion coefcients. The increased rate has been shown to be rst order with respect to the
surfactant concentration indicating micelle mediated ripening. It is hypothesized that an optimum for-
mulation for the sub-micron emulsion with these types of surfactant, will balance surfactant concentra-
tion to minimize droplet size during processing while aiming to minimize or prevent Ostwald ripening.
2009 Elsevier Inc. All rights reserved.

1. Introduction emulsions are of interest primarily because of their droplet size,

In recent years, there has been an increasing interest in the lit-
erature regarding the formation and use of nano- or mini-emul-
sions. These are emulsions that have droplet sizes in the 100 nm
range, but are kinetically stabilized rather than being thermody-
namically stable [1,2], as is the case for micro-emulsions [3]. Such
* Corresponding author.
E-mail address: I.T.Norton@bham.ac.uk (I.T. Norton).
however their potential has also been investigated in such
applications as functional ingredient encapsulation [4] and
nano-structured multiple emulsions [5,6]. Applications studied
for nano-emulsions include polymerization and other reaction
processes [7,8], drug delivery [9], cosmetics [10], and novel foods
[11]. A range of processes have been investigated for the produc-
tion of nano-emulsions, including temperature or concentration
induced phase inversion [12,13] and high-pressure homogeniza-
tion processes [14]. Here, we investigate the formation of

0021-9797/$ - see front matter 2009 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcis.2009.05.077
202 J.V.L. Henry et al. / Journal of Colloid and Interface Science 338 (2009) 201206
nano-emulsions in a high-pressure impinging jet device (the
Microuidizer M110-S), in order to understand the factors that
limit the droplet size. It has been shown [15,16] that the droplet
size produced from such a process is a consequence of the energy
input, driving droplet break-up, and the coalescence rate arising
from droplet collisions combined with the inability of the surfac-
tants to give full coverage of the increased surface area formed in
the process.
It is this coalescence during the emulsication process that is
often the limiting factor in the size reduction in the emulsication
process [17]. Emulsion interfaces are formed in shear ow by three
mechanisms; binary break-up, capillary break-up and tip dropping
[18] and in turbulent ow by the forces experienced within eddies
[19]. Oilwater interfaces formed may then be made stable against
coalescence through surfactant adsorption. If the surfactant cannot
reach and coat the naked interface formed in the break-up event
before droplets collide then coalescence occurs. At low surfactant
concentrations this determines the droplet size, not the amount
of energy dissipated in the process [17].
Of interest to foods and cosmetics industries is the use of essen-
tial oils. These are used to impart avors and aromas. In order to
maximise the positive attributes in a product increased surface
area of the emulsion phase is an advantage. Thus there is a desire
to decrease the droplet size. However, this increases the Laplace
pressure and leads to an increased rate of ripening. In this paper,
decane-in-water nano-emulsions were produced. Decane was cho-
sen as the model oil phase in an experimental study of the behavior
of an edible emulsion systems for three reasons: (1) decane is
chemically stable, being an alkane (this is not the case with essen-
tial oils/volatile aromatic avor compounds (VAFCs) which are of-
ten unsaturated and oxidise over time); (2) the composition of
essential oils/VAFCs vary from batch to batch so any results are
batch specic and (3) decane like essential oils/VAFCs is partially
soluble in water [20]. The use of decane guaranteed the uniform
purity and chemical stability of the oil phase over the course of
the study. It was desirable from the point of view of the study that
the model oil phase used was of low viscosity so as to facilitate the
efcient emulsication of ne droplets, [21] and for the measure-
ment of the emission spectrum of the uorescent dye used to mea-
sure coalescence [22].
The surfactants studied in this paper Tween 20, Tween 80, Brij
96v and sucrose monoesters are commonly used as emulsion sta-
bilizers in a wide range of food and personal care products
[10,17,23]. It is envisaged that nano-emulsions might offer signi-
cant potential as functional ingredients in foods and personal care
products both in terms of their effectiveness to deliver avors and
aromas due to the increased surface area, and as mechanisms of
delivering oil soluble bio-actives. Therefore it would be advanta-
geous to formulate nano-emulsions with surfactants, which have
gained widespread acceptability in formulations of these types.
