Professional Documents
Culture Documents
RAYN E R • DE JME K
Emulsions are found in a wide variety of food products, pharmaceuticals, paints,
and cosmetics, thus emulsification is a truly multidisciplinary phenomenon.
Therefore, understanding of the process must evolve from the combination of (at
Engineering Aspects
least) three different scientific specializations. Engineering Aspects of Food
Emulsification and Homogenization describes the state-of-the-art technology
E N G I N E E R I N G A S P E C T S O F F O O D E M U L S I F I C AT I O N A N D H O M O G E N I Z AT I O N
and brings together aspects from physical chemistry, fluid mechanics, and
chemical engineering. The book explores the unit operations used in emulsifica-
of
tion and homogenization processes, using fundamental theory from different fields
to discuss design and function of different emulsification techniques.
Features
EDITED BY
K16909
Marilyn R ayner
ISBN: 978-1-4665-8043-5
90000 P et r D ejm ek
9 781466 580435
Engineering Aspects
of
Food Emulsification
and
Homogenization
Contemporary Food Engineering
Series Editor
Professor Da-Wen Sun, Director
Food Refrigeration & Computerized Food Technology
National University of Ireland, Dublin
(University College Dublin)
Dublin, Ireland
http://www.ucd.ie/sun/
Marilyn Rayner
Petr Dejmek
CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742
© 2015 by Taylor & Francis Group, LLC
CRC Press is an imprint of Taylor & Francis Group, an Informa business
This book contains information obtained from authentic and highly regarded sources. Reasonable
efforts have been made to publish reliable data and information, but the author and publisher cannot
assume responsibility for the validity of all materials or the consequences of their use. The authors and
publishers have attempted to trace the copyright holders of all material reproduced in this publication
and apologize to copyright holders if permission to publish in this form has not been obtained. If any
copyright material has not been acknowledged please write and let us know so we may rectify in any
future reprint.
Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced,
transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or
hereafter invented, including photocopying, microfilming, and recording, or in any information stor-
age or retrieval system, without written permission from the publishers.
For permission to photocopy or use material electronically from this work, please access www.copy-
right.com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222
Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that pro-
vides licenses and registration for a variety of users. For organizations that have been granted a photo-
copy license by the CCC, a separate system of payment has been arranged.
Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are
used only for identification and explanation without intent to infringe.
Visit the Taylor & Francis Web site at
http://www.taylorandfrancis.com
and the CRC Press Web site at
http://www.crcpress.com
Dedication
ix
x Contents
xi
xii Series Preface
relevant topic in each chapter, with reference lists for further information. Therefore,
each book can serve as an essential reference source to students and researchers in
universities and research institutions.
Da-Wen Sun
Series Editor
Series Editor
Born in southern China, Dr. Da-Wen Sun is a world
authority in food engineering research and education; he
is a member of the Royal Irish Academy (RIA), which is
the highest academic honor in Ireland; he is also a mem-
ber of Academia Europaea (The Academy of Europe) and
a fellow of the International Academy of Food Science and
Technology. His main research activities include cooling,
drying, and refrigeration processes and systems, quality
and safety of food products, bioprocess simulation and optimization, and computer
vision technology. Especially, his many scholarly works have become standard ref-
erence materials for researchers in the areas of computer vision, computational fluid
dynamics modeling, vacuum cooling, and so on. Results of his work have been pub-
lished in over 800 papers, including more than 380 peer-reviewed journal papers (Web
of Science h-index = 60). He has also edited 14 authoritative books. According to
Thomson Reuters Essential Science IndicatorsSM, based on data derived over a period
of 10 years from the ISI Web of Science, there are about 4500 scientists who are among
the top 1% of the most cited scientists in the category of agriculture sciences. For many
years, Dr. Sun has consistently been ranked among the top 50 scientists in the world
(he is at the 25th position in March 2015).
He received a first class in both bachelor’s (honors) and master’s degree programs
in mechanical engineering, and a PhD in chemical engineering in China before
working in various universities in Europe. He became the first Chinese national
to be permanently employed in an Irish university when he was appointed college
lecturer at the National University of Ireland, Dublin [University College Dublin
(UCD)], in 1995, and was then continuously promoted in the shortest possible time
to senior lecturer, associate professor, and full professor. Dr. Sun is now the professor
of Food and Biosystems Engineering and the director of UCD Food Refrigeration
and Computerised Food Technology.
As a leading educator in food engineering, Dr. Sun has significantly contributed
to the field of food engineering. He has trained many PhD students who have made
their own contributions to the industry and academia. He has also delivered lec-
tures on advances in food engineering on a regular basis in academic institutions
internationally and delivered keynote speeches at international conferences. As a
recognized authority in food engineering, he has been conferred adjunct/visiting/
consulting professorships from 10 top universities in China, including Zhejiang
University, Shanghai Jiaotong University, Harbin Institute of Technology, China
Agricultural University, South China University of Technology, and Jiangnan
University. In recognition of his significant contribution to food engineering world-
wide and for his outstanding leadership in the field, the International Commission
of Agricultural and Biosystems Engineering (CIGR) awarded him the CIGR Merit
Award twice in 2000 and in 2006, the Institution of Mechanical Engineers based
xiii
xiv Series Editor
in the United Kingdom named him Food Engineer of the Year 2004. In 2008, he
was awarded the CIGR Recognition Award in honor of his distinguished achieve-
ments as the top 1% of agricultural engineering scientists in the world. In 2007, he
was presented with the only AFST(I) Fellow Award by the Association of Food
Scientists and Technologists (India), and in 2010, he was presented with the CIGR
Fellow Award; the title of Fellow is the highest honor in CIGR and is conferred
to individuals who have made sustained, outstanding contributions worldwide. In
March 2013, he was presented with the You Bring Charm to the World award by
Hong Kong–based Phoenix Satellite Television with other award recipients includ-
ing Mr. Mo Yan—the 2012 Nobel Laureate in Literature and the Chinese Astronaut
Team for Shenzhou IX Spaceship. In July 2013, he received the Frozen Food
Foundation Freezing Research Award from the International Association for Food
Protection for his significant contributions to enhancing the field of food freezing
technologies. This is the first time that this prestigious award was presented to a
scientist outside the United States.
He is a fellow of the Institution of Agricultural Engineers and a fellow of Engineers
Ireland (the Institution of Engineers of Ireland). He is also the editor-in-chief of
Food and Bioprocess Technology—An International Journal (2012 impact fac-
tor = 4.115), former editor of Journal of Food Engineering (Elsevier), and editorial
board member for a number of international journals, including the Journal of Food
Process Engineering, Journal of Food Measurement and Characterization, and
Polish Journal of Food and Nutritional Sciences. He is also a chartered engineer.
On May 28, 2010, he was awarded membership in the RIA, which is the high-
est honor that can be attained by scholars and scientists working in Ireland; at the
51st CIGR General Assembly held during the CIGR World Congress in Quebec
City, Canada, on June 13–17, 2010, he was elected incoming president of CIGR
and became CIGR president in 2013–2014; the term of his CIGR presidency is six
years, two years each for serving as incoming president, president, and past presi-
dent. On September 20, 2011, he was elected to Academia Europaea (The Academy
of Europe), which is functioning as the European Academy of Humanities, Letters
and Sciences, and is one of the most prestigious academies in the world; election to
Academia Europaea represents the highest academic distinction.
Preface
MOTIVATION IN SHORT FOR THIS BOOK
• Emulsions are used widely and produced in large volumes. Thus, emulsion
formation or emulsification is an important unit operation.
• Emulsification is of interest for a broad audience, both because of its influ-
ence on the functionality of emulsion-based products and because it is gen-
erally energy intensive running at low efficiency.
• Today, there is no comprehensive text on emulsification available in English
describing the state-of-the-art technology and bringing together aspects
from physical chemistry, formulation, fluid mechanics, and chemical engi-
neering. Together these aspects are the foundation needed for understanding
emulsification at more than a rudimentary level.
xv
xvi Preface
• Physical chemists have long studied interfaces between liquids and their
relation to surface-active molecules. This will be a key aspect in the process
of coalescence as well as to understanding how surface-active agents can
aid in disruption.
• Fluid mechanics has developed theoretical and experimental methods for
understanding the interplay between hydrodynamic forces and the drop
interface. This is vital for understanding how differences in design can
bring about different emulsification results.
• Research in chemical engineering has led to relations between operating
conditions and emulsion characteristics and devised methods for measur-
ing fragmentation or coalescence rates in bulk. This has obvious practical
importance, but can also be used for comparison with predictions from the
more fundamental disciplines.
• Emulsions are used widely and produced in large volumes. Thus, emulsion
formation or emulsification is an important unit operation.
• Emulsification is of interest for a broad audience, both because of its influ-
ence on the functionality of emulsion-based products and because it is gen-
erally energy intensive running at low efficiency.
• Today, there is no comprehensive text on emulsification available in English
describing the state-of-the-art technology and bringing together aspects
from physical chemistry, formulation, fluid mechanics, and chemical engi-
neering. Together these aspects are the foundation needed for understand-
ing emulsification at more than a rudimentary level.
REFERENCES
Becher, P. (1983). Encyclopedia of Emulsion Technology. Marcel Dekker Inc., New York.
Binks, B.P. (1998). Modern Aspects of Emulsion Science. Royal Society of Chemistry, Cambridge.
Clift, R., Grace, J.R., Weber, M.E. (1978). Bubbles, Drops and Particles. Academic Press,
New York.
Cosgrove, T. (2010). Colloid Science: Principles, Methods and Applications. John Wiley & Sons,
Chippenham, UK.
Dickinson, E., Stainsby, G. (1982). Colloids in Food. Applied Science Publishers, London.
xviii Preface
Friberg, S., Larsson, K. (1997). Food Emulsions. Marcel Dekker Inc., New York.
Goodwin, J.W. (2004). Colloids and Interfaces with Surfactants and Polymers. John Wiley & Sons,
TJ International, Padstow, Cornwall.
Gopal, E.S.R. (1968). Principles of emulsion formation. In: P. Sherman (ed.), Emulsion Science.
Academic Press, London, pp. 2–75.
Holmberg, K., Jönsson, B., Kronberg, B., Lindman, B. (2003). Surfactants and Polymers in
Aqueous Solution. John Wiley & Sons, Guildford, UK.
McClements, J.D. (2005). Food Emulsions. CRC Press, Boca Raton, FL.
Sjöblom, J. (2006). Emulsions and Emulsion Stability. Taylor & Francis, Boca Raton, FL.
Tetra Pak Dairy Index Report, Issue 4, July 2011, “Emerging Middle Class” Tetra Pak AB,
Lund, Sweden.
Walstra, P. (1983). Formation of emulsions. In: P. Beacher (ed.), Encyclopedia of Emulsion
Technology, Volume I: Basic Theory. Marcel Dekker Inc., New York, pp. 57–127.
Walstra, P. (1993). Principles of emulsion formation. Chemical Engineering Science 48,
333–349.
Walstra, P., Smulders, P.E.A. (1998). Emulsion formation. In: B.P. Binks (ed.), Modern Aspects
of Emulsion Science. Royal Society of Chemistry, Cambridge, pp. 56–99.
Petr Dejmek
Lund University
Marilyn Rayner
Lund University
Editors
Marilyn Rayner graduated with a bachelor’s degree in
biological engineering, with a food engineering special-
ization from the University of Guelph, Canada, in 1999
and earned her PhD in food engineering from Lund
University in 2005 on modeling droplet formation in
membrane emulsification. Dr. Rayner was awarded the
Food Engineer of the Year 2005 from the Institution of
Mechanical Engineers, United Kingdom, for her work on
predicting the effects of pore geometry on droplet size
in membrane emulsification processes. Since then, Dr. Rayner has worked in the
area of multiphysics modeling, unit operation, interfacial phenomena, and particle-
stabilized emulsions, and is currently associate professor in food engineering at
Lund University and the founder of two spin-off enterprises in the area of membrane
design and particle-stabilized formulations.
xix
Contributors
The collective of authors contributing to this volume comprise a major portion of the
researchers who have in recent years been most active in fundamental research on
the process of emulsification aimed at food industries.
xxi
Synopsis
SECTION I—EMULSION FUNDAMENTALS
Chapter 1
Scales and Forces in Emulsification
Marilyn Rayner
An overview is provided of the characteristics of the types of equipment used in the
mechanical production of emulsions and the basic governing physics of emulsifica-
tion in the limits of high-energy emulsification in processes such as vat mixers, col-
loid mills, high-pressure homogenizers, microfluidizers, ultrasound transducers, and
low-energy emulsification in channel devices of membrane, sieve, or microchannel.
Droplet size distributions, the classical Taylor droplet breakup criterion, the dimen-
sionless Weber number, timescales of deformation and the Kolmogorov–Hintze
theory of turbulent breakup are introduced.
Chapter 2
Emulsion Formation and Instability
Björn Bergenståhl
This chapter provides a reasoned explanation of the instabilities based on cream-
ing/sedimentation, flocculation, coalescence, and Ostwald ripening. Stokes’s settling
equation is discussed with respect to hindered settling and non-Newtonian viscosity.
The arrested state of both attractive and repulsive colloidal glasses is introduced,
with size and volume fraction effects. Brownian, shear-induced, and sedimentation-
induced collisions are discussed, as well as the interface parameters governing
coalescence stability; van der Waals, solvation, and electrostatic forces; and poly-
mer-induced steric repulsion, depletion attraction, and bridging.
Chapter 3
Formulation of Emulsions
Marie Wahlgren, Björn Bergenståhl, Lars Nilsson, and Marilyn Rayner
An overview is provided of the effects of compositional choices on the end-user func-
tionality of emulsions, including colloidal, microbiological, and oxidative stability; fla-
vor release; and freeze–thaw stability. The properties affecting emulsion functionality
are reviewed in detail for lipids, low-molecular emulsifiers, proteins, gelling and non-
gelling polysaccharides, protein–polysaccharide complexes, and food-based particles.
Emulsion characterization methods such as emulsification capacity, emulsion sta-
bility index, creaming index, accelerated characterization methods based on critical
osmotic pressure, and emulsion rheological test are described together with their
theoretical underpinnings.
xxiii
xxiv Synopsis
Chapter 4
Particle-Stabilized Emulsions
Malin Sjöö, Marilyn Rayner, and Marie Wahlgren
The theoretical basis of the high stability of particle-stabilized emulsions based on
detachment energy is shown, and the literature on the specifics of particle-stabilized
emulsions with respect to creaming, coalescence, Ostwald ripening, surface tension,
and rheology are reviewed in detail.
Examples of food-grade stabilizing particles, including modified starch granules,
lipid particles, soy and zein particles, egg yolk particles, cellulose, chitin, and flavo-
noids are given.
Chapter 6
High-Pressure Homogenizer Design
Fredrik Innings
Starting with the history of high-pressure homogenizers, the pros and cons of design
choices in modern commercial homogenizers are clarified, with a focus on effi-
ciency, process line integration, wear, and investment and running costs.
Chapter 7
High-Pressure Homogenization with Microstructured Systems
Karsten Köhler and Heike Schuchmann
The development of the orifice-type high-pressure homogenization is presented from
both theoretical and technical perspectives, including patent literature. In addition to
the approaches of the previous chapters, attention is paid to the role of spatial varia-
tion of the flow field in different orifice geometries and the methods available for its
study, in particular modeling and simulation. Experimental results from different
orifice geometries are compared.
Synopsis xxv
Chapter 8
Rotor–Stator Devices
Karsten Köhler and Heike Schuchmann
Emulsification in pumps, stirred vessels, colloid mills, toothed-rim dispersers,
double rotor dispersers, and extruders is described. The unifying parameter of
energy density is described, and the positive consequences of batch processing, in
particular for emulsions in which the droplets are only slowly being stabilized, are
pointed out.
Chapter 10
Emulsification with Microsieves and Other Well-Defined
Microstructured Systems
Karin Schroën and Akmal Nazir
In membrane emulsification, one of the main issues is control of the droplet size. Due to
the polydispersity of the membrane pores, polydisperse emulsions are obtained; there-
fore, the use of membranes with equally sized pores is seen to control droplet size. In
this chapter, microsieves and other devices with monodispersed pores are presented;
their performance in cross-flow emulsification is compared to that in regular mem-
branes and more classic emulsification devices. Further, premix emulsification with
metal sieves with uniform pores, and the combination of metal sieves with glass bead
beds, is described as a means to prevent low pore activation, which is a drawback of
using devices with uniform pores. The pros and cons of all methods are summarized.
xxvi Synopsis
Chapter 11
Formation and Modification of Dispersions Using
Shirasu Porous Glass Membranes
Goran T. Vladisavljevic’
Shirazu porous glass (SPG) membrane morphology, chemical composition, and
methods to modify its surface chemistry, as well as the common setups for SPG
membranes are presented. Emulsification performance parameters of the system,
both in cross-flow and direct emulsification and in premix, straight-through emulsi-
fication, are explained and quantified.
Examples are given of the use of PSG membranes for the formation of W/O, O/W,
and multiple emulsions, as well as emulsion-based particles and gas bubbles.
Section I
Emulsion Fundamentals
1 Scales and Forces
in Emulsification
Marilyn Rayner
contents
1.1 Introduction ......................................................................................................3
1.2 Droplet Size Distributions ................................................................................4
1.3 Overview of Emulsification Machines and Homogenization Devices .............7
1.4 Droplet Disruption Fundamentals .................................................................. 11
1.4.1 Interfacial Forces ................................................................................ 12
1.4.2 Disruptive Forces ................................................................................ 13
1.4.3 Flow Types .......................................................................................... 15
1.5 Flow Regimes ................................................................................................. 16
1.5.1 LV Flow Regime ................................................................................. 18
1.5.2 Laminar Interfacial Tension-Driven Regime ..................................... 19
1.5.3 TI Regime ........................................................................................... 23
1.5.4 TV Flow Regime ................................................................................24
1.5.5 Cavitation Inertial Regime .................................................................25
1.6 Comparison of Emulsification Efficiency .......................................................26
1.7 Complications and Concluding Remarks ....................................................... 29
References ................................................................................................................ 30
ABSTRACT This chapter introduces the mechanical aspects of the formation and
breakup of emulsion droplets. An overview of the most common types of emulsifica-
tion machines and homogenization devices is provided. Droplet disruption mecha-
nisms are presented considering the forces, length, and timescales involved in the
various types of flow regimes encountered in the mechanical production of emulsions
by different categories of equipment. Finally, the energy efficiency of emulsification
methods is discussed, highlighting some complications with the theory presented, in
addition to some general comments on the future in the area.
1.1 IntroductIon
In order to make an emulsion, four basic ingredients are required: two immiscible
phases (often oil and water), energy to create the oil–water interface of the emulsion
drops, and surfactants to stabilize this interface to prevent the drops from coalesc-
ing immediately upon contact. By varying the amounts and composition of these
four basic ingredients, we can control key emulsion properties such as emulsion
3
4 Engineering Aspects of Food Emulsification and Homogenization
type (i.e., oil-in-water [O/W] versus water-in-oil [W/O]), droplet volume fraction,
ϕ (also known as “dispersed phase faction”), droplet size and size distribution, and
the nature of the stabilizing layer surrounding the droplets. These properties deter-
mine many of the central organoleptic properties of emulsion-based food products
such as shelf life, appearance, rheology, texture, and flavor. How we convert energy
from bulk mechanical stresses during emulsification to generate and stabilize the
new interfacial area created as a result of the emulsion drops is, of course, central to
the study of emulsification technologies and the focus of much of this book. A wide
variety of processing equipment and technologies are available for the generation of
emulsions, many of which will be thoroughly discussed in Chapters 5 through 11.
In line with the aim of this book, we consider emulsification in the context of food
emulsions, and thus only the mechanical production of emulsions will be discussed.
In the chemical industry, there are other methods to produce emulsions that do not
rely on mechanical generation of droplets, such as precipitation of the dispersed
phase previously dissolved in the continuous phase and the phase inversion tem-
perature method. These and other nonmechanical methods, however, are outside the
scope of this book. This chapter is an overview of the physics and technologies used
in the mechanical production of emulsions. For a more comprehensive treatment of
the topic, the reader is referred to Walstra’s eminent works in the area (Walstra 1993,
2003, Walstra 2005, Walstra and Smulders 1998) as well as Schubert and Ax (2003),
Karbstein and Schubert (1995), and McClements (2005).
The basic physical process of emulsification is illustrated in Figure 1.1, which
depicts the two main types of emulsification processes: high energy and low energy. In
both cases, the four main ingredients of emulsions are shown: (1) a continuous phase,
(2) a dispersed phase, (3) an emulsifier or a stabilizer, and (4) the necessary energy
input to create more oil–water interface. In the case of high-energy emulsification,
intense mechanical input is used to break up a coarse premix of emulsion drops. This
mechanical energy may arise from stirring, shearing, turbulent eddies, or even from
ultrasonic waves. This can be practically achieved through a wide range of different
technical designs and apparatus, however what is similar for all high-energy methods
is that droplet disruption is determined by the strength and duration of the applied
mechanical energy. High-energy emulsification processes are discussed in detail in
Chapters 5 and 6, on high pressure homogenization; Chapter 7, on microfluidization;
and Chapter 8, on rotor stator devices.
In the case of low-energy emulsification processes, a coarse premix is not nec-
essary. However, in some cases, the dispersed phase consists of an emulsion itself.
Droplets are formed directly in the continuous phase by injection through a porous
membrane or microchannel. Low-energy emulsification processes are discussed in
detail in Chapters 9 through 11.
1. Continuous phase
Deformation
2. and disruption
Dispersed Fast stabilization
phase with sufficient
concentrations
3. Emulsifier
stabilizer
Slow stabilization
or deficient
concentrations
Coalescence
(a)
3. Emulsifier/stabilizer
Phases Emulsification in the droplet formation zone and stabilization of new interface—
generally separate uniform drops
(b)
FIGure 1.1 (a) High- and (b) low-energy emulsification processes. (High-energy processes
redrawn based on Karbstein, H., and Schubert, H., Chem. Eng. Process., 34, 3, 205–211, 1995.)
6 Engineering Aspects of Food Emulsification and Homogenization
∫
Sa ≡ d a f ( d )∂d
0
(1.1)
∑
n
di3
d32 = i =1
(1.3)
∑
n
di2
i =1
∑
n
di4
d43 = i =1
(1.4)
∑
n
di3
i =1
where:
di is the diameter of the ith measured drop of a total of n drops measured
Another emulsion variable directly related to the droplet size distribution is the spe-
cific surface area of the emulsion, AS:
6φ
As = (1.7)
d32
Scales and Forces in Emulsification 7
where:
ϕ is the volume fraction of the disperse phase droplets and has the unit m−1, that
is, m2 interfacial area of emulsions droplets per m3 of emulsion
This specific surface area is important as it relates to both the amount of energy put
into the system to create this interface as well as the amount emulsifier or surfactant
required to stabilize the emulsion.
There are many different methods to assess particle size distributions, such as
microscopy (light, confocal, electron, etc.), particle counters (i.e., Coulter counter),
light scattering (i.e., Malvern Mastersizer), nuclear magnetic resonance (via diffusion
times), and sedimentation/centrifugation (Walstra 2005). They vary with respect to
particle size ranges covered, measurement principles, and limitations. The interested
reader is directed to McClements’s (2007) comprehensive review on emulsion char-
acterization techniques for more details.
Inlet (premix)
Stator
Rotor
Outlet
(a) (b)
Outlet
Ultrasound probe
Outlet
Inlet
Valve
Impact ring
Valve seat
Inlet
(c) (d)
Piston pump
Channels
Circulation of the
continuous phase
Impact
zone
Fine emulsion
(e) (f)
of the stirrer. Thus, mixed tanks are generally limited to the production of coarse
emulsions batchwise as a pre-step to other emulsification processes.
In rotor–stator type equipment such as colloid mills and toothed-disc dispersing
machines, drops are disrupted in the gap between the rotating rotor and the station-
ary stator. In colloid mills, drops are disrupted in the conical gap, which can be either
smooth or serrated with various designs. Here, the droplet disruptive stresses are
determined by the gap width (typically 100–3000 µm), rotor radius, rotational rate
(typical peripheral speeds between 5 and 40 m s−1), and the liquid flow rate through
the gap, which can range between 4 and 20,000 l h−1 (Karbstein and Schubert 1995,
McClements 2005). Colloid mills are most suitable for production of intermediate
to high viscosity products and can achieve droplet diameters between 1 and 5 µm
(McClements 2005). Toothed disc dispersing machines are similar to a colloid
mill, except that the flow is not specifically bounded, consisting of single or several
table 1.1
some Features of common methods and machines to produce emulsions
bounded or energy relative energy droplet sizes
homogenizer types Flow regime un-bounded Flow density (J m−3) efficiency achieved viscosity typical volumes
Stirred tanks TI, TV, LV U Low-high Low 2 µm and larger Low-to-medium Batches up to
103–106 several m3
Colloid mill LV (TV) B Low-high Intermediate 1–5 µm Medium-to-high 4–20,000 l h−1
103–108
Toothed-disc disperser TV B Low-high Intermediate 1–10 µm Low-to-medium Batches cm3 up to
Scales and Forces in Emulsification
CI, cavitation inertial; LV, laminar viscous; STB, spontaneous transformation based, TI, turbulent inertial; TV, turbulent viscous.
9
10 Engineering Aspects of Food Emulsification and Homogenization
pairs of concentrically arranged split discs, one of which rotates at high speed, with
typical peripheral speeds up to 40 m s−1 (Karbstein and Schubert 1995).
High-pressure homogenizers, in contrast to mixers and rotor–stator type machines,
do not have moving parts acting in the droplet disruption zone. High-pressure valve
homogenizers are currently the most common method of producing fine emulsions
in the food industry. Here, a high-pressure pump delivers a coarse premix emulsion
through a narrow orifice or valve, which can have various designs. High-pressure
drops are achieved (typically between 3 and 20 MPa) where the emulsion flows
at very high speed (up to 2000 ms−1). Droplets are disrupted due to elongational
flow, eddies, stress fluctuations inside turbulent jets, or in some cases, by cavitation
(Karbstein and Schubert 1995, Schultz et al. 2004). High-pressure homogenization
is most suitable for low-to-medium viscosity products (such as milk and cream) and
submicron-sized droplets can be attained by a single pass through the homogenizer
with industrial production rates ranging from 100 to 20,000 l h−1. Most commercial
homogenizers use a spring-loaded standard valve design such that the gap through
which the emulsion is passed can be adjusted (typically between 15 and 300 µm)
(McClements 2005). Decreasing the gap height increases the pressure drop across
the valve, thereby increasing the intensity of the flow and droplet disruptive forces,
which in turn can produce smaller droplets. The relationship between pressure drop
and droplet size is roughly linear in the log–log scale, that is, log d ∝ log P, the
slope depending on flow conditions within the homogenizer, which in turn are deter-
mined by the valve geometry and fluid properties (Karbstein and Schubert 1995,
Phillips 1985).
Ultrasonic homogenizers use high-intensity sound waves to generate intense
shear and pressure gradients within the liquid that disrupts droplets mainly by
cavitation and turbulent effects (McClements 2005). There are two methods com-
monly used in the industry to produce ultrasonic waves. Piezoelectric transducers
are used for small batch volumes ranging from a few cubic centimeters to a few hun-
dred cubic centimeters, and liquid jet generators are used on a larger scale where
a jet coarse emulsion is pumped to impinge on a sharp-edged blade. This jet flow
causes the blade to vibrate rapidly, thus generating the ultrasonic field that breaks
droplets in its immediate vicinity. Ultrasonic jet-type homogenizers can produce
larger volumes continuously with fluid flow rates ranging from 1 to 500,000 l h−1.
The factors that govern droplet disruption are intensity, duration, and frequency of
the ultrasonic waves in relationship to the volume of emulsion they are applied on
(McClements 2005).
Microfluidizers are similar to high-pressure homogenizers in that they do not
have moving parts in the droplet disruption zone. This type of homogenizer consists
of fluid inlet(s), some type of powerful pumping device, and an interaction chamber
containing channels through which the fluids are forced to flow and interact with
each other in an impact zone. A coarse premix emulsion (or in some cases the indi-
vidual oil and water phases depending on the design) is accelerated to exceptionally
high velocities within the channels of the microfluidizer and then led to simulta-
neously impinge on each other on a solid surface inside the interaction chamber
(McClements 2005). Microfluidizers are available in a wide range of production vol-
umes ranging from 10 ml up to 12,000 l h−1, maintaining an operating pressure of
Scales and Forces in Emulsification 11
up to 270 MPa, thus having the ability to produce very fine emulsions with droplet
sizes <0.1 µm (Lee and Norton 2013, McClements 2005).
Membrane and microchannel emulsification methods involve pressing a dispersed
phase (either as a single phase or as a premix) through a microporous membrane or
a microchannel chip. This process, invented by Nakashima’s group (Nakashima,
Shimizu, and Kukizaki 1991), makes use of microporous membranes, where the
continuous phase flows tangentially to the membrane surface and the dispersed
phase is pressed through the membrane. Droplets of dispersed phase are formed at
the openings of the membrane pores and are continuously swept away by the flowing
continuous phase. The key feature of the membrane emulsification process, which
sets it apart from conventional emulsification technologies, is that the size distri-
bution of the resulting droplets is primarily governed by the choice of membrane
rather than the development of turbulent or extensional droplet breakup (Peng and
Williams 1998). Hydrophilic membranes are used in making O/W emulsions and
hydrophobic membranes are used in making W/O emulsions, the point being that the
dispersed phase should not wet the membrane surface, that is, the oil–membrane–
water contact angle should be >145°. A close relative to membrane emulsification is
microchannel emulsification, which operates on the same principle of extruding the
dispersed phase into the continuous phase, but differs with respect to the membrane.
As the name implies, microchannel emulsification uses a microfabricated channel
array. These plates can be stacked to create a module to increase production output.
Membrane and microchannel emulsification methods are carried out at very low-
energy input rates, and although energy is required for the generation of droplets,
they are not subsequently broken by flow conditions in the continuous phase. Here,
droplet size is determined mainly by the geometry of the pore or channel rather than
by the flow conditions in the continuous phase, thus the energy transfer for creating
a new interfacial area (i.e., the area of emulsion droplets) is much more efficient than
in high-energy processes, which rely on turbulent or viscous shearing by the flowing
continuous phase. For this reason, low-energy emulsification processes often have
narrow droplet sizes and are more suitable for sensitive ingredients, but generally
have larger droplet sizes and much lower production rates.
In the following sections, the scales and forces associated with emulsion formation
will be discussed and several simplifications have been made. Here, the interfa-
cial tension, γ, is assumed constant. This means that either no surfactant is present
(not ever practically the case) or that there is a large excess. Furthermore, we have
assumed the phases to have a Newtonian viscosity (true for oil and dilute water
phases) as well as relatively low volume fractions; thus emulsification processes for
high internal-phase emulsions are not considered.
2γ
pL = (1.9)
R
where:
γ is the interfacial tension
R is the radius of a spherical drop
For any general closed surface, the Laplace pressure is a function of the mean curva-
ture, H, defined by its two principal radii of curvature, R1 and R2:
1 1
pL = 2 γH = γ + (1.10)
R1 R2
One way to understand the mechanical basis of the Laplace pressure is to consider a
differential change in surface energy dEsurf = γdA and the corresponding differential
change in volume for a drop of a given size.
dEsurf γdA
pL = = (1.11)
dV dV
Recall the following geometric relationships for a sphere and the chain rule of
calculus:
Scales and Forces in Emulsification 13
Area:
dA
A = 4πR2 , = 8πR
dR
Volume:
4 3 dV
V= πR , = 4πR2
3 dR
Thus,
dA dR dA 2
⋅ = =
dR dV dV R
Substituting into Equation 1.11 we get the expression of Laplace pressure for a spher-
ical drop:
dEsurf γdA 2 γ
pL = = = (1.12)
dV dV R
To break an existing drop into smaller ones in the context of emulsification, it must
be strongly deformed, and any deformation from a spherical shape increases the
Laplace pressure. As seen in Figure 1.3, R1 and R2 are the principal radii of curva-
ture in Equation 1.10. This deformation can be achieved via shear and elongational
flow in the continuous phase or by extrusion through a noncircular pore. According
to Taylor’s (1934) theory of deformation-induced breakup of droplets, the droplet
deformation can be described by
L − B R1 − R2
δTaylor = = (1.13)
L + B R1 + R2
His work showed that under idealized conditions the breakup of a droplet is only
possible when the deformation achieves a critical value δTaylor > 0.5. If we set
Equation 1.13 equal to 0.5 and solve for L, we get L > 3B; meaning, if a drop is
deformed into an ellipsoid to such an extent that its length is more than three times
its breadth, it can break up into two daughter drops. This criterion seems to hold in
shear stress induced by viscous flows as well as in extrusion through membranes and
microchannels. The latter phenomenon has been termed spontaneous transformation
based (STB) droplet formation. In this case, the interfacial forces are also driving
the droplet formation, rather than being merely a source of resistance that disruptive
forces need to overcome. This is discussed further in Section 1.5.2.
L
R1
R1
R2 B R2
x
dVx
(a) Droplet breakup in simple shear flow, G = = shear rate (velocity gradient)
dy
L
x R1 R1
R2 B
R2
dVx
(b) Droplet breakup in elongational flow, G = = elongation rate
dy
Membrane or
microchannel
FIGure 1.3 Illustration of droplet formation in laminar flows: Simple shear flow (a) and
elongational flow (b). As a spherical droplet is deformed it takes on an ellipsoidal shape with
two principal radii of curvature, R1 and R2. Once deformation exceed a certain extent it will
break into two or more daughter droplets. A similar deformation criteria can been seen in
extruding through pores and channels with high aspect ratio geometries (c).
the machine (see Table 1.1). It is the latter that is the basis for the external stresses that
lead to the generation of disruptive forces for droplet breakup. In most emulsification
processes, the velocity or pressure gradients required for droplet disruption forces are
delivered via flow conditions in the continuous phase; the smaller the desired drops,
the more intense this flow needs to be. There are two types of external forces causing
the disruption of drops during emulsion formation: frictional and inertial.
Frictional forces (also called as viscous or shear forces) are due to the flow
of the continuous phase along the surface of the drop, thus viscous forces
mainly act parallel to the drop surface. The local stress generated is a func-
tion of the strain rate, G (velocity gradient), and the continuous phase vis-
cosity, ηc, that is, σ ∼ Gηc (see Figure 1.3).
Scales and Forces in Emulsification 15
Inertial forces are generated by the local pressure fluctuations caused by local
velocity fluctuations in turbulent flows. They generally act perpendicu-
lar to the surface of drops. The local stress generated is a function of the
continuous phase density and the mean and local velocity differences, that
is, σ ∼ ρcu ∆u .
where:
U is linear velocity
L the characteristic length
ρ the density
η the viscosity of the flowing fluid
The choice of U and L depend on the whether we are considering the flow in relation
to the apparatus it is flowing through, or the flow experienced by a drop in a continu-
ous medium. For example, for the flow in a pipe, U is the mean linear liquid velocity,
that is, u, and L is the diameter of a pipe. In the case of flow around a suspended
drop, U is the drops velocity relative to the continuous phase directly surrounding
it and L is the drop diameter. In the case of a mixer in a vessel, Rei is related to the
diameter and rotational speed of the impellor.
Ni Di2ρ
Rei = (1.15)
η
For each type of flow, there exists one or more flow regimes that can describe the
source and intensity of disruptive forces acting on droplets. Within each flow regime,
16 Engineering Aspects of Food Emulsification and Homogenization
there is an essential variable that describes the intensity of the disruptive forces
(or stresses) acting upon the droplets. This is often a mechanical stress, σ, generated
by the flow. The ratio of the disruptive forces (pulling drops apart) to the interfacial
forces (holding drops together) is represented by the Weber number in turbulent
flows and the capillary number, Ca, in laminar flows (also known as the laminar
Weber number, WeL).
inertial force
We = (1.16)
interfacial tension force
viscous force
Ca = We L = (1.17)
interfacial tension force
The origin, magnitude, and efficacy of the inertial and viscous forces to achieve
droplet disruption are determined by the type of flow conditions and the properties
of the phases.
Forces
3
External stress acting on droplets (σ) ηc G ε ηc ε 2 d 2 ρc
length scales
1/ 5
Mean diameter (d≈)b (2 γ We cr )/(ηc G) γ/( ε ηc ) γ 3 /(ε2 ρc )
Scales and Forces in Emulsification
timescales
Sources: Adapted from McClements: D.J., Food Emulsions: Principles, Practices, and Techniques, CRC Press, Boca Raton, FL, 2005; Walstra, P., Chem. Eng. Sci., 48, 2,
333–349, 1993; Walstra, P., and Smulders, P.E.A., Modern Aspects of Emulsion Science, The Royal Society of Chemistry, Cambridge, pp. 56–99, 1998.
Symbols: We, Weber number (see text); Re, Reynolds number (see text); Γ, surface excess of surfactant (mol m−2); ε, power density (J s−1 m−3); d, droplet diameter (m);
γ, interfacial tension (J m−2); η, viscosity (Pa s); G, velocity gradient (s−1); mc, surfactant concentration in the continuous phase (mol m−3); τ, characteristic time (s);
σ, stress (pa); ρ, density (kg m−3).
Subscripts: d, dispersed phase; c, continuous phase; DEF, deformation; ADS, adsorption; cr, critical value for droplet break-up.
a For d > η2 /( γ ρ ).
c c
d c
b Only if η η .
17
18 Engineering Aspects of Food Emulsification and Homogenization
Here, the droplets are deformed by viscous shear stresses acting on the droplets
generated by a velocity gradient inside a viscous liquid. There are two extremes
with respect to velocity gradients: simple shear and purely elongational (Figures 1.3).
Simple shear causes rotation as the fluid is moving faster on the top of the drop rela-
tive to the bottom of the drop, which also causes the liquid inside the drop to rotate.
For small Weber numbers, the Taylor deformation (Equation 1.13) is equal to the
Weber number in laminar flow:
σ G ηc d
We L = = (1.18)
0.5 pL 2γ
where:
d is the droplet diameter (Walstra 2003)
For larger values of We, the drop becomes deformed and will break if We exceeds a
critical value Wecr. Numerous experimental studies have considered how flow condi-
tions and viscosity ratio of the continuous and dispersed phases affect the critical
Weber number required for the droplet breakup. Figure 1.4 is a typical plot of the
effect of viscosity ratio on Wecr for various types of laminar flow. By taking Wecr
from the plot and inserting it into Equation 1.18, the largest drop with a given viscos-
ity and interfacial tension that could remain unbroken in a given applied flow field
can be estimated. For simple shear the following equation gives a good approxima-
tion for the maximum droplet sizes that can persist under steady-state conditions in,
for example, a simple shear flow found in colloid mills (Walstra 1993, Walstra and
Smulders 1998):
α=0
10 Shear
Critical Weber number, Wecr
α = 0.2
α=1
Elongational α = 0.6
0.1
Viscosity ratio, ηd/ηc
FIGure 1.4 The effect of viscosity ratio ηd/ηc (dispersed phased over continuous phase) on
the critical Weber number required for drop breakup in the laminar flow with varying degrees
of elongation, from simple shear (α = 0) to purely plane hyperbolic elongation (α = 1).
Scales and Forces in Emulsification 19
2γ We cr
dmax = (1.19)
ηcG
However, in practice, the average droplet sizes obtained are somewhat smaller
(Walstra 2003). The viscosity ratio has a significant effect on Wecr in a simple shear
flow and if the dispersed phase has a viscosity more than four times that of the con-
tinuous phase, it does not disrupt at all, but rather rotates in the flow field. The reason
for this lies in relative magnitude the timescales of droplet deformation and rotation.
The time required for deformation in simple shear is as follows:
ηd
τDEF = (1.20)
ηc G
The time required for a drop to turn half a rotation is π/G; thus setting this equal to
Equation 1.20 yields ηd = πηc, which approaches 4ηc (Walstra 2003). Aside from the
strong effect viscosity ratio has on whether the droplet breakup seen in Figure 1.4 can
be achieved, the type of flow also plays a significant role where elongational flows
have much lower critical Weber numbers. In elongational flow (plane hyperbolic),
the denominator of Equation 1.19 would be 2ηcG. Furthermore, there is no internal
rotation of droplets in elongational flow and becomes highly elongated independent
of the viscosity ratio. Laminar flow patterns can also be combinations of shear and
elongation (curves shown in Figure 1.4 for 0 < α < 1), and such types of intermediate
flow are common in stirred vessels and mixers (Walstra 2003).
In many practical situations, different emulsifying equipment is compared by
their power density, that is, how much energy per unit time and volume is available
for droplet disruption. In simple laminar flows, this is a function of the viscosity and
local velocity gradient with the units of J s−1 m−3.
ε = ηcG 2 (1.21)
For the case of a mixer or stirrer in a vessel with laminar flow conditions (Rei < 10)
the power density, ε, is proportional to the revolution rate squared, Ni2 and the
impellor diameter to the power 3 divided by the fluid volume in the vicinity of the
impellor Vi ∼ Di (the term Di cancels out).
3
ε = klηN i2 (1.22)
The constant k l varies depending on the impellor type and Reynolds number. Plots
of these constants are found in texts on industrial mixing (Doran 1995); for example,
at Rei = 1, for Rushton turbines k l equals 70, for paddle type stirrers 35, for pitched
propellers 40, and for helical ribbons 1000.
FIGure 1.5 Images of STB droplet-forming structures. Left: illustration of the pore
structure of a microporous glass membrane. (Courtesy of SPG instruction manual, ISE
Chemical Company, Japan.) Right: microchannel with rectangular-shaped pores. (Courtesy
of Aquamarijn BV.)
Scales and Forces in Emulsification 21
around 3. Droplets formed spontaneously from pores having an aspect ratio above 3,
whereas pores with aspect ratios below 3 had poly-dispersed droplets detaching with
the help of the continuous phase flow. Pores having an aspect ratio at 3 displayed a
mixed mechanism, where it was observed that some of the pores produced droplets
spontaneously whereas others exhibited large droplet inflation. This phenomenon
has been observed in microscope video images of droplets forming from round pores
by van der Graaf et al. (2004) and Kobayashi et al. (Kobayashi, Mukataka, and
Nakajima 2004, Kobayashi, Uemura, and Nakajima 2006, Kobayashi, Yasuno et al.
2002) as well as in computational fluid dynamics (CFD) simulations performed by
Abrahamse et al. (2001). Conversely, Kobayashi et al. (2004) did not observe neck-
ing outside the pore for their oblong straight through microchannels with aspect
ratio 3.8, but did see continuous outflow from pores with aspect ratios of 1, 1.9, and
2.7. The location of droplet necking can explain why markedly different emulsifica-
tion results are obtained depending on the shape of the pore. Pores with aspect ratios
greater than 3 produce droplets of regular size because necking occurs inside the
pore, which is illustrated in Figure 1.3c. Rounder pores with aspect ratio less than
3 produce poly-dispersed droplet-size distributions because the necking takes place
outside the pore exposed to the flow of the continuous phase. This flow, although
often laminar at these dimensions, is still rather chaotic. The shear stress to which any
given droplet is exposed to during necking may be drastically different depending on
when a neighbor droplet is detached and the path it took as it moved away from the
membrane surface, that is, droplets can shield each other from the shearing liquid.
An explanation as to why an aspect ratio of 3 is important can be found again in
Taylor’s (1934) description of the theory of deformation-induced breakup. Taylor
defines the deformation of a spherical oil droplet into an ellipsoid, as the droplet is
subject to shear-inducing flow inside a narrow gap, recall Equation 1.13.
L−B
δTaylor =
L+B
Here, L and B are the lengths of the major and minor axes directions. It was estab-
lished that the droplet breakup is only possible when the deformation achieves a
critical value δcrit > 0.5. If we set δTaylor = δcrit and solve for L, one obtains L > 3B.
Therefore, if Taylor’s ellipsoidal deformation concept is applied to the cross section
oil protruding inside a membrane pore or microchannel, then L/B is, by definition,
its aspect ratio and if it is greater than 3, breakup (i.e., necking) can occur in the pore
or microchannel.
The diameter of the resulting droplets can be estimated based on the interfacial
tensions, the critical pressure of the membrane pore or microchannel determined by
its geometry, and the three-phase contact angle (Rayner 2005, Rayner, Trägårdh, and
Trägårdh 2005). Equations and details are found in Table 1.3. If the aspect ratio is less
than 3, then further deformation of the droplet via drag forces generated by the flow-
ing continuous phase tangential to the pore are required for droplet detachment. This
type of shear-assisted droplet detachment is similar to LV, as it is the deformation of
the droplet via the wall shear stress that leads to droplet breakup (Peng and Williams
1998, Schroder and Schubert 1999, Timgren, Trägårdh, and Trägårdh 2009, 2010).
22
table 1.3
droplet sizes Generated in continuous emulsification processes for various Flow conditions
Flow regimes membrane/microchannel rotor stator/colloid mill high-pressure homogenizer
Sources: Membrane and micro-channel correlations modified from Peng, S.J., and Williams, R.A., Chem. Eng. Res. Des., 76, A8, 894–901, 1998; Rayner, M., and
Trägårdh, G., Desalination, 145, 1–3, 165–172, 2002; Rayner, M. et al., Colloids Surf., A, Physicochem. Eng. Aspects, 266, 1–3, 1–17, 2005. Other from Walstra,
P., and Smulders, P.E.A., Modern Aspects of Emulsion Science, The Royal Society of Chemistry, Cambridge, pp. 56–99, 1998; Schubert, H., and Ax, K., Texture
in Food, Woodhead Publishing Ltd, New York, 2003.
Symbols: d, droplet diameter (m); γ, interfacial tension (J m−2) of the surfactant covered interface; γo, interfacial tension of the bare oil–water interface; Pcrit, critical break-
through pressure of extrusion of oil through the membrane; θ, oil–water–membrane contact angle; H, mean curvature of the pore or channel (see Equation 1.10);
Dpore, the diameter of a membrane pore of the microchannel, kwall, 1.7; σwall, wall shear stress (Pa); Pv, net power density (J s–1 m–3); EV, energy density
(see Equation 1.36 ) (J m−3); t, time in the homogenizer; A1, A2, . . ., A5 are constants that are a function of physiochemical properties of the fluids, that is,
Ai = f (ηd , ηc , ρc , γ,), in general d max = Ai EV− b, where b is an experimentally determined exponent and in this case dmax is the asuter mean diameter, d32.
Engineering Aspects of Food Emulsification and Homogenization
1.5.3 tI regIme
Reflow > 2500, Redrop > 1
Turbulent flow is typified by the occurrence of eddies (swirls and vortices), thus the
local velocity at a given point in the flow has a local velocity that generally differs
from its time-average value of u. This local velocity fluctuates chaotically and the
average difference over time between u and u will equal zero; however, the root
mean square of these fluctuations are finite (Walstra 2003).
u′ ≡ (u − u )21/ 2 (1.23)
It is the magnitude of these fluctuations on the length scale of turbulent eddies that
give rise to disruptive forces on drops. TI breakup takes place when the droplet
deformation is caused by the smallest scale eddies in the system and, therefore, drop-
lets are of the same order of magnitude to the eddy size. In the TV breakup regime
(Section 1.5.4), droplet sizes are reduced below the size of the smallest eddies in the
system from the shearing forces created within these eddies.
Turbulent flow has a spectrum of eddy sizes (l). The larger the eddy, the higher
the value of u′. Large eddies transfer their kinetic energy to smaller eddies. Although
smaller eddies have a lower degree of velocity fluctuations (u′), they have a higher
velocity gradient (u′/l ). Thus, the small eddies have a high-specific kinetic energy
and are called energy-bearing eddies having the size le. As such, the local velocity
on the size scale of these eddies is given by:
Due to the chaotic nature of turbulent flow, the droplet disruption likely occurs via an
abrupt local protrusion or indent on a drop. There is no reason to believe that drops
will break into to two equal volumes, hence the resulting droplet size distributions
are fairly wide, as observed experimentally. For valve-type high-pressure homoge-
nizers, the homogenizing pressure, pH, can be varied and the resulting power density
is proportional to p1H.5. For the case of a mixer or a stirred vessel the power density,
ε, is proportional to the revolution rate cubed, Ni3 the impellor diameter to the power
5 divided by the fluid volume in the vicinity of the impellor Vi ∼ Di . The constant k P
3
is called the power number and varies depending on the impellor geometry, but is
constant for a given geometry at Rei > 105, k P is equal to 5–6 for Rushton turbines, 2
for paddle-type stirrers, and 0.35 for pitched propellers (Doran 1995).
For other types of machines and a more rigorous theoretical treatment of the topic,
the interested reader is referred to Vankova et al. (2007) and Walstra’s works and
references therein.
eddies (Equation 1.25) and drop size. Here, the limits of Redrop can be estimated
using Equation 1.30 (Walstra 2003). It also gives the limit for condition for the larg-
est droplet in the Redrop in the TI regime (d > η2c /γρc ) in Table 1.2.
γ1/ 2ρ1c/ 2 d 1/ 2
Re drop = (1.30)
ηc
In the TV regime, the scale of the droplets is much smaller than that of the energy-
carrying eddies (d le ); however, the local velocity needed at the scale of the
droplet is
because the shear stress acting on the drop is given by viscosity time and the velocity
gradient, that is,
u′(d )
σ = ηc (1.32)
d
As the flow between eddies is likely elongational (i.e., plan hyperbolic), Wecr will
not strongly depend on the viscosity ratio and the maximum droplet size in the TV
regime can be estimated by Equation 1.33 (McClements 2005, Walstra 2003).
Eddies
pressure
fluctuations
Drops are larger than the smallest eddies and deform by the
(a) action of hydrodynamic pressure fluctuations
shockwave
FIGure 1.6 Illustration of droplet formation in turbulent flows: (a) droplet breakup in
turbulent inertial regime, (b) droplet breakup in turbulent viscous regime, and (c) droplet
breakup in cavitation.
Emin
ΦH = 100% (1.34)
EV
6φ
Emin = γ (1.35)
d32
In emulsification, as the droplets are broken into smaller ones, the interfacial area
becomes significantly larger, and the free energy of the system is increased by an
Scales and Forces in Emulsification 27
∫
EV = PV t dt (1.36)
where:
PV is the net power density
t is the duration of the emulsification procedure
However, PV needs to exceed some critical value for droplet disruption to take place,
which depends on the Laplace pressure of the droplet to be broken. For most of the
common types of homogenizers used in the food industry, theoretical or semiempiri-
cal equations are available to estimate the energy density (Karbstein and Schubert
1995, Schubert and Ax 2003, Stang and Schubert 2001). For example, in a high-
pressure valve homogenizer, the energy density is equal to the operating pressure,
that is, EV = PH (Stang and Schubert 2001). Alternatively, the net energy consump-
tion can be found experimentally by measuring the increase in temperature dur-
ing homogenization (as well as the newly created droplet interface) as 99.9% of the
energy lost is lost as heat (Berg and Lundh 1978) or by monitoring the electrical
power requirements of the homogenizer (Abismaïl et al. 1999). The energy density
concept is practically useful in comparing emulsification processes and machine
designs as power input into a machine and volumetric flow rate are readily accessible
operating parameters. A comparison of various emulsification processes showing
the effect of energy density on droplet size is shown in Figure 1.7, which is a basic
illustration of the energy density concept, that given equal energy densities, different
28 Engineering Aspects of Food Emulsification and Homogenization
100
ϕ = 30%
HPH-sev
1
US
HPH-ov
MF
0.1 3
10 104 105 106 107 108
−3
Energy density, Ev (Jm )
FIGure 1.7 Comparison of various emulsifying processes based on the energy density
concept. Vegetable oil-in-water emulsions with excess surfactant. Symbols: ϕ is oil-volume
fraction, HPH refers to high-pressure homogenizers with standard valve (v), sharp-edged
valve (sev), and orifice valve (ov) designs, respectively. US refers to ultrasound; MS refers
to Microfluidizer®. (Redrawn from Schubert, H., and Ax, K., Texture in Food, Woodhead
Publishing Ltd., New York, 2003.)
emulsifying equipment produces very different droplet sizes. Most notable is the
difference between low-energy processes such as membrane emulsification and
high-energy processes such as valve homogenizers, microfluidizers, and ultrasonic
probes. Membrane emulsification has an energy density about 3 orders of magnitude
lower than that of conventional high-energy processes, with an energy efficiency
approaching the theoretical limit. Furthermore, as the majority of the energy sup-
plied to making droplet via membrane or microchannel emulsification processes
goes directly into the generation of interfacial area (rather than to viscous dissipa-
tion), the energy density is strongly correlated with the oil-volume fraction, ϕ, as seen
in Figure 1.7. In Table 1.3, some approximations of the maximum expected droplet
sizes as a function of EV are given for high-energy emulsification processes as well
as equivalent correlations for membrane and microchannel droplet formation under
STB and shear-induced droplet formation mechanisms.
Decreasing energy use and costs is always an objective, and several strategies
to improve the energy efficiency of homogenizers have been suggested, including
(McClements 2005, Walstra and Smulders 1998)
• Combining techniques that are most efficient for a given size range, for
example, a high-speed blender in combination with a high-pressure
homogenizer.
reFerences
Abismaïl, B., J.P. Canselier, A.M. Wilhelm, H. Delmas, and C. Gourdon. 1999. “Emulsification
by ultrasound: Drop size distribution and stability.” Ultrasonics Sonochemistry 6
(1–2):75–83.
Abrahamse, A.J., A. van der Padt, R.M. Boom, and W.B.C. de Heij. 2001. “Process fundamentals
of membrane emulsification: Simulation with CFD.” AIChE Journal 47 (6):1285–1291.
doi:10.1002/aic.690470606.
Berg, E.T., and G. Lundh. 1978. “Functional characterization of protein stabilized emulsions:
Standardized emulsifying procedure.” Journal of Food Science 43 (5):1553–1558. doi:
10.1111/j.1365-2621.1978.tb02541.x.
Brans, G., R.G.M. van der Sman, C.G.P.H. Schroën, A. van der Padt, and R.M. Boom. 2006.
“Optimization of the membrane and pore design for micro-machined membranes.”
Journal of Membrane Science 278 (1–2):239–250.
Charcosset, C., I. Limayem, and H. Fessi. 2004. “The membrane emulsification process—A
review.” Journal of Chemical Technology and Biotechnology 79 (3):209–218.
Doran, P.M. 1995. “7—Fluid flow and mixing.” In Bioprocess Engineering Principles, edited
by P.M. Doran, 129–163. London: Academic Press.
Freudig, B., S. Tesch, and H. Schubert. 2003. “Production of emulsions in high-pres-
sure homogenizers—Part II: Influence of cavitation on droplet breakup.” Chemical
Engineering and Technology 26 (6):266–270.
Gijsbertsen-Abrahamse, A.J., A. Van Der Padt, and R.M. Boom. 2004. “Status of cross-flow
membrane emulsification and outlook for industrial application.” Journal of Membrane
Science 230 (1–2):149–159.
Håkansson, A., C. Trägårdh, and B. Bergenståhl. 2009. “Studying the effects of adsorption,
recoalescence and fragmentation in a high pressure homogenizer using a dynamic simu-
lation model.” Food Hydrocolloids 23 (4):1177–1183.
Hinze, J.O. 1955. “Fundamentals of the hydrodynamic mechanism of splitting in dispersion
processes.” AIChE Journal 1 (3):289–295. doi:10.1002/aic.690010303.
Innings, F., E. Hultman, F. Forsberg, and B. Prakash. 2011. “Understanding and analysis of
wear in homogenizers for processing liquid food.” Wear 271 (9–10):2588–2598.
Karbstein, H., and H. Schubert. 1995. “Developments in the continuous mechanical pro-
duction of oil-in-water macro-emulsions.” Chemical Engineering and Processing 34
(3):205–211. doi:10.1016/0255-2701(94)04005-2.
Kobayashi, I., S. Mukataka, and M. Nakajima. 2004. “Effect of slot aspect ratio on droplet for-
mation from silicon straight-through microchannels.” Journal of Colloid and Interface
Science 279 (1):277–280.
Kobayashi, I., and M. Nakajima. 2002. “Effect of emulsifiers on the preparation of food-grade
oil-in-water emulsions using a straight-through extrusion filter.” European Journal of
Lipid Science and Technology 104 (11):720–727. doi:10.1002/1438-9312(200211)104:
11<720::aid-ejlt720>3.0.co;2-e.
Kobayashi, I., M. Nakajima, K. Chun, Y. Kikuchi, and H. Fukita. 2002. “Silicon array of elonga-
ted through-holes for monodisperse emulsion droplets.” AIChE Journal 48 (8):1639–1644.
doi:10.1002/aic.690480807.
Kobayashi, I., K. Uemura, and M. Nakajima. 2006. “CFD study of the effect of a fluid
flow in a channel on generation of oil-in-water emulsion droplets in straight-
through microchannel emulsification.” Journal of Chemical Engineering of Japan
39 (8):855–863.
Kobayashi, I., G.T. Vladisavljević, K. Uemura, and M. Nakajima. 2011. “CFD analysis of micro-
channel emulsification: Droplet generation process and size effect of asymmetric straight
flow-through microchannels.” Chemical Engineering Science 66 (22):5556–5565.
Scales and Forces in Emulsification 31
Kobayashi, I., M. Yasuno, S. Iwamoto, A. Shono, K. Satoh, and M. Nakajima. 2002. “Microscopic
observation of emulsion droplet formation from a polycarbonate membrane.” Colloids
and Surfaces A: Physicochemical and Engineering Aspects 207 (1–3):185–196.
Kolmogorov, A.N. 1949. “On the breakage of drops in a turbulent flow.” Mathematics and
Mechanics 1:339–343.
Lee, L.L., N. Niknafs, R.D. Hancocks, and I.T. Norton. 2013. “Emulsification: Mechanistic
understanding.” Trends in Food Science and Technology 31 (1):72–78. doi:10.1016/j.
tifs.2012.08.006.
Lee, L., and I.T. Norton. 2013. “Comparing droplet breakup for a high-pressure valve
homogeniser and a Microfluidizer for the potential production of food-grade nanoemul-
sions.” Journal of Food Engineering 114 (2):158–163.
Maan, A.A., R. Boom, and K. Schroën. 2013. “Preparation of monodispersed oil-in-water
emulsions through semi-metal microfluidic EDGE systems.” Microfluidics and
Nanofluidics 14 (5):775–784.
Matos, M., M.A. Suárez, G. Gutiérrez, J. Coca, and C. Pazos. 2013. “Emulsification with
microfiltration ceramic membranes: A different approach to droplet formation mecha-
nism.” Journal of Membrane Science 444:345–358.
McClements, D.J. 2005. Food Emulsions: Principles, Practices, and Techniques, edited by
F.M. Clydesdale, Chapter 6, pp. 233–268. CRC Press Series in Contemporary Food
Science. Boca Raton, FL: CRC Press.
