Journal of Membrane Science 362 (2010) 111

Contents lists available at ScienceDirect

, Karin Schron
1
, Remko Boom
Wageningen University, Food Process Engineering Group, Bomenweg 2, 6703 HD Wageningen, The Netherlands
a r t i c l e i n f o
Article history:
Received 24 February 2010
Received in revised form 15 June 2010
Accepted 27 June 2010
Available online 31 July 2010
Keywords:
Premix membrane emulsication
Energy density
Droplet break-up mechanism
Process parameters
Glass beads
a b s t r a c t
Membrane emulsication is known to be a mild technique that renders narrowly dispersed emulsions
at energy inputs that are orders of magnitude lower than in traditional emulsication techniques. Cross-
owmembrane emulsicationis most investigatedandis knownfor the monodispersity of the emulsions
produced; however, this can only be obtained at relatively low disperse phase fraction. For emulsions
with higher disperse phase fractions, premix membrane emulsication is an interesting alternative that
is in our opinion on the verge of breaking through.
Principally, in this mild process, a coarse premix is pushed through a porous membrane leading to a
ne emulsion having smaller and uniform droplets, at the expense of relatively low energy input. The
mean emulsion droplet size can precisely be tuned by adjusting the pore size, transmembrane pressure
and the number of cycles. The process can be used for a range of applications, including shear sensitive
products such as double emulsions. The present manuscript provides an overview covering the state
of the art, including insights in break-up mechanisms and the preparation of various products, and an
outlook on further improvement of the process.
2010 Elsevier B.V. All rights reserved.
Contents
1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2. Emulsion characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
3. Break-up mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
3.1. Localized shear forces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
3.2. Interfacial tension effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
3.3. Steric hindrance between droplets. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
4. Process parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
4.1. Membrane properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
4.2. Transmembrane pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
4.3. Disperse phase fraction and stabilization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
4.4. Continuous phase viscosity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
4.5. Number of homogenization cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
5. Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
5.1. Single emulsions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
5.2. Multiple emulsions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
5.3. Gel microbeads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
5.4. Polymer microspheres . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
6. Outlook. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Notation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Appendix A. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

Corresponding author. Tel.: +31 317 482240; fax: +31 317 482237.
E-mail addresses: akmal.nazir@wur.nl, akmal.nazir@yahoo.com (A. Nazir).
1
Tel.: +31 317 482231; fax: +31 317 482237.
0376-7388/$ see front matter 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.memsci.2010.06.044
2 A. Nazir et al. / Journal of Membrane Science 362 (2010) 111
1. Introduction
An emulsion is a mixture of two liquids that cannot mix e.g.
oil and water. Emulsions have applications in many industries
like food, pharmaceutical, cosmetic, agriculture, petrochemical and
other chemical industries [1]. An emulsion may be single e.g. oil-
in-water (O/W) and water-in-oil (W/O) emulsions, or it may be
an emulsion of an emulsion e.g. water-in-oil-in-water (W/O/W)
and oil-in-water-in-oil (O/W/O) emulsions, also termed multiple
or double emulsions.
Many different methods for emulsication have been devel-
oped, mostly depending upon the product (and economical)
requirements. Conventionally, the emulsions are prepared by
mechanical disruption of the droplets of the dispersed phase into
the continuous phase. Colloid mills, rotor stator systems, high-
pressure homogenizers, and ultrasonic homogenizers are popular
types of equipments for this, due to their high throughput [2].
These systems, having high energy requirements, apply shear and
extensional stresses to the product that may cause loss of func-
tional properties of heat and shear sensitive components [3], in
addition, they showpoor control over droplet size and distribution
[4,5].
To overcome these problems, new methods for emulsication
using microstructured systems like, cross-ow membrane emulsi-
cation [6] and premix membrane emulsication [7] have received
muchattention. For simplicityreasons we will call these techniques
cross-ow and premix emulsication from now on. Besides these
two shear-basedmethods, also spontaneous emulsicationdevices
have been reported, such as microchannel emulsication [810]
and Edge-based Droplet GEneration (EDGE) emulsication [11],
however, these technologies are still in the development phase,
and will not be discussed further in this review.
In cross-ow (or direct) emulsication, the emulsion is formed
by pushing the to-be-dispersed phase through a membrane
into the cross-owing continuous phase. Ideally, droplet size
can be controlled primarily by the choice of the membrane,
the cross-ow velocity, and the transmembrane pressure; typ-
ically, a factor of 25 is found between pore size and droplet
size. Cross-ow emulsication has advantages such as low shear
stresses, low energy requirement, uniform droplet size, which
allow use of less surfactant, and ease of design and scale-out
[12]. The most commonly used membranes for oil-in-water emul-
sions are hydrophilic Shirasu porous glass (SPG) membranes
[6], ceramic aluminium oxide (-Al
2
O
3
) membranes [13], -
alumina- and zirconia coated membranes [14], macroporous silica
glass membranes [15], and micro-fabricated metal membranes
[16,17]. Further, work has been done on silicon and silicon
nitride microsieves [1821]. For water-in-oil emulsions, polyte-
trauoroethylene (PTFE) membranes [2225], hydrophobized SPG
membranes [2225], hydrophobizedmicro-fabricatedmetal mem-
branes [16] and hydrophobized silicon nitride microsieves [26]
have been used. A limitation in case of cross-ow emulsication
is the low dispersed phase ux through the conventional mem-
branes (like SPG or ceramic membranes which have a relatively
high resistance), and therefore recirculation is often required to
increase the amount of disperse phase. In that case, interactions
of forming droplets with droplets in the emulsion, lead to a con-
siderable polydispersity as was visualized by Abrahamse et al. for
microsieves [18]. Further, the required membrane area is rather
large, and this makes this technology too expensive for large-scale
application. For diluted specialty products that need to meet high
qualitystandards, cross-owemulsicationis however aninterest-
ing technique to consider. Some of the cross-ow emulsication
studies using SPG membranes have been reviewed in Table A.1
in Appendix A, showing emulsion characteristics under different
operating conditions.
Contrary to cross-ow emulsication, premix emulsication
can be used to produce emulsions with high dispersed phase frac-
tion, albeit that the size of the droplets is not as monodisperse as
for cross-owemulsication. In its appearance, premix emulsica-
tion is a modied form of the classic emulsication systems, such
as high-pressure homogenization. These also start with a coarse
premix that is rened upon passage through the emulsication
machine, andinpremixemulsication, as introducedbySuzuki and
coworkers [23], the premix emulsion is passed through a microp-
orous membrane. In most cases, a membrane is used that is wetted
by the continuous phase of the premix and the emulsion is broken
up into smaller droplets. Sometimes the membrane is wetted by
the disperse phase, and in that case phase inversion can take place,
leading to very high disperse phase volume fractions (see Fig. 1).
It has to be mentioned that phase inversion was found to be only
possible for a limited number of products.
The energy costs for premix emulsication are relatively low,
since no cross-ow is needed. The energy needed can be one
order of magnitude lower than for cross-ow emulsication [27]
for highly concentrated products. However in general the desired
emulsion cannot be produced in a single passage. Further homog-
enization by repetitive cycles, commonly termed as repeated or
multi-stage premix emulsication, yields better control of droplet
size and distribution [3], however, at a corresponding increase of
the overall energy input. The drawback of premix emulsication
is membrane fouling that may become serious depending on the
formulation components [28], and related to that their interaction
with the membrane and their ease of removal.
When comparing various emulsication methods, the energy
density, usually dened as energy input per unit volume of
emulsion, is a useful parameter that enables comparison of emul-
sication efciencies. In Fig. 2, the energy efciencies of premix
emulsication [28] with metallic Stork Veco sieves (rectangular
pores of 10m405m, porosity 4%) are compared to homog-
enizers [4], microuidizers [4], and cross-ow emulsication with
a ceramic membrane (pore size 0.2 and0.8m) [4]. We cansee that
for cross-ow or premix emulsication less energy is required to
produce small droplets compared to the classic methods. In cross-
ow emulsication, the energy is applied more efciently than in
premix emulsication, but it should be noted that the pore size of
the metal sieves is far fromoptimal for productionof small droplets
through premix emulsication, and it is expected that considerable
improvement is possible here.
Besides the energy efciency, also the required membrane
area is of great importance for membrane emulsication tech-
niques. Fig. 3 shows a comparison between premix and cross-ow
emulsication for the required membrane area as a function of
transmembrane pressure, given a production of 20m
3
h
1
of a 30%
fat/oil product. Please keepinmindthat these are calculatedvalues,
based on the numbers shown in Table 1. In case of cross-owemul-
sication, microsieves, although having a low porosity, show the
lowest membrane area required [20]. This is due to the extremely
low resistance of these sieves compared to other membranes. The
low porosity of the microsieves was chosen to prevent any hin-
dranceamongthegrowingdroplets that canresult inapolydisperse
emulsion. The experiments for cross-ow and premix emulsica-
tion are not from the same study. There are essential differences
mainly in the membrane pore size, which is in one case consider-
ably higher and in one case comparable to that of the membranes
usedfor cross-owemulsication. Althoughthis restricts the valid-
ity, we would like to stress the importance of trends that are
observed in Fig. 3. In both premix studies, the required area is much
more strongly related to the applied pressure than for cross-ow
emulsication, and this could result in lower required areas when
pressures are used that are comparable to those applied in cross-
owemulsication. For commercial production, of course, the cost
A. Nazir et al. / Journal of Membrane Science 362 (2010) 111 3
Fig. 1. Schematic representation of batch premix emulsication systems: (a) simple emulsion without phase inversion; (b) simple emulsion with phase inversion; (c) double
emulsion.
Fig. 2. Energy efciencies of various emulsifying processes: cross-ow emulsica-
tion [4]: () 1%, () 5%, () 10%, () 20% and () 50%; () premix emulsication (5%)
[28]; high-pressure homogenization [4], () orice valve, () at valve homogenizer
and () Microuidizer (all 30%).
and life span of the membranes and the cost of modules and addi-
tional equipment should be considered together with the required
area, but these details are not readily available, and therefore we
nd the area an indicative starting point for any comparison.
As is clear from Figs. 2 and 3, membrane emulsication holds
a number of advantages over conventional emulsication technol-
ogy. The simplicity of premix emulsication makes it an interesting
option for large-scale production of emulsions, although many
aspects are still not (well) understood. Since its introduction, sev-
eral investigations have been carried out concerning principles,
Fig. 3. Membrane area required to produce a 30% fat/oil product at 20m
3
h
1
as
a function of transmembrane pressure: a comparison of cross-ow and premix
emulsication (see Table 1 for further details used in the calculations).
process parameters, and application of premix membrane emul-
sication. The aim of the present article is to provide a review that
covers all these explorations, and which provides an outlook on
future prospects.
2. Emulsion characterization
As mentioned, premixemulsicationbegins witha coarse emul-
sion, which is then extruded/homogenized through a membrane
under pressure to obtain a ne emulsion. The resulting emulsion is
mostly characterized by the Sauter diameter (d
3,2
), and the droplet
Table 1
Data used for the calculations in Fig. 3.
Code in Fig. 3 Emulsication process Membrane type dm (m) S.D.
a
(m) O/W emulsion Ref.
SPG
Premix
Premix SPG 2.4 0.5
b
25% corn oil [7]
PES Premix Polyethersulfone 0.8 0.5
b
30% sunower oil [29]
SPG
CF
Cross-ow SPG 0.2 0 0.6 30% milk fat [20]
Al
1
Cross-ow -Al
2
O
3
0.2 0 0.35 30% milk fat [20]
Al
2
Cross-ow -Al
2
O
3
0.2 0.1 0.35 30% milk fat [20]
Al
3
Cross-ow -Al
2
O
3
0.2 0.25 0.35 30% milk fat [20]
M
1
Cross-ow Microsieve 0.2 0 0.01
c
30% milk fat [20]
M
2
Cross-ow Microsieve 0.2 0 0.01
c
30% milk fat [20]
a
Standard deviation of the log-normal pore size distribution.
b
Assumed.
c
=0.25(dm/d
dr
)
2
=810
3
110
2
assuming a square array of pores.
4 A. Nazir et al. / Journal of Membrane Science 362 (2010) 111
size distribution; the productivity is related to the transmembrane
ux.
The droplet size distributions are usually measured with laser
light diffraction. The Sauter diameter is dened as the diameter of a
spherical droplet having the same area per unit volume (S
v
) as that
of the total collection of the droplets in the emulsion:
d
3,2
=
6
S
v
=