From a scientic viewpoint, the surfactants studied are of interest
because of their similarities, and fundamental differences. On one
hand, they have similarly low molecular weights. On the other
hand, there are differences in their structures, which might lead
them to behave very differently at the oilwater interface, particu-
larly the sucrose monoesters. Tween-type surfactants possess a
branched hydrophilic region of three polyoxyethylene chains
substituted to a sorbitan ring, while Brij 96v has a far more linear
molecular structure in its hydrophilic region. The sucrose monoes-
ters have a large head group that can potentially aggregate at the
oil/water interface.
The nal droplet size of an emulsion arises from the two com-
peting processes; droplet break-up and coalescence. The question
is then which effect predominates as a determinant of nal droplet
size, and how does this depend on surfactant concentration? One
aim of this work is to measure the degree of coalescence occurring
during the homogenization process in a high-pressure opposed-jet
homogenizer, and how this is affected by the type and concentra-
tion of surfactant present. These investigations are in turn related
to the droplet sizes produced from the device. Previous work has
addressed this question a number of ways; by uorescent dye mix-
ing, Lobo et al. [24,25]; by microscopic determination of mixing of
two dyes in microscopic droplets, Danner et al. [26], using com-
puter simulation, Cristini et al, [27], by turbidity [28] and by addi-
tion of brominate to oil to change the refractive index of half of the
droplets followed by observation of turbidity changes as a result of
changes in refractive index as a consequence of droplet mixing
during emulsication [29]. Although this technique is not particu-
larly useful for nano-emulsions as droplets approach the wave-
length of light.
The common principle of the above experimental techniques is
that the chemical label used to study break-up and coalescence is
soluble only in one phase, and so dye transfer from droplet to drop-
let can only occur via coalescence. Following a previous study of
pyrene uorimetry in alkane-in-water emulsions [30], Lobo et al.
[24,25] used an oil-soluble uorescent probe, 1-undecylpyrene
(UDP) to measure the amount of coalescence that occurred during
the emulsication of tri-2-ethylhexyl phosphate in a high-pressure
homogenizer. The approach we have adopted follows that of Lobo
et al. [24,25] and uses the same water insoluble uorescent dye,
UDP, dissolved in decane, to determine the degree of coalescence
occurring during formation of a decane-in-water emulsion system.
UDP allows the determination of coalescence in emulsions be-
cause when it is dissolved at a sufciently high concentration, it
will exhibit an excimer peak in its uorescence spectrum. At lower
concentrations, however, this excimer peak diminishes and when
sufciently dilute, is not present. Such behavior allows the mea-
surement of the concentration of UDP spectroscopically and thus
it can be used to measure the degree of coalescence (dilution)
occurring in emulsions.
Nanometre droplets are more susceptible to degradation by
Ostwald ripening [31] on storage as a consequence of the increased
Laplace pressure. In a previous study by Hoang et al. [32], ripening
rates of decane-in-water emulsions (containing micron sized drop-
lets) stabilized by Tween 20 were determined at several surfactant
concentrations; the dilute (0.5% and 1% v/v) emulsions showed
that the ripening rates were increased by the presence of surfac-
tant micelles. Ripening rates of emulsions have been shown to be
faster with higher phase volumes [33], and addition of excess sur-
factant [34].
Here we present results for droplet coalescence during homog-
enization and the subsequent ripening rates of decane-in-water
sub-micron emulsions stabilized by polyoxyethylene sorbitan
monolaurate (Tween 20), polyoxyethylene sorbitan monooleate
(Tween 80), decaethylene glycol monooleyl ether (Brij 96v) and a
range of sucrose monoesters. These data are then discussed in
terms of their implications for the inuences of surfactant compo-
sition and structure on the mechanism of emulsion formation and
subsequent stability.