McClements, D.J. 2007. “Critical review of techniques and methodologies for characterization
of emulsion stability.” Critical Reviews in Food Science and Nutrition 47 (7):611–649.
Nakashima, T., M. Shimizu, and M. Kukizaki. 1991. “Membrane emulsification by micropo-
rous glass.” Key Engineering Materials 61–62:513–516.
Nazir, A., K. Schroën, and R. Boom. 2011. “High-throughput premix membrane emulsifica-
tion using nickel sieves having straight-through pores.” Journal of Membrane Science
383 (1–2):116–123.
Peng, S.J., and R.A. Williams. 1998. “Controlled production of emulsions using a cross-
flow membrane. Part I: Droplet formation from a single pore.” Chemical Engineering
Research and Design 76 (A8):894–901. doi:10.1205/026387698525694.
Phillips, L.W. 1985. The High Pressure Dairy Homogenizer. Vol. 6, NIRD Technical Bulletins.
28pp. Reading, England: The National Institute in Dairying.
Rayner, M. 2005. “Membrane emulsification: Modelling interfacial and geometric effects
on droplet size.” PhD Dissertation, Department of Food Technology, Engineering and
Nutrition, Faculty of Engineering, Lund University, Sweden.
Rayner, M., and G. Trägårdh. 2002. “Membrane emulsification modelling: How can we
get from characterisation to design?” Desalination 145 (1–3):165–172. doi:10.1016/
s0011-9164(02)00403-4.
Rayner, M., G. Trägårdh, and C. Trägårdh. 2005. “The impact of mass transfer and interfacial
expansion rate on droplet size in membrane emulsification processes.” Colloids and
Surfaces A: Physicochemical and Engineering Aspects 266 (1–3):1–17.
Schroder, V., and H. Schubert. 1999. “Production of emulsions using microporous, ceramic
membranes.” Colloids and Surfaces A: Physicochemical and Engineering Aspects 152
(1–2):103–109. doi:10.1016/s0927-7757(98)00688-8.
Schubert, H., and K. Ax. 2003. “Engineering food emulsions.” In Texture in Food, edited by
D.M. McKenna. New York: Woodhead Publishing Ltd.
Schubert, H., K. Ax, and O. Behrend. 2003. “Product engineering of dispersed systems.” Trends
in Food Science and Technology 14 (1–2):9–16. doi:10.1016/s0924-2244(02)00245-5.
Schultz, S., G. Wagner, K. Urban, and J. Ulrich. 2004. “High-pressure homogenization as a
process for emulsion formation.” Chemical Engineering and Technology 27 (4):361–368.
doi:10.1002/ceat.200406111.
32 Engineering Aspects of Food Emulsification and Homogenization
contents
2.1 Introduction .................................................................................................... 33
2.2 Interfaces ........................................................................................................34
2.3 Stability and Instability of Droplets ............................................................... 35
2.4 Creaming/Sedimentation ................................................................................ 36
2.5 Flocculation ....................................................................................................40
2.6 Coalescence .................................................................................................... 41
2.7 Ostwald Ripening ........................................................................................... 43
2.8 Surface Interactions in Emulsion Systems .....................................................44
2.8.1 Van der Waals Interactions .................................................................44
2.8.2 Solvation Interactions .........................................................................44
2.8.3 Electrostatic Repulsion .......................................................................44
2.8.4 Polymer-Induced Interactions ............................................................. 47
2.8.5 Bridges due to a Third Phase .............................................................. 48
References ................................................................................................................ 48
2.1 IntroductIon
Emulsions are dispersions of two liquid phases within each other: a more polar phase,
usually termed water phase, as it typically is an aqueous solution, and an oil phase,
as it typically consists of a more or less nonpolar liquid with a low solubility in the
water phase. One of the phases is continuous whereas the other is dispersed. The
structures are commonly named oil in water or water in oil, depending on whether
we have oil droplets in water or water droplets in oil.
Emulsions are, by their nature, unstable and are defined as consisting of at least
two phases. Technically, we want to be able to control the stability, or more cor-
rectly, the kinetics of the destabilization, when the emulsion is formed. The emul-
sion processing and composition is aimed to create an emulsion with a suitable
stability.
33
34 Engineering Aspects of Food Emulsification and Homogenization
2.2 Interfaces
All liquids display cohesiveness through intermolecular interactions. An interface
between two phases displays the step change in composition as well as in the inten-
sity of the intermolecular interactions. An interfacial tension is obtained that reflects
the difference. The interfacial tension is the energy cost of creating the surface. The
interface between two pure liquids appears liquid.
Various molecular species present at the interface may moderate the interaction
contrast and thereby reduce the interfacial tension. The reduction of surface tension
shows the attraction of the interface toward the molecules, their surface activity. The
surface activity leads to a flow of surface active molecules toward the interface until
it becomes saturated. Molecules present at the interface generate a pressure along
the surface. If the concentration is uneven, there are pressure gradients that may
lead to flow along the surface. If the surface displays a significant surface pressure,
it appears more or less as a solid. All together, these basic features create a highly
dynamic situation when new surfaces are formed during an emulsification event.
An efficient emulsification is achieved if we have unstable interfaces and obtain
stable droplets, whereas stable interfaces and unstable droplets make emulsification
difficult.
The stability respective instability, of interfaces respectively droplets, is deter-
mined both by static forces as well as by dynamic forces, as illustrated in Figure 2.1.
Dynamic effects can be summarized on the basis of Gibbs–Marangoni effect,
whereas static effects can be summarized on the basis of interfacial tension and
surface forces (Walstra 1993, 2003). Flow across and along the surfaces contributes
strongly to the emulsification and interfacial instability if the interfacial tension is
low. The flow to newly formed interfaces is obviously essential and critical for the
emulsification efficiency (Maldonado-Valderrama et al. 2008).
Spontaneous emulsification can be obtained due to the dynamic effects in systems
with very low interfacial tension, with low viscosity, and with large flow across the
Interfaces Droplets
Diffusion to and
across interfaces. Flux Flux along
along interfaces interfaces
Dynamic
6φ
A= (2.1)
d
where:
A is interfacial area (m2/m3)
ϕ is volume fraction dispersed phase
d is droplet diameter (m)
Despite the large area, typically 600 m2/liter (in an emulsion of 10% oil and 1 µm in
droplet diameter), the total interfacial energy is just 6 J/liter (assuming an interfacial
tension of 10 mN/m)—indeed is a quite small energy. By experience, we may also
compare emulsions with comparable, large interfacial tensions with emulsions with
comparable, low interfacial tensions; it can be concluded that the interfacial tensions
do not reflect the instability, although it is a major driving force for the instability.
Instability is caused by different instability mechanisms, which describe the loss
of the dispersed state by overcoming the threshold energies that keep the emul-
sions stable. Typically, we distinguish between creaming/sedimentation, floccula-
tion, coalescence, and Ostwald ripening (Figure 2.2) (McClements 2004a, Walstra
1993, 2003).
Coalescence
0.2
ϕ
Flocculation Creaming
1 10
Diameter (μm)
fIGure 2.2 Volume fractions and particle sizes, where different instability processes are
of particular importance.
All the mechanisms occur in parallel in all emulsion systems. However, various factors
influence the rate and thereby determine the mechanisms that are most critical in a
particular system. Most decisive are particle size and volume fraction, as indicated in
Figure 2.2.
2.4 creamInG/sedImentatIon
Creaming is an obvious source of instability, as anybody who has observed the rapid
separation between vinegar and oil when a salad dressing is mixed would confirm.
However, when considering the large variability of technical emulsion, we can see
that there are systems where gravity-induced separation proceeds very readily and
Emulsion Formation and Instability 37
there are systems that remain stable for a long time. Many factors influence the rate
(McClements 2004a). Here, we will look into droplet size, concentration, rheology,
and aggregation state.
Particle size. The role of particle size is clearly expressed in the Stokes law for
settling particles.
g ⋅ d 2 ⋅ ∆ρ
vStokes = (2.2)
18 ⋅ η
where:
vStokes is the settling or creaming velocity (m/s)
g is the gravitational acceleration (m/s2)
∆ρ is the density difference (kg/m3)
η is the viscosity (Pa s)
D
vdiffusion = (2.3)
2⋅t
kB ⋅ T
D= (2.4)
3⋅ π⋅η⋅ d
where:
vdiffusion is the average Brownian velocity (it is provided as the absolute
value) (m/s)
D is the diffusion constant (m2/s)
t is the time (s)
k B is the Boltzmann’s constant (J/K)
T is the absolute temperature (K)
creaming to win over the diffusion much longer; at sizes below 0.5 µm, most
emulsions can be assumed to be stable against creaming.
Concentration. One assumption in Stokes law is that the movement of one
particle is independent of that of other particles. However, in an emulsion,
particles are not alone; with micrometer-sized droplets, the number concen-
tration is about 1014 droplets/liter. The average distance between the par-
ticles can be estimated assuming cubic close packing to
π 13
δ = d 1 − (2.5)
φ ⋅ 6
where:
δ is the distance between the emulsion droplets (surface to surface) (m)
The distance clearly scales against diameter. It also becomes quite small
when the volume fraction approaches 0.5. The interparticle distance in a
typical emulsion with a volume fraction of about 0.1 and a droplet size of
1 μm is about 1 µm. The particles in a typical emulsion are polydisperse and
different particles display different Stokes velocities.
If a particle moves with a creaming velocity of 50 nm/s, it may encounter
another particle within 20 s (if the next particle is significantly smaller).
The frequent particle encounters reduce the average creaming compared
to the prediction according to Stokes law significantly. The phenomenon
is usually referred to as hindered sedimentation. Hunter (1986) made
a prediction of the effect as a function of volume fraction, as showed in
Figure 2.3. The hindered creaming/settling explains the apparent stability
that we observe for various concentrated practical systems.
Rheology. The flow properties of the continuous phase are described by vis-
cosity in the Stokes equation. Therefore, there is an underlying assumption
that the viscosity is Newtonian. In the presence of polymers in the continu-
ous phase, this assumption may be valid (McClements 2004a). For non-
Newtonian fluids, the viscosity is an apparent property that depends on the
shear rate. We may assume that shear is developed over a distance comparable
to the radius of the particle. Thus, the shear rate in our example emulsion,
rising with the velocity of about 50 nm/s, is about 0.1/s. However, the
Brownian motion of the same particle in this timescale is about 10 times
higher and results in a shear rate of 1/s. For larger particles, the shear field
created by the sedimentation will be higher, and more important. But the
present field will never be smaller than about 1–0.1/s. In addition, we may
have non-Newtonian liquids that display yielding properties (if the mol-
ecules in the continuous phase have developed a network that demands a
certain stress to flow). In that case, the yield value needs to be stronger than
the stress generated by the rising force on the particle, which, for our exam-
ple particle, is about 1 mPa. For larger particles, the stress by the rising
movements increases approximately proportionally to the diameter. Thus,
Emulsion Formation and Instability 39
v/vStokes
0.5
fIGure 2.3 The creaming rate of emulsions as a function of volume fraction dispersed
phase. The figure is based on the semiempirical equation by Hunter using the constants
by Chanamai and McClements. (Data from Hunter, R.J., Foundations of Colloid Science,
Oxford University Press, Oxford, 1986; Chanamai, R. and McClements, D.J., Colloids and
Surfaces A: Physicochemical and Engineering Aspects 172, 79–86, 2000.)
Interaction
Repulsive
0.1 μm 10 μm colloidal
Repulsive glass
Colloidal
liquid
Colloidal
liquid
Attractive Attractive
colloidal
glass
Volume fraction
fIGure 2.4 Principal phase diagram showing when colloidal liquids and colloidal glasses
are formed as a function of volume fraction dispersed phase and repulsive attractive interac-
tions, respectively. The effects of particle size are outlined.
2.5 flocculatIon
Flocculation is caused by collision events between moving droplets. The collision
may be caused by different sources of droplet movement: Brownian, shear-, and
gravity-induced collisions. The different collision mechanisms are compared in
Figure 2.5.
Time
10 years
1 year
Gravity
1 month
1 week
1 day
Brownian
1 hour
1 minute
Shear
1s
0.1 0.2 0.5 1 2 5 10 20 50 100
Droplet size in μm
fIGure 2.5 The kinetics of different aggregation mechanisms in emulsions. The diagram
shows the timescale of significant instability (in terms of the timescale for a doubling of
the diameter when assuming immediate droplet fusion after a collision event) as a function
of particle size (droplet diameter). The timescale of instability is shown for Brownian floc-
culation, shear-induced flocculation, and gravity-induced flocculation. Assumptions: Every
encounter leads to coalescence (w = 1), volume fraction 0.1, aqueous environment at 25°C,
10% density difference, a shear field of 10 (s−1).
strongly on the particle size and becomes more rapid as the size increases.
The kinetics of the destabilization process is illustrated in Figure 2.5, similar
to the processes illustrated for Brownian flocculation. It is clear that shear-
induced aggregation is different from Brownian and may lead to the forma-
tion of very large objects if emulsion starts to destabilize.
Gravity-induced aggregation. The creaming-induced movements of droplets
may also lead to collisions (McClements 2004a, Melik and Fogler 1988).
For small droplets, such as the micrometer-sized droplet in our example,
it may take about 20 s to reach the next droplet, whereas the Brownian
flocculation is much faster. However, when the particles are larger, gravity-
induced aggregation becomes more important. For 10 µm-sized particles,
the creaming time to reach an encounter is lower than the time for Brownian
motions leading to the same event.
The outcome of collision events depends on the strength of the collision
energy relative to the interactive forces between the particles. Thus, we will
discuss interparticle interactions between emulsion droplets in Section 2.8.
2.6 coalescence
Coalescence is the process when two separate droplets fuse to form one new large
droplet. This process can be described as being dependent on the stability of the film
separating two adjacent droplets (Kabalnov 1998, McClements 2004a, Walstra 1993).
42 Engineering Aspects of Food Emulsification and Homogenization
Droplet size. Smaller droplets allow for a larger contact surface when counted
on the complete emulsion (Walstra 1993). If we expect that the probability
for coalescence to be proportional to random disturbances, it may be pro-
portional to the contact area, thereby becoming a more important desta-
bilizing mechanism for the finer dispersed emulsions. On the other hand,
smaller droplets have a higher internal pressure that may stabilize them
against disturbances.
Surface interaction. When droplets have aggregated together, the film separat-
ing them may drain. The drainage proceeds until an equilibrium thickness
has been obtained. In the Scheludko cells, a thickness of the semistable
film could be in the range 25–200 nm (Chan et al. 2011, Scheludko 1967).
The equilibrium thickness may correspond to the distance when we have
a balance between attractive and repulsive surface interactions. However,
this distance is in most systems quite short, and the actual films tend to be
thicker as a consequence of slow drainage.
Interfacial tension. A low interfacial tension allows disturbances to create
large protrusions or dents at the interface, which may lead to the formation
of holes or bridges, thereby causing a sudden collapse of the films (Chan
et al. 2011). A higher surface tension reduces disturbances and is therefore
expected to lead to more stable films. This consequence of the interfacial
tension is interesting as a very counter intuitive aspect of the emulsion
technology.
Interfacial viscosity. The molecules or particles forming an interfacial layer
usually appear more or less like a solid structure at the interface. This
solid-like character slows down the drainage, thereby ensuring slow drain-
age. A dense layer also creates surface pressure. The surface pressure will
rapidly heal sudden disturbances that may lead to the rupture of the film.
The flow of emulsifier contributes to the stabilizing action by dragging the
liquid into the film when moving along the surface. The relation between
surface rheological parameters and emulsion stability has recently been
reviewed by Pelipenko et al. (2012).
Emulsion Formation and Instability 43
4⋅γ
Π Laplace = (2.6)
d
where:
γ is the interfacial tension between oil and water
The Laplace pressure of our example emulsion is about 0.4 bar, which is a significant
pressure. The pressure leads to a dissolving force when acting on the molecules in
the internal phase. The energy released due to the pressure is about 4 J/mol that may
increase the solubility about 0.2%. The flux from drop to drop depends on the solu-
bility, the distance between the droplets (the flux over a distance is proportional to
1/d), the Laplace pressure (the effect is proportional to 1/d), and the total area (pro-
portional to 1/d) (Kabalnov and Shchukin 1992). Thus, the rate of the process will
be proportional to 1/d3 and therefore strongly depend on the size. Very fine emulsion
may be coarsened rapidly, although the process decay when the particle size reach
somewhere around 0.5–5 µm. The process depends on solubility of the dispersed
phase. Typically, low molecular oils (hydrocarbons with a molar mass below, maybe,
200 g/mol) are exposed to this process. The presence of a high-molecular component
in the system may delay the process and therefore stabilize the emulsion (Welin-
Berger and Bergenstahl 2000).
44 Engineering Aspects of Food Emulsification and Homogenization
The counterions have a strong affinity, electrostatic as well as by Van der Waals
forces, to such a charged surface and may adsorb into a tight layer. This first tight
layer is termed the Stern layer. A majority of the charges are compensated already
in the Stern layer.
Assuming that 90% of the charges are compensated within the Stern layer, we
obtain a potential at the Stern layer of about 70 mV.
Outside the Stern layer, the accumulation of counterions only depends on the bal-
ance between electrostatic interaction and the thermal diffusion. The charge of the
surface decays by the counterions layer.
To obtain the interaction strength, the surface charge is expressed as surface
potential (ς). The relation between surface charge and surface potential at the Stern
layer (e.g., the outer Helmholtz plane, approximately equal to the experimental zeta
potential) is given by the Grahame equation, which can be approximated to:
σ
ς= (2.7)
ε r ⋅ ε0 ⋅ κ
where:
σ is the surface charge density (C/m2)
εr and ε0 are the relative and vacuum dielectric constants, respectively (C/V m)
κ is the reciprocal Debye layer thickness
The key parameter here is the Debye layer. The attraction is counteracted by entropic
mobility of the ions and the result is a diffused layer of ions, usually termed the dou-
ble layer, schematically illustrated in Figure 2.6. The concentration of counterions
− +
−+ +
−+ + + −
−
− + + + +
−
− + + +
− + −
−
Cx
ψx
Stern layer
fIGure 2.6 The counterion accumulation at a charge surface and the counterion concentration
(Cx) and surface potential (ψx) as a function of the distance (x) (counted from the Stern layer).
46 Engineering Aspects of Food Emulsification and Homogenization
in the double layer decays by increasing the distance from the surface with the decay
constant κ −1, the Debye length. The Debye length is obtained from the counterion
concentration and valence:
ε r ⋅ ε0 ⋅ RT
λ D = κ −1 = (2.8)
2F2 ∑ ci ⋅ zi2
where:
R is the universal gas constant (J/mol K)
F is Faraday constant (C/mol)
ci is the concentration of counterion i (mol/m3)
zi is the valence of counterion i
The Debye length varies from about 0.7 nm for physiological salt solution (0.9%
NaCl) to 5 nm in quite soft water-like normal tap water (0.15 ppm, 3°dH or 2° Clark).
The potential as well as the counterion concentration scale according to the Debye
length are now given:
ς x = ς 0 ⋅ e − ( x / λ0 ) (2.9)
When two planar surfaces with double layers approach each other, the double layers
of both the surfaces overlap and an osmotic repulsion is created due to the excess
concentration in the overlap region. A repulsive pressure is generated (van’t Hoff
equation):
P(h) = RT ⋅
∑c (h) − ∑c (∞)
i i (2.11)
The following expression for the repulsive pressure is obtained as the surplus con-
centration is a function of the Debye layer and the surface charge:
1 2 − ( h / λD )
P(h) = 2ε0ε r ⋅ ⋅ ς0 ⋅ e (2.12)
λ D2
By comparing the repulsive electrostatic force with the attractive van der Waals
force, assuming additivity, Dejaugin, Landau, Vervey, and Overbeek in parallel were
able to create a theory (the DLVO theory) with quantitatively predictive abilities
(Derjaguin and Landau 1941, Verwey and Overbeek 1948).
The outcome of the theory was that at high charge and low ionic strength, the repul-
sion dominates and the system remains stable. At high ionic strength, the attraction
dominates and the system becomes destabilized. At low charge and/or, at intermedi-
ate ionic strength, a repulsive barrier is created that provides an activation energy that
aggregating particles need to overcome. This activation energy leads to slower aggrega-
tion rate as only a fraction of the collisions results in aggregation. The stability factor is
Emulsion Formation and Instability 47
the ratio between slow aggregation and fast aggregation (the aggregation rate obtained
when every collision is assumed to lead to aggregation). A typical stability factor can
be 106 or comparable numbers depending on surface charge and ionic strength.
table 2.1
surface Interactions caused by the presence of polymers
type of Interaction polymer surface Interaction other conditions
Bridging Adsorbing Low surface coverage
Steric repulsion Adsorbing High surface coverage and good solvent
conditions
Depletion attraction Nonadsorbing Intermediate polymer concentration
48 Engineering Aspects of Food Emulsification and Homogenization
references
Bancroft, W.D. 1913. The theory of emulsification. V. Journal of Physical Chemistry 17,
501–519.
Bergenstahl, B. and Claesson, P. 1997. Surface forces in food emulsions, in Food Emulsions.
Friberg S. and Larsson K. (eds.) Marcel Dekker, New York, 57–110.
Bibette, J., Roux, D., and Nallet, F. 1990. Depletion interactions and fluid-solid equilibrium in
emulsions. Physical Review Letters 65, 2470.
Boode, K. and Walstra, P. 1993. Partial coalescence in oil water emulsions. 1. Nature of the aggre-
gation. Colloids and Surfaces A: Physicochemical and Engineering Aspects 81, 121–137.
Butt, H.J. and Kappl, M. 2009. Normal capillary forces. Advances in Colloid and Interface
Science 146, 48–60.
Chan, D., Klaseboer, E., and Manica, R. 2011. Film drainage and coalescence between
deformable drops and bubbles. Soft Matter 7, 2235–2264.
Chanamai, R. and McClements, D.J. 2000. Dependence of creaming and rheology of monodisperse
oil-in-water emulsions on droplet size and concentration. Colloids and Surfaces A:
Physicochemical and Engineering Aspects 172, 79–86.
Emulsion Formation and Instability 49
Claesson, P.M., Blomberg, E., and Poptoshev, E. 2001. Surface forces and emulsion stability,
in Encyclopedic Handbook of Emulsion Technology. Sjoblom J. (ed.) Marcel Dekker,
New York, Chapter 13, pp. 305–327.
Davies, J.T. 1957. A quantitative kinetic theory of emulsion type, I. Physical chemistry of
the emulsifying agent, in Proceedings of the 2nd Congress on Surface Activity, Vol. 1,
Butterworths, London, 426–438.
Dawson, K.E. 2002. The glass paradigm for colloidal glasses, gels, and other arrested states driven
by attractive interactions. Current Opinion in Colloid and Interface Science 7, 218–227.
Derjaguin, B. and Landau, L. 1941. Theory of the stability of strongly charged lyophobic sols and
of adhesion of strongly charged particles in solutions of electrolytes. Acta Physicochimica
URSS 14, 663; Reprinted in: Progress in Surface Science 1993, 43, 30–59.
Fredrick, E., Walstra, P., and Dewettinck, K. 2010. Factors governing partial coalescence in
oil-in-water emulsions. Advances in Colloid and Interface Science 153, 30–42.
Hunter, R.J. 1986. Foundations of Colloid Science, Vol. 1, Oxford University Press, Oxford.
Kabalnov, A.S. 1998. Coalescence in emulsions, in Modern Aspects of Emulsion Science.
Binks B.P. (ed.). The Royal Society of Chemistry, Cambridge, Chapter 7, pp. 205–257.
Kabalnov, A.S. and Shchukin, E.D. 1992. Ostwald ripening theory: Applications to fluorocar-
bon emulsion stability. Advances in Colloid and Interface Science 38, 69.
Lopez-Montilla, J.C., Herrera-Morales, P.E., Pandey, S., and Shah, D.O. 2002. Spontaneous
emulsification: Mechanisms, physiochemical aspects, modelling and applications.
Journal of Dispersion Science and Technology 23, 219–268.
Lyklema, H. 2000. Electric double layer, in Fundamentals of Interface and Colloid Science,
Vol. 2, Liquid-Solid Interfaces. Academic Press, London, Chapter 3, pp. 3.1–3.232.
Maldonado-Valderrama, J., Martin-Rodriguez, A., Gálvez-Ruiz, M.J., Miller, R., Langevin, D., and
Cabrerizo-Vilchez, M.A. 2008. Foams and emulsions of β-casein examined by interfacial
rheology. Colloids and Surfaces A: Physicochemical and Engineering Aspects 323, 116–122.
McClements, D.J., 2000. Comments on viscosity enhancement and depletion flocculation by
polysaccharides. Food Hydrocolloids 14, 173.
McClements, D.J. (ed.) 2004a. Emulsion stability, in Food Emulsions: Principles, Practices,
and Techniques, 2nd ed., CRC Press, Boca Raton, FL, Chapter 7, pp. 267–339.
McClements, D.J. (ed.) 2004b. Emulsion stability, in Food Emulsions: Principles, Practices,
and Techniques, 2nd ed., CRC Press, Boca Raton, FL, Chapter 3, pp. 33–93.
Melik, D.H. and Fogler, H.S. 1988. Fundamentals of colloidal stability in quiescent media.
in Encyclopedia of Emulsion Technology, Vol. 3, Becher P. (ed.), Marcel Dekker,
New York, Chapter 1, pp. 3–78.
Pelipenko, J., Kristl, J., Rosik, R., Baumgartner, S., and Kocbek, P. 2012. Interfacial rheology:
An overview of measuring techniques and its role in dispersions and electrospinning.
Acta Pharmaceutica 62, 123–140.
Richardsson, G., Bergenstahl, B., Langton, M., Stading, M., and Hermansson, A.M. 2004.
The function of alpha-crystalline emulsifiers on expanding foam surfaces. Food Hydro-
colloids 18, 655–663.
Scheludko, A. 1967. Thin liquid films. Advances in Colloid and Interface Science 1, 391.
Walstra, P. 1993. Principles of emulsion formation. Chemical Engineering Science 48, 333–349.
Walstra, P. (ed.) 2003. Changes in dispersity, in Physical Chemistry of Foods. Marcel Dekker,
New York, Chapter 13, pp. 476–547.
Welin-Berger, K. and Bergenstahl, B. 2000. Inhibition of Ostwald ripening in local anesthetic
emulsions by using hydrophobic excipients in the disperse phase. International Journal
of Pharmaceutics 200, 249–260.
Verwey, E.J.W. and Overbeek, J.Th.G. 1948. Theory of the Stability of Lyophobic Colloids.
Elsevier, Amsterdam, the Netherlands.
3 Formulation of Emulsions
Marie Wahlgren, Björn Bergenståhl,
Lars Nilsson, and Marilyn Rayner
Contents
3.1 Introduction .................................................................................................... 52
3.2 Functionality that Ingredients Should Give to Emulsions .............................. 52
3.2.1 Nutrition and Health ........................................................................... 52
3.2.2 Texture and Flavor .............................................................................. 53
3.2.3 Shelf-Life Stability ............................................................................. 54
3.2.3.1 Emulsion Stability................................................................ 54
3.2.3.2 Chemical Stability ............................................................... 56
3.2.3.3 Microbiological Stability ..................................................... 57
3.2.3.4 Freeze–Thaw Stability ......................................................... 58
3.3 Issues to Consider When Choosing Ingredients for Emulsions...................... 59
3.4 Key Ingredients in Emulsions......................................................................... 62
3.4.1 Fats and Oils ....................................................................................... 62
3.4.2 Low Molar Mass Emulsifiers..............................................................64
3.4.3 Proteins ...............................................................................................66
3.4.4 Polysaccharides................................................................................... 70
3.4.5 Protein–Polysaccharide Complexes.................................................... 74
3.4.6 Particles .............................................................................................. 74
3.5 Evaluation of Emulsion Formulation and Ingredient Performance ................ 77
3.5.1 Emulsification Capacity ......................................................................80
3.5.2 Emulsion Stability Index .................................................................... 82
3.5.3 Assessing Gravitational Separation—Creaming Index...................... 82
3.5.4 Accelerated and Environmental Stress Tests ...................................... 86
3.5.5 Evaluation of Texture .......................................................................... 88
References ................................................................................................................90
ABSTRACT In this chapter, we describe some of the main concerns when it comes
to formulating emulsions. This includes the choice of ingredients, such as emulsifiers,
oils, preservatives, and thickeners. This is done with a focus on how these ingredients
can give the desired properties of the emulsions, such as texture, flavor, nutrition,
and stability. Commonly encountered thickeners and emulsifiers are described,
and the methods to characterize the key properties of emulsion and ingredient are
discussed.
51
52 Engineering Aspects of Food Emulsification and Homogenization
3.1 IntroduCtIon
Almost all industrially processed emulsion-based food products are made up of a
wide variety of constituents, including fats and oils, emulsifiers, texture modifiers,
preservatives, antimicrobial agents, antioxidants, pH adjusters, sweeteners, salts,
coloring agents, flavors, and, of course, water. Each of these has been included in the
food product due to its intrinsic function or a combination of functions with other
compounds in the formulation. They are there to provide the overall quality of food
products such as nutritional value, flavor, texture, and shelf life. In this chapter, we
will discuss how the ingredients deliver these quality attributes to emulsions, and
we will also give a more general description of some of the key ingredients in emul-
sions, primarily oils, emulsifiers, and texture modifiers.
Food ingredients can be described on several levels:
or fat crystals. Large particles will give a sandy mouthfeel usually described as
tallowness (Watanabe et al. 1992).
When it comes to flavor, the release of flavoring components from the dispersed
phase is important. The release will be affected by how these molecules are trans-
ported out of the dispersed phase and thus by properties such as diffusion coefficient
of the component, droplet size of the dispersed phase, and interaction with other
ingredients in the emulsions (such as the emulsifier). The release will also be influ-
enced by partitioning of the flavoring ingredient into two phases and thus be affected
by the concentration of the dispersed phase. This is especially important for O/W
emulsions, as it is the concentration of aroma in the water phase and the head space
(gas phase above the emulsion) that influences its taste. Low-fat products can show a
burst of flavor due to the quick release of the oil-soluble components, whereas high-
fat products often display a more continuous release of components that partition
to the oil phase (Bayarri, Taylor, and Hort 2006). When designing and producing
low-fat products, the release profile of the oil-soluble components may have to be
modulated, for example, by encapsulation.
It has also been seen that in systems that have the same release of aroma com-
ponents into the gas phase, changes in the rheology of the emulsion still can affect
taste. This could be attributed to the difference in the release pattern between volatile
aroma compounds and more water-soluble taste compounds such as sugar (Bayarri
et al. 2006), where the latter is more sensitive to the rheology. When it comes to the
water-soluble components, they will predominately be in the water phase, and thus
O/W emulsions will have a quick influence on the taste. However, if taste masking
is desired, the water-soluble components can sometimes be encapsulated in double
emulsions.
the process of coalescence. Increased viscosity of the continuous phase can also
decrease coalescence (as well as the rate of creaming/sedimentation) to some extent.
Another mechanism that drives the evolution of droplet size is caused by the pres-
sure difference between the inside and outside of a curved surface. This so-called
Laplace pressure is higher for a more curved surface, for example, small droplets;
this is the driving force Ostwald ripening, which leads to an increase in particle
size of the emulsions at the expense of smaller droplets. In this case, the solubility
of the dispersed phase in the continuous phase is of major importance; that is, a low
solubility slows down or prevents Ostwald ripening. Hence, Ostwald ripening is
typically not observed in triglyceride O/W emulsions but, for instance, can occur
for more soluble oils such as aromatic and essential oils. Ostwald ripening can also
be decreased by increasing the viscosity in the continuous phase (decreases diffu-
sion) and systems with low curvature. Pickering emulsions, for example, have been
suggested to decrease Ostwald ripening, as they might have a local zero curvature
(Tcholakova, Denkov, and Lips 2008).
Flocculation is the aggregation of droplets. Flocculated systems may have desired
properties for formulation such as beneficial rheology, but extensive flocculation
might lead to increased creaming and thus may lead to coalescence. Changes in the
degree of flocculation can also affect the rheology of the emulsions, changing prop-
erties such as mouthfeel.
The colloidal stability of the emulsion will be governed by the repulsive/attractive
forces between individual droplets of dispersed phase, the energy and rate of droplet
collisions, the viscoelastic properties of the interface between oil and water, and the
solubility of the dispersed phase in the continuous one. The choice of an emulsifier
could influence all of these, and a proper choice of viscosity modifier will influence
all kinetic factors such as collision of droplets and diffusion of dissolved molecules.
The most important repulsive and attractive forces between emulsions droplets
are summarized as follows:
Hydrophobic effect. This is the main reason for the instability of emulsions.
The hydrophobic interaction is based on the exclusion of nonpolar compo-
nents from water.
van der Waals attraction. These forces exist in all systems. Between small mol-
ecules, van der Walls forces are of short range and decay with increasing dis-
tance between the molecules proportional to the distance raised to the power
of minus six. However, in a colloidal system, they can be of a much more
long range, decaying with the reciprocal of distance. Together with the elec-
trostatic forces, it is the basis for the Derjaguin–Landau–Verwey–Overbeek
(DLVO) theory (Verwey and Overbeek 1948).
Electrostatic repulsion. This can be an important stabilizing force for food
emulsions. Both proteins and ionic emulsifiers can be charged, depend-
ing on the pH, and when adsorbed, at the droplet interface giving rise
to electrostatic repulsion. Emulsions stabilized by electrostatic repulsions
are sensitive to salt and, in many cases, sensitive to pH. This sensitivity
toward salt is due to the decay in the range of the electrostatic repulsion
56 Engineering Aspects of Food Emulsification and Homogenization
caused by the presence of ions and the effect strongly increases with the
valance of the ions. Thus, it is the ionic strength that is the key issue when
it comes to stability in emulsions based on ionic emulsifiers. One should
be aware that the ionic strength of buffers changes with pH. In some cases,
there could also be specific ion interactions; for example, an interaction
between calcium ions and casein that leads to aggregation (Dickinson and
Davies 1999).
Depletion attraction and steric repulsion. These interactions are caused by
the presence of macromolecules in the continuous phase. Depletion attrac-
tion is due to the fact that macromolecules (proteins, polymers, and col-
loidal particles) having no affinity toward the interface will be excluded
in the space between two approaching emulsion drops; this will lead to
an osmotic pressure gradient, which then favors aggregation. Depletion
attraction is typically observed in emulsions containing dissolved neutral
polysaccharides. Thus, the addition of polysaccharides to alter the rheology
or to form complexes with emulsifying agents may lead to depletion aggre-
gation (Magnusson and Nilsson 2011). Steric repulsion is induced by mac-
romolecules adsorbed at the interface; this is mainly due to the excluded
volume effect, as adsorbed molecules come close together (Israelachvili
1985). Steric repulsion thus requires not only the affinity of the macromol-
ecule to the interface but also a high solubility of the macromolecule in the
continuous phase. The latter allows parts of the adsorbed macromolecule
to protrude into the continuous phase, giving rise to steric hindrance. Both
nonionic low molecular emulsifiers and polymers might stabilize the emul-
sion through steric repulsion. These systems are less sensitive to salt and pH
than electrostatic stabilized emulsions.
one in an emulsion. This is because the water phase in the emulsion may include
oxidative agents such as transition metals and iron and ingredients such as eth-
ylenediaminetetraacetic acid (EDTA) and iron-binding proteins (e.g., lactoferrin)
that may decrease oxidation in emulsions (Waraho, McClements, and Decker 2011).
Phenolic compounds have been seen to be pro-oxidatives, especially in the presence
of iron (Medina et al. 2012, Sørensen et al. 2008).
Antioxidants can, as mentioned, be added to the formulation, and the activity of
these antioxidants will depend on their location in the emulsion and solution condi-
tions such as pH. It has been shown that nonpolar antioxidants are more effective
in emulsions as compared to nonpolar oxidants, which are more effective in bulk
oils (Frankel 1998). This is called the polar paradox and is probably related to the
fact that the antioxidant has to be close to the lipids that it should protect. There is
a growing interest to use naturally occurring phenolic compounds such as caffeine,
coumaric acid, and rutin as antioxidants (Kikuzaki et al. 2002, Medina et al. 2012,
Sørensen et al. 2008). These compounds have, however, also been seen to, at some
conditions, be pro-oxidative (Sørensen et al. 2008). This phenomena of having both
anti-oxidative and pro-oxidative characteristics depending on the formulation further
highlights the importance to know the function of the specific additive at the condi-
tions used for each food product. Another problem with several phenolic compounds
is their low solubility (Löf, Schillén, and Nilsson 2011) and, in these cases, their
existence as dispersed particles in the continuous phase, which, of course, reduces
their antioxidative capacity.
in between droplets; however, as pointed out by others, W/O emulsions such as mar-
garine and spreads also need to show how microbiological safety is obtained during
the shelf life (Charteris 1996, Delamarre and Batt 1999). For these products, spoilage
is often due to moulds and can be reduced by the addition of preservatives such as
sorbates and benzoates (Delamarre and Batt 1999).
low-molecular emulsifiers lose their solubility below the so-called Kraft point and
thus the function of these emulsifiers will decrease.
taBle 3.1
Comparison of Functional Characteristics and Formulation attributes of various general Classes of emulsifiers
used in Food emulsions
general Class small molecular weight surfactants macromolecules Particles
Approx. size ~0.4 to 1 nm 2–200 nm 10 nm to 10 µm
Surface active Yes Yes Yes—via partial dual wettability
Amphiphilic Yes (head and tail) Yes (hydrophobic and hydrophilic No (unless Janus particles)
regions)
Adsorption kinetics Fast in dynamic equilibrium Medium partially irreversibly Slow but essentially irreversible
Desorption energy Low <10 kT High and increasing if Exceptionally high greater than
conformational changes occurs at several thousand to tens of
the interface, several thousand kT million kT, depending on particle
size and contact angle
Chemical types Nonionic HLB 12–16 Nonionic HLB Ionic Proteins Polysaccharides Colloidal solids Colloidal
7–10 θ < 90° solids θ > 90°
Food examples Polysorbates Monoglycerides Phospholipids Caseinates egg Modified starches, Modified starch Fat crystals
proteins celluloses in and cellulose
solution crystals/particles
Solubility/ Water Oil Water Water Water Water Oil
dispersibility
Emulsion type O/W W/O O/W O/W O/W O/W W/O
Usage level (g/goil) ~0.05 ~0.05 ~0.05 ~0.05 ~1 to 1.5 ~0.02 to 1 ~0.02 to 1
pH stability Good Good Depends on pKa Poor Good Variable Good
Salt stability Good Good Poor at I > CFC Poor at I > CFC Good Variable Good
Source: McClements, D.J. “Emulsion ingredients.” In Food Emulsions: Principles, Practices, and Techniques. CRC Press, Boca Raton, FL, 2005.
CFC, critical flocculation concentration; I, ionic strength.
Engineering Aspects of Food Emulsification and Homogenization
Formulation of Emulsions 61
Water OH
Oil
HO Monoglyceride
O O
Lecithin
O
O
O O
O
O O P O
O O
wO
O OH
O x
HO OH
O O
z y
Water Water
Macromolecules θ Particle
FIgure 3.1 Schematic structure of the various general classes of emulsifiers at the oil–
water interface: (a) nonionic small molecular weight emulsifiers; (b) ionic small molecular
weight emulsifiers; (c) amphiphilic macromolecules with hydrophobic and hydrophilic regions
such as proteins and modified starches; (d) particles such as starch granules, fat crystals, and
whey protein microgels. Relative sizes and other attributes are provided in Table 3.1.
has been shown to decrease its molecular weight during homogenization (Modig
et al. 2006, Nilsson, Leeman et al. 2007). Soy proteins have shown disruption as
well as aggregation induced by high-pressure homogenization (Roesch and Corredig
2003). Several issues such as heat, shearing, and the adsorption into surfaces of the
equipment can affect the ingredients in the emulsion. Another issue could be that
the surface-active components might induce the leakage of components from gas-
kets and other plastic and rubber parts of the equipment. Thus, an incompatibility
between the process and the ingredients used has to be considered during the devel-
opment process.
In food products, a further complication is that many of the ingredients used are
often very complex mixtures, for example, mixture of proteins, polar lipids, and
polysaccharides. One good example here is egg yolk, which is used to stabilize
mayonnaise-type emulsions. In such cases, it may be difficult to know which ingredient
actually contributes to the emulsification and stabilization actions; to complicate things
further, the components at the interface can vary with solution properties such as pH
(Magnusson and Nilsson 2013, Nilsson et al. 2006, Nilsson, Osmark et al. 2007). The
interaction between ingredient components may enhance the stability of emulsions as
well as cause instability, and the effect of adding individual ingredients can be difficult
to predict.
62 Engineering Aspects of Food Emulsification and Homogenization
Finally, one should be aware of the variation and inhomogeneity of the ingredi-
ents and at least have some knowledge if this might affect batch-to-batch variation
of products. Several commercial emulsifiers are mixtures that show a variation in
chain length of the hydrophobic tail (cf. sorbitan esters and ethoxylated sorbitan
esters) or variation in molecular weight (cf. all polymers) and even variation in the
composition, for example, lecithin and whey proteins. These variations can affect
the composition of the molecules at the interface and could influence issues such as
shelf-life stability, rheology, droplet size, and so on.
α β′ β
FIgure 3.2 (a–c) Schematic description of crystalline structures for triglycerides with
unsaturated fatty acid side chains.
taBle 3.2
Common Food oils/lipids
Production vitamin e saturated lipids monounsaturated Polyunsaturated
name [million tons (may 2014)]a (mg/100g)b (g/100g)b (g/100g)b (g/100g)b melting Pointc
Formulation of Emulsions
Sources: a Agriculture, U.S.D.O., World Production, Markets, and Trade Reports: Oilseeds, United States Department of Agriculture, Washington, DC, 2014.
b United States Department of Agriculture—Agricultural Research Service, http://ndb.nal.usda.gov/ndb/foods.
oil will crystallize and the type of polymorph that is formed during crystallization.
Triglycerides exist in several polymorphic forms, where β is the most stable one. Due
to the differences in treatment, for example, cooling rate, the crystals can be locked
into other less stable polymorphic forms, where α and β′ are the most common ones
for triglycerides; this is considered in several textbooks and in a recent review by Sato
et al. (2013). As discussed previously, β′ has more appealing organoleptic properties
than β. The different polymorphic forms also have different melting points, which
can be important for the stability of the emulsions. As a rule of thumb, the less stable
polymorphic forms, the lower the melting temperature. In more complex systems,
the crystallization will be dependent on the mixture of triglycerides and the addition
of other components. One example is the addition of diacylglycerol to blends of palm
super olein to increase the onset temperature for crystallization (Ng et al. 2014).
Apart from triglycerides, most oils also contain traces of numerous other com-
ponents such as diacylglycerol, monoacylglycerols, free fatty acids, phospholipids,
tocopherols, and minerals. As reviewed by Chen, McClements, and Decker (2011),
these trace compounds can affect the stability of the oil when it comes to oxidation
but also when it comes to emulsion stability. For example, free fatty acids, mono- and
diglycerides as well as phospholipids are surface active and can cause foaming upon
mixing, or destabilization of protein-stabilized emulsions. However, the effect on
lipid oxidation of these compounds, as described by Chen, McClements, and Decker
(2011), is not straightforward, although free fatty acids have been seen to accelerate
the oxidation of triacylglycerols. Tocopherols, on the other hand, have a beneficial
effect on reducing the oxidation as they work as antioxidants.
FIgure 3.3 Schematic description of some self-assembled lipid structures and an explana-
tion of packing parameter: (a) CPP < 1/3, (b) CPP > 1/2 and < 2, (c) CPP > 2.
Friberg and Wilton (1970) suggested that the presence of lamellar liquid crystalline
phases is a strong indication of a good emulsifier in simple systems.
The functionality of low-molecular emulsifiers is, in a wide interpretation, deter-
mined by their solution properties. Although the character of low-molecular emulsi-
fiers is such that they contain regions that are water soluble and those that are more
lipophillic, their overall character can make them more soluble in one of the two
phases. Thus, emulsifiers can be found in a range from highly soluble in the oil to
more soluble in the water phase. The effect of solubility on emulsion character was
first expressed in the Bancroft (1913, p. 501) rule, stating that “hydrophilic colloid
will tend to make water the dispersing phase while a hydrophobic colloid will tend
to make water the disperse phase.” To describe the degree of hydrophilicity con-
tra lipophilicity, it is very popular to use the hydrophilic–lipophilic balance system
according to Griffin (1954). The HLB number is expressed as a number based on the
molecular weight of hydrophobic components compared to the molecular weight of
the molecule. The HLB number can also be estimated from the chemical structure
according to molecular group contributions as stated by Davies (1957):
66 Engineering Aspects of Food Emulsification and Homogenization
taBle 3.3
hlB values as Predictor for the use of emulsifiers
hlB value applications example of emulsifiersa
3.5–6 W/O emulsifier Glycerol monostearate
7–9 Wetting agent Sorbitan monolaurate
8–18 O/W emulsifier Tween 80
13–15 Detergent Tween 81
15–18 Solubilization Sodium Oleate
The HLB value of an emulsifier is often used as a rule of thumb (see Table 3.3). However,
one should be aware of the fact that solution conditions might change the HLB balance
of a system; for example, the addition of salt can screen charge groups, making the
system appear less hydrophilic. Another factor of importance is the temperature. The
effective HLB value is strongly temperature dependent. For ethoxylated emulsifiers, the
emulsifier gets less hydrophilic with increasing temperature and finally becomes insoluble
in water at a temperature denoted as the cloud point. In an emulsion system, this can be
followed by the phase inversion temperature (PIT), which corresponds to the temperature
at which the effective HLB is about 6 (Shinoda and Sato 1969). Emulsions stored at a tem-
perature of 25°C–60°C below the PIT are usually more stable. However, in food applica-
tions, this is rarely used, as ethoxylated surfactants are uncommon for food applications.
Another important temperature to consider for emulsifiers is the Krafft point (Krafft and
Wiglow 1895). The Krafft point is the temperature below which the surfactant has low
solubility and, hence, cannot form micelles. Technical functionality (such as foaming
and emulsifying action) is only obtained above the Krafft temperature. High-melting fat
bases (fully hardened C18-dominated fats) or long paraffinic chains creates high-melting
emulsifiers with Krafft temperatures in the range of 40°C–60°C. Precipitating emulsifiers
may contribute to fat crystallization and solid emulsifier may have a textural functional-
ity; however, for most applications, such high melting points are unsuitable. Intermediate
melting fat bases (C14–C18 fats with some unsaturation) give emulsifiers with Krafft or
transition temperatures between 30°C and 50°C. These emulsifiers could be used to cre-
ate stable α-gels and usually display well-performing properties in baking applications.
Low-melting fat (highly unsaturated fat), branched hydrocarbons and inclusion of aro-
matic groups, gives low Krafft points, sometimes below 0°C. Table 3.4 summarizes some
examples and usages of common low-molecular weight emulsifiers.
3.4.3 ProteiNS
Proteins function both as emulsifiers and as rheological modifiers in the formula-
tion of food emulsions. The character of proteins in emulsions will be based on their
taBle 3.4
Common low-molecular weight emulsifiers
number
name eu/usa solubility uses Comment
Lecithin E322/184.1400 Dispersible but insoluble in water, Margarine, chocolate, breads and cakes, bubble gum, salad dressings, Mixture of phosphoric acid,
where it swells on hydration. and sauces choline, fatty acids, glycerol,
Soluble in oils and fats. glycolipids, triglycerides, and
phospholipids
Formulation of Emulsions
Fatty acid salts E470/172.863 Sodium and potassium salts are Baked goods (e.g., bread and cakes), confectionery, dairy products, Charged at normal and low pH
soluble in water. Calcium salts are margarines, spreads, shortenings, salad dressings, and sauces
insoluble in water.
Sodium stearoyl E481/172.846 Dispersible in warm water and Fine bakery wares, emulsified liqueur, fat emulsions, desserts, Negatively charged
lactylate soluble in hot edible oils and fats. beverage whiteners, and minced and diced canned meat products
Citric acid esters E472/172.832 Dispersible in hot water, insoluble Fats for stabilizing, also as synergists for antioxidants, baking fat Negatively charged
of MG in cold water, and soluble in emulsions, bakery margarines and shortening for stabilizing, mar-
edible oils and fats. garine, mayonnaise, salad dressings, sauces, and in low-calorie foods
Mono and E471/184.1505 Oil Baked goods, confectionery (e.g., chewing gum, toffees, and caramels), Nonionic
diglycerides dairy products, creams, desserts, edible ices, margarines, shortenings
Polyglycerol esters E475/172.854 Water Cakes and icings, margarine, and salad oils Nonionic
of FA
Propylene glycol E477/172.856 Oil Whippable icing Nonionic
esters of FA
Polyoxyethylene E435/172.836 Water Fine bakery wares, fat emulsions for baking purposes, milk and cream Nonionic cloud point around
(20) sorbitan analogues, emulsified sauces, soups, dietary food supplements, carriers
monooleate and solvents for colors, fat-soluble antioxidants, and antifoaming agents
Polyoxyethylene E433/172.840 Water Fine bakery wares, fat emulsions for baking purposes, milk and cream Nonionic cloud point around
(80) sorbitan analogues, emulsified sauces, soups, dietary food supplements,
monostearate dietetic foods, carriers and solvents for colors, fat-soluble
antioxidants, and antifoaming agents
67
68 Engineering Aspects of Food Emulsification and Homogenization
tertiary structure in the solution and at the interface, their size, the net charge and
charge distribution, their capability to form gels, and the distribution of hydrophilic
and hydrophobic groups. There are numerous proteins that are used in emulsions,
and the choice of protein emulsifier is often not only based on function but also based
on what food group the emulsion is related. There are several traditional or natural-
occurring emulsions that, at least, are partially stabilized by proteins, such as mayon-
naise, dairy products, and sausages. Table 3.5 presents some of the more common
protein emulsifiers. Most of these emulsifiers are mixtures of several different protein
species. Thus, depending on the production and formulation conditions, the actual
proteins at the interface may differ although the same protein emulsifier is used.
taBle 3.5
Common Commercial Protein emulsifiers and example
of Proteins that are Part of the emulsifier
emulsifier Key Proteins molecular weight IP
Whey proteina β-Lactoglobulin 18.6 5.3
α-Lactalbumin 14.2 4.8
Bovine serum albumin 66 5.1
Caseins b
α1-Casein 23 4.1
α2-Casein 25 5.3
β-Casein 24 5.1
κ-Casein 19 5.6
Egg whitec Ovalbumin 45 4.5
Ovotransferrin 77.7 6.0
Ovomucoid 28 4.1
Lysozyme 14.3 10.7
Egg yolkc Phosvitin 160–190
Low-density lipoproteins 16–135
Cobalamin-binding proteins 39
Riboflavin-binding proteins 37
Biotin-binding proteins 72
α- and β-Lipovitellins 400
Soy proteind α-Conglycinin 18–33
β-Conglycinin 104
σ-Conglycinin 141–171
Glycinin 317–360
Sources: a Kinsella, J.E. and Whitehead, D.M., Advances in Food and Nutrition
Research, Academic Press, San Diego, CA 1989.
b Swaisgood, H.E., J. Dairy Sc., 76, 10, 3054–3061, 1993.
d Clarke, E.J. and Wiseman, J., J. Agr. Sci., 134, 111–124, 2000.
Factors affecting the protein adsorption into the interface during competitive adsorp-
tion from solution are size, charge and hydrophobicity of the protein, the transport
conditions of proteins to the surface during emulsification, if adsorbed proteins can
be exchanged by proteins in solution, and the degree of conformational changes of the
protein at the interface (Nilsson et al. 2006, Nilsson, Osmark et al. 2007, Wahlgren and
Arnebrant 1991). Thus, it is a complex issue to understand what proteins are actually
adsorbed at the interface. Magnusson and Nilsson (2013) reviewed this recently for egg
yolk in high internal phase emulsions and discussed that the main property governing
adsorption was the hydrophobicity of the proteins and that there is a preference for
HDL and LDL proteins to adsorb at the interface. In the case of milk proteins, Surel
et al. (2014) have seen that in mixtures of casein micelles and whey proteins, casein
dominates at the interface when the fraction of casein in the solution is above 25%.
Proteins get their amphiphilic character from the mixture of hydrophilic and
hydrophobic amino acids. The amino acid sequence (secondary structure) also gives
the template for the three-dimensional structure of the protein (tertiary structure).
However, one should be aware that the tertiary structure will vary due to solution
conditions, and that proteins in solution have a well-defined tertiary structure, which
could be considerably changed and even lost when adsorbing at an interface. Proteins
are often divided into different categories based on their tertiary structure. The most
common structures are random coil (casein), globular proteins (whey proteins and
egg proteins), and rod-like structures (fibrinogen, collagen, and gelatin). In many
cases, the protein has a defined molecular weight but for some food proteins such as
gelatin, this is not the case. The distribution of hydrophobic groups within the poly-
mer is important. For globular proteins, the hydrophobic groups are mainly found
inside the core of the protein shielding them from water. Upon adsorption into the
oil–water interface, these hydrophobic groups could orient themselves toward the oil,
which might lead to conformational changes of the protein. A few proteins especially
κ-Casein has very distinctive hydrophilic and hydrophobic domains, which together
with its semirandom coil structure makes them especially suitable as emulsifiers.
The casein proteins α1-, α2-, β-, and κ-caseins form complex called casein micelles.
Although these proteins play a large biological and a technical role, the structure of
the casein micelles is still debated (Dalgleish 2011, Horne 2002).
The main difference between low-molecular weight emulsifiers and proteins is
that while the adsorption of the former is completely reversible, when the concentra-
tion is lowered, proteins have a tendency to adsorb irreversibly. This makes them less
sensitive to changes such as dilution. However, even if the adsorption is irreversible
toward the lowering of concentration, the protein could still be exchanged by other
species (proteins or low-molecular ones) that have higher driving force for adsorp-
tion, for example, a higher reduction of the surface tension at the oil–water interface.
The kinetics of these events and the conformational changes of the proteins can
be slow, on the timescale of hours to days, and can lead to postproduction changes
of the emulsion. Furthermore, proteins are sensitive to heat, enzymes, and solution
conditions such as pH and ionic strength, which lead to degradation, aggregation,
and other protein changes. These events can also lead to long-term change of emul-
sions stabilized by proteins. Another difference between low molar mass (or small)
emulsifiers and proteins is the rheology of the adsorbed layer. Proteins often form
70 Engineering Aspects of Food Emulsification and Homogenization
thicker, more viscous layers than small emulsifiers (Bosa and van Vlieta 2001). This
is in most cases positive for the long-term stability of the emulsion. Proteins might
stabilize emulsions through electrostatic repulsion and, thus, several protein sys-
tems show tendencies to aggregate at pH close to the isoelectric point of the pro-
tein emulsifier. If such aggregation does not lead to coalescence, it could lead to an
increase in the viscoelasticity of the emulsion (Wu, Degner, and McClements 2013).
In systems where both small molecular emulsifiers and proteins are present, there
might be a competition between the components at the interface or there might be
a cooperative adsorption (Maldonado-Valderrama and Patino 2010, Nylander et al.
2008, Rodríguez, García, and Niño 2001, Waninge et al. 2005). The competitive
adsorption of proteins and small emulsifiers are strongly concentration dependent,
and at concentrations below the CMC of the emulsifiers, proteins often dominate at
the interface (Wahlgren and Arnebrant 1992). The order in which the components
reach the surface might also be important as small surface-active components can-
not always remove already adsorbed proteins (Karlsson, Wahlgren, and Trägårdh
1996, Wahlgren 1995). Cooperative adsorption may occur when the protein com-
plexes with the low-molecular emulsifier, which, for example, is common for many
ionic surfactants (Maldonado-Valderrama and Patino 2010). It is often seen that the
adsorption of low-molecular emulsifiers to protein-stabilized emulsions have a detri-
mental effect on the emulsion stability (Wilde et al. 2004). Furthermore, there could
be strong interactions between proteins and surfactants in solution, changing the
structure and behavior of the proteins (Nylander et al. 2008).
Proteins also have the capability to change the rheology of the emulsions, espe-
cially if they are triggered to aggregate and to form a gel. Gel formation is often
induced by heating and denaturation of the proteins but could also be an effect of
pH, for example, the change in the rheology between milk and yoghurt. For example,
increased viscosity through the addition of proteins is important in low-fat products
such as margarines, sausages, and spreads (Chronakis 1997). Common proteins used
to form gel structures are milk-based systems such as whey proteins (Chronakis 1997,
Youssef and Barbut 2011) and soy proteins (Youssef and Barbut 2011).
3.4.4 PolySaccHarideS
Polysaccharides primarily function as viscosity modifiers in emulsions. However,
hydrophobically modified polysaccharides are also used as emulsifiers. A thorough
description of polysaccharides is given in Food Polysaccharides and Their Applications
(Stephen, Phillips, and Williams 2006). The properties of a polysaccharide is given by
the structure of the smallest repeating saccharide units, the degree of branching of the
polymer, and its molecular size. Differing from proteins, polysaccharides typically
have a high degree of polydispersity when it comes to branching and molecular weight.
There can also be a large batch-to-batch variation, which might lead to variation in
performance. Table 3.6 presents some of the more common polysaccharide groups and
these will also be discussed subsequently.
Traditionally, exudate gums, such as gum arabic, have been used as emulsifiers
especially in flavored beverages (Dickinson 2003). These are natural polysaccha-
rides that are produced by plants as a protection against bacteria and dehydration.
taBle 3.6
some Key Polysaccharides
name molecular structure Function viscosity modification
Starch amylase Essentially linear (1 → 4)-α-d-glucan Stabilizers, thickeners, Starch gelatinization; the ordered crystalline regions
Formulation of Emulsions
(Continued)
71
72
Source: Stephen, A.M. et al., Food Polysaccharides and Their Applications, CRC Press, Boca Raton, FL.
Engineering Aspects of Food Emulsification and Homogenization
Formulation of Emulsions 73
One has to be aware that there is a very high variation in the composition between
gums obtained from different species (Stephen, Phillips, and Williams 2006) and that
this variation might affect the emulsion produced. These contain a heterogeneous mix-
ture of highly branched polysaccharides and a small amount of proteins (2% for gum
arabic) covalent attached to the polysaccharides. Gum arabic is thought to have a water
blossom structure built up of an amino acid core of around 400 units onto which bulky
polysaccharide units of 250 kDa are grafted (Stephen, Phillips, and Williams 2006).