ks

1
v
i
d
i

1
(1)
where v
i
is the volume fraction of droplets in the ith range of sizes
having mean diameter of d
i
and k
s
is the number of size ranges.
The size distribution data can be used to calculate coefcients of
variation (spans) to indicate the width of the size distribution. If
the coefcient of variation is less than 0.4, the particles can be
considered to be monodisperse [30].
The transmembrane ux (J) is dened as:
J =
Q
A
(2)
where Q is the volumetric ow rate, and A is the cross sectional
area of the membrane. The actual velocity in the pores, which is
related to local shear forces that are responsible for droplet break-
up [31], is a function of the ux and the porosity of the membrane.
The average wall shear stress (
w,p
) inside the membrane pores can
be dened as [27]:

w,p
=
8
c
J
d
m
(3)
where
c
is the continuous phase viscosity, J is the transmembrane
ux, and , and d
m
are the membrane tortuosity, porosity and
pore diameter, respectively.
3. Break-up mechanisms
In general, it is assumed that shear forces are responsible for
droplet break-up; however, it is far from clear how these forces
operate, and how they can be related to design of a process. One
may expect that more mechanisms operate simultaneously [27].
For example, vander Zwanet al. [32] microscopically visualizedthe
droplet break-up mechanism in O/W premix emulsication using
microuidic devices and found three factors responsible for break-
up.
3.1. Localized shear forces
Break-up due to the shear forces exerted on a droplet coming
close to the tip of a channel branching, or due to divergent ow in
both legs of a branching e.g. Y- or T-shaped branching. Link et al.
[33] also studied the droplet break-up in T junctions, albeit for W/O
emulsions, and found an expression for critical capillary number
(C
cr
) for breaking a drop in the T junction:
C
cr
=
o

2/3
o
1

2
(4)
where is a dimensionless constant, which is a function of the
viscosity contrast of the two uids andthe geometry of the channel.

o
is the droplet initial extensionbefore enteringintothe Tjunction,
dened as the ratio of droplet length to its circumference.
3.2. Interfacial tension effects
Break-up due to deformation inside a channel because of the
channel geometry, which is comparable to the mechanism of
microchannel emulsication. Whena droplet is squeezedthrougha
Fig. 4. Schematic representation of the dumbbell-shaped droplet in a 3D constric-
tion [32].
constrictioninthechannel, thedumbbell-shapeof thedroplet gives
rise toa difference inLaplace pressure betweenthe dispersedphase
inside the constraint (P
c
) and the dispersed phase before (P
d1
)
andafter (P
d2
) theconstriction[34]. Inathree-dimensional, cylin-
drical pore, van der Zwan et al. [32] estimated that the snap-off can
take place when:
P
c
> P
d1