2. Materials and methods
2.1. Materials and emulsion preparation
Several emulsions of decane (Aldrich, purity 99+%) in de-ionized
water stabilized by surfactants were prepared. The surfactants
used were Tween 20 (polyoxyethylene sorbitan monolaurate)
SigmaUltra, P7949, Tween 80 (polyoxyethylene sorbitan monool-
eate) Sigma-Aldrich Cat. No. HO03431EO, Brij 96v (decaethylene
glycol monooleyl ether) Fluka Cat. No. 16011. Sucrose monoesters
were obtained from Mitsubishi Kagaku Foods Corporation, Japan.
 J.V.L. Henry et al. / Journal of Colloid and Interface Science 338 (2009) 201206 203

The sucrose monolaurate (Ryoto sugar ester L-1695) was 84.6% mon-
olaurate and 15.4% dilaurate, the sucrose monomyristate (Ryoto su-
gar ester M-1695) was 74% monomyristate and 21% dimyristate,
and the sucrose monopalmitate (Ryoto sugar ester P-1570) was
56% monopalmitate, 19.5% dipalmitate, 12.5% monostearate and
4.5% distearate. The monomyristate and monopalmitate also con-
tained some moisture and very low levels of residual fatty acids.
The uorescent dye 1-undecylpyrene was synthesized accord-
ing to a synthetic route provided by Lloyd Lobo in a personal com-
munication and discussed in Lobo et al. [22,24,25]: details are
given in the thesis by Henry [35].
The preparation of the pre-emulsion was as follows; decane was
added to the aqueous surfactant solution. After 60 s of mixing with
a Silverson L4R Laboratory mixer, with ne emulsor attachment,
this coarse emulsion was then further homogenized using a single
pass through a M110-S Microuidizer with a Microjet high-shear Fig. 1. Ratio of UV peaks at 477 and 376 nm for UDP in decane versus concentration
of UDP.
chamber at a pressure of 1035 bar.
In this way, emulsions with decane at 0.1 volume fraction and
surfactant concentrations in the range 1.0  1032.5  102 M 2.3. Droplet size measurement
were prepared. Emulsions with 6.0  103 M UDP in the decane
phase were similarly prepared. Samples were taken from both Droplet sizes were determined with Dynamic Light Scattering
labeled and unlabeled emulsions and mixed vigorously for 10 s (DLS) on an ALV/CGS-8F S/N 052 laser goniometer system as sup-
by hand in a 1:1 ratio and then re-microuidized in a single pass plied by ALV GmbH, D-63225 Siemensstrasse 4, Langen/Hessen
for the purposes of measuring competition between droplet Germany. Before each measurement, the sample was prepared by
break-up and coalescence. diluting the original by about 10,000 times with de-ionized water.
Intensity averaged radii were computed from the intensity auto-
2.2. Fluorescence measurements correlation data using the cumulants method and z average values
used for all analysis. Samples were stored in a refrigerator at 3 C
Fluorescence measurements were made on a Perkin Elmer between measurements.
Luminescence Spectrometer LS50B. Because an emulsion was un-
der examination, the front surface accessory was used to obtain
3. Results and discussion
the emission. A 50 ll aliquot of emulsion was diluted in 3 ml of
de-ionized water in a Kartell 4.5 ml PMMA 1961 uorimetry cuv-
3.1. Initial droplet sizes
ette as supplied by Kartell Plastics UK Ltd., Unit C, Broad Lane, Cot-
tenham, Cambridge, CB4 8SW, UK, and measurements made
The dependence of droplet size on surfactant concentration was
immediately to avoid any effects of coalescence as a consequence
studied. Fig. 2, shows results for the freshly made samples. As the
of the dilution (all measurements were made within 1 min of
concentration of surfactant is increased (05 mM) the droplet size
emulsion production). Triplicate measurements were carried out.
initially decreases steeply. At surfactant concentrations between 5
Excitation was at a wavelength of 340 nm. Emission acquisition
and 10 mM the decrease in droplet size with increasing surfactant
wavelength was between 370 nm and 500 nm. The two monomer
concentration is much more gradual. A limiting value is reached at
peaks monitored were at 376 and 397 nm. The excimer peak was
measured at 477 nm. Each spectrum analyzed was corrected by
subtracting a value for the unlabeled emulsion as a baseline.