It is the protein moieties of the exudated gums that make them surface active. Alftrén
showed that for gum arabic and mesquite gum, the amount of protein in the polysac-
charide fraction increased with increasing molecular mass and that these high protein
content/high-molecular-weight fractions were preferentially adsorbed into emulsion
droplets (Alftrén et al. 2012, Evans, Ratcliffe, and Williams 2013). The emulsification
capacity of gum arabic is lost upon heating (Williams, Phillips, and Randall 1990).
Another group of polysaccharide that is dependent on a protein fraction for its emul-
sifying properties are modified pectins (Akhtar et al. 2002, Dickinson 2003). Although
pectin is mainly used as a rheological modifier in emulsions, if they are modified, they
might work as emulsifiers (Dickinson 2003). The modification is, in most cases, acety-
lation; however, depolymerization using acids has also been used (Dickinson 2003).
Akhtar et al. (2002) have shown that depolymerized citrus pectin of 70% esterifica-
tion gives good stable emulsions, although only 25% of the pectin is adsorbed into the
interface and that upon storage, there are some flocculation that increase particle size.
For example, hydrophobic modification of polysaccharides starch can also pro-
duce molecules that are surface active and can be used as emulsifiers. Nilsson and
Bergenståhl (2006, 2007) have done extensive studies of hydrophobically modified
starch and shown that they are good emulsifiers and that the surface load of OSA-
starch can be as high as 16 mg/m2. They have also shown that it is the high molar
mass components of the polymer that are selectively adsorbed to the emulsion drop-
lets (Nilsson, Leeman et al. 2007).
According to Dickinson (2013), xanthan gum, which has high viscosity at low shear,
is established as the first choice when it comes to using them as rheological modifiers
for stabilizing emulsions, but there are a large range of polysaccharides that are used for
improving texture in food emulsions. Polysaccharides can increase the viscosity of an
emulsion either by some gelation mechanism, such as those triggered by the addition of
calcium ions (alginate, pectins, and carrageenan) and the formation of double helices
and crystallization (starch), or by nonspecific chain–chain interactions determining
the viscosity. Nongelling polysaccharides, especially if they are linear and are good
solvents, often behave as random coil polymers. At low concentrations, such polymers
behave more or less as Newtonian liquids but as the concentration increases, the poly-
mer chains start to overlap and the rheological behavior of the polymer changes. Above
the so-called overlap concentration, the viscosity becomes non-Newtonian (shear thin-
ning) and the viscosity increases more steeply with polymer concentration. Gelling of
polysaccharide can also form continuous water-swollen networks at low concentra-
tions. To obtain such systems, the polysaccharides contain both regions that form the
physicochemical bonds between the polymers and the nonbinding regions that primar-
ily hold the water. The regions that form the bonds are usually well ordered, allowing
for helices, egg-box, ribbon–ribbon, and double helix–ribbon structures.
74 Engineering Aspects of Food Emulsification and Homogenization
In complex food products, for example, the interaction with the polysaccharides
with other components of the product might lead to segregated networks when pro-
teins and polysaccharides phase segregate. The polysaccharides, especially starch,
can also form inclusion complex with small emulsifiers such as monoglycerides and
fatty acids (Eliasson 1986, Tufvesson, Wahlgren, and Eliasson 2003). These interac-
tions are often stronger than the tendency for the emulsifier to adsorb into interfaces
and thus it will lower the amount of the emulsifier available (Lundqvist, Eliasson,
and Olofsson 2002). The complexes as formed will also have additional properties
such as melting point than the double helices normally formed in starch.
In complex food systems, these types of interactions might either lead to the stabili-
zation of the emulsion or, especially in the latter case, destabilization.
Lately, there has been an interest in using the protein–polysaccharide complex
as emulsifiers in food systems (Evans, Ratcliffe, and Williams 2013). As discussed
previously, some traditional polysaccharide emulsifiers are probably protein–
polysaccharide complexes; other methods to obtain such complexes could be the
formation of Maillard conjugates (Akhtar and Dickinson 2007, Zhang, Chi, and
Li 2013) or electrostatic complexes (Harnsilawat, Pongsawatmanit, and McClements
2006, Koupantsis, Pavlidou, and Paraskevopoulou 2014, Xu et al. 2014). Protein–
polysaccharide complexes are used in the encapsulation of emulsion droplets, lead-
ing to the protection of sensitive substances (Xu et al. 2014), or the encapsulation of
flavors (Koupantsis, Pavlidou, and Paraskevopoulou 2014). Protein–polysaccharide
complex stabilized emulsions have been shown to be more stable to stress, for exam-
ple, heat than emulsions only stabilized by the protein (Harnsilawat, Pongsawatmanit,
and McClements 2006). They have also been seen to be insensitive to pH and salt
concentration (Zhang, Chi, and Li 2013).
3.4.6 ParticleS
In addition to small molecular-weight surfactants and macromolecules, colloidal
particles can be utilized to stabilize emulsions. Particle-stabilized emulsions (com-
monly referred to as Pickering-type emulsions) are possible as a result of the proper-
ties of the particles, where a combination of size, form, and partial dual wettability
of both the oil and water phases confers Pickering particles several useful properties
and the ability to create emulsion droplets that are highly stable against coalescence
Formulation of Emulsions 75
the particles may form a network or a gel-like structure with increased viscoelastic
moduli (Dickinson 2013). This can also result in the emulsion having a yield stress,
which, even if relatively small, will assist in preventing creaming, drop–drop con-
tact, and coalescence under quiescent storage conditions (Dickinson 2012). This may
also prove to be the reason why particle-stabilized emulsions, even when the surface
coverage of particles at the oil–water interface is much less than a closely packed
monolayer, can remain stable over several years of storage (Timgren et al. 2013).
Rheological properties resulting from particle–particle interactions may also have
the added benefit of reducing the need for additional thickeners and viscosity modi-
fiers in particle-stabilized formulations (Dickinson 2013).
taBle 3.7
Characterization of emulsion Ingredient Properties
Key Properties (reviews) examples of methods Comments
Surface tension of biopolymers, surfactants, Wilhelm plate, drop volume, Pendent drop Pendent drop works well for soluble substances and for
and proteinsa oil–water interface. It is also good for measuring dynamic
change in surface tension. Wilhelm plate are easier to use
for nonsoluble materials
Interfacial rheology that is especially Interfacial rheology can be measured as dilatational The methods listed here are suitable for measurements
important for biopolymers and proteins.b deformation (oscillating increase and decrease of close to equilibrium
Presence of low-molecular emulsifiers often surface, e.g., a pendant drop) Langmuir trough gives both dilatation and shearing
decrease interfacial elasticityc Shearing deformation where the area is constant but the deformation
shape is changed (deep-channel surface viscometer)
Crystallinity and polymorphismd X-ray diffraction is the preferred method for identifying The use of synchrotron radiation enables analyses on a
crystalline structure but DSC and FTIR can be used as short timescale and thus kinetic phenomena can be
complementary techniques studied. X-ray can also be used to study self-associated
structures of lipids and synchrotron radiation can be used
to study structures at interfacese
Melting point and amount of solid materialf DSC, ultrasonics, and heat-controlled microscopy One should be aware that the scanning rate used in DSC
sometimes are too fast to have the sample in equilibrium
and this can affect the measured melting temperature
(Continued)
Engineering Aspects of Food Emulsification and Homogenization
taBle 3.7 (Continued )
Characterization of emulsion Ingredient Properties
Key Properties (reviews) examples of methods Comments
Formulation of Emulsions
Molecular weight/mass distribution is critical There are numerous methods for molecular weight Size exclusion and FFF can be linked to light-scattering
properties especially for proteins and determination; for high-molecular-weight molecules, detectors to get additional information such as shape. FFF
polysaccharides. Information on branching FFF,g size exclusion,h gel electrophoreses, rheology of is especially well suited for larger polymers and proteins
and shape can also be important to diluted solutions, and analytical ultracentrifugation can such as starch molecules. MALDI TOF is primarily used
understand biopolymers be used. Mass spectroscopy such as MALDI TOFi to get composition and structural information
Sources: a Drelich, J. et al., Encyclopedia of Surface and Colloid Science, Marcel Dekker, New York, 2002.
b Pelipenko, J. et al., Acta Pharm., 62, 2,123–140, 2012.
c Maldonado-Valderrama, J. and Patino, J.M.R., Curr. Opin. Colloid Interface Sci., 15, 4, 271–282, 2010.
e Cristofolini, L., Curr. Opin. Colloid Interface Sci., 19, 3, 228–241, 2014.
f McClements, D.J., Food Emulsions: Principles, Practices, and Techniques, CRC Press, Boca Raton, FL, 2005a.
h Hagel, L., Protein Purification, John Wiley & Sons, Inc., Hoboken, NJ, 2011.
DSC, differential scanning calorimetry; FFF, field flow fractionation; FTIR, Fourier transform infrared spectroscopy.
79
80 Engineering Aspects of Food Emulsification and Homogenization
properties are a result of microstructure and molecular interactions within the emul-
sion formulation (Corredig and Alexander 2008) and are closely related to the ability
of the formulation to stabilize and maintain the stability of the emulsion droplets,
as well as the rheology of the resulting dispersion. For this reason, in the develop-
ment of food-based emulsions, the effectiveness of emulsifiers and the evaluation of
texture are often studied. From the material presented in the earlier sections of this
chapter, it is apparent that there is a large variety of emulsifiers that can be used in
formulating food-based emulsions. The fundamental performance of an emulsifier
can be described by the emulsifying capacity (EC) as the minimum amount required
to produce a stable emulsion and its ability to produce small drops during homogeni-
zation. The emulsion stability index (ESI) is a measure of the ability of an emulsifier
to prevent droplets from aggregating, flocculating, and coalescing over time.
via a mass balance of the emulsifier; that is, the amount adsorbed at the oil–water
interface, which is found by considering the initial concentration of emulsifier in the
continuous phase, Cini , minus the amount remaining in the continuous phase after
emulsification, Cend . This is carried out experimentally by carefully separating the
droplets from the continuous phase by centrifugation or filtration and determin-
ing the remaining concentration of the emulsifier (Tcholakova et al. 2002). The
interfacial area to which the emulsifiers are adsorbed is found by measuring the
specific surface area of the emulsion by either microscopy or an automated particle
size analyzer. The specific surface area, S, is determined from the surface mean
diameter, d32
d32 =
∑Nd i
3
i i
(3.2)
∑Nd i
2
i i
where:
Ni is the number of drops with diameter di
Because the specific surface area, S, is the sum of all the surface areas of all drops
divided by the sum of all their volumes (m2/m3), we can calculate S from d32:
4π(d32 /2)2 6
S= = (3.3)
(4/3)π(d32 /2)3 d32
Now, including this into the mass balance or the emulsifier over the continuous phase
and interface, we get
Vdisp
φ= (3.5)
Vcts + Vdisp
Typically, the values of ΓS for molecular food emulsifiers is around a few milligram
per square meter, but is much larger for particles, hundreds to thousands milligram per
square meter, as ΓS is also directly related to the thickness of the interfacial layer. These
estimates of surface load values provide some knowledge with respect to the minimum
amount of emulsifier that is required to make an emulsion having droplets of a given
size and the dispersed phase fraction. However, in practice, an excess of emulsifier is
often used, as not all emulsifiers are ideally adsorbed into the oil–water interface during
homogenization due to kinetic limitations, as well as due to the partitioning equilibrium
conditions between the interface and the continuous phase. Furthermore, the surface
load of some types of emulsifiers is also sensitive to formulation conditions such as ionic
strength, pH, the concentration of macromolecules, temperature, and so on.
82 Engineering Aspects of Food Emulsification and Homogenization
d( 0 )t
ESI = (3.6)
d( t ) − d( 0 )
where:
d( 0 ) is the initial mean droplet diameter of the emulsion
d( t ) is the mean droplet diameter measured after a storage time, t (McClements
2005b)
Some of the main strengths of this method include that the mean droplet diameter
can be readily determined in analytical instruments, the evolution of particle size
microstructure is often a precursor to quality deterioration on the macrostructure
(creaming and phase separation, etc.) and can re-repeated over relevant timescale for
the shelf life of the product. A similar index is also sometimes used that compares
the specific surface area of the emulsions rather than just mean droplet size as a mea-
sure of how much coalescence has taken place. This surface coalescence index (SCI)
is more sensitive to the fate of smaller drops (as they have a relatively larger surface
are to volume ratio) and can be calculated by
S( 0 ) − S( t )
SCI = (3.7)
S( 0 )
where:
S( 0 ) is the initial specific surface area of the emulsion
S( t ) is new specific surface area of the emulsion measured after a storage time,
t (Anton, Beaumal, and Gandemer 2000)
S is calculated directly from 6/d32
It should be noted that there is no compelling evidence that a single index such as
EC, ESI, or SCI can be used to ultimately compare the effectiveness or the stability
of emulsifiers if they have been produced under different homogenization conditions.
Still, these indices are very useful when comparing a series of emulsifiers of emulsion
formulations produced under standardized conditions or in situations when the influ-
ence of specific changes are being made to the formulation, processing conditions, or
functionality of a specific ingredient that is being studied (McClements 2005b).
know at what degree creaming or sedimentation is likely to occur over the shelf life
of a relevant product. Due to the fact that emulsion droplets in the context of food
emulsions typically never have the same density as the continuous phase and are
large enough for the buoyant forces to overcome viscous resistance and Brownian
motion, they allow gravitational separation to be observed on a relevant timescale.
As most edible oils at room temperature have a lower density than aqueous solu-
tions, oil droplets in O/W emulsions will tend to rise to the top of the container in
a process referred to as creaming, leaving the depleted layer by an emulsion drop at
the bottom of the container, often referred to as serum. These terms likely originate
from the prevalence of diary emulsions. For W/O emulsions, the sedimentation of
water droplets is observed, although generally at a much slower rate due to the higher
viscosity of the oil. However, the opposite can be observed, where the sedimentation
of oil droplet can occur if fat crystals or other weighing agents are added to the oil
phase, or in some cases, in particle-stabilized emulsions, where a higher density of
the stabilizing particle layer increases the overall density of the droplet causing them
to settle (Rayner, Timgren et al. 2012).
The rate at which a single spherical droplet or particle will cream (or settle) in a
Newtonian fluid can be predicted by the Stokes velocity:
2 gr 2 (ρ2 − ρ1 )
vStokes = − (3.8)
9η1
where:
g is acceleration due to gravity
r is the particle radius
ρ1 and ρ2 are the densities of the continuous and dispersed phases, respectively
η1 is the continuous phase viscosity
The settling or creaming rate of drops and particles indicated by Stokes equation is
somewhat idealized, as in reality, emulsions drops are not all the same size and will be
interacting during creaming or settling. Furthermore, Stokes law is mainly applicable
at low concentrations of the dispersed phase. However, Stokes equation does provide an
illustration of the factors that have the most impact on the gravitational tendency, specif-
ically the viscosity of the fluid surrounding the droplets, their relative density, and, to a
large degree, the droplet size due to the exponent. For example, an oil droplet creams at
a rate of 0.1 mm/day if its diameter is 0.1 µm, and will cream at a rate of 10 mm/day if the
diameter is 1 µm, if all other conditions remain constant. In many practical situations,
these conditions are not constant and are more complex; for example, there is often an
increase in the effective particle size during creaming due to coalescence, flocculation,
or Ostwald ripening (see Chapter 2), which results in Equation 3.8 under predicting the
rate of gravitational separation (McClements 2007). Therefore, it is often more practical
to directly quantify gravitational separation of the emulsions during storage.
The extent of creaming or sedimentation in an emulsion can be monitored by
visual observation. This method is cheap and straightforward, only requiring the
emulsions to be stored in an appropriate environment in clear glass vials or test
tubes. The layer formed by creamed emulsion droplets can be readily seen, and often
84 Engineering Aspects of Food Emulsification and Homogenization
(a) (b)
Vemuls Hserum
EI = × 100% CI = × 100%
Vtotal Hemuls
FIgure 3.4 (a) Test tube showing creamed emulsions as defined in EI; (b) schematic test
tube showing layers as defined in CI.
the serum layer is transparent or optically distinct to such a degree that its height can
be determined, or its volume estimated. The emulsion index (EI) is a measure of the
volume of an emulsion layer formed relative to the total volume given by the follow-
ing equation, with the volumes defined in Figure 3.4:
Vemuls
EI = × 100% (3.9)
Vtotal
The relative heights of the creamy layer is also used to define the creaming index (CI):
H serum
CI = × 100% (3.10)
H emuls
φ
CI final = 1 − × 100% (3.11)
P
Knowing the expected CI final can be relevant if emulsions with different dispersed
phase volumes are to be compared, as the more dispersed phases, the lesser will
be the serum. This concept had also been extended to adjust for the fact that
particles, and especially the larger food-based ones used in stabilizing Pickering
emulsions, also contribute to the amount of dispersed phase observed. The total
amount of nonseparated emulsion can be expressed as the relative occluding vol-
ume (ROV).
Formulation of Emulsions 85
Vemuls
ROV = (3.12)
Vdisp + Vparticles
where:
Vemuls is the volume of the observed emulsion (i.e., the nonclear fraction) after
emulsification
Vdisp is the known volume of the added dispersed phase
Vparticles is the known volume occupied by the added particle stabilizers
1.1 6
1.0 5
Emulsion index
0.9 4
ROV
0.8 3
0.7 2
0.6 1
0.5 0
0.1 1 10 0.1 1 10
Storage time (weeks) Storage time (weeks)
FIgure 3.5 EI and ROV of quinoa starch granule-stabilized oil-in-water emulsions. (Data
from Timgren, A. et al., Procedia Food Sci., 1, 95–103, 2011.)
86 Engineering Aspects of Food Emulsification and Homogenization
concentrated emulsions and the measurement of both the particle size and phase thick-
ness continuously over time (Mengual et al. 1999), detecting the creaming long before
it is visible to the naked eye. The other major limitation of the gravitational separa-
tion analysis is that it takes a significantly long time to monitor instability that may
occur after weeks of storage and that the storage conditions in reality are not neces-
sarily those found in a controlled laboratory environment. This can be overcome in
two ways: (1) by increasing the gravitational field where the droplets cream/settle to
accelerate storage and (2) by exposing the emulsions to environmental stresses that
may trigger instability.
0 ζ0 ζ1 ζ2 ζ
ω
Cream
Oil Serum
Hrel HC
Z 0
FIgure 3.6 Schematic image of the thickness of the oil layer released during a forced
coalescence test.
axis of rotation (z direction in Figure 3.6) due to their relatively lower density. This
is the case for almost all food emulsions. Initially, the emulsion droplets form a
creamy layer in which they are forced into close proximity but keep their initial
form. As the centrifugal force is increased, they are pressed tighter and tighter
together and eventually the interfacial layer surrounding and stabilizing the drop-
lets will rupture, releasing a layer of oil (Hrel) on the top of the emulsion column in
the tube (see Figure 3.6).
The critical pressure that the emulsion can withstand before the oil is released
CR
when the film ruptures is described as a critical osmotic pressure, POSM . For a full
derivation, refer to Tcholakova et al. (2002, 2006) and references therein.
Hc
(Voil tot − Voil rel )
∫
= ∆ρgk φ( z )dz = ∆ρgk
CR
POSM (3.13)
ATT
0
where:
Δρ is the density difference between the oil and aqueous phases
gk is the centrifugal acceleration
φ(z) is the local volume fraction of oil along the z direction along the centrifugal
field
Voil tot is the total volume of oil in the emulsion
Voil rel is the volume of oil released
ATT is the interior cross-sectional area of the test tube containing the emulsion
After centrifugation, the height of the creamy layer, HC and oil released Hoil rel can
be easily measured, where Hoil rel = Voil rel / ATT . POSM
CR
may be readily calculated from
experimental data, if one assumes that the centrifugal field is homogenous through
the column of creamed emulsion, HC, and can be represented by the square of the
angular frequency, ω times the mean distance of the emulsion layer from the axis of
rotation, ζ:
88 Engineering Aspects of Food Emulsification and Homogenization
ω2 (ζ1 + ζ 2 )
gk ≈ = constant (3.14)
2
Tcholakova et al. (2002, 2006) have proven that this is a reasonable assumption, as
more precise calculations that take into the account the spatial variation of the field
compared to using a mean distance gives a relatively small difference in the result,
and the imposed error is within the experimental accuracy of the measurements. This
method has been demonstrated to be particularly useful for monitoring the coales-
cence stability of different types of protein-stabilized emulsion with various composi-
tions (Denkov, Tcholakova, and Ivanov 2006, Tcholakova et al. 2002, 2003, 2006).
In addition to centrifugation, other types of accelerated coalescence tests include
subjecting the emulsions to other types stress such as mechanical forces (extended
homogenization, pumping, vibration, shearing, extruding, whipping, shaking, and
mixing) environmental stresses (freeze–thaw cycling, thermal processing, and heat
abuse), as well as compositional stresses (drying causing a change in solute com-
position, changes in pH and ionic strength, etc.). All of which with the purpose
of emulating some sort of typical event, environmental stress, or process that the
emulsion under consideration should withstand during processing, transport, shelf
life, and use. The formulation in general—and the performance of its emulsifier in
particular—is evaluated in a variety of conditions depending on its application to
establish a design space, in which a particular emulsifier is expected to successfully
function. Examples of test methods and experimental conditions/protocols can be
found in McClements (2007) and references therein.
where:
η0 is the viscosity of the continuous phase
φc is the is the volume fraction at random close packing of spheres
φe is the volume fraction of oil in the emulsion.
100,000
10,000
1,000
G′ and G′′ (Pa)
100
10
0
0.1 1 10
Frequency, f (Hz)
FIgure 3.7 Elastic modulus (G′, in Pa) and viscous modulus (G″, in Pa) versus frequency
(f, in Hz) for Pickering emulsions with 19% Miglyol oil-in-water (O/W) emulsion stabilized
by chitosan G″ (closed squares), G′ (open squares) with 55% Miglyol O/W emulsion stabilized
by octenyl succinic anhydride (OSA)-modified quinoa starch G″ (closed circles), G′ (open
circles).
90 Engineering Aspects of Food Emulsification and Homogenization
10,000
1,000
100
G′ and G′′ (Pa)
10
0.1
0.01
1.00E−01 1.00E+00
Complex shear strain
FIgure 3.8 Elastic modulus (G′, in Pa) and viscous modulus (G″, in Pa) versus complex
shear strain for Pickering emulsions with 19% Miglyol O/W emulsion stabilized by chitosan
G″ (closed squares), G′ (open squares) with 55% Miglyol O/W emulsion stabilized by OSA-
modified quinoa starch G″ (closed circles), G′ (open circles).
linear region. Typically, in strain tests, the strain is increased until the structure of the
emulsions is broken and the gel starts to flow, which is shown as a rapid decease of
G′. As can be seen in Figures 3.7 and 3.8, even if the gel is stiffer (high G′) and has a
smaller linear gel region (frequency), it might still flow at lower strains.
The rheology of emulsions is especially important to avoid creaming in nons-
pace filled systems. To arrest creaming of an emulsion drop the elastic modulus of
the surrounding media must exceed the stress exerted by the droplet on the fluid
due to buoyancy. The bouncy force on a droplet will be F = VΔρg (where V is the
volume of the drop 4/3 πR3, Δρ the density difference between the phases, and
g gravity). The resulting stress is the force applied over an area, in this case the sur-
face area of the drop (4πR2), thus the stress exerted by the droplet will be RΔρg/3.
The stress asserted by normal emulsions droplets will thus be normally below
0.1 Pa. Therefore, to arrest creaming, gelling polymers only have to withstand
the stress asserted by the droplet. Furthermore, this also means that to predict the
resistance of the system to creaming, the rheology has to be measured at a low
stress, which can be obtained by constant stress or creep measurements. However,
normally we would like the system to flow when handled; therefore, a good rheo-
logical modifier should yield at higher stresses. In this respect, stress curves are
more informative.
reFerenCes
Abee, T., L. Krockel, and C. Hill. 1995. “Bacteriocins: Modes of action and potentials in food
preservation and control of food poisoning.” International Journal of Food Microbiology
28:169–185.
Formulation of Emulsions 91
Agriculture, U.S.D.O. 2014. World Production, Markets, and Trade Reports: Oilseeds.
Washington, DC: United States Department of Agriculture.
Akhtar, M., and E. Dickinson. 2007. “Whey protein–maltodextrin conjugates as emulsi-
fying agents: An alternative to gum arabic.” Food Hydrocolloids 21 (4):607–616.
doi:10.1016/j.foodhyd.2005.07.014.
Akhtar, M., E. Dickinson, J. Mazoyer, and V. Langendorff. 2002. “Emulsion stabilizing prop-
erties of depolymerized pectin.” Food Hydrocolloids 16 (3):249–256. doi:http://dx.doi
.org/10.1016/S0268-005X(01)00095-9.
Alftrén, J., J.M. Peñarrieta, B. Bergenståhl, and L. Nilsson. 2012. “Comparison of molecular and
emulsifying properties of gum arabic and mesquite gum using asymmetrical flow field-flow
fractionation.” Food Hydrocolloids 26 (1):54–62. doi:10.1016/ j.foodhyd.2011.04.008.
Allouche, J., E. Tyrode, V. Sadtler, L. Choplin, and J.L. Salager. 2004. “Simultaneous conductiv-
ity and viscosity measurements as a technique to track emulsion inversion by the phase-
inversion-temperature method.” Langmuir 20 (6):2134–2140. doi:10.1021/la035334r.
Anton, M., V. Beaumal, and G. Gandemer. 2000. “Adsorption at the oil–water interface and
emulsifying properties of native granules from egg yolk: Effect of aggregated state.”
Food Hydrocolloids 14 (4):327–335. doi:10.1016/s0268-005x(00)00009-6.
Aveyard, R., B.P. Binks, and J.H. Clint. 2003. “Emulsions stabilised solely by colloidal par-
ticles.” Advances in Colloid and Interface Science 100–102:503–546.
Awade, A.C. 1996. “On hen egg fractionation: Applications of liquid chromatographyt o the
isolation and the purification of hen egg white and egg yolk proteins.” Z lebensm Unters
Forsch 202:1–14.
Bancroft, W.D. 1913. “The theory of emulsification, V.” Journal of Physical Chemistry
17 (6):501–519.
Bayarri, S., A.J. Taylor, and J. Hort. 2006. “The role of fat in flavor perception: Effect of par-
tition and viscosity in model emulsions.” Journal of Agricultural and Food Chemistry
54:8862−8868.
Benjamins, J., M.H. Vingerhoeds, F.D. Zoet, E.H.A. de Hoog, and G.A. van Aken. 2009.
“Partial coalescence as a tool to control sensory perception of emulsions.” Food
Hydrocolloids 23 (1):102–115. doi:10.1016/j.foodhyd.2007.11.017.
Berasategi, I., M. Garcia-Iniguez de Ciriano, I. Navarro-Blasco, M.I. Calvo, R.Y. Cavero,
I. Astiasaran, and D. Ansorena. 2014. “Reduced-fat bologna sausages with improved
lipid fraction.” Journal of the Science of Food and Agriculture 94 (4):744–751.
doi:10.1002/jsfa.6409.
Binks, B.P. 2002. “Particles as surfactants—Similarities and differences.” Current Opinion in
Colloid & Interface Science 7 (1–2):21–41.
Binks, B.P. 2007. “Colloidal particles at liquid interfaces.” Physical Chemistry Chemical
Physics 9 (48):6298–6299. doi:10.1039/b716587k.
Binks, B.P., A.N. Boa, M.A. Kibble, G. MacKenzie, and A. Rocher. 2011. “Sporopollenin
capsules at fluid interfaces: Particle-stabilised emulsions and liquid marbles.” Soft Matter
7 (8):4017–4024.
Binks, B.P., J.H. Clint, G. Mackenzie, C. Simcock, and C.P. Whitby. 2005. “Naturally occurring
spore particles at planar fluid interfaces and in emulsions.” Langmuir 21 (18):8161–8167.
Bosa, M.A., and T. van Vlieta. 2001. “Interfacial rheological properties of adsorbed protein lay-
ers and surfactants: A review.” Advances in Colloid and Interface Science 91:437–471.
Burt, S. 2004. “Essential oils: Their antibacterial properties and potential applications in
foods—A review.” International Journal of Food Microbiology 94 (3):223–253.
doi:10.1016/j.ijfoodmicro.2004.03.022.
Castro, M.P., A.M. Rojas, C.A. Campos, and L.N. Gerschenson. 2009. “Effect of preservatives,
tween 20, oil content and emulsion structure on the survival of Lactobacillus fructiv-
orans in model salad dressings.” LWT—Food Science and Technology 42 (8):1428–1434.
doi:10.1016/j.lwt.2009.02.021.
92 Engineering Aspects of Food Emulsification and Homogenization
Charoen, R., A. Jangchud, K. Jangchud, T. Harnsilawat, E.A. Decker, and D.J. McClements.
2012. “Influence of interfacial composition on oxidative stability of oil-in-water emul-
sions stabilized by biopolymer emulsifiers.” Food Chemistry 131 (4):1340–1346.
doi:10.1016/j.foodchem.2011.09.128.
Charteris, W.P. 1996. “Microbiological quality assurance of edible table spreads in new prod-
uct development.” Journal of the Society of Dairy Technology 49:87–98.
Chen, B., D.J. McClements, and E.A. Decker. 2011. “Minor components in food oils: A
critical review of their roles on lipid oxidation chemistry in bulk oils and emulsions.”
Critical Reviews in Food Science and Nutrition 51 (10):901–916. doi:10.1080/104083
98.2011.606379.
Chevalier, Y., and M.-A. Bolzinger. 2013. “Emulsions stabilized with solid nanoparticles:
Pickering emulsions.” Colloids and Surfaces A: Physicochemical and Engineering
Aspects 439:23–34. doi:10.1016/j.colsurfa.2013.02.054.
Chronakis, I.S. 1997. “Structural–functional and water-holding studies of biopolymers in low
fat content spreads.” Lebensmittel-Wissenschaft und -Technologie 30:36–44.
Chung, C., and D.J. McClements. 2013. “Structure–function relationships in food emul-
sions: Improving food quality and sensory perception.” Food Structure 1:106–126.
doi:10.1016/j.foostr.2013.11.002.
Clarke, E.J., and J. Wiseman. 2000. “Developments in plant breeding for improved nutritional
quality of soya beans I. Protein and amino acid content.” Journal of Agricultural Science
134:111–124.
Corredig, M., and M. Alexander. 2008. “Food emulsions studied by DWS: Recent advances.”
Trends in Food Science and Technology 19 (2):67–75.
Cristofolini, L. 2014. “Synchrotron X-ray techniques for the investigation of structures and
dynamics in interfacial systems.” Current Opinion in Colloid & Interface Science
19 (3):228–241. doi:10.1016/j.cocis.2014.03.006.
Dalgleish, D.G. 2003. Food Emulsions Their Structures and Properties. F.E. Stig, K. Larsson,
and J. Sjöblom, Chapter 1, pp. 1–44.
Dalgleish, D.G. 2011. “On the structural models of bovine casein micelles—Review and pos-
sible improvements.” Soft Matter 7 (6):2265. doi:10.1039/c0sm00806k.
Davies, J.T. 1957. “A quantitative kinetic theory of emulsion type. I. Physical chemistry of
the emulsifying agent.” Gas/liquid and liquid/liquid interfaces. Proceedings of the 2nd
International Congress Surface Activity, Butterworths, London.
De Folter, J.W.J., M.W.M. Van Ruijven, and K.P. Velikov. 2012. “Oil-in-water Pickering emul-
sions stabilized by colloidal particles from the water-insoluble protein zein.” Soft Matter
8 (25):2807–2815.
Degner, B.M., C. Chung, V. Schlegel, R. Hutkins, and D.J. McClements. 2014. “Factors influ-
encing the freeze-thaw stability of emulsion-based foods.” Comprehensive Reviews in
Food Science and Food Safety 13 (2):98–113. doi:10.1111/1541-4337.12050.
Delamarre, S., and C.A. Batt. 1999. “The microbiology and historical safety of margarine.”
Food Microbiology 16:327–333.
Denkov, N.D., S. Tcholakova, and I.B. Ivanov. 2006. “Globular proteins as emulsion
stabilizers—Similaries and differences with surfactants and solid particles.” 4th World
Congress on Emulsions, Lyon, France.
Derkach, S.R. 2009. “Rheology of emulsions.” Advances in Colloid and Interface Science
151 (1–2):1–23. doi:10.1016/j.cis.2009.07.001.
Destribats, M., M. Rouvet, C. Gehin-Delval, C. Schmitt, and B.P. Binks. 2014. “Emulsions
stabilised by whey protein microgel particles: Towards food-grade Pickering emul-
sions.” Soft Matter 10: 6941–6954. doi:10.1039/C4SM00179F.
Dickinson, E. 1992. Introduction to Food Colloids. Oxford: Oxford University Press.
Dickinson, E. 2003. “Hydrocolloids at interfaces and the influence on the properties of dis-
persed systems.” Food Hydrocolloids 17:25–39.
Formulation of Emulsions 93
Dickinson, E. 2010. “Food emulsions and foams: Stabilization by particles.” Current Opinion
in Colloid & Interface Science 15 (1–2):40–49. doi:10.1016/j.cocis.2009.11.001.
Dickinson, E. 2012. “Use of nanoparticles and microparticles in the formation and stabilization
of food emulsions.” Trends in Food Science and Technology 24 (1):4–12. doi:10.1016/
j.tifs.2011.09.006.
Dickinson, E. 2013. “Stabilising emulsion-based colloidal structures with mixed food ingredients.”
Journal of the Science of Food and Agriculture 93 (4):710–721. doi:10.1002/jsfa.6013.
Dickinson, E., and E. Davies. 1999. “Influence of ionic calcium on stability of sodium casein-
ate emulsions.” Colloids and Surfaces B-Biointerfaces 12:203–212.
Dorobantu, L.S., A.K.C. Yeung, J.M. Foght, and M.R. Gray. 2004. “Stabilization of oil-water
emulsions by hydrophobic bacteria.” Applied and Environmental Microbiology 70
(10):6333–6336.
Drelich, J., Ch. Fang, and C.L. White. 2002. “Measurement of interfacial tension in fluid-fluid
systems.” In Encyclopedia of Surface and Colloid Science, 3152–3166. Marcel Dekker.
Einstein, A. 1906. “Eine neue Bestimmung der Moleküldimensionen.” Annalen der Physik
324 (2):289–306. doi:10.1002/andp.19063240204.
Eliasson, A.-C. 1986. “On effects of Surface active agents on the gelatinization of starch—a
calorimetric investigation.” Carbohydrate Polymers 6:463–476.
Ercelebi, E.A., and E. Ibanoglu. 2010. “Stability and rheological properties of egg yolk gran-
ule stabilized emulsions with pectin and guar gum.” International Journal of Food
Properties 13 (3):618–630. doi:10.1080/10942910902716984.
Eriksson, M. 2013. “Pickering emulsions based on food grade biological particles.” MSc,
Department of Food Technology, Engineering and Nutrition, Lund University (KLT
920, 2013).
Evans, M., I. Ratcliffe, and P.A. Williams. 2013. “Emulsion stabilisation using polysaccharide–
protein complexes.” Current Opinion in Colloid & Interface Science 18 (4):272–282.
doi:10.1016/j.cocis.2013.04.004.
Frankel, E.N. 1998. Lipid Oxidation. Dundee, Scotland: Oily Press.
Fredrick, E., P. Walstra, and K. Dewettinck. 2010. “Factors governing partial coalescence in
oil-in-water emulsions.” Advances in Colloid and Interface Science 153 (1–2):30–42.
doi:10.1016/j.cis.2009.10.003.
Frelichowska, J., M.A. Bolzinger, and Y. Chevalier. 2010. “Effects of solid particle content
on properties of o/w Pickering emulsions.” Journal of Colloid Interface and Science
351 (2):348–356. doi:10.1016/j.jcis.2010.08.019.
Friberg, S., and I. Wilton. 1970. “Liquid crystals—the formula for emulsions.” American
Perfumer and Cosmetics 85:27–30.
Golding, M., and T.J. Wooster. 2010. “The influence of emulsion structure and stability on
lipid digestion.” Current Opinion in Colloid & Interface Science 15 (1–2):90–101.
doi:10.1016/j.cocis.2009.11.006.
Gould, J., J. Vieira, and B. Wolf. 2013. “Cocoa particles for food emulsion stabilisation.” Food
& Function 4 (9):1369–1375. doi:10.1039/c3fo30181h.
Griffin, W.C. 1954. “Calculation of HLB values of non-ionic surfactants.” Journal of the
Society of Cosmetic Chemists 5 (4):249–256.
Gu, Y.X., Y.H. Huang, B. Liao, G.M. Cong, and M. Xu. 2000. “Studies on the characterization
of phase inversion during emulsification process and the particle sizes of water-borne
microemulsion of poly(phenylene oxide) ionomer.” Journal of Applied Polymer Science
76 (5):690–694. doi:10.1002/(sici)1097-4628(20000502)76:5<690::aid-app11>3.0.co;2-q.
Gupta, R., and D. Rousseau. 2012. “Surface-active solid lipid nanoparticles as Pickering
stabilizers for oil-in-water emulsions.” Food & Function 3 (3):302–311. doi:10.1039/
c2fo10203j.
Hagel, L. 2011. “Gel filtration: Size exclusion chromatography.” In Protein Purification,
51–91. John Wiley & Sons.
94 Engineering Aspects of Food Emulsification and Homogenization
Krafft, F., and H. Wiglow. 1895. “Ueber das Verhalten der fettsauren Alkalien und der Seifen
in Gegenwart von Wasser, III. Die Seifen als Krystalloide.” Berichte der deutschen
chenischen Gesellschaft 28:2566–2573.
Krieger, I.M. 1972. “Rheology of monodisperse lattices.” Advances in Colloid and Interface
Science 3:111.
Krog, N. 1997. “Food emulsifiers.” In Food Emulsions, edited by S. Friberg and K. Larsson,
141–188. New York: Marcel Dekker.
Laca, A., M.C. Saenz, B. Paredes, and M. Diaz. 2010. “Rheological properties, stability and sensory
evaluation of low-cholesterol mayonnaises prepared using egg yolk granules as emulsifying
agent.” Journal of Food Engineering 97 (2):243–252. doi:10.1016/ j.jfoodeng.2009.10.017.
Laredj-Bourezg, F., Y. Chevalier, O. Boyron, and M.A. Bolzinger. 2012. “Emulsions sta-
bilized with organic solid particles.” Colloids and Surfaces A: Physicochemical and
Engineering Aspects 413:252–259. doi:10.1016/j.colsurfa.2011.12.064.
Larsson, K. 1986. “Physical properties—Structural and physical characteristics.” In The Lipid
Handbook, edited by F.D. Gunstone, J.L. Harwood, and F.B. Padley, 321–384. London:
Chapman & Hall.
Le Révérend, B.J.D., I.T. Norton, P.W. Cox, and F.Spyropoulos. 2010. “Colloidal aspects of
eating.” Current Opinion in Colloid & Interface Science 15 (1–2):84–89. doi:10.1016/
j.cocis.2009.11.009.
Li, C., Y. Li, P. Sun, and C. Yang. 2013. “Pickering emulsions stabilized by native starch gran-
ules.” Colloids and Surfaces A: Physicochemical and Engineering Aspects 431:142–149.
doi:10.1016/j.colsurfa.2013.04.025.
Li, C., P. Sun, and C. Yang. 2012. “Emulsion stabilized by starch nanocrystals.” Starch−Stärke
64 (6):497–502. doi:10.1002/star.201100178.
Löf, D., K. Schillén, and L. Nilsson. 2011. “Flavonoids: Precipitation kinetics and
interaction with surfactant micelles.” Journal of Food Science 76 (3):N35–N39.
doi:10.1111/j.1750-3841.2011.02103.x.
Lück, E. 1990. “Food applications of sorbic acid and its salts.” Food Additives & Contaminants
7 (5):711–715. doi:10.1080/02652039009373936.
Lundqvist, H., A.-C. Eliasson, and G. Olofsson. 2002. “Binding of hexadecyltrimethylam-
monium bromide to starch polysaccharides. Part I. Surface tension measurements.”
Carbohydrate Polymers 49 (1):43–55.
Luo, Z., B.S. Murray, A.L. Ross, M.J.W. Povey, M.R.A. Morgan, and A.J. Day. 2012. “Effects
of pH on the ability of flavonoids to act as Pickering emulsion stabilizers.” Colloids and
Surfaces B: Biointerfaces 92:84–90. doi:10.1016/j.colsurfb.2011.11.027.
Luo, Z.J., B.S. Murray, A. Yusoff, M.R.A. Morgan, M.J.W. Povey, and A.J. Day. 2011. “Particle-
stabilizing effects of flavonoids at the oil-water interface.” Journal of Agricultural and
Food Chemistry 59 (6):2636–2645. doi:10.1021/jf1041855.
Magnusson, E., and L. Nilsson. 2011. “Interactions between hydrophobically modified starch
and egg yolk proteins in solution and emulsions.” Food Hydrocolloids 25 (4):764–772.
doi:10.1016/j.foodhyd.2010.09.006.
Magnusson, E., and L. Nilsson. 2013. “Emulsifying properties of egg yolk.” In Eggs: Nutrition,
Consumption and Health, 69–85.
Magnusson, E., C. Rosén, and L. Nilsson. 2011. “Freeze–thaw stability of mayonnaise
type oil-in-water emulsions.” Food Hydrocolloids 25 (4):707–715. doi:http://dx.doi.
org/10.1016/j.foodhyd.2010.08.024.
Maldonado-Valderrama, J., and J.M.R. Patino. 2010. “Interfacial rheology of protein–surfactant
mixtures.” Current Opinion in Colloid & Interface Science 15 (4):271–282. doi:10.1016/
j.cocis.2009.12.004.
Malone, M.E., I.A.M. Appelqvist, and I.T. Norton. 2003. “Oral behaviour of food hydrocolloids
and emulsions. Part 1. Lubrication and deposition considerations.” Food Hydrocolloids
17 (6):763–773. doi:10.1016/s0268-005x(03)00097-3.
96 Engineering Aspects of Food Emulsification and Homogenization
Marefati, A., M. Rayner, A. Timgren, P. Dejmek, and M. Sjöö. 2013. “Freezing and freeze-
drying of Pickering emulsions stabilized by starch granules.” Colloids and Surfaces A:
Physicochemical and Engineering Aspects 436:512–520. doi:http://dx.doi.org/10.1016/j.
colsurfa.2013.07.015.
Marku, D., M. Wahlgren, M. Rayner, M. Sjöö, and A. Timgren. 2012. “Characterization
of starch Pickering emulsions for potential applications in topical formulations.”
International Journal of Pharmaceutics 428:1–7.
Mayer, S., J. Weiss, and D.J. McClements. 2013. “Behavior of vitamin E acetate delivery sys-
tems under simulated gastrointestinal conditions: Lipid digestion and bioaccessibility
of low-energy nanoemulsions.” Journal of Colloid and Interface Science 404:215–222.
doi:10.1016/j.jcis.2013.04.048.
McClements, D.J. 2002. “Colloidal basis of emulsion color.” Current Opinion in Colloid &
Interface Science 7 (5–6):451–455. doi:10.1016/s1359-0294(02)00075-4.
McClements, D.J. 2005a. “Characterization of emulsions.” In Food Emulsions: Principles,
Practices, and Techniques. Boca Raton, FL: CRC Press.
McClements, D.J. 2005b. Food Emulsions: Principles, Practices, and Techniques, edited by F.M.
Clydesdale, CRC Press Series in Contemporary Food Science. Boca Raton, FL: CRC Press.
McClements, D.J. 2007. “Critical review of techniques and methodologies for characterization
of emulsion stability.” Critical Reviews in Food Science and Nutrition 47 (7):611–649.
doi:10.1080/10408390701289292.
McClements, D.J., and E.A. Decker. 2000. “Lipid oxidation in oil-in-water emulsions: Impact
of molecular environment on chemical reactions in heterogeneous food systems.”
Journal of Food Science 65 (8):1270–1282. doi:10.1111/j.1365-2621.2000.tb10596.x.
Medina, I., I. Undeland, K. Larsson, I. Storrø, T. Rustad, C. Jacobsen, V. Kristinová, and
J.M. Gallardo. 2012. “Activity of caffeic acid in different fish lipid matrices: A review.”
Food Chemistry 131 (3):730–740. doi:10.1016/j.foodchem.2011.09.032.
Mengual, O., G. Meunier, I. Cayré, K. Puech, and P. Snabre. 1999. “TURBISCAN MA
2000: Multiple light scattering measurement for concentrated emulsion and suspen-
sion instability analysis.” Talanta 50 (2):445–456. doi:http://dx.doi.org/10.1016/
S0039-9140(99)00129-0.
Michalski, M.C. 2009. “Specific molecular and colloidal structures of milk fat affecting
lipolysis, absorption and postprandial lipemia.” European Journal of Lipid Science and
Technology 111:413–431.
Modig, G., L. Nilsson, B. Bergenståhl, and K.G. Wahlund. 2006. “Homogenization-induced deg-
radation of hydrophobically modified starch determined by asymmetrical flow field-flow
fractionation and multi-angle light scattering.” Food Hydrocolloids 20 (7):1087–1095.
Moore, R.L., S.E. Duncan, A.S. Rasor, W.N. Eigel, and S.F. O’Keefe. 2012. “Oxidative
stability of an extended shelf-life dairy-based beverage system designed to contribute to
heart health.” Journal of Dairy Science 95 (11):6242–6251. doi:10.3168/jds.2012-5364.
Ng, S.P., O.M. Lai, F. Abas, H.K. Lim, B.K. Beh, T.C. Ling, and C.P. Tan. 2014. “Compositional
and thermal characteristics of palm olein-based diacylglycerol in blends with palm super
olein.” Food Research International 55:62–69. doi:10.1016/j.foodres.2013.10.035.
Nilsson, L. 2013. “Separation and characterization of food macromolecules using field-flow frac-
tionation: A review.” Food Hydrocolloids 30 (1):1–11. doi:10.1016/ j.foodhyd.2012.04.007.
Nilsson, L., and B. Bergenståhl. 2006. “Adsorption of hydrophobically modified starch at oil/
water interfaces during emulsification.” Langmuir 22:8770–8776.
Nilsson, L., and B. Bergenståhl. 2007. “Adsorption of hydrophobically modified anionic starch
at oppositely charged oil/water interfaces.” Journal of Colloid and Interface Science
308 (2):508–513. doi:10.1016/j.jcis.2007.01.024.
Nilsson, L., M. Leeman, K.G. Wahlund, and B. Bergenståhl. 2007. “Competitive adsorption
of a polydisperse polymer during emulsification: Experiments and modeling.” Langmuir
23:2346–2351.
Formulation of Emulsions 97
Song, X., Y. Pei, W. Zhu, D. Fu, and H. Ren. 2014. “Particle-stabilizers modified from indica
rice starches differing in amylose content.” Food Chemistry 153:74–80. doi:http://
dx.doi.org/10.1016/j.foodchem.2013.12.046.
Sørensen, A.D.M., A.M. Haahr, E.M. Becker, L.H. Skibsted, B. Bergenståhl, L. Nilsson,
and C. Jacobsen. 2008. “Interactions between iron, phenolic compounds, emulsifiers,
and pH in omega-3-enriched oil-in-water emulsions.” Journal of Agricultural and Food
Chemistry 56 (5):1740–1750.
Stephen, A.M., G.O. Phillips, and P.A. Williams, eds. 2006. Food Polysaccharides and Their
Applications. Boca Raton, FL: CRC Press.
Stokes, J.R., M.W. Boehm, and S.K. Baier. 2013. “Oral processing, texture and mouthfeel:
From rheology to tribology and beyond.” Current Opinion in Colloid & Interface
Science 18 (4):349–359. doi:10.1016/j.cocis.2013.04.010.
Surel, C., J. Foucquier, N. Perrot, A. Mackie, C. Garnier, A. Riaublanc, and M. Anton. 2014.
“Composition and structure of interface impacts texture of O/W emulsions.” Food
Hydrocolloids 34:3–9. doi:10.1016/j.foodhyd.2013.06.016.
Swaisgood, H.E. 1993. “Review and update of casein chemistry.” Journal of Dairy Science 76
(10):3054–3061. doi:http://dx.doi.org/10.3168/jds.S0022-0302(93)77645-6.
Tabilo-Munizaga, G., and G.V. Barbosa-Cánovas. 2005. “Rheology for the food industry.”
Journal of Food Engineering 67 (1–2):147–156. doi:10.1016/j.jfoodeng.2004.05.062.
Tadros, T. 2004. “Application of rheology for assessment and prediction of the long-term
physical stability of emulsions.” Advances in Colloid and Interface Science 108–109:
227–258. doi:10.1016/j.cis.2003.10.025.
Tan, Y., K. Xu, C. Liu, Y. Li, C. Lu, and P. Wang. 2012. “Fabrication of starch-based nano-
spheres to stabilize Pickering emulsion.” Carbohydrate Polymers 88:1358–1363.
Tcholakova, S., N.D. Denkov, I.B. Ivanov, and B. Campbell. 2002. “Coalescence in
β-lactoglobulin-stabilized emulsions: Effects of protein adsorption and drop size.”
Langmuir 18 (23):8960–8971. doi:10.1021/la0258188.
Tcholakova, S., N.D. Denkov, I.B. Ivanov, and B. Campbell. 2006. “Coalescence stability
of emulsions containing globular milk proteins.” Advances in Colloid and Interface
Science 123–126:259–293. doi:10.1016/j.cis.2006.05.021.
Tcholakova, S., N.D. Denkov, and A. Lips. 2008. “Comparison of solid particles, globular
proteins and surfactants as emulsifiers.” Physical Chemistry Chemical Physics 10 (12):
1608–1627. doi:10.1039/b715933c.
Tcholakova, S., N.D. Denkov, D. Sidzhakova, I.B. Ivanov, and B. Campbell. 2003.
“Interrelation between drop size and protein adsorption at various emulsification condi-
tions.” Langmuir 19 (14):5640–5649. doi:10.1021/la034411f.
Timgren, A., M. Rayner, P. Dejmek, D. Marku, and M. Sjöö. 2013. “Emulsion stabilizing
capacity of intact starch granules modified by heat treatment or octenyl succinic anhy-
dride.” Food Science & Nutrition 1 (2):157–171. doi:10.1002/fsn3.17.
Timgren, A., M. Rayner, M. Sjöö, and P. Dejmek. 2011. “Starch particles for food based
Pickering emulsions.” Procedia Food Science 1:95–103.
Ting, Y., Y. Jiang, C.-T. Ho, and Q. Huang. 2014. “Common delivery systems for enhancing
in vivo bioavailability and biological efficacy of nutraceuticals.” Journal of Functional
Foods 7:112–128. doi:10.1016/j.jff.2013.12.010.
Tufvesson, F., M. Wahlgren, and A.C. Eliasson. 2003. “Formation of amylose–lipid complexes
and effects of temperature treatment. Part 1. Monoglycerides.” Starch - Stärke 55 (2):61–71.
Tzoumaki, M.V., T. Moschakis, V. Kiosseoglou, and C.G. Biliaderis. 2011. “Oil-in-water emul-
sions stabilized by chitin nanocrystal particles.” Food Hydrocolloids 25 (6):1521–1529.
Tzoumaki, M.V., T. Moschakis, E. Scholten, and C.G. Biliaderis. 2013. “In vitro lipid diges-
tion of chitin nanocrystal stabilized o/w emulsions.” Food & Function 4 (1):121–129.
Verwey, E.J.W., and J.Th.G. Overbeek. 1948. Theory of the Stability of Lyophobic Colloids.
Amsterdam, the Netherlands: Elsevier.
Formulation of Emulsions 99
Wahlgren, M. 1995. “Removal of T4 lysozyme from silicon oxide surfaces by sodium dodecyl
sulfate.” Surface and Colloid Science Symposium, Lund, Sweden, October 18–19.
Wahlgren, M., and T. Arnebrant. 1991. “Protein adsorption to solid surfaces.” Trends in
Biotechnology 9 (1):201–208.
Wahlgren, M., and T. Arnebrant. 1992. “The concentration dependence of adsorption from a
mixture of b-lactoglobulin and sodium dodecyl sulfate onto methylated silica surfaces.”
Journal of Colloid and Interface Science 148:201–206.
Walstra, P. 2005. “8 Emulsions.” In Fundamentals of Interface and Colloid Science, edited by
J. Lyklema, 1–94. Academic Press.
Waninge, R., P. Walstra, J. Bastiaans, H. Nieuwenhuijse, T. Nylander, M. Paulsson, and
B. Bergenståhl. 2005. “Competitive adsorption between β-casein or β-lactoglobulin and
model milk membrane lipids at oil-water interface.” Journal of Agricultural and Food
Chemistry 53 (3):716–724.
Waraho, T., D.J. McClements, and EA. Decker. 2011. “Mechanisms of lipid oxidation in food disper-
sions.” Trends in Food Science and Technology 22 (1):3–13. doi:10.1016/ j.tifs.2010.11.003.
Watanabe, A., I. Tashima, V. Matsuzaki, J. Kurashige, and K. Sato. 1992. “On the formation
of granular crystals in fat blends containing palm oil.” Journal of the American Oil
Chemists’ Society 69:1077–1080.
Wilde, P., A. Mackie, F. Husband, P. Gunning, and V. Morris. 2004. “Proteins and emulsi-
fiers at liquid interfaces.” Advances in Colloid and Interface Science 108–109:63–71.
doi:10.1016/j.cis.2003.10.011.
Williams, P.A., G.O. Phillips, and R.C. Randall. 1990. “Structure-function relationships of
gum arabic.” In Gums and Stabilisers for the Food Industry, edited by G.O. Phillips,
D.J. Wedlock, and P.A. Williams, 25. Oxford: IRL.
Wongkongkatep, P., K. Manopwisedjaroen, P. Tiposoth, S. Archakunakorn, T. Pongtharangkul,
M. Suphantharika, K. Honda, I. Hamachi, and J. Wongkongkatep. 2012. “Bacteria
interface Pickering emulsions stabilized by self-assembled bacteria-chitosan network.”
Langmuir 28 (13):5729–5736. doi:10.1021/la300660x.
Wu, B., B. Degner, and D.J. McClements. 2013. “Microstructure & rheology of mixed colloidal
dispersions: Influence of pH-induced droplet aggregation on starch granule–fat droplet mix-
tures.” Journal of Food Engineering 116 (2):462–471. doi:10.1016/ j.jfoodeng.2012.12.020.
Xu, W., W. Jin, C. Zhang, Z. Li, L. Lin, Q. Huang, S. Ye, and B. Li. 2014. “Curcumin loaded
and protective system based on complex of κ-carrageenan and lysozyme.” Food
Research International 59:61–66. doi:10.1016/j.foodres.2014.01.059.
Yang, Y., and D.J. McClements. 2013. “Vitamin E bioaccessibility: Influence of carrier oil
type on digestion and release of emulsified alpha-tocopherol acetate.” Food Chemistry
141 (1):473–481. doi:10.1016/j.foodchem.2013.03.033.
Youssef, M.K., and S. Barbut. 2011. “Effects of two types of soy protein isolates, native and
preheated whey protein isolates on emulsified meat batters prepared at different protein
levels.” Meat Science 87 (1):54–60. doi:10.1016/j.meatsci.2010.09.002.
Yusoff, A., and B.S. Murray. 2011. “Modified starch granules as particle-stabilizers of oil-in-
water emulsions.” Food Hydrocolloids 25 (1):42–55. doi:10.1016/j.foodhyd.2010.05.004.
Zapico, P., M. de Paz, M. Medina, and M. Nuñez. 1999. “The effect of homogenization of
whole milk, skim milk and milk fat on nisin activity against Listeria innocua.” Inter-
national Journal of Food Microbiology 46: 151–157.
Zhang, B., Y.J. Chi, and B. Li. 2013. “Effect of ultrasound treatment on the wet heating
Maillard reaction between β-conglycinin and maltodextrin and on the emulsifying
properties of conjugates.” European Food Research and Technology 238 (1):129–138.
doi:10.1007/s00217-013-2082-y.
4 Particle-Stabilized
Emulsions
Malin Sjöö, Marilyn Rayner, and Marie Wahlgren
CONTeNTs
4.1 Introduction .................................................................................................. 101
4.2 Particles as Emulsifiers ................................................................................. 102
4.3 Stability of Particle-Stabilized Emulsions .................................................... 106
4.3.1 Coalescence ...................................................................................... 106
4.3.2 Ostwald Ripening ............................................................................. 108
4.3.3 Creaming or Sedimentation .............................................................. 109
4.4 Structure and Rheology of Pickering Emulsions ......................................... 109
4.5 Influence of Other Components in the System ............................................. 112
4.6 Food-Grade Particles for Emulsion Stabilization ......................................... 113
4.7 Additional Functional Properties of Particle-Stabilized Emulsions............. 118
References .............................................................................................................. 118
4.1 INTRODUCTION
One specific type of emulsion that has received increasing attention, especially
since the late twentieth century, is the particle-stabilized emulsion. This inter-
facial phenomenon was initially described by Walter Ramsden (1903) in the first
known publication on the topic. However, this type of emulsion is generally referred
to as Pickering emulsion, named after the second scientist to publish on the topic,
S.U. Pickering (1907). Both Ramsden and Pickering independently observed that
solid particles were able to stabilize the interface between two immiscible phases.
More papers were published on Pickering emulsion during the first decade of the
twenty-first century than during the entire twentieth century. The renewed interest
in Pickering emulsions can be related to the fact that these emulsions have a very
high long-term stability toward coalescence and Ostwald ripening. The ability to
101
102 Engineering Aspects of Food Emulsification and Homogenization
100,000
105 kT
103 kT
10,000 102 kT
Particle radius (nm)
1,000
100
10
1
0 30 60 90 120 150 180
Contact angle, θ (°)
FIGURe 4.1 Contour plot of equal detachment energy from a triglyceride–water interface
for various combinations of particle radii and contact angles.
Particle-Stabilized Emulsions 103
Particles can stabilize both oil-in-water (O/W) and water-in-oil (W/O) emulsions
and there are also examples where particles have been used to stabilize at least one
of the phases in double emulsions (Matos et al. 2013; Spyropoulos, Frasch-Melnik,
and Norton 2011). As is the case with many emulsions systems, phase inversion may
occur when the fraction of the continuous phase is decreased in the system beyond
a certain point. For example, as shown by Binks and Lumsdon (2000), the fraction
of water where catastrophic phase inversion from W/O to O/W occurs is strongly
dependent on the hydrophobicity of the particles, and the fraction required for this
inversion increases with increasing particle hydrophobicity. This is in line with the
theoretical assumption that the contact angle and thus the hydrophobicity will be
important for what type of emulsion is formed at a water–particle contact angle
Water Oil
Oil Water
Water Oil
Oil Water
drop drop
(a) (b)
FIGURe 4.2 The location of a particle at the interface between oil and water is determined
by the contact angle, θ, measured through the water phase. This also determines the type of
emulsion formed, (a) θ < 90° will favor O/W emulsions, whereas (b) θ > 90° will favor W/O
emulsions. (Redrawn from Rayner, M. et al., J. Sci. Food Agric., 92, 9, 1841–1847, 2012.)