R
c1

R
c2
>
2
R
1
P
c
> P
d2


R
c1

R
c2
>
2
R
2

(5)
where R
c1
and R
c2
are the constriction radii as shown in Fig. 4.
Further, R
1
andR
2
are the droplet radii before andafter the constric-
tion. If R
c2
R
c1
, snap-off is induced when 2R
c1
<R
1
and 2R
c2
<R
2
.
Although, shear forces may act simultaneously on the droplet, the
lower value of critical capillary number (around 310
3
) in this
case, indicates that the deformation of the droplet inside the con-
striction already destabilizes the droplets, along the lines of the
interfacial tension-induced snap-off mechanism.
In addition to above mentioned Laplace instabilities, Rayleigh
instabilities may operate in case of higher continuous phase ow
[35]. The droplets after having left the constriction remain elon-
gated, which then may lead to break-up into polydisperse droplets.
3.3. Steric hindrance between droplets
The dispersed phase droplets start accumulating before the
membrane and inside the channels. These accumulating droplets
can inuence each other and thus induce break-up. Break-up in
this case is strongly dependent on the interfacial properties: a sta-
ble emulsion will resist coalescence, and yield net steric break-up;
a less stable emulsion may well coalesce.
In cross-ow emulsication, the forces acting on the forming
droplet are mainly the interfacial tension force (that keeps the
droplet connected to the pore) and the shear force (due the con-
tinuous phase ow that tries to remove the droplet). However, in
case of big droplets (>10m), also the buoyant and inertia forces
need to be considered. The point at which the oil will start to have
a pressure gradient from pore to droplet is:
P
p
P
d


R
p

2
R
d
R
d
2R
p
(6)
where P
p
and P
d
are the Laplace pressure difference in the pore
and of emerging droplet having radius of R
p
and R
d
, respectively,
and is the interfacial tension. So, once the droplet radius is about
twice as large as the pore radius, there is a possibility of sponta-
A. Nazir et al. / Journal of Membrane Science 362 (2010) 111 5
neous snap-off. Nevertheless, at higher transmembrane pressure,
pores may generate a liquid jet instead of single spherical droplets.
And if the shear forces are strong enough, this effect can be used
to produce droplets from a liquid jet emerging from membrane
pore due to Raleigh instabilities, like premix emulsication. So,
in this way, certain similarities may exist between droplet break-
up in premix and cross-ow emulsication depending upon the
operating parameter.
In the next section, the most relevant process parameters are
discussed, together with some examples from various literature
sources.
4. Process parameters
Various parameters inuence the droplet size, such as the
membrane properties (pore size, pore size distribution, etc.), trans-
membrane pressure, disperse phase fraction and stabilization,
continuous phase viscosity, and number of homogenizing cycles,
and these are discussed here.
4.1. Membrane properties
SPG membranes are the most extensively studied membranes
for premix emulsication (see Table A.2), which were reported to
have various advantages like (i) interconnected micropores, (ii) a
wide spectrumof available meanpore sizes (0.0530m) withnar-
rowsize distribution, (iii) a high porosity (5060%) and besides, (iv)
the surface can be hydrophobized by reaction with organic silanes
[3]. However, the effect of these properties is not simple. For exam-
ple, the porosity of the membranes as such may be high, but the
percentage of active pores is often very low; usually below 10% as
demonstrated by Vladisavljevic et al. [36] for cross-ow emulsi-
cation.
The membrane pore size correlates with the droplet size and
the size distribution (and the ux of the emulsion). Zhou et al. [37]
studied the size and uniformity of agarose beads prepared by pre-
mix emulsication using SPG and polyethylene (PE) membranes,
and they found a linear relationship between the number average
diameter of agarose beads and membranes pores size. Besides, it
was notedthat the pore size distributionandthe shape of the open-
ings of the pores do not affect the emulsication results within a
wide range. Probably, the largest pores carry most of the liquid, and
the droplet formation inside these pores decides the droplet size
of the resulting emulsion. This is different from results for cross-
ow emulsication, where clear differences are found: in this case
the droplet formation takes place on the surface of the membrane,
rather thaninside the membrane, as is the case for premix emulsi-
cation. With premix membrane emulsication, thicker membranes
gave more uniformemulsions, again pointing to multiple break-up
inside the membrane. Most important for premix emulsication is
that the contact angle betweenthe continuous phase andthe mem-
brane surface must be lowenough for complete wetting, in order to
obtain uniformly sized particles. Membranes that are incompletely
wetted by the continuous phase often give large polydispersity and
larger average droplet sizes. As mentioned previously, a membrane
that is wetted by the disperse phase may result in phase inver-
sion. Depending on the formulation, this inversion either leads to
demulsication; or, inalimitednumber of cases, inphaseinversion.
Vladisavljevic et al. [31] prepared W/O/Wemulsions by extrud-
ing a coarse W/O/W emulsion through SPG membranes, and found
that the mean outer droplet size increased with increasing pore
size, as was the case for the gel beads mentioned in the previ-
ous section. The ratio between droplet and pore size decreased
with increasing the pore size and the number of passes and was
1.250.68 after ve passes. While for cross-ow emulsication it
was 3.46, which is a relatively high and also independent on the
pore size.
4.2. Transmembrane pressure
The premix emulsication process involves using a transmem-
branepressuretopushthecoarseemulsionthroughthemembrane.
Increasing the transmembrane pressure increases the permeating
ux [7,23,38], according to the Darcys law, if the emulsioncanow
through the membrane as if it were only the continuous phase:
J =
P
tm
R
m

c
(7)
where J is the ux, P
tm
is the transmembrane pressure, R
m
is
the membrane resistance, and
c
is the continuous phase viscosity,
which will be discussed later. In (repeated) premix emulsication,
the transmembrane pressure is utilized to overcome ow resis-
tances inside the pores (P
ow
) and for droplet disruption (P
disr
)
i.e., to overcome interfacial tension forces [27]. This is summarized
in Eq. (8):
P
tm
=
pore
R
m
J
i
. .. .
P
flow
+C
ow

d
i

1
d
i1

. .. .
P
disr
(8)
where
pore
is the emulsion viscosity in the pores, R
m
is the mem-
brane resistance, J
i
is the transmembrane ux corresponding to ith
cycle, C is a constant, is the volume fraction of dispersed phase in
the emulsion,
ow
is the interfacial tension between oil and water,
and d
i
is the resulting mean droplet diameter corresponding to ith
cycle. If fouling occurs, an additional resistance could be added to
account for this.
The pressure that needs to be applied for premix emulsication
is co-determined by various factors. First, the continuous phase of
the premix should be able to intrude the pores, and for a non wet-
ting liquid, a pressure needs to be applied corresponding to the
Laplace pressure (assuming cylindrical pores):
p
c
=
4
ow
cos
d
m
(9)
where p
c
is the critical pressure,
ow
is the oilwater interfacial
tension, is the contact angle, and d
m
is the pore diameter. The
nal droplet diameter, d
2
, may be larger or smaller than the pore
diameter, d
m
, depending on the shear stresses inside the pores.
Further, the local transmembrane pressure acting across the
droplet needs to be higher than the Laplace pressure of the droplets
(p
L,drop
=4
ow
/d
dr
) in order to deform them, and as mentioned
needs to be higher than a critical pressure to allow intrusion in a
pore. If the initial droplet diameter, d
1
, is not much larger than the
membrane pore diameter, d
m
, in other words, the ratio d
1
/d
m
is
close to 1 (but larger), then for the critical pressure the following
equation is given by [39]:
p
c
=