The ratio of the magnitude of the excimer peak intensity at
477 nm to the monomer peak intensity at 376 nm was used for
each spectrum and a calibration graph versus UDP concentration
was constructed between 0.25 mM and 8.0 mM and found to have
a linear relationship (Fig. 1). To establish the extent of coalescence,
two emulsions were prepared as described above, one with 6 mM
UDP in the decane, and one without. The uorescence spectrum of
this mixture was recorded, the mixture was then immediately re-
microuidized by a single pass through the M110S, and the uo-
rescence spectrum was recorded a second time. The uorescence
spectrum of the unlabeled emulsion was recorded, and its value
subtracted from the values of the labeled emulsion, and the unla-
beled emulsion, to remove the contribution to uorescence of the
oil phase and the surfactant in subsequent calculations. Following
the procedure described by Lobo [24], the extent of mixing of the
dyed and un-dyed oil phases was calculated for the emulsion spec-
tra before and after the second microuidization process. The ratio
of the peaks before re-emulsication/peak after re-emulsication
was calculated. This ratio should have values that range between
1 (no mixing between the labeled and unlabeled droplets) and 2
(full mixing between the labeled and unlabeled droplets). The ratio Fig. 2. Initial droplet size of decane emulsions; Tween 20 (d); Tween 80 (s); Brij
was thus used to calculate the percentage of oil re-coalesced in the 96v (j) and sucrose monopalmitate (N). Error bars are standard deviations
second emulsication step. obtained from triplicate runs.
204 J.V.L. Henry et al. / Journal of Colloid and Interface Science 338 (2009) 201206
15 mM concentration for all six emulsiers studied. The minimum
droplets sizes obtained were somewhat different for the various sur-
factants. Tween 80 gave a minimum size of 320 40 nm, Tween 20
minimum size was 250 30 nm, the sucrose esters all had similar
sizes: monolaurate 230 20 nm, monopalmitate 250 20 nm
and monomyristate 260 20 nm. Brij 96v gave the smallest mini-
mum droplet size of approximately 200 20 nm. The similarities
of these minimum droplet sizes and the relationship between the
droplet size and surfactant concentration indicates that the size is
mainly determined by the homogenization process rather than by
either the interfacial tension or molecular structure of the surfac-
tant. The saturated interfacial tensions for Tween 20 and 80 are
5 mN m1 [36], for Brij 4 mN m1 and for the sucrose esters be-
tween 7 and 10 mN m1 [37] which do not correlate with the se-
quence of limiting droplet sizes observed here.
Further support for the argument that the droplet size is not
very dependent on the surfactant can be obtained from considering
the potential surface coverage of the surfactants at all the concen-
trations studied. In order to consider if the droplets sizes we obtain
are dependent upon the amount of surfactant present we can start
by taking the simplistic assumption that all of the surfactant is on
the interface as a monolayer for the emulsions formed at 1 mM.
This is obviously not totally true as there will be an equilibrium be-
tween the surface excess and the bulk. However, by making this
assumption we can calculate the head group size for the surfac-
tants. This gives sizes of 0.97 nm2 for Tween 20, 0.76 nm2 for
Tween 80, (which compare very well with measured sizes reported
in the literature [36] of 0.97 nm2 and 0.61 nm2, respectively),
0.69 nm2 for Brij 96v and 0.56 nm2 for the sucrose esters. Using
these estimated values we can then calculate the proportion of
each surfactant required to give surface coverage for the emulsions
produced in Fig. 2. Fig. 3 shows that the amount of surfactant cal-
culated to give surface coverage decreases with increasing concen-
tration of surfactant reaching about 2030% at 25 mM and that at
surfactant concentrations above 5 mM there is excess surfactant
present. Thus in our study the droplet size is not limited by the
amount of surfactant present at concentrations above 1 mM.