104 Engineering Aspects of Food Emulsification and Homogenization
as measured through the water phase. A contact angle greater than 90° will favor
W/O emulsions, whereas a contact angle lesser than 90° will favor O/W emulsions,
as shown schematically in Figure 4.2. A high energy of desorption for larger par-
ticles means that particles far from a contact angle of 90° will also stabilize emul-
sions, providing a possibility to produce stable emulsions at a very high fraction of
the internal phase (Capron and Cathala 2013; Marku et al. 2012; Midmore 1998).
Such emulsions normally require a high amount of traditional surfactants, but, for
example, Capron and Cathala (2013) have shown that low amounts, less than 0.1% of
cellulose nanoparticles, could also stabilize O/W emulsions containing a dispersed
phase of 92%.
The droplet size of a Pickering emulsion will be governed by either the ratio
between the amount of particles and the dispersed phase volume or the energy avail-
able for emulsification. This means that above a critical concentration, where there
are enough particles to cover the dispersed phase, the droplet size of the emulsion
will decrease linearly with increasing particle-to-dispersed phase ratio, until a pla-
teau is reached where no further reduction in size can be obtained, due to the limita-
tion of the equipment used (Frelichowska, Bolzinger, and Chevalier 2010; Li, Sun,
and Yang 2012; Marku et al. 2012; Rayner, Sjöö et al. 2012; Timgren et al. 2011).
In the linear region, it is possible to theoretically estimate the amount of particles
needed to obtain a specific particle-stabilized emulsion droplet size. To do this, the
following assumptions are made: (1) that there is a strong preferential adsorption of
the particles into an ordered monolayer at the surface of the dispersed phase drop-
lets and (2) that the surface mean radius of the particle and droplet size are good
representatives of the system. These calculations were performed for quinoa starch
granules using a radius of 1.8 µm assuming a hexagonal close packing at the surface
using Equation 4.2 (see Figure 4.3).
250
200
Starch coverage (%)
150
100
50
0
0 200 400 600 800 1000 1200
Starch/oil ratio (mg/ml)
FIGURe 4.3 Starch coverage versus particle concentration for a Pickering emulsion stabi-
lized with quinoa starch particles of 1.8 µm and calculated assuming hexagonal close packing.
Particle-Stabilized Emulsions 105
d32p
mp = 4Voρpϕ (4.2)
d32d
where:
mp is the mass of particles (kg)
Vo is the volume of the dispersed phase (m3)
ρp is the particle density (kg/m3)
φ is the packing density (i.e., the assumed hexagonal close packing φ ≈ 0.907)
d32p and d32d are the diameters of the particle and emulsion drop, respectively, (µm)
The theoretical maximum coverage, ΓM (mg/m2), could be calculated using Equation 4.3.
2
Γ M = ρp d pϕ ⋅ 106 (4.3)
3
Although illustrative, there is no perfect correlation between the measured particle
size and the theoretical plotted in Figure 4.4, as data points lie both above and below
the theoretical line. Furthermore, if one calculates the actual surface coverage of
the droplets, it decreases with decreasing particle/oil ratios and is well below what
would be expected for a closed packed layer for low ratios. There has been an ongo-
ing discussion on whether a closed packed layer is necessary to stabilize Pickering
emulsions. It has also been shown that a surface coverage well below close packing
gives rise to stable Pickering emulsions that do not change drop size either due to
storage (8 weeks to several years) or upon mild centrifugation (Marku et al. 2012).
Horozov and Binks (2006) have suggested that the emulsions could be stabilized
by bridging particle monolayers at concentrations below the full surface coverage.
1
Measured drop surface area (m2/ml oil)
0.1
FIGURe 4.4 The measured drop surface area versus estimated drop surface area for
starch particle-stabilized emulsions. Larger particles tend to produce larger drop area
than predicted, although in the case of starches, the particle shape may also have an
influence. (Open access image from Timgren, A. et al., Food Sci. Nutr., 1, 2, 157–171,
2013.)
106 Engineering Aspects of Food Emulsification and Homogenization
Midmore (1998) showed that flocculation of silica particles could help the stabiliza-
tion at low surface coverage.
In a practical situation, one has to also consider to what extent the hydrophobicity
of particles will lead to particle–particle aggregation and thus several aggregated
particles act as single (larger) particles stabilizing an emulsion, thereby causing a
multilayer of particles, alternatively, a high degree of hydrophobicity may cause
aggregation of emulsion droplets. For example, Rayner, Sjöö et al. (2012) showed
that increasing the hydrophobicity of starch particles above a threshold value sta-
bilized O/W emulsions and led to an increase in droplet size; however, a decrease
in the amount of free starch not attached to the emulsion droplets was indicative of
particle aggregates adsorbed at the surface.
4.3.1 CoalesCenCe
Both classical surfactants and particles can stabilize emulsions from coalescence
over long periods of time at optimal conditions. However, the stability of traditional
surfactants is highly dependent on small droplet sizes whereas Pickering emulsions
can be stable for months or even years for droplet sizes of greater than 100 µm
(Timgren et al. 2013).
Particle-Stabilized Emulsions 107
FIGURe 4.5 The packing of particles between two particle-stabilized droplets. Left droplets
stabilized by monolayers, middle droplets stabilized by bridging particles, and right droplets
stabilized by individual particles and with aggregated particles in the continuous phase.
Coalescence occurs when a thin film of the continuous phase is drained from the
interface between two adjacent droplets. This leads to the nucleation of a pore, and
when this pore reaches a critical size, it starts growing rapidly and the two drops are
fused. Chen et al. (2013) have shown that the coalescence of Pickering emulsions can
occur via either the conventional route, where a liquid bridge grows continuously and
merges two droplets together, or via an oscillating mechanism, where bridges are
formed but fail to grow due to geometric restrictions until a stable bridge is formed
and the droplets merge. In Figure 4.5, the idealized packing of particles between two
droplets is described. If the energy of detachment for the particles is high, coalescence
will not occur due to a mechanism described by Denkov et al. (1992); Kruglyakov
and Nushtayeva (2004); and Tcholakova, Denkov, and Lips (2008). Contact between
the two discontinuous phases could occur through the menisci of the bare oil–water
interface. The stability of the film will, in this case, be the same as the maximum
capillary pressure drop, which can be resisted by the liquid menisci formed between
the adsorbed particles. The shape of the menisci depends on a number of factors
such as particle radius, antiparticle distance and configuration of the particle layer,
three-phase contact angle, oil–water interfacial tension, and the capillary pressure
across the fluid interface. This theory predicts that the maximum capillary pressure
that an emulsion can withstand prior to coalescence is higher (1) when particles have
a contact angle further away from 90°, (2) when the contact angle hysteresis is larger,
and (3) when the particles are small (Denkov et al. 1992). The size and contact angle
dependence is in the opposite direction from the factors that govern the detachment
energy from the surface, and thus theoretically there is an optimal particle size and
contact angle that will favor stable emulsions. To complicate things further, this will
be contingent on whether the particles at the surface form a mono or bilayer protec-
tion against coalescence (Denkov et al. 1992; Kruglyakov and Nushtayeva 2004;
Tcholakova, Denkov, and Lips 2008).
The ability to obtain the so-called partial coalescence is a specific character-
istic of particle-stabilized emulsions. In this case, two droplets start to merge but
instead of forming one larger droplet; the coalescence is arrested giving rise to
two merged droplets, sometimes obtaining peanut- or dumbbell-like shapes. This
occurs for emulsions that initially do not have the full surface coverage of the par-
ticles (Pawar et al. 2011), but when the droplets start to coalesce (thereby reducing
their total interfacial area), the particles do not detach. If the particle concentra-
tion is right, the merging of the drops will then lead to an increase in the surface
coverage. Given the right conditions, the particle concentration at the surface will
108 Engineering Aspects of Food Emulsification and Homogenization
increase until the particles form a cohesive structure that resists the Laplace pres-
sure, hindering further coalescence (Pawar et al. 2011). Partial coalescence will
be important for the structure and rheology of the emulsion. This has been well
known in food systems and has been reviewed in detail by Fredrick, Walstra, and
Dewettinck (2010).
d 3 8γ DC V
dt
( )
rc = ow sol m
9 RT
(4.4)
Some authors have reported that Ostwald ripening is low for Pickering emul-
sions, at least at high particle coverage of the droplets (Ashby and Binks 2000;
Juarez and Whitby 2012). It has been shown, however, that Ostwald ripening
can be the driving force for instability at a low surface coverage of the droplets,
but as the surface coverage increases, Ostwald ripening stops (Ashby and Binks
2000; Juarez and Whitby 2012). Juarez and Whitby (2012) showed that at a low
surface coverage of toluene films, the Ostwald ripening was higher than that
predicted by the LSW theory. This was attributed to the aggregation of droplets
that was seen to be favored by low surface coverage. As the particle content
increased, they saw a drop in the ripening rate attributed to both higher surface
coverage and less aggregation. The surface tension between the oil and water is
an important factor and both proteins and especially surfactants lower the sur-
face tensions considerably, thereby decreasing Ostwald ripening. As described in
detail in a review by Tcholakova, Denkov, and Lips (2008), the surface tension of
a Pickering emulsion can be divided into the surface tension of the bare surface
γ b, which is always the same for each individual system and the macroscopic
apparent surface tension γa , which is dependent on the particle coverage, and that
is used in Equation 4.4. As discussed by Tcholakova, Denkov, and Lips (2008),
when the surface coverage becomes high, for example, due to initial shrinkage
of the drops, Ostwald ripening stops, as the apparent interfacial tension, γa , and
the drop capillary pressure, PC = 2γA/R, approach zero.
Particle-Stabilized Emulsions 109
Another key component for Ostwald ripening is the solubility of the liquids
in each other. For example, triglyceride liquids have a rather low solubility in
water. However, the presence of micelles may increase the solubility of oil in
water. Drelich et al. (2012) showed that silica Pickering emulsions did not favor
transfer of oils between the emulsion droplets, but when a surfactant was added
in concentrations above the critical micelle concentration (CMC) of the surfac-
tant, oil was transferred between the emulsion droplets. Thus, one additional
advantage of Pickering emulsions compared to surfactant-stabilized emulsions
is that they do not change the apparent solubility of oil or an active substance in
the water phase.
thinning are also important for the consumer handling the product before using it,
for example, when being removed from its container.
All types of emulsions will become viscoelastic when they reach a certain level
of the dispersed phase where there is mechanical interaction between the droplets.
Mason, Bibette, and Weitz (1995) showed for surfactant-stabilized emulsions that
even very different emulsions systems follow the same dependence on the volume
fraction of the dispersed phase when the elastic modulus of the system is scaled to
γ/R. The scaled modulus have been seen to increase drastically at a volume fraction
around 0.63, which is equal to random close packing of solid spheres. Arditty et al.
(2004) showed that this behavior is not seen for Pickering emulsions. Instead the
emulsions follow the Mason curve if the scaling factor is the interfacial elasticity
and not the surface tension. This is due to the strong attachment of the particles to
the interface.
The excess particles can form a network in which the oil droplets are dis-
persed. In this case, the rheological properties of the gel will depend on both the
particle network as such, and on the dispersed phase droplets (Abend et al. 1998).
The dispersed phase will function as a filler in the particle network, giving rise to
a more elastic system. This behavior was shown to be formed by Aerosil 130 and
silicon oils (Sugita, Nomura, and Kawaguchi 2008) and by paraffin O/W emul-
sions stabilized by different bentonites and hectorites (Lagaly, Reese, and Abend
1999).
Pickering emulsions, where the adsorbed particles are able to form bridges
between the emulsion droplets, can have structures similar to flocculated particle
gels. Lee, Chan, and Mohraz (2012) showed that the rheology of these gels was
mainly influenced by the volume fraction of particles. In fact, systems with com-
pletely different oil-to-water ratios could have very similar rheological properties
as long as the particle ratio was the same. In contrast to surfactant-stabilized emul-
sions, these systems showed elastic properties even when the volume fraction of the
dispersed phase was below the random-close-packing limit for spheres. Therefore,
Pickering emulsions that are mechanically stable toward creaming at dispersed
phase volume fractions lower than the random-close-packing limit can be obtained.
Effects of particle aggregation causing droplet aggregation in Pickering emulsions
and increased elastic properties have also been observed both for starch particle-
stabilized emulsions (Rayner, Sjöö et al. 2012; Rayner, Timgren et al. 2012) and in
milk protein particles (Dickinson 2001).
Pickering emulsions stabilized by the same particles, that is, quinoa starch
granules, have been shown to have a weak gel behavior in a wide range of par-
ticle and oil concentrations (Figure 4.6). In a weak gel, the elastic modulus, G′,
is larger than the viscous modulus, G″; although being frequency dependent, the
gel strength can further be evaluated by the shear when the phase angle, δ, is
45° (Figure 4.7). As expected, the volume fraction of the dispersed phase mainly
determines the elastic properties and the stability of the gel (Timgren et al. 2011).
Marku et al. (2012) further showed that the viscoelastic behavior is influenced by
the properties of the dispersed phase, especially the melting point of the oil. Shear
can also be used in a controlled way to destabilize emulsions in specific applica-
tions (Whitby et al. 2011).
Particle-Stabilized Emulsions 111
10,000
G′
1,000
G″
100
G′, G″ (Pa)
10
0.1
0.01
0 10 20 30 40 50 60
Oil concentration (%)
FIGURe 4.6 Viscoelastic properties in terms of elastic (G′) and viscous (G″) modulus for
quinoa starch-stabilized emulsions with buffer and medium-chain triglyceride oil. Particle con-
centration 214 mg/ml oil except for at 40% oil: 530 mg/ml oil. (Data combined from different
measurements Marku, D. et al., Int. J. Pharm., 428, 1–2, 1–7, 2012; Rayner, M. et al., Faraday
Discuss., 158, 139–155, 2012; Sjöö, M. et al. Submitted manuscript (under review J. Colloid.
Interf. Sci.), 2014; Timgren, A. et al., Procedia Food Sc., 1, 95–103, 2011.)
1,000
G′
G″
100
Elastic modulus, G′ (Pa)
10
0.1
0.00001 0.001 0.1 10
γ * (strain)
FIGURe 4.7 Elastic (G′) and viscous (G″) modulus as a function of complex strain of qui-
noa starch-stabilized emulsion.
112 Engineering Aspects of Food Emulsification and Homogenization
TABLe 4.1
examples of Food-Based Particles Used as stabilizers for Pickering-Type
emulsions
Formulations/
Particle Type size and shape Applications References
Polysaccharides
Starch granules (quinoa, 0.5–20 µm in size Model systems of Dejmek et al. (2012);
rice, barley, and waxy and various O/W emulsions over Marefati et al. (2013);
maize) with various shapes (round, a wide range of oil Marku et al. (2012);
hydrophobic oblong, contents (5%–75%), Matos et al. (2013);
modifications (octenyl sharp-edged food emulsions, Rayner, Sjöö et al. (2012);
succinic anhydride polyhedral, etc.) salad dressings, Rayner, Timgren et al.
[OSA], or dry heating, depending on mayonnaise-type (2012); Rayner et al.
and in some cases botanical source emulsions, double (2014); Timgren et al.
without) emulsions, (2011, 2013)
encapsulation,
topical formulations,
cosmetics, and dried
emulsions (oil-filled
powders)
Starch nanocrystals Polygonal 50% paraffin O/W, Li et al. (2014);
structures sizes size of emulsions Miao et al. (2014)
40–100 nm drops containing
0.02–6.0 wt% starch
nanocrystals,
droplets were
33–13 µm,
respectively
Starch nanospheres Uniform O/W and W/O Tan et al. (2014)
spheres ~250 nm emulsions were
produced depending
on the water-to-oil
ratio. Using 1%
particles, emulsion
droplet sizes were
between 20–28 µm
(Continued)
114 Engineering Aspects of Food Emulsification and Homogenization
Proteins
Spray-dried soy protein Collapsed spheres O/W with oil phase Liu and Tang (2013);
isolate particles with (raisin shapes) fraction ranging Paunov et al. (2007)
CaP cores 2–10 µm from 0.1 to 0.9. At
6.4 wt% soy particles
emulsion drops
20–60 µm
(Continued)
Particle-Stabilized Emulsions 115
Miscellaneous
Egg yolk granules 0.3–2 µm Model O/W emulsions, Aluko and Mine (1997);
salad dressings-type Anton, Beaumal, and
emulsions, freeze- Gandemer (2000); Anton
thaw stable and Gandemer (1997);
mayonnaises, Ercelebi and Ibanoglu
cosmetics. Emulsion (2010); Eriksson (2013);
drop size of Laca, Paredes, and Diaz
10–70 µm depending (2012); Laca et al. (2010);
on the amount used Rayner et al. (2014)
Cocoa particles 2–5 µm 20% purified sunflower Gould, Vieira, and Wolf
oil, with 2%–10% (2013)
cocoa particles added.
Emulsion drops
~10 µm in diameter
Flavonoids (tiliroside, ~100 nm 20% oil-in-water in Luo et al. (2011, 2012)
rutin, and naringin) 1 mm flavonoids.
Emulsions drops
2–200 µm in
diameter depending
on the pH
Starch has received growing attention to be used as particles for the stabiliza-
tion of Pickering emulsions. In nature, a large variation in sizes and shapes of
starch granules can be found depending on the botanical origin. Most starches are
relatively large (greater than 15 µm or more) and thereby rather large and heavy
in order to stabilize Pickering emulsions (Timgren et al. 2013). However, there
are some plants producing extraordinarily small starch particles, which have been
shown to effectively stabilize emulsions (Timgren et al. 2011). Because starch is
hydrophilic, it is not likely to adsorb into the oil–water interface to a large extent,
although native starch has been used (Li, Li et al. 2013; Timgren et al. 2013). In
other studies, starch has been modified with octenyl succinic anhydride or ther-
mally treated using dry heat to obtain more hydrophobic character (Rayner, Sjöö
et al. 2012; Timgren et al. 2013; Yusoff and Murray 2011). Furthermore, the produc-
tion of starch-based nanospheres provides an additional source of starch particles
for particle stabilization (Tan et al. 2012). Modified starch has also previously been
used for a long time as an emulsifier, although in the molecular form, as reviewed
by Eliasson et al. (2013). In a Pickering emulsion, particles are mainly attached to
the drop surface, although free starch may be present in the system. If particles
are added to an emulsion already stabilized by surfactants, they are though less
likely to adsorb to the interface (see Figure 4.8). This would be the case also for
starch particles added to an emulsion stabilized with molecular starch. However,
an emulsion stabilized by starch particles can be heated, which then releases starch
molecules during the gelatinization process, although not destabilizing the emul-
sion (Marefati et al. 2013).
Fat crystals and solid lipid particles also have the potential of stabilizing
Pickering emulsion food systems such as whipped cream or spread. The mecha-
nisms of Pickering lipid crystal stabilization, fat crystal wetting, and the temperature
effects on fat crystal-stabilized emulsion have been previously reviewed (Ghosh and
Rousseau 2011; Rousseau 2013).
FIGURe 4.8 Particles at the drop interface and as free particles in the continuous phase
of O/W emulsions. (a) starch particle (0.5–15 µm) stabilized emulsion, (b) starch molecule
(40–400 nm) stabilized emulsion, (c) surfactant (1–10 nm) stabilized emulsion.
Particle-Stabilized Emulsions 117
(a) (b)
(c) (d)
FIGURe 4.9 Micrographs of emulsion droplets in mayonnaise stabilized with egg yolk
granules (a, b), quinoa starch granules (c), and liquid egg yolk (d), respectively. Scale bars
50 µm.
118 Engineering Aspects of Food Emulsification and Homogenization
ReFeReNCes
Abend, S., N. Bonnke, U. Gutschner, and G. Lagaly. 1998. “Stabilization of emulsions by het-
erocoagulation of clay minerals and layered double hydroxides.” Colloid and Polymer
Science 276 (8):730–737.
Aluko, R.E., and Y. Mine. 1997. “Competitive adsorption of hen’s egg yolk granule lipo-
proteins and phosvitin in oil-in-water emulsions.” Journal of Agricultural and Food
Chemistry 45 (12):4564–4570.
Anton, M., V. Beaumal, and G. Gandemer. 2000. “Adsorption at the oil–water interface and
emulsifying properties of native granules from egg yolk: Effect of aggregated state.”
Food Hydrocolloids 14 (4):327–335.
Anton, M., and G. Gandemer. 1997. “Composition, solubility and emulsifying properties of
granules and plasma of egg yolk.” Journal of Food Science 62 (3):484–487.
Arditty, S., V. Schmitt, J. Giermanska-Kahn, and F. Leal-Calderon. 2004. “Materials based on
solid-stabilized emulsions.” Journal of Colloid and Interface Science 275 (2):659–664.
Ashby, N.P., and B.P. Binks. 2000. “Pickering emulsions stabilised by Laponite clay par-
ticles.” Physical Chemistry Chemical Physics 2 (24):5640–5646.
Binks, B.P., P.D. Fletcher, M.A. Thompson, and R.P. Elliott. 2013. “Influence of propylene glycol
on aqueous silica dispersions and particle-stabilized emulsions.” Langmuir 29 (19):5723–33.
Binks, B.P., and S.O. Lumsdon. 2000. “Influence of particle wettability on the type and stabil-
ity of surfactant-free emulsions.” Langmuir 16:8622–8631.
Binks, B.P., J.A. Rodrigues, and W.J. Frith. 2007. “Synergistic interaction in emulsions
stabilized by a mixture of silica nanoparticles and cationic surfactant.” Langmuir 23
(7):3626–3636.
Capron, I., and B. Cathala. 2013. “Surfactant-free high internal phase emulsions stabilized by
cellulose nanocrystals.” Biomacromolecules 14 (2):291–296.
Chen, G., P. Tan, S. Chen, J. Huang, W. Wen, and L. Xu. 2013. “Coalescence of Pickering
emulsion droplets induced by an electric field.” Physical Review Letters 110 (6):064502.
De Folter, J.W.J., M.W.M. Van Ruijven, and K.P. Velikov. 2012. “Oil-in-water Pickering emul-
sions stabilized by colloidal particles from the water-insoluble protein zein.” Soft Matter
8 (25):2807–2815.
Dejmek, P., A. Timgren, M. Sjöö, and M. Rayner. 2012. New particle stabilized emulsions and
foams. WO Patent 2,012,082,065. December 15, 2011.
Denkov, N.D., I.B. Ivanov, P.A. Kralchevsky, and D.T. Wasan. 1992. “A possible mecha-
nism of stabilization of emulsions by solid particles.” Journal of Colloid and Interface
Science 150 (2):589–593.
Particle-Stabilized Emulsions 119
Kalashnikova, I., H. Bizot, P. Bertoncini, B. Cathala, and I. Capron. 2013. “Cellulosic nano-
rods of various aspect ratios for oil in water Pickering emulsions.” Soft Matter 9
(3):952–959.
Kalashnikova, I., H. Bizot, B. Cathala, and I. Capron. 2011. “New Pickering emulsions stabi-
lized by bacterial cellulose nanocrystals.” Langmuir 27 (12):7471–7479.
Kalashnikova, I., H. Bizot, B. Cathala, and I. Capron. 2012. “Modulation of cellulose nano-
crystals amphiphilic properties to stabilize oil/water interface.” Biomacromolecules 13
(1):267–275.
Kawazoe, A., and M. Kawaguchi. 2011. “Characterization of silicone oil emulsions stabilized
by TiO2 suspensions pre-adsorbed SDS.” Colloids and Surfaces A: Physicochemical and
Engineering Aspects 392 (1):283–287.
Kruglyakov, P., and A. Nushtayeva. 2004. “Emulsions stabilised by solid particles: The role of
capillary pressure in the emulsion films.” In Interface Science and Technology, edited by
D.N. Petsev, 641–676, Chapter 16. Elsevier, London.
Laca, A., B. Paredes, and M. Diaz. 2012. “Lipid-enriched egg yolk fraction as ingredient in
cosmetic emulsions.” Journal of Texture Studies 43 (1):12–28.
Laca, A., M.C. Saenz, B. Paredes, and M. Diaz. 2010. “Rheological properties, stability and
sensory evaluation of low-cholesterol mayonnaises prepared using egg yolk granules as
emulsifying agent.” Journal of Food Engineering 97 (2):243–252.
Lagaly, G., M. Reese, and S. Abend. 1999. “Smectites as colloidal stabilizers of emulsions.
II. Rheological properties of smectite-laden emulsions.” Applied Clay Science 14
(5–6):279–298.
Lee, M.N., H.K. Chan, and A. Mohraz. 2012. “Characteristics of Pickering emulsion gels
formed by droplet Bridging.” Langmuir 28 (6):3085–3091.
Levine, S., B.D. Bowen, and S.J. Partridge. 1989a. “Stabilization of emulsions by fine par-
ticles. I. Partitioning of particles between continuous phase and oil/water interface.”
Colloids and Surfaces 38 (2):325–343.
Levine, S., B.D. Bowen, and S.J. Partridge. 1989b. “Stabilization of emulsions by fine par-
ticles II. Capillary and van der Waals forces between particles.” Colloids and Surfaces
38 (2):345–364.
Li, C., Y. Li, P. Sun, and C. Yang. 2013. “Pickering emulsions stabilized by native starch
granules.” Colloids and Surfaces A: Physicochemical and Engineering Aspects
431:142–149.
Li, C., Y. Li, P. Sun, and C. Yang. 2014. “Starch nanocrystals as particle stabilisers of oil-in-
water emulsions.” Journal of the Science of Food and Agriculture 94 (9):1802–1807.
Li, C., P. Sun, and C. Yang. 2012. “Emulsion stabilized by starch nanocrystals.” Starch−Stärke
64 (6):479–502.
Li, Z., M. Xiao, J. Wang, and T. Ngai. 2013. “Pure protein scaffolds from pickering high inter-
nal phase emulsion template.” Macromolecular Rapid Communications 34 (2):169–174.
Lifshitz, I.M., and V.V. Slyozov. 1961. “The kinetics of precipitation from supersaturated solid
solutions.” Journal of Physics and Chemistry of Solids 19:35–50.
Liu, F., and C.H. Tang. 2013. “Soy protein nanoparticle aggregates as pickering stabilizers for
oil-in-water emulsions.” Journal of Agricultural and Food Chemistry 61 (37):8888–8898.
Luo, Z., B.S. Murray, A.L. Ross, M.J.W. Povey, M.R.A. Morgan, and A.J. Day. 2012. “Effects
of pH on the ability of flavonoids to act as Pickering emulsion stabilizers.” Colloids and
Surfaces B: Biointerfaces 92:84–90.
Luo, Z.J., B.S. Murray, A. Yusoff, M.R.A. Morgan, M.J.W. Povey, and A.J. Day. 2011. “Particle-
stabilizing effects of flavonoids at the oil-water interface.” Journal of Agricultural and
Food Chemistry 59 (6):2636–2645.
Marefati, A., M. Rayner, A. Timgren, P. Dejmek, and M. Sjöö. 2013. “Freezing and freeze-
drying of Pickering emulsions stabilized by starch granules.” Colloids and Surfaces A:
Physicochemical and Engineering Aspects 436 (5):512–520.
Particle-Stabilized Emulsions 121
Sjöö, M., S. Cem Emek, T. Hall, M. Rayner, and M. Wahlgren. 2014. “Barrier properties of
heat treated starch Pickering emulsions.” Submitted manuscript (under review Journal
of Colloid and Interface Science). doi:10.1016/j.jcis.2015.03.004.
Spyropoulos, F., S. Frasch-Melnik, and I.T. Norton. 2011. “W/O/W emulsions stabilized by
fat crystals—Their formulation, stability and ability to retain salt.” In 11th International
Congress on Engineering and Food, edited by G. Saravacos, P. Taoukis, M. Krokida,
V. Karathanos, H. Lazarides, N. Stoforos, C. Tzia, and S. Yanniotis, 1700–1708.
Amsterdam, the Netherlands: Elsevier Science, B.V.
Sugita, N., S. Nomura, and M. Kawaguchi. 2008. “Rheological and interfacial properties of
silicone oil emulsions stabilized by silica particles.” Journal of Dispersion Science and
Technology 29 (7):931–936.
Tan, Y., K. Xu, C. Liu, Y. Li, C. Lu, and P. Wang. 2012. “Fabrication of starch-based nano-
spheres to stabilize pickering emulsion.” Carbohydrate Polymers 88 (4):1358–1363.
Tan, Y., K. Xu, C. Niu, C. Liu, Y. Li, P. Wang, and B.P. Binks. 2014. “Triglyceride-water emul-
sions stabilised by starch-based nanoparticles.” Food Hydrocolloids 36:70–75.
Tasset, S., B. Cathala, H. Bizot, and I. Capron. 2014. “Versatile cellular foams derived from
CNC-stabilized Pickering emulsions.” RSC Advances 4 (2):893–898.
Tcholakova, S., N.D. Denkov, and A. Lips. 2008. “Comparison of solid particles, globu-
lar proteins and surfactants as emulsifiers.” Physical Chemistry Chemical Physics 10
(12):1608–1627.
Timgren, A., M. Rayner, P. Dejmek, D. Marku, and M. Sjöö. 2013. “Emulsion stabilizing
capacity of intact starch granules modified by heat treatment or octenyl succinic anhy-
dride.” Food Science & Nutrition 1 (2):157–171.
Timgren, A., M. Rayner, M. Sjöö, and P. Dejmek. 2011. “Starch particles for food based
Pickering emulsions.” Procedia Food Science 1:95–103.
Torres, L.G., R. Iturbe, M.J. Snowden, B.Z. Chowdhry, and S.A. Leharne. 2007. “Preparation
of o/w emulsions stabilized by solid particles and their characterization by oscillatory
rheology.” Colloids and Surfaces A: Physicochemical and Engineering Aspects 302
(1–3):439–448.
Tzoumaki, M.V., T. Moschakis, V. Kiosseoglou, and C.G. Biliaderis. 2011. “Oil-in-water emul-
sions stabilized by chitin nanocrystal particles.” Food Hydrocolloids 25 (6):1521–1529.
Tzoumaki, M.V., T. Moschakis, E. Scholten, and C.G. Biliaderis. 2013. “In vitro lipid diges-
tion of chitin nanocrystal stabilized o/w emulsions.” Food & Function 4 (1):121–129.
Vignati, E., R. Piazza, and T.P. Lockhart. 2003. “Pickering emulsions: Interfacial tension, col-
loidal layer morphology, and trapped-particle motion.” Langmuir 19 (17):6650–6656.
Wagner, C. 1961. “Theorie der Alterung von Niederschlägen durch Umlösen (Ostwald-
Reifung).” Z. Elektrochem 65:581–591.
Whitby, C.P., F.E. Fischer, D. Fornasiero, and J. Ralston. 2011. “Shear-induced coalescence
of oil-in-water Pickering emulsions.” Journal of Colloid and Interface Science 361
(1):170–177.
Yang, F., S. Liu, J. Xu, Q. Lan, F. Wei, and D. Sun. 2006. “Pickering emulsions stabilized
solely by layered double hydroxides particles: The effect of salt on emulsion formation
and stability.” Journal of Colloid and Interface Science 302 (1):159–169.
Yusoff, A., and B.S. Murray. 2011. “Modified starch granules as particle-stabilizers of oil-in-
water emulsions.” Food Hydrocolloids 25 (1):42–55.
Zhu, Y., L.H. Lu, J. Gao, Z.G. Cui, and B.P. Binks. 2013. “Effect of trace impurities in triglyc-
eride oils on phase inversion of Pickering emulsions stabilized by CaCO3 nanoparticles.”
Colloids and Surfaces A: Physicochemical and Engineering Aspects 417:126–132.
Section II
High-Energy Processes
5 Droplet Breakup
in High-Pressure
Homogenizers
Andreas Håkansson
CONTeNTs
5.1 Introduction .................................................................................................. 125
5.2 The High-Pressure Homogenizer ................................................................. 126
5.3 An Historical Perspective on Fragmentation ................................................ 128
5.4 Fragmentation and Stabilization Mechanisms ............................................. 129
5.4.1 Fragmentation by Laminar Shear ..................................................... 129
5.4.2 Fragmentation by Turbulence ........................................................... 132
5.4.3 Fragmentation by Cavitation ............................................................ 136
5.4.4 Stabilizing Stresses and Comparisons .............................................. 137
5.5 Influence of Operating Parameters on Emulsification Result ...................... 139
5.5.1 Homogenizing Pressure .................................................................... 139
5.5.2 Second-Stage Effect and Thoma Number ........................................ 140
5.5.3 Viscosity of the Dispersed and Continuous Phases .......................... 141
5.5.4 Volume Fraction of Oil ..................................................................... 142
5.6 The Location of Drop Breakup .................................................................... 144
5.7 Conclusions on the Fragmentation Mechanisms in the HPH ....................... 145
References .............................................................................................................. 146
5.1 INTRODUCTION
The HPH is used for size reduction or the disintegration of dispersed particles such
as cells, macromolecules, or emulsion drops. The by far largest application is the size
reduction of emulsion drops, which will be the focus of this chapter.
125
126 Engineering Aspects of Food Emulsification and Homogenization
Impact
Gap ring
Forcer Gap
entrance exit
Outlet
re Inlet Gap chamber
ri chamber
Seat
FIGURe 5.1 Schematic, not to scale, representation of the high-pressure homogenizer valve.
(Note that the gap height is greatly exaggerated.)
ΔP, is set by adjusting the force applied on the forcer, which, in turn, sets the gap
height. Homogenization pressures are usually in the range of 5–40 MPa for food
applications, such as dairy processing of milk, but can be above 100 MPa for special
applications, such as cell breakage (Middelberg, 1995) or the disruption of macro-
molecules (Floury et al., 2002).
Because the gap height is small (order of magnitude 100 µm) and set implicitly
by the homogenizing pressure, it is generally unknown. The same is true for the
gap velocity, Ug, another important parameter for discussing fragmentation. Because
direct measurements are often not possible, gap height and velocity are often esti-
mated from semiempirical correlations such as (Phipps, 1975)
2 7/ 5 2
1 Q 5ρCν3C/ 5 Q 1 1 1 Q
∆P = + r 2 / 5 − r 2 / 5 + 2 2πr h (5.1)
4 2πri h h 3 2π i e e
where:
ri and re are the gap inlet and exit radii, respectively (see Figure 5.1)
ρC is the continuous phase density
νC is the continuous phase kinematic viscosity
The hydrodynamic description in Equations 5.1 and 5.2 is vital for discussing frag-
mentation as it constitutes a link between the practically accessible parameters such
as Q and ΔP, and the more fundamental parameters h and Ug.
128 Engineering Aspects of Food Emulsification and Homogenization
350 400
300 350
300
250
Ug (m/s)
250
h (μm)
200
200
150
150
100 100
50 50
20 40 60 80 100 20 40 60 80 100
(a) ΔP (MPa) (b) ΔP (MPa)
FIGURe 5.2 Gap height (a) and gap velocity (b) for a production-scale HPH as functions of
homogenizing pressure. See text for parameter values.
The fat and the casein particles in being forced through the capillary tubes or orifice …
are either partially broken up or elongated into capillary filaments, and … are mashed
or squeezed, so far as to completely disintegrate them.
Gaulin, 1904: 2
fragmentation of drops. This theory was disproved by noting that the compressibility
of liquids such as water and oil is very low. Even at a relatively high pressure differ-
ence such as 100 MPa, the volumetric change is just about 5%.
Fragmentation from impingement, that is, when drops of high velocity hit the
impact ring in the outlet chamber, has been another suggestion. It is now known
that a jet is created as the fluid exits the gap; however, the velocity declines rap-
idly as a function of distance to the gap exit, and drop velocities are relatively low
when reaching that far unless the impact ring distances are very short (Innings and
Trägårdh, 2007). Furthermore, fragmentation visualizations have shown breakup to
occur long before the impact ring (Innings, Fuchs, and Trägårdh, 2011). Thus, direct
impact is not responsible for the observed fragmentation.
∂vx
=0
∂x
∂vx
=G
∂y
FIGURe 5.3 Illustration of a spherical particle placed in a shear field and definition of
simple shear, G.
130 Engineering Aspects of Food Emulsification and Homogenization
(Saffman, 1965). If the particle is liquid, on the other hand, three scenarios can occur
(Grace, 1982; Stone, 1994):
1. The drop deforms and elongates due to viscous drag on the interface.
2. The liquid inside the drop starts to rotate.
3. The drop in itself starts to rotate.
The degree of deformation increases with shear, G, but is countered by the interfacial
tension, as any deviations from the spherical shape will increase the surface pressure
and thus the free energy of the system. The extent of deformation is often described
in terms of a capillary number:
2Gµ C d
Ca = (5.3)
γ
where:
γ is interfacial tension
It will become increasingly favorable for the drop to break into two if it becomes
sufficiently elongated. One can show that this occurs approximately when the drop is
twice as long as it is thin, if only the surface energy is taken into consideration. Thus,
after it has been sufficiently elongated, it is expected to break.
Experimental studies of the deformation and breakup of fluid drops placed in
static shear fields were initiated by Taylor in the 1930s. An early conclusion from
this work was that the extent of deformation and point of breakage can be described
in terms of the capillary number. Furthermore, the critical value depends on the
dispersed-to-continuous phase viscosity ratio and the type of shear flow (Taylor,
1934). For a simple shear flow (i.e., a two-dimensional flow with a linear velocity
gradient in one direction), it has been shown experimentally that the critical capil-
lary number tends to infinity as the viscosity ratio tends to 4 (Grace, 1982), implying
that medium-to-high viscosity oils never break from laminar simple shear. However,
it should be noted that no such limiting viscosity ratio has been found for extensional
shear or the more general shear fields to which the HPH flow can be classified. Thus,
laminar shear could not be disregarded as a breakup mechanism per se.
A strong laminar shear can be found in the inlet chamber and boundary layers of
the gap in the HPH valve (Håkansson et al., 2011). It is the acceleration created from
the reducing flow through area when approaching the gap that produces high shear
in the inlet chamber. Experiments and hydrodynamic modeling has shown that the
maximum shear rate in the gap center can be estimated from (Håkansson et al., 2011,
2012; Innings and Trägårdh, 2007)
Ug
max(G) ≈ (5.4)
2h
The total pressure loss will be dominated by the expansion loss (i.e., third term in
Equation 5.1) for a production-scale HPH:
Droplet Breakup in High-Pressure Homogenizers 131
ρCUg2
∆P ≈ (5.5)
2
By substituting Equations 5.2 and 5.5 in Equation 5.4, the maximum shear experienced
by a drop in the inlet chamber can be shown to be proportional to the homogenizing
pressure:
re 2π
max(G) ≈ ∆P (5.6)
QρC
The fragmenting stress, σfrag, on a drop from a shear rate, G, can then be calculated from
σfrag = µ CG (5.7)
However, drops will only experience this high stress for a short period of time, just
before entering the gap. When taking into account the deformation timescale,
µD
τdef = (5.8)
σfrag
The deformation timescale is the minimal time required for deforming and break-
ing a drop. Walstra (1983) argued that only low viscosity drops could ever be frag-
mented by laminar shear, as it is reasonable to assume that the time spent in the high
shear region, τ, scales inversely with G and thus,
µD
τ > τdef => < C1 (5.9)
µC
x νC
δ( x ) = 5 (5.10)
Ug x
The boundary layers will merge if the gap length, L g = re − ri, is long enough to give
δ(x = L g) = h/2, that is,
Q h
Lg ≥ (5.11)
2πreν C 100
132 Engineering Aspects of Food Emulsification and Homogenization
200 0.035
0.03
150 0.025
Lg (mm)
Lg (mm)
0.02
100
0.015
50 0.01
0.005
0 0
5 10 15 20 5 10 15 20
(a) Q (m3/h) (b) Q (L/h)
FIGURe 5.4 Critical gap length, L g, for the laminar boundary layers in the gap to extend
into the center of the gap as a function of volumetric flow at a constant homogenizing pres-
sure (ΔP = 30 MPa). (a) Production-scale homogenizer (re = 16 mm) and (b) laboratory-scale
HPH (re = 3 mm).
Figure 5.4 illustrates the critical gap length needed for boundary layers to merge for a
production and a laboratory-scale homogenizer when fed with varying volumetric flow
rates, Q, while holding the homogenizing pressure constant at 30 MPa. For a production-
scale homogenizer (re = 16 mm), the gap would need to be approximately 50 mm long,
which is roughly 50 times longer than the actual gap length, whereas the boundary layers
would merge already after approximately 10 µm into the gap for a laboratory-scale HPH
(which is a short distance when compared to an actual length, close to 1 mm).
This analysis of boundary layers implies that laminar shear can influence emulsifica-
tion inside the gap for small-scale homogenizers, but not in the case of production-scale
HPHs. This could explain why experimental investigations often see different behavior
between production and laboratory homogenizers (e.g., Walstra and Smulders, 1998
and references therein).
1/ 4
ν3
η= C (5.12)
ε
Droplet Breakup in High-Pressure Homogenizers 133
where:
ε is the rate of dissipation of turbulent kinetic energy, which under static condi-
tions corresponds to the rate of energy input into the system
The rate of dissipation of turbulent kinetic energy can therefore be used as a measure
of the energy available for fragmentation in a turbulent flow.
Turbulent eddies of different length scales influence drops differently, which
makes it important to describe the relative amount of turbulent energy for eddies
of different sizes. This can be done with a spectrum of turbulent kinetic energy,
E(l), describing the contribution of eddies of length scale, l, to the total turbu-
lent kinetic energy. A standard Kolmogorov model spectrum (Pope, 2000) can
be seen in Figure 5.5 for a turbulent flow with ε = 109 m 2/s3, νC = 10−6 m 2/s, and
h = 150 µm (i.e., an oil-in-water emulsion in a production-scale HPH). Most of the
energy is contained in the larger scales close to l = h and eddies smaller than η
contain very little energy. Relevant drop sizes in a production-scale HPH are often
close to 1 µm; a reasonable interval of drop sizes has been inserted in Figure 5.5
as a comparison.
Fragmentation of a drop due to interactions with turbulent eddies was first
described theoretically by Kolmogorov (1949) and was then further developed by
Hinze (1955). The fragmentation is often described in terms of two limiting cases,
or more accurately as two mechanisms: turbulent inertial and turbulent viscous
fragmentation.
h η
E(l)
102 100
l (μm)
FIGURe 5.5 Turbulent spectra showing the turbulent kinetic energy as a function of eddy
length scales. Parameters chosen from the turbulent flow in the outlet chamber of a production-
scale HPH. Range of drop sizes has been inserted as comparison. (Note the logarithmic
scales and inverted horizontal axis.)
134 Engineering Aspects of Food Emulsification and Homogenization
(a) (b)
FIGURe 5.6 Schematic illustration of interactions between drops and turbulent eddies
under turbulent inertial (a) and turbulent viscous (b) drop breakup. (Reprinted from J. Colloid
Interface Sci., 312, Vankova, N. et al., Emulsification in turbulent flow 1. Mean and maxi-
mum drop diameters in inertial and viscous regimes, 363–380, Copyright 2007, with permis-
sion from Elsevier.)
Inertial fragmentation originates from pressure fluctuations over the length scale
of the drop due to interactions with small turbulent eddies (i.e., l < d); see illustration
in Figure 5.6a. The stress on the drop can be estimated from
ρC uu
σfrag = d
(5.13)
2
where:
uu d is the average squared velocity fluctuation for eddies of length scales smaller
than the drop (l < d)
Emulsion drop sizes often have diameters of magnitude 1 µm when entering the
homogenizer. As seen in Figure 5.5, the drops will be in the far left region known
as the inertial subrange. The spectrum in this region can be approximated by (Pope,
2000)
E( κ) = C2 ⋅ ε2 / 3 ⋅ κ −5/ 3 (5.14a)
2π
κ= (5.14b)
l
and thus
∞
uu d
=
∫ E ( κ )d κ = 2 ε
2π/ d
2/3 2/3
d (5.15)
laminar shear (c.f. Figures 5.3 and 5.6b). Thus, the stress on the drop under this
mechanism can be calculated from
σfrag = µ CGd (5.16)
where:
Gd is the turbulent shear created by eddies larger than the drop (l > d)
uu d
Gd = (5.17)
d
which is somewhat contradictory as uu d relates to eddies smaller than d and Gd to
eddies larger than d. However, as uu d increases with eddy size and Gd decreases
with eddy size, it could be argued that both are dominated by eddies of sizes close
to d. This can serve as a motivation for using Equation 5.17 as a first approximation
when a more refined model is unavailable.
Estimation of fragmenting stresses based on Equations 5.15 and 5.17 require an
estimation of the dissipation rate, ε. Here, it should be pointed out that the dissipation
rate varies considerably throughout the valve geometry. Because the full flow field
is rarely available, a mean efficient value, ε, is often calculated instead and seen as
characteristic of the process:
Ug3
ε = C3 (5.18)
h
with C3 = 1/20 (Mohr, 1987) or 1/80 (Innings and Trägårdh, 2007). Attempts have
been made in calculating local values of ε using computational fluid dynamics (e.g.,
Stevenson and Chen, 1997; Kleinig and Middelberg, 1997; Floury, Belletre et al.,
2004); however comparisons to experiments show that none of the proposed models
are able to describe the intense turbulent field accurately (Håkansson et al., 2012).
It should also be remembered that in addition to the spatial distribution of tur-
bulence, there is a stochastic component to the velocity fluctuations at each posi-
tion. The fragmenting stresses in Equations 5.13 and 5.16 are based on the average
velocity fluctuation. However, the stochastic nature of turbulence implies that much
more intense eddies exist temporarily. Because these eddies will give rise to faster
and more efficient breakup, the spread in the eddy energy should also be taken into
consideration, this was pointed out rather early by Kolmogorov (1949). However,
it is still not known if this has a significant effect on drop fragmentation in HPHs.
An important consequence of the Kolmogorov–Hinze theory is that an eddy of
length scale close to the drop diameter is most efficient in breaking the drop. This is
interesting in relation to Figure 5.5, which shows that most of the turbulent kinetic
energy is in eddies of significantly larger sizes. Thus, turbulent breakup of drops
does not occur where the total turbulent kinetic energy is highest but where the eddy
length scales are most efficiently distributed in comparison to drop sizes.
136 Engineering Aspects of Food Emulsification and Homogenization
where:
PV(T) is the vapor pressure at fluid temperature T
P∞ is the upstream static pressure
V∞ is a characteristic fluid velocity
Cavitation generally occurs when NC falls below a critical value and increase in
intensity with decreasing cavitation number (Brennen, 1995). Cavitation num-
ber decreases with increasing homogenizing pressures due to the increase in fluid
velocity.
There is ample experimental evidence showing cavitation taking place in the HPH.
Cavitation has been studied using a variety of techniques based on the different
properties of a cavitating flow, such as ultrasonic emissions (Håkansson et al., 2010;
Kurzhals, 1977), free radical formation (Floury, Legrand, and Desrumaux, 2004;
Shirgaonkar, Lothe, and Pandit, 1998), wear (Innings, Hultman et al., 2011; Phipps,
1985), and light scattering of vapor bubbles (Håkansson et al., 2010; McKillop
et al., 1955).
Droplet Breakup in High-Pressure Homogenizers 137
It is also known that cavitation can have an emulsifying effect as illustrated, for
example, in ultrasonic emulsification. This has led investigators to suggest cavita-
tion to be the dominant mechanism of drop fragmentation during high-pressure
homogenization (Kurzhals, 1977; Loo, Slatter, and Powell, 1950; McKillop, 1955);
see Section 5.5.2 for a more comprehensive discussion.
Due to the complexity of the cavitation process and interaction with the drop
interface, quantitative descriptions of the cavitation mechanisms, for example, in
terms of stresses, is not yet available.
Turning to the location of cavitation in the HPH valve, the gap inlet with the high
local velocity created close to the separation zone is a likely candidate (Phipps, 1974).
Cavitation visualizations agree with this and show cavitation bubbles in the beginning
or inside the gap (Håkansson et al., 2010; Phipps, 1974). These experiments indicate
the occurrence of bubble implosions before exiting into the outlet chamber, which is
also supported by comparisons of stability times of cavitation bubbles obtained from
bubble dynamics simulations with hold-up times (Håkansson et al., 2010).
An opposing view that cavitation extends into the outlet chamber has gained sup-
port by findings of intense wear in this region (Innings, Hultman et al., 2011).
4γ
σstab,1 = (5.20)
d
Second, the deformation of the drop is countered by the viscous forces; more energy
is required to deform a more viscous drop. Hinze (1955) suggested
µD σfrag
σstab,2 = (5.21)
d ρD
where:
ρD is the dispersed phase density
µD is the dispersed phase kinematic viscosity
This formulation has gained experimental support from a large number of studies
(Calabrese, Chang, and Dang, 1986; Davies, 1985; Vankova, Tcholakova, Denkov,
Ivanov et al., 2007).
The total stabilizing stress can be obtained by the summation of Equations 5.20 and
5.21. A comparison of stresses can be seen in Figure 5.7. Fragmentation is assumed
to be due to the turbulent mechanisms with a dissipation rate relevant for production-
scale high-pressure homogenization (ε = 109 m2/s3). Figure 5.7a compares the total
stabilizing and fragmenting stresses for spherical drops of different diameters for a
138 Engineering Aspects of Food Emulsification and Homogenization
1010
109
108
107
σ (Pa)
106
105
σstab = σstab,1 + σstab,2
104 σstab,1
σfrag
103 −2
10 10−1 100 101
(a) d (μm)
107
106
σ (Pa)
105
104
0 10 20 30 40 50
(b) μD (Pas)
FIGURe 5.7 (a) Illustration of fragmenting (turbulent inertial and viscous at ε = 109 m2/s3)
and stabilizing forces for µD = 30 mPas. (b) Stabilizing stress as a function of dispersed phase
viscosity for d = 20 nm (µC = 1 mPas, γ = 20 mN/m, ρD = 700 kg/m3).
viscous drop (µD = 30 mPas, µC = 1 mPas). The largest stable drop size is roughly
150 nm in diameter for this combination of parameters; that is, a drop size distribu-
tion being subjected to these hydrodynamic conditions for a sufficient time period
and without coalescence would be transformed toward a monodisperse size distribu-
tion with diameter 150 nm.
Droplet Breakup in High-Pressure Homogenizers 139
The Laplace pressure has also been included in the figure to show that it is
insufficient for stabilizing the drop by itself even if drops are relatively large.
Figure 5.7b displays the stabilizing stresses as functions of dispersed phase viscosity.
As seen in the figure, the resistance from viscous stabilizing stresses (Equation 5.21)
becomes important already at rather low dispersed phase viscosities.
d32 = C4 ⋅ ∆P q (5.22)
The constant C4 varies between homogenizers and the constant q is often found close
to −0.6 for pilot and production-scale homogenizers for low volume fractions of oil
(Phipps, 1985; Walstra and Smulders, 1998).
The experimentally obtained value of the constant q has been seen as an indi-
cation for turbulent fragmentation ever since Walstra (1969) pointed out that the
combination of Kolmogorov’s fragmentation theory with a reasonable scaling of
homogenizing pressure and dissipation rate of turbulent kinetic energy, ε, implies
that q = −0.6. However, as was pointed out by Phipps (1985), this is not a conclusive
evidence as it may very well be that other fragmentation mechanisms (e.g., cavita-
tion) could also act in accordance with q = −0.6. However, the higher q values often
found for laboratory-scale HPHs do imply that the fragmentations in these HPHs are
dominated by another mechanism (Walstra and Smulders, 1998).
In the last couple of years, it has also become possible to measure fragmenta-
tion rates directly (Vankova, Tcholakova, Denkov, Vulchev et al., 2007). The results
show that the experimentally obtained fragmentation rates are consistent in size and
scaling with turbulent breakup, as discussed in Section 5.4.2. This further supports
turbulence as the dominant mechanism of breakup.
140 Engineering Aspects of Food Emulsification and Homogenization
6φD γ 1 1
ηHPH = − (5.23)
∆Pp dend d0
where:
ϕD is the dispersed phase volume fraction
In other words, the efficiency is higher when obtaining smaller drops using the same
pressure or the same drop size when using a lower pressure difference. In discuss-
ing efficiency, it is important to properly define the term pressure difference. Many
HPHs are designed with two valves, or stages, coupled in series, as indicated in
Figure 5.8. The fluid enters the HPH at a pressure P0, which is often somewhat higher
than the atmospheric pressure due to feed pumps. The piston pump (or pumps) in the
HPH increases the pressure to a high pressure, P1, at which the fluid enters the first
valve. The back pressure to the first valve, P2, is set by adjusting the gap height in the
second valve. The fluid exits the HPH at pressure P3. It is important to note that it is
the pump pressure difference, ΔPp, that needs to be supplied to the system, whereas
it is the pressures over the gaps, ΔP1 and ΔP2, that control fragmentation. However,
if the upstream and downstream pressures of the HPH are assumed to be constant
(P3 = P0), then the pump homogenizing pressure is equal to the total pressure loss
over both stages. This explains why the pump pressure difference is often used to
describe valve pressure in discussing HPH droplet fragmentation.
The use of two consecutive valves is based on experiments reporting higher efficiency
for dual stages. The two-stage settings are often described using the Thoma number:
P2
Th = (5.24)
P1
Optimum efficiency is often reported in the region Th = 0.1 − 0.2 (Kurzhals, 1977;
Mohr, 1987; Pandolfe, 1982). This is interesting in relation to the fragmentation
mechanisms. Dividing the total pressure loss in two parts is expected to give lower
P0 P1 P2 P3 ≈ Patm
Pump Stage 1 Stage 2
FIGURe 5.8 Schematic representation of pressures and pressure differences in the HPH.
Droplet Breakup in High-Pressure Homogenizers 141
laminar and turbulent stresses and thus larger drops and lower efficiency. Thus, if
laminar shear and/or turbulence dominated breakup, optimum efficiency would be
expected at Th = 0, corresponding to a single-stage configuration. However, using
a second valve also alters the absolute pressure in the outlet chamber. Although the
laminar and turbulent mechanisms only depend on the pressure difference over the
gap, cavitation is influenced by absolute pressure levels (see Equation 5.19). The
presence of an optimum efficiency as a function of Thoma number is, therefore, often
seen as evidence of cavitation controlling drop breakup. Studies of cavitation inten-
sity in HPHs show that the amplitude of a cavitation-induced ultrasound is affected
by the Thoma number (Håkansson et al., 2010; Kurzhals, 1977). Kurzhals (1977)
found that the Thoma number giving maximum efficiency also corresponded to the
most efficient emulsification, which supports this theory. A completely opposite view
is that the optimal Thoma number is obtained when suppressing cavitation in order
for it to not interact unfavorably with the turbulence (SPX, 2008). This is not an
implausible explanation as the presence of cavitation bubbles could influence the
structure and intensity of the turbulence (e.g., Iyer and Ceccio, 2002) and the sup-
pression of cavitation at sufficiently high Thoma numbers has been discussed theo-
retically and has been established experimentally (Håkansson et al., 2010; Kurzhals,
1977; Phipps, 1974).
A third explanation for the observed maximum, which does not rely on cavitation,
is based on observations of clustering in emulsions (Ogden, Walstra, and Morris,
1976). This has led investigators to propose that the second stage is needed in order
to break the clusters formed after the first stage. However, this fails to explain why
emulsions extracted after the first stage also seem to indicate optimal efficiency at
intermediate Thoma numbers (Loo and Carleton, 1953).
d32 = C5 ⋅µ Dp (5.25)
where:
p = 0.2 − 0.4 (Calabrese, Chang, and Dang, 1986; Kolb, 2001; Phipps, 1975).
Different suggestions have been made for explaining this value of p. Calabrese,
Chang, and Dang (1986) derived an expression for the limiting case when vis-
cous stresses dominate over surface forces in stabilizing a drop and obtained
p = 3/8. Walstra (1983), on the other hand, attributed the effect of the dispersed
phase viscosity to the timescale needed for drop deformation and breakup (c.f.
Equation 5.8). In this view, the fragmentation of more viscous drops is hin-
dered due to a lack of available time. Walstra and Smulders (1998) derived
p = 3/4 from a deformation timescale that limited drop fragmentation
142 Engineering Aspects of Food Emulsification and Homogenization
decreases with increasing volume fraction of oil (Phipps, 1985). Care must be taken
when interpreting these results. If the total amount of emulsifier is held constant
when increasing the volume fraction of the dispersed phase, this corresponds to a
decrease in the concentration of emulsifier available per volume of oil, and thus
there are two overlapping factors influencing the result: an increase in the amount of
dispersed phase and a decrease in the relative amount of emulsifier. The two factors
must be discussed separately.
Assuming a constant emulsifier-to-oil ratio, the volume fraction of oil is
expected to have two main effects on the fragmentation process. First, the number
of emulsion drops increases with increasing volume fraction of oil, which in turn
is expected to give rise to an increase in the rate of collisions and coalescence.
Coalescence can be described as a second-order process as it requires collisions
between pairs of drops, whereas fragmentation can be described as a first-order
process. Coalescence rate, therefore, increases faster than fragmentation rate with
increasing volume fraction of oil. Specific methods for measuring coalescence rate
during emulsification have been developed and confirm a fast increase in the rate
of coalescence with increasing volume fractions of oil (Lobo, Svereika, and Nair,
2002; Mohan and Narsimhan, 1997).
Second, increasing the volume fraction of the dispersed phase influences the flow
field in the homogenizer. Friction between the flow field and drops reduces the veloc-
ity gradients, which would imply decreased fragmentation rates. For turbulent flow,
the situation is more complex. Many theoretical studies on emulsification assume a
monotonic suppression of turbulent fluctuations, with an increase in volume fraction
of the dispersed phase (e.g., Coulaloglou and Tavlarides, 1977). However, addition
of dispersed phase drops can, depending on the properties of both flow field and
the drop, either enhance or attenuate on a turbulent flow (Gore and Crowe, 1989;
Poelma and Ooms, 2006). Prediction of the direction and size of the effect from
theory is presently not possible for volume fractions of technical relevance. However,
some experimental indications exist. Walstra (1974b) suggested that the smallest
eddy length scales were suppressed by the addition of polymers to the pre-emulsion
from observations of increasing drop sizes with polymer length scale. In a different
study, measurements of flow velocity in a scaled model indicated that the addition of
particles redistributes energy between different eddy length scales, which implies a
shift in the relative importance of the inertial to the viscous turbulent mechanisms
(Håkansson et al., 2013). In conclusion, experiments show a decreased efficiency
with an increased amount of the dispersed phase. However, when emulsifier con-
centration is high, the mechanisms involved are complex and not yet completely
understood.