ow
{2 +2a
6
/

(a
6
1) arccos(1/a
3
) 4a
2
}
r
m
(a +

a
2
1)
(10)
where a is the ratio d
1
/d
m
, r
m
is the pore radius and
ow
is the
oilwater interfacial tension. For larger values of a (d
1
d
m
), the
critical pressure becomes equal to the capillary pressure given in
Eq. (9).
4.3. Disperse phase fraction and stabilization
Another promising feature of premix emulsication is that at
given operating conditions, the mean droplet size is independent
on the dispersed phase content over a wide range (160vol.%);
6 A. Nazir et al. / Journal of Membrane Science 362 (2010) 111
although it should be noted that the transmembrane ux signif-
icantly decreases with increasing dispersed phase content due to
an increase in viscosity [27]. The droplets coming out of the mem-
brane pores may be readily stabilized by surfactants while passing
through the membrane, which ultimately leads to negligible coa-
lescence.
Moreover, the droplet size and uniformity of the coarse emul-
sion do not affect the emulsication results; the mean droplet size
is primarily dependent on the mean pore size and wall shear stress
inside the pores [40,41]. In case of e.g. high-pressure homogeniz-
ers, at constant operating conditions the droplet size is strongly
dependent on the dispersed phase percentage [42]. This is because
the surface area that is created during passage in such machines
cannot be covered in time by the surfactants, leading to instability
of the produced emulsion, and a need for repeated processing. This
can, to some extend, also be the case for premix emulsication, but
here the process allows more time for coverage of producedsurface
area.
4.4. Continuous phase viscosity
The inuence of the continuous phase viscosity on the premix
emulsication process is complex. Primarily, the permeate ux is
inversely proportional to the emulsion viscosity as indicated in Eq.
(7). The emulsion viscosity will be close to the viscosity of the con-
tinuous phase viscosity at lowdispersedphase volume fraction, but
can become considerably higher with higher disperse phase frac-
tions. Further, the continuous phase viscosity inuences the wall
shear stress as indicated in Eq. (3), which will be higher for viscous
liquids, resulting in smaller droplets [31].
4.5. Number of homogenization cycles
In repeated or multi-stage premix emulsication, in addition to
improving monodispersity, the permeate ux also increases with
increasing number of passes [27,41,43], most probably as a result
of the decreased viscosity related to droplet size reduction. Besides,
as mentioned earlier, if the droplet size is similar to the pore size, it
is expected to pass unhindered, and less pressure is needed. Under
non-foulingconditions, thelargest increaseinuxtakes placeinthe
second pass as the largest droplet size reduction occurs in the rst
pass. The increase in ux can be explained by a decrease in P
disr
and an increase in P
ow
(Eq. (8)) and ultimately P
tm
becoming
equal to P
ow
after few passes, usually 35 depending upon the
nature of the coarse emulsion.
Components in the premix emulsion may have negative side
effects when they foul the membrane. Surh et al. [44] studied the
preparation of lecithin-stabilized O/W emulsions by repeated pre-
mix emulsication using SPG membranes. They found that as the
number of passes through the same membrane increased from 1
to 5, the transmembrane ux decreased from 30 to 1m
3
m
2
h
1
because of membrane fouling due to lecithin.
5. Applications
5.1. Single emulsions
Single emulsions play an important role in the formulation of
various products such as foods; examples of O/W emulsions are
dressings, articial milks, creamliqueurs, etc., andexamples of W/O
emulsions are margarines, and low fat spreads. In addition, there
are numerous non-food emulsions like pharmaceutical products,
cosmetics, pesticides, bitumen (for road application), water-based
paints, photographic lms, paper coatings, lubricants, etc. The
method used for the preparation of emulsions has a great inuence
on the physicochemical properties of the nal product; the droplet
size and size distribution are among the most important properties
that have to be considered while preparing a certain type of emul-
sion. Moreover, application of high shear and extensional stresses
during the process may cause loss of functional properties of shear
and heat sensitive components. In regard of droplet size, droplet
size distribution, and low shear stress, premix emulsication is a
good candidate for the preparation of single emulsions, as obvious
fromliteratureontheproductionof simpleemulsions carryingfood
ingredients like corn oil, soybean oil, etc. [7,27,44].
5.2. Multiple emulsions
Recently, several premix emulsication studies have been car-
ried out for the production of multiple emulsions that have
potential applications for controlledrelease of a substance fromthe
inner phase. Vladisavljevic et al. [27] prepared W/O/W emulsions
withanarrowdroplet sizedistribution(span=0.28) at highproduc-
tion rates (transmembrane ux=1.837m
3
m
2
h
1
) by repeated
premix emulsication using SPG membranes. In another study,
Vladisavljevic et al. [31] found that the mean size of the outer drops
was unaffectedbythe volume fractionof inner droplets inthe range
of 0.30.5, and the encapsulation efciency of a hydrophilic marker
(CaNa
2
EDTA) was virtually independent of the number of passes.
Also Shima et al. [45] reported on repeated premix emulsica-
tion for the production of W/O/W emulsions prepared as a carrier
system for the daily uptake of a bioactive substance. They passed
the premix through a cellulose acetate membrane to produce a ne
emulsionwitha meanoil droplet diameter of <1mwithanencap-
sulation efciency of >90%. During preparation of the premix in a
rotor stator system, inclusionof the outer water phase solutioninto
the oil phase was observed; however, the includedwater phase dis-
appeared during membrane emulsication, most probably because
it wetted the membrane well, and was captured in this way. Unlike
the internal phase, external water phase is not stabilized with a
surfactant suitable for stabilization a W/Oemulsion, and therefore,
when captured it will re-coalesce with the external phase inside
the membrane.
Surh et al. [46] studied the preparation of W/O and
W/O/W emulsions containing gelled internal water droplets. They
compared emulsication methods and observed that with a
high-pressure valve homogenizer smaller droplets were obtained
compared to premix emulsication, but the membrane produced a
narrower droplet size distribution, at high encapsulation efciency
of the internal phase (>95%).
Kukizaki [47] prepared hydrophilic drug-encapsulating solid
lipid microcapsules (SLMCs) for drug delivery with a narrow par-
ticle size distribution via solid-in-oil-in-water (S/O/W) dispersions
by premix emulsication using SPG membrane with a mean pore
diameter of 14.8m. Subsequent solidication of the oil phase in
the S/O/Wdispersion resulted in SLMCs with a mean particle diam-
eter of 15.4m and a high encapsulation efciency (up to 93.5%).
5.3. Gel microbeads
Monodisperse beads of e.g. agarose are important for a vari-
ety of chromatographic applications such as gel ltration [48],
ion-exchange chromatography [49], hydrophobic interaction chro-
matography [50] and afnity chromatography [51]. Investigations
have beencarriedout toproduce agarose beads using premixemul-
sication. Conventional methods like suspension gelation [52],
or spraying gelation [53], are not efcient enough for producing
uniform-sized beads. Zhou et al. [54] were the rst who reported
the production of uniform agarose beads by premix emulsication
andpreparedbeads withdiameters ranging from15to60musing
membranes with different pore size. Later on, they reported the
production of uniform-sized agarose beads with smaller diameter
A. Nazir et al. / Journal of Membrane Science 362 (2010) 111 7
(less than10m) andhighagarose content (more than14%); which
was not possible by regular cross-ow emulsication [37].
5.4. Polymer microspheres
For drug-loaded microspheres, the solvent evaporation method
involving high-speed homogenization, mechanical stirring, or
ultrasonication, have been studied extensively [5559]. Various
biodegradable polymers were considered (such as polylactide,
poly(glycolide), poly(-caprolactone), poly(saccharides), or albu-
min); however, the size of the particles prepared by these methods
is difcult to control and also the size distribution is very
broad. Sawalha et al. [60] prepared narrowly dispersed polylac-
tide (PLA) hollow microcapsules with sizes 0.355m by premix
emulsication and found that particles of dened size and size
distribution can be produced. Wei et al. [61] prepared uniform-
sized poly(lactide-co-ethylene glycol) (PELA) microspheres with
high encapsulation efciency of antigen by premix emulsication.
Under optimum conditions, they obtained a particle size of about
1m and reported that the polymer properties and solidication
rate are two effective strategies to yield high encapsulation. Even
a few studies target biodegradable nanoparticles; Wei et al. [62]
prepared uniform-sized PLA nanoparticles by combining premix
emulsication and solvent removal, starting from larger droplets
that are subsequently reduced in size due to solvent removal. They
obtained meansizes of about 321669nmdepending uponthe vol-
ume ratio of the phases in the emulsion, and mentioned that this
method has high productivity, simplicity, and is suitable for easy
scale-up.
In a recent investigation, Kooiman et al. [63] studied the
synthesis and characterization of novel polymeric microcapsules
for ultrasound-triggered delivery of lipophilic drugs. Microcap-
sules (having mean number-weighted diameter in the range of
1.221.31m) with a shell of uorinated end-capped poly(l-lactic
acid) were prepared through premix emulsication and contained,
apart froma gaseous phase, different amounts of hexadecane oil as
a drugcarrier reservoir. The partially oil-lled microcapsules with
high drug loads and well-dened acoustic activation thresholds
were reported to have a great potential for ultrasound-triggered
local delivery of lipophilic drugs under ultrasoundimage-guidance.
6. Outlook
Although the knowledge base for premix emulsication does
not seem to be as wide as for other emulsication methods; var-
ious effects that occur in e.g. cross-ow emulsication are also of
relevance for premix emulsication, and can be used to design pro-
cesses. It should be mentioned, that the interaction of droplets
and its inuence on the actual size of the obtained droplets is
still uncharted territory. It is obvious from the available data,
that it is an interesting technique for the controlled production
of small sized emulsions, and all kinds of related products. Var-
ious aspects, such as the narrow droplet size distribution, high
productivity, and robustness, make premix emulsication not only
suited for shear sensitive emulsions and related products, but also
for emulsions in general as long as the membrane is not fouled
during operation. If this is the case, and alternative system could
be used consisting of a packed layer of glass beads instead of a
membrane, as introduced by van der Zwan et al. [28]. Such type
of dynamic membrane, having morphology very similar to the
conventional premix emulsication membranes, has the advan-
tage that the system can be easily cleaned after emulsication,
and therefore, such a system could be interesting for emulsions
having ingredients that cause (depth) fouling of conventional
membranes.
Notation
A cross sectional area of the membrane (m
2
)
a ratio of emulsion size to membrane pore
C constant in Eq. (8)
C
cr
critical capillary number
d
1
mean initial droplet diameter (m)
d
1,0
mean number-weighted droplet diameter (m)
d
2
mean nal droplet diameter (m)
d
3,2
mean Sauter diameter (m)
d
dr
mean droplet diameter (m)
d
i
meandroplet diameter intheithcycle/rangeof sizes (m)
d
m
membrane pore diameter (m)
J transmembrane ux (ms
1
)
J
i
transmembrane ux corresponding to ith cycle (ms
1
)
k
s
number of size ranges
p
c
critical pressure (Pa)
Q volumetric ow rate (m
3
s
1
)
R
1
droplet radius before constriction (m)
R
2
droplet radius after constriction (m)
R
c1
/R
c2
constriction radii (m)
R
d
radius of emerging droplet (m)
r
m
membrane pore radius (m)
R
m
membrane resistance (m
1
)
R
p
membrane pore radius (m)
S
v
area per unit volume (m
1
)
v
i
volume fraction of droplets in the ith cycle/range of sizes
(m
3
)
Greek letters
P
c
Laplace pressure of dispersed phase inside the constraint
(Pa)
P
d
Laplace pressure of emerging droplet (Pa)
P
d1
/P
d2
Laplace pressure of dispersed phase before and after
the constriction (Pa)
P
p
Laplace pressure of dispersed phase inside the pore (Pa)
P
tm
transmembrane pressure (Pa)
constant in Eq. (4)