It is clear from Figs. 2 and 3 that neither the surfactant concen-
tration nor the surfactant type is limiting the droplet size produced
in the microuidizer, suggesting that the homogenization process
is determining the droplet size. As reported [17] the droplet size
seems to be determined initially by the rate at which the surfactant
reaches and stabilises the interface as it forms. Thus it would be ex-
pected that the droplet size would decrease with increasing surfac-
Fig. 3. Calculated fraction of surfactant at the oilwater interface; assuming
monolayer coverage. Tween 20 (h), Tween 80 (d) and Brij 96v (s) sucrose
monopalmitate (j), sucrose monomyristate (N) and sucrose monolaurate (M).
tant concentration until the limiting value is reached where break-
up of droplets in the homogeniser (the forward reaction) is
matched by the re-coalescence process (the back reaction). This
is dependent upon the rate at which the surfactant can enter the
interface and stabilise naked interface as it is formed in the pro-
cess. At surfactant concentrations of 15 mM it appears that the
break-up dominates. In order to check this we used a concentration
dependent dye to investigate coalescence in the microuidizer.
3.2. Coalescence studies
Following on from this previous work it is important to disen-
tangle the steps in the emulsication process. We investigated this
using a technique developed by Lobo [22,24,25] in which we stud-
ied the extent of coalescence after droplet break-up in a Microjet
high-shear chamber as a function of surfactant type and concentra-
tion. An oil phase volume of 0.1 was used in these experiments.
Fig. 4, shows the results in the form of percentage coalescence that
has occurred in the process for the Tweens, Brij 96v and sucrose
monopalmitate. The results obtained for the sucrose monoesters
overlaid the results shown and have been excluded for clarity.
Fig. 4 shows a steady decline in the calculated coalescence
events with increasing concentration of the surfactants used. All
six surfactants follow the same trend with an almost linear de-
crease in the calculated amount of coalescence occurring with
increasing concentration of surfactant until the data suggests that
no measurable coalescence occurs above 15 mM.
The amount of surfactant required to stop coalescence occur-
ring in the process is very similar to the concentration to give
the minimum droplet size in the process. This shows that at
15 mM surfactant concentrations and above the process is deter-
mining the droplet size and the surfactant properties have little
or no effect, i.e. the forward rate is dominant. At concentrations be-
low 15 mM the nal droplet size produced from the process is a
consequence of multiple coalescence and break-up events.
3.3. Evolution of droplet size with time
Fig. 5 illustrates droplet growth data for 10 vol.% decane-in-
water emulsions stabilized by Tween 80. As can be seen there is
considerable ripening of the system over the 200 h of the study
and the rate of ripening increases as the concentration of the sur-
factant is increased.
The ripening rate can be obtained by plotting the increase in
volume of the droplets as a function of time, Fig. 6. If Ostwald rip-
ening is the predominant mechanism for the increase in droplet
Fig. 4. Droplet coalescence of mixed 1:1 ratio of decane-in-water emulsion on re-
homogenization using the microuidizer. Brij (N),Tween 20 (j),Tween 80 (s) and
sucrose monopalmitate (h).Error bars are standard deviations obtained from
multiple runs.
 J.V.L. Henry et al. / Journal of Colloid and Interface Science 338 (2009) 201206
Fig. 5. Z average droplet radius versus time for decane emulsions at different
Tween 80 concentrations; () 2.5 mM, (h) 5 mM, (d) 10 mM and (M) 20 mM. Error
bars are standard deviations obtained from multiple measurements of the stored
emulsions.
size, as opposed to coalescence, then the mean volume should in-
crease as a linear function of time. As can be seen from Fig. 6 the
increase in mean volume observed at all surfactant concentrations
give reasonably good straight lines and the ripening rate increases
with surfactant concentration.
The ripening rate can be calculated using LifshitzSlesovWag-
ner (LSW) theory [38,39]
x dr3N =dt 8=9C1cV m D=qRT
where x is the Ostwald ripening rate; rN the droplet radius; C(1)
the bulk phase solubility of the oil; c the interfacial tension; Vm
the molar volume of the oil; q the density of the dispersed phase;
D the diffusion coefcient of the dispersed phase in the bulk phase;
R the gas constant and T is the temperature (K).