Experiments show a decrease in resulting drop size as a function of increasing
emulsifier concentration as long as the concentration is low. For higher concentra-
tions, the effect is reduced in magnitude and the drop size levels out (see Walstra,
2005 and references therein). This has led investigators to suggest two different
regions: an emulsifier-poor region, where the surface chemistry of the drop emulsi-
fier system dominates, and an emulsifier-rich region, where the effects discussed
in Section 5.3 dominate (Tcholakova, Denkov, and Lips, 2008). The emulsifier
144 Engineering Aspects of Food Emulsification and Homogenization
Impact
ring
Cavitation
Turbulence
Laminar shear
Seat
Fragmentation
FIGURe 5.9 Illustration of the location of a high laminar shear, cavitation, and intense tur-
bulence according to hydrodynamic studies, compared to the region of drop breakup accord-
ing to visualization experiments.
Droplet Breakup in High-Pressure Homogenizers 145
drops was consistent with a combination of turbulent inertial and turbulent viscous
fragmentation.
ReFeReNCes
Bergenståhl, B.A., Claesson, P.M., 1997. Surface forces in emulsions. In: Food Emulsions,
Friberg, S.E., Larsson, K. (eds.), Marcel Dekker Inc., New York, 57–110.
Brennen, C.E., 1995. Cavitation and Bubble Dynamics. Oxford University Press, New York.
Calabrese, R.V., Chang, T.P.K., Dang, P.T. 1986. Drop breakup in turbulent stirred-tank con-
tactors. Part I: Effect of dispersed-phase viscosity. AIChE Journal 32(4), 657–666.
Coulaloglou, C.A., Tavlarides, L.L., 1977. Description of interaction processes in agitated
liquid-liquid dispersions. Chemical Engineering Science 32, 1289–1297.
Davies, J.T., 1985. Drop sizes of emulsions related to turbulent energy dissipation rates.
Chemical Engineering Science 40, 839–842.
Floury, J., Belletre, J., Legrand, J., Desrumaux, A., 2004. Analysis of a new type of high
pressure homogeniser. A study of the flow pattern. Chemical Engineering Science 59,
843–853.
Floury, J., Desrumaux, A., Axelos, M.A.V., Legrand, J., 2002. Degradation of methylcellulose
during ultra-high pressure homogenisation. Food Hydrocolloids 16, 47–53.
Floury, J., Legrand, J., Desrumaux, A., 2004. Analysis of a new type of high pressure
homogeniser. Part B. study of droplet break-up and recoalescence phenomena. Chemical
Engineering Science 59, 1285–1294.
Gaulin, A., 1904. Process of treating milk or similar liquids. US Patent No. 753,792, March 1.
Gore, R.A., Crowe, C.T., 1989. Effect of particle size on modulating turbulent intensity.
International Journal of Multiphase Flow 15, 279–285.
Goulden, J.D.S., 1958. Light transmission by dilute emulsions. Transactions of the Faraday
Society 54, 941–945.
Grace, H.P., 1982. Dispersion phenomena in high viscosity immiscible fluid systems and
application of static mixers as dispersion device in such systems. Chemical Engineering
Communications 14, 225–277.
Håkansson, A., Fuchs, L., Innings, F., Revstedt, J., Bergenståhl, B., Trägårdh, C. 2010. Visual
observation and acoustic measurement of cavitation in an experimental model of a high-
pressure homogenizer. Journal of Food Engineering 100(3), 504–513.
Håkansson, A., Fuchs, L., Innings, F., Revstedt, J., Trägårdh, C., Bergenståhl, B., 2011. High
resolution experimental measurement of turbulent flow field in a high pressure homoge-
nizer model and its implications on turbulent drop fragmentation. Chemical Engineering
Science 66(8), 1790–1801.
Håkansson, A., Fuchs, L., Innings, F., Revstedt, J., Trägårdh, C., Bergenståhl, B., 2012.
Experimental validation of k-ε RANS-CFD on a high-pressure homogenizer valve.
Chemical Engineering Science 71, 264–273.
Håkansson, A., Fuchs, L., Innings, F., Revstedt, J., Trägårdh, C., Bergenståhl, B., 2013.
Velocity measurements of turbulent two-phase flow in a high-pressure homogenizer
model. Chemical Engineering Communications 200, 93–114.
Hinze, J.O., 1955. Fundamentals of the hydrodynamic mechanism of splitting in dispersion
processes. AIChE Journal 1, 289–295.
IDF, 2010. The world dairy situation 2010. Bulletin of the International Dairy Federation 446.
Innings, F., Fuchs, L., Trägårdh, C., 2011. Theoretical and experimental analyses of drop
deformation and break-up in a scale model of a high-pressure homogenizer. Journal of
Food Engineering 103, 21–28.
Innings, F., Hultman, E., Forsberg, F., Prakash, B. 2011. Understanding and analysis of wear
in homogenizers for processing liquid food. Wear 271, 2588–2598.
Innings, F., Trägårdh, C. 2005. Visualization of the drop deformation and break-up process
in a high pressure homogenizer. Chemical Engineering & Technology 28(8), 882–891.
Innings, F., Trägårdh, C., 2007. Analysis of the flow field in a high-pressure homogenizer.
Experimental Thermal and Fluid Science 32, 345–354.
Droplet Breakup in High-Pressure Homogenizers 147
Iyer, C.O., Ceccio, S.L., 2002. The influence of developed cavitation on the flow of a turbulent
shear layer. Physics of Fluids 14, 3414–3431.
Kawaguchi, T., 1971. Entrance loss for turbulent flow without swirl between parallel discs.
Bulletin of the Japan Society of Mechanical Engineers 14, 355–363.
Kleinig, A.R., Middelberg, A.P.J., 1996. The correlation of cell disruption with homoge-
nizer valve pressure gradient determined by computational fluid dynamics. Chemical
Engineering Science 41(23), 5103–5110.
Kleinig, A.R., Middelberg, A.P.J., 1997. Numerical and experimental study of a homogenizer
impinging jet. AIChE Journal 43(4), 1100–1107.
Kolb, G., 2001. Zur Emulsionsherstellung in Blendensystemen. Doctoral Thesis, University
of Bremen, Germany.
Kolmogorov, A.N., 1949. On the breakage of drops in a turbulent flow. Doklady Akademii
Nauk SSSR 66, 825–828. (Originally in Russian. Reprinted and translated in Selected
Works of A.N. Kolmogorov, Volume 1: Mathematics and Mechanics, Tikhomirov, V.M.
[ed.], 1991, 339–343.)
Kurzhals, H.-A., 1977. Untersuchungen über die physikalisch-technischen Vorgänge beim
Homogenisieren von Milch in Hochdruck-Homogenisiermaschinen. Doctoral Thesis,
University of Hannover, Germany.
Lee, L., Norton, I. 2013. Comparing droplet breakup for a high-pressure valve homogenizer
and a microfluidizer for the potential production of food-grade nanoemulsions. Journal
of Food Engineering 114, 158–163.
Lobo, L., Svereika, A., Nair, M., 2002. Coalescence during emulsification. 1. Method develop-
ment. Journal of Colloid and Interface Science 253, 409–418.
Loo, C.C., Carleton, W.M., 1953. Further studies of cavitation in the homogenization of milk
products. Journal of Dairy Science 36, 64–75.
Loo, C.C., Slatter, W.L., Powell, R.W., 1950. A study of the cavitation effect in the homogeni-
zation of dairy products. Journal of Dairy Science 33, 692–702.
McKillop, A.A., Dunkley, W.L., Brockmeyer, R.L., Perry, R.L., 1955. The cavitation theory of
homogenization. Journal of Dairy Science 38, 273–283.
Middelberg, A.P.J. 1995. Process-scale disruption of microorganisms. Biotechnological
Advances 13(3), 491–551.
Mohan, S., Narsimhan, G., 1997. Coalescence of protein-stabilized emulsions in a high-
pressure homogenizer. Journal of Colloid and Interface Science 192, 1–15.
Mohr, K.-H., 1987. High-pressure homogenization. Parts II. The influence of cavitation on
liquid-liquid dispersion in turbulence fields of high energy density. Journal of Food
Engineering 6, 311–324.
Nakayama, Y., 1964. Action of the fluid in the air-micrometer. Bulletin of the Japan Society of
Mechanical Engineers 7, 698–707.
Ogden, L.V., Walstra, P., Morris, H.A., 1976. Homogenization-induced clustering of fat glob-
ules in cream and model systems. Journal of Dairy Science 59, 1727–1737.
Pandolfe, W.D., 1981. Effect of dispersed and continuous phase viscosity on droplet size of
emulsions generated by homogenization. Journal of Dispersion Science and Technology
2, 459–474.
Pandolfe, W.D., 1982. Development of the new gaulin micro-gap™ homogenizing valve.
Journal of Dairy Science, 65, 2035–2044.
Phipps, L.W., 1974. Cavitation and separated flow in a simple homogenizing valve and
their influence on the break-up of fat globules in milk. Journal of Dairy Research
41, 1–8.
Phipps, L.W., 1975. The fragmentation of oil drops in emulsion by a high-pressure homo-
genizer. Journal of Physics D: Applied Physics 8, 448–462.
Phipps, L.W., 1985. The High Pressure Dairy Homogenizer. The National Institute for
Research in Dairying, Reading.
148 Engineering Aspects of Food Emulsification and Homogenization
CONTENTS
6.1 History of the Homogenizer ......................................................................... 149
6.2 The Homogenizer ......................................................................................... 152
6.3 The Homogenization Device ........................................................................ 155
6.4 Wear in the Homogenization Gap ................................................................ 157
6.5 Scale Up and Gap Height.............................................................................. 161
6.6 Economic Aspects ........................................................................................ 164
6.7 The Homogenizer in a Processing Line ....................................................... 166
6.7.1 Split Homogenization ....................................................................... 166
6.7.2 Full Stream Homogenization ............................................................ 167
6.7.3 Partial Homogenization .................................................................... 167
References .............................................................................................................. 167
149
150 Engineering Aspects of Food Emulsification and Homogenization
(a)
(b)
FIGURE 6.1 Patent sketch from the first US homogenization patent: (a) side view and
(b) top view. (Data from Gaulin, A., System for intimately mixing milk. US Patent US756953,
filed September 30, 1902, issued April 12, 1904.)
Di
Do
In 1925, the two-stage homogenizer was patented by Gaulin, but it took quite
some time before it reached widespread use. For some products, the wide gap design
reduces the need for a second stage, as the outer part of the gap serves as a pressure
controller for the upstream part. This effect was optimized by Gaulin, and in 1955,
they patented the Liquid Whirling homogenizing valve, where the product is subject
to a number of rapid pressure and velocity changes as it flows through the gap.
During the following years, a lot of similar devices were invented, for example,
the perforated metal discs by Cherry-Burrell and the Multi-Flo compressed wire by
Crepaco. About 30 years ago, the knife-edge gap (Figure 6.3) was developed with a
large outer diameter in the order of 100 mm and a short gap in the order of 1 mm,
which is today’s standard design.
The last step in gap design was started by Gaulin patenting the MicroGap™ in
1983, US4383769. Here, the flow is distributed in six parallel gaps, making very high
production capacities possible (discussed further in Section 6.5).
Up to the 1960s, the homogenizers in the dairies were used to homogenize pas-
teurized milk with a shelf life of less than a week. To achieve this, a homogenization
pressure of 120–150 bar is required. First launched in Switzerland in 1961, asep-
tic milk with a shelf life of many months demanded a homogenization pressure of
220–250 bar, ensuring that the fat droplets are small enough to be mixed around by
the Brownian motion. To reduce the risk of recontamination of the milk, the homo-
genizer was placed upstream in the nonaseptic part of the process. In 1963, Alfa Laval
(now Tetra Pak) launched the VTIS™ aseptic process, where the milk was heated
by direct steam injection. The VTIS process demanded that the homogenizer be
placed downstream of the sterilizing section, thereby defining the need for an aseptic
homogenizer. There are two demands on an aseptic machine; first, it must be able to
handle the presterilization process, where the complete processing line is heated up
to 120°C–140°C for 20–40 minutes, killing any microorganisms remaining after the
cleaning; second, great care must be taken so that no microorganisms can enter the
sterile milk and recontaminate it. The presterilization is taken care of by upgrading
all seals and gaskets to qualities that can handle the high temperatures. The only
position microorganisms can contaminate the product in a homogenizer is along the
pistons. As the pistons move in and out they can bring microorganisms into the milk.
To counter this Alfa Laval prolonged the pistons and doubled the seals. Between the
seals, a sterilization zone was formed by flushing it with steam or hot condensate,
normally at 120°C.
(a) (b)
FIGURE 6.5 The wet end of the homogenizer: (a) Cut through the check valves seen from
the front, and (b) cut through a piston seen from the side. (Courtesy of Tetra Pak.)
−10
−15
−20
−25
−30
−35
−40
Cycle
12
10
8
Inlet (bar)
0
0 0.25 0.5 0.75 1
−2
Cycle
FIGURE 6.7 Pressure variations for a three-piston homogenizer in a typical process line.
The damper is just a closed pipe filled with air. If placed closed to the homogenizer,
they will work very well but have two main disadvantages. First, they need to be dis-
mantled and cleaned manually; second, the air will slowly dissolve into the product,
which limits the runtime of the dampers. There have been many attempts to solve
these two problems, for example, by adding cleaning in place (CIP) or compressed
air inlets to the dampers, but none has been totally successful in incorporating all the
requirements of the modern food industry.
High-Pressure Homogenizer Design 155
FIGURE 6.10 The wear pattern created on the forcer by tomato paste after 30 minutes of
homogenization. The material is Wallex 20. The wear pattern contains many grooves in the
radial direction, so that a wavy pattern is created in the circumferential direction. The typical
reflection is also seen as a wavy pattern in the radial direction, with the first crater being the
deepest. (Courtesy of Tetra Pak.)
FIGURE 6.11 A modern seat and a forcer that can be easily exchangeable and turnable.
(Courtesy of Tetra Pak.)
158 Engineering Aspects of Food Emulsification and Homogenization
The amount of wear of the gap is totally dependent on the type of particles and
the amount of cavitation in the gap. Figure 6.12 shows the relative amount of wear for
a couple of particles/product types (Innings et al. 2010). Note that the scale is loga-
rithmic and the worst particles give a 5-log increase of the wear. Corundum particles
were added as a scientific reference, but the calcium particles are common in food
products in, for example, fortified milk.
The different products results in quite different wear patterns in the gap. Figure 6.13
shows a stainless steel seat that has been running on tap water for 600 hours at 600
bar and with a Thoma number of 0.2. The flow is from right to left, the length of the
undamaged gap was 0.5 mm, and the image is 0.7 mm wide. The seat had a surface
finish of Ra = 0.8 µm and on the unworn surfaces the machine groves can easily be
seen. The presence of the machined groves is a good measure, indicating that the
surface has not worn out. The wear pattern on the outlet is porous and rough, which
is typical for cavitation–erosion and is completely different compared to the case
where particles were added. Note that the inlet is totally undamaged, which shows
that a Thoma number of 0.2 suppresses the cavitation at the inlet-separation bubble.
Figure 6.14 shows a stainless steel seat that has been running on tomato paste for
2 hours at 280 bar and with a Thoma number of 0.2. A wide crater has been formed
100
Tomato
10
Calcium, 10 μm
1
0.1 Milk
0.01
0.001
0.0001
FIGURE 6.12 The weight loss for different particle types and with cavitation without added
particles as reference. Note that the scale is logarithmic.
FIGURE 6.13 Stainless steel seat that has been running for 600 hours at 600 bar on tap
water and with a Thoma number of 0.2.
High-Pressure Homogenizer Design 159
FIGURE 6.14 Wear from tomato paste on a stainless steel seat with Th = 0.2. The surface
has a corrosive appearance.
FIGURE 6.15 Wear from calcium particles. Low cavitation (Th = 0.2). This wear has
removed all the machining marks in the gap region and has also caused a lot of deep craters on
the inlet and the outlet areas. The area between the craters has a matte appearance.
on both the forcer and the seat. The surface is smooth and has a matte look. No outlet
damages are appearing on either the forcer or the seat. The radial machine grooves
have been worn away also at the regions outside of the crater.
Figures 6.15 and 6.16 show the wear from calcium particles at the standard and a
low Thoma number, respectively. It shows that there is a great synergy between the
particles and the cavitation. The mechanism is likely to be that the imploding cavita-
tion bubbles accelerate the particles, so that they hit the gap surfaces at high velocities.
Figure 6.17 shows the profiles of the worn-out gaps from Figures 6.15 and 6.16.
The amount of wear is extensive in Figure 6.17b, but it should be noted that it was
still possible to keep the homogenization pressure, so it is not in any case worse than
a worn-out commercial valve.
This insensitivity to wear is one of the main advantages of the variable-gap type
homogenization devices, where it is the force from the hydraulic piston that keeps
the gap height low. As the gap is worn out, the force on the forcer closes the gap
slightly, making sure that the homogenization pressure is always constant. This is
of course only true if the wear is fairly even and not too severe. When the wear is
too severe, channels will be worn down into the seat and the forcer, and when all
the product flows in the channels, it will not be possible to close the gap anymore;
the machine cannot withstand the homogenization pressure, which indicates that it
160 Engineering Aspects of Food Emulsification and Homogenization
FIGURE 6.16 Wear from calcium particles. High cavitation (Th = 0); the crater near the
inlet has increased considerably in depth and width and the reflection between the two has
started, see the crater in Figure 6.17b.
Forcer
Inlet
Seat
(a)
Forcer
Inlet
(b) Seat
FIGURE 6.17 (a, b) Profile of the worn-out gaps from calcium particles. (a) Standard cavita-
tion (Th = 0.2). (b) High cavitation (Th = 0). The length of the gap is always 0.5 mm and the
gap height is in the order of 100 µm.
is time to replace the parts. A fixed orifice does not have this automatic adjusting
function, so when a fixed orifice is worn, the pressure drop decreases and the flow
rate must be increased to keep the homogenization pressure and this is normally not
possible in an industrial processing line. The other advantage of the variable gap
is, of course, that the flow rate and the homogenization pressure can be controlled
High-Pressure Homogenizer Design 161
1.0
Sauter mean diameter (μm)
0.8
0.6
0.4
0.2
0.0
0.000 0.050 0.100 0.150 0.200
Gap height (mm)
FIGURE 6.18 Sauter diameter as a function of gap height. Data gathered from a large num-
ber of industrial installations.
162 Engineering Aspects of Food Emulsification and Homogenization
250
200
Device diameter (mm)
Gap height (μm)
150
D theoretical
D industrial
100
Gap industrial
50
0
0 5,000 10,000 15,000 20,000 25,000 30,000
Flow rate (l/h)
FIGURE 6.19 Theoretical device diameter as a function of capacity for a 50 µm gap (―);
industrial relevant device diameter restricting the device between 5 and 120 mm (- - -). Gap
height for an industrial homogenizer (•••).
(a) (b)
FIGURE 6.20 The force balance of the forcer for (a) a standard homogenization device and
(b) a balanced force-type device.
forcer, as can be seen in Figure 6.20a. Second, the weight of the seat and the forcer
would be in the order of 100 kg, and the weight of the housing needed to withstand
the pressure would be in the order of 1000 kg. As this is not industrially feasible,
the solution is to restrict the device diameter to, for example, 120 mm, shown by the
dashed line in Figure 6.19. The same is also true on the other end of the scale: it is not
possible to manufacture a device with a very small diameter. A 100 L/h homogenizer
would need a device with a diameter of 0.9 mm, and it is not possible to manufacture
if the gap shape should mimic that of an industrial machine. With the low-capacity
High-Pressure Homogenizer Design 163
machines, the gap heights are therefore normally too small. This can be seen in the
black line in Figure 6.19, where the smallest device has been restricted to 5 mm. For
low capacities, gap height will be very small and for capacities larger than 13,000 L/h,
the gap will increase to more than 100 µm for a capacity of 30,000 L/h (Innings 2005).
Quite a few ideas have come up over the years to overcome this problem. The first
idea in commercial production was maybe the easiest one: to balance the forces by
restricting the force from the product on the forcer. This can be done by reducing the
area of the high-pressure product that is affecting the forcer by extending it and sealing
the center part of the forcer from the high-pressure product (Figure 6.20b). This is done
in the Soavi NanoVALVE™ and it can double the possible diameter and thus achieve a
high efficiency up to higher capacities. The disadvantage is that you get a complicated
high-pressure seal that will wear quickly and would need to be replaced regularly.
Another idea to increase the efficiency is to increase the circumference of the gap by
multiplying it. One way of doing it is shown in Figure 6.21, where two (Figure 6.21a)
24 26
1 2
20 17
23 21 16
8 25
7 6
12
1 4 8
2 3
A 18
4
22
1 A 21 23
9 8
14 3 24 19
15 5
25
12 16
7
h 20
13 7
6 10 9
4 6 8
11 12
10 16 11a 11c
5 13
(a) 2 (b) 15 11b
FIGURE 6.21 Homogenization device with (a) two gaps—one inside the other—and (b) three
gaps inside each other.
164 Engineering Aspects of Food Emulsification and Homogenization
Hydraulic piston
Outlet
Gaps
Inlet
FIGURE 6.22 Five-parallel gap HD EnergyIQ™ concept from Tetra Pak. (Courtesy of
Tetra Pak.)
the surface area of the drops. More than 99.9% of the energy is instead spent heating
up the product, giving it a heat increase of about 1°C for each 50 bar of homogeni-
zation pressure.
When looking into the economic aspects of a homogenizer, we have three main
cost types:
1. Investment cost
2. Electricity cost
3. Spare parts and maintenance cost
A homogenizer lasts for more than 50 years but is normally replaced after 30–40
years; this practice, however, is clearly not relevant to a calculation of yearly cost.
The cost for a standard industrial machine capable of a flow rate of 20,000 L/h
and a homogenization pressure of 200 bar is in the order of €120,000. Figure 6.23
shows the cost for the machine for the first year given that the machine runs
20 hours a day.
From Figure 6.23, it can be seen that electricity cost for one-year production
is about half of the cost of the brand new machine. The maintenance and spare
parts cost is a substantial €16,000 per year, but is dwarfed by the electricity cost.
It should be noted that this is only the case when running milk and other non-
abrasive products. If the machine is running an abrasive product such as tomato
juice or ketchup, the maintenance cost can be in the same order as the electric-
ity cost. The only cost not taken up in Figure 6.23 is the cooling and seal water
cost. Water is used to cool the crankcase and in nonaseptic machines also for
the piston seals. In aseptic machines, hot water is used and if it is not controlled
carefully, the cost for the water spent will be surprisingly high. In dairies, you
can sometimes see an open 6 mm pipe feeding water at 95°C to the seals. This
will give a fairly moderate flow of 15 L/min, but the yearly cost for this hot water
is about €15,000.
120,000
100,000
80,000
Cost/year (€)
60,000
40,000
20,000
0
Investment Electricity Maintenance
IW
Raw milk
Surplus
cream
Skim milk
REFERENCES
Bylund, Gösta. 2003. Dairy Processing Handbook. Lund, Sweden: Tetra Pak Processing
Systems AB.
Gaulin, Auguste. 1904. System for intimately mixing milk. US Patent US756953, filed
September 30, 1902, issued April 12, 1904.
Håkansson, Andreas, Laszlo Fuchs, Fredrik Innings, Johan Revstedt, Björn Bergenståhl,
Christian Trägårdh. 2010. “Visual observations and acoustic measurements of cavitation
in an experimental model of a high-pressure homogenizer.” Journal of Food Engineering
100 (3): 504–513.
Hansson, Rikard, Rolf Malmberg. 1998. Homogenizer valve. European Patent
WO1998047606A1, filed April 21, 1998, issued October 29, 1998.
Innings, Fredrik. 2005. “Drop break-up in high-pressure homogenisers.” PhD Thesis, Lund
University, Sweden.
168 Engineering Aspects of Food Emulsification and Homogenization
Innings, Fredrik, Erik Hultman, Fredrik Forsberg, Braham Prakash. 2010. “Understanding
and analysis of wear in homogenizers for processing liquid food.” Wear 271 (2011):
2588–2598.
Innings, Fredrik, Rolf Malmberg. 2000. A method of homogenization. European Patent
WO2000015327A1, filed September 13, 1999, issued March 23, 2000.
Kurzhals Hans-Albert. 1997. “Undersuchungen Uber die physikalisch-technichen Vorgänge beim
Homogenisiren von Milch in Hochdruck-Homogenisiermaschinen.” PhD Dissertation,
Technichen Universität Hannover, Germany.
Malmberg, Rolf, Jozo Valencuk. 2010. Homogenizer valve. US Patent, US 20140177382, filed
December 20, 2011, issued June 26, 2010.
7 High-Pressure
Homogenization with
Microstructured Systems
Karsten Köhler and Heike Schuchmann
Contents
7.1 Introduction .................................................................................................. 170
7.2 Technical Equipment .................................................................................... 170
7.2.1 Valves................................................................................................ 170
7.2.2 Orifices and Nozzles ......................................................................... 170
7.2.3 Flow Conditions ................................................................................ 172
7.3 Emulsification Mechanism ........................................................................... 173
7.3.1 Deformation ...................................................................................... 173
7.3.2 Breakup............................................................................................. 174
7.3.3 Disruption as Result of Breakup of Deformed Droplets .................. 174
7.3.4 Process Functions ............................................................................. 175
7.3.5 Droplet Stabilization ......................................................................... 178
7.4 Process and Material Parameters Influencing Emulsification ...................... 179
7.4.1 Geometry .......................................................................................... 179
7.4.2 Multiple Stage ................................................................................... 184
7.4.3 Multiple Passage ............................................................................... 185
7.4.4 Viscosity Ratio.................................................................................. 186
7.4.5 Stability ............................................................................................. 187
7.4.6 Scale-Up ........................................................................................... 189
7.5 Conclusion .................................................................................................... 189
References .............................................................................................................. 190
169
170 Engineering Aspects of Food Emulsification and Homogenization
7.1 IntRoDUCtIon
High-pressure homogenizers (HPHs) were developed approximately 100 years ago
(Gaulin, 1899) during the age of industrialization. The basic idea of combining a
high-pressure pump and a disruption system, such as a valve, was presented at the
1900 World Exposition in Paris and endures to this day. However, we still see ongo-
ing developments of pumps and disruption systems stemming from either daily
application problems or new product challenges.
Current techniques permit volume streams of up to 50,000 L/h and pressures
of up to 10,000 bar. However, homogenization pressures in industrial applications
today are in the range of 50–2000 bar. Piston pumps mainly serve as high-pressure
pumps. In bench-scale equipment, a single-piston pump commonly delivers the vol-
ume stream—or rather the pressure—but in production plants, up to eight piston
pumps are found. The disadvantage of the single-piston pump is that the pressure
and the volume stream can vary significantly over time, which results in a pulsation
of stresses on the product. Inhomogeneous stresses act on the droplets, thus mak-
ing product properties difficult to control. To reduce pulsation, several pistons are
combined in a phase-shifted manner. Valves are used to control the different pistons,
which are not usually influencing the quality of the emulsion.
Section 7.2 presents the current state of technical equipment. Section 7.3 dis-
cusses the mechanism of droplet breakup in these devices. Section 7.4 discusses the
influence of the main parameters on emulsification.
7.2.1 ValVes
Valves, also known as radial diffusors, are commonly used high-pressure disruption
systems. The fundamental idea is to reduce the flow diameter with a valve plunger,
which is pushed to a valve seat forming a small gap. These systems were introduced
and discussed in Chapters 5 and 6. Over the years, the geometry of flat valves has also
been developed. In this chapter, we will limit our discussion to the conventional old
flat valve and the new flat valve with a tapering inlet. The size range of the smallest
gap of the valves is generally 100 nm to several micrometers. Thus, valves are also
microstructured systems. In this chapter, however, we will use the definition of micro-
structured systems only for disruption systems with a fixed geometry, called orifices.
Ed a b
Bd2
Ed
(a) (b)
(c) (d)
FIGURe 7.1 Schematic drawing of a simple orifice (a) and modified types (b–d).
(Data as published by Stang, M., Zerkleinern und Stabilisieren von Tropfen beim mecha-
nischen Emulgieren, Dissertation, Universität Karlsruhe, 1998; Freudig, B., Herstellen von
Emulsionen und Homogenisieren von Milch in modifizierten Lochblenden, Dissertation,
Universität Karlsruhe, Germany, 3-8322-3147-1, 2004; Aguilar, F.A. et al., Chem. Ing. Tech.,
80, 5, 607–613, 2008.)
disturbances (Aguilar et al., 2008). Similar effects are found for orifices with internal
steps deflecting the flow (Figure 7.1d) (Cook, 1985, 1990; Penth, 2000; Aguilar
et al., 2008).
Generally, in all microstructured devices, the same flow patterns arise as with
conventional flat valves. There are regions of laminar and turbulent flow, as well as
cavitation. They differ only in how long the different flow patterns exist and in the
height of tensions created.
7.3.1 defOrmatiOn
Droplets are deformed by tensions that result from different flow conditions or vibra-
tions, and these have an effect on the droplet, especially the interfaces. The coun-
teracting tensions are the surface tension and the viscosity—or elastic effects—of
the surface and the dispersed phase (Arai et al., 1977). If the deforming tensions
exceed the counteracting tensions, the droplet starts to deform. To describe the pro-
cess of deformation, the three dimensionless numbers Weber (We), capillary (Ca),
and Ohnesorge (Oh) are established using the Reynolds number (Re):
v⋅x σ⋅ x We η ⋅ v We η
Re = , We = , Ca = = , Oh = = (7.1)
ν γ Re γ Re ρ⋅γ ⋅ x
Depending on the flow conditions and material parameters, the We, Ca, or Oh num-
ber can better describe the deformation. Besides the acting tensions, the duration of
the acting tensions is also important (Walstra, 1983).
Most work is done on the deformation of the whole droplet. Droplet deformation
due to laminar shear flow has been widely investigated (Walstra, 1983; Bentley and
Leal, 1985; Stone, Bentley, and Leal, 1986). However, it is restricted to a narrow
range of viscosity ratios between the dispersed and the continuous phases ηd/ηc for
174 Engineering Aspects of Food Emulsification and Homogenization
single-droplet disruption, or between the dispersed phase and the emulsion ηd/ηe for
emulsions, respectively (Armbruster, 1990; Jansen, Agterof, and Mellema, 2001).
Laminar elongation flow is advantageous if highly viscous dispersed phases have
to be disrupted (Grace, 1982). It is usually found in the inlet of disruption systems.
Specific disruption systems are designed for increased elongation in the inlet flow.
Turbulence can be described as eddies of varying sizes (see flow conditions).
Depending on the relation of droplet size to eddy size, both the viscous and the
inertial regimes can be differentiated (Hinze, 1955). In the turbulent viscous regime,
in a Lagrange approach, the droplet is subject only to laminar shear and elongation
tensions due to the fact that the droplet is smaller than the eddies (Vankova et al.,
2007). Thus, the whole droplet is deformed as in laminar flow.
Two mechanisms are known to deform the interface of the droplet: capillary waves
or the turbulent inertial regime. Behrend (2002) showed that the smallest achievable
droplets by capillary waves are in the range of microns. This effect is thus negligible
for the production of submicron emulsions. The deformation of droplets in the tur-
bulent inertial regime is of great interest. This regime was first described by Hinze
(1955). In this regime, the droplets are in the same range as eddies or larger. This
results in a deformation of the surface due to several eddies acting on the surface.
7.3.2 Breakup
A breakup of droplet occurs, if the deformation reaches a critical value in deforma-
tion and time. The breakup of a formed filament can be explained by instabilities
or mechanical tensions (Tcholakova et al., 2011). The most often discussed case in
the literature is the Plateau–Rayleigh instability (Plateau, 1873; Eggers, 1997). This
effect is driven by a perturbation of the flow, which results in small difference in the
diameter of the filament. These differences produce differences in the local capillary
pressures and enforce the growth of large filament parts and the reduction of small
parts until the breakup occurs. Instead of perturbations, the filament surrounding
flow can also produce differences in the diameter of the filament.
Droplets are disrupted if they are deformed over a period of time, tdef, which
is longer than a critical deformation time, tdef,cr, and, if the deformation exceeds a
critical value. Both criteria must be fulfilled. The necessary deformation as well as
the time needed for droplet breakup depends on the external tension, σ; the droplet
diameter, x; the interfacial tension, γ; and the viscosities of the dispersed phase, ηd;
and the continuous phase, ηc. The last three parameters are material parameters,
which today can be measured offline using established methods (Walstra, 1993).
ηd
tdef,crit = (7.2)
σ − pk
The second approach is to measure the size of the remaining droplet, which has
resisted against the external forces and thus can be found in the final emulsion. Using
this approach, it is possible to measure the tensions appearing in the process by
inserting it into the Weber or capillary number.
E P ∆p ⋅V (7.3)
EV = PV ⋅ t res = = = = ∆p
V V V
When passing the emulsion several (n) times through the disruption system, this
must be multiplied by the number of passes, n. This can be assumed as long as
the disruption is going on and the duration of tensions is the limiting factor, which
176 Engineering Aspects of Food Emulsification and Homogenization
is often the case in HPHs. Once all droplets that cannot survive the stresses are
disrupted, this concept does not work anymore. An equilibrium droplet diameter is
found, which is determined by the maximum volume related power density, PV, also
called ε in Kolmogorov’s, Hinze’s, and Davies’s theory of droplet breakup in turbu-
lent flow (Kolmogorov, 1949; Hinze, 1955; Davies, 1972).
With the specific disruption energy, EV, a mean droplet diameter such as the
Sauter mean diameter x32 can be calculated in case of well-defined flow conditions
by the following process functions:
Laminar shear flow (Grace, 1982) is given by the following equation:
η η
x32 ∝ EV−1 ⋅ f d or x32 ∝ (n ⋅ ∆p)−1 ⋅ f d (7.4)
ηe ηe
Laminar elongational flow (Walstra, 1983; Bentley and Leal, 1986; Chesters, 1991)
is given by the following equation:
Isentropic turbulent flow (Arai et al., 1977) and microturbulences in the cavitational
flow (Bechtel, Gilbert, and Wagner, 1999, 2000; Behrend, Ax, and Schubert, 2000;
Behrend, 2002) are given by the following equation:
In Equation 7.6, the exponent of the viscosity of the dispersed phase, ηd, is equivalent
to 0 for lowly viscous dispersed phase (ηd < 10 mPa s), whereas it is 0.75 for highly
viscous dispersed phases.
In defined flow conditions, the exponent of the disruption energy is either −1 or
between −0.25 and −0.4. For industrial homogenization valves, exponents around −0.6
have been published, for example by Walstra (1983), depicting the mix of flow condi-
tions found in homogenization valves. The overall equation describing the process
functions for high-pressure homogenization processes thus is:
where:
b is in the range of 0.25–1
c is lesser than 0.75
This means that using Equation 7.7, we can predict the droplet size of an emulsion of
a specific recipe produced in a homogenizer with a specific valve.
In all approaches a change in the valve geometry is often a challenge, due to the fact
that changes in flow conditions are difficult to predict, as it stands today. In different
geometries, the tensions are distributed differently in location, time, type, and intensity
(Equation 7.7) using a mean tension over time, type, and location, thereby limited to
explaining local mechanisms or predicting the efficiency of a geometric modification.
High-Pressure Homogenization with Microstructured Systems 177
> mm μm nm
Macro Micro
Mesoenergy
energy density density tension
FIGURe 7.2 Graphical explanation of the mesoenergy density. (Data from Köhler, K.,
Simultanes Emulgieren und Mischen, Logos Verlag, Berlin, Germany, 2010.)
In conclusion, models describing specific devices and flow conditions are avail-
able, and investigations are conducted to predict droplet size distributions resulting
from a high-pressure homogenization process. However, slight changes in the geom-
etry or other parameters often cause the models to fail. This is explainable by the fact
that most investigations in disruption are of single flow types, with homogenous flows
in space and enough time to break the droplet. As a result, most models are based on
these experiments. Yet, in most HPHs, the different flow patterns are often superposed
and inhomogeneous in time and space. This means that the flow conditions are often
transient and the flow on the walls differ completely from the main stream. In addi-
tion, droplet–droplet interactions resulting, for example, in their coalescence and thus
in a change of droplet sizes are usually not considered at all. As a result, most models
used to describe droplet size fail to calculate the final droplet size, in particular the
droplet size distribution especially arising through modifications.
Ω = pcoal ⋅ C (7.8)
The coalescence probability mainly depends on material parameters such as the sur-
face tension and elasticity, as well as droplet–droplet interaction forces of electro-
static, hydrodynamic, or steric nature. These values change over time through the
homogenization process due to the newly produced and emulsifier-reduced surfaces,
which are again covered with other emulsifier molecules adsorbing at them. This,
however, is a process that takes some time. This time depends on the emulsifier
molecule structure and chemical nature of the phases (Miller, 1990). Therefore, the
kinetics of the emulsifier(s) also influence the probability of coalescence and thus
the coalescence rate (Stang, Karbstein, and Schubert, 1994; Vankova et al., 2007).
The collision frequency, C, depends on the local flow conditions and surface area
per volume. In an isentropic turbulent flow, it can be calculated using Equation 7.9.
C = K ⋅ u ⋅ x 2 ⋅ n2 (7.9)
where:
K is the collision coefficient
u the velocity of the droplet (being dependent on the droplet’s diameter, x)
n is the number of droplets
As can be easily derived from Equation 7.9, the collision frequency increases with the
number of droplets (n2) and their size (x2). Even when the droplet size is reduced in
emulsification, the surface area and the number of droplets increase significantly and
have a strong influence on the collision frequency. Also, the flow has an important
impact on the droplet coalescence, as it generates the necessary acceleration of the drop-
lets. Both parameters increase with the specific energy input. Coalescence rates found
for typical emulsification processes can be mathematically described using first-order
kinetics. Kinetic rate constants derived from these can be used to describe the influence
of process and material characteristics on resulting droplet sizes. Detailed information
regarding droplet coalescence is found in Danner (2001) and Vankova et al. (2007).
7.4.1 GeOmetry
In Section 7.2, we introduced several disruption devices. In this section, we discuss
the influence of geometry on the emulsification result. In Figure 7.3 disruption sys-
tems are compared by producing corn oil droplets in water. As the droplet fraction
180 Engineering Aspects of Food Emulsification and Homogenization
4.0 O/W
CE = 5% SDS
3.8 φ = 0.25 vol%
Δp = 100 bar
3.6 Orifice one stage
Droplet diameter, x903 (μm)
3.4
Bd1 = 1.5 mm
3.2 Ed = 2 mm
Ed = 4 mm
3.0 Ed = 8 mm
2.8 Bd1 = 4 mm
Ed = 2 mm
2.6 Ed = 8 mm
2.4
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Orifice diameter, Bd2 (mm)
FIGURe 7.3 Droplet diameter, x903; plotted over the valve diameter, Bd2; for different inlet,
Bd1; and outlet diameters, Ed2, by a constant homogenization pressure, Δp = 100 bar.
is rather low (1%), we can ensure that droplet–droplet coalescence rates are low, and
we have a disruption-dominated emulsification process.
In all cases, the droplet diameter decreases linearly with increasing specific
energy input (corresponding to the pressure difference applied), as predicted by
Equation 7.7. However, the position and the slope of the resulting curves differ. To
understand the differences, we have to discuss the geometries in detail.
The new-generation flat-valve geometry allows for droplet elongation in the inlet
as well as higher turbulence and cavitation in the outlet. Elongation of droplets to
filaments results in decreased mean Sauter diameters, as disruption of elongated fila-
ments is easier compared to spherical droplets. This results in decreased Sauter mean
diameters and a decreased slope of the curve in double logarithmic diagrams.
Results found for the simple orifice geometry are similar, indicating similar
local flow conditions. Regarding the geometry of the valve (see Figure 7.1) and flow
conditions already published by simulation, droplet elongation followed by a breakup
in local turbulences is a realistic scenario to be assumed. The Microfluidizer® geom-
etry equipped with the double-stage disruption system—thus applying back pressure—
allows for a further improvement of homogenization results. Comparable slopes of the
curves indicate that comparable flow conditions are responsible for droplet breakup.
Similar results are also found for double valves, which also apply back pressure to
the first stage (Kolb, 2001; Freudig, 2004; Karasch and Kulozik, 2008). All authors
hypothesize that the back pressure is responsible for an improved breakup. The reason
for this has to be investigated in more detail in future.
High-Pressure Homogenization with Microstructured Systems 181
This shows that the geometry of microstructured valves is a factor of main influ-
ence on the homogenization result. We thus concentrate on this using the simple
basic geometry, as shown in Figure 7.1. We changed the three design parameters: the
inlet, outlet, and valve diameter. We produced emulsions and ensured that coales-
cence did not appear and lead to a misinterpretation of disruption results. For this,
we used sodium dodecyl sulfate (SDS) in excess, known to be a very fast stabilizing
emulsifier, and a low dispersed phase fraction of φ = 0.25 vol%. In Figure 7.3, the
maximum droplet diameter of the volume collective x903 is plotted over the valve
diameter, Bd2, for a different inlet, Bd1, and outlet diameter, Ed, at a constant homog-
enization pressure, Δp = 100 bar.
Remarkable is the fact that the droplet diameter can be modified by a factor of 2
in changing the geometry. The largest droplets are achieved by a valve diameter Bd1
around the 0.4 mm level. This effect can be seen in the maximum droplet size x903, as
well as in the mean Sauter droplet diameter x32 (see Köhler, 2010).
Furthermore, we see that the combination of large inlet and outlet diameters leads
to smaller droplets. However, this correlation is not significant and, therefore, cannot
be used for further optimization. Yet this is consistent with results reported by Stang
(1998, S. 123–124). He found that the homogenization result is not significantly influ-
enced by the ratio of inlet diameter to valve diameter.
We simulated the flow for the different modifications of the geometry and
compared the resulting tensions in each cell of the VOF-mesh. Laminar tensions
(or mesoenergy densities EV, m), were achieved by multiplying the elongation
and shear rate with the corresponding viscosity. The turbulence was modeled
using the RNG-k-ε-model. Turbulent tensions were calculated by the following
equation:
The tensions calculated were compared on the symmetry axis of the orifice. A discus-
sion of the local distribution of the tension can be found in Köhler (2010). As vegeta-
ble oil (viscosity: 60 mPa s) was used as the dispersed phase and water (viscosity of
1 mPa s) as the continuous phase, the breakup due to laminar shear can be neglected.
At this viscosity, ratio droplets will only rotate in pure laminar shear flow, but not
deform (Stone, Bentley, and Leal, 1986). Elongation in y-direction is negligible due
to low mesoenergy densities. Thus, only turbulent dissipation ε and elongation in the
x-direction are considered (see Figure 7.4). In the following diagram, the mesoen-
ergy densities resulting from the simulations are plotted over time for the symmetry
axis of orifices with a diameter Bd2 of 0.1 and 0.8 mm. The time is set to zero at the
moment when a fluid element is at the entrance of the orifice.
Both droplet devices are subjected to stresses prior to entering the orifice, as they
are accelerated and elongated. The tensions increase to maximum values in the first
microseconds after entering the orifice. The resulting tensions are comparable, and only
the absolute value and the duration change. With a decrease in the orifice diameter, Bd2,
elongation is increased. The main difference between the orifices is in the duration of the
tensions, especially those resulting from turbulence. This shows that droplet deforma-
tion and breakup kinetics may not be neglected in homogenization. The time between
182 Engineering Aspects of Food Emulsification and Homogenization
One-phase simulation
104 water
RNG-k-ε turbo model
Δp = 100 bar
Bd1_Bd2_Ed
1,5_Bd2_2
103
Ev,m (kg/ms2)
Mesoenergy density
102 Bd2 (mm)
0.1 0.8
ε
dvx/dx
101
−20 0 20 40 60
Time, t (μs)
FIGURe 7.4 Comparison of the mesoenergy densities resulting from turbulence (ε) and
elongation in x-direction (dvx /dx) in the orifices of diameters Bd2 = 0.1 and 0.8 mm plotted
over time on the symmetry axis of the orifice (t = 0: entrance of the orifice).
the maximum tension by elongation and turbulence decreases from around 30–10 µs.
Thus, a droplet deformed by elongation before entering the orifice has less time to relax
before being subjected to the tensions resulting from turbulence.
Droplets are disrupted in the laminar elongational flow as soon as the Weber
number exceeds the value of 0.1 and the deformation time reaches a critical value
(Walstra, 1993). We calculated the droplet diameter at which the tension on the sym-
metry axis exceeds the critical Weber number. This calculated maximum droplet
diameter is indicated in the following diagram with elongation (see Figure 7.5). The
droplet diameter at which the tension persists long enough to break up the droplets
is indicated by time.
Only at small orifice diameters, the elongational tensions are high enough to
deform droplets up to their breakup. The duration of the deformation, however, is
not long enough to achieve droplets in a size of the experiment.
To describe the droplet disruption in turbulence, models working with a turbulent
Weber number are also available (Hinze, 1955; Arai et al., 1977). But so far, no equa-
tion for the critical turbulent Weber number is known. Thus, we used the approach of
Kolmogorov and calculated the maximum droplet diameter surviving the dissipation
rate. This maximum droplet diameter decreases slowly with increasing orifice diam-
eter, Bd2. For an orifice diameter of 0.8 mm, the model predicts the droplet size well,
although differences are found for the small diameters. Also, models of other authors
(Hinze, 1955; Davies, 1985; Vankova et al., 2007) are not able to predict the droplet
size found in our experiments. For small orifice diameters, elongation seems to play
a pronounced role in the breakup. In summary, this confirms that models describing
High-Pressure Homogenization with Microstructured Systems 183
250.0
One-phase simulation
water
100.0 Δp = 100 bar
RNG-k-ε turbo model
Bd1_Bd2_E
Droplet diameter, x (μm)
1,5_Bd2_2
10.0
Calculated:
Elongation
Time
3.0 Turbulence
Experiment:
x903
1.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Orifice diameter, Bd2 (mm)
FIGURe 7.5 Calculated maximum droplet diameters surviving elongation (calculated for a
dominating effect of the elongation rate or elongation time) and turbulent energy dissipation
compared to the maximum droplet diameter achieved in the experiments.
4
Milk
φ = 3.5 vol%
3
ϑ = 60–65°C
Full-stream
One stage
2
Droplet diameter, x903 (μm)
1
Slit orifice
Orifice
U
UD
Y-chamber
Z-chamber
0.4
45 100 500 1000
Homogenization pressure, Δp (bar)
FIGURe 7.6 Maximum volume-rated milk fat globule droplet sizes x903 after homogeni-
zation with different orifices at homogenization pressures up to 1000 bar. Homogenization
temperature 65°C.
3.3
3 Milk
φ = 3.5 vol%
ϑ = 65°C
2 Full-stream
Droplet diameter, x903 (μm)
Two stage
YZ
Y
Z
ZR YZ
0.3
45 100 500 1000
Homogenization pressure, Δp (bar)
FIGURe 7.7 Maximum volume-rated milk fat globule droplet sizes x903 after homogeni-
zation with different orifices at homogenization pressures up to 1000 bar. Homogenization
temperature 65°C.
High-Pressure Homogenization with Microstructured Systems 185
besides local stresses, inner stresses in the droplets have to be considered when
analyzing droplet breakup. Milk fat globules, for example, are known to resist elon-
gation due to their specific multilayer membrane composition (Kessler, 2002) and
will thus react differently to external stresses than rapeseed oil droplets.
Future research, therefore, has to concentrate not only on the continuous phase
flow in microstructured orifices but also on droplet deformation behavior.
1.2
Orifice
Veg. oil-in-water
Volume distribution, q3 (μm−1)
0.8
Number of passages n
φ = 30%
1
φ = 1%
0.4 8
1
φ = 30%
8
0
10−2 10−1 100 101
Droplet diameter, d (μm)
FIGURe 7.8 Multiple passage: droplet size distribution depending on the dispersed
phase content and the number of passages. (Data from Tesch, S., Charakterisieren mecha-
nischer Emulgierverfahren: Herstellen und Stabilisieren von Tropfen als Teilschritte beim
Formulieren von Emulsionen, Dissertation, Universität Karlsruhe, Germany, 2002.)
2
Orifice
Water, veg. oil
O/W emulsifier: Tween 80
Sauter mean diameter, x12 (μm)
1
W/O emulsifier: Triodan
φ (vol%):
W/O O/W
30% 30%
50% 60%
70%
0.1
106 107 2.107
Energy density, Ev (J/m3)
FIGURe 7.9 Influence of the dispersed phase fraction on the Sauter mean diameter is depen-
dant on the energy density for different emulsion types. (Data from Tesch, S., Charakterisieren
mechanischer Emulgierverfahren: Herstellen und Stabilisieren von Tropfen als Teilschritte
beim Formulieren von Emulsionen, Dissertation, Universität Karlsruhe, Germany, 2002.)
for W/O emulsions. At the same energy density (or homogenization pressure), the
droplets of the W/O emulsions are smaller by a factor of 2. This can be explained by
the breakup mechanism. Breakup is improved at the viscosity ratio ηd/ηe between
0.1 and 1, as is found in O/W emulsions. In W/O emulsions, the viscosity ratio ηd/ηe
is <<0.1, resulting in higher critical capillary numbers.
7.4.5 staBility
As discussed earlier, homogenization in HPHs is a fast process. Droplets deform and
break within milliseconds and less. Thus, the homogenization is only crowned by
success, if the emulsifier is fast enough to stabilize the newly formed surfaces.
One example for a critical homogenization in terms of stabilization is the homog-
enization of concentrated milk, also called cream. Dairy processes are one of the old-
est industrial high-pressure homogenization processes and up to now, the ones with
the biggest volume streams. In conventional processing, raw milk is separated prior to
homogenization into a low-fat phase (0.03–0.3 vol% fat, skim milk) and a fat-enriched
phase (13–42 vol% fat, cream) using a separator (Kessler, 2002). In the conventional
full stream homogenization process, the milk is first standardized to the final product fat
content by mixing these two phases. The product of this process is then homogenized at
pressures around 100 bar. Also, conventionally applied are partial stream homogeniza-
tion processes. Here, the cream is diluted with skim milk to a fat content of 13–17 vol%,
then homogenized, and afterward standardized again to the target fat concentration of,
for example, 3.5 vol% in full-cream milk. This reduces the required energy as less con-
tinuous phase has to be compressed to nearly the same homogenization pressure.
188 Engineering Aspects of Food Emulsification and Homogenization
Full-stream homogenization
φmilk = 3.5 vol%
6 φmilk = 32 vol%
Droplet diameter, x903 (μm)
SEM
φcream = 42 vol%
4 φcream = 32 vol%
φSM = 0.3 vol%
Distance, l = 3 mm
Pressure ratio
2 Cream/SM = 10/1.5
Volume ratio
Cream/SM = 1/10
Homogenization temperature:
ϑ = 65°C
0 One stage
40 50 100 150 200 250 300 350
Homogenization pressure, Δp (bar)
FIGURe 7.10 Influence of the fat content of homogenized cream and the homogenization
pressure on the characteristic maximum droplet diameter, x903, for full-stream and partial
homogenization processing. When SEM valves are used, cream of 32 and 42 vol% fat were
mixed with skim milk (SM, φ = 0.3 vol%) within the valve. In the conventional full-stream
homogenization milk (volume fat content φ = 3.5 vol%) and cream (φ = 32 vol%) were homoge-
nized as full stream. (Data from Köhler, K. et al., Chem. Eng. Technol., 31, 12, 1863–1868, 2008.)
7.4.6 scale-up
A scale-up of disruption units can generally be realized by a geometric magnifica-
tion (scaling up) or by adding several units at same scale (numbering up). A scale-up
by a geometric magnification of orifices is usually limited as the local flow condi-
tions responsible for droplet breakup are changed. It was shown that, for example, on
simple orifices, a geometric magnification of the orifice diameter is only possible up
to a diameter d > 0.8 mm (Aguilar et al., 2008). A scale-up by increasing the number
of orifices (numbering up) is possible and technically realized, for example, in the
jet disperser® of Bayer Technology Services (BTS, Leverkusen, D) (Bayer AG, 1997,
2001). Important in this context is the distance between the different orifices, thus
the maximum number of holes per unit area. Aguilar et al. (2008) showed that a ratio
of the distance between two holes and the hole diameter has to be over 6.
Today, HPHs with flat valves are available with flow rates up to 50,000 L/h, for
example, in the dairy industry. In orifice systems, flow rates up to 5,000 L/h are
commercially available.
7.5 ConCLUsIon
HPHs are broadly used today in homogenization process. Besides valves, micro-
structured devices such as orifices of simple geometry (slit or hole) or one with
deflecting streams (e.g., Microfluidizer) are well established. Laminar, turbulent
190 Engineering Aspects of Food Emulsification and Homogenization
and cavitational flow is found in those orifices producing shear, elongational, and
Reynolds tensions, which deform and break droplets in the timescale of microsec-
onds. Today, many models are available to describe the homogenization result for
specific homogenization devices qualitatively. However, a detailed understanding of
the homogenization process and the influence of especially geometric orifice param-
eters on droplet breakup is missing. A better understanding of local flow conditions
and a focus of ongoing research are required.
For large-scale production (e.g., the dairy process) commercial homogeniza-
tion valves are still state of the art. However, especially when specific solutions, for
example, improved breakup or stabilization, are required, microstructured devices
offer real alternatives. SEM-type valves offer the chance to change local conditions
during the moment of droplet breakup and stabilization such as temperature, pH,
chemical composition, or droplet concentration.
ReFeRenCes
Aguilar, F.A., Freudig, B., Schuchmann, H.P.: Herstellen von Emulsionen in Hoch-
druckhomogenisatoren mit modifizierten Lochblenden, Chemie Ingenieur Technik,
76 (4), 396–399, 2004.
Aguilar, F.A., Köhler, K., Schubert, H., Schuchmann, H.P.: Herstellen von Emulsionen in ein-
fachen und modifizierten Lochblenden: Einfluss der Geometrie auf die Effizienz der
Zerkleinerung und Folgen für die Maßstabsvergrößerung, Chemie Ingenieur Technik,
80 (5), 607–613, 2008.
Arai, K., Konno, M., Matinaga, Y., Saito, S.J.: Effect of dispersed-phase viscosity on the maxi-
mum stable drop size for breakup in turbulent flow, Chemical Engineering of Japan,
10 (4), 325–330, 1977.
Armbruster, H.: Untersuchungen zum kontinuierlichen Emulgierprozeß in Kolloidmühlen unter
Berücksichtigung spezifischer Emulgatoreigenschaften und der Strömungsverhältnisse
im Dispergierspalt, Dissertation, Universität Karlsruhe (TH), Germany, 1990.
Baldyga, J., Orciuch, W., Makowski, L., Malski-Brodzicki, M., Malik, K.: Break up of nano-
particle clusters in high-shear devices, Chemical Engineering and Processing, 46 (9),
851–861, 2007.
Bayer AG, Patentnr.: US4996004, Preparation of pharmaceutical or cosmetic dispersions,
February 26, 1991.
Bayer AG, Verfahren und Vorrichtung zur Herstellung einer parenteralen Arzneistoffzubereitung,
1997.
Bayer AG, Dispersion nozzle with variable throughput, July 4, 2001.
Bayer MaterialScience AG, Homogenizing nozzle and method for the production of an aque-
ous two-component polyurethane lacquer emulsion, 2006.
Bechtel, S., Gilbert, N., Wagner, H.G.: Grundlagenuntersuchungen zur Herstellung von
Emulsionen im Ultraschallfeld, Chemie Ingenieur Technik, 71 (8), 810–817, 1999.
Bechtel, S., Gilbert, N., Wagner, H.G.: Grundlagenuntersuchungen zur Herstellung von
Emulsionen im Ultraschallfeld Teil 2, Chemie Ingenieur Technik, 72 (5), 450–459,
2000.
Behrend, O.: Mechanisches Emulgieren mit Ultraschall., Dissertation, Universität Karlsruhe
(TH), Germany, 2002.
Behrend, O., Ax, K., Schubert, H.: Influence of continuous phase viscosity on emulsification
by ultrasound, Ultrasonics Sonochemistry, 7 (2), 77–85, 2000.
High-Pressure Homogenization with Microstructured Systems 191
Bentley, B.J., Leal, L.G.: An experimental investigation of drop deformation and breakup
in steady, two-dimensional linear flows, Chemical Engineering Department, CIT,
December 21, 1985.
Bentley, B.J., Leal, L.G.: An experimental investigation of drop deformation and breakup
in steady two-dimensional linear flows, Journal of Fluid Mechanics, 167, 241–283,
1986.
Bondy, C., Söllner, K.: On the mechanism of emulsification by ultrasonic waves, Journal of
the Chemical Society Transactions of the Faraday Society, 31, 835–843, 1935.
Casoli, P., Vacca, A., Berta, G.L.: A numerical procedure for predicting the performance of
high pressure homogenizing valves, Simulation Modelling Practice and Theory, 18 (2),
125–138, 2010.
Chesters, A.K.: The modelling of coalescence processes in fluid- liquid dispersions: A review
of current understanding, Chemical Engineering Research and Design, 69 (4), 259–270,
1991.
Cook, E.J.: Patentnr.: 4533254, Microfluidizer (Teil I), August 6, 1985.
Cook, E.J.: Patentnr.: 4908154, Microfluidizer (Teil II), March 13, 1990.
Cook, E.J., Lagace, A.P.: Patentnr.: 4533254, Apparatus for forming emulsions, Biotechnology
Development, 1985.
Dalgleish, D.G., Tosh, S.M., West, S.: Beyond homogenization: The formation of very small
emulsion droplets during the processing of milk by a Microfluidizer, Netherlands Milk
and Dairy Journal, 50 (2), 135–148, 1996.
Danner, T.: Tropfenkoaleszenz in Emulsionen, Dissertation, Universität Karlsruhe (TH), 2001.
Darling, D.F., Butcher, D.W.: Milk-fat globule membrane in homogenized cream, Journal of
Dairy Research, 45 (2), 197–208, 1978.
Davies, J.T.: Turbulence phenomena at free surfaces, American Institute of Chemical Engineers
Journal, 18 (1), 169–173, 1972.
Davies, J.T.: Drop sizes of emulsions related to turbulent energy-dissipation rates, Chemical
Engineering Science, 40 (5), 839–842, 1985.
Eggers, J.: Nonlinear dynamics and breakup of free-surface flows, Reviews of Modern Physics,
69 (3), 865–930, 1997.