ow
oilwater interfacial tension (Nm
1
)
membrane porosity

o
droplet initial extension (ratio of droplet length to its cir-
cumference)

c
continuous phase viscosity (Pa s)

pore
emulsion viscosity in the pore (Pa s)
contact angle (

)
mean tortuosity factor of pores (the ratio of mean pore
length and membrane thickness)
interfacial tension (Nm
1
)

w,p
shear stress inside pore wall (Pa)
volume fraction of dispersed phase in the emulsion
Appendix A.
Tables A.1 and A.2.
8 A. Nazir et al. / Journal of Membrane Science 362 (2010) 111
Table A.1
Cross-ow emulsication studies using SPG membranes.
Membrane characteristics,
system design
Emulsion characteristics
(1: continuous phase; 2:
dispersed phase)
Droplet characteristics Pressure, P (kPa)
ux, J (m
3
m
2
h
1
)
Ref.
Flat SPG membrane disks
(3.01 and 9.83m)
O/W, (1) water +Tween 20
or SDS +polyethylene
glycol, (2) decane +liquid
parafn
With 9.83m membrane
using 2% SDS:
d
dr
=29.98m,
d
dr
/dm =3.05, span=0.53
J =1410
5
P=3.517.4
[64]
Tubular SPG membrane
(4.8m)
O/W, (1) commercial
skimmed milk, (2)
soy/rapeseed oil
Using soy oil: d
dr
=12.4m J =0.05
P=90
[65]
Tubular SPG membranes
0.2 and 0.4m
hydrophilic 0.4 and 1m
hydrophobic
O/W (solid lipid particles),
(1) water +Tween 20 or
Pluronic F68, (2) Gelucire
or Compritol
d
dr
=50750nm J =0.0080.84
P=400 or 600
[66]
Tubular SPG membrane
(7.0m)
O/W, (1) water +Tween 60
or Tween 20, (2) sunower
oil (20%)
d
dr
=around 32m at 30