Using this equation, the Ostwald ripening rate, x, for decane in
water emulsion with no surfactant (interfacial tension
52.6 mN m1) would be 72 nm3 s1. As the interfacial tension is
lower in the presence of surfactant (in all the cases studied here
10 mN m1) the ripening rates calculated for the surfactants
studied here is approximately 1520 nm3 s1.
The measured ripening rates are plotted in Figs. 7 and 8 for the
surfactants studied at concentrations of 2.5 mM and above. The
data obtained at 1 mM is excluded as the starting droplet sizes
Fig. 6. Droplet volume versus time for decane emulsions stabilised with Tween 80.
() 2.5 mM, (h) 5 mM, (N) 10 mM and (s) 20 mM. Error bars calculated from Fig. 5.
205
Fig. 7. Ripening rates (x) as a function of surfactant concentration for surfactant
stabilized decane emulsions, sucrose monoplamitate (h), sucrose monomyristste
(N), sucrose monolaurate (d) and Brij 96v (s).
were large and creaming and coalescence on storage was a prob-
lem. The lines shown in Fig. 8 are an exponential t to the data.
Although there are signicant errors in the data they do seem to
t an exponential dependency and a plot of log rate against log sur-
factant concentration gave straight lines with gradients close to 1.
Thus the process is rst order. This is in contrast to what might be
expected, as the concentration of emulsier is increased the Ost-
wald ripening rate should initially decrease until the critical mi-
celle concentration is reached, at which point the interfacial
tension remains constant and therefore so should the ripening rate.
The CMCs for the different surfactants are 0.15 mM for Tween 20
and 0.6 mM for Tween 80 [36], 0.07 mM for Brij [40], and
0.5 mM for sucrose monolaurate, 0.04 mM for sucrose monomyri-
state and 0.02 mM for sucrose monopalmitate [37], indicating that
for all surfactant concentrations studied here that concentrations
in the aqueous phase are above the CMC.
Fig. 7 shows that for the sucrose monoesters the ripening rates
are independent of the concentration of surfactant and are approx-
imately 20 nm3 s1. The ripening rate measured for Brij 96v are
approximately 60 nm3 s1, with some indication that the rate in-
creases as the surfactant concentration was increased. These obser-
vations may be explained in the former case as simple ripening
caused by the solubility of the oil in water (for decane this is
2.05  108 v/v), and in the latter case enhance ripening as a con-
sequence of the presence of micelles [41]. This type of dependency
can be more clearly seen in Fig. 8, which shows that increasing the
concentration of Tween leads to an increase in ripening rates.
Tween 80 reaches a plateau with a ripening rate of approximately
Fig. 8. Ripening rates (x) as a function of surfactant concentration for nonionic
surfactant stabilized decane emulsions. Tween 20 (j) and Tween 80 ().
206 J.V.L. Henry et al. / Journal of Colloid and Interface Science 338 (2009) 201206
400 100 nm3 s1, and Tween 20 reaches a plateau ripening rate of
800 200 nm3 s1. Plots of increase in droplet volume as a function
of time were shown to be linear, indicating a genuine ripening
process.
The measured ripening rate for Brij 96v is approximately three
times the calculated value which is similar to the observations re-
cently reported by Wooster et al. [2]. These authors implied that
such increases are not unexpected and indeed the lack of depen-
dency on surfactant concentration suggests a simple dependency
on the decane solubility in water. The 40 fold increase with Tween
80 and 80 fold increase with Tween 20 are considerably larger than
and increases reported by Wooster et al. [2]. In addition the rate of
ripening tting to a rst order effect of surfactant concentration,
indicates that with Tween the ripening is now dominated by the
presence of micelles that transport decane from the smaller drop-
lets to the larger ones. As all of the studies are carried out at surfac-
tant concentrations at least 50 times the CMC of the surfactant,
then in order to understand the differences observed, we need to
consider how the different surfactant micelles can inuence the
transport of oil between droplets.