EN ISO 5167-1: Durchflussmessung von Fluiden mit Drosselgeräten in voll durchströmten
Leitungen mit Kreisquerschnitt Teil 1: Allgemeine Grundlagen und Anforderungen
(ISO 5167-1:2003); Deutsche Fassung EN ISO 5167-1:2003, 2003.
Floury, J., Bellettre, J., Legrand, J., Desrumaux, A.: Analysis of a new type of high pres-
sure homogeniser. A study of the flow pattern, Chemical Engineering Science, 59 (4),
843–853, 2004.
Freudig, B.: Herstellen von Emulsionen und Homogenisieren von Milch in modifizierten
Lochblenden, Dissertation, Universität Karlsruhe, Germany, 3-8322-3147-1, 2004.
Gaulin, A.: Patentnr.: Brecet nr. 295596, US625497, Appareil et Procédé pour la Stabilisation
du Lait, May 23, 1899.
Grace, H.P.: Dispersion phenomena in high-viscosity immiscible fluid systems and appli-
cation of static mixers as dispersion devices in such systems, Chemical Engineering
Communications, 14 (3–6), 225–277, 1982.
Hinze, J.O.: Fundamentals of the hydrodynamic mechanism of splitting in dispersion pro-
cesses, American Institute of Chemical Engineers Journal, 1 (3), 289–295, 1955.
Innings, F.: Drop Break-up in High-Pressure Homogenisers, Dissertation, Lund University,
Faculty of Engineering, 91-628-6646-X, 2005.
Jansen, K.M.B., Agterof, W.G.M., Mellema, J.: Droplet breakup in concentrated emulsions,
Journal of Rheology, 45, 227–236, 2001.
Johansen, F.C.: Flow through pipe orifices at low Reynolds Numbers, Proceedings of the
Royal Society of London. Series A, 126 (801), 231–245, 1930.
192 Engineering Aspects of Food Emulsification and Homogenization
Karasch, S., Kulozik, U.: Hochdruckhomogenisierung von Milch mit modifizierten Lochblenden
im Vergleich zu konventionellen Flachventilen, Chemie Ingenieur Technik, 80 (8), 1117–
1124, 10.1002/cite.200800079, 2008.
Karbstein, H.: Untersuchungen zum Herstellen und Stabilisieren von Öl-in-Wasser-
Emulsionen, Dissertation, Universität Karlsruhe, 1994.
Kessler, H.G.: Food and Bio Process Engineering—Dairy Technology, Verlag A. Kessler,
München, Germany, 2002.
Kissling, K., Schütz, S., Piesche, M.: Numerical investigation on the deformation of drop-
lets in high-pressure homogenizers. in High Performance Computing in Science and
Engineering ‘10 (Editors: Nagel, Wolfgang E., Kröner, Dietmar B., and Resch, Michael
M.), 287–294, Springer-Verlag, Berlin, Germany, 978-3-642-15748-6, 2011.
Kleinig, A.R., Middelberg, A.P.J.: Numerical and experimental study of a homogenizer
impinging jet, American Institute of Chemical Engineers Journal, 43 (4), 1100–1107,
1997.
Köhler, K.: Simultanes Emulgieren und Mischen, Logos Verlag, Berlin, Germany, 978-3-8325-
2716-7, 2010.
Köhler, K., Aguilar, F.A., Hensel, A., Schubert, K., Schubert, H., Schuchmann, H.P.: Design
of a microstructured system for homogenization of dairy products with high fat content,
Chemical Engineering & Technology, 30 (11), 1590–1595, 10.1002/ceat.200700266,
2007.
Köhler, K., Aguilar, F.A., Hensel, A., Schubert, K., Schubert, H., Schuchmann, H.P.: Design of
a microstructured system for the homogenization of dairy products at high fat content.
Part II: Influence of process parameters, Chemical Engineering & Technology, 31 (12),
1863–1868, 2008.
Köhler, K., Karasch, S., Schuchmann, H.P., Kulozik, U.: Energiesparende Homogenisierung
von Milch mit etablierten sowie neuartigen Verfahren, Chemie Ingenieur Technik,
80 (8), 1107–1116, 10.1002/cite.200800070, 2008.
Kolb, G.E.: Zur Emulsionsherstellung in Blendensystemen, Dissertation, Universität Bremen,
Germany, 3-8265-9204-2, 2001.
Kolmogorov, A.N.: O Droblenii Kapel V Turbulentnom Potoke, Doklady Akademii Nauk
SSSR, 66 (5), 825–828, 1949.
Meinhart, C.D., Wereley, S.T., Santiago, J.G.: PIV measurements of a microchannel flow,
Experiments in Fluids, 27 (5), 414–419, 1999.
Mielnik, M.M., Saetran, L.R.: Micro particle image velocimetry—An overview, International
Workshop on Size Effects in Microfluidics and Heat Transfer—Fundamental and
Practical Aspects, 10, 83–90, September 16, 2004.
Miller, R.: Adsorption kinetics of surfactants at fluid interfaces: Experimental conditions and
practice of application of theoretical models, Colloids and Surfaces, 46, 75–83, 1990.
Muschiolik, G., Roeder, R.-T., Lengfeld, K.: Patentnr.: DE 19530247, Druckhomogenisator,
August 17, 1995.
Ogden, L.V., Walstra, P., Morris, H.A.: Homogenization-induced clustering of fat globules in
cream and model systems, Journal of Dairy Science, 59 (10), 1727–1737, 1976.
Penth, B., Patentnr.: WO/2000/061275, Method and device for carrying out chemical and
physical processes, April 7, 2000.
Plateau, J.A.F.: Statique expérimentale et théorique des liquides soumis aux seules forces
moléculaires, Acad. Sci. Brux. Mem., 2, 1873.
Schlichting, H., Gersten, K.: Grenzschicht-Theorie, 10th ed. (Editors: Schlichting, Hermann
and Gersten, Klaus), Online-Ressource, Springer-Verlag, Berlin, Germany, 978-3-540-
32985-5, 2006.
Sheng, J., Meng, H., Fox, R.O.: A large eddy PIV method for turbulence dissipation rate estimation,
Chemical Engineering Science, 55 (20), 4423–4434, 2000.
High-Pressure Homogenization with Microstructured Systems 193
Silver, R.S.: The theories of stress due to collapse of vapour bubbles in a liquid, F. Inst. P.
Engineering (London), 154, 501ff., December 25, 1942.
Stang, M.: Zerkleinern und Stabilisieren von Tropfen beim mechanischen Emulgieren,
Dissertation, Universität Karlsruhe, Germany, 1998.
Stang, M., Karbstein, H., Schubert, H.: Influence of emulsifier adsorption kinetics and emulsi-
fication machine construction on disparity of oil-in-water emulsions in Food Colloids—
Proteins, Lipids and Polysaccharides (Editors: Dickinson, E. and Bergenstähl, B.),
382–392, 1994.
Steiner, H., Teppner, R., Brenn, G., Vankova, N., Tcholakova, S., Denkov, N.: Numerical simu-
lation and experimental study of emulsification in a narrow-gap homogenizer, Chemical
Engineering Science, 61 (17), 5841–5855, 2006.
Stevenson, M.J., Chen, X.D.: Visualization of the flow patterns in a high-pressure homogeniz-
ing valve using a CFD package, Journal of Food Engineering, 33 (1–2), 151–165, 1997.
Stone, H.A., Bentley, B.J., Leal, L.G.: An experimental-study of transient effects in the
breakup of viscous drops, Journal of Fluid Mechanics, 173, 131–158, 1986.
Taylor, G.I.: The viscosity of a fluid containing small drops of another fluid, Proceedings of the
Royal Society of London. Series A, 138 (834), 41–48, 1932.
Tcholakova, S., Lesov, I., Golemanov, K., Denkov, N.D., Judat, S., Engel, R., Danner, T.:
Efficient emulsification of viscous oils at high drop volume fraction, Langmuir, 27 (24),
14783–14796, 2011.
Tesch, S.: Charakterisieren mechanischer Emulgierverfahren: Herstellen und Stabilisieren
von Tropfen als Teilschritte beim Formulieren von Emulsionen, Dissertation, Universität
Karlsruhe (TH), Germany, 2002.
Vankova, N., Tcholakova, S., Denkov, N.D., Ivanov, I.B., Vulchev, V.D., Danner, T.:
Emulsification in turbulent flow, 1. Mean and maximum drop diameters in inertial and
viscous regimes, Journal of Colloid and Interface Science, 312, 363–380, 2007.
Walstra, P.: Effect of homogenization on the fat globule size distribution in milk, Netherlands
Milk and Dairy Journal, 29 (2–3), 279–294, 1975.
Walstra, P.: Formation of emulsions. in Encyclopedia of Emulsion Technology, Vol. 1 (Editor:
Becher, P.), 57–128, Marcel Dekker Inc., New York, 1983.
Walstra, P.: Principles of emulsion formation, Chemical Engineering Science, 48 (2), 333–349,
1993.
Walstra, P.: Casein sub-micelles: Do they exist?, International Dairy Journal, 9 (3–6), 189–192,
1999.
Walstra, P., Oortwijn, H.: The membranes of recombined fat globules. 3. Mode of formation,
Netherlands Milk and Dairy Journal, 36 (2), 103–113, 1982.
Wengeler, R.: Hydrodynamic Stress Induced Dispersion of Nanoscale Agglomerates by a High
Pressure Process, Cuvillier Verlag, Göttingen, Germany, 978-3-86727-182-0, 2007.
Wolf, F., Schuch, A., Köhler, K., Schuchmann, H.P.: Ansatz zur Beschreibung der zerkleiner-
ungsrelevanten Strömungsverhältnisse beim Emulgieren von W/O-Emulsionen mit
Lochblenden, Chemie Ingenieur Technik, 84 (12), 2215–2220, 10.1002/cite.201100065,
2012.
8 Rotor–Stator Devices
Karsten Köhler and Heike Schuchmann
contents
8.1 Introduction .................................................................................................. 195
8.2 Technical Equipment .................................................................................... 196
8.2.1 Stirred Vessel .................................................................................... 196
8.2.2 Colloid Mill ...................................................................................... 197
8.2.3 Toothed-Rim Dispersing Machines .................................................. 198
8.2.4 Extruder ............................................................................................ 198
8.2.5 Rotor–Rotor Devices ........................................................................ 199
8.2.6 Pumps ............................................................................................... 199
8.3 Emulsification Mechanism ........................................................................... 199
8.4 Emulsification Parameters ............................................................................200
8.4.1 Droplet Disruption: Processing Parameters and Geometrical
Design of RSDs ................................................................................200
8.4.2 Emulsion Recipe: Viscosity Ratio and Dispersed Phase Fraction....200
8.4.3 Stabilization of Droplets against Coalescence .................................202
8.5 Scale-Up .......................................................................................................204
8.6 Conclusion ....................................................................................................204
References ..............................................................................................................205
8.1 IntroductIon
Rotor–stator devices (RSDs) are probably the most widely used emulsifying sys-
tem. They consist of a minimum of one rotating part. The complexity of these com-
ponents ranges from simple stirrer systems, such as propeller stirrers rotating in a
vessel as a stator, to rotor–rotor systems with two rotating parts, but with no stator.
Section 8.2 explains in detail about their structural design and operation.
One of the main benefits of RSDs is the fact that they can be run in batch, semi-
batch, and alternating as well as in a continuous mode, each having its respective
merits (Figure 8.1). The batch mode offers the advantage of realizing many process
operation steps in parallel. Thus, products are mixed, pasteurized, homogenized,
195
196 Engineering Aspects of Food Emulsification and Homogenization
Continuous
Batch
Semibatch
FIGure 8.1 Operation modes in the homogenization process realized in RSDs: (a) batch
and semibatch modes and (b) continuous and alternating methods.
and cooled in one vessel, which is used for mayonnaise-type products or sauces
in food industry. However, it cannot be ensured that the product volume in total is
processed on equal terms, which presents the main disadvantage in batch process-
ing. Especially the broad distribution of stresses acting on emulsion droplets and
residence time results in a broad distribution of droplet sizes and development of
by-products, often unwanted. Therefore, extreme process conditions as required; for
example, emulsions with droplets in the submicron-size range have to be realized by
continuous process.
By the movement of the rotor, the fluid in the device is accelerated, thereby form-
ing certain flow patterns. These flow patterns lead to tensions that deform and eventu-
ally break up emulsion droplets. Flow patterns typically found in RSDs are discussed
in Section 8.3.
In Section 8.4, influences of various processes and material parameters on the
emulsification result are described. These influences are discussed and illustrated
by way of resulting droplet size distributions in the manufacture of both oil-in-water
(O/W) and water-in-oil (W/O) emulsions.
(a) (b)
(c) (d)
FIGure 8.2 Rotating part in various rotor–stator and rotor–rotor devices: (a) stirred vessel,
(b) colloid mill, (c) toothed-rim dispersing machine, and (d) extruder.
is the use of several stirred tanks. The second one is to use a single stirred tank with
different stirrers (e.g., anchor and propeller stirrer), or a stirred tank combined with
a colloid mill (see Figure 8.1). An alternative is to use different stirrer plates on a
single stirrer rod.
Dispersion discs are proposed especially for high-viscous emulsions (viscosity
around 10 Pa s) to be produced with slowly adsorbing emulsifiers (Urban et al.,
2006). Geometric modifications start with simple teeth and end with complex geom-
etries (see Figure 8.3).
The broad distribution of tensions acting on droplets and residence time is a main
disadvantage of stirred vessels. The following RSDs have the benefit of a controlled
volume in which the droplets are disrupted.
(a) (b)
(c) (d)
FIGure 8.3 Geometric modifications of various stirred vessels: (a) propeller stirrer, (b) dis-
persion disc, (c) anchor stirrer, and (d) helical ribbon agitators. (a, b: Courtesy of EKATO®; c, d:
Courtesy of Turbo-Rührwerke.)
the benefit of using one motor for both applications; however, this poses a problem:
the emulsification and pumping effect cannot be adjusted separately.
8.2.4 extruder
An extruder consists of a housing and a minimum of one screw and one die. Typical
extruders are single-screw or twin-screw extruders, equipped with one or two screws,
respectively. If two screws are mounted, they intermesh and run in a counter or coro-
tating mode (Riaz, 2000). In general, extruders of up to eight screws are fabricated,
but often the benefit of, for example, a higher and a more constant volume stream is
Rotor–Stator Devices 199
not worth the problems and costs (Guy, 2001).The length of the extruder screw can
vary from some centimeters to several meters. Extruders can only be operated in a
continuous mode. Depending on geometric parameters and screw speed, volume flow
rates are between several tens of kilogram per hour and tons per hour. Extruders are
usually the device of choice if very high viscous phases, such as polymers or waxes,
have to be mixed and emulsified. The extruder screw is equipped with elements of
different design. As some elements are designed to essentially pump the liquid to the
die, others are adapted for improved mixing or emulsification effects. Simple extrud-
ers (also called mono pumps) have screws designed only for forward pumping a fluid.
Higher tensions, as required for emulsification, are produced through backward or
kneading elements, the latter existing in neutral, forward, and backward flow direc-
tion. In backward elements, the fluid is mostly pressed back through the gap of the
screw and the housing, which also increases the residence time of the fluid in the
extruder. At the end of the extruder barrels, a die is often mounted. This die is often
designed as either a simple hole or an orifice of spherical shape, reducing the cross-
sectional area by a high degree. If a die is mounted at the end of the extruder, both the
fluid residence time and the tensions acting on the fluid increase. The gaps between
the screw elements and the housing are in the size of some micrometer to some mil-
limeter. Screw rotational speeds usually are in the range of some hundred revolutions
per minute (rpm), with high-throughput extruders running at up to 1800 rpm.
8.2.6 puMpS
The centrifugal pump is an emulsification device that is broadly distributed. The
blades of the pumps are comparable with the plates of a stirrer and have the same
effect. However, these pumps are most often adjusted to hauling a fluid, which results
in low tensions for emulsification. In some cases, these tensions are high enough for
the required emulsification effect. In this chapter, however, emulsification in or by
pumps is not further discussed, as in most cases, the pumping effect is dominant and
the emulsification effect is negligible.
Depending on the geometry, as well as circumferential and axial speeds (Taylor and
axial Reynolds number), Taylor vortices are found or flow turns turbulent (Taylor,
1923; Wimmer, 1988; Karbstein, 1994; Atiemo-Obeng, Penney, and Armenante,
2004). Some RSDs are designed for inducing additional elongational flow. Cavitation
is also found in some devices, however, often unwanted, and therefore neglected in
this chapter. Laminar and turbulent flow result in elongational or shear tensions act-
ing on droplets. Emulsification mechanisms resulting from these are explained in
detail in Section 7.3.
10
Collid mill, 30% O/W
emulsifier: LEO-10
Sauter mean diameter, x12 (μm)
FIGure 8.4 Linear trend between the energy density and Sauter mean diameter indepen-
dent of the rotor geometry of a colloid mill. Process parameters varied: rotor speed and gap
width. (Data from Karbstein, H., Untersuchungen zum Herstellen und Stabilisieren von Öl-in-
Wasser-Emulsionen, Dissertation, Universität Karlsruhe, Germany, 1994.)
of both the dispersed and the continuous phases has a major effect on droplet
break-up. As long as the dispersed phase content is low, the single-droplet scenario
can also be assumed. However, at higher droplet numbers, as found in typical
emulsions, the viscosity ratio of the dispersed phase to the emulsion is relevant
(Armbruster, 1990; Jansen, Agterof, and Mellema, 2001). As RSDs often produce
flow patters similar to those of a four-roll apparatus, this basic research result
also applies. Figure 8.5 depicts the impact of the viscosity ratio on the Sauter
mean droplet size for a toothed-rim RSD. Emulsions of O/W and W/O types were
produced with varying process parameters (rotor speed and toothed-rim design).
The dispersed phase fraction was set to 30% and a fast adsorbing emulsifier was
chosen to ensure that droplet breakup dominates coalescence. The parameter given
in the diagram for the different dots is the specific energy, Ev, that resulted from
the different processing parameters and the emulsion viscosity. For its definition,
see Chapter 7.
Independent of the specific energy input and the emulsion type, the small drop-
lets are achieved at a viscosity ratio λ in the range 0.1–1, as found in the basic
studies (Bentley and Leal, 1986). As the mean droplet size increase strongly with
the increasing viscosity ratio, we can hypothesize that droplet breakup is mainly
induced by laminar shear tensions. Elongational tensions seem to play a second-
ary role.
With this knowledge, many effects of recipe or process parameters published
in the literature for RSDs can be explained. O/W emulsions, as often found for
life science applications, are characterized by λ values well above 1. An increase
in the viscosity of the continuous phase (and thus in the viscosity of the emul-
sion, respectively) is in most cases beneficial, as λ values decrease and are closer
202 Engineering Aspects of Food Emulsification and Homogenization
20
Toothed rim dispersing machine
10
Sauter mean diameter, x12 (μm)
1
30% W/O emulsifier: PGPR-90
Ev (J/m3): 1⋅106
2⋅106
5⋅106
1⋅107
0.1
0.01 0.1 1 10 100
Viscosity ratio, λ = ηd/ηe (−)
FIGURE 8.5 Influence of the viscosity ratio between the dispersed phase and the emulsion
on the Sauter mean droplet diameter of O/W and W/O emulsions produced by a toothed-rim
dispersing machine. (–) means no dimension. (Data taken from Karbstein, H., Untersuchungen
zum Herstellen und Stabilisieren von Öl-in-Wasser-Emulsionen, Dissertation, Universität
Karlsruhe, Germany, 1994; Tesch, S., Charakterisieren mechanischer Emulgierverfahren:
Herstellen und Stabilisieren von Tropfen als Teilschritte beim Formulieren von Emulsionen,
Dissertation, Universität Karlsruhe, Germany, 2002.)
to the optimal 0.1–1 range. Same is true for increasing the dispersed phase frac-
tion (see Figure 8.6), as the emulsion viscosity increases (see Figure 8.5). Both an
increase in the dispersed phase content and the continuous phase viscosity result in
an increase in specific energy density at constant process parameters (gap width and
circumferential speed). Therefore, emulsification in RSD is improved by these two
parameters.
100
Collid mill
O/W emulsifier: LEO-10
W/O emulsifier: PGPR-90
Sauter mean diameter, x12 (μm)
10
Dispersed phase fraction
O/W:
Open symbol φ = 10%–40%
Filled symbol φ = 50%–80%
W/O:
Filled symbol φ = 1%–70%
1 4
10 105 106 107 108
Energy density, Ev (J/m3)
FIGure 8.6 Sauter mean droplet size of emulsions as function of the energy density: parame-
ters changed are the dispersed phase content φ (1%–80%) and the emulsion type (W/O and O/W).
Emulsions were produced on a colloid mill. (Data taken from Karbstein, H., Untersuchungen
zum Herstellen und Stabilisieren von Öl-in-Wasser-Emulsionen, Dissertation, Universität
Karlsruhe, Germany, 1994; Tesch, S., Charakterisieren mechanischer Emulgierverfahren:
Herstellen und Stabilisieren von Tropfen als Teilschritte beim Formulieren von Emulsionen,
Dissertation, Universität Karlsruhe, Germany, 2002.)
100
O/W emulsion
Dispersed phase content of 30%
Sauter mean diameter, x12 (μm)
10
LEO-10
Egg yolk
1
RSD
0.1
104 105 106 107 108 109
Energy density, Ev (J/m3)
FIGure 8.7 Sauter mean droplet size of emulsions as function of the energy density:
changed parameters include the production device and the emulsifier type at the constant
dispersed phase content and emulsion type. (Data from Karbstein, H., Untersuchungen
zum Herstellen und Stabilisieren von Öl-in-Wasser-Emulsionen, Dissertation, Universität
Karlsruhe, Germany, 1994.)
204 Engineering Aspects of Food Emulsification and Homogenization
8.5 scale-up
Scale-up of emulsification processes in RSDs today is well established. Nearly
all devices are available from laboratory scale up to production scale. Stirrers are
usually scaled geometrically for vessels of some milliliter to some thousand liters.
Adapting the residence time distribution is a major challenge. Solutions are summa-
rized in Zlokarnik (1999).
For scaling up continuous emulsification processes in colloid mills and toothed-
rim dispersing machines, the rotor and stator diameters are scaled accordingly.
Process parameters are adapted until the specific energy density is in the same range
(Karbstein, 1994). Figure 8.8 gives an example for colloid mills.
8.6 conclusIon
RSDs are often used in industrial emulsification, for example, for the production
of mayonnaise, sauces, creams, and lotions. They are available from laboratory
to production scale. In terms of geometry, four different devices can be differen-
tiated. Depending on production defaults and product viscosity, stirred vessels,
colloid mills, toothed-rim dispersing machines, and extruders are used. Stirred
vessels as such or in combination with a continuous working RSDs such as a col-
loid mill or a toothed-rim device is the method of choice whenever batch process-
ing is required or emulsification is only one of many processing steps in emulsion
production.
The main parameters responsible for droplet breakup are the rotor diameter, rota-
tional speed, gap width, and the geometry of the RSD, as well as emulsion viscosity.
10
Manufacturer 1
Rotor diameter
dR = 50 mm
Manufacturer 2
Rotor diameter
dR = 110 mm
1
106 107 108
Specific energy, Ev (J/m3)
FIGure 8.8 Sauter mean diameter as function of specific energy for colloid mills of differ-
ent rotor diameter and thus volume throughput. Diameter 1: pilot scale; diameter 2: produc-
tion scale. (Data from Karbstein, H., Untersuchungen zum Herstellen und Stabilisieren von
Öl-in-Wasser-Emulsionen, Dissertation, Universität Karlsruhe, Germany, 1994.)
Rotor–Stator Devices 205
reFerences
Armbruster, H.: Untersuchungen zum kontinuierlichen Emulgierprozeß in Kolloidmühlen unter
Berücksichtigung spezifischer Emulgatoreigenschaften und der Strömungsverhältnisse
im Dispergierspalt, Dissertation, Universität Karlsruhe (TH), Germany, 1990.
Atiemo-Obeng, V.A., Penney, W.R., Armenante, P.: Rotor-stator mixing devices. in Handbook of
Industrial Mixing (Editor: Paul, E.L., Atiemo-Obeng, V.A., and Kresta, S.M.), 479–505,
John Wiley & Sons, Hoboken, NJ, 2004.
Bentley, B.J., Leal, L.G.: An experimental investigation of drop deformation and breakup in
steady two-dimensional linear flows, Journal of Fluid Mechanics, 167, 241–283, 1986.
Grace, H.P.: Dispersion phenomena in high viscosity immiscible fluid systems and appli-
cation of static mixers as dispersion devices in such systems. Chemical Engineering
Communications, 14, 225–277, 1982.
Guy, R.: Extrusion Cooking: Technologies and Applications, CRC Press, Boca Raton, FL,
9781855735590, 2001.
Jansen, K.M.B., Agterof, W.G.M., Mellema, J.: Droplet breakup in concentrated emulsions,
Journal of Rheology, 45 (1), 227–236, 2001.
Karbstein, H.: Untersuchungen zum Herstellen und Stabilisieren von Öl-in-Wasser-Emulsionen,
Dissertation, Universität Karlsruhe (TH), Germany, 1994.
Riaz, M.N.: Extruders in Food Applications, Technomic Publishing Co., Inc., CRC Press,
Miami Shores, FL, 2000.
Symex GmbH&Co KG, Patent nr.: EP1825907A1, Homogenisator (“co-twister”), Veröffentli-
chungsdatum, August 29, 2007; Eingetragen, February 16, 2001; Prioritätsdatum,
February 18, 2000.
Taylor, G.I.: Stability of a viscous liquid contained between two rotating cylinders, Philosophical
Transactions of the Royal Society of London. Series A, 223 (605–615), 289–343, 1923.
Tesch, S.: Charakterisieren mechanischer Emulgierverfahren: Herstellen und Stabilisieren
von Tropfen als Teilschritte beim Formulieren von Emulsionen, Dissertation, Universität
Karlsruhe (TH), Germany, 2002.
Urban, K., Wagner, G., Schaffner, D., Röglin, D., Ulrich, J.: Rotor-stator and disc systems for
emulsification processes, Chemical Engineering & Technology, 29 (1), 24–30, 2006.
Wimmer, M.: Viscous flows and instabilities near rotating bodies, Progress in Aerospace
Sciences, 25 (1), 43–103, 1988.
Zlokarnik, M.: Rührtechnik—Theorie und Praxis, Springer-Verlag, Berlin, Germany, 3-540-
64639-6, 1999.
Section III
Low-Energy Processes
9 Microchannel
Emulsification
Aspects of Droplet
Generation, Channel Materials,
Operating Conditions, and
Scaling-Up Strategies
Isao Kobayashi, Marcos A. Neves,
and Mitsutoshi Nakajima
Contents
9.1 Introduction .................................................................................................. 210
9.2 Principle of Droplet Generation for MC Emulsification............................... 212
9.2.1 Droplet-Generation Mechanism ....................................................... 212
9.2.2 Effect of the Flow of the Dispersed and Continuous Phases............ 214
9.3 Effect of Channel Structure and Dimensions on Emulsification .................. 216
9.3.1 Grooved MC Arrays ......................................................................... 216
9.3.2 Straight-Through MC Arrays............................................................ 217
9.3.3 Effect of Channel Dimensions .......................................................... 218
9.4 Effect of Channel Materials and Their Surface Properties on
Emulsification ............................................................................................... 220
9.4.1 Silicon MC Arrays ............................................................................ 220
9.4.2 Polymer MC Arrays .......................................................................... 222
9.4.3 Stainless Steel MC Arrays ................................................................ 223
9.5 Effect of Process Factors on Emulsification ................................................. 223
9.5.1 Temperature ...................................................................................... 223
9.5.2 Viscosities of Dispersed and Continuous Phases ............................. 226
9.5.3 Electrolyte Concentration ................................................................. 226
9.6 Scaling-Up Strategies ................................................................................... 228
9.6.1 Grooved MC Array Chip .................................................................. 228
9.6.2 Asymmetric Straight-Through MC Array Chip ............................... 229
9.7 Concluding Remarks .................................................................................... 231
References .............................................................................................................. 232
209
210 Engineering Aspects of Food Emulsification and Homogenization
9.1 IntRoDUCtIon
Emulsions are immiscible liquid-liquid systems that consist of numerous droplets
dispersed in a continuous phase. Emulsions form the basis for many kinds of tradi-
tional foods, for example, milk, cream, beverages, dressings, dips, sauces, butters,
and desserts (Dickinson and McClements, 1996), and have been increasingly used in
nutritional beverages designed for infants, elderly people, and athletes, among oth-
ers (McClements, 2005). Many emulsion-based products are used in our daily life
in the form of foods (milk, butter, mayonnaise, etc.), cosmetics (facial creams, body
lotions, hair products, etc.), pharmaceuticals (drug-delivery systems, vitamin drops,
etc.), and chemicals (paints, sprays, lubricants, etc.). Among others, appearance, rhe-
ology, and stability depend on the size of emulsion droplets, as well as on their size
distributions. Traditional emulsification techniques generally require intense energy
input, resulting in polydisperse emulsions, whereas most of the energy applied is lost
as heat, which may be detrimental to heat-sensitive bioactive compounds, generally
loaded into emulsions for targeted delivery.
The physical stability of emulsions and their underlying interfacial aspects have
been extensively investigated (Walstra, 1996). Emulsions become kinetically stable
against droplet coalescence for a finite period when surfactants are added prior to
emulsification. During storage, emulsions can separate into their original constitu-
ents, oil, and aqueous phases, either due to coalescence of the droplets or due to
creaming/sedimentation resulting from density differences between the two phases.
Coalescence can be prevented by adding a surface-active compound, which reduces
coalescence by adsorbing to the interface. In addition, creaming or flocculation can
be reduced by controlling physicochemical properties, such as decreasing the droplet
size, increasing the continuous phase viscosity, or matching the densities of both the
continuous and dispersed phases.
Microchannel Emulsification 211
FIGURe 9.1 Optical micrographs of protein-stabilized O/W emulsion droplets: (a) mono-
disperse emulsion prepared using microchannel emulsification and (b) polydisperse emul-
sion prepared using a rotor–stator homogenizer. (Reproduced from Food Hydrocolloids,
20, Saito, M. et al., Comparison of stability of bovine serum albumin-stabilized emulsions
prepared by microchannel emulsification and homogenization, 1020–1028, Copyright 2006,
with permission from Elsevier.)
Monodisperse emulsions with uniform droplets are useful in both scientific and
industrial fields, as they have advantages such as higher stability and offer simpler
understanding and control of physical properties (McClements, 2004). Monodisperse
emulsions are also promising templates for preparing monodisperse microparticles
and microcapsules (Vladisavljević and Williams, 2005). Figure 9.1 illustrates poly-
disperse and monodisperse emulsion droplets prepared using different techniques
(Saito et al., 2006). The average size and size distribution of an emulsion affect the
stability and properties of emulsion-based products. For instance, fine emulsions
were reported to have higher viscosities as compared to coarse emulsions due to
their narrower size distribution, decreased droplet deformability, and stronger col-
loidal interactions (Nisisako, Torii, and Higuchi, 2004). The visual appearance of an
emulsion depends on the droplet size as well; for example, an emulsion with droplets
greater than 10 μm appears turbid, and the turbidity increases with decreasing drop-
let size (Steegmans, Schroën, and Boom, 2009), until the droplet size is less than
the wavelength of visible light; the emulsion then becomes transparent (Steegmans
et al., 2010). Further, a wide size distribution in polydisperse emulsions promotes
coalescence (Nie et al., 2008) and Ostwald ripening.
Microfabrication technology, which originated from semiconductor technology,
makes it possible to precisely fabricate micrometer-sized channels on a microfluidic
chip. Microfluidic devices enable the formulation of emulsions with well-controlled
droplet sizes and at high energy efficiency, which improves product properties and
shelf life, but also reduces the total energy expenditure. Not only single emulsions
but also multiple emulsions can be successfully produced, although their produc-
tion scale is still limited. Microfluidic emulsification of immiscible fluids is of inter-
est to several different application areas because of the high monodispersity of the
obtained emulsions. The microfluidic channels have very narrow size distribution
(typically lesser than 1%), which is a necessary condition for the direct production
of monodisperse emulsions. In contrast to the traditional emulsification devices (e.g.,
colloid mills and high-pressure homogenizers), which use intense force fields to
disrupt large, premixed droplets into smaller and relatively polydisperse droplets
212 Engineering Aspects of Food Emulsification and Homogenization
(McClements, 2004), over the last two decades, novel techniques have been devel-
oped for the gentle and controlled formation of droplets from two immiscible liquids,
without preemulsification. These techniques involve the injection of a pure dispersed
phase through microchannels (MCs) or membrane pores into a continuous phase,
so that enabling the production of uniformly sized droplets of controllable average
size, due to the possibility of regulating the droplet size according to the size of the
pores or channels, rather than turbulent flow or cavitation areas, which often occur in
traditional emulsification devices (Vladisavljević, Kobayashi, and Nakajima, 2012).
Among those novel emulsification techniques, MC emulsification, which was ini-
tially proposed by Kawakatsu, Kikuchi, and Nakajima (1997), is a very mild emulsi-
fication process driven by interfacial tension. This process can produce monodisperse
emulsions using a microfluidic channel array microfabricated on a silicon chip. The
MC array devices can afford the integration of hundreds of thousands of channels on
a single chip (Kobayashi, Mukataka, and Nakajima, 2005a), and the diameter of the
generated droplets can range from 1 μm (Kobayashi, Uemura, and Nakajima, 2007)
to several millimeter (Kobayashi, Wada et al., 2008). MC arrays can be fabricated
parallel to the MC plate surface as open microgrooves (Kawakatsu, Kikuchi, and
Nakajima, 1997) or normal to the plate surface as straight-through holes (Kobayashi
et al., 2002). Hereinafter, we will describe major features and outline recent develop-
ments on MC emulsification technique.
Slot Oil–water
interface
Channel
Flow direction 20 μm
(dispersed phase)
(a) (b) (c)
Droplet
Neck
20 μm
(d) (e) (f)
FIGURe 9.2 Visualized CFD result for droplet generation from an asymmetric straight-
through MC (tdet is the detachment time): (a) tdet: −40.0 ms, (b) tdet: 0.0 ms, (c) tdet: 35.3 ms,
(d) tdet: 41.8 ms, (e) tdet: 48.7 ms, and (f) tdet: 51.9 ms. (Data from Kobayashi, I. et al., Chem.
Eng. Sci., 66, 5556–5565, 2011.)
becomes smaller, as the dispersed phase increases over the slot outlet, and then their
pressure balance becomes opposite (Figure 9.3b, II). Rapid flow of the dispersed
phase toward the domain over the slot outlet was triggered by the change in the
internal pressure balance, forming a neck in the slot (Figure 9.2d and e). The internal
pressure at this neck sharply increased soon after the pressure balance became oppo-
site, as shown in Figure 9.3b, III. In this stage, the dispersed-phase flux at the neck
exceeds than that present in the front of the neck (van Dijke, Shroën, Boom, 2008),
leading to the instantaneous pinch-off of the neck and the completion of the detach-
ment process (Figure 9.2e and f). Such droplet-generation process, which does not
require a cross-flowing continuous phase, is periodically repeated to produce mono-
disperse emulsions by MC emulsification. Droplet generation for MC emulsification
is also a highly energy-efficient process with a typical energy input of 103–104 J/m3
(Sugiura et al., 2001; Kobayashi, Takano et al., 2008).
214 Engineering Aspects of Food Emulsification and Homogenization
ΔPd,slot,over
ΔPd,slot,in
ΔPd,MC
(a)
ΔPd,MC
12
ΔPd,slot,in
ΔPd,slot,over
ΔPd (kPa)
I II III
0
20 30 40 50 60
(b) tdet (ms)
FIGURe 9.3 (a) Three-dimensional snapshot of the oil–water interface in the detachment
process. Here, ΔPd,MC is the internal pressure of the dispersed phase at the channel outlet,
ΔPd,slot,in is the internal pressure of the dispersed phase in the slot, and ΔPd,slot,over is the inter-
nal pressure over the slot. (b) Variation of ΔPd,MC, ΔPd,slot,in, and ΔPd,slot,over as a function of
detachment time (tdet). (Data from Kobayashi, I. et al., Chem. Eng. Sci., 66, 5556–5565, 2011.)
Critical value
FIGURe 9.4 Variations of the droplet size and the droplet production rate for MC emulsifi-
cation as a function of the flow rate of the dispersed phase (pressure applied to the dispersed
phase).
Above the critical value, the resulting droplet size is sensitively varied by the flow
rate of the dispersed phase, and the droplet production rate becomes lower due to the
increase in the droplet size (Figure 9.4).
The flow transition of the dispersed phase during MC emulsification can be well
explained by a dimensionless number called the capillary number that expresses the
balance between the viscous and interfacial forces (Sugiura, Nakajima et al., 2002).
In MC emulsification, capillary number of the dispersed phase that flows in the chan-
nel (Cad,MC) is used to analyze the flow state of the dispersed phase during droplet
generation. Previous studies (Sugiura, Nakajima et al., 2002; Kobayashi, Mukataka,
and Nakajima, 2005a; Kobayashi et al., 2011) reported the critical Cad,MC values of
lesser than 0.05, below which droplets of uniform size were stably generated from
the channels. The interfacial tension is considered to be the dominant force in the
range below these critical Cad,MC values. The critical Cad,MC depends hardly on the
channel diameter, but depends greatly on the liquid viscosity, surface properties, and
channel length, as well as the operating temperature (Sugiura, Nakajima, and Seki,
2002b; Butron Fujiu, Kobayashi, Uemura et al., 2011). Above the critical Cad,MC, the
influence of the viscous force cannot be neglected, stabilizing the neck. This results
in the delayed pinch-off of the neck and the generation of nonuniformly sized large
droplets (Kobayashi et al., 2011). The internal pressure balance in and around the
neck is assumed to be a key factor for determining the droplet generation.
The influence of the flow rate of the cross-flowing continuous phase on the drop-
let size in MC emulsification has been investigated by Kawakatsu et al. (1999) and
Kobayashi, Hori et al. (2010). Kawakatsu et al. (1999) used grooved MC array
plates, and Kobayashi, Hori et al. (2010) used the asymmetric straight-through MC
array plates. The maximum Reynolds numbers of the cross-flowing continuous phase
were 15 for the grooved MC arrays and 28 for the asymmetric straight-through MC
arrays, suggesting its laminar state. For grooved MC arrays, the resulting droplet size
is not influenced by the cross-flowing continuous phase in the applied range, except in
the case of generating droplets whose size is similar to the well depth. The droplets
216 Engineering Aspects of Food Emulsification and Homogenization
as large as the well depth become smaller as the flow rate of the continuous phase
increases. For asymmetric straight-through MC arrays, the resulting droplet size that
ranged between 75 and 180 μm was much smaller than the continuous-phase domain
over the slot outlets (1000 μm height). The size of droplets smaller than 100 μm was inde-
pendent of the flow rate of the cross-flowing continuous phase (Qc) applied in this study.
On the contrary, the size of droplets larger than 100 μm became sensitive to Qc in its range
over a critical value. An analysis of the force balance during droplet generation suggests
that the decrease in the resultant droplet size above a critical Qc may be attributed to the
fact that the buoyant force becomes effective for promoting the pinch-off of the neck and
droplet detachment. The critical Qc values were sufficiently high to appropriately sweep
away the droplets from the outlet of an MC array; therefore, large droplets of uniform size
can be collected from the module by applying Qc below the critical value.
Channel
Terrace
Grooved MC array
Inlet hole
for CP
8 mm
Outlet hole
(c) 22.5 mm for CP
FIGURe 9.5 Grooved MC array chips of the dead-end and cross-flow types. (a) Three-
dimensional drawing of droplet generation from part of a grooved MC array. (b) Schematic view
of a grooved MC array chip of the dead-end type. DP stands for the dispersed phase. (c) Schematic
view of a simple grooved MC array chip of the cross-flow type. CP stands for the continuous phase;
DP stands for dispersed phase.
Microchannel Emulsification 217
devices in the world (Kikuchi et al., 1992). At least two grooved MC arrays are posi-
tioned on each MC emulsification chip. Grooved MC arrays are sealed by physically
attaching them to a flat glass plate in a module during MC emulsification. Currently
available grooved MC arrays can produce monodisperse emulsions with droplet sizes
of 1–550 μm (Kobayashi, Uemura, Nakajima, 2007; Kobayashi, Hori et al., 2012).
Modules equipped with the grooved MC array chips can be classified as both dead-
end and cross-flow types (Figure 9.5b and c). Grooved MC array chips of the dead-end
type were initially used for MC emulsification. Four grooved MC arrays were arranged
on all four sides of a chip (Kawakatsu, Kikuchi, and Nakajima, 1997). A major advan-
tage of these dead-end grooved MC array chips includes an easy understanding of the
droplet-generation behavior from the channels. However, it is usually not straightfor-
ward to collect the emulsions produced when using such chips. To make a continu-
ous operation of MC emulsification possible, Kawakatsu et al. (1999) developed the
grooved MC array chips of the cross-flow type. Cross-flow grooved MC array chips
enable the collection of the generated droplets driven by the cross-flow of a continu-
ous phase in a well between two parallel-grooved MC arrays. On a simplest cross-flow
chip, two grooved MC arrays were arranged at both longitudinal sides of the central
well used as the cross-flow channel. The purpose of cross-flow was to collect droplets
from the module and not to control the droplet size. Zhang et al. (2013) demonstrated
a successful long-term production of monodisperse oil-in-water (O/W) emulsions for
seven days using a cross-flow grooved MC array chip. Cross-flow grooved MC array
chips are more suitable for higher droplet production rates, as many parallel MC arrays
can be incorporated on a chip (Kobayashi, Hori et al., 2010). However, the cross-flow
type chips still have a low throughput of monodisperse emulsions.
Kobayashi, Hori et al. (2012) proposed emulsification using grooved nanochannel
(NC) array chips of the dead-end type for producing monodisperse submicron emul-
sions. Monodisperse submicron emulsions with droplets as small as 0.5 μm were
produced by NC emulsification.
Droplet
24 mm
Continuous
phase 10 mm
A A′
FIGURe 9.6 (a) Three-dimensional drawing of a droplet generation from part of an asym-
metric straight-through MC array. (b) Schematic view of an asymmetric straight-through MC
array chip of standard size.
(bottom) side and microslots on the downstream (top) side (Figure 9.6). The use of
these arrays enabled the stable production of monodisperse O/W and water-in-oil
(W/O) emulsions using two phases with viscosities of lesser than 1 mPa s (Kobayashi,
Mukataka, and Nakajima, 2005b; Kobayashi et al., 2009). Asymmetric straight-
through MC arrays are currently available for producing monodisperse emulsions
with droplet sizes of 10–180 μm (Kobayashi, Hori et al., 2010). When an asymmet-
ric straight-through MC array chip (Figure 9.6b) was used, the monodisperse emul-
sion droplets of n-tetradecane were successfully produced at a maximum flux of the
dispersed phase of 2700 L/(m2 h) (Vladisavljević, Kobayashi, and Nakajima, 2011).
A maximum droplet throughput per unit area of an asymmetric straight-through MC
array is independent of the channel diameter, as determined by CFD simulations
(Kobayashi et al., 2011). Asymmetric straight-through MC arrays are one of the most
commonly used microstructures for current MC emulsification researches.
10
dMC Critical Ud,MC
8 5 μm
10 μm
6 30 μm
ddrop (−)
50 μm
100 μm
4
(a)
0
dMC
8 150 μm
200 μm
6 300 μm
ddrop (−)
400 μm
4
(b)
0 2 4 6 8 10 12
Ud,MC (mm/s)
FIGURe 9.7 Variation of dimensionless droplet diameter (ddrop) obtained from the
CFD simulations as a function of dispersed-phase velocity (Ud,MC) that flows in a
channel; effect of channel size. Here dMC is the channel diameter. (a) dMC of 5 to 100 μm.
(b) dMC of 150 to 400 μm. − in the graphs stands for the lines fitted the plots. (Data
from Kobayashi, I. et al., Chem. Eng. Sci., 66, 5556–5565, 2011.)
Maximum droplet-generation rate
Droplet size
FIGURe 9.8 Variations of the droplet size as functions of terrace length (slot depth) (a) and
channel length (b).
220 Engineering Aspects of Food Emulsification and Homogenization
unstable, and the droplet size decreases with the increasing terrace length. The small-
est droplet size can be obtained by a ratio of terrace length to channel diameter of
about 2. When the terrace length exceeded the channel diameter by more than 2 folds,
the droplet size increased with the increasing terrace length. This increasing slope
became lower in a range of very high ratios of terrace length to channel diameter. It
should be stated that the terrace length is determined by the mask design and cannot be
varied during chip fabrication. In contrast, the channel height can be readily tuned by
etching time in the case of fabricating the silicon-grooved MC array chips. The droplet
size gradually increases with the increasing channel height, if the terrace length is suf-
ficiently larger than the channel size. There is a maximum channel height for a given
channel width due to wet anisotropic etching described in Section 9.4.1. Analytical
models for predicting the size of the droplets generated from the grooved MC arrays
have been proposed by Sugiura, Nakajima, and Seki (2002a, 2004) and van der Zwan,
Shroën, and Boom (2009). The size of the droplets generated using the asymmetric
straight-through MC arrays is somewhat tunable by the slot depth, which is normally
controlled by the etching period. The droplet size varies with the slot depth, as shown
in Figure 9.8a. The diameter of circular channels is defined by the mask design, sug-
gesting that it is difficult to change the channel diameter during chip fabrication.
The channel length, which does not affect the droplet size, is a major param-
eter affecting the maximum droplet-generation rate from each channel. Figure 9.8b
shows the variation of the maximum droplet-generation rate as a function of the
channel length under a certain channel diameter. For a given channel diameter, lon-
ger channels lead to the production of monodisperse emulsions at higher droplet-
generation rates (Sugiura, Nakajima, and Seki, 2002b). Appropriately, narrow and
long channels are preferable for MC emulsification, as they have a greater pressure
drop of the dispersed phase, which is given by Fanning’s equation:
ρ U 2 l
∆Pd,MC = 4 f d d MC (9.1)
2 dMC
where:
ΔPd,MC is the pressure drop of the dispersed phase in a channel
f is Fanning’s friction factor
ρd is the density of the dispersed phase
Ud is the velocity of the dispersed phase
lMC is the channel length
dMC is the channel diameter
FIGURe 9.9 Photographs of MC emulsification chips made of silicon (a), PMMA (b), and
stainless steel (c). Top and bottom photographs in each frame are grooved MC array chips and
(asymmetric) straight-through MC array chips, respectively.
tABLe 9.1
Materials and typical techniques Used for Fabrication of Microchannel
Array Chips
Inherent surface
Materials Fabrication Methods Affinity Reference
Surface- Anisotropic wet Hydrophilic Kawakatsu, Kikuchi, and
oxidized etching, chemical dry Nakajima (1997); Kobayashi,
silicon etching Mukataka, and Nakajima (2005b)
Poly(methyl Injection molding, Hydrophobic Liu et al. (2005a); Kobayashi,
methacrylate) synchrotron Hirose et al. (2008)
lithography
Stainless steel Mechanical dicing, end Hydrophilic Tong et al. (2000b); Kobayashi,
milling Hori et al. (2012)
to the inlet side of the chip (Kobayashi, Mukataka, and Nakajima, 2005b). During the
DRIE process, deep channels can be fabricated by including passivation with polymer
that prevents lateral etching of the sidewalls. The fabrication procedures of these MC
arrays are reviewed in detail in our previous publication (Vladisavljević et al., 2013).
Although silicon is an inherently hydrophobic material, its surface can be easily
modified with hydrophilic and hydrophobic treatments (Kobayashi, Takano et al.,
2008). The surface of the silicon MC arrays becomes hydrophilic after plasma-
oxidation to form a thin silicon dioxide layer. Hydrophobic MC arrays can be obtained
by surface oxidization and subsequent treatment with a silane coupling reagent.
In MC emulsification, the surface property of MC arrays has to be appropriately con-
trolled to achieve a successful production of monodisperse emulsions. A key factor
for a successful MC emulsification is to prevent wetting of the dispersed phase to the
MC array surface; that is, hydrophilic and hydrophobic channels are commonly used
to produce O/W and W/O emulsions, respectively. The contact angle of the dispersed
phase to the channel wall is an indicator that is useful for understanding the interaction
between the oil–water interface stabilized by surfactant molecules and the MC array
surface. Butron Fujiu, Kobayashi, Neves et al., (2011) demonstrated that droplets of
uniform size are stably generated via MC arrays when the interface that advances in
the channel has a high contact angle of the dispersed phase and a symmetrical shape.
Silicon MC arrays have a negatively charged surface in the water after hydrophilic
treatment, which is due to the presence of silanol group (–SiOH). To avoid the attrac-
tive interface/channel interaction, anionic and selected ionic surfactants (or proteins)
have to be used for producing monodisperse O/W emulsions by MC emulsification
(Tong et al., 2000a; Kobayashi, Nakajima, and Mukataka, 2003; Yin et al., 2005).
Silicon MC arrays after hydrophobic treatment have an uncharged surface in organic
solutions. There is a nonstrong repulsive interaction between the interface covered by
hydrophobic surfactant molecules and hydrophobic channel surface, although mono-
disperse W/O emulsions can be produced by appropriately selecting emulsion compo-
sitions and operating temperature (Kobayashi et al., 2009; Butron Fujiu et al., 2012).
Soybean oil
100 Sodium oleate
Tween 20
PGM
Viscosity, ηd,ηc (mPa s)
10
(a)
0.1
100
Viscosity ratio, ξ (−)
10
Sodium oleate
Tween 20
PGM
(b)
0.1
0 10 20 30 40 50 60 70 80
Temperature (°C)
FIGURe 9.10 (a) Viscosity of the dispersed phase (soybean oil) and various continuous
phases (sodium oleate, Tween 20 or pentaglycerol monolaurate; 1 wt% in Milli-Q water),
measured at various temperatures. (b) Variation of the viscosity ratio as a function of tem-
perature. ηd is the viscosity of the dispersed phase, ηc is the viscosity of the continuous phase,
and ξ is the viscosity ratio, defined as ηd/ηc. Lines in (a) and (b) are fitted based on the expo-
nential equations presented over the graphs.
Microchannel Emulsification 225
the viscosity of the continuous phase (ηc) at each temperature was independent of the
type of emulsifier. ηc values ranged between 0.3 and 1.4 mPa s and decreased exponen-
tially as the operating temperature increased from 10°C to 70°C. The viscosity ratio
(ξ), defined as ηd/ηc, where ηd is the viscosity of the dispersed phase, also decreased
exponentially as the operating temperature was increased, within the range applied in
this study. Figure 9.11 depicts the oil–water interface stabilized by sodium oleate, and
by Tween 20 in a rectangular channel at temperature of 70°C. The contact angle of the
dispersed phase to the channel wall (θd) was measured using a new MC array method
proposed by Butron Fujiu, Kobayashi, Neves et al. (2011), where the contact angle is
obtained by averaging the two contact angles (θd1 and θd2) indicated in Figures 9.11a
and b. The contact angle of the dispersed phase on the channel wall exceeded 130° and
decreased gradually as the temperature increased. Monodisperse O/W emulsions were
produced independent of the temperature applied and the type of emulsifier used. The
resultant droplet diameters, which ranged between 30 and 36 μm, decreased gradually
as the temperature increased. They also reported that the critical flow velocity increased
with the increasing operating temperature. Butron Fujiu et al. (2012) also evaluated the
influence of temperature on the production characteristics of W/O emulsions by MC
emulsification, using a hydrophobic MC array chip that was controlled between 10°C
and 55°C. In that case, the continuous phase included decane oil containing 5 wt% tetra-
glycerin monolaurate condensed ricinoleic acid ester, a hydrophobic emulsifier, and the
dispersed phase consisted of an aqueous solution containing 1 wt% of sodium chloride
and 5 wt% of polyethylene glycol. They reported that, at each operating temperature,
the resultant droplet diameter was also almost constant below a critical flow velocity of
the dispersed phase. Moreover, the maximum droplet-generation rate from a channel
gradually increased with the increasing operating temperature due to the decrease in
viscosity of both phases.
θd1 ≈ θd2
θd1
θd2
20 μm
(a)
θd1 > θd2
θd1
θd2
20 μm
(b)
80 60 50 50
Coefficient of variation (%)
Coefficient of variation (%)
70 50 40 40
Droplet diameter (μm)
Droplet diameter (μm)
40 40
30 30
30 30
20 20
20 20
10 10
10 10
0 0 0 0
0 0.2 0.4 0.6 0.8 1.0 1.2 0 0.2 0.4 0.6 0.8 1.0 1.2
(a) NaCl concentration (mol/L) (b) NaCl concentration (mol/L)
FIGURe 9.12 Influence of NaCl concentration on the average droplet diameter and coef-
ficient of variation of O/W emulsions produced using MC emulsification (WMS1-3 chip),
stabilized by SDS (a) and Tween 20 (b).
228 Engineering Aspects of Food Emulsification and Homogenization
(a) 100 μm
(b) 100 μm
FIGURe 9.13 Optical micrographs of oil droplets formation from the slot outlets at differ-
ent NaCl concentrations (between 0.1 and 1.0 mol/L). (a) SDS-stabilized droplets. (b) Tween
20-stabilized droplets.
Through- Through-
holes holes
(inlet) (outlet)
Grooved
MC array
50 μm
(a) (b)
FIGURe 9.14 (a) Photograph of a large grooved MC array chip of the cross-flow type. (Data
from Kobayashi, I. et al., Microfluid. Nanofluidics, 8, 252–262, 2010.) (b) Optical micrograph
showing the generation of soybean oil droplets of uniform size via part of a grooved MC
array. The continuous phase was Milli-Q water containing 1.0 wt% SDS. The flow rate of the
dispersed phase was 1.5 mL/h.
Asymmetric
straight-
40 mm
through
MC
200 μm
40 mm
(a) (b)
5-L tank
(dispersed phase)
Emulsification
module
a flow rate of the dispersed phase over 10 L/h and monodisperse emulsions with
a throughput over 100 L/h.
ReFeRenCes
Butron Fujiu, K., Kobayashi, I., Neves, M.A., Uemura, K., Nakajima, M. 2011. Effect of
temperature on production of soybean oil-in-water emulsions by microchannel emulsi-
fication using different emulsifiers. Food Science and Technology Research, 17, 77–86.
Butron Fujiu, K., Kobayashi, I., Uemura, K., Nakajima, M. 2011. Temperature effect on
microchannel oil-in-water emulsification. Microfluidics and Nanofluidics, 10, 773–783.
Butron Fujiu, K., Kobayashi, I., Uemura, K., Nakajima, M. 2012. Influence of temperature
on production of water-in-oil emulsions by microchannel emulsification. Colloids and
Surfaces A: Physicochemical and Engineering Aspects, 10, 773–783.
Dickinson, E., McClements, D.J. 1996. Advances in Food Colloids, Chapman & Hall, London,
p. 333.
Iwamoto, S., Nakagawa, K., Sugiura, S., Nakajima, M. 2002. Preparation of gelatin micro-
beads with a narrow size distribution using microchannel emulsification. AAPS
Pharmaceutical Science and Technology, 3, 25.
Kawakatsu, T., Kikuchi, Y., Nakajima, M. 1997. Regular-sized cell creation in microchan-
nel emulsification by visual microprocessing method. Journal of the American Oil
Chemists’ Society, 74, 317–321.
Kawakatsu, T., Komori, H., Nakajima, M., Kikuchi, Y., Yomemoto, T. 1999. Production of
monodisperse oil-in-water emulsion using cross-flow type silicon microchannel plate.
Journal of Chemical Engineering of Japan, 32, 241–244.
Kikuchi, Y., Sato, K., Ohki, H., Kaneko, T. 1992. Optically accessible microchannels formed in
a single-crystal silicon substrate for studies of blood rheology. Microvascular Research,
44, 226–240.
Kobayashi, I., Hirose, S., Katoh, T., Zhang, Y., Uemura, K., Nakajima, M. 2008. High-aspect-
ratio through-hole array microfabricated in a PMMA plate for monodisperse emulsion
production. Microsystem Technologies, 14, 1349–1357.
Kobayashi, I., Hori, Y., Neves, M.A., Uemura, K., Nakajima, M. 2012. Controlled produc-
tion of monodisperse submicron emulsions by nanochannel emulsification. 12th
International Conference on Microreaction Technology, February 20–22, Lyon, France.
Kobayashi, I., Hori, Y., Uemura, K., Nakajima, M. 2010. Production characteristics of large
soybean oil droplets by microchannel emulsification using asymmetric through-holes.
Japan Journal of Food Engineering, 11, 34–48.
Kobayashi, I., Mukataka, S., Nakajima, M. 2004a. CFD simulation and analysis of emulsion
droplet formation from straight-through microchannels. Langmuir, 20, 9868–9877.
Kobayashi, I., Mukataka, S., Nakajima, M. 2004b. Effect of slot aspect ratio on droplet for-
mation from silicon straight-through microchannels. Journal of Colloid and Interface
Science, 279, 277–280.
Kobayashi, I., Mukataka, S., Nakajima, M. 2005a. Production of monodisperse oil-in-
water emulsions using a large silicon straight-through microchannel plate. Industrial
Engineering & Chemistry Research, 44, 5852–5856.
Kobayashi, I., Mukataka, S., Nakajima, M. 2005b. Novel asymmetric through-hole array
microfabricated on a silicon plate for formulating monodisperse emulsions. Langmuir,
21, 7629–7632.
Kobayashi, I., Murayama, Y., Kuroiwa, T., Uemura, K., Nakajima, M. 2009. Production of
monodisperse water-in-oil emulsions consisting of highly uniform droplets using asym-
metric straight-through microchannel arrays. Microfluidics and Nanofluidics, 7, 107–119.
Kobayashi, I., Nakajima, M., Chun, K., Kikuchi, Y., Fujita, H. 2002. Silicon array of elongated
through-holes for monodisperse emulsion droplets. AIChE Journal, 48, 1639–1644.
Kobayashi, I., Nakajima, M., Mukataka, S. 2003. Preparation characteristics of oil-in-water
emulsions using differently charged surfactants in straight-through microchannel emulsi-
fication. Colloids and Surfaces A: Physicochemical and Engineering Aspects, 229, 33–41.
Microchannel Emulsification 233
Kobayashi, I., Neves, M.A., Wada, Y., Uemura, K., Nakajima, M. 2012. Microchannel emul-
sification using stainless-steel chips: Oil droplet generation characteristics. Green
Processing & Synthesis, 1, 353–362.
Kobayashi, I., Takano, T., Maeda, R., Wada, Y., Uemura, K., Nakajima, M. 2008. Straight-
through microchannel devices for generating monodisperse emulsion droplets several
microns in size. Microfluidics and Nanofluidics, 4, 167–177.