C P=3.34.8 [67]
Asymmetric tubular SPG
membrane, consisting of
an inner skin layer
(0.67m) and a support
layer (4.7m)
O/W, (1) water +SDS (0.3%,
w/v), (2) Soybean oil
d
dr
=2.182.22m J =0.0110.039
P=35120
[68]
Hydrophobic modied SPG
membranes (1.8, 2.0, 2.5,
4.8 and 11.1m)
W/O, (1) kerosene oil
(0.15.0wt.%) +PGPR 90,
(2) water +NaCl
(0.0170.855mol/l)
d
dr
/dm =3.110.13,
span=around 0.28
P=0.516.5 [30]
Tubular SPG membrane
(0.46.6m)
O/W, (1) demineralized
water +Tween 80 (2%,
w/w), (2) rapeseed oil
d
dr
/dm =3.5,
span=0.260.45
With 4.8m membrane:
J =0.08 at P=40
[40]
SPG membrane (15m) O/W, (1) water +SDS
(0.3wt.%), (2) soybean oil
d
dr
=30m, CV=1125% J =0.5810
6
5.810
6
[69]
Cylindrical SPG membrane
(1, 2.94m)
O/W, (1) water +polyvinyl
alcohol +sodium lauryl
sulfate, (2) styrene +divinyl
benzene +hexadecane
d
dr
/dm =6.6 J =3.2410
6
2.5210
5
m
3
h
1
P=12.868.7
[7072]
Table A.2
Premix emulsication studies.
Membrane
characteristics,
system design
Emulsion characteristics
(1: continuous phase; 2:
dispersed phase)
Droplet
characteristics
Pressure, P (kPa)
ux, J (m
3
m
2
h
1
)
Ref.
Tubular SPG (2.7
and 4.2m),
cross-ow
O/W, (1) water, (2) corn oil,
PGPR and PGFE as
emulsiers for oil and
water phase, respectively
d
dr
/dm =1.42.1,
span=0.40.62
P=10100
J =0.033.5
[7]
Flat PTFE (1.0m),
dead end
O/W and W/O d
dr
/dm =24.1 J =up to 9 [23]
Flat PTFE (1.0m),
dead end, phase
inversion
O/W and W/O, (1) water,
(2) corn oil, PGPR and PGFE
as emulsiers for oil and
water phase, respectively
d
dr
/dm =2.84.0 P=100800
J =15
[24]
Flat polycarbonate
(0.33, 0.38, 0.47,
0.6 and 1.0m),
dead end,
multi-stage
(n=118)
O/W, (1) water +SDS
(0.2wt.%), (2) kerosene oil
d
dr
/dm 1.6 for
n>12
P=100
J =0.20.6
[39]
Flat PTFE (1.0m),
dead end,
multi-stage
(n=13)
O/W d
dr
/dm =1.22.6,
span=0.550.9
J =218 [73]
Tubular SPG
(1.1m), dead
end, multi-stage
(n=3)
S/O/W, (1)
water +surfactant l-1695
(1wt.%) +sodium cholate
(1wt.%) +d-glucose
(1wt.%), (2)
surfactant-coated insulin
dispersed in soybean oil to
form S/O
d
dr
/dm =1.0 J =1.6 [74]
Flat cellulose
acetate (0.2, 0.45,
0.8 and 3.0m),
Dead end
W/O/W, (1) Hanks
solution, (2) 10
4
mol/l
PTSA sol. +C8TG containing
hexaglyceryl condensed
ricinoleate (110%, w/v) to
form W/O emulsion
d
dr
/dm =1.03.5 P=300440 [45]
A. Nazir et al. / Journal of Membrane Science 362 (2010) 111 9
Table A.2 (Continued)
Membrane
characteristics,
system design
Emulsion characteristics
(1: continuous phase; 2:
dispersed phase)
Droplet
characteristics
Pressure, P (kPa)
ux, J (m
3
m
2
h
1
)
Ref.
Tubular SPG
(10.7m), dead
end, multi-stage
(n=15)
W/O/W, (1) water +Tween
80 (0.5wt.%) +d-glucose
(5wt.%) +sod. alginate
(1wt.%), (2) 5wt.%
d-glucose aqueous sol.
dispersed in soybean oil
having 5wt.% PGPR
d
dr
/dm =0.411.2,
span=0.280.6
P=20300
J =1.837
[27]
Tubular -alumina
(1.5m), stirring
O/W, (1) water +SDS
(2wt.%), (2) toluene
d
dr
/dm =1.51.8,
span=11.2
P=200
J =0.420.62
[75]
Tubular SPG
(5.420.3m),
dead end,
multi-stage
(n=15)
W/O/W, (1) water +Tween
80 (0.5wt.%) +glucose
(5wt.%) +sod. alginate
(1wt.%), (2) aqueous sol
having glucose (5wt.%) and
CaNa2
EDTA (5wt.%)
dispersed in Soybean oil
containing 5wt.% PGPR
d
dr
/dm =0.371.2,
span=0.280.93
P=70150
J =2240
[31]
Tubular SPG
(8.0m), dead
end, multi-stage
(n=15)
O/W, (1) water +emulsier
(a combination of both SDS
and Tween 20), (2) corn oil
d
dr
/dm =0.51.4,
span=0.330.77
at n=7
P=100
J =360
[76]
Flat polycarbonate,
dead end,
multi-stage
(n=5)
W/O/W, (1) water +SDS
(1cmc) +NaCl (0.1M), (2)
aqueous sol. having NaCl
(0.1M) and dextrain
(910
5
M) dispersed in
dodecane having Arlacel
P135 as surfactant
d
dr
=0.72.5m J =3.714.7 [77]
Tubular SPG
(8.0m), dead
end, multi-stage
(n=5)
W/O/W, (1) water +Tween
20 (0.5wt.%) +phosphate
buffer, pH 7 (5mM) +NaCl
(100mM) +NaN3
(0.02wt.%), (2) water with
or without WPI dispersed
in corn oil having 8wt.%
PGPR
d
dr
/dm =0.200.29 P=100
J =70 at n=5
[46]
Glass lter
(1.0m), dead
end, multi-stage
(n=11)
Polymer (PLA)
microspheres
d
dr
=1.0m,
span=0.7
[60]
Glass lter
(1.0m), dead
end, multi-stage
(n=115)
Polymer (PLA)
microspheres
d
dr
=0.355.0m [78]
SPG (5m),
continuous
membrane
module run for
100min
O/W, (1)
water +SDS +phosphate
buffer, pH 7+PVA, (2)
Isooctane +racemic
naproxen methyl ester
d
dr
=1.32m P=120 [38]
SPG (8m), dead
end, multi-stage
(n=15)
O/W, (1) aqueous sol (pH
3) containing 100mM
acetic acid, 0.02wt.% NaN3
and 1.6 or 1.8wt.% lecithin,
(2) corn oil up to 10% or
20% of emulsion
d
3,2
=upto
5m
P=100150
J =301
[44]
SPG (5.2m), dead
end
Polymer (PELA)
microspheres
d
dr
=1.0m P=300 [61]
SPG (1.4m), dead
end
Polymer (PLA)
nanoparticles
d
dr
=321669nm P=1000 [62]
SPG (10.2m),
dead end,
multi-stage
(n=3)
W/O, agarose beads, (1)
liquid parafn/petroleum
ether (7:5,
v/v) +Hexaglycerin penta
ester (4wt.%), (2) 10wt.%
agarose sol. +0.9wt.% NaCl
d
dr
=10m P=98 [41]
Glass beads (beads
d
4,3
=75.9m
and
span=0.677),
dead end,
multi-stage
(n=6)
O/W, (1) water +Tween 20
(0.5%, v/w), (2)
n-hexadecane upto 5% of
emulsion
[28]
10 A. Nazir et al. / Journal of Membrane Science 362 (2010) 111
Table A.2 (Continued)
Membrane
characteristics,
system design
Emulsion characteristics
(1: continuous phase; 2:
dispersed phase)
Droplet
characteristics
Pressure, P (kPa)
ux, J (m
3
m
2
h
1
)
Ref.
SPG (10.2m) and
PE (11.8 and
25.6m), dead
end, multi-stage
(n=3)
W/O, agarose beads, (1)
liquid parafn/petroleum
ether (7:5,
v/v) +Hexaglycerin penta
ester (4wt.%), (2) 10wt.%
agarose sol.
d
dr
=3.069.02m P=98 [37]
Polycarbonate
(1.0m), nylon
(0.8m),
polyethersulfone
(0.8m) and
nitrocellulose
mixed ester
(0.8m), dead
end, multi-stage
(n=17)
O/W, (1) water +Tween 20
(2%, w/w) or BSA (12%,
w/w), (2) sunower oil
d3,2
=112m P=100900
J =upto 46 depending upon
membrane and pressure
applied
[29]
Tubular SPG
membrane (5.4,
7.6, 9.9 and
14.8m), dead
end
S/O/W, (1) water +Tween
40 (1.0wt.%), (2) Vitamin
B12
(0.21.1wt.%)
dispersed in glycerol
trimyristate containing
5wt.% PGPR
d
dr
(S/O) =15.5m
P=25200
J =11.8114.2
[47]
Glass lter
(1.0m), dead
end, multi-stage
(n=10)
Polymer (pLApFO)
microcapsules
d
1,0
=1.221.31m [63]
References
[1] M. Chappat, Some applications of emulsions, Colloids Surf. A 91 (1994)
5777.
[2] H. Karbstein, H. Schubert, Developments inthe continuous mechanical produc-
tion of oil-in-water macro-emulsions, Chem. Eng. Process. 34 (1995) 205211.
[3] C. Charcosset, I. Limayem, H. Fessi, The membrane emulsication processa
review, J. Chem. Technol. Biotechnol. 79 (2004) 209218.
[4] U. Lambrich, H. Schubert, Emulsication using microporous systems, J. Membr.
Sci. 257 (2005) 7684.
[5] S.M. Joscelyne, G. Trgrdh, Membrane emulsicationa literature review, J.
Membr. Sci. 169 (2000) 107117.
[6] T. Nakashima, M. Shimizu, M. Kukizaki, Membrane emulsication by microp-
orous glass, in: Inorganic Membranes, ICIM2-91, 1991, pp. 513516.
[7] K. Suzuki, I. Shuto, Y. Hagura, Characteristics of the membrane emulsication