The observed differences in ripening behavior must lie in the
different molecular structures. The molecular structures of Tween
20 and Tween 80 are branched with a hydrophilic [42] region,
which is centered about a sorbitan ring esteried to 4 short-chain
oligomers of polyethylene oxide. Brij 96v and the sucrose monoes-
ters have no branching in this mono-functional structure [23,42], a
fact that may have several consequences to their behavior. It is
likely that the un-branched surfactants can form a more closely-
packed monolayer at the oilwater interface than either of the
Tween-type surfactants, thus resulting in less micelle mediated
transport of decane between oil droplets. This will form the basis
of further studies.
4. Conclusions
4.1. Droplet break-up versus coalescence
The uorescent probe technique has been used to gain insight
into the homogenization process in a high-pressure impinging jet
device; in particular the role of different, low molecular weight
surfactants in creating small droplets has been investigated. It
has been shown that above an optimal surfactant concentration
(15 mM) in the formulation, droplet coalescence in the device is re-
duced to a minimum. Measurements of the droplet size immedi-
ately after homogenization show that a limiting droplet size is
reached (at 15 mM), with increasing surfactant concentration.
These observations show that once the back coalescence is halted
the process itself mainly determines the droplet size. Droplets pro-
duced with lower levels of surfactant are a consequence of multi-
ple break-up and coalescence steps within the homogeniser.
4.2. Ostwald ripening study
For the sucrose monoesters the ripening rates were very close
to those calculated using the LSW theory taking into account the
interfacial tension. For the Brij 96v the ripening rate was depen-
dent of surfactant concentration at surfactant concentrations be-
low 10 mM and at surfactant concentrations above 10 mM a
factor of three higher than calculated using the LSW theory. Both
these factors point to a micelle enhanced ripening process. With
the Tween surfactants, as their concentration was increased, the
Ostwald ripening rate increased following a rst order dependency
and the observed rate was 2040 times higher than calculated
using the LSW theory. The increase in Ostwald ripening rates for
the Tween surfactants occurs at the concentration where there is
excess surfactant, indicating that the Ostwald ripening is micelle
mediated with the Tweens.
There is an optimum concentration of surfactant that balances
the desire for the formation of the smallest possible droplets, and
the apparently conicting aim of the formulation also being stable
over long periods. This concentration lies at the point where the
droplet size is governed by the ability of the process to create small
droplets, rather than by the ability of the available surfactant to
cover the droplets.
Acknowledgments
We would like to thank EPSRC for provision of a formulation
engineering EngD studentship to JVLH and Unilever Research, Col-
worth House for providing the facilities where the work was car-
ried out. We also acknowledge the assistance of Christison
Scientic in providing information regarding their Microjet High-
shear chamber and Dr. F. Spyropoulos for helpful discussion and
advice.
References
[1] J.V. Henry, W.J. Frith, P.J. Fryer, I.T. Norton, Foods Food Ingred. J. Jpn. 213 (3)
(2008) 192.
[2] T.J. Wooster, M. Golding, P. Sanguansri, Langmuir 24 (22) (2008) 12758.
[3] T.G. Mason, J.N. Wilking, K. Meleson, C.B. Chang, S.M. Graves, J. Phys.: Condens.
Matter 18 (2006) R635.
[4] D.J. McClements, E.A. Decker, J. Food Sci. 65 (2000) 1270.
[5] N. Garti, A. Benichou, Encyclopedic Handbook of Emulsion Technology, Marcel
Dekker, New York, 2001. 377407.
[6] N. Garti, A. Benichou, in: Food Emulsions, fourth ed., Marcel Dekker, New York,
2004, p. 353.
[7] K. Landfester, Macromol. Rapid Commun. 22 (12) (2001) 896.
[8] K. Landfester, N. Bechtold, F. Tiarks, M. Antonietti, Macromolecules 32 (1999)
5222.
[9] R.J. Prankerd, V.J. Stella, J. Parenteral Sci. Technol. 44 (1990) 139.
[10] F. Zlli, F. Suter, Seifen Oele Fette Wachse 123 (1997) 880.