Kobayashi, I., Uemura, K., Nakajima, M. 2007. Formulation of monodisperse emulsions using
submicron-channel arrays. Colloids and Surfaces A: Physicochemical and Engineering
Aspects, 296, 285–289.
Kobayashi, I., Vladisavljević, G.T., Uemura, K., Nakajima, M. 2011. CFD analysis of micro-
channel emulsification: Droplet generation process and size effect of asymmetric
straight flow-through microchannels. Chemical Engineering Science, 66, 5556–5565.
Kobayashi, I., Wada, Y., Hori, Y., Neves, M.A., Uemura, K., Nakajima, M. 2012. Microchannel
emulsification using stainless-steel chips: Oil droplet generation characteristics.
Chemical Engineering & Technology, 35, 1865–1871.
Kobayashi, I., Wada, Y., Uemura, K., Nakajima, M. 2008. Generation of uniform drops
via through-hole arrays micromachined in stainless-steel plates. Microfluidics and
Nanofluidics, 5, 677–687.
Kobayashi, I., Wada, Y., Uemura, K., Nakajima, M. 2010. Microchannel emulsification for
mass production of uniform fine droplets: Integration of microchannel arrays on a chip.
Microfluidics and Nanofluidics, 8, 252–262.
Kobayashi, I., Zhang, Y., Neves, M.A., Hori, Y., Uemura, K., Nakajima, M. 2014. Influence of elec-
trolyte concentration on microchannel oil-in-water emulsification using differently charged
surfactants. Colloids and Surfaces A: Physicochemical and Engineering Aspects, 440, 79–86.
Kobayashi, I., Zhang, Y., Neves, M.A., Wada, Y., Uemura, K., Nakajima, M. 2013. Production
of monodisperse O/W and W/O emulsions using PMMA straight-through microchan-
nel arrays. 3rd Asia-Pacific Chemical and Biological Microfluidics Conference, August
19–22, Seoul, Korea.
Liu, H., Nakajima, M., Nishi, T., Kimura, T. 2005a. Effect of channel structure on prepara-
tion of a water-in-oil emulsion by polymer microchannels. European Journal of Lipid
Science and Technology, 107, 481–487.
Liu, H., Nakajima, M., Nishi, T., Kimura, T. 2005b. Hydrophilic modification of polymer
microchannel for preparation of oil-in-water emulsion. Nippon Shokuhin Kagaku
Kogaku Kaishi, 52, 599–604.
McClements, D.J. 2005. Food Emulsions: Principles, Practice, and Techniques. 2nd ed. CRC
Press, Boca Raton, FL, p. 301.
Neves, M.A., Ribeiro, H.S., Fujiu, K.B., Kobayashi, I., Nakajima, M. 2008. Formulation of con-
trolled size PUFA-loaded oil-in-water emulsions by microchannel emulsification using
β-carotene-rich palm oil. Industrial Engineering & Chemistry Research, 47, 6405–6411.
Neves, M.A., Ribeiro, H.S., Kobayashi, I., Nakajima, M. 2008. Encapsulation of lipophilic
bioactive molecules by microchannel emulsification. Food Biophysics, 3, 126–131.
Nie, Z., Seo, M., Xu, S., Lewis, P., Mok, M., Kumacheva, E., Whitesides, G., Garstecki, P.,
and Stone, H. 2008. Emulsification in a microfluidic flow-focusing device: Effect of the
viscosities of the liquids. Microfluidics and Nanofluidics, 5, 585–594.
Nisisako, T., Torii, T., Higuchi, T. 2004. Novel microreactors for functional polymer beads.
Chemical Engineering Journal, 2004, 101, 23–29.
Saito, M., Yin, L.-J., Kobayashi, I., Nakajima, M. 2005. Preparation characteristics of mono-
disperse oil-in-water emulsions with large particles stabilized by proteins in straight-
through microchannel emulsification. Food Hydrocolloids, 19, 745–751.
Saito, M., Yin, L.-J., Kobayashi, I., Nakajima, M. 2006. Comparison of stability of bovine
serum albumin-stabilized emulsions prepared by microchannel emulsification and
homogenization. Food Hydrocolloids, 20, 1020–1028.
234 Engineering Aspects of Food Emulsification and Homogenization
Steegmans, M.L.J., Ruiter, J. De, Schroën, K.G.P.H., Boom, R.M. 2010. A descriptive force-
balance model for droplet formation at microfluidic Y-junctions. AIChE Journal, 56,
2641–2649.
Steegmans, M.L.J., Schroën, K.G.P.H., Boom, R.M. 2009. Characterization of emulsification
at flat microchannel Y junctions. Langmuir, 25, 3396–3401.
Sugiura, S., Nakajima, M., Iwamoto, S., Seki, M. 2001. Interfacial tension driven monodis-
persed droplet formation from microfabricated channel array. Langmuir, 17, 5562–5566.
Sugiura, S., Nakajima, M., Kumazawa, N., Iwamoto, S., Seki, M. 2002. Characterization of
spontaneous-transformation-based droplet formation during microchannel emulsifica-
tion. Journal of Physical Chemistry B, 106, 9405–9409.
Sugiura, S., Nakajima, M., Seki, M. 2002a. Effect of channel structure on microchannel emul-
sification. Langmuir, 18, 5708–5712.
Sugiura, S., Nakajima, M., Seki, M. 2002b. Prediction of droplet diameter for microchannel
emulsification. Langmuir, 18, 3854–3859.
Sugiura, S., Nakajima, M., Seki, M. 2004. Prediction of droplet diameter for microchannel
emulsification: Prediction model for complicated microchannel geometries. Industrial
Engineering & Chemistry Research, 43, 8233–8238.
Sugiura, S., Nakajima, M., Tong, J., Nabetani, H., Seki, M. 2000. Preparation of monodis-
persed solid lipid microspheres using microchannel emulsification technique. Journal
of Colloid and Interface Science, 227, 95–103.
Tong, J., Nakajima, M., Nabetani, H., Kikuchi, Y. 2000a. Surfactant effect on production
of monodispersed microspheres by microchannel emulsification method. Journal of
Surfactants and Detergents, 3, 285–293.
Tong, J., Nakajima, M., Nabetani, H., Kikuchi, Y. 2000b. Production of oil-in-water micro-
spheres using a stainless steel microchannel. Journal of Colloid and Interface Science,
237, 239–248.
van der Zwan, E., Schroën, K., Boom, R. 2009. A geometry model for the dynamics of micro-
channel emulsification. Langmuir, 25, 7320–7327.
van Dijke, K.C., Isao Kobayashi, Karin Schroën, K., Uemura, K., Nakajima, M., Boom, R.
2010. Effect of viscosities of dispersed and continuous phases in microchannel oil-in-
water emulsification. Microfluidics and Nanofluidics, 9, 77–85.
van Dijke, K.C., Schroën, K.C.P.G.H., Boom, R.M. 2008. Microchannel emulsification:
From computational fluid dynamics to predictive analytical model. Langmuir, 24,
10107–10115.
Vladisavljević, G.T., Khalid, N., Neves, M.A., Kuroiwa, T., Nakajima, M., Uemura, K.,
Sato, S., Ichikawa, S., Kobayashi, I. 2013. Industrial lab-on-a-chip: Design, applica-
tions, and scale-up for drug discovery and delivery. Advanced Drug Delivery Reviews,
65, 1626–1663.
Vladisavljević, G.T., Kobayashi, I., Nakajima, M. 2011. Effect of dispersed phase viscosity on
maximum droplet generation frequency in microchannel emulsification using asymmet-
ric straight-through channels. Microfluidics and Nanofluidics, 10, 1199–1209.
Vladisavljević, G.T., Kobayashi, I., Nakajima, M. 2012. Production of uniform droplets using
membrane, microchannel and microfluidic emulsification devices. Microfluidics and
Nanofluidics, 13, 151–178.
Vladisavljević, G.T., Williams, R.A.T. 2005. Recent developments in manufacturing emul-
sions and particulate products using membranes. Advances in Colloid and Interface
Science, 113, 1–20.
Walstra, P. 1996. Emulsion stability. In: Becher, P. (Ed.) Encyclopedia of Emulsion Technology,
Vol. 4, Marcel Dekker, New York, pp. 1–62.
Zhang, Y., Kobayashi, I., Neves, M.A., Uemura, K., Nakajima, M. 2013. Long-term continu-
ous production of soybean oil-in-water emulsions by microchannel emulsification. Food
Science and Technology Research, 19, 995–1001.
10 Emulsification with
Microsieves and Other
Well-Defined
Microstructured Systems
Karin Schroën and Akmal Nazir
Contents
10.1 Introduction ................................................................................................. 236
10.2 Microsieves ................................................................................................. 238
10.2.1 Microsieve Structure and Pore Activation .................................... 238
10.2.2 Process Parameters and Droplet Formation Mechanism .............. 241
10.2.3 Toward Practical Application? ...................................................... 242
10.2.4 Metal Sieves .................................................................................. 243
10.2.5 Metal Sieves in Combination with Packed Beds ..........................246
10.3 Comparison of Emulsification with Microstructured Systems ...................248
10.4 Concluding Remarks................................................................................... 250
References .............................................................................................................. 250
235
236 Engineering Aspects of Food Emulsification and Homogenization
10.1 IntRoDUCtIon
Since Nakashima and Shimizu filed their first Japanese patent in 1986, various
designs for and products made by membrane emulsification have been reported.
Mostly, cross-flow emulsification and premix emulsification have been used, and
various good reviews are available in literature. The interested reader is referred to
Joscelyne and Trägårdh (2000; general review); Charcosset, Limayem, and Fessi
(2004; general review); Vladisavljević and Williams (2005; general overview with
many products); van der Graaf, Schroën, and Boom (2005; double emulsions);
Charcosset (2009; specific for food); Nazir, Schroën, and Boom (2010; specific for
premix emulsification); Maan, Schroën, and Boom (2011; for spontaneous droplet
formation systems); and Vladisavljević, Kobayashi, and Nakajima (2012; for an
extensive comparison also with microfluidic devices).
Membrane emulsification has been used in various applications, and, especially,
the lower energy intensity needed for production of emulsions has been celebrated as
one of the benefits of this technology, as illustrated in Figure 10.1, which is adapted
after a figure from the PhD thesis of Schröder from the Technical University of
Karlsruhe in Germany, into which data from the research group of Professor
Schubert—Karbstein and Schubert (1995), Schröder (1999), Behrend and Schubert
(2001), and Lambrich and Schubert (2005)—have been added by the last author.
The energy required in membrane emulsification using a vegetable oil at 30 vol%
with a typical viscosity of 60 mPa s can be orders of magnitude lower than that of
more classic emulsification devices such as high pressure homogenizers and colloid
mills. What also distinguishes membrane emulsification from the other technologies
is that the energy density is a function of the volume fraction of oil that is required
100
Toothed colloid mill
Cross-flow membrane Ultrasound
30 vol% oil
emulsification (ceramic High-pressure homogenizers:
Mean droplet size, x (μm)
FIGURe 10.1 Energy density for various emulsification devices. The emulsion comprises
of 30% vegetable oil with a viscosity of 60 mPa s. (Based on Schröder, V., Herstellen van
Öl-in-Wasser Emulsionen mit Microporösen Membranen. PhD thesis, Technische Hochschule
Karlsruhe, Germany, 1999 and extended by Lambrich, U. and Schubert, H., J. Membr. Sci., 257,
76–84, 2005.)
Emulsification with Microsieves 237
10,000
1,000
Required area (m2)
100
FIGURe 10.2 Required membrane surface area for the production of 30% oil emulsion at
20 m3/h. (Data were collected by Gijsbertsen-Abrahamse, A.J. et al., J. Membr. Sci., 230, 149–159,
2004, from: [1] Schröder, V. et al., J. Colloid Interface Sci., 202, 334–340, 1998; [2] Nakashima,
T. et al., Key Eng. Mater., 61–62, 513–516, 1991; [3] Fuchigami, T. et al., J. Sol-Gel Sci. Technol.,
19, 337–341, 2000; [5] Abrahamse, A.J. et al., J. Membr. Sci., 204, 125–137, 2002.)
238 Engineering Aspects of Food Emulsification and Homogenization
10.2 MICROSIEVES
Through photolithographic techniques used in the microchip industry, it is possible
to make pores of uniform size in silicon wafers. For an extensive description of the
technology and various applications, please consult the book by van Rijn (2004) or
the papers written by van Rijn and his coworkers (van Rijn and Elwenspoek, 1995;
van Rijn et al., 1997). The geometry, spacing, and size of the pores can be designed at
will (see Figure 10.3). These sieves are known for their very low resistance because
the actual layer in which the pores are placed is extremely thin (1 µm, as is also vis-
ible in the images in Figure 10.3), which results in high fluxes. Both aspects were
expected to be beneficial for emulsification, since the required surface area could be
much smaller as for other membranes; therefore, these sieves have been tested exten-
sively in our lab, within the PhD project of Anneke Gijsbertsen-Abrahamse (2003).
Cross-flow
16 Cross-flow
velocity (m/s)
Fraction active pores (%)
12 0.011
0.017
0.028
8
0.039
0
3 6 9 12 15
(b) Transmembrane pressure (kPa)
FIGURe 10.4 (a) Microscopic observation of droplet formation on microsieves (top); three
snapshots taken in fast succession showing droplets growing to sizes much larger than the
underlying pores that push neighboring droplets of the pore inducing polydispersity, and
that are not formed from all pores. (b) Active pores as function of applied pressure on a
hydrophilic microsieve with round pores of 7 μm and thickness 5 μm (bottom). (Data from
Abrahamse, A.J. et al., J. Membr. Sci., 204, 125–137, 2002.)
The layout of the microsieve is very typical and shown in Figure 10.5, with a sub-
structure onto which the actual sieve is placed.
The resistances of the substructure and sieve are indicated by Rs and Rp(pore),
and the various pressures in the system are the pressure in the continuous phase
(p0), in the substructure (ps), and the applied pressure at the feed side of the to-
be-dispersed phase (p1). Besides, there is an activation pressure that needs to be
exceeded in order to make droplets, and this pressure is related to the Laplace
pressure of the pore. The flow of the to-be-dispersed phase through the sieve is
determined by the pressures in the system and the resistance of both substructure
and sieve, which in turn is determined by the number of active pores (Nact), as
shown in Equation 10.1:
Rp p1 − p0
N act = − 1 (10.1)
Rs ps − p0
240 Engineering Aspects of Food Emulsification and Homogenization
p0
ps − pact
Φp = Nact Rp
η Rp
ps
p1 − ps Rs
Φs =
η Rs
p1
FIGURe 10.5 Schematic representation of the microsieve structure with its typical sub-
structure onto which the actual sieve is placed (right), and an overview of the various pressure
gradients that are relevant in the microsieve (left). (After Gijsbertsen-Abrahamse, A.J. et al.,
J. Membr. Sci., 230: 149–159, 2004.)
128Lp 24 ∞
Rp =
πdp
4
+ 3 1 −
dp ∑a ε
i =1
i
i +(1/ 2 )
(10.2)
The line in Figure 10.4b is a result of this model, leading to a typical activation
pressure of just over 3 bar and an increasing number of active pores at increasing
pressure. The resistance of the substructure and the pores are in the same order
of magnitude, and this also implies that if a pore starts forming droplets, locally
the pressure under that pore will be lower leading to preferential flow toward that
pore, which also does not allow neighboring pores to become active unless the
applied pressure is increased dramatically. Similar effects also play a role in regu-
lar membrane emulsification, where active pores will keep neighboring pores from
becoming active due to these pressure differences. If the pore resistance were to
be increased, the applied pressure as a whole needs to be higher to activate pores
because of increased flow resistance, but pore activation would be much more
gradual. This was also tested in the work of van der Graaf et al. (2004), and they
found that if the sieves were much thicker, all pores became active producing simi-
lar sized droplets.
Obviously, the flux of the microsieves is reduced if the longer pores are used
because of the additional resistance. As a matter of fact, the sieves used by van der
Graaf et al. (2004) are on the border of the limiting aspect ratio (width vs. length
of the pore) that technically can be made. The overall process for these thick sieves
is stable at 100% pore activation. Based on these data, and allowing for sufficient
spacing between the pores (typical porosity is 1%), the required area of the micro-
sieve was calculated for the same system that is described in Figure 10.2, and it is
clear that the required surface area is much lower, that is, in the order of 1–10 m2
Emulsification with Microsieves 241
(Gijsbertsen-Abrahamse, van der Padt, and Boom, 2004). Whether this is the deci-
sive step toward practical application is described in the respective section.
In this capillary number defined for the continuous phase, ηc is the dynamic viscos-
ity of the continuous phase, vc is the characteristic average velocity of the continuous
phase, and γow is the interfacial tension between oil and water phase. At high Ca,
either the high shear (which may be a result of high continuous phase velocity or
high viscosity or both) and/or low interfacial tension leads to small droplets. Various
others have followed this approach, and van Rijn (2004) used it for microsieves.
Rayner and Trägårdh (2002) also used the capillary number to predict the droplet
size, and found reasonable agreement, but at the same time mentioned that there
was room for improvement in order to be able to capture the operating parameters
better. Also, Abrahamse et al. (2002) started from the force balance when analyz-
ing droplet formation with microsieves and identified various factors that influenced
the droplet size beyond what was expected from the basic force balance. This was
confirmed by Lepercq-Bost et al. (2008), who used a ceramic membrane, and found
similar effects.
Real breakthroughs in understanding of the droplet formation mechanism resulted
from extensive modeling studies and investigations with microfluidic devices. As
early as 2001, Abrahamse et al. reported computational fluid dynamics results that
showed that interfacial tension and surface properties, as reflected in the contact
angle, are very relevant for the droplet size. Rayner et al. (2004) used the surface
evolver under conditions for which the force balance does not hold, and they could
predict droplet size with an average error of 8%. When combining modeling with
observation with microfluidic T-junctions, which are a model for cross-flow emulsi-
fication, van der Graaf et al. (2004) and van der Graaf, Steegmans et al. (2005) came
to the conclusion that the flow of both the continuous and to-be-dispersed phases
determines the size of the droplets that are formed. In Figure 10.6, a comparison
is shown of the microscopic observation and the result obtained through Lattice
Boltzmann simulation (van der Graaf et al., 2004; van der Graaf, Steegmans et al.,
2005; van der Graaf et al., 2006).
van der Graaf and colleagues (2005) found that droplet formation takes place in
two parts. First, the droplet needs to obtain a certain size inside the continuous phase
channel, after which the still attached droplet starts detaching. However, during this
242 Engineering Aspects of Food Emulsification and Homogenization
FIGURe 10.6 Comparison of Lattice Boltzmann simulations and microfluidic observations with
T-shaped junctions that are used as a model for cross-flow emulsification. In the bottom images,
the to-be-dispersed phase is pressurized through the vertical channel, and in the middle images this
phase is pushed toward the exit, while in the top images the droplet is snapped off from the feed.
(Data from van der Graaf, S. et al., Langmuir, 22, 4144–4152, 2006.)
phase, the droplet will still receive additional liquid from the feed channel, which
makes it grow larger. Because of this, the droplet size is not only a function of the
applied shear rate but also of the applied oil flow rate (transmembrane pressure),
which makes droplet size much less simple to predict than originally assumed, and
also makes process design more complex, since the flow of both phases needs to be
controlled accurately, unless the detachment time does not significantly contribute to
the total volume of the droplet.
To make matters even more complex, under the process conditions at which drop-
let formation takes place, dynamic interfacial tension effects are also expected to
take place, which implies that the actual value of the interfacial tension can be any-
thing between the equilibrium value that represents a surfactant saturated surface
and that of an empty interface. This was confirmed as early as 1998 by Schröder,
Behrend, and Schubert through the bursting membrane technique. van der Graaf
et al. (2004) have used droplet volume tensiometry to extrapolate interfacial tension
values to the conditions applied for droplet formation on a microsieve, and they also
showed that dynamic interfacial tension effects play a role there. Measuring dynamic
interfacial tensions under the conditions used in emulsification is far from trivial, but
also here there seems to be a breakthrough thanks to microfluidic investigations with
Y-shaped junctions as described by Steegmans et al. (2009).
even cease to occur as illustrated in the simulation work of Abrahamse et al. (2001).
Therefore, surface modification of the sieves to prevent adhesion of especially pro-
teins has been investigated extensively in the works of Arafat et al. (2004, 2007) and
Rosso et al. (2009) in which they show that adsorption of bovine serum albumin
and fibrinogen could be prevented by mildly attaching ethylene oxide chains to the
surface.
The price of the low porosity, high resistance microsieves that are required for
stable operation during emulsification will be higher as for regular microsieves and
much higher as for other membranes. But given the reduction in required surface
area, they still may be interesting, although in our opinion it is more likely that
application will not take place for a bulk product but for a niche market. Further
reduction in price may take place simply through the scale of production, but alter-
natively also production of polymeric microsieves, as proposed by Vogelaar et al.
(2005), is an interesting development. They used a template that is made by the
same techniques that are used to make the SiN microsieves, but now they make a
negative image from which repeatedly polymeric casts can be made. This consid-
erably reduces the costs of the microsieves; if, for example, 100 casts can be made
from one template, the cost would roughly reduce to about 1% of that of the SiN
microsieve.
Besides sieve design, different process designs have also been published,
mostly focusing on application of shear through more than just the flowing liquid.
For example, Stillwell and coworkers (2007) investigated a stirred cell with a
membrane at the bottom resulting in emulsions that were rather polydisperse due
to the differences in shear rates across the membrane. Rotating systems using
(metal) membranes with uniform pores were presented by Verena (Eisner)-
Schadler (2006), Verena (Eisner)-Schadler (2007), Aryantia et al. (2006), and
Yuan and coworkers (2008). This resulted in better control over droplet size; how-
ever, the droplets were rather large due to the fact that pores in metal sieves can-
not be made as small as in other microstructured systems.
Pore size: 4 μm × 4 μm
Thickness: 60 μm
Porosity: 2.65%
Supporting mesh on back side
FIGURe 10.7 Examples of metal sieves (Courtesy of Stork Veco B.V., Eerbeek, the
Netherlands) with squared and rectangular pores that were used in premix emulsification by
Nazir et al. (Data from Nazir, A. et al., J. Membr. Sci., 383, 1–2, 116–123, 2011; Nazir, A.
et al., Chem. Eng. Sci., 93, 173–180, 2013a.)
35
30
Droplet diameter, d32 (μm)
25
20
15
10
0
0 1 2 3 4 5 6
No. of passes, N (−)
FIGURe 10.8 Droplet size as function of number of passes through a metal sieve with rect-
angular pores with 11.6 µm width operated at (⚪) 50, (⬦) 100, and (Δ) 200 kPa pressure. The
sieves were operated from the front (empty mark) and back (filled mark); for more informa-
tion see Figure 10.7. (Data from Nazir, A. et al., J. Membr. Sci., 383, 1–2, 116–123.)
Emulsification with Microsieves 245
300
250
Droplet Weber number, Wed (−)
200
150
100
50
0
0 100 200 300 400 500
Droplet Reynolds number, Red (−)
FIGURe 10.9 Droplet Weber number versus droplet Reynolds number: (◻) 4 μm square
pores, ( ◼) 7.1, (○) 10.6, (⚫) 11.6, (Δ) 12.8, and (▴) 13.2 μm sieve; the last five sieves all
have rectangular pores and only the width of the pore is given here; the length is approximately
300 μm for all.
As expected, the droplet size decreased with increasing number of passes and also
the applied pressure as illustrated in Figure 10.8 for a sieve with rectangular pores.
What was further noted was that the droplet size is independent of the way the pre-
mix is pushed through the sieve, and this indicates that the structure of the sieve is
not that relevant for droplet breakup. The actual fluxes at which the emulsions were
produced are very high (can be well beyond 1000 m3/m2 h depending on the applied
pressure), and that is orders of magnitude higher as reported for cross-flow emul-
sification, for example, by Vladisavljević, Surh, and McClements (2006). Besides,
the flux stays constant during operation, indicating that the system is not sensitive
to fouling. On the downside, the span values of the emulsions are on the high side,
typically around 1.2–1.4 (Nazir, Schroën, and Boom, 2011, 2013a).
For squared pores, the droplet size reduction was much less in spite of their much
smaller size, and this is due to a difference in droplet breakup mechanism, as illus-
trated in a dimensionless plot (Figure 10.9) in which the droplet Weber number (ratio
of shear and interfacial tension) is plotted versus the droplet Reynolds number for
data for the first pass.
The following definitions were used by Nazir, Schroën, and Boom (2013a). The
hydraulic Reynolds number inside the pore, Rep, is
ρvdh
Re p = (10.4)
ηc
with ηc being the continuous phase viscosity and dh, the pore hydraulic diameter,
being 2lw/(l + w), with l the length of the pore, and w its width. The droplet Reynolds
number, Red, was defined by Nazir (2013) using the pore Reynolds number suggested
by Van Dinther et al. (2012):
246 Engineering Aspects of Food Emulsification and Homogenization
2
d
Re d = Re p 32,in (10.5)
2 dh
where:
d32,in is the ingoing Sauter mean droplet diameter
The droplet Weber number, Wed, which is a ratio between the inertial forces (as a
result of local pressure fluctuations) and interfacial tension forces on a droplet, was
defined in Nazir, Schroën, and Boom (2011) as
d32,outρv 2
We d = (10.6)
2σ
in which d32,out is the Sauter mean diameter of the droplet that is produced and σ is
the droplet interfacial tension, for which the equilibrium value of 5.8 mN/m is used.
The sieves with rectangular pores have about five times higher droplet Weber
number, and all behave similarly but different from square pores. The inertial forces
seem to be more important for sieves having rectangular pores and are facilitated
at higher droplet Reynolds number. The flow through the rectangular pores was not
well developed, and partial turbulent conditions may have existed after the pores
therewith contributing to droplet breakup (Nazir, Schroën, and Boom, 2011). For
the square pores, at low values for Reynolds number, spontaneous droplet breakup
due to Laplace pressure differences may be more important, indicating that under
these conditions there is an effect of pore geometry on the mean droplet size.
Basically, Figure 10.9 contains all parameters needed to design emulsification pro-
cesses with metal sieves, but as mentioned the size distribution of the prepared emulsions
is not that sharp at span values between 1.2 and 1.4. To improve on this, Nazir (2013) and
Nazir, Schroën, and Boom (2013b) decided to deposit a layer of beads on the metal sieves
as was previously suggested by van der Zwan, Schroën, and Boom (2008), in order to
allow for more breakup mechanisms to occur as was deduced from microfluidic observa-
tions done by the same author (van der Zwan et al., 2006). The results of the investiga-
tions with packed beds are summarized in the next section.
0.5 1.2
0.4
1.0
d32/dv (−)
0.3
δ (−)
0.8
0.2
FIGURe 10.10 (a) Dimensionless droplet diameter, d32/dv, and (b) droplet span, δ, as a func-
tion of Reynolds number; Re: (Δ) particle size (related to void size dv) and (○) bed height varied, H.
numbers. Both data series start off separately at relatively low Reynolds number but
then merge at Reynolds number above 40. At low Reynolds numbers, the void size
has a greater influence on the droplet size than the velocity, indicating that the con-
striction needed for spontaneous droplet snap-off is here dominant in the size reduc-
tion; this is corroborated by the reduction of the span. With increasing Reynolds
number, both curves are the same, which indicates that the breakup mechanism
becomes similar (a decreasing bed height or an increasing void size), and this is due
to the dominance of inertial effects (i.e., droplet breakup due to local shear forces).
This inertial droplet breakup region is characterized by a decrease in droplet unifor-
mity (Nazir, Schroën, and Boom, 2013b). The size of the droplets is also influenced
by the viscosity ratio of both phases, and that is depicted in Figure 10.11a, in which
the ratio of droplet size over void size is shown.
The droplet size consistently became smaller with increasing number of passes
and was smallest at low viscosity ratio because the droplets can be effectively elon-
gated by the highly viscous continuous phase, after which the liquid threads will break
into small droplets. The flux that is shown in the right-hand side plot in Figure 10.11
increases till a viscosity ratio of 3 due to a decrease in continuous phase viscos-
ity, which also results in flow conditions that are less effective for droplet breakup,
because the exerted shear on the dispersed phase will decrease. At a viscosity ratio of
0.4 600
0.3
J (m3 m−2 hr−1)
400
d32/dv (−)
0.2
200
0.1
0 0
0.01 0.1 1 10 100 1000 0.01 0.1 1 10 100 1000
(a) ηd/ηc (−) (b) ηd/ηc (−)
FIGURe 10.11 Dimensionless droplet diameter, d32 /dv, (a) and flux, J, (b) obtained after
different passes as a function of viscosity ratio: (Δ) 1st pass, (□) 3rd pass, and (♦) 5th pass.
(Data from Nazir, A. et al., Chem. Eng. Sci., 116: 547–557, 2014.)
248 Engineering Aspects of Food Emulsification and Homogenization
3, the flux reaches its maximum after which it decreases again, most probably because
of the emulsion droplets that now have to be broken up by a different mechanism, as
they cannot be easily elongated by the continuous phase. Interaction of the droplets
with the pore walls (glass beads) and constriction effects now become important, most
probably leading to higher effective viscosity inside the beds and lower fluxes. Still the
actual flux values that were found are high, and as was the case with the metal sieves,
the values stay constant, indicating that the packed beds are not prone to fouling.
At a low viscosity ratio, it is expected that droplet breakup will take place upon exit
of the bed, as was also found for metal sieves (Nazir, Schroën, and Boom, 2011, 2013a).
At higher viscosity ratio, the dispersed phase will still be able to intrude into the porous
bed, but breakup will mainly take place due to constriction inside the bed, leading to a
constant droplet size, as the droplet size now is determined by the average pore size in
the packed bed. This is also reflected in the slightly lower fluxes at high viscosity ratio,
indicating that the dispersed phase inside the packed bed impedes the flow. This con-
clusion is strengthened by the observation that the flux increases systematically, albeit
slightly, at higher number of passes corresponding to smaller droplets.
All droplet sizes were much smaller than the pore size. The droplet-to-pore size
ratio was as low as 0.1, which is the lowest ratio already reported for premix mem-
brane emulsification (Vladisavljević, Kobayashi, and Nakajima, 2012), whereas in
cross-flow membrane emulsification, the droplet size is typically 2–10 times the pore
size (Charcosset, Limayem, and Fessi, 2004). This also indicates that pore sizes need
to be much smaller for cross-flow emulsification to produce droplets of the same size,
which is also reflected in the typical fluxes in cross-flow membrane emulsification
that are orders of magnitude lower than the fluxes observed for the metal sieves and
the combination with packed beds. Besides, the span of the emulsions was consider-
ably reduced compared to the use of metal sieves only, and depending on the emul-
sion composition and process condition, this could be as low as 0.75.
For both the effects of viscosity, and the variation of bed height and bead size, scaling
relations were successfully derived (Nazir, 2013). The general shape of the equation is
γ ζ
d32 H η
= αEV−β d (10.7)
dv D ηc
with Ev being the energy density; H the bed height; D the particle diameter; ηc and
ηd the viscosity of the continuous and dispersed phases, respectively; and α, β, γ,
and ζ fit parameters. These equations can be used to estimate the droplet size pro-
duced with all systems based on metal sieves and cover all the process conditions
and product properties (Nazir, 2013; Nazir, Schroën, and Boom, 2013b).
tABLe 10.1
A Comparison of Conventional and Microstructured emulsification systems
with Premix emulsification
Premix emulsification
Conventional Microstructure
emulsification emulsification Membrane sieve Packed Bed
Small droplet size ++ + + + +
Droplet uniformity − +++ ++ + ++
Energy efficiency −− ++ ++ + +
Productivity ++ −− + ++ +
Fouling issues −− ++ ++ − −
Note: Number of + or − signs indicates how well a technique performs on a specific characteristic.
100
10
Volume (%)
0.1
0.01 0.1 1 10 100
Droplet diameter (μm)
FIGURe 10.12 Droplet size distributions obtained using different emulsification systems:
(□) Microfluidizer, 1100 bar, n = 2; (×) Microfluidizer, 50 bar, n = 1; (Δ) cross-flow mem-
brane emulsification, dp = 0.4 μm, 3.3 bar. (All three from Vladisavljević, G.T. et al., Colloids
Surf., A, 232, 199–207, 2004.); (●) packed bed system, dp = 23 μm, 2 bar, n = 5. (Data from
Nazir, A. et al., Chem. Eng. Sci., 92, 190–197, 2013b.); (⬦) EDGE system, 1.2 μm. (Data from
van Dijke, K.C. et al., Am. Inst. Chem. Eng. J., 56, 3, 833–836, 2010.); (+) Microchannel,
16 μm. (Data from Vladisavljević, G.T. et al., Colloids Surf. A, 232, 199–207, 2004.)
ReFeRenCes
A.J. Abrahamse, R. van Lierop, R.G.M. van der Sman, A. van der Padt, R.M. Boom. 2002.
Analysis of droplet formation and interactions during cross-flow membrane emulsifica-
tion. Journal of Membrane Science 204: 125–137.
A.J. Abrahamse, A. van der Padt, R.M. Boom, W.B.C. de Heij. 2001. Process fundamentals
of membrane emulsification: Simulation with CFD. American Institute of Chemical
Engineers Journal 47 (6): 1285–1291.
A. Arafat, M. Giesbers, M. Rosso, E.J.R. Sudhölter, C.G.P.H. Schroën, R.G. White et al. 2007.
Covalent biofunctionalization of silicon nitride surfaces. Langmuir 23: 6233–6244.
A. Arafat, K. Schroën, L. de Smet, E. Sudhölter, H. Zuilhof. 2004. Tailor-made functionalization
of silicon nitride surfaces. Journal of the American Chemical Society 126: 8600–8601.
Emulsification with Microsieves 251
A.A. Maan, K. Schroën, R. Boom. 2011. Spontaneous droplet formation techniques for mono-
disperse emulsions preparation—Perspectives for food applications. Journal of Food
Engineering 107 (3–4): 334–346.
T. Nakashima, M. Shimizu. 1986. Porous glass from calcium alumino boro-silicate glass.
Ceramics Japan 21: 408.
T. Nakashima, M. Shimizu, M. Kukizaki. 1991. Membrane emulsification by microporous
glass. Key Engineering Materials 61–62: 513–516.
A. Nazir. 2013. Premix emulsification systems. PhD thesis, Wageningen University, the Netherlands.
A. Nazir, R.M. Boom, K. Schroën. 2014. Influence of the emulsion formulation in premix
emulsification using packed. Chemical Engineering Science 116: 547–557.
A. Nazir, K. Schroën, R. Boom. 2010. Premix emulsification: A review. Journal of Membrane
Science 362 (1–2): 1–11.
A. Nazir, K. Schroën, R. Boom. 2011. High-throughput premix membrane emulsification
using nickel sieves having straight-through pores. Journal of Membrane Science 383
(1–2): 116–123.
A. Nazir, K. Schroën, R. Boom. 2013a. The effect of pore geometry on premix membrane
emulsification using nickel sieves having uniform pores. Chemical Engineering Science
93: 173–180.
A. Nazir, K. Schroën, R. Boom. 2013b. Droplet break-up mechanism in premix emulsification
using packed beds. Chemical Engineering Science 92: 190–197.
S.J. Peng, R.A. Williams. 1998. Controlled production of emulsions using a crossflow mem-
brane. Part I: Droplet formation from a single pore. Chemical Engineering Research and
Design 76: 894–901.
M. Rayner, G. Trägårdh. 2002. Membrane emulsification modelling: How can we get from
characterisation to design? Desalination 145: 165–172.
M. Rayner, G. Trägårdh, C. Trägårdh, P. Dejmek, P. 2004. Using the surface evolver to
model droplet formation processes in membrane emulsification. Journal of Colloid and
Interface Science 279 (1): 175–185.
C.J.M. van Rijn (ed.) 2004. Nano and Micro Engineered Membrane Technology. Membrane
Science and Technology Series 10. Elsevier, Amsterdam, the Netherlands. ISBN 0444514899,
9780444514899, p. 384.
C.J.M. van Rijn, M.C. Elwenspoek. 1995. Micro filtration membrane sieve with silicon micro
machining for industrial and biomedical applications. Proceedings of the Institute of Electrical
and Electronics Engineers 29: 83–87.
C.J.M. van Rijn, M. van der Wekken, W. Nijdam, M. Elwenspoek. 1997. Deflection and
maximum load of microfiltration membrane sieves made with silicon micromachining.
Journal of Microelectromechanical Systems 6: 48–54.
M. Rosso, M. Giesbers, A. Arafat, K. Schroën, H. Zuilhof. 2009. Covalently attached organic
monolayers on SiC and SixN4 surfaces: Formation using UV light at room temperature.
Langmuir 25 (4): 2172–2180.
V. Schröder. 1999. Herstellen van Öl-in-Wasser Emulsionen mit Microporösen Membranen.
PhD thesis, Technische Hochschule Karlsruhe, Germany.
V. Schröder, O. Behrend, H. Schubert. 1998. Effect of dynamic interfacial tension on the
emulsification process using microporous, ceramic membranes. Journal of Colloid and
Interface Science 202: 334–340.
M.L.J. Steegmans, A. Warmerdam, C.G.P.H. Schroën, R.M. Boom. 2009. Dynamic interfacial
tension measurements with microfluidic Y-junctions. Langmuir 25 (17): 9751–9758.
M.T. Stillwell, R.G. Holdich, S.R. Kosvintsev, G. Gasparini, and I.W. Cumming. 2007. Stirred
cell membrane emulsification and factors influencing dispersion drop size and unifor-
mity. Industrial & Engineering Chemistry Research 46: 965–972.
S. Sugiura, M. Nakajima, S. Iwamoto, M. Seki. 2001. Interfacial tension driven monodis-
persed droplet formation from microfabricated channel array. Langmuir 17: 5562–5566.
Emulsification with Microsieves 253
Contents
11.1 Introduction ................................................................................................. 256
11.1.1 Membrane Dispersion Processes................................................... 257
11.1.2 Membrane Treatment of Dispersions ............................................ 258
11.1.3 Comparison of Membranes with Other Methods to Generate
and Treat Dispersions ....................................................................260
11.2 SPG Membrane ........................................................................................... 261
11.2.1 Fabrication of SPG Membrane ...................................................... 261
11.2.2 Properties of SPG Membrane........................................................264
11.2.3 Surface Modification of SPG Membrane ......................................266
11.3 Emulsification Using SPG Membrane ........................................................ 267
11.3.1 Factors Affecting Droplet Size in DME........................................ 270
11.3.1.1 Influence of Transmembrane Pressure and Flux .......... 271
11.3.1.2 Influence of Membrane Pore Size and Shear Stress
on the Membrane Surface ............................................. 273
11.3.1.3 Influence of Surfactant ................................................. 275
11.3.2 Factors Affecting Droplet Size in PME ........................................ 275
11.3.3 Applications of Direct and Premix Membrane Emulsification
Using SPG Membrane ................................................................... 278
11.4 Gas Dispersion Using SPG Membrane ....................................................... 285
11.5 Conclusions ................................................................................................. 286
References .............................................................................................................. 286
255
256 Engineering Aspects of Food Emulsification and Homogenization
ABSTRACT This chapter deals with the production, properties, and macrofluidic
applications of Shirasu porous glass (SPG) membrane. The first section provides an
overview of the membrane microfluidic processes used for production and modification
of liquid–liquid and gas–liquid micro and nanodispersions, such as direct and premix
membrane emulsification with and without phase inversion, membrane demulsifica-
tion, membrane micromixing/direct precipitation, and micro and nanobubbling. In the
last section of this chapter, SPG membranes are compared with conventional homog-
enizers and microfluidic drop generators in terms of production rate, droplet size uni-
formity, and applied shear stresses. The second section deals with the fabrication of
SPG membrane by spinodal decomposition in Na2O─CaO─Al2O3─B2O3─SiO2 type
glass and morphological, mechanical, and hydrodynamic properties of SPG mem-
brane. This chapter also covers modification of surface charge, contact angle, and
porosity of SPG membrane using different physical and chemical methods, such as
deposition of silica nanoparticles onto membrane surface, coating with silicon resin,
filling the pores with solvent-responsive polymer chains, and chemical modification
with silane coupling agents. The fourth section is focused on the effects of physical
properties of the dispersed and continuous phases, operating parameters, and mem-
brane properties on the droplet size in direct and premix SPG membrane emulsifi-
cation. In addition, the most common classes of micro and nanoparticles fabricated
using SPG membrane are reviewed, and their fabrication routes are discussed. It was
concluded that a broad variety of different chemical and physicochemical processes
can be combined with SPG membrane emulsification to convert droplets into uniform
particle. The last section briefly discusses the generation of micro and nanobubbles
using SPG membrane.
Keywords: membrane emulsification, Shirasu porous glass membrane, nanoparticles,
polymeric microspheres, microbubbles, Janus particles, core-shell particles
11.1 IntroduCtIon
Synthetic membranes are mainly used for separation purposes and to achieve a
chemical or biochemical conversion. Membrane separation processes are character-
ized by the fact that a feed stream is divided into two product streams of differ-
ent chemical compositions: retentate and permeate (Figure 11.1a) (Mulder, 1996).
A shear rate is applied at the retentate–membrane interface to limit concentration
polarization and accumulation of the rejected solids on the high pressure side of the
membrane. In the last two decades, microfluidic applications of membranes (forma-
tion of droplets and bubbles, micromixing of miscible liquids, droplet breakup, and
coalescence, etc.) are gaining in popularity as a result of rising global interest in
microfluidic technologies. Membrane microfluidic processes can be classified into
two groups: (1) formation of dispersions (gas–liquid, liquid–liquid, and solid–liquid)
(Figure 11.1b) and (2) treatment of dispersions (demulsification, homogenization,
and phase inversion). In a membrane dispersion process (Figure 11.1b), phase I is
injected through a microporous membrane into phase II for the purpose of (1) mix-
ing of two miscible fluids, usually two liquid phases; (2) forming droplets or bubbles
of phase I into phase II. Membrane treatment of dispersions (Figure 11.1c) involves
Formation and Modification of Dispersions 257
Product
Feed Retentate Phase II Product
P1 P2 P2
P2 P1 P1
(a) Permeate (b) Phase I (c) Feed
Water or
water-
Oil Gas soluble
organic
solvent
FIGure 11.2 Membrane dispersion processes with hydrophilic membrane: (a) production
of O/W emulsion by DME (Data from Nakashima, T. et al., Adv. Drug Deliv. Rev., 45, 47−56
2000.); (b) production of microbubbles (Data from Kukizaki, M. and Goto, M., Colloids Surf. A,
293, 87−94, 2007.) and nanobubbles (Data from Kukizaki, M. and Goto, M., J. Membr. Sci.,
281, 386−396, 2006.); and (c) production of nanoparticles by membrane micromixing/direct
precipitation method. (Data from Chen, G.G. et al., Powder Technol., 139, 180−185, 2004.)
258 Engineering Aspects of Food Emulsification and Homogenization
are needed to produce water-in-oil (W/O) emulsions (Cheng et al., 2008; Jing et al.,
2006), and hydrophilic membranes are required to prepare oil-in-water (O/W)
emulsions (Figure 11.2a). In membrane microbubbling, a pressurized gas is forced
through a hydrophilic membrane into aqueous continuous phase, leading to the for-
mation of microbubbles (1 μm < dbubble < 1 mm) or nanobubbles (dbubble < 1 μm),
depending on the pore size of the membrane (Figure 11.2b). Micromixing is the
interpenetration of miscible solutions at the molecular level, and it is a crucial step
in any homogeneous reaction (Okhonin et al., 2011). In membrane micromixing,
an organic solution containing water-miscible organic solvent or an aqueous solu-
tion penetrates through a hydrophilic membrane into another aqueous phase for the
purpose of mixing two solvents rapidly with each other. Membrane micromixing can
be combined with direct precipitation to produce inorganic (Chen, Luo, Xu et al.,
2004) and organic (Laouini et al., 2011) nanoparticles. Precipitation of inorganic
compounds requires dispersion of an aqueous solution of water-soluble salt A into
an aqueous solution of water-soluble salt B, and nanoparticles are formed as a result
of chemical reaction between the two salts: A + B → C + D, where one of the
products is sparingly soluble in water (Table 11.1). Precipitation of organic com-
pounds requires dispersion of a water-miscible organic solvent containing particle-
forming organic compounds into an aqueous phase (antisolvent), and precipitation
occurs as a result of the lower solubility of the organic solutes in the aqueous phase
(Figure 11.2c). Particle-forming organic compounds in pharmaceutical nanodisper-
sions are active principle ingredients (APIs); excipients and typical organic phase
compositions are listed in Table 11.2.
tABLe 11.1
Formation of Inorganic nanoparticles by Membrane Micromixing/direct
Precipitation Method
Membrane
salt A salt B and Pore size references
BaSO4 nanoparticles 0.1−0.3M BaCl2 0.1M NaSO4 5 μm stainless steel, Chen, Luo, Xu
(d = 20−200 nm) 0.2−0.9 μm Ni et al. (2004)
Anatase-TiO2 0.03−0.15M 0.1−0.3M 0.2 μm Ni Chen, Luo, Yang
nanoparticles Ti(SO4)2 NH4HCO3 et al. (2004)
(d = 9−20 nm)
ZnO nanoparticles 0.2−1.2M ZnSO4 2.25M 5 μm stainless steel Wang et al.
(d = 9.4−14 nm) NH4HCO3 (2010)
Formation and Modification of Dispersions 259
tABLe 11.2
Formation of organic nanoparticles by Membrane Micromixing/direct
Precipitation Method
Membrane
excipients solvent and API and Pore size references
BDP-loaded 20−60 mg ml–1 Lipoïd® Ethanol + 0.4 0.4−10.2 μm Jaafar-Maalej
liposomes E80 + 4−12 mg mg/ml BDP SPG et al. (2011)
(d = 60−200 nm) ml–1 Chl
SPL-loaded 20−80 mg/ml Lipoïd® Ethanol + 3 mg 40 nm PP Laouini et al.
liposomes E80 + 4−16 mg ml–1 SPL hollow fiber (2011)
(d = 110−190 nm) ml−1 Chl
Vitamin E-loaded 5 mg/ml PCL Acetone + 4 mg 0.2−10.2 μm Khayata
PCL nanoparticles ml–1 vitamin E SPG et al. (2012)
(d = 250−350 nm)
Caffeine and 105 mM Tw + 105 mM Ethanol + 10 mg 0.9 μm SPG Pham et al.
SPL-loaded Chl + 23.3 mM DCP ml–1 caffeine or (2012)
niosomes 3 mg ml–1 SPL
(d = 111−115 nm)
API, Active principle ingredient; BDP, beclomethasone dipropionate; Chl, cholesterol; DCP, dicetyl
phosphate; Lipoïd® E80, egg yolk lecithin from Lipoïd GmbH; PCL, polycaprolactone; PP, polypropyl-
ene; SPL, spironolactone; Tw, Tween 60.
O/W W/O
emulsion emulsion Oil
layer
Hydrophilic Hydrophobic
membrane membrane Hydrophilic
membrane
O/W
W/O
premix
emulsion
FIGure 11.3 Treatment of emulsions using membranes: (a) production of O/W emulsion
by PME (Data from Suzuki, K. et al., Food Sci. Technol. Int. Tokyo, 2, 43−47, 1996.);
(b) production of W/O emulsion by PME with phase inversion (Data from Suzuki, K. et al.,
Food Sci. Technol. Int. Tokyo, 5, 234−238, 1999.); and (c) demulsification of W/O emulsion.
(Data from Kukizaki, M. and Goto, M., J. Membr. Sci., 322, 196−203, 2008.)
valve homogenizers it ranges from 50 to over 2000 bar. In addition, energy input
in conventional dispersion devices is not spatially uniform; for example, in rotor-
stator devices, shear forces are high in close proximity to a rotor and low in “dead
zones,” leading to the production of polydispersed emulsions. On the other hand, in
the majority of membrane dispersion processes, shear is uniformly distributed over
the membrane surface.
Another advantage of membrane emulsification compared to conventional emul-
sification devices is that membrane systems allow integration of emulsification step
and emulsion post-processing to achieve simultaneous drop generation and sepa-
ration, chemical/biochemical conversion, or physicochemical transformation. The
examples include integration of DME or PME with liquid–liquid extraction (Chen,
Luo, Sun et al., 2004, Xu et al., 2005), biphasic enzymatic transformation (Li and
Sakaki, 2008; Mazzei et al., 2010), pervaporation (Chang and Hatton, 2012), and
complex coacervation (Piacentini et al., submitted).
tABLe 11.3
typical Mixing ratios of raw Materials in the Production
of sPG from na2o–Cao–Al2o3 –B2o3 –sio2 Mother Glass
wt%
Shirasu 51
Limestone 23
Boric acid 22
Soda ash 4
tABLe 11.4
Composition of Primary Glass,a sPG,a and Porous Vycor
Glass and Pyrex Glass
Primary Glass sPG Vycor® Glass Pyrex® Glass
for sPG (wt%) (wt%) (wt%) (wt%)
SiO2 49 69 94−99.5 81
Al2O2 10 13 0−0.5 2
CaO 17 2 – –
B2O3 16 7 0.2−6.0 13
Na2O 5 5 < 0.1 4
K2O 2 4 – –
Fe2O3 1 0.4 – –
Sources: Nakashima, T. et al., J. Ceram. Soc. Jpn. Int. Ed., 100, 1389−1393,
1992; Nakashima, T., 38th International SPG Forum on Membrane
and Particle Science and Technology in Food and Medical Care,
Sadowara, Japan, 2002.
a Based on proportions of raw materials given in Table 11.3.
Cooling to 760°C–750°C
SPG membrane
FIGure 11.4 A flow diagram of different steps involved in the fabrication of Shirasu
porous glass (SPG) membrane.
1
T1
UCST Binodal Interconnected
-type pores
Temperature
Spinodal
2
T2 Droplet
-type pores
Xs X1 Xi
Molar fraction of acid insoluble phase
FIGure 11.5 Spinodal decomposition of glass induced by cooling mother glass from an initial
temperature T1 to temperature T2 lying in the spinodal region (within the spinodal line). To pre-
vent phase separation via nucleation, a transition from the stable to the spinodal region of the
phase diagram must proceed quickly or through the upper critical solution temperature (UCST).
264 Engineering Aspects of Food Emulsification and Homogenization
V 1/ 2 − Ea
d p = 4 K 1/ 2 p t exp (2 RT ) (11.1)
mm 2
where:
K is a constant depending on the composition of mother glass
Ea is the activation energy for diffusion during phase separation [400−600 kJ
mol–1 according to Nakashima (2002) and Kukizaki (2010)]
R is the universal gas constant
Vp/mm is the total pore volume per unit mass of dry membrane
Therefore, the mean pore diameter of SPG membrane is proportional to the square
root of the heating time at any constant temperature, whereas a logarithm of the
mean pore diameter is inversely proportional to 1/T2 at constant heating time.
N 0.56
= 2 (11.2)
Am dp
where:
N/Am and dp are in m–2 and m, respectively
32ξ2δm
Rm,sym = (11.3)
d p2ε
Formation and Modification of Dispersions 265
tABLe 11.5
Properties of Commercial Isotropic (symmetric) sPG Membrane
shape tubes or Flat discs
Thickness, δm 0.4−1 mm
Compressive strength 200−280 Mpa
Pore diameter, dp 0.04−20 μm
Porosity, ε 50%−60%
True density 2000−2500 kg m–3
Zeta potential at pH = 3−10 and CNaCl = 1−100 mol m–3 −15−(−45) mV
Pore tortuosity, ξ 1.25−1.4
Number of pores per unit cross-sectional area, N/Am 109−1014 m–2
Specific pore volume, Vp/mm 0.5−0.6 dm3 kg–1
Hydraulic resistance, Rm,sym 108−1012 m–1
Sources: Kukizaki, M., Sep. Sci. Technol., 69, 87−96, 2009; Nakashima, T., 38th International
SPG Forum on Membrane and Particle Science and Technology in Food and Medical
Care, Sadowara, Japan, 2002; Nakashima, T. et al., J. Ceram. Soc. Jpn. Int. Ed., 100,
1389−1393, 1992; Vladisavljević, G.T. et al., J. Membr. Sci., 250, 69−77, 2005.
200 μm
(a) (b)
FIGure 11.6 (a) Scanning electron micrograph of the surface of SPG membrane polished
with diamond paste and used for visualization of membrane emulsification by metallographic
microscope; (b) X-ray microtomography image of SPG membrane. (Data from Vladisavljević,
G.T. et al., J. Membr. Sci., 302, 243−253, 2007.)
where:
δm is the membrane thickness
ε is the membrane porosity
The hydraulic resistance of isotropic SPG membrane is relatively high (Table 11.5),
due to its substantial thickness of 400−1000 μm, but can be reduced if the membrane
is fabricated with anisotropic structure (Kukizaki and Goto, 2007b). Assuming that
the pore tortuosity and porosity, ξ and ε, are independent of the pore size, the hydrau-
lic resistance of anisotropic SPG membrane is given by (Kukizaki and Goto, 2007b):
266 Engineering Aspects of Food Emulsification and Homogenization
32ξ2δskin 32ξ2δsup
Rm,asym = + 2 (11.4)
2
d p,skin ε d p,supε
where:
δskin and δsup are the thicknesses of the skin and support layer, respectively
dp,skin and dp,sup their mean pore diameters
According to Kukizaki and Goto (2007b), the thickness of the skin layer is 6% of
the overall membrane thickness and the ratio of the pore diameters in the skin and
support layer is around 7, so it can be written as:
ξ2 δ δ 0.06 0.94
Rm,asym = 32 2skin + 2sup = Rm,sym + 2 ≈ 0.08 Rm,sym
ε d p,skin d p,sup 1 7
−OH −OSi(CH3)2R
Cl−Si(CH3)2R
SiO2 −OH SiO2 −OSi(CH3)2R
TMS: R = CH3
ODS: R = C18H37
−OH −OSi(CH3)2R
(a)
RO
−OH −O
RO Si NH2
RO +
SiO2 −OH + H2O SiO2 −O Si NH3
APTMS: R = CH3
−
−OH APTES: R = C2H5 −O + OH
(b)
FIGure 11.7 Chemical modification of SPG surface by treatment with organosilane com-
pounds: (a) hydrophobic treatment with monochlorosilanes (TMS—trimethylchlorosilane,
ODS—octadecyldimethylchlorosilane) (Data from Kai, T. et al., J. Polym. Sci., Part A-1:
Polym. Chem., 44, 846−856, 2006.); (b) introduction of amino groups by amino trialkoxysi-
lanes to render the surface positively charged [APTMS—(3-aminopropyl)-trimethoxysilane,
APTES—(3-aminopropyl)-triethoxysilane].
and hydraulic resistance of SPG membrane can be modified over a wide range by
incorporating dextran macromolecules within the pores (Kawakita et al., 2009; Seto
et al., 2011). Dextran can be synthesized by in situ enzymatic reaction between dex-
transucrase immobilized within the pores and sucrose from an aqueous solution that
is passed through the membrane. A reversible change in the hydraulic resistance of
the modified SPG membrane is a consequence of reversible extension and shrinkage
of solvent-responsive dextran chains inside the pores.
The surface of untreated SPG surface has a negative zeta potential between
–15 and –45 mV within a pH range of 2−8 due to dissociation of silanol groups
(≡Si-OH Δ ≡SiO− + H+) (Kukizaki, 2009b). A positive charge on the membrane sur-
face can be induced by treating the membrane with amino trialkoxysilanes, such as
(3-aminopropyl)-trimethoxysilane (APTMS) and (3-aminopropyl)-triethoxysilane
(APTES) (Figure 11.7b). Amino trialkoxysilanes undergo hydrolysis in aqueous
solution resulting in the formation of silanol groups, which can be condensed with
a silanol group on the SPG surface to form stable siloxane bond (Si–O–Si) required
for surface modification.
30 μm
(a) (b)
FIGure 11.8 (a) A micrograph of droplets formed on the surface of SPG membrane in
DME (Data from Vladisavljević, G.T. et al., J. Membr. Sci., 302, 243−253, 2007.); (b) micro-
graphs of droplets before PME and after passing five times through 8-μm SPG membrane.
(Data from Vladisavljević and McClements, 2010.)
Various SPG membrane devices have been used in DME: (1) cross-flow module with
tubular SPG membrane with an effective length of up to 500 mm, (2) a short SPG mem-
brane tube with an effective length of 7−15 mm in a stirred vessel (internal or external
pressure micro kit), and (3) rotating SPG membrane tube in a stagnant continuous phase.
In the cross-flow DME system, a continuous phase liquid circulates from a storage tank
through the bore of the membrane tube and back to the tank (Figure 11.9). A dispersed
phase-forming liquid stored in a pressure vessel is fed to the outside of the membrane
tube and force to penetrate through the membrane under the pressure difference which
is 1.1–5 times higher than the capillary pressure (Vladisavljević and Schubert, 2003a).
The apparatus is operated continuously until a desired dispersed phase concentration is
achieved in the emulsion. A transmembrane flux in cross-flow DME should be kept
below 1−30 l m−2 h−1 to obtain uniform droplets with a relative span factor of droplet size
distribution of 0.25−0.45. To increase transmembrane flux by 2 orders of magnitude,
the continuous phase can be introduced into SPG membrane tube radially, as shown in
Figure 11.10. A tangential introduction of the continuous phase generates spiral stream-
lines in the axial direction (swirl flow) that exert a strong centrifugal force onto the inner
surface of the membrane helping to sweep away droplets from the membrane surface
(Shimoda et al., 2011). At the swirl-flow velocity of 0.85−5.4 m s−1 and the transmem-
brane flux of 0.3−3 m3 m−2 h−1, a relative span factor of droplet size distribution of
0.45−0.64 was achieved with an oil/water phase volume ratio in single-pass operation
of up to 0.4 (Shimoda et al., 2011). Insertion of static turbulence promoters is an alter-
native method of increasing shear at the membrane surface in cross-flow DME, while
maintaining a low shear in the recirculation loop (Koris et al., 2011).
Cross-flow systems are easy to scale up and offer a constant shear stress along
the membrane surface. However, at least several hundred millilitres of the continu-
ous phase is required in the system to provide recirculation. The SPG test kit shown
in Figure 11.11a requires a much smaller amount of continuous phase (<50 ml) and
can be operated with a very low hold-up volume of both phases, which is useful
for expensive samples such as medical preparations (Higashi and Setoguchi, 2000).
Formation and Modification of Dispersions 269
Vent
1 Membrane
module
Continuous
phase
Compressed Dispersed
gas phase
FIGure 11.9 An apparatus for cross-flow DME using tubular SPG membrane. During
initial start-up, a valve 1 is open to remove any trapped air from the module. (Data from
Nakashima, T. et al., Monodisperse single and double emulsions and method of producing
same, US Patent 5,326,484, 1994.)