method combined with preliminary emulsication for preparing corn oil-in-
water emulsions, Food Sci. Technol. Int., Tokyo 2 (1996) 4347.
[8] T. Kawakatsu, Y. Kikuchi, M. Nakajima, Regular-sized cell creation in
microchannel emulsication by visual microprocessing method, J. Am. Oil
Chem. Soc. 74 (1997) 317321.
[9] S. Sugiura, M. Nakajima, S. Iwamoto, M. Seki, Interfacial tension driven
monodispersed droplet formation from microfabricated channel array, Lang-
muir 17 (2001) 55625566.
[10] I. Kobayashi, M. Nakajima, K. Chun, Y. Kikuchi, H. Fukita, Silicon array of elon-
gated through-holes for monodisperse emulsion droplets, AIChE J. 48 (2002)
16391644.
[11] K.C.v. Dijke, G. Veldhuis, K. Schron, R.M. Boom, Simultaneous formation of
many droplets in a single microuidic droplet formation unit, AIChE J. 9999
(2009).
[12] S. van der Graaf, C.G.P.H. Schroen, R.G.M. van der Sman, R.M. Boom, Inu-
ence of dynamic interfacial tension on droplet formation during membrane
emulsication, J. Colloid Interface Sci. 277 (2004) 456463.
[13] V. Schroeder, H. Schubert, Inuence of Emulsier and Pore Size on Membrane
Emulsication, Special Publication, vol. 227, Royal Society of Chemistry, 1998,
pp. 7080.
[14] S.M. Joscelyne, G. Trgrdh, Food emulsions using membrane emulsication:
conditions for producing small droplets, J. Food Eng. 39 (1999) 5964.
[15] T. Fuchigami, M. Toki, K. Nakanishi, Membrane emulsication using solgel
derived macroporous silica glass, J. Sol-Gel Sci. Technol. 19 (2000) 337341.
[16] M.J. Geerken, M.N.W. Groenendijk, R.G.H. Lammertink, M. Wessling, Micro-
fabricated metal nozzle plates used for water-in-oil and oil-in-water
emulsication, J. Membr. Sci. 310 (2008) 374383.
[17] G.T. Vladisavljevic, R.A. Williams, Manufacture of large uniform droplets using
rotating membrane emulsication, J. Colloid Interface Sci. 299 (2006) 396402.
[18] A.J. Abrahamse, R. van Lierop, R.G.M. van der Sman, A. van der Padt, A. Boom,
Analysis of droplet formation and interactions during cross-ow membrane
emulsication, J. Membr. Sci. 204 (2002) 125137.
[19] M.J. Geerken, R.G.H. Lammertink, M. Wessling, Tailoring surface properties for
controlling droplet formation at microsieve membranes, Colloids Surf. A 292
(2007) 224235.
[20] A.J. Gijsbertsen-Abrahamse, A. van der Padt, R.M. Boom, Status of cross-ow
membrane emulsication and outlook for industrial application, J. Membr. Sci.
230 (2004) 149159.
[21] J. Zhu, D. Barrow, Analysis of droplet size during crossow membrane
emulsication using stationary and vibrating micromachined silicon nitride
membranes, J. Membr. Sci. 261 (2005) 136144.
[22] N. Yamazaki, H. Yuyama, M. Nagai, G.H. Ma, S. Omi, A comparison of mem-
brane emulsication obtained using SPG (Shirasu Porous Glass) and PTFE
[poly(tetrauoroethylene)] membranes, J. Dispersion Sci. Technol. 23 (2002)
279292.
[23] K. Suzuki, I. Fujiki, Y. Hagura, Preparation of corn oil/water and water/corn oil
emulsions using PTFE membranes, Food Sci. Technol. Int. 4 (1998) 164167.
[24] K. Suzuki, K. Hayakawa, Y. Hagura, Preparation of high concentration O/W and
W/O emulsions by the membrane phase inversion emulsication using PTFE
membranes, Food Sci. Technol. Res. 5 (1999) 234238.
[25] N. Yamazaki, K. Naganuma, M. Nagai, G.-H. Ma, S. Omi, Preparation of W/O
(water-in-oil) emulsions using a PTFE (polytetrauoroethylene) membranea
new emulsication device, J. Dispersion Sci. Technol. 24 (2003) 249257.
[26] M.J. Geerken, R.G.H. Lammertink, M. Wessling, Interfacial aspects of water
drop formation at micro-engineered orices, J. Colloid Interface Sci. 312 (2007)
460469.
[27] G.T. Vladisavljevic, M. Shimizu, T. Nakashima, Preparation of monodisperse
multiple emulsions at high production rates by multi-stage premix membrane
emulsication, J. Membr. Sci. 244 (2004) 97106.
[28] E.A. van der Zwan, C.G.P.H. Schron, R.M. Boom, Premix membrane emulsica-
tion by using a packed layer of glass beads, AIChE J. 54 (2008) 21902197.
[29] A. Trentin, M. Ferrando, F. Lpez, C. Gell, Premix membrane O/W emulsica-
tion: effect of fouling when using BSA as emulsier, Desalination 245 (2009)
388395.
[30] C.J. Cheng, L.Y. Chu, R. Xie, Preparation of highly monodisperse W/O emulsions
with hydrophobically modied SPG membranes, J. Colloid Interface Sci. 300
(2006) 375382.
[31] G.T. Vladisavljevic, M. Shimizu, T. Nakashima, Production of multiple emul-
sions for drug delivery systems by repeated SPG membrane homogenization:
inuence of mean pore size, interfacial tension and continuous phase viscosity,
J. Membr. Sci. 284 (2006) 373383.
[32] E. van der Zwan, K. Schron, K. van Dijke, R. Boom, Visualization of droplet
break-up in pre-mix membrane emulsication using microuidic devices, Col-
loids Surf. A: Physicochem. Eng. Aspects 277 (2006) 223229.
[33] D.R. Link, S.L. Anna, D.A. Weitz, H.A. Stone, Geometrically mediated breakup of
drops in microuidic devices, Phys. Rev. Lett. 92 (2004).
[34] J.G. Roof, Snap-off of oil droplets in water-wet pores, Soc. Pet. Eng. J. 10 (1970)
85.
[35] T.G. Mason, J. Bibette, Shear rupturing of droplets in complex uids, Langmuir
13 (1997) 46004613.
[36] G.T. Vladisavljevic, I. Kobayashi, M. Nakajima, R.A. Williams, M. Shimizu, T.
Nakashima, Shirasu porous glass membrane emulsication: characterisation
of membrane structure by high-resolution X-ray microtomography and micro-
scopic observation of droplet formation in real time, J. Membr. Sci. 302 (2007)
243253.
A. Nazir et al. / Journal of Membrane Science 362 (2010) 111 11
[37] Q.Z. Zhou, G.H. Ma, Z.G. Su, Effect of membrane parameters on the size and
uniformity in preparing agarose beads by premix membrane emulsication, J.
Membr. Sci. 326 (2009) 694700.
[38] N. Li, K. Sakaki, Performance of an emulsion enzyme membrane reactor com-
bined with premix membrane emulsication for lipase-catalyzed resolution of
enantiomers, J. Membr. Sci. 314 (2008) 183192.
[39] S.H. Park, T. Yamaguchi, S. Nakao, Transport mechanismof deformable droplets
in microltration of emulsions, Chem. Eng. Sci. 56 (2001) 35393548.
[40] G.T. Vladisavljevic, U. Lambrich, M. Nakajima, H. Schubert, Production of O/W
emulsions using SPG membranes, ceramic [alpha]-aluminium oxide mem-
branes, microuidizer and a silicon microchannel platea comparative study,
Colloids Surf. A 232 (2004) 199207.
[41] Q.Z Zhou, L.Y. Wang, G.H. Ma, Z.G. Su, Multi-stage premix membrane emulsi-
cation for preparation of agarose microbeads with uniform size, J. Membr. Sci.
322 (2008) 98104.
[42] S. Brosel, H. Schubert, Investigations on the role of surfactants in mechanical
emulsication using a high-pressure homogenizer with an orice valve, Chem.
Eng. Process. 38 (1999) 533540.
[43] L.G.R. Henelyta, S. Ribeiro, BadolatoF Gabriela G., H. Schubert, Production of
O/W emulsions containing astaxanthin by repeated premix membrane emul-
sication, J. Food Sci. 70 (2005) E117E123.