[11] J. Weiss, P. Takhistov, D.J. McClements, J. Food Sci. 71 (9) (2006) R107.
[12] R. Ponsa, I. Carreraa, J. Caellesa, J. Rouchb, P. Panizzab, Adv. Colloid Interface
Sci. 106 (2003) 129.
[13] M. Minana-Perez, C. Gutron, C. Zundel, J.M. Anderez, J.L. Salager, J. Dispersion.
Sci. Technol. 20 (3) (1999) 893.
[14] S. Schultz, G. Wagner, K. Urban, J. Ulrich, Chem. Eng. Technol. 27 (4) (2004)
361.
[15] H. Schubert, R. Engel, Chem. Eng. Res. Des. 82 (A9) (2004) 1137.
[16] J.R. Canselier, H. Delmas, A.M. Wilhelm, B. Abismail, J. Dispersion Sci. Technol.
23 (2002) 333.
[17] I.T. Norton, F. Spyropoulos, P.W. Cox, Food Hydrocolloids 27 (6) (2009) 1521.
[18] R.A. de Bruijn, Deformation and break-up of drops in simple shear ows, Ph.D.
Thesis, Technische Univ., Eindhoven (Netherlands), 1989.
[19] P. Walstra, Chem. Eng. Sci. 48 (2) (1993) 333.
[20] C. Druaux, A. Voilley, Trends Food Sci. Technol. (8) (1997) 364.
[21] G. Kolb, K. Viardot, G. Wagner, J. Ulrich, Chem. Eng. Technol. 24 (2001) 293.
[22] L. Lobo, Synthetic Route of 1-undecylpyrene, Personal Communication, 2004.
[23] S. Makino, S. Ogimoto, S. Koga, Agric. Biol. Chem. 47 (2) (1983) 319.
[24] L. Lobo, J. Colloid Interface Sci. 254 (1) (2002) 165 (October 1).
[25] L. Lobo, A.J. Svereika, Colloid Interface Sci. 261 (2003) 498.
[26] T. Danner, M. Stang, H. Schubert, in: Proceedings of the 3rd World Congress in
Emulsions, Lyon, 2002.
[27] V. Cristini, J. Blawzdziewicz, M. Loewenberg, J. Comput. Phys. 168 (2001) 445.
[28] G. Narasimham, J.P. Gupta, D. Ramkrishna, Chem. Eng. Sci. 34 (1979) 257.
[29] L. Taisne, p. Walstra, B. Cabane, J. Colloid Interface Sci. 184 (1996) 378.
[30] T. Sakai, K. Kamogawa, K.O. Kwon, H. Sakai, M. Abe, Colloid Polym. Sci. 280
(2002) 99.
[31] I. Capek, Adv. Colloid Interface Sci. 107 (2004) 125.
[32] T.K.N. Hoang, V.B. La, L. Deriemaeker, R. Finsy, Langmuir 18 (2002) 1485.
[33] A.S. Kabalnov, E.D. Shuchukin, Adv. Colloid Interface Sci. 38 (1992) 69.
[34] D.J. Mc Clements, S.R. Dungan, J. Phys. Chem. 97 (1993) 7304.
[35] J.V.L. Henry, Engineering Doctorate Thesis, University of Birmingham, 2008.
[36] R.K.O. Apenten, Q.H. Zhu, Food Hydrocolloids 10 (1) (1996) 27.
[37] N. Becerra, L.R. de la Nuez, A.L. Zanocco, E. Lemp, G. Gurther, Colloids Surf. A
272 (2006) 2.
[38] I.M. Lifshitz, V.V. Slyozov, J. Phys. Chem. Solids 19 (1961) 35.
[39] C.Z. Wagner, Electrochemistry 65 (1961) 581.
[40] K.A. Yoon, D.J. Burgess, Pharm. Res. 13 (3) (1996) 433.
[41] T.K.N. Hoang, V.B. La, L. Deriemaeker, R. Finsy, Langmuir 19 (15) (2003) 6019.
[42] A. Yaghmur, A. Aserin, N. Garti, Colloids Surf. A 209 (1) (2002) 71.