The continuous phase is kept under agitation by a magnetic stir bar, while the
dispersed phase is injected through the membrane tube from outside to inside. The
membrane tube serves as a draft tube, which results in more effective circulation of
the continuous phase than in an internal pressure SPG kit.
In addition to DME with static SPG membrane, where shear stress is controlled
by fluid flow over the membrane surface, dynamic SPG membrane systems have
been investigated, where shear is controlled by rotating the membrane within a static
continuous phase (Pawlik and Norton, 2012, 2013). Rotating membrane systems can
be operated batchwise or continuously. In a continuous flow operation, surface shear
is decoupled from the cross-flow velocity, which means that sufficient shear on the
membrane surface can be achieved no matter how small the flow rate of the continu-
ous phase may be. Therefore, emulsions with a high dispersed phase concentration
can be produced without emulsion recycling that can help prevent damage to shear
sensitive components and secondary breakup of the drops formed by the membrane.
270 Engineering Aspects of Food Emulsification and Homogenization
Streamline of the
continuous phase
Tubular
SPG membrane
Continuous
phase
(a)
Continuous Dispersed
phase phase
Emulsion
(b)
FIGure 11.10 Introduction of continuous phase in cross-flow module from the tangential
direction to improve a dispersed phase flux through the membrane (a) cross-sectional view of the
module in the tangential direction, that is, perpendicular to the direction of cross-flow; (b) cross-
sectional view of the module in the axial direction, that is, parallel to the direction of cross-flow.
(Data from Shimoda et al., 2011.)
An SPG membrane rig used for PME is shown in Figure 11.11b. A pressurized
premix from a pressure vessel is passed through the membrane tube from outside
to inside under the driving force of pressure difference ranging from several bars
(for a 10-μm membrane) to more than 10 bar (for 1-μm membrane) and up to 50
bar for the membrane with submicron pore sizes. The product emulsion is collected
inside the membrane tube and discharged from the bottom of the tube. In order to
reduce the droplet size additionally and improve droplet size uniformity, product
emulsion is passed repeatedly through the membrane (Vladisavljević, Shimizu et al.,
2004, 2006; Vladisavljević, Surh et al., 2006). Membrane homogenization using
repeated cycles was first developed by Olson et al. (1979) and used for homogeniza-
tion of lipid vesicles using track-etch polycarbonate filters.
Dispersed
phase Vent
Continuous
phase
(a)
Premix
Emulsion
(b)
FIGure 11.11 External pressure type micro kits available by SPG Technology Co., Ltd.
(Sadowara, Japan) for (a) DME and (b) PME. The kits are supplied with SPG membrane tube
with an effective length of 10−15 mm.
and continuous phases, nature of the surfactant used and the surfactant concentra-
tion, and emulsion formulation (Joscelyne and Trägårdh, 2000).
Oil
Water phase phase
Oil Water
phase phase
θ θ
γmw γmo γmo γmw
γwo γwo
FIGure 11.12 Typical contact angles through the water phase and phase pressures encoun-
tered in membrane emulsification: (a) production of O/W emulsion (θ < 90°, Po > Pw); (b) pro-
duction of W/O emulsion (θ > 90°, Po < Pw). The contact angle θ is the angle measured through
the water phase, where a liquid–liquid interface meets a membrane surface (γmw = interfacial
tension between the membrane and water phase, γmo = interfacial tension between the mem-
brane and oil phase, γwo = interfacial tension between the water and oil phase).
where:
γwo is the equilibrium interfacial tension between the water and oil phase
θ is the contact angle; that is, the angle formed by a water phase at the three-
phase boundary where the water phase, oil phase, and membrane intersect
(Figure 11.12)
A hydrophilic membrane (θ < 90°) is used in the production of O/W emulsion; thus
Pcap > 0 and Po > Pw. A hydrophobic membrane (θ > 90°) is used in the production
of W/O emulsion, and thus Pcap < 0 and Po < Pw; that is, the water phase pressure
should be higher than the oil phase pressure by Pcap to drive the water phase through
the membrane. The droplet-generation regime is determined by capillary number
given by Ca = Udηd /γwo, where Ud is the velocity of the dispersed phase in a pore
and μd is the viscosity of the dispersed phase. For low capillary numbers in the pores
(Ca < Cacr), droplets are formed in the dripping regime. In this regime, the interfa-
cial tension force dominates the viscous force (Sugiura et al., 2002), and the droplet
size is virtually independent on the transmembrane flux (Figure 11.13). For high
capillary numbers (Ca > Cacr), droplets grow to large sizes (dd > 10dp) before being
detached from the membrane surface, which is termed as continuous outflow regime
(Kobayashi et al., 2003). In this regime, the viscous force dominates the interfacial
tension force and the droplet size sharply increases with increasing the dispersed
phase velocity. The critical flux, Jcr, that is, the transmembrane flux at which the
transition from dripping to continuous outflow regime occurs, is independent of
the pore size and increases with decreasing the viscosity of the dispersed phase.
Emulsions produced in the continuous outflow regime are highly polydisperse due
Formation and Modification of Dispersions 273
Jcr
Transmembrane flux, J
FIGure 11.13 Mean droplet size in DME as a function of transmembrane flux, J. Dripping
regime is characterized by the formation of small droplets at high frequency and occurs at
J < Jcr. Continuous outflow regime is characterized by the formation of large droplets at low
frequency and occurs at J > Jcr.
to the random nature of droplet formation process. In addition, flow transition from
dripping to continuous outflow does not occur simultaneously for all pores, leading
to the large variations in the droplet size for the pores operating in the dripping and
continuous outflow regime.
Fca = πd p γ (11.6)
274 Engineering Aspects of Food Emulsification and Homogenization
Pore size, dp
Droplet size, dd
FIGure 11.14 Mean droplet size in DME as a function of mean pore size of SPG membrane
and shear stress on the membrane surface.
2
d (11.7)
Fd = 9πσw dd d − rp2
2
where:
rp is the pore radius
σw is the shear stress on the membrane surface
The equation for the droplet diameter can be obtained by solving Equations 11.6 and
11.7 for dd:
For laminar flow inside the membrane tube (Re < 2300): λ = 64/Re and σw = 8ηcUc/dmi,
where dmi is the inner diameter of the membrane tube and ηc is the viscosity of continu-
ous phase. For the rotating SPG membrane, σw can be estimated from (Vladisavljević
and Williams, 2006):
2ηc rmo
2
ω
σw = (11.10)
rb − rmo
2 2
where:
ω is the angular velocity of the membrane
rmo is the outer radius of the membrane tube
r b is the inner radius of the cylinder in which the membrane tube is rotating
Pore size, dp
DME
Droplet size, dd
PME
Transmembrane flux, J
FIGure 11.15 Mean droplet size in PME as a function of mean pore size of SPG membrane
(at σw,p = const) and transmembrane flux (at dp = const). For comparison, a relationship between
mean droplet size and mean pore size in DME is shown by the dashed line.
shear stress on the pore walls of 200 Pa (Vladisavljević, Shimizu et al., 2006). The
critical pressure in PME is given by (Park et al., 2001):
If d1 /dp >> 1, the capillary pressure is given by Equation 11.5. In PME, the transmem-
brane pressure resulting in the most uniform droplets is typically 10−50 times larger
than Pcap (Vladisavljević, Shimizu et al., 2004). The mean droplet size decreases
with increasing the mean shear stress on the pore walls, given by:
8ηe J ξ
σw,p = (11.12)
εdp
where:
ηe is the viscosity of emulsion inside the pores
According to Equation 11.12, the mean droplet size decreases with increasing trans-
membrane pressure, as shown in Figures 11.15 and 11.17b. The pressure energy is
used for flow through the membrane pores and droplet disruption:
Formation and Modification of Dispersions 277
1 1
∆ptm = ηe (Rm + Rf )J + Cφ − γ
d1 d p
(11.13)
∆p
flow
∆pdisr
where:
C is a constant
ϕ is the volume fraction of the dispersed phase
Rm is the membrane resistance
Rf is the fouling resistance
The second term in Equation 11.13 is based on the assumption that the energy needed
for droplet disruption is proportional to the resultant increase in surface area. The
fouling resistance occurs as a result of accumulation of the dispersed phase on the
membrane surface (external fouling) and inside the pores (internal fouling). External
fouling dominates at high dd/dp ratios in the feed emulsion and low transmembrane
pressures, whereas internal fouling dominates at high transmembrane pressures
and small droplet sizes relative to the pore size. In repeated PME (Vladisavljević,
Shimizu et al., 2004):
1 1
∆ ptm = ηe (Rm + Rf, i )Ji + Cφ − γ
di di −1 (11.14)
∆p
flow ∆pdisr
where:
Ji and Rf,i are J and Rf during ith pass through the membrane
di and di−1 are the mean droplet diameters after ith and (i−1)th pass, respectively
The effect of varying droplet size on the viscosity of emulsion was disregarded
in Equation 11.14. As the number of passes through the membrane increases at
Δptm = const, the mean droplet size tends to a constant minimum value, that is, di → di−1
(Figure 11.17b), which means that Δpdisr → 0 and Δpflow → Δptm. Therefore, the term
accounting for droplet disruption (Δpdisr) becomes progressively less important than
the flow term (Δpflow), and pressure energy of the feed mixture is increasingly used for
providing emulsion flow through the membrane (Figure 11.16). As a consequence of
redistribution of pressure terms in Equation 11.14, the transmembrane flux at constant
operating pressure increases after each pass through the membrane until a maximum
flux is established. The maximum transmembrane flux in PME is limited by the mem-
brane resistance, emulsion viscosity, and transmembrane pressure (Figure 11.17a).
The effect of continuous phase viscosity, dispersed phase concentration, trans-
membrane pressure on the mean droplet size, and transmembrane flux in repeated
PME is shown in Figure 11.17. The largest increase in flux between the two passes was
observed between the first and second pass, because the most significant reduction in
the mean droplet size was observed in the first pass. Under the same conditions, the
limiting flux was substantially lower at the higher dispersed phase content, which was
a consequence of both the higher viscosity of emulsion, ηe, and the higher Δpdisr term
278 Engineering Aspects of Food Emulsification and Homogenization
150
140 Δpflow
130
Pressure difference, Δpflow or Δpdisr/kPa
20
10 Δpdisr
0
1 2 3 4 5
Number of emulsification cycles, n
FIGure 11.16 The pressure difference used to overcome the hydraulic resistances in the
system and interfacial tension force as a function of the number of passes through the mem-
brane at two different transmembrane pressures. Production of W/O/W emulsion using PME
at the viscosity of the continuous phase of 126 mPa⋅s, the concentration of W/O drops in
W/O/W emulsion of 10 vol%, the concentration of inner water phase in the W/O emulsion of
10 vol%, and the mean pore size of the membrane of 10.7 μm. (Data from Vladisavljević, G.T.
et al., Colloids Surf. A, 232, 199−207, 2004.)
240
200
160
120
Transmembrane flux, J/m3 m−2 h−1
80
20
18
16
ηc = 1 mPas, φo = 1 vol%, Δptm = 150 kPa
14 ηc = 1 mPas, φo = 20 vol%, Δptm = 150 kPa
4
1 2 3 4 5
(a) Number of passes through membrane, n
3
2
Span
0.5
0.4
0.3
11 dp = 10.7 μm
Mean size of outer drops d50/μm
10
1 2 3 4 5
(b) Number of passes through membrane, n
FIGure 11.17 The effect of the number of passes through the membrane on: (a) transmembrane
flux and (b) median diameter and relative span factor of W/O drops. Production of W/O/W
emulsion using PME at different transmembrane pressures (100 or 150 kPa), viscosities of the
continuous phase (1 or 126 mPa s), and concentrations of W/O drops in W/O/W emulsion (1 or
20 vol%). The mean pore size of the SPG membrane was 10.7 μm and the concentration of inner
water phase in the W/O emulsion was 30 vol%. (Data from Vladisavljević, G.T. et al., J. Membr.
Sci., 284, 373−383, 2006.)
280 Engineering Aspects of Food Emulsification and Homogenization
(Surh et al., 2007); nano and microemulsions (Choi et al., 2012; Koga et al., 2010;
Laouini et al., 2012; Oh et al., 2011, 2013; Pradhan et al., 2013); emulsions with
droplets laminated with multilayered biopolymer films (Gudipati et al., 2010; Nazir
et al., 2012; Vladisavljević and McClements, 2010); microbubbles (Kukizaki and
Goto, 2007a); nanobubbles (Kukizaki and Goto, 2006); micro and nanoparticles
(Vladisavljević and Williams, 2005, 2010); and vesicles (liposomes and niosomes)
(Hwang et al., 2011; Pham et al., 2012).
Some examples of particles fabricated by DME or PME using SPG membrane are
given in Table 11.6 and Figure 11.18. Emulsion droplets were transformed into solid
particles by implementing a variety of chemical reactions or physicochemical processes
within the droplets, such as cross-linking of hydrogel-forming polymers (Wei et al.,
2013), polymerization of monomer mixtures (Omi et al., 1995), solidification from a
melt (Kukizaki and Goto, 2007c), polymer precipitation induced by solvent evaporation
or extraction (Liu, Ma, Meng et al., 2005), redox reaction (Kakazu et al., 2010), com-
plex coacervation (Kage et al., 1997), and thermal coagulation (El-Mahdy et al., 1998).
tABLe 11.6
examples of Microparticles Fabricated using dMe and PMe with sPG
Membrane
secondary reaction/Process
Product type example after dMe or PMe references
Ceramic particles Silica nano or Polymerization of silicic acids Kandori et al. (1992)
microparticles by interfacial or internal
reaction
Liquid crystal Thermochromic Melt crystallization in O/W Segura et al. (2013)
particles liquid crystal emulsion
particles
Carbon particles Carbon cryogel Sol–gel polycondensation Yamamoto et al.
followed by freeze-drying and (2010)
carbonization
Metal particles Solder metal Solidification of liquid metal in Torigoe et al. (2011)
microparticles M/W or M/O emulsion
Silver nanoparticles Reduction of silver ions in W/O Kakazu et al. (2010)
microemulsions
Solid lipid W/S microcarrier Melt crystallization in W/O/W Kukizaki and Goto
particles emulsion (2007c)
S/S microcarrier Melt crystallization in S/O/W Kukizaki (2009c)
emulsion
Coherent particles Melt crystallization in O/W D’oria et al. (2009);
emulsion Li et al. (2011)
Gel micro and Ca alginate Cross-linking of sodium Liu et al. (2003);
nanoparticles alginate with Ca2+ in W/O You et al. (2001);
emulsion Akamatsu et al.
(2011)
(Continued)
Formation and Modification of Dispersions 281
(Continued)
282 Engineering Aspects of Food Emulsification and Homogenization
(Continued)
Formation and Modification of Dispersions 283
AA, acrylic acid; CAP, chloramphenicol; CMS, chloromethylstyrene; DFB, decafluorobutane; DMAEMA,
dimethylaminoethyl methacrylate; DVB, divinylbenzene; EDMA, ethylene dimethacrylate; ENB
5-ethylidene-2-norbornene; HTCC, N-[(2-hydroxy-3-trimethylammonium) propyl] chitosan chloride; GP,
glycerophosphate; MAA, methacrylic acid; MMA, methyl methacrylate; mPEG, poly(monomethoxypoly
ethylene glycol); NIPAM, N-isopropylacrylamide; PAAm, PAAm: polyacrylamide; PE2CA, poly(ethyl
2-cyanoacrylate); PLA, polylactic acid or polylactide; PLGA, poly(lactic-co-glycolic acid); P(M-co-U-co-F),
Poly(melamine-co-urea-co-formaldehyde); PNaAMPS, poly[sodium 2-(acrylamido)-2-methylpropane-
sulfonate]; PPC, poly(propylene carbonate); P(S-BIEM), poly[styrene-2-(2-bromoisobutyryloxy)ethyl
methacrylate]; PUU, polyurethaneurea; St, styrene; TPP, tripolyphosphate; VP, vinyl polymer.
284 Engineering Aspects of Food Emulsification and Homogenization
40°C
SiO2
25°C
PDMAEMA
PMMA
(a) (b) (c) (d)
Porous
Silica nanoparticles SiO2 shell
− −
−
− −
− + + ++ −
− PLGA −
+ + ++
− −
− −
(e) − − (f) (g) (h)
Hydrophilic
PS-HEMA Fe3O4 drug
Hydrophilic
drug nanoparticle
solution
Hydrophobic
Solid
surfactant
lipid
(i) (j) (k) (l)
The mean size of nanobubbles was 8.6 times larger than the mean pore size and
unaffected by the flow velocity of air in the pores within a range of 0.5–3.7 m s−1
(Kukizaki and Goto, 2006). Microbubbles generated by SPG membranes can find
applications in the production of aerated food products (Zúñiga and Aguilera,
2008), ultrasound contrast agents for ultrasonic examinations (Hou et al., 2009),
and aerobic wastewater treatment (Liu, Tanaka, Zhang et al., 2012, 2013), which
can be combined with UV irradiation (Tasaki et al., 2009) or activated sludge pro-
cess (Liu, Tanaka, Ma et al., 2012).
11.5 ConCLusIons
SPG membranes are increasingly being used in microfluidic applications aiming at
generating uniform micro and nanodroplets, nanobubbles, and nanoparticles. They
have also been used for modification of emulsions (phase inversion, demulsification,
and homogenization) as well as in micromixing/direct nanoprecipitation processes
for production of inorganic and organic nanoparticles. SPG membranes can over-
come low throughput limitations of conventional microfluidic junctions and flow
focusing devices by providing countless number of pores that serve as massively
parallel T-junctions. Direct and premix membrane emulsification (DME and PME)
are two main modes of operation of SPG membrane emulsification devices. In DME,
the mean droplet size is proportional to the mean pore size and the proportionality
constant is typically around 3, whereas in PME, the ratio of the mean droplet size
to the mean pore size is less than 1.5 and can be below unity. To form uniformly
sized particles, DME or PME can be combined with a variety of physicochemical
or chemical processes that can be applied individually or in combination, such as
polymerization, cross-linking, solvent evaporation, electrostatic deposition, internal
phase separation, coagulation, calcination, sol-gel chemistry, crystallization, etc.
reFerenCes
Ahmad, H., Saito, N., Kagawa, Y. and Okubo, M. (2008) “Preparation of micrometer-sized,
monodisperse ‘Janus’ composite polymer particles having temperature-sensitive poly-
mer brushes at half of the surface by seeded atom transfer radical polymerization,”
Langmuir, 24: 688−691.
Akamatsu, K., Chen, W., Suzuki, Y., Ito, T., Nakao, A., Sugawara, T., Kikuchi, R. and Nakao, S.
(2010) “Preparation of monodisperse chitosan microcapsules with hollow structures
using the SPG membrane emulsification technique,” Langmuir, 26: 14854−14860.
Akamatsu, K., Ikeuchi, Y., Nakao, A. and Nakao, S. (2012) “Size-controlled and monodis-
perse enzyme-encapsulated chitosan microspheres developed by the SPG membrane
emulsification technique,” J. Colloid Interface Sci., 363: 707−710.
Akamatsu, K., Maruyama, K., Chen, W., Nakao, A. and Nakao, S. (2011) “Drastic difference
in porous structure of calcium alginate microspheres prepared with fresh or hydrolyzed
sodium alginate,” J. Colloid Interface Sci., 371: 46−51.
Anna, S.L., Bontoux, N. and Stone, H.A. (2003) “Formation of dispersions using ‘flow focus-
ing’ in microchannels,” Appl. Phys. Lett., 82: 364−366.
Chang, E.P. and Hatton, T.A. (2012) “Membrane emulsification and solvent pervaporation
processes for the continuous synthesis of functional magnetic and Janus nanobeads,”
Langmuir, 28: 9748−9758.
Formation and Modification of Dispersions 287
Chen, G.G., Luo, G.S., Sun, Y., Xu, J.H. and Wang, J.D. (2004) “A ceramic microfiltration
tube membrane dispersion extractor,” AIChE J., 50: 382−387.
Chen, G.G., Luo, G.S., Xu, J.H. and Wang, J.D. (2004) “Membrane dispersion precipitation
method to prepare nanopartials,” Powder Technol., 139: 180−185.
Chen, G., Luo, G., Yang, X., Sun, Y. and Wang, J. (2004) “Anatase-TiO2 nano-particle prepa-
ration with a micro-mixing technique and its photocatalytic performance,” Mater. Sci.
Eng. A, 380: 320−325.
Cheng, C.J., Chu, L.Y., Xie, R. and Wang, X.W. (2008) “Preparation of highly monodisperse
W/O emulsions with hydrophobically modified SPG membranes,” Chem. Eng. Technol.,
31: 377−383.
Cho, Y.H., Lee, J.J., Park, I.B., Huh, C.S., Baek, Y.J. and Park, J. (2005) “Protective effect of
microencapsulation consisting of multiple emulsification and heat gelation processes on
immunoglobulin in yolk,” J. Food Sci., 70: E148−E151.
Choi, Y.K., Poudel, B.K., Marasini, N., Yang, K.Y., Kim, J.W., Kim, J.O., Choi, H.G. and Yong,
C.S. (2012) “Enhanced solubility and oral bioavailability of itraconazole by combining
membrane emulsification and spray drying technique,” Int. J. Pharm., 434: 264−271.
Chu, L.Y., Park, S.H., Yamaguchi, T. and Nakao S. (2002) “Preparation of micron-sized mono-
dispersed thermoresponsive core-shell microcapsules,” Langmuir, 18: 1856−1864.
Chu, L.Y., Rui, X., Zhu, J.H., Chen, W.M., Yamaguchi, T. and Nakao, S. (2003) “Study of
SPG membrane emulsification processes for the preparation of monodisperse core-shell
microcapsules,” J. Colloid Interface Sci., 265: 187−196.
Doan, T.V.P., Couet, W. and Olivier, J.C. (2011) “Formulation and in vitro characterization
of inhalable rifampicin-loaded PLGA microspheres for sustained lung delivery,” Int.
J. Pharm., 414: 112–117.
D’oria, C., Charcosset, C., Barresi, A.A. and Fessi, H. (2009) “Preparation of solid lipid par-
ticles by membrane emulsification—Influence of process parameters,” Colloids Surf. A,
338: 114−118.
Dragosavac, M.M., Holdich, R.G., Vladisavljević, G.T. and Sovilj, M.N. (2012) “Stirred cell
membrane emulsification for multiple emulsions containing unrefined pumpkin seed oil
with uniform droplet size,” J. Membr. Sci., 392−393: 122−129.
El-Mahdy, M., Ibrahim, E.S., Safwat, S., el-Sayed, A., Ohshima, H., Makino, K., Muramatsu,
N. and Kondo, T. (1998) “Effects of preparation conditions on the monodispersity of
albumin microspheres,” J. Microencapsulation, 15: 661−673.
van der Graaf, S., Schroën, C.G.P.H., Van der Sman, R.G.M. and Boom, R.M. (2004) “Influence
of dynamic interfacial tension on droplet formation during membrane emulsification,”
J. Colloid Interface Sci., 277: 456−463.
Gudipati, V., Sandra, S., McClements, D.J. and Decker, E.A. (2010) “Oxidative stability and
in vitro digestibility of fish oil-in-water emulsions containing multilayered membranes,”
J. Agric. Food Chem., 58: 8093–8099.
Hao, D.X., Gong, F.L., Hu, G.H., Lei, J.D., Ma, G.H. and Su, Z.G. (2009) “The relation-
ship between heterogeneous structures and phase separation in synthesis of uniform
PolyDVB microspheres,” Polymer, 50: 3188−3195.
Higashi, S. and Setoguchi, T. (2000) “Hepatic arterial injection chemotherapy for hepatocel-
lular carcinoma with epirubicin aqueous solution as numerous vesicles in iodinated
poppy-seed oil microdroplets: Clinical application of water-in-oil-in-water emulsion
prepared using a membrane emulsification technique,” Adv. Drug Deliv. Rev., 45:
57−64.
Hou, Z., Lin, C. and Zhang, Q. (2009) “Preparation and characterization of PLA ultrasound
contrast agents by combining an ultrasound method and a Shirasu Porous Glass (SPG)
membrane emulsification technique,” 3rd International Conference on Bioinformatics
and Biomedical Engineering, June 11–13, 2009. Beijing, People’s Republic of China:
ICBBE.
288 Engineering Aspects of Food Emulsification and Homogenization
Hu, J., Hiwatashi, K., Kurokawa, T., Liang, S.M., Wu, Z.L. and Gong, J.P. (2011) “Microgel-
reinforced hydrogel films with high mechanical strength and their visible mesoscale
fracture structure,” Macromolecules, 44: 7775–7781.
Hwang, T., Park, T.J., Koh, W.G., Cheong, I.W., Choi, S.W. and Kim, J.H. (2011) “Fabrication
of nano-scale liposomes containing doxorubicin using Shirasu porous glass membrane,”
Colloids Surf. A, 392: 250−255.
Ito, F., Kanakubo, Y. and Murakami, Y. (2011) “Rapid preparation of monodisperse biodegrad-
able polymer nanospheres using a membrane emulsification technique under low gas
pressure,” J. Polym. Res., 18: 2077–2085.
Ito, F., Uchida, Y. and Murakami, Y. (2010) “Facile technique for preparing organic–inorganic
composite particles: Monodisperse poly(lactide-co-glycolide) (PLGA) particles having
silica nanoparticles on the surface,” Colloids Surf. A, 361: 109−117.
Jaafar-Maalej, C., Charcosset, C. and Fessi, H. (2011) “A new method for liposome prepara-
tion using a membrane contactor,” J. Liposome Res., 21: 213−220.
Jing, C., Chin, C.Y. and Xie, R. (2006) “Preparation of highly monodisperse W/O emulsions
with hydrophobically modified SPG membranes,” J. Colloid Interface Sci., 300: 375−382.
Joscelyne, S.M. and Trägårdh, G. (2000) “Membrane emulsification—A literature review,”
J. Membr. Sci., 169: 107−117.
Kage, H., Kawahara, H., Ogura, H. and Matsuno, Y. (1997) “Microencapsulation of mono-
dispersed droplets by complex coacervation method and membrane thickness of gener-
ated capsules,” Kagaku Kogaku Ronbun., 23: 652−658.
Kai, T., Suma, Y., Ono, S., Yamaguchi, T. and Nakao, S. (2006) “Effect of the pore surface modifi-
cation of an inorganic substrate on the plasma-grafting behavior of pore-filling-type organic/
inorganic composite membranes,” J. Polym. Sci., Part A-1: Polym. Chem., 44: 846−856.
Kakazu, E., Murakami, T., Akamatsu, K., Sugawara, T., Kikuchi, R. and Nakao, S. (2010)
“Preparation of silver nanoparticles using the SPG membrane emulsification technique,”
J. Membr. Sci., 354: 1–5.
Kanakubo, Y., Ito, F. and Murakami, Y. (2010) “Novel one-pot facile technique for preparing
nanoparticles modified with hydrophilic polymers on the surface via block polymer-
assisted emulsification/evaporation process,” Colloids Surf. B, 78: 85−91.
Kandori, K., Kishi, K. and Ishikawa T. (1992) “Preparation of uniform silica hydrogel par-
ticles by SPG filter emulsification method,” Colloids Surf., 62: 259−262.
Karbstein, H. and Schubert H. (1995) “Developments in the continuous mechanical produc-
tion of oil-water macro-emulsions,” Chem. Eng. Process., 34: 205−211.
Kawakita, H., Hamamoto, K., Seto, H., Ohto, K., Harada, H. and Inoue, K. (2009) “Porosity
estimation of a membrane filled with dextran produced by immobilized dextransucrase,”
AIChE J., 55: 275−278.
Kawashima, Y., Hino, T., Takeuchi, H., Niwa, T. and Horibe, K. (1991) “Shear-induced phase
inversion and size control of water/oil/water emulsion droplets with porous membrane,”
J. Colloid Interface Sci., 145: 512−523.
Khayata, N., Abdelwahed, W., Chehna, M.F., Charcosset, C. and Fessi, H. (2012) “Stability
study and lyophilization of vitamin E-loaded nanocapsules prepared by membrane con-
tactor,” Int. J. Pharm., 439: 254−259.
Kobayashi, I., Nakajima, M. and Mukataka, S. (2003) “Preparation characteristics of oil-in-
water emulsions using differently charged surfactants in straight-through microchannel
emulsification,” Colloid. Surf. A, 229: 33−41.
Koga, K., Takarada, N. and Takada, K. (2010) “Nano-sized water-in-oil-in-water emulsion
enhances intestinal absorption of calcein, a high solubility and low permeability com-
pound,” Eur. J. Pharm. Biopharm., 74: 223–232.
Koltuniewicz, A.B., Field, R.W. and Arnot, T.C. (1995) “Cross-flow and dead-end microfiltra-
tion of oily-water emulsion. Part I: Experimental study and analysis of flux decline,”
J. Membr. Sci., 102: 193−207.
Formation and Modification of Dispersions 289
Kong, S.D., Choi, C., Khamwannah, J. and Jin, S. (2013) “Magnetically vectored delivery of cancer
drug using remotely on–off switchable nanocapsules,” IEEE Trans. Magn., 49: 349−352.
Kong, S.D., Zhang, W., Lee, J.H., Brammer, K., Lal, R., Karin, M. and Jin, S. (2010)
“Magnetically vectored nanocapsules for tumor penetration and remotely switchable
on-demand drug release,” Nano Lett., 10: 5088–5092.
Kong, S.D., Zhang, W., Lee, J.H., Choi, C., Khamwannah, J., Karin, M. and Jin, S. (2012)
“Externally triggered on-demand drug release and deep tumor penetration,” J. Vac. Sci.
Technol. B: Microelectron. Nanometer Struct., 30: 02C102−02C102-7.
Koris, A., Piacentini, E., Vatai, G., Bekassy-Molnar, E., Drioli, E. and Giorno, L. (2011)
“Investigation on the effects of a mechanical shear-stress modification method during
cross-flow membrane emulsification,” J. Membr. Sci., 371: 28−36.
Kosvintsev, S.R., Gasparini, G. and Holdich, R.G. (2008) “Membrane emulsification: Droplet
size and uniformity in the absence of surface shear,” J. Membr. Sci., 313: 182−189.
Kosvintsev, S.R., Gasparini, G., Holdich, R.G., Cumming, I.W. and Stillwell, M.T. (2005)
“Liquid-liquid membrane dispersion in a stirred cell with and without controlled shear,”
Ind. Eng. Chem. Res., 44: 9323−9330.
Kou, X., Li, Q., Lei, J., Geng, L., Deng, H., Zhang, G., Ma, G., Su, Z. and Jiang, Q. (2012)
“Preparation of molecularly imprinted nanospheres by premix membrane emulsification
technique,” J. Membr. Sci., 417–418: 87–95.
Kukizaki, M. (2009a) “Shirasu porous glass (SPG) membrane emulsification in the absence
of shear flow at the membrane surface: Influence of surfactant type and concentration,
viscosities of dispersed and continuous phases, and transmembrane pressure,” J. Membr.
Sci., 327: 234−243.
Kukizaki, M. (2009b) “Relation between salt rejection and electrokinetic properties on Shirasu
porous glass (SPG) membranes with nano-order uniform pores,” Sep. Sci. Technol., 69:
87−96.
Kukizaki, M. (2009c) “Preparation of solid lipid microcapsules via solid-in-oil-in-water dis-
persions by premix membrane emulsification,” Chem. Eng. J., 151: 387−396.
Kukizaki, M. (2010) “Large-scale production of alkali-resistant Shirasu porous glass (SPG)
membranes: Influence of ZrO2 addition on crystallization and phase separation in
Na2O─CaO─Al2O3─B2O3─SiO2 glasses; and alkali durability and pore morphology of
the membranes,” J. Membr. Sci., 360: 426−435.
Kukizaki, M. and Goto, M. (2006) “Size control of nanobubbles generated from Shirasu-
porous-glass (SPG) membranes,” J. Membr. Sci., 281: 386−396.
Kukizaki, M. and Goto, M. (2007a) “Spontaneous formation behavior of uniform-sized micro-
bubbles from Shirasu porous glass (SPG) membranes in the absence of water-phase
flow,” Colloids Surf. A, 296: 174−181.
Kukizaki, M. and Goto, M. (2007b) “Preparation and characterization of a new asymmetric
type of Shirasu porous glass (SPG) membrane used for membrane emulsification,”
J. Membr. Sci., 299: 190−199.
Kukizaki, M. and Goto, M. (2007c) “Preparation and evaluation of uniformly sized solid lipid
microcapsules using membrane emulsification,” Colloids Surf. A, 293: 87−94.
Kukizaki, M. and Goto, M. (2008) “Demulsification of water-in-oil emulsions by permeation
through Shirasu-porous-glass (SPG) membranes,” J. Membr. Sci., 322: 196−203.
Kukizaki, M. and Goto, M. (2009) “A comparative study of SPG membrane emulsification in
the presence and absence of continuous-phase flow,” J. Chem. Eng. Jpn., 42: 520−530.
Kukizaki, M. and Nakashima, T. (2004) “Acid leaching process in the preparation of porous
glass membranes from phase-separated glass in the Na2O–CaO–MgO–Al2O3–B2O3–
SiO2 system,” Membrane, 29: 301−308.
Kukizaki, M. and Wada, T. (2008) “Effect of the membrane wettability on the size and size
distribution of microbubbles formed from Shirasu-porous-glass (SPG) membranes,”
Colloids Surf. A, 317: 146−154.
290 Engineering Aspects of Food Emulsification and Homogenization
Laouini, A., Fessi, H. and Charcosset, C. (2012) “Membrane emulsification: A promising alterna-
tive for vitamin E encapsulation within nano-emulsion,” J. Membr. Sci., 423–424: 85−96.
Laouini, A., Jaafar-Maalej, C., Sfar, S., Charcosset, C. and Fessi, H. (2011) “Liposome prepa-
ration using a hollow fiber membrane contactor—Application to spironolactone encap-
sulation,” Int. J. Pharm., 415: 53−61.
Lee, S.H., Baek, H.H., Kim, J.H. and Choi, S.W. (2009) “Core-shell poly(d,l-lactide-co-
glycolide)/poly(ethyl 2-cyanoacrylate) microparticles with doxorubicin to reduce initial
burst release,” Macromol. Res., 17: 1010−1014.
Lee, J., Hwang, D.R. and Shim, S.E. (2010) “Controlling morphology of polymer micro-
spheres by Shirasu porous glass (SPG) membrane emulsification and subsequent polym-
erization: From solid to hollow,” Macromol. Res., 18: 1142−1147.
Li, Y., Fessi, H. and Charcosset, C. (2011) “Preparation of indomethacin-loaded lipid particles
by membrane emulsification,” Adv. Sci. Lett., 4: 591−595.
Li, N. and Sakaki, K. (2008) “Performance of an emulsion enzyme membrane reactor com-
bined with premix membrane emulsification for lipase-catalyzed resolution of enantio-
mers,” J. Membr. Sci., 314: 183−192.
Liu, X.D., Bao, D.C., Xue, W.M., Xiong, Y., Yu, W.T., Yu, X.J., Ma, X.J. and Yuan, Q. (2003)
“Preparation of uniform calcium alginate gel beads by membrane emulsification cou-
pled with internal gelation,” J. Appl. Polym. Sci., 87: 848−852.
Liu, Y., Feng, X.J., Bao, D.C., Li, K.X. and Bao, M. (2010a) “Preparation of microcapsule-
supported palladium catalyst using SPG (Shirasu Porous Glass) emulsification tech-
nique,” Chin. Chem. Lett., 21: 979–982.
Liu, Y., Feng, X.J., Bao, D.C., Li, K. and Bao, M. (2010b) “A new method for the preparation
of microcapsule-supported palladium catalyst for Suzuki coupling reaction,” J. Mol.
Catal. A: Chem., 323: 16−22.
Liu, X., Lee, J.K. and Kessler, M.R. (2011) “Microencapsulation of self-healing agents with
melamine-urea-formaldehyde by the Shirasu porous glass (SPG) emulsification tech-
nique,” Macromol. Res., 19: 1056−1061.
Liu, R., Ma, G.H., Meng, F.T. and Su, Z.G. (2005) “Preparation of uniform-sized PLA micro-
capsules by combining Shirasu Porous Glass membrane emulsification technique and
multiple emulsion-solvent evaporation method,” J. Controlled Release, 103: 31−43.
Liu, R., Ma, G.H., Wan, Y.H. and Su, Z.G. (2005) “Influence of process parameters on the
size distribution of PLA microcapsules prepared by combining membrane emulsifica-
tion technique and double emulsion-solvent evaporation method,” Colloids Surf. B, 45:
144−153.
Liu, C., Tanaka, H., Ma, J., Zhang, L., Zhang, J., Huang, X. and Matsuzawa, Y. (2012) “Effect
of microbubble and its generation process on mixed liquor properties of activated sludge
using Shirasu porous glass (SPG) membrane system,” Water Res., 46: 6051−6058.
Liu, C., Tanaka, H., Zhang, L., Zhang, J., Huang, X., Ma, J. and Matsuzawa, Y. (2012) “Fouling
and structural changes of Shirasu porous glass (SPG) membrane used in aerobic waste-
water treatment process for microbubble aeration,” J. Membr. Sci., 421−422: 225−231.
Liu, C., Tanaka, H., Zhang, J., Zhang, L., Yang, J., Huang, X. and Kubota, N. (2013) “Successful
application of Shirasu porous glass (SPG) membrane system for microbubble aeration
in a biofilm reactor treating synthetic wastewater,” Sep. Purif. Technol., 103: 53–59.
Ma, G.H., An, C.J., Yuyama, H., Su, Z.G. and Omi, S. (2003) “Synthesis and characterization
of polyurethaneurea-vinyl polymer (PUU-VP) uniform hybrid microspheres by SPG
emulsification technique and subsequent suspension polymerization,” J. Appl. Polym.
Sci., 89: 163−178.
Ma, G.H., Chen, A.Y., Su, Z.G. and Omi, S. (2003) “Preparation of uniform hollow poly-
styrene particles with large voids by a glass-membrane emulsification technique and a
subsequent suspension polymerization,” J. Appl. Polym. Sci., 87: 244−251.
Formation and Modification of Dispersions 291
Ma, G.H., Nagai, M. and Omi, S. (1999a) “Study on preparation and morphology of uniform
artificial polystyrene–poly(methyl methacrylate) composite microspheres by employ-
ing the SPG (Shirasu Porous Glass) membrane emulsification technique,” J. Colloid
Interface Sci., 214: 264−282.
Ma, G.H., Nagai, M. and Omi, S. (1999b) “Effect of lauryl alcohol on morphology of uniform
polystyrene-poly(methyl methacrylate) composite microspheres prepared by Porous
glass membrane emulsification technique,” J. Colloid Interface Sci., 219: 110−128.
Ma, G.H., Nagai, M. and Omi, S. (1999c) “Preparation of uniform poly(lactide) microspheres
by employing the Shirasu porous glass (SPG) emulsification technique,” Colloids Surf.
A, 153: 383−394.
Ma, G.H., Nagai, M. and Omi, S. (2001) “Study on preparation of monodispersed poly(styrene-
co-N-dimethylaminoethyl methacrylate) composite microspheres by SPG (Shirasu
porous glass) emulsification technique,” J. Appl. Polym. Sci., 79: 2408−2424.
Ma, G.H., Omi, S., Dimonie, V.L., Sudol, E.D. and El-Aasser, M.S. (2002) “Study of the
preparation and mechanism of formation of hollow monodisperse polystyrene micro-
spheres by SPG (Shirasu Porous Glass) emulsification technique,” J. Appl. Polym. Sci.,
85: 1530−1543.
Ma, G.H., Sone, H. and Omi, S. (2004) “Preparation of uniform-sized polystyrene-polyacrylamide
composite microspheres from a W/O/W emulsion by membrane emulsification technique
and subsequent suspension polymerization,” Macromolecules, 37: 2954−2964.
Mazzei, R., Drioli, E. and Giorno, L. (2010) “Biocatalytic membrane reactor and membrane
emulsification concepts combined in a single unit to assist production and separation of
water unstable reaction products,” J. Membr. Sci., 352: 166−172.
Meng, T., Xie, R., Chen, Y.C., Cheng, C.J., Li, P.F., Ju, X.J. and Chu, L.Y. (2010) “A thermo-responsive
affinity membrane with nano-structured pores and grafted poly(N-isopropylacrylamide) sur-
face layer for hydrophobic adsorption,” J. Membr. Sci., 349: 258−267.
Meng, T., Xie, R., Ju, X.J., Cheng, C.J., Wang, S., Li, P.F., Liang, B. and Chun, L.Y. (2013)
“Nano-structure construction of porous membranes by depositing nanoparticles for
enhanced surface wettability,” J. Membr. Sci., 427: 63–72.
Mulder, M. (1996) Basic Principles of Membrane Technology, Dordrecht, the Netherlands:
Kluwer Academic Publishers.
Muramatsu, N. and Kondo, T. (1995) “An approach to prepare microparticles of uniform size,”
J. Microencapsulation, 2: 129−136.
Muramatsu, N. and Nakauchi, K. (1998) “A novel method to prepare monodisperse micropar-
ticles,” J. Microencapsulation, 15: 715−723.
Nakashima, T. (2002) “Porous glass material and its recent applications,” 38th International
SPG Forum on Membrane and Particle Science and Technology in Food and Medical
Care, November 21–22, Sadowara, Japan.
Nagashima, S., Ando, S., Tsukamoto, T., Ohshima, H. and Makino, K. (1998) “Preparation of
monodisperse poly(acrylamide-co-acrylic acid) hydrogel microspheres by a membrane
emulsification technique and their size-dependent surface properties,” Colloids Surf. B,
11: 47−56.
Nakashima, T. and Kuroki, Y. (1981) “Effect of composition and heat treatment on the phase
separation of NaO─B2O3─SiO2─Al2O3─CaO glass prepared from volcanic ashes,”
Nippon Kagaku Kaishi, 8: 1231−1238.
Nakashima T. and Shimizu M. (1986) “Porous glass from calcium alumino boro-silicate
glass,” Ceramics Japan, 21: 408−412.
Nakashima T., Shimizu M. and Kukizaki, M. (1991) “Membrane emulsification by micropo-
rous glass,” Key Eng. Mater., 61−62: 513−516.
Nakashima, T., Shimizu, M. and Kukizaki M. (1992) “Mechanical strength and thermal resis-
tance of porous glass,” J. Ceram. Soc. Jpn. Int. Ed., 100: 1389−1393.
292 Engineering Aspects of Food Emulsification and Homogenization
Rayner, M., Trägårdh, G. and Trägårdh, C. (2005) “The impact of mass transfer and interfacial
expansion rate on droplet size in membrane emulsification processes,” Colloids Surf. A,
266: 1−17.
Schröder, V., Behrend, O. and Schubert H. (1998) “Effect of dynamic interfacial tension on
the emulsification process using microporous, ceramic membrane,” J. Colloid Interface
Sci., 202: 334−340.
Segura, R., Cierpka, C., Rossi, M., Joseph, S., Bunjes, H. and Kähler, C. (2013) “Non-encapsulated
thermo-liquid crystals for digital particle tracking thermography/velocimetry in microfluid-
ics,” Microfluid. Nanofluid., 14: 445–456.
Seto, H., Ohto, K. and Kawakita, H. (2011) “Reversible extension and shrinkage of solvent-
responsive dextran chains produced by enzymatic reaction,” J. Membr. Sci., 370: 76−81.
Shimoda, M., Miyamae, H., Nishiyama, K., Yuasa, T., Noma, S. and Igura, N. (2011) “Swirl-
flow membrane emulsification for high throughput of dispersed phase flux through
Shirasu porous glass (SPG) membrane,” J. Chem. Eng. Jpn., 44: 1−6.
Si, T., Wang, Y., Wei, W., Lv, P., Ma, G. and Su, Z. (2011) “Effect of acrylic acid weight per-
centage on the pore size in poly(N-Isopropyl acrylamide-co-acrylic acid) microspheres,”
React. Funct. Polym., 71: 728–735.
Sugiura, S., Nakajima, M., Kumazawa, N., Iwamoto, S. and Seki, M. (2002) “Characterization
of spontaneous transformation-based droplet formation during microchannel emulsifi-
cation,” J. Phys. Chem. B, 106: 9405−9409.
Supsakulchai, A., Ma, G.H., Nagai, M. and Omi, S. (2002a) “Microencapsulation of fine tita-
nium dioxide powders from (S/O)/W emulsion with subsequent solvent evaporation,”
ACS Symp. Ser., 801: 260−275.
Supsakulchai, A., Ma, G.H., Nagai, M. and Omi, S. (2002b) “Uniform titanium dioxide (TiO2)
microcapsules prepared by glass membrane emulsification with subsequent solvent
evaporation,” J. Microencapsulation, 19: 425−449.
Surh, J., Jeong, Y.G. and Vladisavljević, G.T. (2008) “On the preparation of lecithin-stabilized
oil-in-water emulsions by multi-stage premix membrane emulsification,” J. Food Eng.,
89: 164−170.
Surh, J., Vladisavljević, G.T., Mun, S. and McClements, D.J. (2007) “Preparation and char-
acterization of water/oil and water/oil/water emulsions containing biopolymer-gelled
water droplets,” J. Agric. Food Chem., 55: 175−184.
Suzuki, K., Fujiki, I. and Hagura, Y. (1999) “Preparation of high concentration of O/W and
W/O emulsions by the membrane phase inversion emulsification using PTFE mem-
branes,” Food Sci. Technol. Int. Tokyo, 5: 234−238.
Suzuki, K., Shuto, I. and Hagura, Y. (1996) “Characteristics of the membrane emulsification
method combined with preliminary emulsification for preparing corn oil-in-water emul-
sions,” Food Sci. Technol. Int. Tokyo, 2: 43−47.
Tanaka, T., Okayama, M., Kitayama, Y., Kagawa, Y. and Okubo, M. (2010) “Preparation of
‘mushroom-like’ janus particles by site-selective surface-initiated atom transfer radical
polymerization in aqueous dispersed systems,” Langmuir, 26: 7843–7847.
Tasaki, T., Wada, T., Fujimoto, K., Kai, S., Ohe, K., Oshima, T., Baba, Y. and Kukizaki, M.
(2009), “Degradation of methyl orange using short-wavelength UV irradiation with
oxygen microbubbles,” J. Hazard. Mater., 162: 1103−1110.
Thorsen, T., Roberts, R.W., Arnold, F.H. and Quake, S.R. (2001) “Dynamic pattern formation
in a vesicle-generating microfluidic device,” Phys. Rev. Lett., 86: 4163−4166.
Torigoe, K., Shimizu, M., Yamamoto, K., Mizozoe, M., Takahashi, H., Suzuki, T. and Murase,
M. (2011) “Method and apparatus for manufacturing low melting point metal fine par-
ticles,” US patent 7,976,608, July 12.
Vladisavljević, G.T., Kobayashi, I. and Nakajima, M. (2012) “Production of uniform droplets
using membrane, microchannel and microfluidic emulsification devices,” Microfluid.
Nanofluid., 13: 151−178.
294 Engineering Aspects of Food Emulsification and Homogenization
Vladisavljević, G.T., Kobayashi, I., Nakajima, M., Williams, R.A., Shimizu, M. and Nakashima,
T. (2007) “Shirasu Porous Glass membrane: Characterisation of microstructure by high
resolution x-ray microtomography and visualisation of droplet formation in real time,”
J. Membr. Sci., 302: 243−253.
Vladisavljević, G.T., Lambrich, U., Nakajima, M. and Schubert, H. (2004) “Production of
O/W emulsions using SPG membranes, ceramic α-Al2O3 membranes, microfluidizer
and a microchannel plate: A comparative study,” Colloids Surf. A, 232: 199−207.
Vladisavljević, G.T. and McClements, D.J. (2010) “Modification of interfacial characteristics of
monodisperse droplets produced using membrane emulsification by surfactant displace-
ment and/or polyelectrolyte electrostatic deposition,” Colloids Surf. A, 364: 123−131.
Vladisavljević, G.T. and Schubert, H. (2002) “Preparation and analysis of oil-in-water emul-
sions with a narrow droplet size distribution using Shirasu-porous-glass (SPG) mem-
branes,” Desalination, 144: 167−172.
Vladisavljević, G.T. and Schubert, H. (2003a) “Preparation of emulsions with a narrow par-
ticle size distribution using microporous α-alumina membranes,” J. Dispersion Sci.
Technol., 24: 811−819.
Vladisavljević, G.T. and Schubert, H. (2003b) “Influence of process parameters on droplet
size distribution in SPG membrane emulsification and stability of prepared emulsion
droplets,” J. Membr. Sci., 225: 15−23.
Vladisavljević, G.T., Shimizu, M. and Nakashima, T. (2004) “Preparation of monodisperse
multiple emulsions at high production rates by multi-stage premix membrane emulsifi-
cation,” J. Membr. Sci., 244: 97−106.
Vladisavljević, G.T., Shimizu, M. and Nakashima, T. (2005) “Permeability of hydrophilic and
hydrophobic Shirasu-porous-glass (SPG) membranes to pure liquids and its microstruc-
ture,” J. Membr. Sci., 250, 69−77.
Vladisavljević, G.T., Shimizu, M. and Nakashima, T. (2006) “Production of multiple emul-
sions for drug delivery systems by repeated SPG membrane homogenization: Influence
of mean pore size, interfacial tension and continuous phase viscosity,” J. Membr. Sci.,
284: 373−383.
Vladisavljević, G.T., Surh, J. and McClements, D.J. (2006) “Effect of emulsifier type on drop-
let disruption in repeated Shirasu porous glass membrane homogenization,” Langmuir,
22: 4526−4533.
Vladisavljević, G.T. and Williams, R.A. (2005) “Recent developments in manufacturing emul-
sions and particulate products using membranes,” Adv. Colloid Interface Sci., 113, 1−20.
Vladisavljević, G.T. and Williams, R.A. (2006) “Manufacture of large uniform droplets using
rotating membrane emulsification,” J. Colloid Interface Sci., 299: 396−402.
Vladisavljević, G.T. and Williams, R.A. (2008) “Recent developments in manufacturing
particulate products from double-emulsion templates using membrane and microflu-
idic devices,” in Aserin, A. (ed.), Multiple Emulsions: Technology and Applications,
Hoboken, NJ: John Wiley & Sons, Inc., pp. 121−164.
Vladisavljević, G.T. and Williams, R.A. (2010) “Recent developments in manufacturing
nanoparticles from emulsion droplets,” in Starov, V. (ed.), Nanoscience: Colloidal and
Interfacial Aspects, Boca Raton, FL: CRC Press, pp. 437–491.
Wang, G., Dou, H. and Sun, K. (2012) “Facile synthesis of hollow polymeric microparticles
possessing various morphologies via seeded polymerization,” Colloid Polym. Sci., 290:
1867−1877.
Wang, G., Leng, Y., Dou, H., Wang, L., Li, W., Wang, X., Sun, K. et al. (2013) “Highly effi-
cient preparation of multiscaled quantum dot barcodes for multiplexed hepatitis B detec-
tion,” ASC Nano, 7: 471−481.
Wang, L.Y., Ma, G.H. and Su, Z.G. (2005) “Preparation of uniform sized chitosan micro-
spheres by membrane emulsification technique and application as a carrier of protein
drug,” J. Controlled Release, 106: 62−75.
Formation and Modification of Dispersions 295
Wang, Y., Qin, J., Yi, W., Li, C. and Ma, G. (2013) “Preparation strategies of thermo-sensitive
P(NIPAM-co-AA) microspheres with narrow size distribution,” Powder Technol., 236:
107−113.
Wang, Y., Zhang, C., Bi, S. and Luo, G. (2010) “Preparation of ZnO nanoparticles using the
direct precipitation method in a membrane dispersion micro-structured reactor,” Powder
Technol., 202: 130−136.
Wei, W., Lv, P.P., Chen, X.M., Yue, Z.G., Fu, Q., Liu, S.Y., Yue, H. and Ma, G.H. (2013)
“Codelivery of mTERT siRNA and paclitaxel by chitosan-based nanoparticles promoted
synergistic tumor suppression,” Biomaterials, 34: 3912–3923.
Wei, W., Ma, G.H., Wang, L.Y., Wu, J. and Su, Z.G. (2010) “Hollow quaternized chitosan
microspheres increase the therapeutic effect of orally administered insulin,” Acta
Biomater., 6: 205–209.
Wei, Y., Wang, Y., Wang, L., Hao, D. and Ma, G.H. (2011) “Fabrication strategy for amphiphi-
lic microcapsules with narrow size distribution by premix membrane emulsification,”
Colloids Surf. B, 87: 399−408.
Wei, Q., Wei, W., Tian, R., Wang, L.Y., Su, Z.G. and Ma, G.H. (2008) “Preparation of uni-
form-sized PELA microspheres with high encapsulation efficiency of antigen by premix
membrane emulsification,” J. Colloid Interface Sci., 323: 267−273.
Wu, J., Wei, W., Wang, L.Y., Su, Z.G. and Ma, G.H. (2008) “Preparation of uniform-sized
pH-sensitive quaternized chitosan microsphere by combining membrane emulsification
technique and thermal-gelation method,” Colloids Surf. B, 63: 164−175.
Xu, J.H., Luo, G.S., Chen, G.G. and Tan, B. (2005) “Mass transfer performance and two-phase
flow characteristic in membrane dispersion mini-extractor,” J. Membr. Sci., 249: 75−81.
Yamamoto, T., Ohmori, T. and Kim, Y.H. (2010) “Synthesis of monodisperse carbon cryo-
gel microspheres using membrane emulsification of a phenol–formaldehyde solution,”
Carbon, 48: 912−928.
Yamashita, N., Konishi, N., Tanaka, T. and Okubo, M. (2012) “Preparation of hemispherical
polymer particles by cleavage of a Janus poly(methyl methacrylate)/polystyrene com-
posite particle,” Langmuir, 28: 12886−12892.
Yang, J., Hao, D.X., Bi, C.X., Su, Z.G., Wang, L.Y. and Ma, G.H. (2010) “Rapid synthesis of
uniform magnetic microspheres by combing premix membrane emulsification and in
situ formation techniques,” Ind. Eng. Chem. Res., 49: 6047–6053.
Yasuda, M., Goda, T., Ogino, H., Glomm, W.R. and Takayanagi, H. (2010) “Preparation of
uniform monomer droplets using packed column and continuous polymerization in tube
reactor,” J. Colloid Interface Sci., 349: 392−410.
Yobas, L., Martens, S., Ong, W.L. and Ranganathan, N. (2006) “High-performance flow-focusing
geometry for spontaneous generation of monodispersed droplets,” Lab Chip, 6: 1073−1079.
You, J.O., Park, S.B., Park, H.Y., Haam, S., Chung, C.H. and Kim, W.S. (2001) “Preparation
of regular sized Ca-alginate microspheres using membrane emulsification method,”
J. Microencapsulation, 18: 521−532.
Yue, H., Wei, W., Fan, B., Yue, Z., Wang, L., Ma, G. and Su, Z. (2012) “The orchestration of
cellular and humoral responses is facilitated by divergent intracellular antigen traffick-
ing in nanoparticle-based therapeutic vaccine,” Pharmacol. Res., 65: 189−197.
Yue, Z.G., Wei, W., Lv, P.P., Yue, H., Wang, L.Y., Su, Z.G. and Ma, G.H. (2011) “Surface
charge affects cellular uptake and intracellular trafficking of chitosan-based nanopar-
ticles,” Biomacromolecules, 12: 2440–2446.
Yuyama, H., Hashimoto, T., Ma, G.H., Nagai, M. and Omi, S. (2000) “Mechanism of suspen-
sion polymerization of uniform monomer droplets prepared by glass membrane (Shirasu
Porous Glass) emulsification technique,” J. Appl. Polym. Sci., 78: 1025−1043.
Yuyama, H., Yamamoto, K., Shirafuji, K., Nagai, M. and Ma, G.H. (2000) “Preparation of
polyurethaneurea (PUU) uniform spheres by SPG membrane emulsification technique,”
J. Appl. Polym. Sci., 77: 2237−2245.
296 Engineering Aspects of Food Emulsification and Homogenization
Zhang, Y., Wei, W., Lv, P., Wang, L. and Ma, G. (2011) “Preparation and evaluation of alginate–
chitosan microspheres for oral delivery of insulin,” Eur. J. Pharm. Biopharm., 77: 11−19.
Zhou, Q.Z., Ma, G.H. and Su, Z.G. (2009) “Effect of membrane parameters on the size and
uniformity in preparing agarose beads by premix membrane emulsification,” J. Membr.
Sci., 326: 694–700.
Zhou, Q.Z., Wang, L.Y., Ma, G.H. and Su, Z.G. (2007) “Preparation of uniform-sized agarose
beads by microporous membrane emulsification technique,” J. Colloid Interface Sci.,
311: 118−127.
Zhou, Q.Z., Wang, L.Y., Ma, G.H. and Su, Z.G. (2008) “Multi-stage premix membrane emul-
sification for preparation of agarose microbeads with uniform size,” J. Membr. Sci., 322:
98–104.
Zhou, Q., Zhang, M.C., Shuang, C.D., Li, Z.Q. and Li, A.M. (2012) “Preparation of a novel
magnetic powder resin for the rapid removal of tetracycline in the aquatic environment,”
Chin. Chem. Lett., 23: 745–748.
Zúñiga, R.N. and Aguilera, J.M. (2008) “Aerated food gels: Fabrication and potential applica-
tions,” Trends Food Sci. Technol., 19: 176−187.
van der Zwan, E.A., Schroën, C.G.P.H. and Boom, R.M. (2008) “Premix membrane emulsifi-
cation by using a packed layer of glass beads,” AIChE J., 54: 2190−2197.
Food and Culinary Science
RAYN E R • DE JME K
Emulsions are found in a wide variety of food products, pharmaceuticals, paints,
and cosmetics, thus emulsification is a truly multidisciplinary phenomenon.
Therefore, understanding of the process must evolve from the combination of (at
Engineering Aspects
least) three different scientific specializations. Engineering Aspects of Food
Emulsification and Homogenization describes the state-of-the-art technology
E N G I N E E R I N G A S P E C T S O F F O O D E M U L S I F I C AT I O N A N D H O M O G E N I Z AT I O N
and brings together aspects from physical chemistry, fluid mechanics, and
chemical engineering. The book explores the unit operations used in emulsifica-
of
tion and homogenization processes, using fundamental theory from different fields
to discuss design and function of different emulsification techniques.
Features
EDITED BY
K16909
Marilyn R ayner
ISBN: 978-1-4665-8043-5
90000 P et r D ejm ek
9 781466 580435