[44] J. Surh, Y.G. Jeong, G.T. Vladisavljevi, On the preparation of lecithin-stabilized
oil-in-water emulsions by multi-stage premix membrane emulsication, J.
Food Eng. 89 (2008) 164170.
[45] M. Shima, Y. Kobayashi, T. Fujii, M. Tanaka, Y. Kimura, S. Adachi, R. Matsuno,
Preparation of ne W/O/W emulsion through membrane ltration of coarse
W/O/W emulsion and disappearance of the inclusion of outer phase solution,
Food Hydrocolloids 18 (2004) 6170.
[46] J. Surh, G.T. Vladisavljevic, S. Mun, D.J. McClements, Preparation and
characterization of water/oil and water/oil/water emulsions containing
biopolymer-gelled water droplets, J. Agric. Food Chem. 55 (2007) 175184.
[47] M. Kukizaki, Preparation of solid lipid microcapsules via solid-in-oil-in-water
dispersions by premix membrane emulsication, Chem. Eng. J. 151 (2009)
387396.
[48] H. Pertoft, A. Halln, Preparation of silica-agarose beads for gel chromatogra-
phy, J. Chromatogr. 128 (1976) 125131.
[49] B.L. Johansson, M. Belew, S. Eriksson, G. Glad, O. Lind, J.L. Maloisel, N. Norrman,
Preparation and characterization of prototypes for multi-modal separation
aimed for capture of positively charged biomolecules at high-salt conditions, J.
Chromatogr. A 1016 (2003) 3549.
[50] S. Hjerten, J.l. Liao, High-performance liquid chromatography of proteins
on compressed, non-porous agarose beads. I. Hydrophobic-interaction chro-
matography, J. Chromatogr. 457 (1988) 165174.
[51] S. Hjerten, J.P. Li, High-performance liquid chromatography of proteins on
deformed non-porous agarose beads. Fast boronate afnity chromatography
of haemoglobin at neutral pH, J. Chromatogr. 500 (1990) 543553.
[52] S. Bengtsson, L. Philipson, Chromatography of animal viruses on pearl-
condensed agar, BBA Specialised Section On Biophysical Subjects 79 (1964)
399406.
[53] K. Hosoya, H. Ohta, K. Yoshizako, K. Kimata, T. Ikegami, N. Tanaka, Preparationof
uniformlysizedpolymeric separationmediapotentiallysuitablefor small-scale
high-performance liquid chromatography and/or capillary electrochromatog-
raphy, J. Chromatogr. A 853 (1999) 1120.
[54] Q.Z. Zhou, L.Y. Wang, G.H. Ma, Z.G. Su, Preparation of uniform-sized agarose
beads by microporous membrane emulsication technique, J. Colloid Interface
Sci. 311 (2007) 118127.
[55] J.W. Winkelmann, M.D. Kenner, R. Dave, R.H. Chandwaney, S.B. Feinstein, Con-
trast echocardiography, Ultrasound Med. Biol. 20 (1994) 507515.
[56] A. Budhian, S.J. Siegel, K.I. Winey, Haloperidol-loaded PLGA nanoparticles:
systematic study of particle size and drug content, Int. J. Pharm. 336 (2007)
367375.
[57] B.S. Zolnik, D.J. Burgess, Evaluationof invivoinvitrorelease of dexamethasone
from PLGA microspheres, J. Controlled Release 127 (2008) 137145.
[58] T.W. Chung, S.S. Wang, W.J. Tsai, Accelerating thrombolysis with chitosan-
coated plasminogen activators encapsulated in poly-(lactide-co-glycolide)
(PLGA) nanoparticles, Biomaterials 29 (2008) 228237.
[59] R.A. Jain, The manufacturing techniques of various drug loaded biodegrad-
able poly(lactide-co-glycolide) (PLGA) devices, Biomaterials 21 (2000)
24752490.
[60] H. Sawalha, N. Purwanti, A. Rinzema, K. Schron, R. Boom, Polylactide micro-
spheres prepared by premix membrane emulsicationeffects of solvent
removal rate, J. Membr. Sci. 310 (2008) 484493.
[61] Q. Wei, W. Wei, R. Tian, L.Y. Wang, Z.G. Su, G.H. Ma, Preparationof uniform-sized
PELA microspheres with high encapsulation efciency of antigen by premix
membrane emulsication, J. Colloid Interface Sci. 323 (2008) 267273.
[62] Q. Wei, W. Wei, B. Lai, L.Y. Wang, Y.XWang, Z.G. Su, G.H. Ma, Uniform-sized PLA
nanoparticles: preparation by premix membrane emulsication, Int. J. Pharm.
359 (2008) 294297.
[63] K. Kooiman, M.R. Bhmer, M. Emmer, H.J. Vos, C. Chlon, W.T. Shi, C.S. Hall,
S.H.P.M. de Winter, K. Schron, M. Versluis, N. de Jong, A. van Wamel, Oil-lled
polymer microcapsules for ultrasound-mediated delivery of lipophilic drugs, J.
Controlled Release 133 (2009) 109118.
[64] M. Kukizaki, Shirasu porous glass (SPG) membrane emulsication in the
absence of shear ow at the membrane surface: inuence of surfactant type
and concentration, viscosities of dispersed and continuous phases, and trans-
membrane pressure, J. Membr. Sci. 327 (2009) 234243.
[65] G. Gutierrz, M. Rayner, P. Dejmek, Production of vegetable oil in milk emul-
sions using membrane emulsication, Desalination 245 (2009) 631638.
[66] C. Doria, C. Charcosset, A.A. Barresi, H. Fessi, Preparation of solid lipid particles
by membrane emulsicationinuence of process parameters, Colloids Surf.
A 338 (2009) 114118.
[67] Y. Zhang, G. Lian, S. Zhu, L. Wang, W. Wei, M. Guanghui, Investigation on the
uniformityandstabilityof sunower oil/water emulsions preparedbyaShirasu
porous glass membrane, Ind. Eng. Chem. Res. 47 (2008) 64126417.
[68] M. Kukizaki, M. Goto, Preparation and characterization of a new asymmetric
type of Shirasu porous glass (SPG) membrane used for membrane emulsica-
tion, J. Membr. Sci. 299 (2007) 190199.
[69] M. Yasuno, M. Nakajima, S. Iwamoto, T. Maruyama, S. Sugiura, I. Kobayashi, A.
Shono, K. Satoh, Visualization and characterization of SPG membrane emulsi-
cation, J. Membr. Sci. 210 (2002) 2937.
[70] H. Yuyama, T. Watanabe, G.H. Ma, M. Nagai, S. Omi, Preparation and analysis
of uniform emulsion droplets using SPG membrane emulsication technique,
Colloids Surf. A 168 (2000) 159174.
[71] H. Yuyama, T. Hashimoto, G.H. Ma, M. Nagai, S. Omi, Mechanismof suspension
polymerization of uniform monomer droplets prepared by class membrane
(Shirasu Porous Glass) emulsication technique, J. Appl. Polym. Sci. 78 (2000)
10251043.
[72] H. Yuyama, K. Yamamoto, K. Shirafuji, M. Nagai, G.H. Ma, S. Omi, Preparation
of polyurethaneurea (PUU) uniform spheres by SPG membrane emulsication
technique, J. Appl. Polym. Sci. 77 (2000) 22372245.
[73] J. Altenbach-Rehm, K. Suzuki, H. Schubert, Production of O/W-emulsions with
narrowdroplet size distributionby repeatedpremix membrane emulsication,
in: 3rd World Congress on Emulsions, Lyon, France, 2002.
[74] E. Toorisaka, H. Ono, K. Arimori, N. Kamiya, M. Goto, Hypoglycemic effect of
surfactant-coated insulin solubilized in a novel solid-in-oil-in-water (S/O/W)
emulsion, Int. J. Pharm. 252 (2003) 271274.
[75] W.H. Jing, J. Wu, W.H. Xing, W.Q. Jin, N.P. Xu, Emulsions prepared by two-stage
ceramic membrane jet-ow emulsication, AIChE J. 51 (2005) 13391345.
[76] G.T. Vladisavljevic, J. Surh, J.D. McClements, Effect of emulsier type on droplet
disruption in repeated Shirasu porous glass membrane homogenization, Lang-
muir 22 (2006) 45264533.
[77] W. Yafei, Z. Tao, H. Gang, Structural evolution of polymer-stabilized double
emulsions, Langmuir 22 (2006) 6773.
[78] H. Sawalha, Y. Fan, K. Schron, R. Boom, Preparation of hollow polylactide
microcapsules through premix membrane emulsicationeffects of nonsol-
vent properties, J. Membr. Sci. 325 (2008) 665671.